.P8 H24 Copy 1 A Study, by the Crop Survey Method, of Factors Influencing the Yield of Potatoes A THESIS PRESKNTED TO THE FACUIvTY OF THE GRADUATE SCHOOIv OF CORNELL UNIVERSITY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY BY EARLE VOLCART HARDENBURG Published as Memoir 57 Cornell University Agricultural Experiment Station, June 1922. A Study, by the Crop Survey Method, of nfluencing t of Potatoes Factors Influencing the Yield A THESIS PRRSRNTRD TO THK FACUIvTV OF THK GRADUATE SCHOOL OF CORNEIvIv UNIVERSITY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY BY EARLE VOLCART HARDENBURG Published as Memoir 57 Cornell University Agricultural Experiment Station, June 1922. ^8 M a4 .1^ 3 CONTENTS PAGE The crop survey as a method of research 1145 Biometry as apphed to crop-survey data 1147 fO The taking cf records 1147 Description of regions surveyed 1148 QJ^ Long Island 1149 Steuben County 1153 IVIonroe County 1154 Frankhn and CHnton Counties 1155 ]\Iethod of studying survey data 1156 The study of factors 1157 CHmate 1157 Elevation 1159 Crop rotation 1 162 Value of land 1166 Soil 1169 Plowing .' 1171 Time of plowing 1171 Depth of plowing ? 1173 Manure and fertihzer 1177 Home-mixed fertilizers 1179 Quantity and value of fertilizer 1180 Manure and fertilizer used in the four regions 1182 Analysis of fertihzer 1188 Method of appljang fertihzer 1191 Use and influence of hme on potato land 1196 Adaptation and yield of varieties 1197 Source of seed 1205 Sun-sprouting of seed 1208 Chemical treatment of seed 1210 Interval between cutting and planting 1211 Dusting cut seed 1214 Type of seed 1215 Large as compared with small tubers for seed 1215 Whole as compared with cut seed 1216 Large as compared with small seed pieces 1217 Number of eyes 1220 Types of seed used in the four regions surveyed 1221 Relation of amount of seed to yield 1224 Date of planting 1231 Hand as compared with machine planting 1232 Checkrow as compared with drill planting 1234 Depth of planting 1237 Depth of cultivation 1242 Ridge as compared with level culture 1244 Frequency of cultivation 1246 Spraying 1255 Relation of date of harvest to yield 1263 ^Method of harvesting in the four regions surveyed 1265 Types of storage in the four regions surveyed 1269 Length of storage period 1269 Summary 1270 Conclusions ."■.... 1272 Author's acknowledgment . 1273 Bibliography 1274 Survey blaiJk 1280 1139 A STUDY, BY THE CROP SURVEY METHOD, OF FACTORS INFLUENCING THE YIELD OF POTATOES A STUDY, BY THE CROP SURVEY METHOD, OF FACTORS INFLUENCING THE YIELD OF POTATOES ^ Earle V. Hardenburg Almost from the date of their estabHshment, practically all state and federal experiment stations in this country, as well as many foreign stations, have tested, by diverse methods, the relative influences of factors affecting the yield of potatoes. A review of the abundant literature of the subject shows that a majority of these tests concern the influence of seed and fertilizers on yield. This fact, further substantiated by the results of the study herein reported, indicates that, with the exception of climate and soil, seed and fertilizers are the most vital factors affect- ing yield. Because of the widely differing environmental conditions under which the tests have been conducted, it is possible in only a limited degree to draw definite conclusions from a summary of the results. Furthermore, a large part of the literature fails to supply much detailed information as to the methods used in the experiments, and gives little if any considera- tion to factors affecting yield other than the one principally concerned in the respective tests. This means that most of the evidence available to date is of only limited application. A comparison of the conclusions reached and the recommendations made by experiment stations, with those warranted by actual practice as found on farms in a potato-growing region, is therefore of considerable value. Such a comparison is, to some extent, made possible by the use of the survey method of collecting and studying data on the influ- ences affecting the yield of potatoes. The survey method has accordingly been applied to the study of such factors in several potato sections of New York, and the results ^re herein compared with those obtained experimentally. As an additional check on the conclusion's drawn, the biometrical method as applied by Rietz and Smith (1910)- has also been used in studying those factors which, according to the sur- vey method, appear to affect the yield to the greatest extent. The survey method of studying crop production, wholly aside from the agricultural methods involved, has proved to be a most valuable means of determining the actual practice thruout the State, and has aided in correcting many false ideas of long standing concerning cultural methods used with this crop. The collection of data was begun in the smiimer of 1913 and continued thru the summer of 1914. In 1913, 330 records of the 1912 potato crop were taken from as many potato farms on Long Island, and 360 records ' Also presented to the Faculty of the Graduate School of Cornell University, February, 1919, as a major thesis in partial fulfillment of the requirements for the degree of doctor of philosophy. = Dates in parenthesis refer to Bibliography, page 1274. 1143 1144 Earle V. Hardenburg of the same year's crop were obtained from that number of farms in northern Steuben County. In 1914, 300 records were similarly taken for the 1913 crop in Monroe County, and 300 in Franklin and Clinton Counties combined. This gives a total of 1290 records for the crops of 1912 and 1913. Each record was in the form of a filled-out survey blank, a sample of which is included at the end of this paper, and was as complete as possible in the details listed. Because of the similarity of regional conditions and of cultural practices, the counties surveyed were studied Fig. 125. regions included in the potato SURVEY as^four distinct sections, as follows: (1) Long Island, including the potato- growing areas of Suffolk and Nassau Counties; (2) Steuben County; (3) Monroe County; (4) Franklin and Clinton Counties. The location of these areas is shown in figure 125. These regions were selected, not because they include the counties of highest total production, but because they represent typical and distinct centers of potato production in the State. The importance of potato production in a region is probably best indicated by figures showing the percentage of total crop acres devoted to this crop and the average potato acreage per farm. A summarj^ of A Study of Factors Influencing the Yield of Potatoes 1145 the scope of the survey and the status of the industry during the years 1912 and 1913 is given in table 1. Of the four regions surveyed, the potato crop is regarded as of most importance on Long Island and of least importance in Franklin and Clinton Counties. TABLE 1. Summary of the Four Regions Surveyed Number Acreage surveyed Average size of Per cent of total acreage Per cent of crop acreage Per cent of crop acreage .\verage potato acreage Average yield per acre Region Year of records farm surveyed potatoes, potatoes. m notatoes, for the per farm for farms farms sur- veyed farms sur- veyed county (1909 census) veyed veved (bushels) Long Island, in- cluding parts of Suffolk and Nassau Coun- ties 1912 330 8,188.16 65.0 37 44 23.0 24.8 175.5 Steuben County. . 1912 360 5,301.10 145.8 10 18 8.3 14.7 136.4 Monroe County . . 1913 300 3,728.25 112.1 11 8.0 12.4 126.2 Franklin and Clin- ton Counties. . . 1913 300 2,160.00 169.5 4 10 5.7 7.2 179.3 THE CROP SURVEY AS A METHOD OF RESEARCH ^ From its inception in this country, agricultural teaching has depended largeh^ on textbooks, collateral references, and the published results of experiments. There is still a considerable lack of practical information which can be supplied only by protracted experimentation or by the study of large numbers of survey records in the regions concerned. Fre- quently problems arise which local experiments fail to solve because of the impossibility of handling the work on a sufficiently extensive scale. Large numbers of records might very often be the means of discovering the common causal factor prevailing thruout a region, thus furnishing the solution of the problem or at least a working basis for its solution. A typical illustration of this is furnished in the investigations on pecan rosette by McMurran (1919). Pathologists had previously been unable to account for the cause or to recommend measures for the control of this disease, which w^as so prevalent thruout the pecan orchards of the Southern States. McMurran, by taking records of many orchards in the various pecan regions of the South, discovered that the disease was almost entirely absent in the orchards of the rich river bottom- ands, and from this observation he deduced that the cause of the disease laj^ in certain soil deficiencies. The farm-crops survey aims first of all to search out the actual facts concerned in the production of a given crop in a given area. This informa- tion, obtained in sufficient quantity, may then be regarded as statistics 1146 Earle V. Hardenburg from which to determine the most beneficial influences and practices. The survey idea was first launched in New York by Dr. L. H. Bailey, under whose direction horticultural studies were made thruout the State. In 1903 Professor John Craig started an orchard survey campaign in west- ern New York. Such of these surveys as were completed have been published as Cornell bulletins (Warren, 1905, a and b; Cummings. 1909; Martin, 1911). Under the direction of Dr. G. F. Warren, the survey idea was extended to include the farm as a whole, with the result that whole farming areas, with the farm as a unit, have been studied in what are called farm-management surveys. The results of such studies have also been published as Cornell bulletins (Warren and Livermore, 1911; Thompson, 1915). Montgomery (1913), in discussing crop surveys, states that their primary function is to determine how to grow the crop, while farm-management surveys aim to determine when to grow the crop. Warren (1914) attests the value of agricultural surveys by saying that there are many kinds of agricultural information that can be found onlj- by svirvey methods, since the conditions in question exist only on the farms. He states further that agricultural knowledge, to be of most value, should be the result of both survey studies and experimental tests. The accuracy of survey methods depends very largely on such factors as the personality of the man procuring the records, the manner in which questions are asked, the number of records obtained for each region studied, the unit used as a basis in the study of a factor, accuracy in tabulation, and the final interpretation of results. The more extensive the record to be obtained, the greater is the number of records necessary for final accuracy. The principal faults in much of the survey work to date lie in the attempt to include too much detail and in the use of too few records. Warren (1914) is of the opinion that ordinarily 1000 records should be used, tho 500 may be enough in some eases. However, the necessity of such large numbers depends somewhat upon the scope of the survey. By the law of averages, large numbers tend to eliminate individual errors. Spillman (1917) has said that the accuracy of any average depends on three things: first, on freedom from bias; secondly, on the number of items from which an average is obtained; and thirdly, on the accuracy of the individual items averaged. Considering the limitations of much of the experimental evidence to date, large numbers of survey records are undoubtedly pro- ductive of as nearly accurate results as are obtained by experimental work. As emphasized by Warren (1914), the region siu'veyed should be an agricultural, not a political, unit. Furthermore, the records should be taken only during a normal year unless records are to be obtained for con- secutive years. Unfortunately for this study, the year 1912 was at first drier than normal, but the abundant rain which fell late in the growing season caused some blight rot; 1913, however, was a more nearly normal year. A Study of Factors Influencing the Yield of Potatoes 1147 BIOMETRY AS APPLIED TO CROP-SURVEY DATA Biometry as a science is beginning to have wide application, wherever sufficient data make its apphcation possible, in the solution of problems involving the study of the interrelation of factors or the study of cause and effect. Until the present time, biometry has been used mainly only in the study of inheritance and in the correlation of characters in large popu- lations of plants and animals. Its use has been thus limited because only in such studies have conditions been so controlled that none but the factor or factors under observation could affect the results, and because it has been possible to use large numbers of individuals for such investigations. Biometry should have a place in the study of crop-survey data wherever large numbers of records are involved, in order that the coefficient of correlation may serve as a check on the conclusions otherwise drawn and that it may furnish, thru its frequency table, a description of the prevailing practice in the region in question. Tolley (1917) states that the coefficient of net correlation affords a good means of determining the net effect of each of several factors bearing on a result, or of eliminating the effect of other factors when it is desired to find the true relationship between any two. Applying biometrical methods to farm-survey data on fattening baby beef, Tolley has shown how the gross apparent correlation between any two or more factors may be substituted in a derived formula and the net correlation of any two factors thereby deduced. A biometrical analysis of some of the more influential factors involved in this study has been made, altho, owing to the relatively large numbers of records used in each study, only the gross correlation has been computed. Aside from the actual significance of the coefficients obtained, much information of descriptive value relative to the frequency of a given practice may be foinid in the frequency distribution tables. One of the chief functions of biometry is description. It affords a means of classif^ang a group of individuals not possible by any other means. THE TAKING OF RECORDS Five men constituted the party employed in the taking of records in 1913. This made it possible for four of the party to travel thru the potato regions in pairs while the fifth man copied and checked each clay's records. In this way, any discrepancies in the records could be checked up by a return visit to the grower or by discussion within the party. The data on the 1913 crop were taken in 1914 by two men. As previously noted (Spillman, 1917), the value and accuracy of survey data depend largely on freedom from bias. This may well apply to the selection of farms to be observed. Therefore it was decided that for these surveys the only lunitation in the selection of a farm was to be in the acreage of the crop produced the previous year. This limitation was set 1148 Earle V. Hardenburg at a minimum of 5 acres, tho a very few records were taken on farms having a production area of only 4 acres. The reason for the establish- ment of this minimum limit lies in the assumption that growers of acreages smaller than 5 are probably not growing potatoes in a manner comparable to the average of the region. The data on cost of production, tho obtained at the same time and indicated on the survey blank, are not a part of this study. The subject of cost has been studied by Fox (1919), formerly of the Department of Farm Management at Cornell Universit3^ DESCRIPTION OF REGIONS SURVEYED For a better understanding of the environmental conditions under which the potato crop was produced, a brief description of climate, soil, topography, elevation, length of growing season, market facilities, type of farming, land values, and status of potato production, is given for each of the four regions surveyed. Unfortunately, of the regions concerned, only Monroe and Clinton Counties have been soil-surveyed by the United States Department of Agriculture. More detailed knowledge as to these environmental influences may be obtained from figures 126 to 129. Fig. 126. elevations of the regions surveyed A Study of Factors Influencing the Yield of Potatoes 1149 LONG ISLAND Most of the potato crop in Suffolk County is grown east of Riverhead on both the north and the south shores of Long Island. The Long Island Railroad furnishes the transportation facilities for practically all of the surplus crop of this region. Most of the roads are improved to a high degree. Thus the time required for shipments to reach New York City need not be over one day and no delay is necessitated by transfers to other railroads. Fig. 127. growing-season rainfall (in inches) in regions surveyed Nearly all of the crop in Nassau County is grown north of a line drawn east and west thru the central part of the county. Most of the surplus crop of this county is transported directly, in heavy wagons and motor trucks, to the Wallabout and Harlem Markets in Brooklyn. The greater part of Long Island is of marine deposit formation, the elevation ranging from a point at about sea level, in the Hampton section, to an altitude of nearly 300 feet in some places on the north shore. The average elevation of the potato fields surveyed was 65.5 feet. Due to the low elevation of the south shore, the crop is exposed to heavy sea fogs which make conditions favorable to the development of late blight. The 1150 Earle V. Hardenburg topography is in general fairly level, tho the slightly rolling lands along the north shore gradually rise until they merge into prominent hills along the Sound. The potato sections of Long Island show an average growing-season rain- fall of from 16 to 20 inches, which is somewhat higher than that of most of the potato sections of New York. Rainfall seldom limits production here. The tempering influence of the Atlantic Ocean affords a growing season of approximately 200 days between killing frosts, which is greater than that Fig. 128. LENGTH OF GROWING SEASON (iN DAYs) IN REGIONS SURVEYED of any other section of New York. The growing season on Long Island is fully a month earlier than that in the other three regions under dis- cussion. The soil of most of the potato-growing areas of Long Island is of a sandy texture, topped by silty loam in layers of varying thickness. This is counter to a rather common impression that the Long Island crop is pro- duced in sandy soil. The greater part of the central section of the island does consist of sand, and this supports little vegetation aside from scrub oak and pine. That the potatoes are grown mainly on the Sassafras series of soil is shown in figure 129. A Study of Factors Influencing the Yield of Potatoes 1151 1152 Earle V. Hardenburg The importance of the crop on Long Island is shown by the fact that no regular system of crop rotation is practiced, potatoes being grown for several successive years on the same land. In order to maintain the humus content, cover crops of rye are turned under each spring. The coimnonest practice is two to four years of potatoes, the land being cover- cropped to rye over winter. Along the north shore, where a rotation is sometimes used, wheat seeded to clover and timothy follows potatoes, the hay being grown from one to two years before the sod is plowed for corn, cabbage, and cauliflower. Potatoes then follow these cultivated Fig. 130. harvesting irish cobbi.kk- i\ nassau county in july The large immature vines should be noted crops. Wheat and hay are the principal rotation crops on the southern shore. Much double-cropping is practiced in Nassau County, the early potatoes being harvested in July and the second crop in late August and early Sep- tember. Land producing a first crop of potatoes is commonly planted to turnips, beets, carrots, or other root crops, or is set to cabbage for the fall market. Rye is used here also as a cover crop. A field in which Cobblers were harvested one day and turnips were planted the next day, is shown in figure 130. A Study of Factors Influencing the Yield of Potatoes 1153 Land values are higher on L6np; Island than in the other potato sections, partly because much of the land in Nassau County is held for real-estate purposes and partly bec^ause of its geographical advantages and adapt- ability for potato production. The values range from $100 an acre in Suffolk County, to $1000 an acre, real-estate value, in Nassau County. The average size of the farms surveyed was 65 acres, of which 37 per cent was in potatoes. On the average, 44 per cent of the total crop acreage was in potatoes, while the average potato acreage per farm was 24.8. The potato crop is relatively more important in the farming system here than elsewhere among the regions surveyed. The number of records taken on Long Island was 330, representing a total of 8188.16 acres planted to potatoes in 1912. The average yield per acre, on the farms surveyed, was 175.5 bushels. STEUBEN COUNTY The area of most intensive production in Steuben County lies in its northeastern part, along the Cohocton River valley and in the hill sections on each side. The Delaware, Lackawanna & Western and the Erie Rail- road handle the potato shipments. Local buyers take most of the crop from the grower, buying it either at harvest time or on contract. They store it in temporary warehouses along the railroads or ship it direct. Because of the unevenness of topography and the heavy nature of the soil in this county, the highways are often so poor that the movement of the crop from field or cellar to the shipping point is seriously handicapped. For this reason, most of the crop is moved at the times when the roads are in the best condition. Much of it is shipped to New York and Philadelphia, but the variety Spalding's Rose 4 is sent to Florida as seed. The elevation of the surveyed fields ranged from 1200 to 2100 feet, the average being 1659.2 feet. This wide range in elevation has considerable influence on the development of the potato crop, as is indicated by this study. A large part of the total crop is produced on hillsides of varying slope, the incline often being so steep as to limit the use of heavy machinery; on the other hand, many of the best potato fields are found on the level table-lands at the highest elevations. Northern Steuben County has an average growing-season rainfall of from 16 to 18 inches, which is sufficient for maximum crops. Because of the heavy nature of the soil, years of abnormally large rainfall often cause much loss from blight rot. The growing season between killing frosts averages 150 days, and is usually sufficient to mature the crop. Because of better air drainage and cooler average temperatures, the crop is often later and the yields are larger on the farms at the higher elevatio'iis. This was not the case in 1912, however, as is shown later in the discussion of the influence of elevation. 1154 Earle V. Hardenburg Five soil scries are principally concerned in the area studied in Steuben County, as shown by figure 129. Nearly half of the crop of 1912 was grown on Lordstown silt loam, which gave a higher average yield than any other series. In elevation this soil series is next to the Volusia series, which is found only at the highest elevations. Tho both of these soil series are naturally low in fertility, the highest average yield was obtained on the Lordstown series, while the lowest average yield was produced on the Volusia series. The soils on the hilltops are largely derived from .shale and sandstone; the valley soils, altho naturally higher in fertility, contain more stone and gravel. Relatively long and fixed rotations are used in Steuben County, the com- monest being potatoes, oats, hay two years. Frequently the sod is left until long past its profitable stage for hay, with the result that the humus content remaining for the potato crop to follow is seriously depleted. Farms on which the sod was left down for the shortest period of years showed the highest yield, and vice versa. Sometimes wheat followed oats in the rotation, giving two successive years of grain. The wheat was used as the nurse crop. These farms showed .a higher average yield of potatoes than did the farms using only one year of grain. This may have been due to the additional residual fertilizer left from the second year of grain, or possibly to production on better soil than is ordinarily devoted to potatoes. Buckwheat, in which Steuben Countj^ ranks second according to the United States census of 1909, is commonly used to follow old sod land that is being broken for potatoes or to break virgin land recently cleared. On the smaller potato farms, corn for grain or silage is grown in the rotation with potatoes. Land values are as low in Steuben County as anywhere in New York, for much of the land is infertile and rough, and little of it has been sold or rented in recent years. The estimated values ranged from $25 to $80 and more an acre, the average being about $50. The average size of the farms surveyed was 145.8 acres, 10 per cent of this being in potatoes. The per cent of total crop acres per farm in potatoes was 18. The average yield per acre on the 360 farms surveyed, which represented a total of 5301.1 acres of potatoes, was 136.4 bushels, MONROE COUNTY The potato section of Monroe County covers most of the region east, west, and south of Rochester. Potatoes are an important crop on most of the farms south of the fruit belt that extends across the northern border of the county abutting on Lake Ontario. Excellent railroad facilities pro- vide transportation for the marketing of the crop, loading stations being located on the New York Central, the Lehigh Valley, the Delaware, Lackawanna & Western, the Erie, and the Buffalo, Rochester, & Pittsburg Railroad. A Study of Factors Influencing the Yield of Potatoes 1155 Elevation is not an influential factor in this region, since its variation is only between 400 and 1000 feet, the average being 592.5 feet. In general the topography is gently rolling, and in only a very few places is it suffi- ciently uneven to affect production or the usual cultural practices. The growing-season rainfall is somewhat less than that of the other regions, ranging normally from 14 to 16 inches. However, it is seldom insufficient for maximum production. Due to the tempering influence of Lake Ontario, the average growing season is 165 days, which is somewhat longer than that of the other regions except Long Island. The soils on which the potato crop is produced are principally of the Dunkirk and Ontario series, as shown in figure 129. Altho both of these soils are naturally fairly fertile, a study of comparative yields shows that, other things being equal, the Dunkirk soils gave the higher production. The soil map of Monroe. County shows an especially wide range in soil types within each of these series. The cropping system of this region is usually a four-years rotation of potatoes and grain or another crop, oats, wheat, hay. Corn is most commonly chosen as the additional cultivated crop to be raised with potatoes, tho beans and cabbage are sometimes used. The value of potato land ranged from $50 to $250 an acre, the average acre value being $150. The farms surveyed averaged 112.07 acres in size, 11 per cent of the total acreage being in potatoes; and the importance of the crop is emphasized by the fact that 15 per cent of the crop acreage is in potatoes. The average yield per acre of the 1913 crop, for the 3728.25 acres of potatoes on the 300 farms, was 126.2 bushels. FRANKLIN AND CLINTON COUNTIES The areas of production in Franklin and Clinton Counties are two : one consists of a broad, level stretch of fertile valley land along the St. Law- rence River, extending across the northern end of Franklin County and over into Clinton County; the other consists of hill and valley farms on each side of the Saranac River, in central Clinton County. In both these areas the potato lands extend back into the foothills of the Adirondack Mountains. Most of the production centers in Franklin County are located along the Rutland Railroad, while the Delaware and Hudson Railroad handles most of the crop of Clinton County. The greater part of the sur- plus is marketed in the eastern seaboard markets after the early crops of Long Island, New Jersey, and the South have been sold. A thriving trade in seed potatoes has been developed with Long Island, New Jersey, and southern points. Being in close proximity to the Adirondack Mountains, this region has a wide range in elevation. It varies from 300 to 1850 feet, the^average for the farms survej^ed being 1038.2 feet. The excellent yields obtained at the higher altitudes are due largely to the cool climate there alTorded. 