THE UNIVERSITY OF ILLINOIS LIBRARY ^> 50.1 vn 5^ cl4 C_o^> . 2. Digitized by the Internet Archive in 2016 https://archive.org/details/sulphurrequireme1424hart (?5A. r f \ A o A \A^‘ c^(Pv 2* AifllVERSiry 0F mjtiQjm t AGRICULTURE library g^ctUMi^ ... * iq Sulphur Requirements of Farm Crops in Relation to the Soil and Air Supply BY E. B. HART AND W. H. PETERSON. It is a well known fact that the wool of the sheep is of a protein nature. This protein, keratin, belongs to the class of proteins known, as albuminoids and is rich in sulphur, contain- ing about two per cent in the air dried condition. Besides the sulphur in organic combination in the pure wool fiber, there is a certain amount of sulphur in other forms in the crude wool. In one hundred pounds of crude wool there may be, approxi- mately, two pounds of total sulphur. This considerable amount of surphur necessary for building the sheep’s fleece has raised the problem of the relative amounts and forms of this element in our common feeding materials and the efficiency of such forms for wool production. Preliminary to the special investigation on the relation of the form and supply of sulphur to wool production which will ap^ pear in future publications, it was necessary to determine the total sulphur in a number of our common feeding materials. The results secured led to a further consideration of the ques- tion of the adequacy of the natural sources of supply of this element for continuous crop production and the data relating to that problem are presented in this paper. It is generally recognized today by agricultural chemists that the amount of sulphur in plant materials, as determined in the ash, is in most cases entirely too low; that in the process of ashing, sulphur is lost and the residual amount found in the ash may represent but a fraction of that originally present in the plant tissue. For that reason it is unfortunate that writers should speak of the sulphur in the ash, so carefully determined 2 Wisconsin Experiment Station. by Wolff 1 as representing that originally contained in the var- ious feeding materials. Such losses of sulphur by ignition have been the subject for study by a number of investigators. Con- tributions to this phase of the question have been made by Berthelot , 2 Barlow , 3 Fraps , 4 Goss , 5 Beistle , 6 Sherman , 7 and others. The work of these several investigators has served to emphasize the inaccuracy of determining the total sulphur of plant tissue by an estimation of that element in the ash. For the purpose of our work a large number of total sulphur determinations have been made on our common farm products. For this work the peroxide method, as outlined by Osborne 8 has been used. The modification of this method, as given in the Official Methods of the Association of Agricultural Chemists , 9 was used for duplicate determinations in a number of cases. Since these determinations were in no closer agreement than duplicates by the Osborne method, the latter was preferred. The addition of sodium carbonate in the official method makes the oxidation slower, the fusion harder to remove from the crucible and, unless care is taken to have the sample well moistened, flashing and burning are very liable to occur when the peroxide is added. Instead of completely boiling off the water from the hydroxide, as directed in the Osborne method, a little should be left. It helps to prevent frothing when the sample is added and secures a more complete saturation of the material with the hydroxide. No hydrochloric acid was used in removing the fusion from the crucible as it was found that hot water would do this com pletely. The black coating on the inside of the crucible is thus retained and the life of the crucible measurably prolonged. Amounts of Sulphur Trioxide in Feeding Materials. The results secured by the Osborne method are reported in Table I. All determinations were made on the air dried ma- terials. For convenience of reference the results are reported both as elemental sulphur and as sulphur trioxide, although in 1 Wolff’s Aschen Analysen. 2 Compt. Rend. 128; 17. s Jour. Amer. Chem. Soc., 1904, 26: 341. 4 N. C. Expt. Sta. Ann. Rept., 1901-3. s N. Mex. Expt. Sta. Bui., 44 e Jour. Amer. Chem. Soc., 24: 1093. 7 Jour. Amer. Chem. Soc., 24: 1100. 8 Jour. Amer. Chem. Soc., 24: 142. 9 U. S. Dept. Agr. Bur. Chem. Bui. 107 . Sulphur Requirements of Farm Crops 3 our discussion of the subject the latter formula for expressing the amount of sulphur will be used. In addition to the total amount of sulphur trioxide reported there is given in a num- ber of cases the amount contained in the ash as secured oy Wolff. Table I.— Sulphur in Feeding Stuffs — Air Dry. Material. ' Sulphur. | Sulphur | Trioxide. Sulphur Trioxide | in ash.* Per cent Per cent Per cent Alfalfa hay 0.287 0.717 0.42-5 0.292 0.730 Bariev 0.153 0.E82 0.060 Barley straw 0.147 0,367 0.207 Beans (white garden) 0.232 0,580 0.130 0.136 0.341 Cabbage 0.819 2.047 1.317 Clover, red | 0.164 0.410 0.222 Corn, white 0.170 0.425, 0.010 0.139 0.347 0.139 0.347 Corn stover, sample 1 0.126 0.315 0.282 0.116 0.290 0.115 0.287 0.128 0.320 <~!nrn silagp 0.122 0.305 Cotton seed meal 0.487 1.217 0.092 Flnnr whpat, 0.180 0.450 Flnnr, graham 0.183 0.457 Glntpn fppd, sample 1 0.560 1.399 Glnten feed, sample 9. 0.5C0 1.249 Hay, mixed 0.160 0.400 0.354 Linseed meal, sample 1 0.404 1.010 0.190 Linsppd meal, sample 2’ 0.375 0.937 Oats, sample 1 0.189 0.472 0.055 Oats, sample 2 0.180 0.450 Oatmpal. sample 1 0.223 0.557 Oatmeal, sample 2 0.228 0.570 Oat straw, sample 1 0.195 0.487 0.230 Oat straw, sample 2 0.218 0.545 Onions 0.568 1.419 0.300 Potatoes 0.137 0.343 Rape tops 0.988 2.470 | 1.132 Rapeseed meal 0.456 1.140 0.381 Rice 0.126 0.315 0.003 Rice bran 0.181 | 0.452 1 0.0.22 Rutabagas, sample 1 0.817 2.041 | Rutabagas, samnle 2 0.632 1.579 Rye 0.123 ; 0.309 | Rye straw 0.049 0.123 Soy bean 0.341 0.852 0.085 Sugar beet, late field sample 0.128 0.320 0.160 Sugar beet, stored roots 0.089 0.222 Sugar beet, early field sample 0.039 0.172 Sugar beet tops 0.433 1.082 0.790 Timothy 0.190 0.475 0.195 Turnips 0.740 1.849 0 897 Turnip tons 0.900 2.249 1.095 Wheat, sample 1 0.170 0.425 0.007 Wheat, sample 2 0.164 0.410 Wheat, sample 3 0.176 0.440 Wheat bran, sample 1 0.200 0.499 0.005 Wheat bran, sample 2 0.224 0.559 Wheat gluten 0.860 2.149 0.011 Wheat straw, sample 1 0.119 0.297 0.132 Wheat straw, sample 2 0.160 0.399 [ — i Wolff’s Aschen Analysen. 4 Wisconsin Experiment Station Table I discloses some very interesting facts. The results are in harmony with those of other investigators in showing a much higher content of total sulphur trioxide in plant ma- terials than is found in the ash. This is particularly true in the seeds where the sulphur exists largely in organic form in the protein molecule. There the loss by ignition is very large. There is one hundred times as much sulphur trioxide in rice grain as in the ash of that grain and forty times as much in the corn grain as in its ash. In wheat the same condition is true, while in the oat, cottonseed and soy bean the total sulphur trioxide recovered in the grain by the fusion method, is about ten times that found in the ash. In onions, cabbage and rutabagas the amount found in the green tissue is from two to four times that recovered in the ash. These materials contain volatile sul- phur oils which are in part responsible for the sulphur lost on ignition. In the case of the hays and straws the losses of sulphur by ignition have not been so large as in the seeds and in those plants rich in sulphur oils. Presumably a considerable part of the sulphur exists in plant stems as sulphates and consequently is permanent on ignition. In mixed hay, for example, the amount of sulphur trioxide recovered in the ash is nearly as large as that found in the original material, but in alfalfa and clover hay the recovery amounts to about 50 per cent. Of course the probable variation in the total sulphur content of different sam- ples of material must be taken into consideration, as the sulphur content of the ash is here quoted from Wolff’s analyses. This would apply especially to such parts of the plant as the stem and leaf, but not in the same degree to the seed the data serve to emphasize the general truth of the large loss of sulphur on ignition from seeds and considerable loss from the tissues of members of the Cruciferae and a somewhat smaller