1156 Earle V. Hardenburg In spite of the range in elevation, very little of the crop is produced on anything but level land. The farms along the St. Lawrence River valley are generally level or gently sloping toward the river, and most of the crop in Clinton County is also grown on fairly level fields, either in the Saranac River valley or on top of the foothills of the Adirondacks. Due to the northerly latitude of this region the growing season is rela- tively short, the average period between killing frosts being 150 days. Elevation and latitude are jointly contributing factors for an ideal potato chmate conducive to late maturity of the crop. As a rule the growth is stopped by frost, resulting in a crop more or less innnature at harvest time. This gives a product of excellent seed value and keeping qualities. The growing-season rainfall averages from 14 to 18 inches, the mountain areas receiving the greater precipitation. The rainfall is uniform thruout the growing season, each month averaging 3 or more inches. Most of the soils of this region are a fine sandy loam and are included in the Ontario, Caloma, and Terrace soil series. The Ontario series com- prises the area along the St. Lawrence River, and the Caloma and Terrace soils comprise most of the area in central Clinton County (fig. 129) . The Ontario series is largely of sedimentary origin and its fertility is rather higher than the average; while the Caloma and Terrace soils are mainly of glacial drift formation and are of only mediocre fertility. The commonest system of cropping is a five-years rotation of potatoes and corn, oats, ^hay three years. The corn is used mainly for silage. Hops have been regarded as a relatively important cultivated crop in the Franklin County area until recently, when low prices, disease, and com- petition with the western crop caused a decided decrease in acreage. At present, potatoes are the chief source of cash income in this district. Land values here are similar to those in Steuben County, the range being from $10 to $100 an acre, with the average at about $50. The average size of the farms surveyed was 169.5 acres. Only 4 per cent of the total acreage, and 10 per cent of the crop acreage, was in potatoes. The average yield per acre on the 300 farms surveyed, repre- senting 2160 acres, was 179.3 bushels. METHOD OF STUDYING SURVEY DATA As previously pointed out, one of the handicaps in any effort to de- termine, by an analysis of survey data, the absolute influence of a single factor on yield, lies in the difficulty of separating the influence of other factors from that of the one in question. This is a necessary step, how- ever, in insuring accuracy and a correct interpretation of results. A preliminary study of factors influencing potato yield in Steuben County in 1912 (Hardenburg, 1915 b) indicated that the most important factors were the amount of seed used per acre, the value of manure and fertilizer employed per acre, and the frequency of bordeaux spraying. The results A Study of Factors Influencing the Yield of Potatoes 1157 of the present study have borne out that conclusion. Therefore, in con- sidering the influence of a given factor on yield, an effort has been made to eliminate as far as possible, or at least to give due credit to, other con- tributing factors. Since the study of each region concerns but a single year, too definite conclusions must not be drawn in interpreting the data presented. De- pending upon the normahty of the season in which the crop was grown, the degree of influence of a given factor may or may not be maintained under average conditions. Tho cultural practices are not usually varied radically from one year to another, differences in the length of the grow- ing season, in the average growing-season temperature, and in rainfall, tend to affect the influence of those practices. Therefore it will not be possible to answer, in any appreciable degree, many of the questions that will be raised. The consideration of experimental results is there- fore of value in furnishing background for the study of «ach factor. As stated by Warren (1914), there are questions that can be answered only by a study of the results obtained on farms, and other questions that can be answered only by the results of experiments. Little attempt has been made to discuss any potato literature except that pertaining to seed, fertilizers, and planting, these being obviously the most influential factors under the grower's control. In reviewing the literature, one is impressed by the large quantity available and by the meagerness and unreliability of the data given to substantiate the state- ments, THE STUDY OF FACTORS CLIMATE A brief review of climatic conditions in each of the surveyed areas has been given, not because of any definite influence on the crop under con- sideration, but to make clearer the normal conditions to which the crop is subject. Facilities for taking weather data in each of these regions are not yet sufficient to allow of any attempt at the correlation of rainfall and temperature with yield for a given year. In general, the average growing-season temperature to which the crop is subject has a marked influence on the vitahty of that crop as used for seed. Briefly, high temperatures tend, to produce devitalization. Long Island growers obtain average yields ranging from 150 to 250 bushels per acre from new Maine seed, but the use of the same stock for seed a second year results in greatly inferior yields, as is indicated in figure 131. The same principle is demonstrated in the rather common practice of introduc- ing seed from northerly latitudes, a practice which is justified on the basis of better yields, as is shown in the tests cited under the caption Source of seed. 1158 Earle V. Hardenburg Fig. 131. growth variation between new and one-year-old maine seed stock on LONG island The photograph .shows also the characteristic topography of potato lands in Suffolk County Valuable studies of the influence of, weather on the yiekl of potatoes in Ohio for a period of fifty-five years have been made by Smith (1915), and a similar study for a period of twenty-six years has been made in New York by Fox (1916). The relationship of both growing-season rainfall and temperature, in both States, is expressed in terms of the coefficient of correlation (r). A comparison of these coefficients shows that .July is by far the most critical month with respect to these factors, in both Ohio and New York. The coefficient of correlation between tem- perature and yield is in most cases negative for both States, indicating that yield is inversely proportional to increase in temperature. So far as rainfall is concerned, the correlation for Ohio is positive and fairly large, indicating that rainfall is ordinarily a limiting factor in yield. The correlation of rainfall and jdeld in New York, on the other hand, is negative, _ showing that years of high rainfall are years of low yield. The average growing-season rainfall for the potato sections of New York, previously given as ranging from 14 to 20 inches, is evidently sufficient for this crop. The negative coefficient of correlation is probably a reflection of the fact that years of highest rainfall in New York have been years of severe loss from blight rot. A Study of Factors Influencing the Yield of Potatoes 1159 ELEVATION Elevation as a factor influencing production has been determined from the figures shown on the topographic sheets of the surveyed areas pub- Hshed by the United States Geological Survey. As far as possible, the location of the potato fields for which data were taken was indicated on these topographic sheets at the time of taking the data. The chief difficulty in determining the absolute influence of elevation lies in the fact that ■ increase or decrease in ek^vation is usually accompanied by a difference- in soil type. A study of elevation, therefore, really involved also the consideration of both climate and soil. The writer is not aware that any test has ever been made in which either one or the other of these factors was studied with the other factor eliminated. Progressive increases in altitude and in latitude are similar in that each is accompanied by a reduction in temperature. T^e United States Weather Bureau, in computing temperature equivalents, makes use of the principle that every 300 feet rise in altitude is accompanied by a reduc- tion in temperature of one Fahrenheit degree. Influence of elevation on Long Island Elevation cannot be considered a potent factor in the Long Island area, for its highest point does not greatly exceed 200 feet. Many farms along the south shore of Suffolk County are below sea level, the sand dunes alone keeping out the sea. A typical Long Island potato field is shown in figure 131. The relation of elevation to yield in 1912 is shown in table 2: TABLE 2. Relation of Elevation to Yield on 327 Long Island Farms in 1912 Elevation (feet) Number of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertilizer per acre Average elevation (feet) 1- 50 157 87 53 22 8 178.3 157.5 184.3 188.8 196.3 12.9 12.0 12.4 12.2 13.4 $35.31 27.71 32.20 33.01 31.65 26.4 50 - 100 64.8 100-150 117.2 150 - 200 167.7 200 and over 218.1 Total 327 175.3 12.5 $32.39" 65.5 1160 Earle V. Hardenburg Altho there is no proof in table 2 that the yield increases with an increase in elevation, there is a slight indication that this may be true. The farms located at 50 to 100 feet elevation had a lower yield than those at the lowest elevation, partly because they received less seed and fertilizer than any other group. Furthermore, the farms at the lowest elevation received slightly more than the average amount of seed and fertilizer per acre. It is improbable, however, that the wide difference in yield between the two groups at the lowest elevations was due entirely to .differences in amount of seed and fertilizer. There may have been some basic reason why the 87 growers at the 50-to-lOO-feet elevation used the least seed and the least fertilizer, which would account in part for the lower yield. No such reason is apparent, however, from the data at hand. Influence of elevation in Steuhen County The average elevation of the farms visited in Steuben County is greater than in any other of the regions concerned in this survey, it being 1659.2 feet. The elevation varies from 1200 to 2100 feet, a range of 900 feet, and within this range there is a considerable variation in the soil types, as is shown later in table 13 (page 1770). A sununary of the average yields obtained at various elevations is given in table 3: TABLE 3.. Relation of Elevation to Yield on 355 Steuben County Farms in 1912 Elevation (feet) Number of farms Average yield per acre (bushels) Average unharvested yield per acre (bushels) Average elevation (feet) 1200-1300 9 36 39 46 46 63 61 34 21 148.8 156.6 129.7 136.4 133.9 131.6 138.3 134.6 124.7 24.6 13.1 20.5 20.0 24.1 29.0 30.7 27.9 17.0 1,243.6 1300- 1400 1,336.0 1400-1500 1500-1600 1600- 1700 1,426.2 1,530.4 1 , 630 . 2 1700-1800 1,732.1 1800-1900 1,829.7 1900 - 2000 1,920.4 2000 - 2100 2,033 4 Total 355 Average. . . 136.4 , 24.2 1,659 2 A general tendency for yields to decrease as elevation increases is ndicated by table 3. This is counter to the expected influence of altitude, and may be explained by the fact that the soil at the higher altitudes of A Study of Factors Influencing the Yield of Potatoes 1161 this region is heavier and of lower natural fertility. Further evidence of this condition is found in the figures showing a greater percentage of unhar vested yield due to blight rot, which is so common in these heavier soils, at the higher elevations. The Green Mountain, or white-sprout, type of potato withstands less heat than does the Rural, or blue-sprout, type. Where the growing- season temperature is relatively cool, as in Franklin and Clinton Counties and on Long Island, the white-sprout type is therefore more common. In Steuben County, of 94 farms growing the white-sprout potatoes, 61 per cent were located above 1660 feet elevation and only 39 per cent were located below this level. Of 239 farms growing the blue-sprout type, 50 per cent were above and 50 per cent were below 1660 feet elevation. There is some tendency, therefore, to grow more of the white-sprout type at the cooler altitudes. Influence of elevation in Monroe County The range of elevation in Monroe County is between 400 and slightly over 800 feet. Little opportunity is therefore afforded to study the influence of this factor in this region. The figures in table 4, interpreted in the light of average seed and fertilizer used, show that elevation has some tendency to increase 3deld. TABLE 4. Relation of Elevation to Yield on 296 Monroe County Farms in 1913 Elevation (feet) Number of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertihzer 400-500 500-600 600-700 700-800 800 and over .30 107 129 23 7 130.5 122.8 116.0 165.2 225.4 12.6 12.5 12.4 12.5 14.7 114.03 11.20 10.86 11.56 11.39 Total 296 127.1 12.5 $11.34 Influence of elevation in Franklin and Clinton Counties A variation of over 1500 feet elevation in the farms in Fra:nklin and Clinton Counties affords excellent opportunity for the study of the influence of elevation on yield. The summary given in table 5 shows a rather marked 1162 Earle V. Hardenburg •rABLE 5. Relation of Elevation to Yield on 290 Franklin and Clinton County Farms in 1913 Elevation (feet) Number of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertilizer Average elevation (feet) 300 - 600 .... 33 19 31 47 101 42 17 154.8 154.9 185.1 184.0 179.1 191.0 193.4 11.6 10,9 11.5 12.1 12.2 12.5 11.8 $12.90 11 12 13.09 13.10 12.72 13.22 14.13 437.7 600 - 800 ... 697.1 800-1000 895.5 1000- 1200 1,094.4 1200-1400 1,296.9 1400- 1600 1,468.6 1600- 1850 1,709.7 Total 290 177.3 11.9 $12.91 1,038.2 infliienco of this factor. With the amounts of seed and the vahie of manure and fertihzer used approximately equal, the best yields were produced at the higher mountain elevations. Since the increase in elevation for this region is accompanied by a con- siderable variation in soil type, a part of the increase in yield at the higher levels may be due to the latter factor. However, since Franklin County has not been soil-surveyed, it is impossible here to measure accurately the influence of the soil. Very little difference in soil type was evident between the Dover fine sandy loam of the lower elevations and the Caloma fine sandy loam of the higher elevations in Clinton County. CROP ROTATION The benefits of crop rotation to a heavy-feeding cultivated crop such as potatoes have long been recognized. The crop survey as a means of com- paring various rotations in a given region, however, has very limited possi- bilities, for in the older farming regions the same general type of rotation is followed thruout. Very few tests have thus far been made by the experi- ment stations to determine the most suitable place in the rotation and the best length of rotation for potatoes in a given region. Probably the most valuable work has been done by Hartwell and Damon (1916) in their twenty-years comparison of different rotations of corn, potatoes, rye, and grass, at the Rhode Island Station. The principal feature of this work lies in a comparison of four-, five-, and six-j^ears rotations of potatoes, rye and rowen, grass, corn, the grass being left down for from one to three years. No stable manure was used, but complete commercial fertilizers A Study of Factors Influencing the Yield of Potatoes 1163 were added to the sod each year. In the matter of fertihzers, Hartwell and Damon's experiment is not comparable to farm practice in New York, where httle or no commercial fertilizer is ever used, stable manure being generally applied, instead, as a top dressing, during the last year of sod or perhaps just before plowing for corn or potatoes. The average yields per acre of potatoes obtained by Hartwell and Damon, in the rotations including grass for one, two, and three years, were 200, 199, and 223 bushels, respectively. It appears that their commercial-fertilizer treat- ments were sufficient to maintain a maximum condition of sod thruout the three years. A test on the influence of various fertilizers on potatoes, conducted at the Rothamsted station, is reported by Hall (1905). In this test the crop was grown for twenty-six consecutive years on the same land, and under each treatment the yields declined during the later as compared to the earlier years of the test. Long Island is the only section in New York in which the crop is grown without rotation, and it is only the in- creased use of fertilizers that has maintained yields there. Not only is it difficult to get sufficient stable manure for the potato crop on Long Island, but many growers do not find it economical to haul fertilizer in this form so great a distance as would often be necessary. Consequently, each year more than a third of the growers sow a cover crop of rye after potatoes. Some use the cover crop every year, while others use it only every second or third year, and some not at all. In the consideration of the influence of cover crops on yield, only those fields are included on which a cover crop was grown in the fall and winter preceding the potato crop. In table 6 the average yields that are obtained directly after cover crops, are com- pared with those obtained when no previous cover crop had been used. TABLE 6. Relation of Cover Crop to Yield on 313 Long Island Farms in 1912 Treatment Number of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertilizer per acre Cover crop No cover crop . Total. 131 182 174.1 177.3 12.6 12.5 313 $32.61 32.25 Average . 175.5 12.5 $32.40 The figures given in table 6 should not be construed to mean that cover crops are not beneficial to the potato crop on Long Island, because the 1164 Earle V. Hardenburg yields obtained in the group listed as not using a cover crop may have been produced on farms which used a cover crop two or three years previously or on farms whose soil was naturally higher in organic content. Granting this, the data on cover crops for Long Island are not sufficient to indicate either advantage or disadvantage accruing from its use. It is true that in 1912 growers who had not sown a cover crop the previous fall did not attempt to supplement the soil fertility by using more fertilizer. This in itself may indicate that, in the main, only those growers who actually needed the cover crop to maintain yields were the ones who used it. The rotations followed in Steuben County, consisting usually of potatoes, grain, and hay, vary principally in the number of successive years that the hay and the grain are left on the same ground. Commercial fertilizer is applied lightly at the time of planting potatoes, and, altho what stable manure is available is put on the sod to be plowed for potatoes, there is seldom enough to cover the entire potato acreage. The yields of hay are largely dependent on the residual fertilizer left from that applied directly to the grain crops. Thus in the longer rotations, in which sod is left down for three or more years, only a poor supply of root and stubble residue is left to supply humus to the potato crop. A comparison of the influence on the yield of various types of rotations in this region is shown in table 7: TABLE 7. Relation of Rotation to Yield on 240 Steuben County Farms in 1912 Manure or fertihzer on part of acreage Manure or fertilizer on entire acreage No manure nor fertilizer used Rotation Num- ber of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Num- ber of farms Average yield per ■ acre (bushels) Average cost of manure and ferti- lizer Num- ber of farms Average Average amount yield of seed per used acre per (bushels) acre (bushels) Potatoes, grain, hay. . Potatoes, grain, hay. 13 117 62 25 11 177.0 134.9 122.7 150.1 143.0 10.9 10.3 10.0 9.2 8.8 8 58 26 12 5 189.1 150.1 135.0 160.9 160.2 $10.51 1 13.98 7 11.95 1 3 150.0 106.2 103.6 60.0 7.5 9.6 Potatoes, grain, hay. 10.5 Potatoes, grain, grain, hay, hay Potatoes, grain, grain, hay, hay, hay 13.34 20.34 1 8.8 Ehminating the factors of seed and fertilizer as given in table 7, the yield consistently decreased with each successive year that the sod remained in rotation. This shows the tendency of the seeding to become thinner and of less value as a source of humus for the potato crop, the older it becomes. The figures for the last two rotations in the table — which differ from the first three in that they contain two years of grain instead of one, and from A Study of Factors Influencing the Yield of Potatoes 1165 each other only in the number of years of successive hay crops — show a higher average yield of potatoes with them than with the first three. This may be due to the additional residual fertilizer left from that applied to the extra year of grain, or to the factor of naturally better soil as indicated by the tendency to produce more grain. The type of rotation commonest in each region is indicated by the figures in table 8 on the percentage of total crop acres occupied by each crop listed. No fixed rotation is indicated for Long Island, where potatoes are grown for a varying number of successive years on the same land. The TABLE 8. Relative Importance of Crops on Farms Surveyed Crop Per cent of crop acres Long Steuben County Monroe County Eranklin and Clinton Counties Average Hay Potatoes Oats Corn for grain . Wheat Orchard Rye Corn for silage . Cabbage Sweet corn .... Beans Buckwheat. . . . Cauliflower .... Barley Garden truck. . Brussels sprouts Corn for fodder , Alfalfa Peas Oats and barley Sugar bush 34.25 21.50 14.75 5.75 5.50 2.75 2.50 1.75 1.75 1.50 1.25 1.25 1.25 1.00 1.00 0.50 0.50 0.50 0.25 0.25 0.25 figures for Steuben County indicate a rotation of potatoes, oats, hay two years; those for Monroe County, a rotation of potatoes with corn or beans or cabbage, oats, wheat, hay one to two years; and those for Franklin and Clinton Counties, a rotation of potatoes with corn, oats, hay three years. A review of the experimental literature on the influence of crop rotation in potato production shows a striking preference for either grass, or a 1166 Earle V. Hardenburg legume productive of considerable vegetative growth, as a crop to precede potatoes. This is evidence of the efficient use which the potato crop is able to make of this form of organic material. Such legumes as cowpeas, soybeans, and crimson clover commonly precede potatoes in the Southern and the South Atlantic S.tates, while timothy, in combination with red or alsike clover, is used generally thruout the principal potato States. Alfalfa is considered the ideal legume to precede potatoes in the alfalfa belt of the West. The root and stubble residue from these crops not only contributes to the food requirements of the potato, but also improves the aeration, the temperature, and the moisture-holding ability of the soil. Generally speaking, the rotations of the three regions aside from Long Island are long enough not to serve as factors limiting yield except as the type of rotation may affect fertilizing practices. Inasmuch as the avail- able stable manure is not usually applied for the benefit of the hay crops, and the residual organic fertility is not thereby maintained or improved, the sod residue commonly turned under before potato planting is usually less valuable after a three-years stand than after a stand of shorter duration. VALUE OF LAND The farmer's estimate of farm land values is very often not based on productive value, altho this factor, together with the distance from rail- road and city and the salabihty of the farm, usually enters into the appraise- ment. A correlation of estimated value with average yields will show, in a measure, the extent to which productive ability of potato land enters into its evaluation. App (1916), studying the factors that influence farm profits on potato farms in Monmouth County, New Jersey, found a consistent tendency for farm acre values to decrease as distance from the railroad increased. His similar conclusions with respect to crop acre values and labor income, however, do not seem warranted from the data given. The figures obtained on land values in the regions surveyed represent the estimated selling value of potato land only. A more important factor than the distance from the post office, which was ascertained and used in making this estimate, would have been that of the distance from the nearest city, village, or railroad. Apparently, on Long Island, land valued up to $550 an acre is yielding an increased crop with the increase in value (table 9). However, it is true also that the increase in land values is accompanied by the use of more seed and more fertilizer, and by more spraying for blight. These combined factors would easily account for the consistent increase in yield. The farms showing a land value of over $550 an acre are located prin- cipally in Nassau County, at a considerable distance from the post office, and are appraised at their real-estate value. In fact, much of the land has been sold at fabulous prices for real-estate purposes and is now rented A Study of Factors Influencing the Yield of Potatoes 1167 TABLE 9. Relation of Value of Land to Yield on 330 Long Island Farms in 1912 Value Num- ber of farms Average yield per acre (bushels) Average distance from post office (miles) Amount of seed used per acre (bushels) Value of manure and fertilizer per acre Per cent of farms using bordeaux Less than $250 118 L32 27 53 147.0 184.7 196.7 191.9 2.7 2.3 2.8 5.6 12.1 12.6 13.0 12.7 $30.24 33.50 35.24 24 $250-$400 45 $400-$550 52 32.73 11 Total 330 1 175.5 3.0 12.5 , $32.40 i 32 back to the original owner who is again growing potatoes on it. Some- what less seed and fertilizer are used on these farms, and less spraying is done on them. In Steuben County, potato land valued up to $80 an acre gives increased yields with the increase in value (table 10). Tho the amount of seed used TABLE 10. Relation of Value of Land to Yield on 360 Steuben County Farms IN 1912 Value Num- ber of farms Average yield per acre (bushels) Average distance from post office (miles) Amount of seed used per acre (bushels) Value of manure and fertilizer per acre Per cent of farms using bordeaux $25-$40 $40-$50 $50-$60 ■... $60-$70 46 111 76 44 42 41 110.2 134.1 139.4 144.4 148.2 145.1 5.3 4.2 3.6 3.6 2.9 2.5 9.6 10.4 10.1 10.4 10.3 9.5 $ 7.93 9.01 10.86 11.80 12.29 10.31 4 6 8 $7O-$80 $80 and over 5 Total 360 Average ... 136.4 3.8 10.1 $10.14 5 per acre is about the same thruout, there is a tendency to spend more in manure and fertilizer for the higher-priced land. The real reason for 1168 Earle V. Hardenburg this is doubtless the relative cheapness with which manure and fertilizer can be handled by the growers nearest the villages. The increased yield of the higher-priced land may be due in part to this increase in the value of manure and fertilizer used. Land values decrease as the distance from the post office increases, in Steuben County. Distance, in fact, may largely determine the valuation of potato land. In Monroe County, as in Steuben County, the yields increased with the increase in land values (table 11), the yield increase being accompanied by. TABLE 11. Relation of Value of Land to Yield on 297 Monroe County Farms in 1913 Value Num- ber of farms Average yield per acre (bushels) Average distance from post office (miles) Amount of seed used per acre (bushels) Value of manure and fertilizer per acre Per cent of farms using bordeaux $ 50-1100 $100- $150 $150- $200 $200- $250 27 145 69 56 103.4 128.3 130.6 129.3 4.3 2.6 2.3 2.3 11.1 12.5 12.7 13.0 $10.86 10.09 12.02 14.10 19 23 35 18 Total 297 Average 127.0 2.6 12.5 $11.33 24 TABLE 12. Relation of Value of Land to Yield on 300 Franklin and Clinton County Farms in 1913 Value Num- ber of farms Average yield per acre (bushels) Average distance from post office (miles) Amount of seed used per acre (bushels) Value of manure and fertilizer per acre Per cent of farms using bordeaux $ 10-$ 25 29 105 114 42 10 160.2 173.4 189.0 178.2 176.9 4.2 3.5 3.7 2.5 1.6 10.2 11.3 12.9 12 3 13.0 $12.35 13.73 12.79 11.83 15.02 $ 25-$ 50 $ 50-$ 75 2 $ 75 -$100 10 Total 300 Average 179,3 3.4 12.0 $13.01 1 A Study of Factors Influencing the Yield of Potatoes 1169 and doubtless largely due to, an increase in the amount of seed and in the value of manure and fertilizer used. Here also the land values tend to decrease as the distance from the post office increases. In Franklin and Clinton Counties there is a tendency to spend more for seed, fertilizer, and spraying, on the farms having the higher- valued potato land (table 12). This expenditure is apparently justified on land valued up to $75 an acre. Above that point, the average yield did not increase even with increased expenditure. This may be taken as an indication that such land was valued at more than its productive ability would justify. These more valuable farms are situated near Peru, in Clinton County, and are thus highly valued because of their location in the apple section of northern New York rather than on the basis of their adaptability to potato culture. It has not been possible to study the influence of soil on yield on Long Island and in Franklin and Clinton Counties. Altho Clinton County has been soil-surveyed, it is included in the tabulations with Franklin County, and the records taken were insufficient to justify the making of such a study on Clinton County alone. The Monroe County soils map, published by the United States Bureau of Soils, has been used in correlat- ing yield and other factors with the soils of that region. Professor E. 0. Fippin, formerly of the Department of Soil Technology at Cornell University, accompanied by the writer, made a reconnaissance soil survey of the surveyed area in Steuben County in the summer of 1916. Professor Pippin's familiarity with soil mapping in New York enabled him therefor to sketch the boundaries of the various soil types and series on the topographic sheets previously used in locating the surveyed potato fields. Nearly half of the crop in Steuben County is grown on the Lordstown soil series at an average elevation of 1718.2 feet. As indicated in table 13, the highest average yield was obtained on this soil series in spite of the fact that only an average amount of seed was used and somewhat less than the average value of manure and fertilizer. In contrast to this, the Volusia soil series, located on the hilltops at an average elevation of 1785.5 feet, yielded the lowest average yield of any series in spite of the fact that about the average amount of seed was used and more than the average value of manure and fertilizer. The principal difference between these two soil series lies in the somewhat darker color and the better ox'dized condition of the Lordstown series. It is true that the soils of this region become lighter in texture and of higher gravel and stone content as the valleys are approached. This condition is probably blamable, at least in part, for the higher percentage of blight rot on the heavier soils 1170 Earle V. Hardenburg at the higher elevations, as indicated in table 13. Altho very little spray- ing for blight was done in 1912, it was noted that the fields which were sprayed returned a higher average yield than those which were not sprayed, irrespective of the soil type. Apparently, depth of planting and date of planting are not influenced by soil type. TABLE 13. Relation of Soil Type to Yield on 293 Steuben County Farms in 1912 Average Average Average Average unhar- value amount Per Num- yield vested of of seed cent of depth f^^ Average Soil type ber of per yield manure used farms of elevation farms acre per and per usmg (feet) (bushels) acre (bushels) ferti- lizer acre (bushels) bordeaux (inches) ing Volusia silt loam and loam 36 115.9 32.7 $10.79 10.2 3.4 iMay 17 1.785.5 Lordstown silt loam 151 144.8 25.2 9.35 10.9 6 3.1 'May 19 1,718.2 Wooster gravelly 19 126 9 3.2 iMay 17 Rodman gravelly loam 59 142.1 24.0 11.76 10.1 2 3.0 Mav 22 1,496.5 Chenango gravelly loam 28 140.4 23.7 9.25 11.0 7 3.1 May IS 1,364.6 Total 293 Average 139.6 25.1 $10.08 10.6 * 3.1 May 19 1,642.8 In the area surveyed in Monroe County, four soil series are concerned — Ontario, Dunkirk, Clyde, and Genesee. In all, fourteen soil types are involved, but because of the small number of farms on some of these types, only those shown in table 14 are used in correlating soil with yield and other factors. On the basis of seed used, of value of manure and ferti- lizer, and of percentage of farms using bordeaux, the Dunkirk fine sandy loam and the Dunkirk fine sand are naturally the best for potatoes from the standpoint of yield, among the types considered. Altho in 1913 nearly half of the crop in the surveyed area was grown on Ontario fine sandy loam, under at least average cultural treatment, it gave the lowest average yield per acre of any series studied. As indicated in the summary of table 14, the soil types of the Dunkirk series seem to give higher yields than those of the Ontario series. The average amount of seed, fertilizer, and spraying was about the same in both series. Here, as in Steuben County, depth and date of planting do not seem to be infiuenced by any difference in soil type. The average elevation of the two soil series sum- marized is almost identical. In brief, with more seed, more fertilizer, and a higher percentage of area sprayed with bordeaux, the Ontario soils yielded 20 bushels per acre less than did the Dunkirk soils. A Study of Factors Influencing the Yield of Potatoes 1171 TABLE 14. Relation of Soil Type to Yield on 253 Monroe County Farms in 1913 Soil type Num- ber of farms Average yield per acre (bushels) Average value of manure and fertii- lizer Average amount of seed used per acre (bushels) Per cent of farms using bordeaux Average depth of planting (inches) Average date of planting Average elevation (feet) Dunkirk fine sandy loam Ontario fine sandy 20 124 32 52 11 14 186.6 122.0 123.1 124.2 123.7 137.2 $11.45 11.09 9.83 12.99 9.01 11.24 12.4 12.6 12.0 13.0 11.0 12.4 15 19 25 6 27 21 2.9 2.9 3.1 3.3 3.3 3.7 June 14 June 3 June 7 June 6 June 8 June 7 659.3 595.2 Dunkirk gravelly 586 2 Ontario loam Dunkirk gravelly sandy loam Dunkirk fine sand. . . 