amount of loss from the hays and straws. Amounts of Sulphur Trioxide Removed by Crops. The very important fact which the above data furnishes is that farm crops remove much more sulphur from the soil than has been supposed. Based on Wolff’s ash analyses a 100 bushel corn crop per acre (grain only) would remove about one-half of a pound of sulphur trioxide, while the actual total sulphur • Sulphur Requirements of Farm Crops o trioxide removed, according to our analysis, would be over 20 pounds. For the purpose of clearly showing what amounts of sulphur trioxide are removed by average farm crops Table II has been constructed. Table II — Pounds of Sulphur Trtoxide and Phosphorus Pentoxide Removed Per Acre by Average Crops. Crop. A 1 j Dry 1 weight. 1 1 From Wolff’s ash analyses sulphur trioxide. Actual sulphur trioxide. Phosphorus pentoxide. Pounds Pounds Pounds Pounds Wheat grain, 30 bushels 1,530 0.15 6.4 14.2 Wheat straw 2,653 3.40* 9.3 6.9 Total crop 4,183 3.55 15.7 1 21.1 Barley, grain, 40 bushels 1,747 1.0 6.6 16.0 Barley straw 1 2,080 j 4.1 7.7 4.7 Total crop 3, 827 5.1 14.3 20.7 Oat grain, 45 bushels 1,625 0.8 7.5 13.0 Oat straw 2,353 5.4 12.2 6.4 Total crop 3,978 6.2 19.7 19.4 Corn grain, 30 bushels 1,500 0.15 6.4 10.0 Corn stalk 1,877 5.20 5.6 8.0 Total crop 3,377 5.35 12.0 18.0 Meadow hay 2,822 9.8 11.3 12.3 Red clover hay 3,763 8.2 15.4 24.9 Alfalfa hay 9,000 37.8 64.8 39! 9 Bean grain, 30 bushels 1,613 9 4 <>•7 Q Bean straw 1,848 4.9 6.3 Total crop 3,461 14.3 29.1 Turnips, root 3,126 27.8 57.8 •>? 4 Turnips, leaf 1,531 16.6 34.4 l6l7 Total crop 4,657 44.4 92.2 33.1 Sugar beet root 4,320 6.9 9.5 90 9 1 Sugar beet leaf 1,848 14.5 20.0 13.1 Total crop 6,i68 21.4 29.5 33.3 Potatoes 3,360 1 1 CTO 1 Tobacco leaf 1,800 11. 5Z 1 PL 21.5 Tobacco stalk S’ 200 lo. 8. 5. 8. Total crop 5,000 1 Cabbage 4,800 1 62.8 | Zl. 98. 16. 61. For purposes of comparison, the amounts removed as calcu- lated from Wolff’s ash analyses are given in a number of cases. In addition to the above figures there are also included the amounts of phosphorus pentoxide removed by average crops. This is done for the purpose of comparing the amounts of these 6 Wisconsin Experiment Station two very essential constituents removed by plant growth. It will also serve as a basis for the proper treatment of the ques- tion of sulphur fertilization. A study of Table II reveals the fact that the average crop of seed from the cereal plants removes from the soil about half as much sulphur trioxide as phosphorus pentoxide and that the straws remove a somewhat larger quantity. The hays are not widely different in the proportion of these important ingredients removed, although alfalfa removes annually twice as much sul- phur trioxide as phosphorus pentoxide. The Cruciferae are heavy sulphur-using plants and the total sulphur trioxide re- moved by these is large. An average acre crop of turnips or cabbage appropriates nearly 100 pounds of this compound. Amounts of Sulphur Trioxide in Soils. No one questions the absolute necessity of sulphur for plant growth. It is necessary for the production of plant proteins and all the plant proteins that have been investigated contain sulphur. Only one class of proteins is known to be free from sulphur and that is the class of protamines of animal origin which, however, have not as yet been isolated from plant tissue. The imperative necessity of maintaining an ample supply of this element for plant production is as important as maintain- ing a supply of phosphorus, nitrogen or any of the other ele- ments essential for plant development. The apparent reason why so little attention has been given to this element in the schemes of fertilization for plant production has been due to the fact that it was believed that crops removed but little from the soil and consequently the supply was ample for continuous pro- duction. Bogdanov 10 in 1899 called attention to .the practical impor- tance of sulphur in agriculture and believed that' it should be applied occasionally as a sulphate for the express purpose of maintaining in the soil an adequate supply of this element. Dymond. Hughes and Dupe, * 11 have also touched upon this sub- ject and concluded that there was not sufficient sulphur in the soil for the greatest yield of crops rich in protein, but that for cereal crops and permanent pasture the soil and rain would pro- 10 Abstract Expt. Sta. Rec., 11: 723. 11 Jour, of Agr. Sci., 1905, 1: 217. Sulphur Requirements of Farm Crops vide a sufficient quantity. So far as we can find in the litera- ture no further treatment of the subject has been made. Investigation of the forms of sulphur in soils has been made by Berthelot and Andre. 12 According to these investigators sul- phur exists in soil as (1) sulphates and sulphides; (2) ethereal sulphur; (3) in organic compounds. Very probably the chief forms in all normal soils will be those existing as sulphates and those in the organic matter of the soil. Sulphides would rarely exist except in waterlogged soils where reducing processes are prominent. Ethereal sulphates would probably be found only after fertilization with the urine of animals, and then only in small amounts. There are some difficulties in the estimation of the total sul- phur of soils. Fusion of the soil itself with an alkali is some- what tedious, although probably the most accurate procedure. The large amount of silicates and their incomplete removal be- fore final precipitation with barium chloride, may endanger the absolute accuracy of the method. Extraction in the wet wa^ with strong oxidizing agents, while probably not absolute, will nevertheless give the amount of sulphur that can reasonably be expected to become available to the crop in future years. Van Bemmelen 13 compared several methods for the estimation of sul- phur in soils, obtaining the highest results by extraction with aqua regia. Trials by the method of ignition with sodium car- bonate and potassium nitrate and the method of ignition in a stream of oxygen as used by Berthelot gave him slightly lower results than with aqua regia. It is safe to assume that diges- tion and extraction with a strong oxidizing agent will remove all sulphates, oxidize sulphides to sulphates, and split up any ethereal sulphates present to form sulphuric acid, and at least partly oxidize the sulphur existing in organic forms. It prob- ably will not give the total sulphur present in the organic ma- terial of the soil. Most of the determinations of sulphur in soils carried out heretofore have been made by extraction with strong hydro- chloric acid. This has been done either by the long time ex- traction method as recommended by Hilgard or by the method adopted by the Association of Official Agricultural Chemists. 12 Ann. Chem. et Phys., 1892, 25 : 305. is Land. Vers. Stat., 1890, 37 : 284. 8 AVisconsin Experiment Station These methods will presumably give all sulphates and ethereal sulphur, but not that existing as sulphides, or the sulphur in combination with the organic matter of the soil. Hilgard found from his analyses of many types of soil, that the average amount of sulphur trioxide in sandy soils was 0.055 per cent wdiile in the clay soils examined, it amounted to 0.075 per cent. This, on the basis of an acre foot of three million pounds, would amount in the first instance to 1,650 pounds, and in the latter case to 2,250 pounds. The average amount of phosphorus pentoxide as given by Hilgard for the same sandy soils was 0.087 per cent and 0.141 per cent for the clay soils. On the same basis of cal- culation an acre foot would contain in the first case 2,610 pounds and in the second 4,230 pounds of phosphorus pentoxide. AVhitson 14 and Stoddart also have shown that the average phosphorus pentoxide contents for the surface eight inches of many AVisconsin soils is about 2,940 pounds. These figures are introduced for the purpose of showing that the amount of sulphur in all normal soils is comparatively low. Of course the quantity would not be considered low if the amount removed by crops was also relatively small; but com- pared with phosphorus the amount of sulphur in normal soils is on the average percentagely lower, while the amount of sul- phur removed by crops is, relative to the supply, quite as large. In some cases, at least, as with the turnip, cabbage and other plants of the Cruciferae family, the amount of sulphur removed is very much