582.0 573.6 515.0 Total 253 Average 128.3 $11.31 12.5 17 3.1 June 6 591.0 Summary Ontario series Dunkirk series 176 77 122.7 142.6 $11.65 $10.42 12.7 12.0 27 17 3.0 3.2 June 5 June 10 591.3 590.5 The study of plowing as to its bearing on yield is treated under two head- ings — time of plowing and depth of plowing. The average date of plow- ing was obtained by averaging the actual dates of plowing for each region. It is recognized, of course, that this date will vary from year to year, depending on seasonal conditions. The only region in the State in which fall plowing was done to any appreciable extent was Franklin and Clinton Counties. To determine the depth of plowing, the grower was in each case asked to give his estimate of the actual depth, in inches, which he plowed for potatoes. Time of plowing For many years, such advantages as increased liberation of plant food, elimination of injurious insects, and improved soil texture, have been pointed out in favor of fall plowing. Very little experimental evidence has been presented, however. Dickens (1914) has furnished data covering two years of work at five substations in Kansas, all showing a decided increase in yield on fall-plowed as compared to spring-plowed land for potatoes. A summary of the data on the season at which potato land was plowed in each of the surveyed regions in this study is given in table 15: 1172 Earle V. Hardenburg TABLE 15. Per Cent of Potato Land Fall- or Spring-plowed in the Areas Surveyed Time of plowing Long Steuben Monroe Island, County, County, 1912 1912 1913 1.5 1.7 0.33 2.1 8.0 0.00 96.4 90.3 99.67 Franklin and Clinton Counties, 1913 Fall Part fall and part spring Spring 57.0 16.0 27.0 The average dates of spring plowing in 1912 for Long Island and for Steuben County were April 3 and April 29, respectively. The average dates of spring plowing in 1913 for Monroe County and for Franklin and Chnton Counties were May 15 and May 12, respectively. No speculation as to the reason for the greater proportion of fall plowing in Franklin and Clinton Counties is offered, but it is presumed to be due as much to weather conditions for the year as to labor competition with work on other crops. The comparative yields on land plowed at the different seasons in Franklin and Clinton Counties are shown in table 16: TABLE 16. Relation of Time of Plowing to Yield on 300 Franklin and Clinton County Farms in 1913 Time of plowing Num- ber of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertilizer Average acreage of potatoes Fall Part fall and part spring Spring 171 48 81 183.7 173.9- 173.2 12.4 11.6 11.5 $12.81 14.47 12.46 7.2 7.2 7.2 Total 300 Average 179.3 12.0 $13.01 7.2 In view of the fact that about a bushel more of seed and a slightly higher value of manure and fertilizer were used on the fall-plowed land, the dif- ference in yield of 10.5 bushels per acre in favor of fall plowing may not be entirely due to a difference in the time of plowing. There is no indication that fall plowing is commoner on the larger potato acreages, since the average acreage was the same in all three groups. A Study of Factors Influencing the Yield of Potatoes 1173 Depth of vlounng It might be expected that a crop such as the potato, which develops underground and is subject to varying conditions of soil moisture and soil texture, would be influenced by the factor of depth of plowing. How- ever, no actual experiments with this problem have come to the writer's attention. Dickens (1914) states that shallow plowing has given the best results on loamy soil at the Kansas station. Stone (1905) states that at the Cornell station, deep plowing on the Dunkirk gravelly soil gave the best results. Generall}'^ speaking, shallow plowing has been recommended for heavy soils and deep plowing for light soils. The possibility of drawing definite conclusions from a study of a factor depending so much on the grower's estimate and on only one year's results, is necessarily limited. This is one of the factors that for its ultimate solu- tion must depend upon carefully controlled experiments on a given soil type in each region concerned. From the following discussion of the regions herein considered, it would appear that depth of plowing is an important factor only on soils of either extreme of texture — deep planting increasing the area for tuber development in heavy soils, and providing for planting at the moisture-table depth in light soils. Depth of plowing on Long Island Altho the soils of Long Island are noticeably lighter than those of any other potato region in the State, a marked correlation of depth of plowing with yield is evident in table 17. An increase in the depth of plowing was accompanied by the use of more seed and fertilizer per acre and a greater TABLE 17. Relation of Depth of Plowing to Yield on 328 Long Island Farms in 1912 Depth of plowing (inches) Num- ber of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertihzer Average number of times sprayed with bordeaux Less than 5 . . 26 60 119 65 36 22 163.7 162.9 173.8 174.1 188.6 202.1 11.4 11.6 12.8 12.9 12.7 12.9 $30.01 30.53 32.15 33.65 35.26 32.13 2.5 5-6 3.0 6-7 3,0 7-8 3.3 8-9 3.4 3.6 Total 328 175.6 12.5 $32.42 3.1 1174 Earle V. Hardenburg frequency of spraying. These factors contributed in some degree to the greater yield apparently resulting from the deeper plowing. In answer to the question whether the deeper plowing was accompanied by deeper planting, it may be stated here that, whereas the Long Island fields were plowed at about the same average depth as those of the similarly light soils of Franklin and Clinton Counties, the average depth of planting was deepest on Long Island, and shallowest in Franklin and Clinton Counties, ot the four regions. Apparently, deeper plowing on Long Island is to be advised. By virtue of its descriptive value as well as its value as a means for measuring correlation, the biometrical method has been applied to this factor of depth of plowing for Long Island, as also for the other regions, and the result is shown in figure 132. The correlation coefficient, r, here Yield per acre, in bushels § K 8 g 1 lO o ^1 s g S (M (M (M CO ci J_ i 1 3 1 t^ 1 1 1 J^ (M lO o § s O Ol ^ t^ o 3 ""* CI (M ) lO t^ O (M 1— I rt 1— I •?>' ^2 t- £"5 T) ;^ (U ID ^od 2g' A Study of Factors Influencing the Yield of Potatoes 1187 point of diminishing returns was reached. The yields increased sufficiently, up to the highest value of manyre and fertihzer used, to warrant the cost. Evidently it would be safe to recommend the use of larger amounts on potatoes in this region. The positive coefficient of correlation shown in figure 138 is 0.258 ± 0.036, a value significant and consistent with the coefficients for the other regions. Yield per acre, in bushels IQ O lO O lO OiOOioOiCO»CO ^ ^ ,-i,-iC<)(N(MIMCOCOCg CO ^H CO 1-1 CO ^H CD 3 ® $ 1-$ 5 „ $ 6-$10 £ =« $11-$15 $16-$20 $21-$25 $26-$30 $36-$40 $41-$4.5 O r^l >o t~- o ^ lO o o5 lO i^ o c^ |iil OOO to ill SS 2 CO 1 i III! ON ^. i^ij S2 ". ill ^« 00 2 o i as s 1 on b'^3 ^^^ gg § " ! IrJ 00 00 O'-H !^- im M CO & ^1^ 2g s 1 1 : 1 c c 1 < Amount of seed used (bushels) d-i o o t <5 l||| a. CO, ... 00 11^ gS 2 o 03 i |||| <°^|. dco 00 o nil ON in CO ill ■*t^ -1 " 2 1 o d^ d o .2 t>.0 (N So CO 00 III 8S ' s o llii tot^ CO do ^ d 14 S^ § 00 ^S5 ;^ s < 1 1 1 1 h 1 1 ' A Study of Factors Ixflvexcixg the Yield of Potatoes 119." i"2 <° B = 5*3 4a 1!° S 2 t. g s DT) a. ^1 C<3CO ^ ^ s g2 111! ;cow ^ II (NMO 2 1 ;.U,S 2J^2 : IIJ L-l ^^ B'i ocoo c-c-^- '^ 1 i4 - ■^ 111! eocOiO o <° e. 5 ^ S9 m §S§ 2' fi '-' 1 II ^s£ z^.s i 1 : ^ M-^ 1 1 s "S ^S s ^^ &^ !? ga22 n S a a H H f 5 II 1196 Earle V. Hardenburg interest to note that more than half of the 211 growers hsted here apphed their fertihzer by the broadcasting method. The study of the methods of applying fertilizer, as treated in this paper, does not allow any definite conclusions to be drawn. The question is apparently one of local application, probably depending on such factors as seasonal rainfall, amount of fertilizer used, and soil type, as previously Use and influence of lime on potato kmd Lime has been given little prominence in use either as a plant food or as a soil amendment for potatoes. This is due to the fact that, being an alkaline agent, its use is conducive to the development of common scab {Actinoviyces chromogenus) on potato tubers whenever the causal organism is present in the soil or introduced on the seed tubers. Therefore lime is usually applied in the rotation as far removed from the potato crop as is possible, while its major benefits to this crop, in the improvement of soil texture and the growth of legumes, are as well accomplished. Wheeler and Adams (1909) reported an increase in the proportion of tubers of mer- chantable size from the use of lime. There seems to be little or no con- clusive evidence available that lime has increased the yield of potatoes except indirectly thru the benefits just mentioned. In spite of the fact that lime is advised for most potato farms outside of the surveyed areas of Long Island and Monroe County, very few growers reported its use in the rotation which they were using at the time when the survey was made. A summary of the use of Imie and its place in the rotation followed in the surveyed regions is given in table 33: TABLE 33. Use of Lime, and Its Place in the Four Areas Surveyed Region Average number of years removed from potatoes Long Island, 1912 Steuben County, 1912 Monroe County, 1913 Franklin and Clinton Counties, 1913 3.4 3.1 3.0 3.7 Very few growers apply lime regularly in their rotation. The commoner practice is to use it only when necessity demands it as a means of producing legumes. On Long Island, where potatoes are grown for so many suc- cessive seasons on the same land, it is unsafe to use lime at all. The number of years from the potato crop that hme is applied, indicated in table 33, shows the consistent effort of the growers to apply it as far from potatoes as is possible, in order to avoid scab. A Study of Factors Influencing the Yield of Potatoes 1197 ADAPTATION AND YIELD OF VARIETIES The crop survey offers one of the best means of determining the relative merits of potato varieties for a given locahty. Too often the experiment stations have made generaHzed recommendations solely on the basis of the performance of a few strains tested for only a few years at the station grounds. A correct knowledgie of the adaptabihty of various types and varieties to given soil and climatic conditions can be obtained only by cooperative controlled tests under varying conditions, or by a crop survey of the performance of the varieties growing over a wide area with diverse condi- tions. Nearly every state experiment station in the United States has at some time conducted a yield test of potato varieties, the results of which are to be found in the published literature. These results are in most cases of very local significance and pertain only to the strains of seed that were available for the test. Because of the wide variation in yield of the dif- ferent strains of a given variety, no absolute recommendations for any variety should be made on the basis of such tests. A comparison by survey methods of the average yields of strains of the varieties within a region, furnishes the best criterion of the merits of such varieties for that region. Stuart (1915) has classified the standard American varieties into groups containing varieties similar in tuber and foliage characters. It is now well known that the varieties within each group conform fairly closely to one another in their adaptation to specific soil and climatic conditions. This has made it possible to determine the type or group of varieties best adapted to certain regions. It remains, then, only to choose high-yielding strains of standard varieties within this group. The status of varieties within each of the surveyed areas has been studied on this basis. Varieties and types have been tabulated in the order of their extent of production in each region. Potato varieties on Long Island Of the four regions surveyed, Long Island presented the greatest varietal standardization by growing the fewest varieties and the fewest types. Growers in this region are convinced that varieties of the Rural -group yield less, are poorer in quality, and are less popular in the New York market, than varieties of the Green Mountain, or white-sprout, type. Generally speaking, for the medium late crop, only varieties of the Green Mountain group are raised on Long Island, and the early varieties for this region are chosen from the Cobbler, the Early Ohio, the Rose, and the Triumph groups (table 34). Altho Green Mountain was only one of several varieties of this group grown in the three years from 1911 to 1913 inclusive, its popularity is shown by the fact that two-thirds of the average total acreage during this period was given to this variety. Irish Cobbler was the leading early variety produced, and most of the acreage of this variety was grown in Nassau 1198 Earle V. Hardenburg TABLE 34. Varieties Grown on 330 Long Island Farms in 1911, 1912, and 1913 Variety Color of sprout Average number of farms raising variety in 1911, 1912, and 1913 Average yield per acre in 1911 and 1912 (bushels) Per cent of total acreage grown in 1911, 1912, and 1913 Green Mountain. Irish Cobbler Carman No. 1 . . . Delaware Mills Pride Uncle Sam Early Ohio Norcross World's Fair . . . . Genesee Seedling Rose Bliss Triumph . . . Early Rose Bagley White Pink.. White White White Blue.. Pink.. White White White Pink . Pink. Pink. White 200.7 119.7 34.0 19.3 16.3 9.7 9.7 8.7 3.7 5.3 5.7 13.0 7.0 4.3 180 169 189 210 164 220 216 212.1 185.8 248.1 172.8 126.9 104.5 126.2 66.0 13.0 10.0 3.0 2.0 1.6 1.6 1.0 0.5 0.3 0.3 0.3 0.2 0.2 County. A summary of the varieties belonging to each group, as classi- fied by Stuart (1915), is given in table 35: TABLE 35. Summary of Types on 330 Long Island Farms in 1911 and 1912 Type Per cent of total acreage in 1911 Per cent of total acreage in 1912 Average jneld per acre in 1911 (bushels) Average Average yield amount per acre of seed in used in 1912 1912 (bushels) (bushels) Average value of manure and fertilizer in 1912 Green .Moimtain Cobbler Triumph Early Ohio Rose Rural , 186.6 189.0 i99'6' 166.8 227.7 179.7 157.2 128.6 231.5 123.6 216.2 12.9 12.2 11.4 12.3 13.4 10.1 $32.74 32.00 28.71 44.99 38.72 27.66 Since the Green Mountain group is the only one of importance in this region, no comparison of relative merits is made between the types. Because most of the Early Ohio acreage was produced near or at Orient Point, under high fertilization and with ideal moisture conditions, its high average yield must be discounted when compared to that of the Cobbler group. Owing to its high average yield and its white skin, Cobbler has proved to be the best early variety for this section. A Study of Factors Influencing the Yield of Potatoes 1199 Potato varieties in Steuben County In contrast to Long; Island, Steuben County showed the least standard- ization of varieties of the four regions. The thirty-five varieties listed in table 36 do not represent all that were found in the region, but only those comprising an average of at least 0.1 per cent of the total acreage for three years. TABLE 36. Varieties on 360 Steuben County Farms in 1911, 1912, and 1913 "Variety Color of sprout Average number of farms raising variety in 1911, 1912, and 1913 Average yield per acre in 1911 and 1912 (bushels) Per cent of total grown m 1911, 1912, and 1913 Number 9 Ruloff Spaldings Rose 4 White Pearl Sir Walter Raleigh Gold Coin Dooley Carman No. 3 White Giant State of Wisconsin Rural New Yorker No. 2 Green Mountain McKinley Pan American White Granger Pearhne German Queen Planet MiUion Dollar Early Manistee Uncle Sam Charles Downing American Banner Goldstein Admiral Dewey California White Mammoth Knoxall Carlisle Norcross Ward's Seedling Scotch Mane Mix Best Early Burpee Clustic Beauty Blue.. White Pink.. Blue. . Blue. . White Blue. Blue . . Blue.. White Blue. . White Blue.. Blue.. Blue.. Blue . . Pink.. Blue . . Blue.. Pink . . Blue.. White Blue.. Blue.. Blue.. White White Blue.. Blue. . White Blue.. Pink.. Blue.. Pink.. White 76.7 41.3 66.6 20.7 16.0 21.0 15 3 15.3 13.0 12.0 7.7 7.3 8.0 4.0 6.3 4.0 4.7 3.7 3.0 5.7 3.0 3 2.7 ,7 .7 ,7 ,7 3 2.7 2.3 2.0 2.0 0.7 143 5 126.5 139.0 148.5 156.8 135.0 148.8 149.5 135. 134. 134. 115. 142. 121 147. 114. 154. 126.8 118.7 163.7 134.0 118.2 127.6 118.0 167.2 135.6 133.8 159.6 156.0 143.0 136.4 127.0 168.9 108.3 106.2 22.6 12.3 11.2 6.6 4.7 4.7 4.5 4.2 4.2 3.1 2.6 1.8 1.7 1.4 1.4 1.3 1.1 1.0 1.0 0.9 0.9 0.8 0.8 0.7 0.7 0.7 0.6 0.5 0.5 0.4 0.3 0.3 0.2 0.1 0.1 1200 Earle V. Hardenburg It is significant that the three most popular varieties of this region, repre- senting nearly half of the average total acreage during 1911 to 1913 inclu- sive, are each of a distinct type. Yet each may have its proper place in Steuben County farming. Number 9, representing a high-yielding strain selected from Rural New Yorker No. 2, heads the list in table 36 and is an ideal blue-sprout variety, adapted to the heavy soils and narrow-valley farms of this I'egion. Spaldings Rose 4, a medium early variety of the pink- sprout type, is profitably raised for a special seed trade with the Hastings potato section of Florida. Ruloff is a variety of the Green Mountain type which is well adapted to the lighter soils of the northern part of the county. Altho there may be this apparent justification for diversity of type, there is surely no justification for so many varieties. Among the more popular varieties listed in the table, such standard varieties as Sir Walter Raleigh and Carman No. 3 would, on the basis of yield, appear to justify their more exclusive use in this region. The average yield of the more popular standard white-sprout varieties in this list is considerably inferior to that of the Rural varieties, which apparently indicates that, in general, this type is not so well adapted here as is that represented by Number 9. A comparison of the types produced in this region is shown in table 37: TABLE 37. Summary of Types on 360 Steuben County Farms in 1911 and 1912 1 1 Per cent , of total Type acreage in 1 1911 1 Per cent of total acreage in 1912 Average yield per acre in 1911 (bushels) Average yield per acre in 1912 (bushels) Average amount of seed used in 1912 (bushels) Average value of manure and fertilizer in 1912 Rural 76 14 9 1 75 15 9 1 138 126.3 113.8 116.7 142.4 123.1 142.2 194.4 8.8 15.5 9.7 9.8 f 8.90 13.57 Rose 8.89 Hebron 12.82 In both 1911 and 1912, the Rural group of varieties outyielded those of the Green Mountain type by an average difference of 12 and 19 bushels per acre, respectively, in spite of the fact that in both years they were grown with considerably less seed and fertilizer. As is shown later, in table 42, a part of this difference was due to a greater average loss per acre in the unharvested yield, due to late blight rot, in the Green Mountain varieties. It therefore appears certain that the Rural type of potato is better adapted to the prevailing conditions of this region. A Study of Factors Influencing the Yield of Potatoes 1201 Potato varieties in Monroe County Altho a very large number of varieties were grown in Monroe County during the years 1912 to 1914, inclusive, it is seen in table 38 that the TABLE 38. Varieties on 300 Monroe County Farms in 1912, 1913, and 1914 Variety Color of sprout Average number of farms raising variety in 1912, 1913, and 1914 Average yield per acre in 1912 and 1913 (bushels) Per cent of total acreage grown in 1912, 1913, and 1914 Sir Walter Ruileigh . . . . Carman No. 3 Number 9 Peerless Jr White Giant MilHpn Dollar Perfection World's Wonder White Flyers White Grant Rural New Yorker No Gold Coin American Giant Granger Twentieth Century . . . Isle of Jersey Hundred Fold Irish Cobbler Number 6 Pan American McKinley Green Mountain Early Michigan Livingston American Banner Number 8 Blue. Blue. Blue. Blue. Blue. Blue. Blue. Blue. ? Blue . Blue . White White Blue.. Blue.. Blue.. ? Pink.. Blue. Blue . , Blue.. White Pink.. Pink.. Blue.. Blue . . 59.3 37 3 45 3 21.0 22.3 21.0 17.0 22.3 10.0 14.0 7.7 3 3 6.3 3 7 5 6 3 4 11.7 103.9 128.2 1?5.6 124.5 129.9 147.0 144.1 138.9 119.2 130.2 124.1 111.0 133.4 109.2 164.9 110.8 141.2 124.6 146.5 138.5 156.5 174.9 115.0 90.6 153.7 150.8 20.7 13.5 12.3 8.3 6.8 6.1 5.5 4.5 2.6 2.4 2.3 1.6 1.4 1.4 1.3 1.2 1.2 1.2 1.2 1.1 0.9 0.8 0.5 0.5 0.4 0.3 prevaihng type was that of the Rural group. Nearly half of the average total acreage of this region during the years 1912 to 1914, inclusive, was planted to the three standard blue-sprout varieties. Sir Walter Raleigh, Carman No. 3, and Number 9. Altho the three-years average yield for these varieties was slightly less than the average yield for the region in 1913, it would still seem desirable, for the sake of standardization, to sedect high- yielding strains from, and to retain, these few varieties to the exclusion of most of the other varieties of the Rural type Hsted in table 38. It is note- worthy that of the twenty-six varieties listed, only three are of the Green 1202 Earle V. Hardenburg Mountain type. The relatively light seasonal rainfall and the heavy soils of the Dunkirk and Ontario series have resulted in the survival of Rural varieties at the expense of other types. Here, as on Long Island, Irish Cobbler was found to be the leading early variety. The status of varietal types in this region is summarized in table 39: TABLE 39. Summary of Types on 300 Monroe County Farms in 1912 and 1913 Type Per cent of total acreage in 1912 Per cent of total acreage in 1913 Average yield per acre in 1912 (bushels) Average yield per acre in 1913 (bushels) Average amount of seed used in 1913 (bushels) Average value of manure and fertiUzer in 1913 Rural Green Mountain 94.0 3.4 1.0 1.0 0.3 0.3 94.0 3.0 1.0 0.5 O.S 0.7 141.4 153.3 135.9 121.0 128.6 121.4 125.0 109.4 117.1 138.8 80.0 100.0 12.8 14.3 11.6 14.5 10.6 11.1 $11.48 10.97 Cobbler Hebrcn 12.75 9.65 Rose 4.40 5.56 The year 1913 was one of low seasonal rainfall in Monroe County, and the crop suffered from the drought. As shown in table 39, varieties of the Green Mountain group yielded, in that year, an average of about 15 bushels per acre less than those of the Rural type, in spite of the use of more seed and about the same amount of fertilizer. This is further evi- dence that varieties of the Green Mountain group, which set tubers earlier than do those of the Rural group, suffer the effects of reaching their critical growth period during the drought season of midsummer in Monroe County. Potato varieties in Franklin and Clinton Counties Altho more varieties are grown in Franklin and Clinton Counties than on Long Island, this region is similar to the Long Island area in that nearly all the varieties were found to be of the Green Mountain type, as is shown in table 40. A' notable absence of early varieties was found among those grown from 1912 to 1914, inclusive. This may be explained in general by the fact that the possibilities for profitable yields are much greater in the main crop varieties. The four Green Mountain varieties hsted at the head of table 40 — White Beauty, Selfic Beauty, World's Fair, and Immense — comprised more than half of the average total acreage of this region during the three years con- cerned. Type has become well standardized here, as the list shows only four varieties of the Rural, or blue-sprout type, the remainder all being of the Green Mountain, or white-sprout, type. While these four Rural varieties are standard varieties, their average yield per acre and extent of A Study of Factors Influencing the Yield of Potatoes 1203 TABLE 40. Varieties on 300 Franklin and Clinton County Farms in 1912, 1913, AND 1914 Variety Color of sprout Average number of farms raising variety in 1912, 1913, and 1914 Average yield per acre in 1912 and 1913 (bushels) Per cent of total grown m 1912, 1913, and 1914 White Beauty Selfic Beauty World's Fair Immense White Mountain Green Mountain Jumbo White Lady Number 9 Mill's Prize , Carman No. 1 National Rural New Yorker No. 2 . Lincoln Scott Mountain King Eggswaire Carman No. 2 Success International Gold Coin Silver Dollar Carman No. 3 White . White. White. White . White. White. White. White. Blue. . . White. White . White . Blue. . White . ? White. White . Blue. . White. White . White . Blue.. 47.3 39.3 32.3 17.7 20.3 18.7 14.0 8.7 11.3 7.7 10.7 6.7 6.3 6.3 4.7 5.3 5.0 5.3 3.7 3.0 2.0 4.7 1.7 198.1 200.7 179,9 154.5 199.1 179.5 17B.8 197.9 168.9 189.9 204.5 184.8 135.6 221.8 160.6 199.7 165.6 191.9 169.9 197.3 147.7 187.4 184.0 16.4 15.0 12.3 7.1 6.7 6.6 4.7 4.1 4.0 3.0 3.0 2.4 2.0 1.9 1.8 1.7 1.7 1.6 1.2 0.9 0.7 0.7 0.5 production do not seem to justify their competition with the Green Moun- tain type A comparison of these two types for the region in 1912 and 1913 is shown in table 41. The averages for 1912 and 1913 show that with TABLE 41. Summary of Types on 300 Franklin and Clinton County Farms in 1912 AND 1913 Type Per cent of total acreage in 1912 Per cent of total acreage in 1913 Average yield per acre in 1912 (bushelsl Average yield per acre in 1913 (bushels) Average amount of seed used in 1913 (bushels) Average value of manure and fertiUzer in 1913 Green Mountain 86 14 92 8 194.5 168.9 183.8 164.9 12.0 13.1 $13 44 Rural 14 42 1204 Earle V. Hardenburg A Study of Factors Influencing the Yield of Potatoes 1205 a bushel of seed less and a dollar's worth less of fertilizer per acre, the varieties of the Green Mountain group yielded approximately 26 and 19 bushels per acre more, respectively, than did the Rural varieties. Under the cool climatic and the good soil conditions of this region, therefore, the Oreen Mountain type of potato is the more profitable. Summary of varieties in all regions A comparison of the principal potato groups produced in all four regions during the years for which survey data were obtained, is shown in table 42. It should be understood that no specific recommendations for a given region can be made on the basis of the comparisons drawn in this table. Certain facts of interest, however, are evident. On the average, the Green Mountain type is grown under more favorable climatic and soil conditions than the other types, with more seed and more manure and fertilizer per acre. Therefore, for the State as a whole, the Green Mountain varieties have yielded more than have those of the Rural type. The table shows that in the blight epiphytotic of 1912 in Steuben County, the rotting of the white-sprout varieties in the field exceeded that of the blue- sprout varieties by about 24 bushels per acre. For practically the same reasons the Cobbler type has outjdelded the Rose as an early potato. The Early Ohio owes its high average jdeld in 1912 to the large amount of seed and fertilizer used. SOURCE OF SEED In the mind of the average grower, the source of his potato seed is of small concern unless of necessity he is compelled to periodically change his seed stock by obtaining it outside his home county. The only section in New York where this is the case is Long Island. In table 43 are shown the sources from which the seed supply was obtained for each of the surveyed regions in 1912 and 1913. TABLE 43. Source of Seed in the Four Regions Surveyed, in 1912 and 1913 Region Farms raising their own seed Farms raising part and buying part of their seed Farms buying all their seed Number Per cent Number Per cent Number Per cent Long Island Steuben County .... Monroe County .... Franklin and Clinton Counties 9 358 248 215 2,7 99.4 82.7 71.7 151 2 38 50 45.8 0.6 12.6 16.6 170 14 35 51.5 0.0 4.7 11.7 1206 Earle V. Hardenburg For many years, the Long Island growers have been getting most of their seed from Maine. In recent years, the decreasing qiiahty of Maine seed has resulted in the use of some seed from Vermont and New York. Seed for the other three regions is almost entirely used within the county where it is grown. Occasionally there is an exchange of seed between growers within the neighborhood. Long Island growers have learned that it is not profitable to use, as seed, stock that has been grown on Long Island for more than one year. An experimental plot at Southampton is shown in figure 131 (page 1158), which demonstrates the difference in results to be expected between new stock from Maine, and Maine stock grown for one year on Long Island. The 45.8 per cent of growers on Long Island using part home-grown seed and part bought seed, shown in table 43, represent the extent to which one-year Long Island stock originally imported from Maine was used in 1912. Most of the seed stock from Maine is purchased in the fall to be shipped in the very early spring, since storage facihties on Long Island are very meager and the crop is planted late in March or early in April. The seed stock of the other three regions is in most cases stored at home along with the bulk of the harvested crop. The necessity of a change of seed on Long Island is due to several factors. The abnormally long period between harvest and planting, much of which includes the warm or hot temperatures of late summer and early fall, is not favorable to potato storage. The soil temperature of this region during the later growing season is apparently so high as to seriously reduce the vigor of the stock for seed, in spite of the high average yields obtained. The problem is therefore one of soil, of growing-season and storage tem- peratures, and of length of storage season as influencing the condition of the seed at planting time. Emerson (1914) compared yields from seed cultivated for some time in Nebraska, with those from seed raised under straw mulch and from seed recently introduced from the Red River Valley of North Dakota. He found that, whereas the cultivated seed of Nebraska rapidly deteriorated under hot growing-season temperatures, by mulching with straw between planting and blossoming time he was able to so reduce the soil temperature as to maintain vigor and obtain practically as good yields from seed so raised as from seed just introduced. Stuart (1913 a) studied the influence of environment on seed by introducing seed of thirteen English varieties of identical origin from both England and Scotland, and growing it con- tinuously in Vermont for six years. The average increase in yield of the Scotch seed over that of the seed from England varied from 10.9 to 2713.9 per cent. Results similarly striking were obtained by Macoun (1905) in four varieties grown over a period of twenty years at Ottawa. He had practiced seed selection each year for twelve years and had thus been able to maintain the original yield of the stock. Four years of adverse con- ditions followed, 'during which the yield of each variety decreased decidedly. A Study of Factors Influencing the Yield of Potatoes 1207 Seed of these same varieties, introduced during the last four unfavorable years from Nappan, Nova Scotia, and subjected to the same unfavorable conditions, maintained its yield much better, showing that the vigor of northern-grown seed enabled it to withstand adversity better than home- grown seed similarly treated. Stone (1905) reported an experiment by Fraser in which tubers that had been stored in a cool cellar until May 1 were taken out and stored for thirty-six days under four different con- ditions. Yields from seed stored in a dark cellar at from 50° to 60° F., in a coldframe at 80°, at a Vjarn window at out-of-door temperatures, and in a greenhouse at from 70° to 90°, showed about equally good results from the coldframe and the barn-window storage. The seed stored in the dark cellar gave decidedly inferior average yields, showing that moderate light and temperature for from four to six weeks before planting improves production over the usual method of dark-cellar storage up to planting time. The potato crop of Maine, Vermont, and northern New York is almost invariably harvested before the maturity of the plants. The vines are usually killed by frost. Therefore the use of northern-grown seed for Long Island means the use of immature seed; and, since immature seed is closely associated with the dormancy, or rest period, of the potato tubers, this question also is concerned. According to Appleman (1918), the rest period varies with the variety but is fairly constant within each variety. Appleman (1912) has shown that three processes go on in potato tubers during the rest period: (1) respiration, or the consumption of sugar by reversion to carbon dioxide and water; (2) conversion of starch to sugar by diastase; and (3) change of the sugar back to starch. Since these after-ripening processes are greatly influenced by temperature, it follows that storage conditions have much to do with the condition of the seed tubers when they are removed from storage for planting. The value of seed harvested in an immature condition, which has been shown by the experiments of Macoun (1905), Shepperd and Churchill (1911), Stuart (1913 b), Zavitz (1916), Ballou (1910), and Gourley (1910), is due prin- cipally to the fact that, the seed being immature, the after-ripening processes leave it in a less devitalized condition than that of seed that has entered storage fully matured. The symptom of curled skin so conamon at harvest time on northern-grown potatoes indicates a lesser degree of suberization of the epidermis than occurs in mature tubers. Appleman (1914 and 1918) has shown how the rest period may be shortened or broken, at almost any time, by the use of anaesthetics or of oxidizing agents to facilitate increased oxygen absorption. He further showed (1918) how the rest period of the southern second-crop seed may be short- ened by harvesting the seed immature, spreading it on the ground, and covering it with excelsior or burlap to prevent suberization. 