greater than the amount of phosphorus removed. Effect of Continuous Cropping on the Sulphur Content of the Soil. To determine definitely the effect of continuous cropping on the sulphur content of soils, a number of analyses of both cropped, virgin and manured soils have been made. Two methods were used for the determination of sulphur trioxide. The first was a slight modification of the Yan Bemmelen method. That investigator used aqua regia for the extraction. Our modification was as follows : — 10 grams of the air-dried soil were placed in a flask, fitted with a ground glass stopper, in which was fused a glass tube about two feet' long to serve as 14 Wis. Expt. Sta. Res. Bui. 2, 1909, p. 42. Sulphur Requirements of Farm Crops 9 a return condenser. Two to three cc. of bromine were added to the soil and then 50 cc. of strong nitric acid, sp. gr. 1.42; the flask and contents were then thoroughly shaken and finally placed on the steam bath and digested for 24 hours. Bromine was added occasionally to replace that lost by evaporation. The whole operation was, of course, carried on in a well ventilated hood. After digesting 24 hours the flask was cooled, fil- tered and the residue thoroughly washed with hot w^ater until at least a volume of A 200 cc. had been collected. The clear fil- trate was then evaporated to dryness and the residue treated with hydrochloric acid. This solution w 7 as again evaporated to dryness and the operation with hydrochloric acid repeated twice in order to insure complete removal of all nitric acid. The residue was then taken up with water slightly acidulated with hydrochloric acid, digested hot for half an hour, filtered and thoroughly washed. The solution was neutralized with am- monia, and then made acid with 4 cc. of hydrochloric acid brought to boiling and precipitated with barium chloride. Af- ter standing w^arm for 24 hours the precipitate was filtered, ignited and weighed. The second method consisted of fusion with sodium peroxide. Ten grams of soil w r ere placed in a 100 cc. nickel crucible, moist- ened with water, about 10 grams of a weighed 20 gram portion of sodium peroxide added, and the mixture thoroughly stirred with a platinum rod. The crucible was placed over an alcohol flame and heated moderately until the mass was dry. The re- mainder of the sodium peroxide was then added, the cover placed on the crucible, strong heat applied until the mass melted, and kept in this condition for 10 minutes. It was then allowed to stand over a lower flame for 1 hour. The crucible was re- moved, cooled, placed in a 600 cc. casserole, hot water added and the fused mass removed. It was neutralized with hydro- chloric acid and then further acidified with 10 cc of hydro- chloric acid. The volume was made up to about 450 cc. and boiled for 15 minutes, or until no undecomposed portion of the fused mass remained on the bottom. The covered casserole w^as allowed to stand on the steam bath over night, filtered through a ‘ ‘ nutsche ’ ’ and the residue thoroughly w r ashed with successive small portions of hot water. The filtrate and washings, if over 500 cc., were evaporated below that volume, refiltered and the 10 Wisconsin Experiment Station volume made up to 500 cc. Aliquots of 250 cc. each were heated to boiling, barium chloride added, boiled for 5 minutes and set aside on a steam bath for 24 hours. The volume was not allowed to decrease as silicic acid may be precipitated if much evaporation takes place. After standing for this length of time the barium sulphate was filtered off, washed, ig- nited and weighed. In the determinations made by this method the precipitate was free from silica as demonstrated by the hydrofluoric acid test. Both these methods were compared with the method of the Official Agricultural Chemists with the following results : The soil used was from the surface 8 inches of one of the experi- mental plots at the University Hill Farm. Per cent of SO r Official method 0.019 Fusion method 043 Nitric acid-bromine method 037 Since the official method gave lower results than either the fusion or wet extraction method it was not used for our work. Instead, for the purpose of greater completeness, all the soils