1208 Earle V. Hardenburg SUN-SPROUTING OF SEED The practice of sun-sprouting seed is one which, tho recommended for many years by experiment stations, has been very Httle practiced by potato growers. It requires the bringing of the seed stock from dark cellars into a place of moderate light and higher temperatures for a period of from four to six weeks prior to planting. The main objects are to improve the stands and increase the yields by (1) the development of short, thick, green sprouts on which tuber-bearing rhizomes develop close together, (2) the opportunity to rogue diseased and otherwise inferior seed, and (3) increasing the earliness thru the starting of healthy growth before planting. Flagg, Towar, and Tucker (1896), in Rhode Island, using duplicate plots and harvesting at two different dates, obtained increased yields from sprouted seed ranging from .32 to 54 bushels per acre. Fraser (1912) sun-sprouted seed of the varieties Sir Walter Raleigh and Carman No. 3 for thirty-six days prior to planting, and obtained increases in yield ranging from 0.9 to 73.7 per cent. Hutcheson and Wolfe (1917), in a two-years comparison, obtained a difference in market- able yield of about 8 bushels per acre in favor of sun-sprouted seed. The extent to which sun-sprouting is practiced in the areas surveyed is shown in table 44. Altho earliness is much desired by Long Island, growers, TABLE 44. Groweks'Who Sun-sprouted Their Seed in the Four Regions Surveyed Long Island, 1912 Steuben County, 1912 Monroe County, 1913 Franklin and Clinton Counties, 1913 Number of growers. Per cent of growers . 166 5.5 no one was found who sun-sprouted seed for the 1912 crop. Much care is used in choosing seed at the time of its purchase in the North. The several days required for cutting the large amount of seed handled by the average grower in this region affords some opportunity for sprout development in the meantime. Furthermore, since nearly all of the Long Island crop is planted by machine, care would be necessary that none of the sprouts so formed would be knocked off in going thru the planter. The 4 per cent of growers practicing sun-sprouting in Steuben County in 1912 furnishes too small a number to study the influence of sun-sprouting on yield in this region. The influence of sun-sprouting on yield in Monroe County in 1913 is shown in table 45. The difference of about 8 bushels per acre in favor of A Study of Factors Influencing the Yield of Potatoes 1209 TABLE 45. Relation of Sun-sprouting to Yield on 166 Monroe County Farms in 1913 Treatment Number of farm.s Yield per acre (bushels) Average amount of seed used (bushels) Average value of manure and fertilizer Not sun-sprouted Sun-sprouted 134 166 121.7 129.5 12.4 12.6 $10.81 11.64 Total 300 126.2 12.5 $11.28 the sprouted seed is probably no more than might be due to the increased amounts of fertihzer and seed used on these same farnis. Therefore it cannot be said, on the basis of this difference in yield, that in 1913 sun- sprouting paid for the extra labor involved. A similar study of this factor for 40 farms in Franklin and Clinton Counties in 1913 is shown in table 46. Here the results of sprouting were TABLE 46. > Relation of Sun-sprouting to Yield on 40 Franklin and Clinton County Farms in 1913 'Treatment Number of farms Yield per acre (bushels) Average amomit of seed used (bushels) Average value of manure and fertilizer Not sun-sprouted 260 40 179.0 180.6 11.4 12.6 $13.25 Sun-sprouted 11.65 Total 300 179.3 12.0 $13.01 even less marked than in Monroe County. The difference of 1.6 bushels per acre in favor of sun-sprouting, while within experimental error, may easily be due to the increased amount of seed used by growers in this group. The smaller value of fertilizer, however, would partially offset the difference in the amount of seed used. Tho sun-sprouting apparently is not justified by the results shown for the last two regions in 1913, it must not be concluded that this is not< a profit- able procedure when it is properly done. It is impossible to determine by survey methods the true merits of this phase of potato growing, because of the difficulty of taking into account the actual methods followed. 1210 Earle V. Hardenburg CHEMICAL TREATMENT OF SEED The chemical treatment of seed potatoes to rid their surfaces of the organ- isms causing common scab (Actinomyces chromogeniis) and rhizoctonia (Corticium vaguum) has been sufficiently tested scientifically to warrant its practice wherever these diseases occur. Tho such treatments as immersion in corrosive sublimate or formaldehyde, or fumigation with formaldehyde gas, are not warranted to insure the crop against either of these diseases in the following crop, yet they have invariably reduced the infection to a profitable extent. Ballou (1910) and Gourley (1910), using duphcate plots of untreated seed, seed treated with formalin, and seed fumigated with formaldehyde gas, reduced the scab infection from an average of 58.5 per cent in untreated seed to 16.7 per cent in formalin-treated seed and to 18.4 per cent in fumigated seed. The writer (Hardenburg, 1917) reported a reduction of rhizoctonia, in the crops of fifty-eight growers in New York who used corrosive sublimate, to 1.8 per cent infection as compared with 12.7 per cent infection of the crops grown by the remaining twenty-two growers considered. He reported a similar reduction of scab infection, thru formalin treatment by sixty-two growers, to 7 per cent as compared with 10.7 per cent in the crops grown from untreated seed. In spite of these tests and the recommendations based on them, a rela- tively small proportion of growers in the four surveyed regions treat their seed. The percentage doing this in each region is reported in table 47: TABLE 47. Per Cent of Growers Treating Seed Chemically in the Four Regions Surveyed Treatment Long Island, 1912 Steuben County, 1912 Monroe County, 1913 Franklin and Clinton Counties, 1913 Formalin . 1.0 0.0 1.0 0.0 0.003 0.000 0.000 0.000 7.3 1.7 0.0 3.7 0.0 Corrosive sublimate ... 0.0 Formaldehyde gas 0.0 Sulfur 0.0 Total 2.0 0.003 12.7 0.0 It has not been possible in this study to correlate the apparent need of seed treatment with the actual practice as indicated in this table. This is due partly to incomplete data from the four regions, and partly to a lack of familiarity with diseases on the part of growers. Such treatment, however, is universally recommended because of the attested saving to the crop. For the four regions, an average of 8.3 per cent of the growers reported scab, and an average of 4.6 per cent reported rhizoctonia, in the crop from which these data were taken. A Study of Factors Influencing the Yield of Potatoes 1211 INTERVAL between CUTTING AND PLANTING An interval of from one to ten days sometimes elapses between the time when seed potatoes are cut and the time when they are planted. Weather conditions unsuited to planting after the seed is cut sometimes make this delay necessary, while in some sections the large amount of seed to be cut makes it seem advisable to cut it several days early in order to facilitate the earliest possible planting of the crop. To some extent this is the case on Long Island. In a few cases, growers have cut seed several days in advance of planting because of an assumed benefit from the drying of the cut surface of the seed pieces to be planted. The object of the present discussion is to determine the relation of this interval of time to the yield. Adams (1887), using two varieties in a single-year test, obtained an average difference of 26 bushels per acre in favor of planning immediately after cutting, between seed cut and planted fresh and seed cut twelve days before planting. Green (1888), on the contrary, using three varieties in a single-year test, reported increased yields for two varieties from a five- days interval, for three varieties from a nine-days interval, and for one variety from a twelve-days interval, over the yields obtained by planting freshly cut seed. These tests, he reported, agreed with those of Goff, of Geneva, who recommended the benefits of drying cut seed for periods not exceeding ten days before planting. T. C. Johnson (1912) , tho not reporting yields, published cuts of fields planted from freshly cut seed and from seed held for ten days after cutting. The outstanding feature of Johnson's test of this factor, carried out under carefully controlled conditions, was the strikingly poorer stand grown from the stored cut seed. Zavitz (1916), in a test covering eight years at the Ontario station, obtained an average difference of 8 bushels per acre in favor of planting freshly cut seed rather than seed held for only four or five days. Furthermore, he obtained an increase of 1 per cent of marketable tubers from the unstored seed. As previously stated, the period between cutting and planting is fre- quently longer on Long Island than in any of the other three regions surveyed. The relation between this interval and the yield is shown in table 48. Altho the relation is not clearly apparent, it is evident that as the interval is increased, the practice of dusting is also increased. Dusting tends to eliminate any of the deleterious effects caused by the drying out or heating of seed cut and stored over the longer periods of time. The average length of the interval between cutting and planting in this region in 1912 was 5.7 days, and more than half of the growers dusted their cut seed. The average interval between cutting and planting in Steuben County being only two days, little correlation between this factor and yield would be expected. This is borne out by table 49. Also, here, as on Long Island, 1212 Earle V. Hardenburg TABLE 48. Relation of Interval between Cutting and Planting, to Yield, on 272 Long Island Farms in 1912 Interval (days) Number of farms Yield per acre (bushels) P?r cent of growers dusting cut seed ' Average amoimt of seed used (bushels) Average value of manure and fertiUzer Less than 2 . 39 33 52 32 13 43 60 161.4 185.4 183.5 171.1 195.4 178.4 177.3 41 55 46 59 46 56 73 12.0 12.8 12.5 12.0 12.5 12.5 12.9 $32 69 2-3 31 73 3-4 33 33 4-5 •... 5-7 30.45 32.39 32 41 33 53 Total 272 Average, 5.7 days 180.0 57 12.6 $32.62 TABLE 49. RelatioiJ of Interval between Cutting and Planting, to Yield, on 354 Steuben County Farms in 1912 Interval (days) Number of farms Yield" per acre (bushels) Per cent of growers dusting cut seed Average amount of seed used (bushels) Average value of manure and fertiUzer Less than 1 103 84 82 55 30 125.0 145.3 140.4 132.4 149.6 1^ 11 25 27 10.0 10.5 10.1 • 9.9 10.2 $ 8 65 1-2 2-3 3-5 10.27 11.39 10.21 5-15 - 11.62 Total 354 Average, 2 days 136.6 14 10.1 $10 17 dusting was commonest where the mterval between cutting and planting was the greatest. This being true, the average yields for each group reported in the table appear to vary according to the average amounts of seed and fertilizer used, rather than according to the length of the interval between cutting and planting. The average interval between cutting and planting in Monroe County was 2.2 days, as shown in table 50. The range in interval was too small to show any marked influence of this factor on yield. In the main, dusting was commonest where the interval was the greatest. As in Steuben A Study of Factors Influencing the Yield of Potatoes 1213 TABLE 50. Relation of Interval between Cutting and Planting, to Yield, on 260 Monroe County Farms in 1913 Interval (days) Number of farms Yield per acre (bushels) Per cent of growers dusting cut seed Average amoimt of seed used (bushels) Average value of manure and fertilizer Less than 1 58 66 64 42 30 141.7 145.3 121.8 132.0 135.2 7 21 22 . 10 27 13.1 13.0 12.6 12.1 11.6 111 68 1-2 2-3 13.32 10 80 3-4 10 79 4 and more 10.87 Total 260 Average, 2.2 days 135.7 17 12.6 $11.66 County, the amount of seed and fertilizer used was so influential as to obscure any slight influence that the factor of the interval between cutting and planting might have had. Growers in the Franklin and Clinton County areas, like those of Steuben and Monroe Counties, aim to plant their seed as quickly as possible after cutting. The average interval in 1913 was only two days. Since only about ten per cent of the growers held seed as long as four days after cutting, no significant relation of this factor to yield was found. However, as shown in table 51, the most dusting was done by the growers who held their seed the longest. TABLE 51. Relation op Interval between Cutting and Planting, to Yield, on 264 Franklin and Clinton County Farms in 1913 Interval (days) Number of farms Yield per acre (bushels) Per cent of growers dusting cut seed Average amoimt of seed used . (bushels) Average value of manure and fertilizer Less than 1 . . 113 51 37 30 33 174.2 193.1 178. 7 173.6 186.5 4 2 5 30 12.2 12.4 11 4 11.6 12.6 $13 27 1-2 12 42 2-3 .. . 13 14 3-4 . 12 64 4 and more 15 11 Total 264 Average, 2 days 180.1 6 12.0 $13.32 1214 Earle V. Hardenburg DUSTING CUT SEED Dusting of cut seed has for its objects the prevention of drying out by- undue bleeding of the cut surfaces, and the prevention of heating which may occur in cut seed stored in large volume under conditions of poor ventilation. Dusting is not commonly practiced except when conditions require the storage of large quantities of cut seed for several days before planting, as on Long Island in 1912. Stone (1905) compared the yields from five varieties cut only one or two days before planting, one lot of each variety being dusted with land plaster and the other lot not being dusted. Four of these varieties gave increased yields from dusting, ranging from 5 to 26 bushels per acre, while the fifth showed a decreased yield of 7 bushels. The writer believes that a greater interval between cutting and planting would have resulted in a greater increase in yield due to dusting. Zavitz (1916) reported on an average of twenty-two tests, covering five years, in which yields from undusted freshly cut seed were compared with yields from seed treated with land plaster and with slaked lime. In all these tests, the seed was planted immediately after cutting and dusting. The average yield from the seed treated with plaster was 23.6 bushels per acre higher than that from the untreated seed, and the average yield from the seed treated with lime was 9.8 bushels per acre more than that from the untreated seed. In another experiment, in which the effects of treating cut seed with road dust, with ground brick, and with land plaster, were compared with the results from no treatment, Zavitz reported yields of 186, 189, 191, and 179 bushels per acre, respectively. Land plaster has been most commonly used as dust because of its cheapness and its adhesive qualities. On the basis of all the comparative tests reported, it has proved the most efficient. Sulfur and air-slaked lime have been used to a lesser extent. The degree to which dusting was practiced in the surveyed regions, as related to the length of time the cut seed was stored, is shown in table 52, and the extent to which various dust materials were used is shown in table 53. Unfortunately, the material used for TABLE 52. Relation of Dusting to Length op Storage Period of Cut Seed in the Four Regions Surveyed Long Island, 1912 Steuben County, 1912 Monroe County, 1913 Franklin and Clinton Counties, 1913 Average number of days seed was 5.7 57 2.0 14 2.2 17 2.G Per cent of growers dusting cut seed 6 A Study of Factors Influencing the Yield of Potatoes 1215 TABLE 53. Per Cent of Growers Dusting Seed Who Used Each Material Material Long Island, 1912 Steuben County, 1912 Monroe County, 1913 Franklin and Clinton Counties, . 1913 Land plaster 49 43 1 3 4 72 4 2 22 20 16 6 2 56 11 Sulfur 6 6 67 Ashes Unnamed 10 dusting was not ascertained from all the growers who followed this practice. In the cases in which it was not ascertained, it was listed as Unnamed. Land plaster was shown to be the material most commonly used, with sulfur second. About two-thirds of the growers in Franklin and Clinton Counties who dusted, used air-slaked lime, which was not used in any other region. Evidently, because of the likelihood that they cause scab, neither form of lime has been popular, nor have wood ashes. TYPE OF seed Probably no other phases of potato culture have received more experi- mental attention than those bearing on the relation of type and amount of seed planted, to yield. In attempting to correlate type of seed with yield, however, the all-important factor of amount of seed used has too often been neglected, both in the method of experimentation and in the conclusions. This has led to confusion, false conclusions, and a con- glomeration of published data of questionable value. To be of real value, such tests must take into account both the types of seed used and the amount of each planted. The relation of type of seed to yield may be treated under the following headings: (1) large as compared with small tubers for seed, (2) whole as compared with cut seed, (3) large as compared with small seed pieces, and (4) number of eyes. In reviewing the litera- ture on each of these points, it has been difficult to reach conclusions, because of the confusion, in most cases, of one factor with one or more of the others; and very often the amount of seed planted has been entirely neglected. Large as compared with small tubers for seed More than thirty years ago, Adams (1889) compared the yields from large, medium, and small whole seed, and found an increase both in total and in marketable yield with each increase in size of tubers used. With an 1216 Earle V. Hardenburg equidistant spacing of 38 inches for each type of seed, more per acre of the large than of the small type was planted. Aicher (1917) compared the yields from three lots of whole seed of a given variety having an average tuber weight of 8, 4, and 3 ounces, respectively. Tho Aicher did not state the amount of seed per acre used for each lot planted, the sets of each lot were planted 16 inches apart in rows 3 feet apart. Aicher reported that as the size of the whole seed was increased, the number of stalks per hill and the total yield per acre were increased. There was little difference in the percentage of marketable yield between the three lots. It must be concluded from this that large whole seed is better than small whole seed, under equidistance of planting, only because of the greater weight of seed used. Welch (1917), in a similar experiment and under similar conditions, duplicated Aicher's results almost exactly, except that he obtained a decrease in the percentage of marketable yield as the size of the whole seed increased. It is clear that the greater yields obtained by Welch from the larger whole seed were due primarily to the almost doubled amount of seed involved in each increase in size of seed used. Harwood and Holden (1893) have brought together the results obtained at the New York, Maryland, Tennessee, Kentucky, Louisiana, Wisconsin, and Utah stations, in all of which it was shown that in many cases not only the gross but also the net yield was greatest from the largest whole seed tubers. However, in all cases this increased yield was obtained from the greatly increased amount of seed used. Zavitz (1916) selected continuously, for eleven years, seed of small unmarketable, small marketable, medium marketable, and large marketable, whole seed tubers, and planted each lot in duplicate plots. With but one slight exception in the eleven years, his average marketable yield from the four types of seed showed an increase with each increase in size of seed used. The average yields per acre from the smallest to the largest seed, for the eleven years, were 105, 145, 181, and 203 bushels, respectively. Zavitz credits these differences in yield to the difference in the weight of .the seed tubers, that is, to the difference in the amount of seed planted per acre. Plumb (1890) planted whole seed of Early Rose varying in tuber weight from 14 ounces down to 1 ounce, and concluded from his results that the larger the seed tuber, the greater was the total yield, the earlier the bloom, the taller the plant, and the later the maturity. Plumb obtained a consistent decrease in net yield, however, as the size of the seed tuber increased. He failed to duplicate his test plots. Whole as compared with cut seed The advisability of cutting seed potatoes depends on three factors: the cost of labor, the cost of seed, ancl the relative merit of whole as com- pared with cut seed in the effect on yield. Literature reporting experi- mental data on this question considers only the last-named factor. Cutting seed tubers at once facilitates loss of sap by bleeding, and the entrance of A Study of Factors Influencing the Yield of Potatoes 1217 rot-producino- bacteria and fungi. Appleman (1918) has shown that the transverse cutting of seed tubers permits development in the median and basal eyes which would remain more or less dormant if the tubers were planted whole, due to the exclusive development of the terminal eyes. This would indicate a more economical use of seed than is the case if whole tubers are planted. Aicher (1917) and Welch (1917), in a three- years duplicate experiment on irrigated land in Idaho, compared whole and cut seed as to stand, earliness, and yield. Aicher obtained an earlier growth, a more prolific top growth, a larger total yield, and a smaller marketable yield, from whole than from cut tubers. Welch obtained a better stand from the whole tubers, but one that was not commensurate with the extra amount of seed recjuired. • He also obtained a smaller marketable yield from whole than from cut tubers, but he does not agree with Aicher that wholeness has anything to do with earliness. The writer is of the conviction that, so far as the relative merit of whole and of cut seed is concerned, these tests of Aicher and Welch are no true criterion. The much greater rate of planting of whole seed over cut seed invalidates any conclusion that may be drawn relative to differences in yield. Har- wood and Holden (1893) reported an interesting experiment attempted by the Tennessee station to test the comparative yields from whole tubers and from an equal weight of halved tubers from the same lot of seed. This eliminated any possible difference in rate of planting, so far as total weight of seed was concerned, but the fact that the half tubers were planted on twice as much area as the whole tubers vitiated the object of the experiment. Nearly two and a half times as much merchantable yield was obtained from the half tubers as from the whole. These same authors report a more nearly accurate test made by the New York station, in which equal-weight seed pieces of whole and cut types were used. The average total yield per acre was the same from the whole and the cut seed, but the marketable yield per acre from the cut seed was nearly double that from the whole seed. Large as compared with small seed pieces The question of size of the seed piece in relation to yield has been treated experimentally on the basis of both weight and proportion of tuber. Neither basis can be considered quite separately, and both naturally involve the question of rate of planting, the spacing of the seed remaining constant. The hundreds of experiments recorded for the study of size of seed piece are virtually tests of the most efficient rate of planting, tho the objects of and the conclusions for them have been related to size of piece alone. Plumb (1890), comparing the yields from whole tubers, halves" quarters, and single-eye pieces, obtained an increased yield with every increase in size of seed, tho the greatest merchantable yield was obtained with the half 1218 Earle V. Hardenburg tuber. As early as 1886, Samuel Johnson (1886) compared the yields from whole and half tubers and from three-, two-, and one-eye pieces, and found that, whereas the whole tuber gave the highest total yield, the three-eye piece gave the highest marketable yield. Two years later, comparing the same types of seed, Johnson (1888) obtained a consistent gain in yield with each increase in size of seed, and a decrease in percentage of stand with each decrease in size of seed. Johnson did not report whether his highest total yields were also the highest net yields. Taft (1892), in a three-years test, compared the efficiency of various rates of planting, by planting whole, half, quarter, and eighth tubers, and single-eye pieces, equidistant in the row. His net yields increased up to and including the half tuber, altho the highest total yield came froirJ whole seed. Adams (1889), using whole, half, two-eye, and one-eye seed pieces, obtained an increased total yield up to and including whole seed, with the greatest marketable yield from two-eye pieces. He did not report in terms of net yield. Green (1887) reported a two-years average test of the yields from one-eye, two-eye, half, and whole seed pieces as increasing with the size of piece used, but made no mention of the net or the market- able yields or of the rate of planting. Hutcheson and Wolfe (1917) made a three-years comparison of the yields from single-eye, half-ounce, one- ounce, and two-ounce pieces. Whereas both total and marketable yield increased up to and including the two-ounce piece, the increased yield from the two-ounce over that from the one-ounce piece was not sufficient to warrant the use of pieces larger than one ounce in weight. Aicher (1917) and Welch (1917), in their duplicate experiment covering three years, concurred in the results showing the highest total yield to be from whole seed and the highest marketable yield from quartered seed pieces, in a comparison of whole, halved, and quartered seed pieces. These investigators were agreed also that the number of stalks per hill increased with the size of piece planted, a fact which probably accounts for the smaller percentage of marketable tuljers from the largest seed. Appleman (1918) tested the influence of weight of seed piece on yield by varying the weight from 0.08 to 1.75 ounces in the variety McCormick and from 0.61 to 1.46 ounces in the variety Irish Cobbler. To give due consideration to rate of planting in such a test, he showed how this vari- ation in McCormick increased the amount of seed from 1.1 to 24.96 bushels per acre. He obtained, in both varieties, an increased total yield with each increase in weight of the seed piece. Zavitz (1916) has furnished perhaps the best contribution to the study of this factor. In ten tests, covering five years, he compared one-sixteenth-, one-eighth-, one-quarter-, one-half-, one-, and two-ounce seed pieces, the rate of planting varying from 1.3 to 41.2 bushels per acre and the number of eyes in each set remaining constant. With no seed piece weighing more than two ounces, Zavitz found increased net, marketable, and total A Study of Factors Influencing the Yield of Potatoes 1219 yields for each increase in size of seed pieces used. But here he also failed to consider the factor of rate of planting. In another five-years experi- ment, however, testing the efficiency of various spacings of seed, he has, apparently unconsciously, furnished some much-needed information. This test showed that with the same weight of seed planted per acre, the one- ounce sets, planted twice as close as the two-ounce sets, gave greater total, marketable, and net yields. Hume, Champlin, and Oakland (1914) compared large, medium, and small seed pieces, eye frequency being constant, and observed an average increase of total yield, in two varieties, of 70.9 per cent in large seed pieces and 55.5 per cent in medium seed pieces, over that from the small seed pieces. Emerson (1907) conducted a very comprehensive and accurate test of the relation of size of seed piece to yield under a constant rate of planting. Planting eighth, quarter, and half tubers, 6, 12, and 24 inches apart, respectively, he used 18 bushels of seed per acre in each plot. This gave him the highest total yield per acre from the quarter-tuber pieces and the lowest total yield from the half tubers. This, together with the test by Zavitz (1916) previously cited, would indicate that with the same rate of planting per acre, smaller pieces, down to one ounce, planted closer, are likely to give larger yields than larger pieces planted farther apart. Schweitzer (1896), with twelve varieties in a one-year test, compared the total yields and the yields of small tubers from one-eye, two-eye, quarter-tuber, half-tuber, and whole-tuber pieces. This gave a variation in rate of planting of from 4 to 87 bushels. His total yield increased with each increase in size of seed piece and in rate of planting. Also, the per- centage of small potatoes increased from 8.9 for one-eye pieces to 26.4 for whole tubers as seed. Harwood and Holden (1893) brought together a compilation of experi- ments from thirteen stations designed to test the relation of size of seed and rate of planting, to yield. In summarizing the comparative value of whole and of half tubers as seed, they showed that the results of a sub- stantial majority of these experiments were in favor of the whole tuber, not only for total and marketable yield, but also for net marketable yield and net value of the crop. Similarly, a majority of the compari- sons of the half-tuber and the two-eye piece favored the former thruout. In drawing such conclusions, it must be borne in mind that these dif- ferences in yield were due to an increase in the amount of seed used because of the larger size of the seed piece. These authors called at- tention to the conclusions of the Ohio station, that "despite the fact that whole potatoes give more small potatoes than one and two eye cut- tings, it is also true that they give more large potatoes." The foregoing review of the question of large as compared with small seed shows that few tests have actually proved any superior merit of large seed, except as the amount used per acre was increased. The few tests of 1220 Earle V. Hardenburg a more comprehensive nature have indicated that equivalent amounts of smaller seed pieces, down to a minimum weight of one ounce, planted closer, may give even more efficient results. With an expensive and limited seed supply, the latter type of seed and system of planting would seem advisable. Number of eyes Many of the older potato growers attach considerable importance to the number of eyes to be left, in cutting seed potatoes. While a few growers cut single-eye pieces, the majority prefer pieces containing two eyes. Whether or not there is any significance in the relation of this factor to yield or to quality of the crop, it is automatically controlled, in practice, by the size of the seed piece, the importance of which has already been discussed. Arthur (1892) showed, in very definite terms, that the yield of large tubers decreases with the use of seed tubers weighing more than four and one-half ounces. His results are in accord with those of many other experiments which show that increasing the number of eyes on the seed piece tends to reduce the average size of the resulting tubers. Whipple (1915) studied the influence of thinning to one stalk per hill, in a two-years test of nine varieties planted from two-ounce pieces irre- spective of the number of eyes. Thinning to one sprout improved the market shape and the uniformity of the crop, but Whipple's results do not justify any conclusion that either total or marketable yield was increased by thinning. The cost of thinning was therefore not warranted. Ballou (1910) has shown that varieties differ in the number of stalks per hill which will develop from a given size of seed piece. Bovee, having frequent eyes, developed more stalks per hill per unit of seed piece than did Carman No. 3, a variety of few eyes. Ballou obtained an increase in the unmarketable yield from every increase in size of seed piece or nmnber of eyes in both varieties. The most profitable net yield in the Bovee was obtained from two-eye pieces planted at the rate of 15 bushels per acre, and in the Carman No. 3 from half-tuber pieces planted at the rate of 25 bushels per acre. Again it remained for Zavitz (1916) to contribute the real test of the influence of eye frequency on yield, by eliminating the factor of size of seed piece. Using one-ounce seed pieces thruout a five-years test, he compared the results from seed pieces containing one, two, three, four, and five eyes, respectively, and found that as the number of eyes increased, the average total yield increased and the percentage of marketable yield decreased. The difference in marketable yield, however, was in no case more than 5 per cent. It is therefore evident from Zavitz 's work that the yield is proportional to the number of stalks per hill, as well as to the size of the seed piece, and that nothing is to be gained by cutting to a certain minimum number in preparing seed for planting. A Study of Factors Influencing the Yield of Potatoes 1221 Types of seed used in the four regions surveyed Obviously there are many difficulties in the way of attempting to deter- mine by survey methods the relation of type of seed to yield. Whether or not a grower decides to cut his seed rather than plant it whole, depends principally on the size of the tubers he has for seed, because, in cutting for seed, most growers have a definite size of seed piece in mind. Growers in Steuben County, more than in other regions, showed a tendency to plant seed of low market value. Much cull seed was therefore planted whole in 1912. Since practically all seed used on Long Island is bought and is of a grade higher than the average, growers there find it economical to cut nearly all of it. Good seed is more cheaply produced and more plentiful in Franklin and Clinton Counties, and therefore relatively large seed is used there, and more of it is planted whole than in the other regions. The proportion of whole and of cut seed used in the four surveyed regions is shown in table 54: TABLE 54. Pee Cent of Growers Using Whole and Cur Seed, in the Four Regions Surveyed Type of seed Long Island, 1912 Steuben County, 1912 Monroe County, 1913 Franklin and Clinton . Counties, 1913 Whole Whole and cut 100 1 40 59 26 74 10 39 Cut 51 Evidently there are very few growers who feel that their seed is small enough, cheap enough, or low enough in quality to warrant planting it whole. However, it is not possible to jvidge from table 54, by the amount of each type of seed used, which region used the best seed in the year for which the data were taken. It has not been possible to study the influence of the size of seed piece on yield, in these regions, because of the impossi- bility of determining even the average size of the seed used. In studying the influence of the size or the degree of wholeness of the tubers used for seed, definite conclusions cannot be drawn because of the lack of uniformity in the opinions of the growers as to the meaning of the terms large, medium, and small, and furthermore because in many cases more than one type of seed was used. The data are presented here for whatever significance they may have. On Long Island, as already stated, all the seed used in 1912 was cut. A comparison of the yields from large tubers cut and from medium-sized tubers cut, is given in table 55. The difference of 8.6 bushels per acre in 1222 Earle V. Hardenburg TABLE 55. Relation of Type of Seed to Yield on 171 Long Island Farms in 1912 Type of seed . Number of farms Average yield per acre (bushels) Average amount of seed used (bushels) Average value of manure and fertilizer Large cut 99 72 163.1 171.7 12.0 13.0 $31 24 33.18 Total 171 Average 167.0 12.4 $31 48 favor of the medium-sized tubers cut is no more, and is probably less, than should be expected from the increased amount of seed and fertilizer used by this group of growers. Apparently there was little choice between the two sizes of tubers used. The practice of cutting seed of egg size for planting was probably com- moner in Steuben County than in the other regions. Some whole seed which might be classed as cull was also used by some of the less progressive growers, as shown in table 56. The yields given in table 56 correlate TABLE 56. Relation of Type of Seed to Yield on 217. Steuben County Farms in 1912 Type of seed Num- ber of farms Average yield per acre (bushels) Average amount of seed used (bushels) Average value of manure and fertilizer 98 11 102 4 2 149.7 139.1 135.7 95.5 181.1 9.9 10.1 10.3 9.4 11.5 $10.94 Large cut Medium-sized cut . . 12.70 10.83 Small whole . . 6.84 Medium-sized whole 9.53 Total 217 146.5 10.4 $10.89 rather closely with the rate of planting and the value of manure and ferti- lizer used. The four growers who used small whole seed also used the least seed and fertilizer, and, as a result, harvested the lowest average yield. In view of the amount of seed and fertilizer used, the growers A Study of Factors Influencing the Yield of Potatoes 1223 who used a combination of both large and medium-sized cut seed obtained a more profitable net yield than did those who used either large or medium- sized cut seed alone. So far as is possible from the data taken, the relation of type of seed to yield in Monroe County is shown in table 57. More than a third of the TABLE 57. Relation of Type of Seed to Yield Farms in 1913 257 Monroe County Type of seed Num- ber of farms Average yield per acre (bushels) Average amount of seed used (bushels) Average value of manure and fertilizer Medium-sized cut 120 18 30 52 17 2 18 143.5 132.5 132.1 131.5 - 129.5 122.9 120.5 12.0 12.8 12.8 13.1 12 3 11.0 11.4 $11 75 Medium-sized whole and large cut Small whole and medium-sized cut 13.26 9.56 12.56 12.66 12.28 Medium-sized whole and medium-sized cut 9.11 Total 257 136.2 12.3 $11.63 growers in this region used medium-sized cut seed in 1913. Judging from the average amount of seed and fertilizer used by these same growers, they obtained a more profitable net yield than did the growers who used either large cut seed or large and medium-sized cut seed. The real expla- nation as to why the group using medium-sized whole and medium-sized cut seed obtained the lowest average yield, lies in the fact that these growers used considerably less than the average amount of seed and fertilizer. Except that more whole seed was used in Franklin and Clinton Counties, the types of seed used there correspond fairly closely to those reported for Monroe County. The fact that about a third of the growers in this region claimed to have used large cut seed indicates that these growers use seed of larger average size than is used in any of the other regions. The relation of type of seed to yield here is shown in table 58. The com- parative yields from large cut and medium-sized cut seed agree very well with the majority of the experiments previously cited, which showed greater yields from the larger seed, due to the greater amount of seed planted. The difference of about 8 bushels per acre of average yield between medium-sized cut and medium-sized whole seed, however, does 1224 Earle V. Hardenburg TABLE 58. Relation of Type of Seed to Yield on 292 Franklin and Clinton County Farms in 1913 Type of seed Num- ber of farms Average yield per acre (bushels) Average amount of seed used (bushels) Average value of manure and fertilizer Large cut Medium-sized whole and large cut ]\ledium-sized whole an ' medium-sized cut 97 G4 21 42 32 8 6 22 189. 1 182.7 181.5 174.6 166.7 166.1 157.8 157.3 12.9 12.4 11.8 11.1 11.0 12.0 11.1 9.9 $13.61 12 .to 10 49 Medium-sized cut ... 12 91 Medium-sized whole ... .... 13 02 Medium-sized and large cut Small whole and large cut 14.08 15 81 Small whole and medium-;ized cut 12.01 Total 292 . 178.5 11.9 $12.96 not agree with the experiments previously cited for the factor of whole as compared with cut seed. The apparent discrepancy cannot be attributed to differences in amount of seed and in value of manure and fertilizer, as these factors average approximately the same for both groups. A rather striking effect of the rate of planting is shown in table' 58, where the types of seed are arranged according to the average yields obtained from each. In fact, thruout these studies of the relation of type of seed to yield, it has been shown that those types which required the highest rate of planting were productive of the highest average yields. RELATION OF AMOUNT OF SEED TO YIELD The very marked effect of rate of planting on yield has been shown in the previous discussions of the effect of manure and fertilizer and of size of seed piece. A majority of the very large number of tests which have been conducted to determine the optimum number of bushels per acre to plant, indicate that in this country too little seed is generally planted, under average conditions. In contrast to this, the large yields obtained by European growers, who commonly plant from 30 to 40 bushels of seed per acre, are often cited. It does not follow, however, that similar rates of planting in New York would he productive of such yields. Land, labor, and climatic conditions in northern Europe are such that high rates of planting are not only possible but also profitable. As shown by the major- ity of experiments testing this factor, the most profitable rate of planting A Study of Factors Influencing the Yield of Potatoes 1225 has seldom been exceeded in this country. Rate of planting is increased either by closer planting or by increasing the size of the seed piece. Harwood and Holden (1893) compiled the results of thirteen experiments conducted at the Michigan station to determine the optimum rate of planting as well as the optimum size of seed piece. The rates of planting varied from 2.7 to 58.9 bushels per acre. The net yields showed that the optimum rates of planting varied from 10.8 to 48 bushels per acre. In only four tests was the best rate of planting higher than 24 bushels. Emerson (1907) tested rates of planting varying from 6 to 36 bushels per acre when eighth, quarter, and half tubers were used. He found that 12, 18, and 36 bushels per acre were the best amounts to use for these respective sizes of seed pieces. Macoun (1905), by varying the spacing of the seed from 10 to 18 inches and thereby varying the rate of planting from 35 to 19 bushels per acre, obtained the highest net yield from 25 bushels of seed. Kohler (1910), using the varieties Early Ohio and vSir Walter Raleigh, varied the rate of planting in each by 3 and by 2j bushels, respectively, from 6 to 21 bushels in the Early Ohio and from 5 to 20 bushels in the Sir Walter Raleigh. He obtained the highest marketable yield in the Early Ohio from 12 bushels of seed, and in the Sir Walter Raleigh from 17.5 bushels. Zavitz (1916), using only one-eye pieces, varied the rate of planting from 1.3 to 41.2 bushels by increasing the size of the seed piece. Here both the highest net yield and the highest marketable yield were obtained from using 41.2 bushels of seed. Zavitz's test covered a five-years period, and is therefore more significant in this respect than the other tests just reported. The wide variation in the optimum rate of planting shown by the experiments here reported, indicates that the available soil moisture and the fertility have much influence in limiting the stand of plants which will develop to maximum productivity. In other words, the potato soils of Steuben County, which are naturally low in fertility, cannot be expected to produce the average yields that are obtained in Monroe, FrankUn, and Clinton Counties. From a review of the data available up to this time, it seems a safe assumption that, under at least average conditions, the rate of planting may be profitably increased from its present average to from 15 to 18 bushels per acre. Where weed control is important, labor scarce and expensive, and land relatively cheap, checki'owing is a common system of planting. Tho this system naturally lowers both the rate of planting and the resulting yield, both may be increased by increasing the size of seed piece used. Rate of planting on Long Island In spite of the fact that the growers on Long Island pay relatively high prices for nearly all of their seed every year, they have apparently learned 1226 Earle V. Hardenburg that it is not profitable to plant less than the average of 12.5 bushels per acre. The influence of the rate of planting on the yield for this region in 1912 is shown in table 59. It may be seen in this table that there was TABLE 59. Relation o- Rate of Planting to Yield on 330 Long Island Farms in 1912 Rate of planting (bushels per acre) Num- ber of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertilizer Less than 11 71 68 77 49 33 32 148.9 181.4 161.7 179.5 193.0 202.0 9.9 11.3 12.2 13.4 14.1 15.6 $28.79 11-12 32 34 12-13 .- 31 50 13-14 33 52 14-15 33 06 15 and more 36.63 Total 330 Average 175.5 12.5 $32.40 Yield per acre, in bushels ci . a. CO ^ 03 9 ^ 11 S 12 ^ 13 o 14 •I 15 -S 16 m 17 18 1 1 1 1 1 1 4 3 1 1 1 1 2 8 8 15 9 7 2 1 2 2 3 8 20 7 11 6 6 3 1 - 2 6 12 23 10 11 4 4 1 1 2 5 12 5 11 6 4 2 1 1 1 4 3 13 3 6 2 2 1 2 1 1 3 3 1 5 1 4 1 1 1 2 27 81 23 13 330 43 60 27 r = 0.273 ± 0.034 Fig. 140. correlation of rate of planting and yield on 330 long island farms in 1912 A Study of Factors Influencing the Yield of Potatoes 1227 a tendency among the growers who planted the most seed to use also somewhat more than the average value of manure and fertilizer. Also, there was more spraying for blight among these growers. However, the same marked influence on the yield from the rate of planting is shown in table 24, under the discussion of manure and fertihzer, where each of these factors is treated under the various subheads. The single discrepancy in table 59, in the group of growers using from 11 to 12 bushels of seed, is due, at least in part, to the larger value of manure and fertilizer used by this group. Since the average number of eyes per piece gradually in- creased from the lowest to the highest rate of planting, it may be con- cluded that the rate of planting varied not so much according to closeness of planting as according to size of seed pieces used. From table 59, it is apparent that the growers in this region who used as high as 15 bushels of seed per acre in 1912, did not use more than was profitable. Applying the biometrical measure of correlation of this factor with'yield (fig. 140) shows the significant coefficient 0.275 ± 0.034. Rate of 'planting in Steuben County In the four regions surveyed, Steuben County growers used the least seed, planting an average of 'only 10.1 bushels per acre in 1912 (table 60). The highest rate reported by any of the 360 growers was 18 bushels, and TABLE 60. Relation of Rate of Planting to Yield on 360 Steuben County Farms IN 1912 Rate of planting (bushels per acre) Num- ber of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertilizer 6-8 19 138 126 57 20 117.8 123.1 138,3 147.7 191.3 6.9 8.6 10.2 12.5 15.4 $ 9.17 8-10 9.27 10-12 10.17 12-14 10.96 14-18 12.63 Total 360 136. ! 10.1 $10.06 there was a profitable increase in yield from every increase of 2 43ushels per acre planted, up to 18 bushels. Steuben County growers did not exceed the optiimmi rate of planting in 1912, and it is safe to recommerid a considerable increase in the average amount of seed used in that region. 1228 Earle V. Hardenburg Altho the increased jdelds were due partly to the increased values of manure and fertilizer used, this factor has been studied in a separate grouping in table 25 under the discussion of the value of manure and fertilizer for the region. The low average rate of planting for this county in 1912 was due partly to the planting of a considerable area in check- rows and partly to the use of small and relatively inferior seed. The data show that the amount of seed used probably did' not exceed the maximum which the relatively low soil fertility could support. The coefficient of correlation between rate of planting and yield for this region, 0.374 ± 0.031 (fig. 141), is the highest found for any of the four regions. Yield per acre, in busliels lO O O O iQ O lO O-i-O OiOOiOOkO IM lO t^ O OJ "O t- O 0 l^ ^ ^ i. O iM lO 1 ci g 10 « 11 l!^ " 15 ° 16 ■I 17 -S 18 n 19 20 21 22 23 Fig. 142 1 I 1 1 2 1 1 4 7 2 5 1 1 1 6 1 1 1 1 1 8 4 8 5 9 3 11 2 2 6 1 1 2 10 22 15 9 9 5 3 1 1 1 2 4 8 12 3 1 1 1 4 7 6 5 4 3 1 1 1 3 1 6 3 4 1 1 9 3 1 1 1 8 2 1 4 1 1 1 1 1 1 V 1 7 38 70 50 61 34 24 r = 0.247 ± 0.037 1 1 1 300 correlation of rate op PLANTING AND YIELD ON 300 MONROE COUNTY FARMS IN 1913 1230 Earle V. Hardenburg Rate of planting in Fr-anklin and Clinton Counties The growers in Franklin and Clinton counties who were interviewed concerning their 1913 crop planted an average of 12 bushels of seed per acre, the rate varying from less than 10 to more than 18 bushels. The relation of this factor to yield in 1913 is shown in table 62. A study of TABLE 62. Relation of Ra.te of Planting to Yield on 300 Franklin and Clinton County Fabms in 1913 Rate of planting (bushels per acre) Num- ber of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertihzer Less than 10 38 151.7 167.0 188.1 185.6 226.1 8.5 10.4 12.3 14.6 17.9 $12 72 10-12 88 105 52 13.20 12-14 12.48 14-16 14.06 16 and more 17 12.45 Total 300 Average 179.3 12.0 $13.01 Yield per acre, in bushels O iQ O i^ C s ^ g J^ ci ^ Fig. 143. 6 1 1 7 2 1 1 4 8 2 2 3 1 3 1 12 9 2 3 7 1 7 1 21 10 5 8 13 14 9 11 2 1 63 11 2 3 4 1 9 2 2 1 24 12 6 8 18 22 11 6 5 1 77 13 1 6 2 7 4 6 1 27 14 2 4 6 6 1 3 1 ' 23 15 1 3 6 8 6 4 2 1 31 16 1 1 3 1 6 17 2 2 18 1 1 3 5 19 20 1 1 2 4 12 27 50 52 76 38 26 14 5 300 r = 0.367 ±0.034 correlation of rate of planting and yield on 300 FRANKLIN AND CLINTON COUNTY FARMS IN 1913 A Study of Factors Influencing the Yield of Potatoes 1231 this table shows that there was less tendency in this region than in the others for growers using the most seed to use also the most manure and fertilizer. The figures are therefore all the more conclusive in showing the marked influence of rate of planting on yield. More seed than the amount indicated by the highest rate of planting here reported might have been used with profit in raising the 1913 crop. The coefficient of correlation, 0.367 ± 0.034 (fig. 143), is altogether significant and is the second largest value found for any of the four regions surveyed. date of planting The average date of planting potatoes in any region is determined pri- marily by the average date of the last killing frost in the spring, altho elevation, soil type, and the type of potato grown, are also important questions varying with different localities. Thus it is possible to plant earlier on light soils and at lower elevations, than on heavy soils and at higher elevations. However, because of the higher prices which usually obtain early in the harvest season, the earliest possible planting and harvest of early varieties is desirable. Zavitz (1916) reported results from thirty-six tests which consisted of planting two early, two medium, and two late varieties on four dates, two weeks apart, extending from May 31 to July 12. He carried this experiment thru a period of six years. Without exception, in all six varieties, both marketable and total yield increased directly with the earliness of planting. A continuation of this test the following year, with the plantings made on six dates instead of on four, gave the same general results. These tests were conducted on ordinary clay loam soil ■at the Guelph station. Champlin and Winright (1917) compared for two years the yields from planting at intervals of fifteen days from April 1 to July 1. For early digging the April 1 planting, and for late digging the May 15 planting, gave the best average yield for the two years. Such results as these are of value, even locally, only when the tests cover a period of several years. Because of the small variation in date of planting within each region surveyed for the one year, and because of the fact that conclusions on the best time to plant cannot be drawn from the yield of only one season, no attempt has been made to correlate the date of planting and the yield. The average date of planting in 1912 and in 1913, and the average date of the last killing frost in the spring, for the four regions, are shown in table 63. The dates shown in this table indicate that Long Island is the only region in which the crop is planted before the average date of the last killing spring frost. It is evident that the Long Island growers ar^e willing to risk possible damage to the crop from frost in order to enhance the earliness of harvest. The planting season of this region is shown to be at least six weeks earlier than that of the others. 1232 Earle V. Hardenburg TABLE 63. Average Date of Planting, and Average Date of Last Killing Frost IN Spring, for the Four Regions Surveyed Region Average date of planting Average date of last killing frost in spring Long Island, 1912 Steuben County, 1912 Monroe County, 1913 Franklin and Clinton Counties, 1913 April 9 May 20 May 24 May 30 April 25 May 10 May 1 May 10 HAND AS COMPARED WITH MACHINE PLANTING The extent to which the potato crop of any region is planted by machine planters is determined principally by the average acreage, the system of spacing hills in the row, and the amount of large stones present in the fields. The writer (Hardenburg, 1915 a) found that in Steuben County, in 1912, when the average acreage of potatoes per farm was at least 5, the saving in labor cost by machine planting more than overbalanced the interest, depreciation, and repair costs of planting by this method. In regions where checkrowing is practiced, machine planting is impossible because potato planters cannot be used to plant in checkrows. As is shown later, much of the acreage in Steuben County was planted in this way in 1912. Some growers in Franklin and Clinton Counties find it impracticable to use planters because there are so many large stones in their fields. Conditions on Long Island, however, are almost ideal for machine planting, and it is done there almost entirely. The extent to which the acreage in each region was planted by hand and by machine, in the two years concerned, is shown in table 64: TABLE 64. Method of Planting and Type of Machine Used in the Four Regions Surveyed Region Per cent of total acreage planted by Per cent of total machine- planted acreage planted by Machine Hand 2-man planter 1-man planter Long Island, 1912 98 25 74 82 2 75 26 18 23 GO 87 56 77 Steuben County, 1912 40 Monroe County, 1913 13 Franklin and Clinton Counties, 1913 44 70 30 56 44 A Study of Factors Influencing the Yield of Potatoes 1233 It might be presumed that growers of the larger acreages in each region would be more inclined to use planters than those having a smaller acreage. Whether this was true in these four regions is shown in table 65 : TABLE 65. Relation of Size of Potato Acreage to Method of Planting Region Average potato acreage per farm Average potato acreage planted by machine Average potato acreage planted by hand Long Island, 1912 - Steuben County, 1912 Monroe County, 1913 . 24.8 14.6 12.4 7.2 24.6 17.9 13.1 6.5 33.5 13.7 10.6 Franklin and Clinton Counties, 1913 8.6 It is evident that growers in none of these region's find it unprofitable to plant by machine so far as average acreage of the crop is concerned. In Steuben and Monroe Counties there was a tendency to use more planters on the larger acreages. The 2 per cent of acreage on Long Island planted by hand averaged higher per farm than the balance which was machine- planted. The same relation held with the 18 per cent of hand-planted acreage in Franklin and Clinton Counties. It is clear, from tables 64 and 65, that the average potato acreage per farm, considered in the light of percentage of total acreage planted by hand and by machine in each region, has no bearing on the extent of machine planting in these four regions. As indicated in table 64, two types of planters were commonly used. One was of the picker type, employing only one man, while the other was usually of the platform type and required two men for its opera- tion. As the second man on a two-man planter is charged with the duty of seeing that there are no skips, better stands of potatoes are expected from this type of planter. It is shown in table 64 that, whereas about three- fourths of the Long Island acreage was planted with one-man planters, the two-man type predominated in the other three regions. A study of the relative yields obtained from the acreage planted with each type in the four regions is shown in table 66. Of the total of 635 growers using machine planters — about one-half of all the farmers visited — the numbers using each type of planter were approximately equal. . The weighted averages in table 66 sliow that with about the same amount of seed per acre used in each planter, the yield was 19.3 bushels per acre higher from the acreage planted with the one-man planter. This average is not a true criterion of the two types of planters, however, because a large proportion of the total machine-planted acreage was on Long Island, 1234 Earle V. Hardenburg TABLE 66. Relation of Type op Planter to Yield in the Four Regions Surveyed Two-man planter One-man planter Region Number of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Number of farms Average jaeld per acre (bushels) Average amount of seed used per acre Long Island, 1912 Steuben County, 1912.. . Monroe County, 1913'. . . Franklin and Clinton Counties, 1913 74 43 175 22 171.6 158.1 132.1 196.4 12.5 11.8 13.2 14.7 249 28 28 16 175.6 148.8 116.3 181.8 12.5 12.9 13.2 14.4 Total.. 314 321 Weighted average Unweighted average '.'.'.'.'.'. 151.4 164.6 12.8 13.1 170.7 155.6 12.6 13.2 where 77 per cent of the machine-planted area was planted with the one- man type of planter. On the contrary, the difference in average yield, even on Long Island, was only 4 bushels per acre in favor of the picker planter, a difference so small as to lie within expected probable error. In the other three regions, where approximately equal rates of planting were used, the average yields favored the two-man planter by ditt'erences ranging from 9.3 to 15.8 bushels per acre. The unweighted averages in the table furnish a true comparison of the efficiency of the two types of planters, and may be accepted as an indication of the increased yield from a more perfect stand resulting from the use of the extra man on the plat- form planter. CHECKROW AS COMPARED WITH DRILL PLANTING A decision as to whether to plant potatoes in checkrows or in drills involves such factors as the cost of labor, available soil fertility and mois- ture, land value, weed control, and the use of machine planters. Of these factors, weed control is probably the most influential. The statements of many growers in Steuben County concerning their reason for checkrow planting emphasized the facility of weed control by cross cultivation in times when hand labor is scarce or when the pressure of other farm work or a wet season might make weed control otherwise difficult. The cost of labor as a determining factor is debatable, for, while checkrowing may reduce to a minimum the cost of taking care of the crop, the seed may be planted at less cost when planters are used, and planters cannot be used A Study of Factors Influencing the Yield of Potatoes 1235 to plant the hills in checkrows. Moreover, while a marker is carried by the machine planter, extra labor is necessary to mark the field in checks for checkrow planting. As is shown in later tables, less seed per acre is usually planted by the checkrow system than by the drill method, since the seed pieces are spaced farther apart. Therefore, from the standpoint of land economy, checkrowing is the less desirable method where land is high in value, as on Long Island. An ample spacing between hills is shown by the typical checkrowed field in Steuben County illustrated in figure 144. Fig. 144. a checkkowed potato field, common in steuben county The smaller amount of seed planted per unit of space in the checkrow system may be desirable wherever soil moisture and fertility are likely to be taxed to their limit. However, the foregoing studies on the relation between rate of planting and yield do not indicate that this point was reached in Steuben County in 1912. Harwood (1893) reported comparative yields from twenty-four experi- ments conducted at the Michigan station, in which the varieties Early Ohio -and Rural New Yorker No. 2 were planted in hills and in drills. These tests are especially valuable because equal amounts of seed per acre were used in both systems of planting. Altho drill planting did not always give the higher yield, the general average showed a difference of 12 bushels per acre for the Early Ohio and 29 bushels for Rural New 1236 Earle V. Hardenburg Yorker No. 2 in favor of this method. These differences were due, not to a difference in rate of planting, but to the system of spacing the seed pieces. Shepperd and Churchill (1911), using the variety Early Ohio and varying the rate of planting according to the space between seed pieces, compared the yields from planting at distances of from 10 to 36 inches in the row. Here the yield decreased directly as the interspace increased, the 10-inch spacing giving the best yield. Zavitz (1916), in a six-years test, using equal-sized large whole seed and equal-sized medium whole seed, compared the results of spacing the seed 1, 2, and 3 feet. While his total yields increased directly as the spacing decreased, he obtained the largest net yield from the 2-foot spacing of large whole seed and from the 1-foot spacing of medium whole seed. In another test, running for nine years and with the same rate of planting in l)oth systems of spacing, he compared the yields from planting in checks 33 inches apart and from planting in the drill row with the seed pieces 1 foot apart. The results showed a nine-years average difference of 39.8 bushels per acre in favor of the drill-planted seed. From a review of the tests here reported, it appears that, irrespective of rate of planting, the yield from planting in drills is generally better than that from planting in checkrows. None of the 330 Long Island growers who were questioned regarding their 1912 crop had planted in checkrows. The almost universal use of planters in this region precludes the possibility of planting by the check- row method. Furthermore, the greater land values encourage economy of space, and the better yields from closer planting have convinced the growers of this region that drill planting is the better method. The method of planting most common in each of the four surveyed regions is indicated in table 67: TABLE System of Planting in the Four Regions Surveyed Region Per cent of growers planting in drills Per cent of growers planting by checkrow Long Island, 1912 100 29 74 18 Steuben County, 1912. 