investigated were subjected to both the fusion and nitric acid- bromine methods. There was close agreement between the re- sults secured by the two methods, but with a uniform tendency for the fusion method to give slightly higher results. This is to be expected as that method should include the total sulphur of the soil, while by the wet method some of the organic sulphur may escape complete oxidation. In the analyses reported only those results secured by the fusion method are recorded. Sulphur Trioxide in Soil Samples. Ten different soils from several parts of this state were first investigated. They were both virgin and the same soils, cropped but generally unmanured. Some of these were kindly furnished us by the Soils Department of this Station and had already been the subject for investigation by that department of losses in phosphorus. The determinations of sulphur trioxide were made on pairs of samples, one from the cropped field and the other from the adjacent virgin soil. In each case the surface 8 inches were taken and every precaution to secure virgin soils of drain- Sulphur Requirements of Farm Crops 11 age and topography similar to that of the cropped soil was ob- served. Analyses are also given in the later part of this article of virgin soils and the same soils heavily manured, but in two instances continuously cropped to such heavy sulphur using plants as the cabbage. The following is a brief history of the unmanured samples. Samples 1, 2 and 3 furnished by the Soils Department, were identical with those described in Research Bulletin No. 2 15 of this Station, page 44, ,s under numbers 7, 8 and 9, respectively. No. 1 (Cropped).' “Janesville. Cropped 63 years. Dur- ing the first 34 years wheat was grown almost continuously. Since 1878 it has been rotated to corn, barley, oats and rye. It has never been seeded down or manured and is in a badly ex- hausted condition.” No. 2 (Cropped). “Edgerton. Cropped about 60 years, largely to wheat at first, but during the last 40 years it has been farmed in a rotation, consisting of two crops of corn, one of oats and two of clover and timothy, of which the first was cut and the second pastured. The field has not produced good crops during the last 10 years. It has been manured but once.” No. 3 (Cropped). “Milton Jet. Cropped 52 years, chiefly to wheat during the first 15 years, since then it has been rotated to oats and barley with a few years of timothy. It has been manured 10 times with an average of 10 loads per acre each time. The field dees not now produce good crops.” No. 4 (Cropped). “Milton Jet. Cropped 50 years. Rotated to corn and oats for about 40 years, then grew one crop of bar- ley, one of corn, one of beets and now in alfalfa. Alfalfa doing poorly. Soil acid. Manured once.” No. 5 (Cropped). “Evansville. Cropped 60 years, early years chiefly in wheat. History for last 28 years — 10 years timothy and clover ; 12 crops of corn ; 6 crops of oats. Manured three times. Manure usually sold.” Table No. 3 gives the results of the analyses of the virgin and cropped soils ; the loss of sulphur trioxide by cropping as deter- mined from the analyses; the estimated amount of sulphur tri- oxide removed by the crops; the amount added in the manure; and the amount removed by the crop in excess of that added in Whitson, A. R. and Stoddart, C. W. “Factors Influencing the phos- phate content of soils.” Wis. Expt. Sta. Research Bui. 2, 1909, p. 44. 12 Wisconsin Experiment Station the manure. It is estimated that the average amount of sul- phur trioxide removed every year by corn, the small grains, clover and timothy was 12 pounds. This amount is somewhat below the figure given in Table II for average crops, but the smaller amount is taken because of the falling off of crop yields in the later years of the field’s history. The average amount of sulphur trioxide added in the manure was estimated on the basis of several analyses, at two pounds per ton of manure with an application of 10 tons yearly. Table III indicates that on the average about 40 per cent of the sulphur trioxide has been lost from the cropped soils. In every case there is a lower percentage of sulphur trioxide in the cropped soil