71 Monroe County, 1913. 26 Franklin and Clinton Counties, 1913 82 In contrast to the conditions on Long Island, approximately three-fourths of the crop in Steuben and in Franklin and Clinton Counties was planted in checkrows. This may be accepted as evidence that relatively cheap land and scarcity of labor make this the better method for these regions. About three-fourths of the Monroe County crop was planted in drills in 1913. A Study of Factors Influencing the Yield of Potatoes 1237 The influence of the planting system on yield and on the amount of seed used is strikingly shown, for these three regions, in table 68. As is indicated in this table, in all three regions the drill system of planting gave an average TABLE 68. Relation of Planting System to Yield in Three of the Regions Surveyed Planted in drills Planted by checkrow Region Num- ber of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertilizer Num- ber of farms Average per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertilizer Steuben County, 1912. . . Monroe County. 1913, . . Franklin and Chnton Counties, 1913 101 221 54 153,0 128.5 188.4 12.2 13.2 14.5 $12,09 11.91 12.94 251 77 243 129.2 120.6 177.2' 9.2 10.2 11.4 $ 9.35 9.68 13.20 Total 376 571. Average 156.6 13.3 $12.31 142.3 10.3 $10 74 yield varying from 7.9 to 23.8 bushels per acre higher than that from plant- ing in checkrows. It is shown further that invariably about 3 bushels more of seed per acre was required to plant by the drill system. The yield was sufficiently higher from this method, however, to more than pay for the extra seed necessary. On the basis of yield alone, these results favor the drill system of planting for all the regions surveyed. However, it is possible that there are seasons in which weed control is largely dependent on the possibility of cross cultivation. The choice at such times becomes one of producing a medium yield by checkrow planting with a minimum of labor, or a much smaller yield by drill planting. depth of PLANTING Depth of planting is a factor which, tho given some experimental atten- tion by various stations for many years, has received little consideration from the potato grower. The depth at which the seed piece is usually placed depends principally on the soil type and the method .of planting. Just as plowing and tillage are normally more shallow in heavy than in light soil, so potatoes are normally planted less deeply in heavy than in light soil. However, the depth at which the potato root system is allowed to develop depends not alone on the depth of planting, but also on the system of tillage employed. This is a factor too often neglected in the ^tudy of the influence of depth of planting on yield. For example, the seed may be planted shallow and the crop given ridge culture, or the seed may be planted deeper and the crop receive level culture. With either method 1238 Earle V. Hardenburg the root system might develop at exactly the same depth. Therefore, in studying this factor by reviewing experimental data, false conclusions may easily be drawn. In studying it by survey methods, however, the problem is not so complicated, because of the fact that approximately the same system of culture is used thruout a given locality. Harwood (1893) reported a test of depth of planting conducted at the Michigan station, using three varieties and planting on sandy loam soil. The depth of planting was varied from 2 to 6 inches. The highest total yield came from the 3-inch planting, while the 4-inch depth was second best. The highest marketable yield came from planting 5 inches deep, altho there was practically no difference between this and the 4- and 6-inch depths. Emerson (1907) compared the yields from plantings at from 1- to 6-inch depths, and obtained a constant increase in yield with each increase in depth up to and including 5 inches. The 6-inch depth gave the second highest yield. Emerson concluded that the better qualit}^ and shape of the tuber resulting from planting from 4 to 5 inches deep, more than offset the extra labor of digging necessary for these depths. Sandsten and Delwiche (1909) harvested the highest total yield from the 4-inch depth of planting, the yield decreasing with the increase in depth below that level. Shepperd and Churchill (1911) compared the yields from plantings at depths of 3, 4, 5, 6, 7, and 8 inches. The 4-inch depth gave 8 per cent higher yield than any deeper planting, and 4.5 per cent more than the 3-inch depth. These investigators did not mention soil type, but reported the greatest yield of marketable tubers and the highest quality from the deeper plantings. Emerson (1914), studying the influence of depth of planting on the value of the harvested crop for seed purposes, planted at 1, 4, and 7 inches. Seed from the 7-inch planting yielded the best, both in total and in market- able yield, in both of the tests he conducted. Also, seed from the 4-inch planting yielded much better than did that from the 1-inch depth. Accord- ing to Emerson, the higher cfuality of the seed from such deep planting is probably due to the fact that it was produced under soil conditions which fluctuated very little in temperature and moisture. Clement and Werner (1917) did not mention soil type in reporting a six-years test on planting at depths of 3, 4, 5, 6, 8, and 10 inches. They obtained tha highest marketable yield from the 4-inch depth, and there was a fairly consistent decrease in yield from plantings above and below that depth. Macoun (1905) made a thoro test of the influence of depth of planting, by comparing the yields for six years, on sandy loam soil, from planting at depths of from 1 to 8 inches. In every one of the six years he obtained the best yield from the 1-inch depth of planting. The second- best average yield came from the 3-inch planting. Since Macoun explained that cultivation during each season eventually placed the seed at a depth of about 2^ inches, it cannot be correctly concluded that 1 inch was a A Study of Factors Influencing the Yield of Potatoes 1239 better depth to plant than 3 inches. Macoun's yiekls decreased rapidly in the plantings that were deeper than 6 inches. Zavitz (1916), using the clay loam soil at the Ontario station for seven years, planted seed 1, 3, 5, and 7 inches deep, and practiced level cultivation. He obtained the best, and practically equivalent, yields from planting 3 and 5 inches deep. He noted that when his seed was planted either shallower or deeper than 4 inches, the new tubers showed a tendency to develop nearer the 4-inch level than at the depth of planting. The evidence presented in the foregoing experiments indicates that, depending to some extent on the soil type and the kind of tillage, the yields are usually better when the seed is planted about 4 inches deep than when it is planted either shallower or deeper. The fact that tubers tend to form near the 4-inch level, irrespective of depth of planting, is in itself an indication that soil moisture and temperature are the most favorable at this depth. While seed planted deeper is normally subjected to tempera- tures too cool for rapid growth, and the resulting crop forms too deep to be dug easily, seed planted less deep is subjected to a greater fluctuation in moisture and temperature, resulting in ill-shaped tubers and very often in a high proportion of sunburned or even blighted tubers. In this investigation an attempt was made to determine whether soil type and method of planting have any influence on depth of planting. The average depth of planting in each region, by machine, by hand, and for the region, is shown in table 69: TABLE 69. Depth of Planting in the Foub Regions Surveyed Region Average depth planted by hand (mches) Long Island, 1912 Steuben County, 1912 Monroe County, 1913 Franklin and Clinton Counties, 1913 3.4 3.1 3.0 2.5 Of the four regions, the deepest planting is found on Long Island and the shallowest in Franklin and Clinton Counties. Inasmuch as the potato soils of these two regions are lighter than those of either Steuben or Monroe County, no influence of soil type on depth of planting is evident in this study. The only significant influence of method of planting on depth is in Franklin and Clinton Counties, where machine-planted potatoes were placed, on the average, 0.7 inch deeper than those planted by hand. Whether or not the average depth of planting shown for each region 1240 Earle V. Hardenburg approximates the optimum depth is considered in the following paragraphs and tables. The relation of this factor to yield on Long Island in 1912 is shown in table 70. The depth of planting appears to have influenced the yield in TABLE 70. Relation of Depth of Planting to Yield on 329 Long Island Farms in 1912 Depth planted (inches) Number of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertilizer Less than 3 68 144 99 18 166.7 178.0 176.7 182.2 12.1 12.5 12.9 12.8 $30.25 3-4 32.60 4-5 33.41 5 and more 33.65 Total .... 329 175.6 12.5 $32.42 this region very little. Altho the average yield increased slightly as the depth increased, the increase in yield beyond the 3-inch depth was no greater than would probably be due to the slight increase in seed and in the value of manure and fertilizer used. Apparently, the average depth of 3.3 inches for 1912 was approximately the best. The results of a similar study in Steuben County are shown in table 71 : TABLE 71. Relation of Depth of Planting to Yield on 360 Steuben County Farms IN 1912 Depth planted (inches) Number of farms Average yield per acre (bushels) Avera-e amount of seed used per acre (bushels) A erage value of manure and fertlHzer Per cent of total yield rotted in field 1-2 6 83 179 92 148,1 139.6 134.4 136.0 10.1 10.2 10.1 10.1 $ 9.39 11.09 9.79 9.97 1 .0 2-3 14.5 3-4 14.7 4 and more 17.4 Total. 360 • Average, 3.1 inches.. 136.4 10.1 $10.06 15.2 A Study of Factors Influencing the Yield of Potatoes 1241 Altho six growers are too few to permit of the drawing of definite conclusions as to the shallowest planting found, there is considerable evidence that planting not more than 2 inches deep in the heavy soils of Steuben County- is desirable, at least in a year as wet as was 1912. Under a constant rate of planting at all depths, and with the least manure and fertilizer used at the shallowest depth of planting, this depth gave the highest average yield and the smallest percentage of field-rotted tubers in 1912. In fact, the percentage of field rot increased with the depth of planting. Assuming that there had been no rot from blight and we't weather that year, the average yield of the fields planted at the shallowest depth would still have been the highest. The relation of depth of planting to yi^ld in Monroe County in 1913 is shown in table 72. It is clear from this table that in 1913, planting shal- TABLE 72. Relation of Depth of Planting to Yield on 263 MonbBe County Farms IN 1913 Depth planted (inches) Number of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertilizer 1-2 ... : 11 67 110 72 151.5 143.1 131. '. 132.6 11.5 12.8 12.6 12.5 $10 43 -3 11 11 3-4 4 f.nd more 11.83 11.91 Total . 260 135.6 12.6 $11.61 lower than the average of this region would have given more than average yields. With the least seed and fertilizer, the shallowest-planted fields gave the highest yields. Since the potato soils of this region are heavier than those of either Long Island or Franklin and Clinton Counties, and, in fact, are rather heavier than ideal potato soil should be, this gives further evidence that potatoes should be planted shallower on heavy than on light soils. With an increase in the value of manure and fertilizer, and an approximately constant amount of seed used per acre, an increase in depth of planting was accompanied by decreased yield on fields planted deeper than 2 inches. The importance of depth of planting as influencing yield in FranMin and Chnton Counties is shown in table 73. .The average yields in this region increased with the depth of planting, down to a depth of 4 inches. A part of this increase must be attributed to an increased use of seed and fertilizer. 1242 Earle V. Hardenburg TABLE 73. Relation op Depth of Planting to Yield on 300 Franklin and Clinton County Farms in 1913 Depth planted (inches) Number of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertilizer Less than 2 44 111 91 54 172.3 175.4 184.9 182.6 11.2 11,9 12.3 12.4 $12 27 2-3 12 95 3-4. . . 13 25 4 and more . 13 25 Total 300 179.3 12.0 $13.01 The evidence in table 73 indicates that the yields began to decrease from planting at depths greater than 4 inches. Evidently, in soils as light as those of this locality, planting at greater depths than the average of 2.6 inches for 1913 may be recommended. Depth of planting is a factor which would require controlled experi- ments covering several years in order to determine the most effective depth for any given region. The evidence of a single year from the sur- veyed regions indicates that in the heavier soils it is safe to plant shallower, and in the lighter soils deeper, than the average depth for 1912 and 1913. DEPTH OF CULTIVATION The term cultivation has been used so promiscuously in agricultural literature that it seems well to define its limitations as used in this study before entering on any discussion of its influence on the yield of potatoes. Cultivation has for its primary objects, weed control and moisture con- servation. Any operation on the crop after it is up, which stirs the soil for either or both of these purposes, is therefore included within the meaning of the term as here used. Such operations as using the weeder, pulling weeds, hoeing, and hilling or ridging the crop, are comprehended by the term. This will account for the great frequency of cultivation noted in the studies of the influence of this factor on yield. Harwood (1893) reported the results of forty-four tests on the influence of depth of cultivation on yield. Considering 1.5 inches as shallow and 5 inches as deep culture, forty of the forty-four tests gave total and marketable yields favoring deep culture. As a rule, the greatest yield of small and sunburned tubers was obtained from shallow culture. Schweit- zer (1896), in a one-year test on potatoes planted 4 inches deep, compared deep and shallow tillage. Altho his yields were almost equal, he obtained A Study of Factors Influencing the Yield of Potatoes 1243 a slightly higher marketable yield from deep cultivation and a slightly higher total yield from shallow cultivation. Information on depth of cultivation in the four regions surveyed, was obtained m relative terms, denoting general depth rather than actual inches The variation m depth thruout the growing season was noted also. It has therefore not been possible to make any definite correlation of this factor with yield m these studies, because of the fact that much variation in opmion may have existed among growers as to just what constituted deep medium, or shallow culture. Furthermore, it is a common practice in all four regions to ridge the rows more or less late in the growing season Ihis practice really amounts to a deep cultivation at the center of the row while httle or none is given close to the plants. Altho considerable varia- tion m the depth of cultivation was found at different times during the growing season, a plurality of the growers practiced relatively deep early- season cultivation and shallow late-season cultivation. Tliis would seem to be good cultural practice, inasmuch as deep tillage early would enlarge the soil area suited to tuber and root development, while shallower tillage later would avoid undue root pruning and disturbance after tuber forma- lion. The practice with respect to this factor, and its apparent influence on yield m each of the regions, is shown in table 74. Of the 1290 growers TABLE 74. Relation of Seasonal Depth of Cultivation to Yield in the Four Regions Surveyed Long Island, 1912 Steuben Coimty, 1912 Monroe County, 1913 Franklin and Chnton Counties, 1913 Type of cultivation Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Deep early and shallow late Medium early and medium late .... Deep early and deep late Deep early and medium late .... 267 9 34 12 172,9 165.1 195.0 149.3 164 112 38 23 141.5 123.3 149.5 121.6 106 21 67 29 146.0 130.3 128.5 122.9 89 58 35 30 183.7 172.1 ■ 175.4 196.2 Total 322 337 223 212 ^ Average 175.2 135.7 136.7 181.2 1244 Earle V. Hardenburg concerned in the whole survey, approximately 1100 practiced one of the four types of cultivation listed in this table. Each of the remainder prac- ticed one of the other possible five types of deep, medium, and shallow early- and late-season cultivation. On the basis of weighted averages, the best average yields were obtained from deep cultivation early in the season and shallow cultivation late. RIDGE AS COMPARED WITH LEVEL CULTURE The system of potato culture in vogue almost universally thruout the New England and Middle Atlantic States has been that of varying degrees of ridging or hilling. An extreme ridging, comparable to that followed in Aroostook County, Maine, is practiced in Franklin and Clinton Counties. In only a few limited areas, notably the Long Island potato areas, is any- thing approaching level cultivation common in these sections of the United States. Regardless of the fact that several station experiments have shown superior merit in point of yield from level culture, ridge culture is by far the commoner. The more obvious advantages of ridge culture consist in (1) greater ease in digging, (2) more efficient weed control by covering rather than by removing and disturbing the root system close to the row, (3) more friable soil for tuber development, (4) protection of tubers from the spores of the late-blight fungus {Phytophthora infestans), and (5) greater surface evaporation of moisture, a factor of special value on heavy soils in regions of possible excess, or poorly distributed, growing-season rainfall. Geismar (1905) compared hill and level culture on both fall- and spring- planted potatoes. His yields favored level culture for both the fall- and the spring-planted crops by 5 and 7 per cent increases, respectively. Geis- mar very mistakenly added these increases and credited the total to the advantage of level culture. This was a blight year at the Michigan station, and, altho Geismar stated that the damage from the disease was confined to the tops, it is possible that some protection from ridging was furnished the ridged plots, and that in a dry year the advantage in level culture would have been even greater. Stone (1905), at the Cornell station, compared various frequencies of hilled and level culture for five years on medium light soil. In each of these years, the yields were best under level culture, the differences ranging from 1 to 37 bushels per acre, the average favoring level culture by 14 bushels. Stone did not explain why the smallest differences in yield occurred in the two driest years, when the greatest advantage from level culture might have been expected. During three of these five years, he compared continuous level culture up to nine cultivations, with laying by and ridging the crop after from three to five cultivations. In these tests, the continuous level culture gave an average advantage of 54 bushels per acre. Shepperd and Churchill (1911), altho reporting no data, stated that level culture has given far better A Study of Factors Influencing the Yield of Potatoes 1245 results than ridging in North Dakota, even in sections having the heaviest rainfall. Since the system of cultivation as well as the depth of planting may have a part in determining the ultimate depth of the seed piece, this factor of depth of seed piece should be controlled in all tests of the influence of system of cultivation on yield. Woods and Bartlett (1909) and Woods (1911) reported a comparison of the yields from shallow planting and high ridging and from medium planting and medium ridging. In these tests the depth of seed piece was constant. The medium-ridge culture gave a three- years average yield of 10 bushels per acre more than the high-ridge culture. Woods (1914), reporting a continuation of these tests but including deep planting and level culture, showed a four-years average yield for the years 1910 to 1913, inclusive, of 276 bushels per acre from medium ridging, 261 bushels per acre from level culture, and 232 bushels per acre from high ridging. Thus, over a long period of years, in a region of relatively high rainfall and with the depth of seed piece constant, the yields favored a system ranging from medium-ridge to level culture. Because of the greater difficulty of harvesting the crop from level culture, however, Woods concluded that, for Maine conditions of soil and climate, there is little choice between these three methods. Macoun (1905) reported four-years average yields, from level and from ridge culture, favoring ridge culture by 22 bushels per acre. Level culture proved the better in one of the four years, and, altho one of the four was a drought year, this was not the year in which the level culture yielded the best. Macoun's results were obtained at the Ottawa station, in a moist sandy loam not subject to drying out. Zavitz (1916), in a nine-years test, obtained an average difference of 7.6 bushels per acre in favor of ridge culture. He stated that three of the nine years were comparatively dry, and in these three years level culture gave the higher yields. Clinton (1916), in a six-years comparison of ridge and level culture, obtained yields slightly favoring ridge culture during three years and yields slightly favor- ing level culture during the remaining three years. He concluded that the only difference in the influence of these two systems on yield is in an advantage from a lower percentage of blight rot under the ridge-culture system. It must be concluded from the above review of experiments that, in general, level culture has given slightly better yields than has ridging. The advantage has been most marked in dry years and in the hghter soils. Depending on regional soil type and seasonal rainfall, however, the advantages generally conceded to ridging should be considered in choos- ing the best system to fit a specific localit3^ Owing to the lack of variation in tillage methods within each of the regions studied, it was not possible to correlate this factor with yield by survey methods. Altho some variation in the degree of ridging exists 1246 Earle V. Hardenburg within each region, lack of information on a definite measurement of this degree makes its use in these studies impossible. Level culture is the system generally understood to be practiced on Long Island. However, nearly all the growers there, while maintaining level culture thruout most of the season, cultivate a slight ridge toward the row late in the season, at either the last or the last two cultivations. The reasons given by growers of the 1912 crop for this practice, were (1) that digging was made easier and (2) that the tubers were protected from the spores of the late-blight fungus. Altho the potato soils of Long Lsland are relatively light in texture, the growing-season rainfall of this region, as shown in figure 127 (page 1149), is relatively high. All growers of the 1912 crop reported the practice of level culture. In Steuben County a system of relatively high ridging is practiced. A ridge is gradually worked toward the row at each cultivation thruout the season, and this is increased late in the season by a speciahzed implement called a hiller. Because of the heavy soil of this region, ridge culture is doubtless of some merit due to the greater ease in harvesting and the protection from blight rot which it affords. All the growers whose 1912 crop was studied practiced ridge culture. Of the 300 growers in Monroe County, 272 reported the practice of level culture, with a slight ridging toward the end of the growing season. The other 28 growers in this region practiced continuous level culture in 1913. Only 1 of the 300 growers in Franklin and Clinton Counties practiced level culture in 1913. Ridging is here begun as soon as the crop is up, the tops, and such weeds as have grown since planting, being covered at that time. By the end of the growing season an extreme ridge has been developed, greater than that used in Steuben County. Altho the grow- ing-season rainfall of this region is almost as high as that of Long Island, the light soils which prevail in most of the section do not seem to warrant such extreme ridging. This is a problem apparently impossible of solution by survey methods and one requiring years of test. FREQUENCY OF CULTIVATION Cultivation as a prime requisite of good crop yields thru its resulting in weed control, moisture conservation, and increased availability of plant food, is one of the oldest known practices of agriculture. However very few experiments of value have been conducted for the express pur- pose of determining the optimum frequency of cultivation. The value of such tests is, of course, dependent on such other factors as duration of the experiment, condition of the seed bed, replication, and time of cultiva- tion. Conclusions drawn must give due consideration to the available soil moisture and fertility and the soil type under which the test is conducted. A Study of Factors Influencing the Yield of Potatoes 1247 Stone (1905), in a carefully controlled experiment covering six years at the Cornell station, compared the yields of potatoes from cultivating three, four, five, six, seven, eight, nine, eleven, and thirteen times during the season. During these years he obtained average yields favoring seven, eight, and nine cultivations, by from 8 to 100 bushels per acre. The plots were replicated from two to four times. Stone's tests showed clearly that under the conditions of his experiment it was possible to culti- vate beyond the limit of maximum production. Emerson (1907) compared yields from what he called poor, medium, and thoro cultivation. Under poor tillage, the land was harrowed three times and cultivated twice, the land not being kept free from weeds even early in the season. Under medium tillage, the land was harrowed three times and cultivated four times, the weeds growing only in the rows after the crop was nearly ripe. Under thoro tillage, four harrowings and six cultivations were given, no weeds being allowed to grow. The yield of the medium-cultivated crop exceeded that from poor tillage by 60 per cent and that from thoro tillage by about 9 per cent. Emerson concluded (1) that tillage can be overdone, (2) that cultivation to control weeds only is sufficient, and (3) that, in eastern Nebraska, two or three harrowings and five or six cultivations are sufficient for potatoes. The high frequency of cultivation recorded for some of the regions included in this study must not be construed to mean that this frequency applies only to operations with a cultivator. As already explained, all operations which stir the soil and control weeds after planting are included. Inasmuch as the rate of planting and the value of manure and fertihzer have already been shown to be very influential on yield, frequency of cultivation is here studied for each region in connection with these factors. Frequency of cultivation on Long Island Long Island is the only region, of the four surveyed, in which the Hal- lock weeder is used extensively. This implement is used principally just before or just after the crop comes up. Having a broad sweep, it removes very efficiently those small weed seedlings which develop between planting time and the time at which the plants come up. Since the entire crop in this region is planted in drills, cross cultivation is impossible and much hand hoeing is therefore done to remove the weeds that develop during the growing season. Many growers reported also hand pulling of weeds. These operations, in addition to the usual cultivations between the rows, resulted in the highest frequency of cultivation in this region, the average in 1912 being 10.9 times. The relation of this factor to yield, under constant rates of {planting, is shown in table 75. The averages for the 329 farms listed in this table indicate that in 1912 it did not pay to cultivate more than ten times. In fact, these averages seem to indicate that frequency of cultivation above 1248 Earle V. Hardenburg TABLE 75. Relation of Frequency of Cultivation and Rate of Planting, to Yield, ON 329 Long Island Farms in 1912 Amount of seed planted Number of times Less than 12 bushels From 12 to 14 bushels 14 bushels and more Average cultivated Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Less than 10 10-13 44 59 35 183.7 158.7 155.7 42 52 32 165.5 183.9 157.4 28 24 13 190.9 194.1 222.8 114 135 80 176.6 177 1 13 and more 168.6 Total 138 126 65 329 Average 166.2 170.2 197.8 175 7 ten was not influential on yield. A further study of the table, however, shows that as the rate of planting increased, the efficiency of the higher frequencies of cultivation increased. This phenomenon may possibly be explained })y considering it in connection with the figures in table 76, TABLE 76. Relation of Frequency of Cultivation and Value of Manure and Ferti- lizer, TO Yield, on 330 Long Island Farms in 1912 Value of manure and fertilizer- Number of times Less than $30 From $30 to $40 $40 and more Average cultivated Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Less than 10 10-13 13 and more 32 57 40 175.3 155.7 151.1 65 62 26 181.2 191.2 163.1 18 16 14 166.5 191.9 217.9 115 135 80 178.4 177.1 168.6 Total 129 153 48 330 Average 158.7 182.2 194.4 175.5 showing the relation of frequency of cultivation and value of manure and fertilizer to yield. The variation in the average yields in tables 75 A Study of Factors Influencing the Yield of Potatoes 1249 and 76 exhibits a very obvious similarity. It is evident in both tables that more than ten cultivations were efficient only when the greater amounts of seed and of manure and fertilizer were used. This may indicate either that a greater amount of tillage was necessary to control the greater weed growth produced by the increased fertility, or that more tillage was necessary to make available sufficient plant food to support the increased stand of potatoes. It may be concluded that, on the average, it did not pay to cultivate potatoes more than ten times in this region in 1912. The coefficient of correlation between this factor and yield, as shown in figure 145, is - 0.087 ± 0.037. This shghtly § Yield per acre, in bushels g 15 8 S r-l O ^ ^ 3- 4 5- 6 7- 8 9-10 11-12 13-14 15-16 17-18 19-20 21-22 23-24 25-26 27-28 1 _ 1 2 3 1 2 1 1 5 5 19 3 11 11 6 3 1 1 11 9 17 10 17 8 7 3 1 1 6 10 20 17 18 2 6 4 4 2 1 3 10 11 5 6 4 1 2 3 2 2 1 5 5 1 1 6 1 1 2 2 1 1 1 1 1 23 13 8 1 10 64 85 90 46 19 10 1 2 1 1 330 1 6 27 38 81 43 60 27 r = -0.087± 0.037 Fig. 145. correlation of frequency of cultivation and yield on 330 long island FARMS IN 1912 negative value indicates that the average frequency was a little too high for maximum production. However, the relatively high probable error renders the coefficient insignificant. Frequency of cultivation in Steuben Comity Cultivation was not so thoroly practiced in Steuben County in 1912, as on Long Island, the average frequency being 7.6 cultivations. Weed control is much more of a problem here than elsewhere because of the fact that potatoes are usually grown on sod land of several 3^ears standing and on land containing a considerable quantity of weed seed or stubble. Furthermore, the seed bed is here more poorly prepared than in most other regions because of the susceptibility of the soil to extreme puddling. Under these conditions, the factor of frequency of cultivation would be I 1250 Earle V. Hardenburg expected to have a direct positive influence on jaeld. The relationship of this factor and the rate of planting, to yield, is shown in table 77: TABLE 77. Relation of Frequency of Cultivation and Rate of Planting, to Yield, ON 349 Steuben County Farms in 1912 Amount of seed planted Number of times cultivated From 6 to 10 bushels From 10 to 14 bushels From 14 to 18 bushels Average Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) 1-4 78 f: 67 2 122.5 123.5 108.3 70 97 15 135.6 141.5 164.2 4 10 6 168.6 201.3 192.1 152 174 23 130.3 4-7 138.0 7-13 168 4 Total 147 .182 20 349 Average 122.8 141.4 191.3 .... 136.4 Four of the groups studied in table 77 contain too few farms to give reliable results, yet the study shows clearly enough that the average yields increased as the frequency of cultivation increased, irrespective of the TABLE 78. Relation of Frequency of Cultivation and Value of Manure AND Fertilizer, to Yield, on 147 Steuben County Farms in 1912 Value of manure and fertihzer From 14 to $12 From $12 to $20 From $20 to $50 Average cultivated Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) 1- 4. . 23 35 10 136.5 142.4 166.0 20 24 4 150.1 172.5 169.9 5 21 5 165.9 165.7 173.5 48 80 19 145 4- 7 . 157 7-13 168.4 Total 68 48 31 147 Average 145.0 162.9 167.1 155.1 A Study of Factors Influencing the Yield of Potatoes 1251 rate of planting. To further allay suspicion that the increased yields, apparently due to increased cultivation, were not in part due to corre- sponding increases in the value of manure and fertilizer, a study of this factor in connection with value of manure and fertilizer is presented in table 78. Altho comparatively few farms are involved in the study in this table, the positive influence of frequency of cultivation on yield is well shown. In contrast to Long Island, it is apparent that growers in this region did not exceed the profitable limit in number of cultivations in 1912. This statement is further proved by the coefficient of correla- tion, 0.231 ± 0.034, shown in figure 146. Fig. 146. Yield per acre, in bushels lO O lO O lO M lO I^ O C-1 -I rt ,-( (M (M eg lO f^ o S LO r- S oi fo 1^ ' O « lO (M (M (M CO CO CO 1 2 4 1 5 2 10 1 3 10 1 10 2 1 1 2 8 9 12 17 9 10 4 2 6 18 15 11 5 5 1 1 1 5 16 6 14 7 10 16 15 5 8 2 1 5 2 5 6 4 4 2 2 3 4 1 1 1 1 1 1 1 1 2 1 1 1 2 2 1 1 11 40 73 65 69 39 23 19 8 4 1 4 1 1 2 5 26 64 73 78 52 38 9 10 2 1 360 r = 0.231 ± 0.034 CORRELATION OF FREQUENCY OF CULTIVATION AND YIELD ON 360 STEUBEN COUNTY FARMS IN 1912 Frequency of cultivation in Monroe County The common rotation of one to two years of cultivated crops, followed by two years of grain, followed by only one to two years of hay, makes the problem of weed control less of a limiting factor to jdeld in Monroe County than in Steuben County. The growing-season rainfall for this region, however, as shown in figure 127, is lower than that for the other three areas, and, because of this, cultivation for moisture conservation might be presumed important. The average frequency of cultivation 1252 Earle V. Hardenburg in 1913 was 8.1 times. The relation of this factor, in connection with the rate of planting and the value of manure and fertilizer, to yield, is shown in tables 79 and 80, respectively: TABLE 79. Relation of Frequency op Cultivation and Rate op Planting, to Yield, on 300 Monroe County Farm.s in 1913 Amount of seed planted Number of times Less than 12 bushels From 12 to 15 bushels 15 bushels and more Average Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Less than 7 7-9 34 41 25 117.1 115.3 113.2 29 30 54 117.9 114.4 133.6 15 23 21 112.1 145.6 172.5 76 124 100 116.6 120.8 9 and more 138.8 Total 100 143 57 300 Average 115.4 122.7 150.1 .... 126.2 TABLE Relation of Frequency of Cultivation and Value op Manure and- Fertilizer, to Yield, on 300 Monroe County Farms in 1913 Value of manure and fertilizer Number of times cultivated Less than $10 From $10 to $20 $20 and more Average Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Less than 7 7-9 9 and more 36 59 41 106.8 114.2 120.6 33 56 47 122.6 125.9 150.1 7 9 12 137.6 133.9 159.2 76 124 100 116.6 120.8 138.8 Total . 136 136 28 300 Average 114.4 134.3 145.5 126.2 Altho increased frequency of cultivation did not produce increased yields for growers using less than 12 bushels of seed per acre, this does A Study of Factors Influencing the Yield of Potatoes 1253 not apply to the growers who used more seed. Furthermore, since the group of growers using less than 12 bushels of seed per acre included those who planted in checkrows, it was doubtless possible to control weeds with fewer cultivations than were necessary for fields planted in drills. A fairly consistent positive influence of frequency of cultivation on yield, irrespective of the value of manlire and fertihzer used, is shown in table 80. It is evident that the highest frequencies of cultivation were pro- ductive of profitably increased yields except for the few growers who, because of checkrow planting which made cross cultivation possible, were able to control the weeds with fewer cultivations. The coefficient 0.169 ± 0.038 (fig. 147), while small, is positive and is significant in value. t- o Yield per acre, in bushels >1 g t2 I I " S I 2 2 1 1 1 1 1 5 6 4 3 1 10 16 8 9 2 2 1 9 14 8 12 4 1 3 1 10 16 14 14 6 2 2 4 6 7 4 1 1 1 1 7 3 4 3 1 2 3 4 1 1 1 1 2 1 1 2 1 1 2 2 1 1 1 1 7 38 70 50 61 34 24 8 4 1 1 1 300 r = 0.169 ±0.038 Fig. 147. correlation of frequency of cultivation and yield on 300 monroe county FARMS IN 1913 Frequency of cultivation in Franklin and Clinton Counties Cultivation in Franklin and Clinton Counties is not generally continued as late in the growing season as in most other regions of the State. On the other hand, ridging is begun early in the season and the crop is given the final ridging soon after blossoming. The average number of cultivations in this region in 1913 was 6.3. As shown in table 15, this was the only region of the four in which a very significant proportion of the total acreage 12.54 Earle V. Hardenburg was plowed in the fall. This practice allows earlier and better seedbed preparation in the spring than would otherwise be possible, and makes later cultivations during the growing season less necessary. The relation of this factor in connection with the amount of seed and the value of manure and fertihzer used, to yield, in 1913, is shown in tables 81 and 82, respec- ABLE 81. Relation of Frequency of Cultivation and Rate of Planting, to Yield, ON 300 Franklin and Clinton County Farms in 1913 Amount of seed planted Number of times Less than 12 bushels From 12 to 14 bushels 14 bushels and more Average Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Less than 6 6-8 46 41 38 164.1 162.5 159.6 49 35 20 184.0 192.0 191.2 30 30 11 189.0 204.8 125 106 177.2 187.7 8 and more 184.5 69 172.0 Total .... 125 104 71 300 Average 162.1 188.1 195.0 179.3 TABLE 82. Relation of Frequency of Cultivation and Value of Manure and Fertilizer, to Yield, on 297 Franklin and Clinton County Farms in 1913 Value of manure and fertihzer Number of times cultivated Less than $10 From $10 to $14 i $14 and more Average Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Less than 6 6-8 44 31 19 173.0 177.7 142.5 41 38 27 177.5 38 188.4 1 37 177.0 1 22 189.0 188.4 188.9 123 106 68 179.1 185.2 8 and more 171.5 Total M 106 1 97 297 Average 168.2 181.3 .... 188.7 179.5 A Study of Factors Influencing the Yield of Potatoes 1255 lively. As shown in these tables, there was no apparent gain in yield in J913 from cultivating more than seven times. In fact, it is question- able whether the small gain shown in most cases from cultivating more than five times was sufficient to pay the extra cost of the labor involved. This means that the average frequency for the year concerned was not far from optmium. The coefficient 0.055 ± 0.039 (fig. 148), while positive, is Yield per acre, in bushels 8 ?i ^ ^ i o 4 > 5 6 7 § 8 •^ 9 o lU Lh II ^ 12 ^ 13 14 t^ o ^ iQ t^ S ^ CM (N 1 3 1 3 2 10 2 3 4 9 6 5 3 32 3 5 14 17 15 18 5 4 ,9 83 3 4 12 7 19 5 9 2 3 64 1 4 6 8 16 2 3 2 42 1 5 6 3 8 3 2 1 29 5 2 3 1 1 3 15 1 2 3 2 1 1 10 1 1 1 1 3 1 1 3 2 1 1 2 1 6 5 12 27 50 26 14 300 52 76 38 r = 0.055 ± 0.039 Fig. 148. correlation of frequency of cultivation and yield on 300 franklin and clinton county farms in 1913 too small and insignificant to indicate any real correlation of frequency of cultivation with yield. It must be concluded, therefore, that in practically all cases sufficiently frequent cultivation was given so that it was not a factor limiting yield. Spraying as a factor in potato production must have been first practiced in this country sometime after 1859. This is the date when the Colorado potato beetle (Leptinotarsa decemlineata) began its movement eastward from the Rocky Mountains (Fraser, 1912). The potentiahty of this pest to cause complete defoliation has since resulted in the extensive use of such arsenical insecticides as paris green, arsenate of lead, and arsenite of soda, for its control. The extent to which insecticides have been used in a given locality has depended on the prevalence of the beetles. The fact that growers in a certain locality did not spray for insects during a certain season, is evidence that insects were scarce or ahiiost absent. A study of the influence of spraying with insecticides on yield in a given region, there- I 1256 Earle V. Hardenburg fore, may not be expected to show positive results. On the contrary, positive influence on the yield from spraying with a fungicide for th^ simultaneous control of late blight {Phytophthora infestans), early blight (Alternaria solani), and tipburn, as well as for the control of flea beetles (Epitrix cucumeris), may be expected in most potato regions every year if the spraying is done thoroly. As a standard fungicide, bordeaux mix- ture has been used for this purpose for about thirty-six years, the practice having begun in France in 1885 (Macoun, 1905). Probably the first systematic and continuous series of spraying experiments with bordeaux conducted in this country, were begun by Jones at the Veimont station in 1891. Lutman (1911) has reported a twenty-years summary of these experiments. During this period, late blight occurred fifteen years out of the twenty, the loss in yield from the resulting rot varying from year to year. These tests showed a gain in yield every year from spraying, the percentage of gain per acre ranging from 18 in a year of no blight, to 215 in a year of much blight, and the average gain per acre for the twenty years being 64 per cent. Altho the frequency of spraying in these tests varied from one to five times a season, the influence of spraying cannot be studied because different frequencies were not used within any one year. Second in importance to the Vermont experiments are those of ten years duration conducted by the New York station at Geneva, under the direction of Stewart, French, and Sirrine (1912). These tests were dupli- cated, one series being conducted on heavy clay loam soil at Geneva, and the other on light sandy loam soil at Riverhead. During the test, late blight occurred six years out of the ten at Geneva and only three years out of the ten at Riverhead. As might be expected, therefore, the greater average gain from spraying was obtained from the Geneva plots. How- ever, there was not one year out of the ten on either series of plots in which a gain from spraying was not obtained. In years of no blight this gain was attributed to the control of such factors as flea beetles, early blight, bugs, and tipburn. Spraying every two weeks during the growing season was each year compared to spraying but three times. With but one slight exception, the more frequent spraying resulted in the higher yield. The ten-years average difference in yield due to this difference in frequency of spraying was 28.5 bushels per acre for Geneva, and 20.7 bushels per acre for Riverhead. Spraying every two weeks gave a ten-years average gain in yield of 97.5 bushels per acre at Geneva and 45.7 bushels per acre at Riverhead. Clinton (1916) reported the results of spraying in a thirteen-year s test at the Connecticut station. Altho no data are presented on the influence of the various frequencies of spraying, increased yields ranging from 10 to 101 bushels per acre (the average being 38 bushels per acre) were reported. At this station, also, increased yields due to spraying were obtained every year, including years of no bhght, in which the average A Study of Factors Influencing the Yield of Potatoes 1257 increase was 29 bushels per acre. In all but four of twenty-two tests, the increased yield more than paid the cost of spraying, the average net gain per acre being $15. Stone (1905) varied the frequency of bordeaux spraying from year to year in a six-years experiment in which imsprayed checki'ows were used. But he did not vary the frequency between plots in any single year. There- fore, no data on frequency of spraying, of any value, are available from this source. A gain in yield from spraying, ranging from 7 to 83 bushels per acre, was obtained during five of the six years. Stone did not attempt to explain the one year of loss apparently clue to spraying, altho the loss averaged more than 30 bushels per acre. Somewhat conflicting data were obtained by Sandsten and Milward (1906) in a two- plot experiment of one year duration. Comparing the results from two, three, five, and six applications of bordeaux to each plot, they found, on one plot, a constant increase in yield with each increased application up to and including five, the increase dropping off slightly with six applications. The second plot showed a general tendency for yields to increase with the frequency of applications, but the data are inconsistent, three applications resulting in a yield lower than that of the check plot, while five applications resulted in a 172-per-cent increase. Testing the influence of frequency of spraying on yield in a blight-free year, Kohler (1909) compared yields from plots duphcated four times and sprayed two, three, four, and six times, respectively. His results showed a decrease in yield of marketable tubers, of 0.7 bushel per acre, from spraying two times, and increased yields of 8.4, 15.8, and 18.7 bushels per acre, respectively, from spraying three, four, and six times, as compared to check plots. A year later, when again there was no occurrence of blight, Kohler (1910) obtained an average increase in yield of 17.4 and 18.8 bushels per acre over the yields of the unsprayed plots, from four and six applications, respectively. Kohler therefore concluded that, irrespective of late blight, better yields may be expected from sprayed plots because of the healthier condition of the foliage. The value of thoroness in applying bordeaux has been well demonstrated by Zavitz (1916) in his report of a seven-years test at the Ontario station. In five of the seven years, no blight rot occurred. In spite of this, both total and marketable yields increased directly with the frequency of spraying, in spraying three, four, and five times during each of the five years. Zavitz found that spraymg both the tops and the bottoms of the plants in all three applications, rather than spraying only the tops, gave an increase in total yield of 13.5 bushels per acre, thus demonstrating the value of thoroness. In 1910, a year of blight rot, spraying from^two to six times gave a proportionate increase in the yield of sound potatoes, a constant increase in the length of the growing period of the plants, and a constant decrease in the percentage of rot in the crop. 1258 Earle V. Hardenburg Macoun (1905) did not test the influence of frequency of spraying, but by spraying four tinies in 1901 and in 1902, and five times in 1904, he obtained an average increased yield, for the three years, of 94.5 bushels per acre. In spite of the loss of millions of dollars to the growers in New York caused by the occurrence of blight every two or three years, and in spite of the proved value of bordeaux mixture as a preventive of this disease, rela- tively few growers make a practice of using a fungicide. Altho blight has frequently been epiphytotic in all four of the regions surveyed except Franklin and Clinton Counties, only one-third of the growers on Long Island in 1912, 5 per cent in Steuben County in 1912, and 25 per cent in Monroe County in 1913, sprayed their crops with a fungicide. Occasional attacks of blight have been observed in Franklin and Clin- ton Counties, but epiphytotics are practically unknown; and even when the fungus is present on the foliage, it seldom attacks the tubers to any serious extent in this region. Only 1 per cent of the growers in this region sprayed for blight prevention in 1913. So few growers used fungicide in Steuben, Monroe, and Franklin and Chnton Counties, that the influence of frequency on yield could not be studied in detail. The extent to which insecticides and fungicides were used in the four regions, and the average yields per acre under the various treatments, are given in table 83. In TABLE 83. Summary of Spraying in the Four Regions Surveyed Long Island, 1912 Steuben County, 1912 Monroe County, 1913 Franklin and Clinton Counties, 1913 Treatment Per cent of farms Average yield per acre (bushels) Per cent of farms Average yield per acre (bushels) Per cent of farms Average yield per acre (bushels) Per cent of farms Average yield per acre (bushels) No spraying Insecticide only 3 64 33 190.7 161.7 197.1 51 44 5 130.2 137.3 171.5 12 63 25 150.3 121.5 126.2 36 63 1 186.1 177.3 152.3 Average 100 175.5 100 136.4 100 126.2 100 179.3 this table, a reliable criterion of the beneficial effects of spraying with fungi- cide is not evident in those cases in which the percentage of growers following any one of the three practices was below 15. Spraying on Lo7ig Island Spraying was done to a greater extent on Long Island than in the other three districts surveyed. Only 9 growers out of 316 did not spray at all in 1912. About one-third of all the growers used a fungicide for the control of blight, tipburn, and flea beetles. Ten per cent of the growers reported A Study of Factors Influencing the Yield of Potatoes 1259 their crops affected with bhght, while nearly one-fourth reported flea- beetle injury. The spraying practices and the average yields, together with the amount of seed and the value of the fertilizer used, are given in table 84. The number of growers not spraying at all was too small TABLE 84. Relation of Spraying Practice to Yield on 316 Long Island Farms in 1912 Spraying practice Number of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertilizer No spraying 9 204 103 190.7 161.7 197.1 12.2 11.8 13.5' S33.08 Insecticide only .... 31.42 Fungicide . . 33.65 316 175.3 12.5 $32.27 to allow of attaching any significance to the average yield obtained by these growers. The increase in yield per acre of over 35 bushels, resulting from the use of a fungicide rather than an insecticide, is significant, altho a part of this difference in yield was due to the use of more seed and fertilizer. Results of further studies on the influence of frequency of spraying with fungicide, on yield, are shown in tables 85 and 86, in which the factors of seed TABLE 85. Relation of Number of Times Sprayed with Fungicide, and Rate of Planting, to Yield, on 109 Long Island Farms in 1912 Amount of seed planted Number of times Less than 12 bushels From 12 to 14 bushels 14 bushels and more Average sprayed Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Less than 5 5-7 11 11 4 154.6 197.2 217.2 15 19 8 189.7 196.5 197.5 7 18 16 181.5 197.0 220.5 33 48 28 179.5 196.9 7 and more 213.8 Total 26 42 41 ■ 109 Average 187.5 194.2 .... 201.6 196.8 1260 Earle V. Hardenburg TABLE Relation of Number of Times Sprayed with Fungicide, and Value of Manure and Fertilizer, to Yield, on 109 Long Island Farms in 1912 Value of murane and fertilizer Number of times Less than $30 From $30 to $40 $40 and more Average sprayed Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Num- ber of farms Average yield per acre (bushels) Less than 5 5-7 15 13 2 161.4 182.8 170.0 12 25 17 177.9 202.8 191.3 6 10 9 236.3 188.6 248.6 33 48 28 179.5 196.9 7 and more 213.8 Total 30 54 25 109 •Average .... 170.0 195.4 228.1 196,8 and fertilizer, respectively, are separated. It appears from table 85 that the yield increased directly with the frequency of spraying, irrespective of the rate of planting. In table 86, the influence of frequency of spraying does not appear to have been so marked. This is due partly, however, to the insufficient number of growers in some of the groups. As a whole, these data indicate that the growers who sprayed the greatest number of times, obtained at least enough increase in yield to pay the extra cost of the labor and materials involved. The correlation of frequency of bordeaux spraying with yield is further shown in figure 149. The coeffi- Yield per acre, in bushels 8 ^ X ?r o ^) u3 o (M (M 1 1 1 9 1 1 2 3 5 1 2 1 1 3 2 4 1 1 3 4 3 3 1 1 3 3 4 3 2 1 2 1 1 5 1 2 2 5 3 2 4 2 1 1 1 1 2 2 1 1 1 16 11 18 27 r = 0.133 zt 0.065 Fig. 149. correlation of frequency of bordeaux spraying, island farms in 1912 11 6 5 3 1 and yield, on 105 Loi A Study of Factors Influencing the Yield of Potatoes 1261 cient, 0.133 ± 0.065, while positive, is not significant because of the relatively high probable error. Since all factors influencing yield are involved in this correlation, such a coefficient need not detract from the real measure of efficiency of bordeaux spraying. Much of interest regard- ing the actual practice of spraying thruout the region may be observed in the frequency table shown in figure 149. Spraying in Steuben County The year 1912 was a year of blight epiphytotic in Steuben County, many of the growers reporting more than half their crop left rotted and un- harvested in the field. More than 93 per cent of the growers found that the late-blight fungus affected either tops or tubers, or both. Such con- ditions should afford excellent means for determining the influence of frequency of bordeaux spraying on yield. The practice of spraying in this region in 1912, the average yield, the rate of planting„and the value of maniu'e and fertilizer per acre, are shown in table 87. The two facts TABLE 87. Relation of Spraying Pbactice to Yield on 360 Steuben County Farms IN 1912 Spraying practice Number of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertilizer 184 160 16 130.2 137.3 171.5 9.9 10.2 11.5 % 9.97 10.23 Fungicide ' 12.96 Total 360 136.4 10.1 $10.06 most clearly set forth in this table are: (1) that, whereas less than half the growers of this region did any spraying in 1912, only about 4.5 per cent used a fungicide for blight control; and (2) that those growers who sprayed the most thoroly also used more seed and fertilizer per acre than the average, and obtained correspondingly higher jaelds. Spraying in Monroe County The treatment accorded the potato crop for blight, tipburn, and flea- beetle control in Monroe County in 1913 is shown in table 88. These data indicate no advantage whatever, as to yield, from fungicidal spraying in 1913. Evidently there was none. The explanation doubtless lies in the fact that the principal functions of bordeaux mixture lie in the protec- tion of the tubers from blight rot and in the prolongation of the plant's 1262 Earle V. Hardenbitrg TABLE 88. Relation of Spraying Practice to Yield on 2S2 Monroe County Farms IN 1913 Spraying practice Number of farms Average yield per acre (bushels) Average amount of seed used per acre (bushels) Average value of manure and fertilizer 33 177 72 150.3 121.5 126.2 12.2 12.2 13.3 $11 40 11 07 11 90 Total 282 Average . 126.0 12.5 $11 33 growing season, thereby increasing yield. The latter function naturally is asserted late in the growing season. In .1913, Monroe County ex- perienced one of the earliest killing fall frosts in its history. As a result, not only the bean crop, but the potato crop as well, was cut down, causing serious loss to the grower. Only about 7 per cent of the potato growers reported the occurrence of late blight up to the time of this frost. It is therefore evident that the possible advantages from fungicidal spraying this year were almost entirely nullified. Under these conditions, frequency of spraying could not be expected to shoAv a normal influence on yield. The correlation coefficient (fig. 150) is 0.084 ± 0.081. This shows insignificant correlation in respect to both the coefficient and its probable error. Yield per acre, in bushels O lO O LQ c ^ ci s^ 4 3 3 4 1 1 3 8 1 4 1 1 1 1 3 2 4 3 1 1 1 3 4 1 1 1 1 2 1 1 11 18 7 17 11 3 r = 0.084 ± 0.081 Fig. 150. correlation of frequency of bordeaux spraying, and yield, on 68 monroe COUNTY FARMS IN 1913 A Study of Factors Influencing the Yield of Potatoes 1263 Spraying in Franklin and Clinton Counties As previously stated, the late-blight fungus (Phijtophthora infestans) seldom attacks the potato crop in Franklin and Clinton Counties. Very probably the reason for this is that muggy atmospheric conditions, so conducive to the disease, seldom prevail here after rains. On the contrary, wide spacing of plants and the frequent breeze that follows rain afford the plants ideal air circulation, thus preventing conditions favorable to blight. Only 3 per cent of the growers reported blight in 1913. The extent to which spraying is practiced in this region is shown in table 89. Altho the three growers who used fungicide used more than the TABLE 89. Relation of Spraying Practice to Yield on 273 Franklin and Clinton County Farms in 1913 Spraying practice Number of farms Average yield per acre (bushels) Average , amount of seed used per acre (bushels) Average value of manure and fertilizer No spraying Insecticide only 98 172 3 186.1 177.3 152.3 12.1 11.8 U.3 •112 59 13 29 15.38 Total 273 179.9 11.9 $13 08 average quantity of seed and fertilizer, they obtained less than the average yield. However, no significance can be attached to this fact, because of the extremely small number of farms. For this same reason, no correla- tion study of the frequency of spraying with the yield in this region has been made. relation of date of harvest to yield The date of harvest of the potato qrop is dependent on such factors as (1) the date of maturity of the crop, (2) the date of the first killing frost in the region, (3) the influence of early market prices, (4) the relation of the potato harvest to other farm work, and (5) the weather. The relative importance of each of these factors varies with the region, in New York State. There are sufficient experimental data available to prove that ordinarily the crop should not be harvested until the foliage is entirely dead because of natural maturity. The basis of this proof lies in the fact that the yield is increased rapidly during the last stages of growth of the plant. Jones (1899) tested the influence of the date of harvest on the yield 1264 Earle V. Hardenburg of potatoes planted on May 20, by digging every ten days from August 2 to September 22. The yield increased from 30 bushels per acre on August 2, to 353 bushels per acre on September 22. Of this increase, 119 bushels came after September 1, and 50 bushels developed during the last ten days. Kohler (1910), working with the variety Early Ohio planted on June 3, similarly tested the rapidity of development of the yield by digging about every seven days from July 31 to August 30. During that period, the foliage developed from an entirely green condition to complete maturity, and the marketable yield increased from 10.9 bushels to 226.8 bushels per acre. There was a gain in marketable yield, of about 7 bushels a day thruout the period, the yield increasing 44.7 bushels per acre during the last week. These data emphasize the possible mistake which some grow- ers make, of digging the crop prior to maturity in order to avoid unfavor- able weather or to take advantage of the relatively high early-market prices. It was not possible, for four reasons, to study by survey methods the influence of date of harvest on yield in the four regions surveyed. First, the information concerning the date of harvesting for Long Island was insufficient; secondly, about 93 per cent of the growers in Steuben County reported the crop more or less affected with late blight; thirdly, a large proportion of the growers in Monroe County reported a killing frost in 1913 which cut down their crop exceptionally early, long before maturity, and reduced the yield much below the average; and fourthly, in Franklin and Clinton Counties the foliage is almost always killed by frost before it is mature, as was the case in 1913. The average date of harvest in the four regions for the years concerned in the survey, and the average date of the first killing fall frost for each region, are given in table 90: TABLE 90. Average Date of Harvest, and Average Date of First Killing Frost, IN THE Four Regions Surveyed Region Number Year of farms 1912 37 1912 348 1913 269 1913 295 Average date of harvest Average date of first killing fall frost Long Island Steuben County Monroe County Franklin and Clinton Counties September 1 September 27 October 12 September 24 October 1-25 October 5 October 15 October 1-10 As is seen in table 90, the Long Island crop was harvested nearly a month before that of any of the other regions, the average date of planting being correspondingly earlier in this region due to climatic conditions. With the exception of Irish Cobbler and other early varieties grown in Nassau County, the crop in this district is usually mature before it is dug. These early varieties are often harvested and marketed before A Study of Factors Influencing the Yifld of Potatoes 1265 maturity in order to reap the benefit of the early-market prices. Further- more, growers of early varieties in Nassau County harvest early in order to be able to follow the potato crop with a crop of vegetables or root crops for the fall market. A harvest of Cobbler potatoes in Nassau County, in the middle of July, is shown in figure 130 (page 1152). It can be noted that the foliage, as separated from the tubers, is not yet mature. On the following day this same field was ridged for turnip planting. As is evident from table 90, the Long Island crop is seldom affected by a killing fall frost. The Long Island growers aim to market their crop as soon after maturity as is possible, in order to supply the New York City market l^efore the earliest crop of other sections of the State is ready to harvest. For the years concerned in this study, the crop in the other three regions was harvested at an average date earlier than the average date of the first fall frost because of the severe blight epideinic of Steuben County, the early and severe frost in Monroe County, and the early frost in Franklin and Clinton Counties. Partly because of the tempering influence of Lake Ontario, the average date of harvest and the average date of the first fall frost in Monroe County are considerably later than for the other regions. Similarly, because of the influence of Lake Cham- plain, the growers located around Peru, in Clinton County, harvested their 1913 crop approximately two weeks later than did other growers in the county. method of harvesting in the four regions surveyed The factors that ordinarily determine whether potatoes shall be dug by hand or by machine are, size of acreage, available labor supply, and soil conditions affecting the efficiency of machine diggers. The author (Har- denburg, 1915 a) found, for Steuben County in 1912, that when the potato acreage per farm was at least 5, the saving in labor cost by machine digging more than outweighed the interest, depreciation, and repair costs of digging by this method. Since the minimum acreage of potatoes per farm recorded in these studies was 5, the factor of economy in the use of machines for digging is probably of no concern in any of the other three regions. There are many farms in Steuben County with fields so steep as to limit the use of heavy elevator diggers. In 1912 the writer (Harden- burg, 1915 a) found the average slope of potato fields dug by hand to be somewhat steeper than that of machine-dug fields. A special type of digger, known as the Boss, or Keeler, which removes the tubers by a rotating reel, has been adapted to the hilly sections of Steuben County because of its exceptionally light draft and its adaptation to slopes too steep for elevator diggers. A study of the influence of slope of field on the type of digger used in 1912 in Steuben County revealed the fact that the fields dug with the reel digger had a higher average slope than those dug with the elevator type (Hardenburg, 1915a). ]266 Earle V. Hardenburg The method of harvesting, and the average potato acreage per farm, for the four regions surveyed, is shown in table 91. It is evident that the TABLE 91. Method of Harvesting, and Average Potato Acreage per Farm, THE Four Regions Surveyed Long Island, 1912 Steuben County, 1912 Monroe County, 1913 Franklin and CHnton Counties, 1913 harvesting Per cent of farms Average potato acreage Per cent of farms Average potato acreage Per cent of farms Average potato acreage Per cent of farms Average potato acreage Elevator digger Reel digger Shovel plow shaker Hand . . 