than in the virgin soil. The estimated amount removed by the crops, minus that supplied in the manure, is in two cases in excess of that indicated to have been lost by the analysis. Such discrepancies are due, probably, to variations in the soil itself, inaccurate estimates of the amounts of crop removed, and the influence of the influx of sulphates with the upward movement of soil water. On the other hand the esti- mated average amounts of sulphur trioxide removed by the crops and tflat indicated by the analyses, are almost identical. This estimate cn the soil is based on the surface 8 inches and the weight of two million pounds per acre. Little importance, however, should be attached to these estimated amounts removed, as there must still be losses of sulphur trioxide due to drainage which the figures do not include. The important thing that the table teaches is, that continuous cropping without adequate fer- Table III. — Influence of Exhaustive Cropping on the Sulphur Trioxioe Content of Soil. Soil sample No. ! SOn in virgin s. il. SO a in cropped soil. Amount Per Acre (Estimated). Loss by crop ping'arcot-ding - to analyses. Removed by crops. Added in manure. Remo ed by crop in excess of that added in manure. Per cent. Percent. Pounds. Pounds. Pounds. Pounds. 1 0 107 | 0 . 0fi6 8?0 756 7^6 0.101 0*058 860 720 20 700 3 0.033 0.010 280 624 200 424 4 0.127 0.088 780 600 20 580 0.055 0.032 460 720 1,0 660 Average. 0.084 0.052 640 684 60 624 Sulphur Requirements of Farm Crops 13 tilization shows a large decrease in the snlphnr trioxide content of the soils examined. Effect of Liberal Manuring on the Sulphur Trioxide Con- tent of Soils. In order to determine what would be the effect of liberal manuring with farm manure on the sulphur trioxide content of soils, several samples of known history were subjected to an examination. The corresponding virgin soils were also exam- ined. The history of the cropped samples is as follows: 1 No. 6. “Evansville. Cropped 60 years. Wheat up to 1860. Since 1860 general rotation of two years in corn, one to two years in meadow, and one year in pasture. Manured every year ' while in corn at the rate of 10 loads to the acre. Crop yields good. ? ’ No. 7. “Berryville. Continuously in cabbage until the crop failed some 10 years ago. Manured heavily with stable manure until three years ago. The last three years fertilized annually with 1,100 pounds per acre of Homestead fertilizer. Surface two inches rejected in taking samples.” No. 8. “Kenosha. Past 15 years in cabbage, alternating every year with corn or potatoes. Chicago stable manure freely applied with yearly applications of about 10 loads. None used in 1909.” The results are given in Table IV. Table IV. — Influence of Liberal Manuring With Farm Manure on the Sulphur Trioxide Content of Soils. Soil sample No. SO s in virgin soil. SO :> in cropped and manured soil. 6 Per cent. 0.061 .103 .110 Per cent. 0.075 .05 .140 7 >8 Average .096 .110 The results show that the sulphur content of these soils was maintained and even slightly increased by the liberal applica- tion of farm manure. 14 Wisconsin Experiment Station Other Factors in the Gain and Loss of Sulpher Trioxide from Soils. In addition to the sulphur added to the soil with applications- of farm manure or certain commercial fertilizers there is also a small amount brought to the land by rain water. This sul- phur in the atmosphere has its origin almost wholly in the burn- ing of fuel, especially soft coal. Where soft coal is abundantly consumed the amount may be expected to be somewhat larger than where anthracite coal is the chief fuel used. The sulphur exists in the atmosphere as sulphur dioxide and trioxide and is brought down in the rain either as the free acids or salts of these acids. It is very probable that part of the sulphur diox- ide is oxidized to a trioxide and reaches the soil as free sulphuric acid or 'sulphate. The amount of sulphur trioxide brought to an acre surface yearly has been determined at Rothamsted, Eng- land. WaringtoiW gives the amount at about IT pounds and adds further that sulphates in rains “will, to a considerable ex- tent, meet the demands of most cultivated crops.” Later