84 5 5 1 5 24.3 21.8 19.1 41.6 11 48 12 23 6 17.0 14.9 13.4 15.4 71 11 3 15 12.9 "l0'7' 9.2 15 1 76 8 8.0 7.5 6 9 Various Average .... 24.8 14.7 12.4 7.2 reel digger is not popular outside of Steuben County, probably decause this digger leaves the tubers in a more or less bruised and scattered condition. As indicated by the figures for both Long Island and Monroe County, the elevator digger was used extensively in these regions, where the soil is relatively light and the fields vary from rolling to level. Growers in Franklin and Clinton Counties have not used the elevator digger exten- sively because of relatively small acreages per farm and an abundance of large boulders, which make the use of such a digger next to impossible. More than three-fourths of the crop in this region was dug by hand in 1913. As a whole, the figures in table 91 show that the average acreage dug by hand was smaller than that dug by machine, and that the average acreage dug by the elevator digger was greater than that dug by any other type of machine. In cases of close planting and heavy top growth, it is often desirable to remove the tops from the tubers before picking them up. In figure 151, a view taken in Franklin County, two men are shown using forks for this purpose, behind an elevator digger drawn by four horses. This illustrates the necessity of using more than two horses because of the heavy draft of these machines. Three types of carriers were found in common use in the regions visited — the standard bushel slatted crate, a hamper basket, and fertilizer bags of various sizes. On Long Island, the commonest carrier in Suffolk County is the fertilizer bag, and that in Nassau County is the fertilzer bag supplemented by hamper baskets of about a bushel capacity. These A Study of Factors Influencing the Yield of Potatoes 1267 .^iT- Fig. 151. removing vines from tubers to facilitate picking up the crop hampers as used in Nassau County, with the owner's initial painted on them, are shown in figure 152. This illustration shows also the common practice in this region of throwing from three to five rows together before picking up the tubers. The prime tubers are then picked up first, the culls remaining until later, as illustrated. Most of the crop of Nassau County — which is marketed directly from the field — is taken, either in those baskets or in bags, by wagon or motor truck, to the Wallabout Markets of Brooklyn, as shown in figure 153. In Suffolk County the crop is taken from the field mainly in bags and is hauled in them to the car or the storehouse, where the potatoes are dumped on the grader if they were not already graded when they were picked up, and are thence emptied into the car for shipment in bulk. The commonest carrier used in the other- three regions is the bushel crate, in which the crop is taken to storage, and there it is either dumped into piles or stored in the crate. B}^ far the greater part of the crop is stored in bulk. In these three regions, the crop is taken to the car either loose in wagon boxes, or in bags, or both ways, with the bags piled on top of the load. 1268 Earle V. Hardenburg Pig. 152. the bushel hampers commonly used for both picking up and hauling to market in nassau county Fig. 153. a Nassau county road wagon, loaded for the wallabout markets of Brooklyn A Study of Factors Influencing the Yield of Potatoes 1269 TYPES of storage IN THE FOUR REGIONS SURVEYED Since Appleman (1912) has shown the importance of low temperature in proper potato storage, it is of interest to note the types of storage used for the crops concerned in these studies. In no case was any grower's crop kept in a refrigerated storage. The proportion of the total stored crop in each region which was stored in various types of storage facihties is given in table 92. The reason for the larger number of farms indicated in Steuben TABLE 92. Types of Potato Storage in the Four Regions Surveyed Region Number larms Per cent of stored crop stored in House cellar Barn cellar Special storage B'arn shed Pit storage Long Island, 1912 . . 231 378 320 300 61 85 72 98 22 11 24 1 12 3 3 1 3 9 Steuben County, 1912 Monroe Comity 1913 1 Franklin and Clinton Counties, 1913 and Monroe Counties than were actually visited in the survey, is due to the fact that a number of the growers in these regions stored their crop in more than one type of storage. Practically all of the Franklin and Clinton County crop of 1913 was stored in the house cellar. In fact, this was the principal type of storage, tho to a lesser extent, in the other regions studied. The next most popular type of storage was the barn cellar. As a rule, both house and barn cellars were constructed with stone walls and dirt floors. Wherever a barn cellar was used, it was generally in close proximity to the stable, advantage being thus taken of the animal heat therefrom to prevent freezing. This was not considered a safe practice in Franklin and Clinton Counties because of the greater severity of the winters in that region. A number of special storage houses were found on Long Island. Since only a small proportion of the Long Island crop is held for more than a few days, these special storages were built not so much to store the harvested crop as for a place of storage for the seed supply brought in from the North to be held until planting time. LENGTH OF STORAGE PERIOD In determining the length of time that the crop was held by the growers in each region, the actual date of sale of all or of parts of the crop was taken as an indicator of the storage period. It was found that a large part of the crop in all four regions was marketed either directly from the field, or 1270 Earle V. Hardenburg after only a few days of holding for proper grading and bagging at the barn. In table 93, the proportion of the crop so handled is considered as not stored. TABLE 93. Length of Storage Period in the Four Regions Surveyed Per cent of crop stored Region For For For For For For For None one two three four five SIX seven month months months months months months months LoiiR Island, 1912. 88 5 3 2 1 1 Steuben County, 1912 64 6 12 6 8 3 1 Monroe County, 1913 38 3 11 21 11 10 5 1 Franklin and Clin- ton Counties, 1913 42 1 12 21 17 1 5 1 As is indicated in table 93, a larger proportion of the crop is stored for one or more months in the three regions of western and northern New York than on Long Island. Except on Long Island, the general practice is to market at least that part of the crop for which there is insufficient storage capacity, at harvest time, the remainder being disposed of as prices warrant and as weather and country roads perm.it. Much of the crop in Steuben County is grown under contract for local buyers. The grower's delivery of this crop mainly at harvest time accounts for the relatively high propor- tion of the crop not stored in this region. SUMMARY Climate, elevation, and soil, as factors influencing yield, were found by this study to be so closely and inseparably related as to make difficult the determination of the influence of each one. The study of available data shows that, whereas the climate for potatoes is generally best at the highest elevations, soil fertility is generally the greatest at the lower eleva- tions. In a year of blight, farms at high elevations are likely to show the best yields ; while in years of no blight, better yields may be expected from the more fertile soils at the lower elevations. The value of potato land as appraised by the growers, proved to be corre- lated with yield up to the point at which the land was affected by real -estate valuation. This point was reached for a few farms located in close proxim- ity to cities or villages. The appraised valuation of these farms was evidently beyond the valuation justified by their productive ability. A Study of Factors Influencing the Yield of Potatoes 1271 From information obtained in the survey of the four regions, it has been possible to determine the status of many factors which, tho not studied as to their influence on yield, have nevertheless a vital relation to production. Among these may be listed the time of plowing, the home-mixing of fertilizer, the analysis of fertilizer, the use of lime, the source of seed, the chemical treatment of seed, the date of planting, the method of planting, the date and method of harvesting, the type of storage, and the length of the storage period. Both biometrical and tabulation studies have shown the amount of seed used and the value of manure and fertilizer per acre to be the most influen- tial factors as relating to yield of all the factors studied. Second to these are depth of plowing, frequency of cultivation, and frequency of spraying. The influence of these five factors, expressed biometrically in terms of r, is summarized for each region in table 94. For obvious reasons, based TABLE 94. Summary of Coefficients of Correlation for Five Factors in the Regions Surveyed Depth of plowing Value of manure and fertilizer.. . Bushels of seed used Frequency of cul- tivation Frequency of bor- deaux spraying. Long Island, 1912 r = 0.159 ± 0.036 r = 0.244 ± 0.035 r = 0.275 ± 0.034 r = -0.087 ± 0.037 r = 0.133 ± 0.065 Steuben County, 1912 : 0.190 ± 0.034 ■■ 0.289 ± 033 : 0.374 ± 0.031 '■ 0.231 ± 0.034 Monroe County, 1913 r = 0.006 ± 0.039 r = 0,258 ± 0.033 r = 0.247 ± 0.037 r = 0.169 ± 0.038 r = 0.084 ± O.OSl Franklin and Clinton Counties, 1913 r = 0.028 ± 0.039 r = 0.169 ± 0.038 r = 0.387 ± 0.034 r = 0.055 rt 0.039 chiefly on environmental differences between the four regions, consider- able variation in the value of coefficients is shown in the table. In a single case, that for frequency of cultivation on Long Island, the coefficient' is negative. Five coefficients out of the eighteen given are too small to be significant, the probable reasons for this being, in most cases, explained in the foregoing text. In but two cases is the probable error greater than the coefficient, these being the probable errors of the coefficients for depth of plowing in Monroe County and in Franklin and Clinton Counties. Probably a more reliable measure of the true influence of these five factors on yield may be obtained from the tabulation studies for ea^h region. In view of the proved importance of these factors, a compari- son of the averages of some of them for the fifty highest- and the fifty lowest-yielding farms in the four regions is given in table 95. In general, the values given in this table confirm the results shown in the discussion of these factors. 1272 Earle V. Hardenburg TABLE 95. Comparison of Fifty Highest- and Fifty Lowest- Yielding Farms of the Four Regions Surveyed, in Average Yield, Potato Acreage, Seed Used, and Fertilizer Used, and Percentage of Growers Spraying with Fungicide Region Average yield per acre (bushels) Average potato acreage per farm Average amount of seed used per acre (bushels) Average value of manure and fertilizer Per cent of growers spraj'ing with fungicide High- est Low- est High- est Low- est High- est Low- est High- est Low- est High- est Low- est Long Island, 1912. . Steuben County, 1912 254.6 204.6 205.1 247.8 95.6 72.4 64.0 114.8 29.8 15.1 13.2 7.0 18.2 12.0 12.8 7.5 13.6 11.2 13.6 13.6 11.6 9.0 11.7 10.7 $35.35 13.65 14.12 15.09 $29.56 7.78 9.08 11.78 50 16 20 16 Monroe Count y, 1913 22 Franklin and Clin- t n Counties, 1913 2 Factors of less, but by no means negligible, influence on yield, as developed by these survey studies, are: method of applying fertilizer, varietal type of potatoes, sun-sprouting of seed, interval between cutting seed and planting, dusting cut seed, type of seed, system of planting, depth of planting, system of cultivation. CONCLUSIONS The foregoing study of crop production by survey methods has, wholly apart from the facts brought out, shown the broad possibilities of this method of research. It does have limitations, however, as is evidenced by certain conflicting data and by the occasionally inconclusive results reported herein. It cannot be used as a substitute either for the present carefully executed research of the state and federal experiment stations, or for more generally localized controlled experiments. On the basis of faets and indications revealed in this study, however, the survey method can and should play a more prominent part in supplementing the present scope of research. In general, too much emphasis has been placed on conclusions drawn from limited experimentation without due attention to their application to local concUtions. Too little research of regional application has been done. Cooperative experiments have been tried, but they have not been sufficiently extensive in duration. A crop survey, to be of greatest value, should be replicated in a given region, depending on the normality of seasonal conditions. The year 1912, while possibly normal for Long Island, was a year of severe loss from blight to the potato crop in Steuben County. The year 1913, while possibly normal for Franklin and Clinton Counties, was a year with an extraordinarily early killing fall frost in Monroe County. These factors have doubtless vitiated to some degree the results of the present study A Study of Factors Influencing the Yield of Potatoes 1273 of the influence of certain factors on yield. A larger number of records for each region, and replication of the survey, may be suggested as the best and probably the only means of obviating these conditions. Whereas sufficient records were not available for the detailed study of some factors, the number used has, on the whole, afforded means for fairly definite conclusions. For as extensive a study of details as has been pursued in the present investigations, not less than three hundred, and preferably four hundred, records should be used. Aside from the relative influence of various factors on yield as revealed in these studies, it has been possible to correct, as well as to verify, many popular ideas of long standing. Altho the " what," the " why," and the " how " of crop production have for years been projected to the farmer, the regional study of actual cause and effect by survey methods has at least contributed to the knowledge as to the " how much." AUTHOR'S ACKNOWLEDGMENT The writer wishes to make acknowledgment of the help given him by Professor E. G. Montgomery, under whose direction this investigation was made, and to express appreciation to Messrs. D. S. Fox, W. M. Peacock, M. F. Abell, R. F. Pollard, and R. H. Cross, who assisted in the taking of records. Grateful acknowledgment is made also to the many farmers who furnished the information necessary, and to Miss Nellie Van Dyne for her capable assistance in the compilation of the data. 1274 Earle V. Hardenburg BIBLIOGRAPHY Adams, L. H. Potatoes. In Report on wheat, oats, barley, corn, and potatoes, for 1887. Wisconsin Agr. Exp. Sta. Bill. 13:8-16. 1887. Report on potatoes for 1889. In Report on oats, barley. and potatoes for 1889. Wisconsin Agr. Exp. Sta. Bui. 22: 7-12. 1889. AicHER, L. C. Whole vs. cut potato tubers for planting on irrigated land. Amer. Soc. Agron. Journ. 9:217-223. 1917. App, Frank. Farm profits and factors influencing farm profits on 370 potato farms in Monmouth County, New Jersey. New Jersey Agr. Exp. Sta. BuL 294:1-103. 1916. Appleman, Charles O. Changes in potatoes during storage. Mary- land Agr. Exp. Sta. Bui. 167:327-334. 1912. Biochemical and physiological study of the rest period in the tubers of Solamim tuberosum. Maryland Agr. Exp. Sta. Bui. 183:181-226. 1914. Physiological basis for the preparation of potatoes for seed. Maryland Agr. Exp. Sta. Bui. 212:79-102. 1918. Arthur, J. C. The potato: relation of the number of eyes qn the seed tuber to the product. Purdue Univ. (Indiana) Agr. Exp. Sta. Bui. 42:103-118. 1892. Ba'lentine, Walter. Investigation on the foraging powers of some agricultural plants for phosphoric acid. Maine Agr. Exp. Sta. Ann. rept. 1893:13-25. 1894. Ballou, F. H. The status of the potato growing industry in Ohio. In Ohio Agr. Exp. Sta., Bui. 218:561-595. 1910. Brooks, William P. Fertilizers for potatoes. Massachusetts Agr. Exp. Sta. Circ. 42. 1914. Champlin, Manley, and Winright, George. Potato culture in South Dakota. South Dakota Agr. Exp. Sta. Bui. 176:694-764. 1917. Clement, P. E., and Werner, H. O. Potato culture in North Dakota. North Dakota Agr. Exp. Sta. Ext. bul. 10:1-12. 1917. Clinton, G. P. Potato spraying experiments, third report. Connecti- cut Agr. Exp. Sta. Rept. 39 (1915):470-487. 1916. A Study of Factors Influencing the Yield of Potatoes 1275 Conner, Chas. M. A preliiuinarv- report on growing Irish potatoes. Florida Agr. Exp. Sta. Bui. 82:391-406. 1906. CuMMiNGS, M. B.' Apple orchard survey of Niagara Count3% Cornell Univ. Agr. Exp. Sta. Bui. 262:277-320. 1909. Dickens. Albert. Potato culture. Kansas Agr. Exp. Sta. Bui. 194: 473-491. 1914. Emerson, R-. A. Potato experiments. Nebraska Agr. Exp. Sta. Bui. 97:1-26. 1907. 'Home mulched vs. northern seed potatoes for eastern Nebraska. Nebraska Agr. Exp. Sta. Bui. 146:1-36. 1914. Flagg, Charles O., Towar, J. D., and Tucker, George M. Experi- ment with seed tubers budded and not budded. Rhode Island Agr. Exp. Sta. Bui. 36:16-19. 1896. Fox, D. S. A farm crops survey in New York, with especial reference to potato production. Thesis for degree of Ph. D., Cornell University. (Unpubhshed.) 1916. An analysis of the costs of growing potatoes. Cornell Univ. Agr. Exp. Sta. Mem. 22:525-627. 1919. Eraser, Samuel. The potato, p. 1-185. 1912. Geismar, Leo M. Report of the Upper Peninsula sub-station foi; the year 1904. In Forty-fourth annual report of the Secretary of the State Board of Agriculture of the State of Michigan, and eighteenth annual report of the Experiment Station from July 1, 1904, to June 30, 1905. Michigan Agr. Exp. Sta. Sp. bul. 31:483-509. (Reference on p. 493- 496.) 1905. GouRLEY, Joseph H. Seasonal report on potatoes, 1909. In Ohio Agr. Exp. Sta., Bul. 218:597-603. 1910. Green, W. J. Experiments with potatoes. In Report of the horti- culturist. Ohio Agr. Exp. Sta. Rept. 6:196-210. 1887. Potatoes. Ohio Agr. Exp. Sta. Rept. 7:116-127. Hall, A. D. Experiments on the continuous growth of potatoes on the same land, Hoos Field, 1876-1901. In The book of the Rothamsted experiments, p. 122-127. 1905. Hardenburg, E. V. Potato machinery efficiency as applied to produc- tion in Steuben County. Thesis for degree of M. S. in Agr., Cornell University. (Unpublished.) 1915 a. 1276 Earle V. Hardenburg Factors influencing potato production in Steuben County studied by the survey method. Thesis for degree of M. S. in Agr., Cornell University. (Unpublished.). 1915 b. Report of potato inspection work in 1917. New York State Potato Assn. Proc. 4:5-9. 19i: Hartwell, Burt L., and Damon, S. C. A twenty-year comparison of different rotations of corn, potatoes, rye, and grass. Rhode Island Agr. Exp. Sta. Bui. 167:1-38.- 1916. Harwood, p. M. Potatoes. Michigan Agr. Exp. Sta. Bui. 95:1-20. 1893. Harwood, P. M., and Holden, P. G. Potatoes. Amounts of seed. Michigan Agr. Exp. Sta. Bui. 93:1-53. 1893. Hume, A. N., Champlin, Manley, and Oakland, I. S. Selection and preparation of seed potatoes — size of seed piece and bud variation. South Dakota Agr. Exp. Sta. Bui. 155:97-112. 1914. HuTCHESON, T. B., and Wolfe, T. K. Potato culture. Virginia Agr. Exp. Sta. Bui. 217:1-16. 1917. Johnson, Samuel. Potato culture. — Varieties. — Experiments with dif- ferent amounts of seed. Michigan Agr. Col. Bui. 13:1-8. 1886. Experiments with potatoes and oats. Michigan Agr. Col. Bui. 34:1-8. 1888. Johnson, T. C. Truck crop potatoes. Virginia Truck Exp. Sta. Bui. 7:131-154. 1912. Truck crop potato fertilizers. Virginia Truck Exp. Sta. Bui. 21:431-452. 1916. Jones, L. R. Certain potato diseases and their remedies. Vermont Agr: Exp. Sta. Bui. 72:1-32. 1899. Jordan, W. H. Commercial fertilizers for potatoes. New York (Geneva) Agr. Exp. Sta. Bui. 187:215-232. 1900. Jordan, W. H., and Sirrine, F. A. Potato fertihzers: method of applica-" tion and form of nitrogen. New York (Geneva) Agr. Exp. Sta. Bui. 327:283-304. 1910. KoHLER, A. R. Potato experiments and studies at Univcrsitv Farm. In Univ. Minnesota Agr. Exp. Sta., Bui. 114:289-333. 1909. " Potato experiments and studies at University Farm in 1909. Univ. Minnesota Agr. Exp. Sta. Bui. 118:65-141. 1910. A Study of Factors Influencing the Yield of Potatoes 1277 LuTMAN, B. F. Plant diseases. Twenty * years' spraying for potato diseases. Potato diseases and the weather. Vermont Agr. Exp. Sta. Bui. 159:213-296. 1911. McMuRRAN, S. M. Pecan rosette in relation to soil deficiencies. U. S. Agr. Dept. Bui. 756:1-11. 1919. Macoun, W. T. The potato and its culture. Cent. Exp. Farm, Canada Agr. Dept. Bui. 49:1-48. 1905. Martin, H. M. An apple orchard survey of Ontario County. Cornell Univ. Agr. Exp. Sta. Bui. 307:161-215. 1911. Montgomery, E. G. Farm crop surveys. Amer. Soc. Agron. Journ. 5:232-233. 1913. MuNSON, W. M. Notes of potatoes. In Report of th§ horticulturist. Maine Agr. Exp. Sta. Ann. rept. 1893:121-124. 1894. Plumb, C. S. Experiments in growing potatoes. Tennessee Agr. Exp. Sta. Bui. 3^:1-24. 1890. Rane, F. William, and Hall, H, F. Ten experiments with potatoes and potato culture for New England. New Hampshire Agr. Exp. Sta. Bui. 111:107-130. 1904. Rane, F. William, and Hunt, Leigh. Potatoes. New Hampshire Agr. Exp. Sta. Bui. 41 : 1-14. , 1897. Rietz, Henry L., and Smith, Louie H. On the measurement of cor- relation with special reference to some characters of Indian corn. Illinois Agr. Exp. Sta. Bui. 148:291-316. 1910. Sandsten, E. p., and Delw^iche, E. J. Potato culture in northern Wis- consin. Wisconsin Agr. Exp. Sta. Bui. 177:1-17. 1909. Sandsten, E. P., and Milward, J. G. The spraying of potatoes for pre- vention of leaf bhght and rot. Wisconsin Agr. Exp. Sta. Bui. 135: 1-24. 1906. Schw^eitzer, p. Enquiry into the principles of potato growing, and tests of varieties. Missouri Agr. Exp. Sta. Bui. 33 : 1-24. 1896. Shepperd, J. H., AND Churchill, 0. 0. Potato culture. North Dakota Agr. Exp. Sta. Bui. 90:81-126. 1911. Smith, J. Warren. The effect of weather upon the yield of potatoes. Monthly weather rev. 43 : 222-236. 1915. Spillman, W. J. Validitv of the survey method of research. U. S. Agr. Dept. Bui. 529:1-15. 1917. 1278 Earle V. Hardenburg Stewart, F. C, French, G. T., and Sirrine, F. A. Potato spraying experiments, 1902-1911. New York (Geneva) Agr. Exp. Sta. Bui. 349:99-139. 1912. Stone, J. L. Potato growing in New York. Cornell Univ. Agr. Exp. Sta. Bui. 228:425-450. 1905. Stuart, William. Place-effect influence on seed potatoes. Vermont Agr. Exp. Sta. Bui. 172:197-216. 1913 a. Good seed potatoes and how to produce them, U. S. Agr. Dept. Farmers' bul. 533:1-16. 1913 b. Group classification and varietal descriptions of some Ameri- can potatoes. U. S. Agr. Dept. Bul. 176:1-56. 1915. Taft, L. R. Potato tests. Michigan Agr. Exp. Sta. Bul. 85:1-21. 1892. Taft, L. R., and Coryell, R. J. Potatoes and the potato scab. Michi- gan Agr. Exp. Sta. Bul. 108:29^7. 1894. Thompson, A. L. Cost of producing milk on 174 farms in Delaware County, New York. Cornell Univ. Agr. Exp. Sta. Bul. 364:109-179. 1915. ToLLEY, H. R. The theory of correlation as applied to farm-survey data on fattening baby beef. U. S. Agr. Dept. Bul. 504:1-14. 1917. Van Slyke, L^ L. Comparative field-test of commercial fertilizers used in raising potatoes. New York (Geneva) Agr. Exp. Sta. Bul. 93(n.s.) : 262-278. 1895. Warren, G. F. The apple industry of Wayne County, New York. Cornell Univ. Agr. Exp. Sta. Bul. 226:227-362. 1905 a. An apple orchard survey of Orleans County. Cornell Univ. Agr. Exp. Sta. Bul. 229:457-499. 1905 b. Agricultural surveys. Cornell Univ. Agr. Exp. Sta. Bul. 344:417-433. 1914. Warren, G. F., and Livermore, K. C. An agricultural survey: town- ships of Ithaca, Dryden, Danby, and Lansing, Tompkins County, New York. Cornell Univ. Agr. Exp. Sta. Bul. 295:375-569. 1911. Welch, John S. Whole vs. cut potato tubers for planting on irrigated land. Amer. Soc. Agron. Journ. 9:224-230. 1917. Wheeler, H. J., and Adams, G. E. Further results in a rotation of potatoes, rye, and clover. Rhode Island Agr. Exp. Sta. Bul. 135: 99-126. 1909. A Study of Factors Influencing the Yield of Potatoes 1279 Wheeler, H. J., To war, J. D., and Tucker, G. M. The effect of liming upon the development of potato tubers. In Fertilizers. Potatoes. Potato scab. Rhode Island Agr. Exp. Sta. Bui. 33 : 46-50. 1895. Whipple, 0. B. Thinning experiments with potatoes. Montana Agr. Exp. Sta. Bui. 106:1-8. 1915. Whitney, Milton. FertiHzers for potato soils. U. S. Soils Bur. Bui. 65:1-19. 1910. Woods, Charles D. Field experiments. Maine Agr. Exp. Sta. Bui. 188:25-32. 1911. Field experiments. Maine Agr. Exp. Sta. Bui. 224:25-48. 1914. Barn and field experiments in 1916. Maine- Agr. Exp. Sta. Bui. 260:85-120. 1917. Barn and field experiments in 1917. Maine Agr. Exp. Sta. Bui. 269:1-44. 1918. Woods, Charles D., and Bartlett, J. M. Field experiments in 1906-8. Maine Agr. Exp. Sta. Bui. 167:85-104. 1909. Zavitz, C. A. Potatoes. Ontario Agr. Dept. Bui. 239:1-88. 1916. Memoir 52, Studies in Pollen, with Special Reference to Longevity, the fifth pieced'ns number in this series of publications, was mailed on March 9, 1922. NEW YORK STATE COLLEGE OF AGRICULTURE CORNELL UNIVERSITY, ITHACA, N. Y. DEPARTMENT OF FARM PRACTICE AND FARM CROPS Farm No Potato Record for 1912. Date . Operator Age P. O County Location Miles to shipping point Soil types Elevation Drainage Exposure 191 Acres farmed Value of land per acre Acres potatoes 4912 Tenure Years owner . Years renter Rotation 1 :2 3 4 5 6 CROP PRODUCTION FOR 1912 Crop Corn, grain Corn, silage Com, other Wheat Rye Oats Barley Buckwheat Hay Alfalfa Oat Hay Oats and Barley Oats-and Peas Field Beans Cabbage Cauliflower Brussels Sprouts - Apples Bearing Apples not Bearing crates bu POTATO PRODUCTION FOR 1911, 1912 AND 1913. Year Varieties Acres Yield per acre Total yield Early 1911 Late ToUl Early 1912 Late . ToUl Early 1913 Late ::::;::::::... -™.^ - -■ — _ Tout ,t.., ^ DISPOSAL OF 1912 CROP Dale BushelB Price per bn. Total Sold Seed Feed Home use SPECIAL EQUIPMENT C: Digger :::::::: Cutter Other equip Total ,. z' ziiz'i. EXPENSES 1912 Amount Price per unit Total RenUl value of land acres .$ tons Manure from preceding crop tons.... Manure, used by 1912 crop tons.... Seed, farm and bought bu lbs lbs Copper sulphate Lime (form for spray) .lbs Insecticide (kind) lbs Equipment Rented Repairs on machinery Depreciation on machinery Man labor hrs. Horse labor hrs. . Equipment labor hrs. Total SUMMARY Total receipts S Total expenses S Profit 1912 crop S Profit per acre $ . LABOR ITEMS Date Acres Per acre ToUl 1 Plowing, Fall 1 i Rolling, times 1 • Removing sprouts, times Starting sprouts ' Marking - Planting, machine _ Planting, hand Fertilizer Hilling, times .::::: Spraying, times Digging and picking up, (hand) i Harrowing after digging : Hauling to storage Hauling to market ... Hauling from storage to market Work on equipment Work on storage Total ■MISCELLANEOUS FACTORS Manure. Kind used Tons per acre Where applied Name of spreader, often in rotation Piled or spread Value of residual manure on potatoes, 1912 per cent Plowing. Depth of plowing Seed. Source Amt. used Seed Treatment. Corrosive sublimate. Formalin. Formaldehyde gas. Flow ,. How treated Satisfactory? Consequent injury to vitality ? Starting Sprouts. Increase in yield noted Cutting Seed. Amt. cut Name of the Satisfactory ? Type of seed planted : 1. Small whole 2. Medium whole 3. Large cut, 3 or 4 pieces . 4. Medium cut, 2 or 3 piece? 5. No. eyes to piece Source, it bought used Plowed in or harrowed ; of sulphur. Formula Increased earliness noted How long cut before planted Fertilizer. Amt. per acre Source of N Home mixed Formula Source of P Ingredients used Cut seed dusted Brand Source of K Amt. r used Insecticide used Amount used Time of application Above, with or below seed Place and frequency in rotation Planting. Date, Early planter How covered Depth planted Coltivating. Type of cultivator No. times in row at each cultivation Hilling. Checkrowing, ridging or level culture '.. Typ xd ised . Spray injury noted ? Yield on unsprayed Digging. How dug What kind would you buy Sorting. Sorted directly Times Bagging or Barreling. Type of carrier used Marketing. Commission rates Shipping rates Storage. Type of storage . Diseases Evident. Effect of manure or lime on diseases Labor. Wages per month Wages per month and board Wages per day and board Work done by women and children Jiills, broadcast, fert. drill, planter, Amt. lime per acr Late , "12. By hand What kind would you buy Distance apart of rows . planter ? Type of Marker used rows Distance apart in ro«s Deep or shallow Cultivated both ways or one Increased yield noted Date of first spraying Name of digger boxes, sorting table, or sort* - Reason Id on sprayed Last spraying, -Type of digger Capacity System of marketing Where shipped , , Description of storage Wages per day ..Board FACTORS ON 1912 COST OF PRODUCTION Total acreage Crop acres Per cent in crops Acres in potatoes Per cent of crop acres in potatoes Value of land per acre Rental value of land per acre Crop index Value of potato crop per acre Potato crop index Total potato production, 1912 Average yield per acre Total man hours per acre Total horse hours per acre Cost of fertilizer per acre Cost of manure per acre Cost of cultivation per acre Cost of spraying per acre , Cost of digging per acre Cost of special equip, per acre Cost per bu Receipts per acre Cost per acre Profit per acre Profit per man hour Profit per horse hour Profit per bushel Per cen of Percen of Per cen of Per cen of Per cen of Per cen of Percen of Per cen of Percen of cost, man labor cost, horse labor cost, land labor cost, manuring cost, fertilizing , cost, cultivation cost, spraying cost, digging cost, special equip. LIBRftRY OF CONGRESS IIM 002 685 241 9 •