determinations made at Rothamsted and furnished us by Director Hall, show an average annual precipitation per acre of about 1 Sy 2 pounds of sulphur trioxide. In England large amounts of soft coal are burned. It was believed not improbable that the amount brought to a surface acre annually by the rain in the open country of our northern states, especially where soft coal is used to but a limited extent, would not be as large as the amount found in England. For the purpose of collecting data on this point a rain-gauge with a collecting surface equal to one square foot was set up at the Uni- versity Hill Farm located three miles west of the city of Wadi- son, Wisconsin. It was placed in an open field and at least a mile from any chimney using considerable quantities of coal. The nearest source of burning coal was that used in running a stone crusher about one mile east of the location chosen for the gauge. As the prevailing winds in this part of the country are from the west or south, the direct influence of this chimney as a source of sulphur was doubtless minimized. The rain water was collected in a copper bottle, containing a trace of sodium bi- carbonate in order to fix any sulphurous acid. A fine wire ic Chemistry of the Farm, 9th Ed., p. 19. Sulphur Requirements of Farm Crops 15 screen was placed above the funnel of the gauge for the purpose of excluding any material accidentally blown or falling into the gauge. The samples of water collected were analyzed for total sul- phur trioxide at the end of each month. For this determination the water was evaporated to dryness in a platinum dish, treated with bromine and nitric acid, again evaporated to dryness, the nitric acid displaced by evaporating several times with hydro- chloric acid and the* sulphates finally precipitated with barium chloride. The results are given in Table V and reported as pounds per acre of sulphur trioxide ; the monthly rain fall at the time the analyses were made is also recorded. Table V.— Amount in Pounds of Sulphur Trioxide Broug-ht to Surface Acre Monthly. June. 1910 July. 1910 A u trust. 1910 September. 1910 October. 1910.... Total Month. SOs Pounds. Rain fall. Inches. 2.3G 1.31 0.60 0.81 4.47 6.56 2.66 1.83 0.61 0.63 10.70 11.14 Not enough data are at hand to warrant establishing an aver- age annual figure, but until such data are accumulated the ten- tative statement can be made that the amount of sulphur tri- oxide brought yearly by rain to an acre surface of land at this location will probably not be less than that found at Rotham- sted, England, which is 17 to 18 pounds. The total amount brought down in the five months recorded was 10.70 pounds with a rain fall of 11.11 inches. Some variations in the sul- phur content of the atmosphere may be expected from season to season, as well as variations incident to frequency of rains; for this reason a yearly figure must be based on actual deter- minations over that period rather than on calculations of pre- cipitation per inch of rain fall based on determinations over only short periods of time. .16 Wisconsin Experiment Station Losses of Sulphur Trioxide from Soils by Drainage « While the atmosphere can serve as a considerable source of sulphur trioxide to the soil, nevertheless as a compensating fac- tor, it is probably more than offset by the losses that soils sustain by drainage. It is well known that most river and lake waters, which in part represent the drainage waters from our soils, con- tain considerable quantities of sulphates. These have been dis- solved out of the soil. The Kothamsted Experiment Station gives some definite data on the losses of sulphur trioxide in the drainage water from soils. The drainage waters from the wheat plots of Broadbalk field have been collected from time to time and completely analyzed by Yoelcker and Frankland. From the unmanured plots they report the quantity of sulphur trioxide in the drainage water at 24.7 parts per million, while the quan- tity varied from 41.0 to 106.1 parts per million in the waters from the plots receiving various fertilizer treatments. Assuming, as Hall does 17 in his discussion of losses by drain- age from these plots, a mean annual drainage equal to 10 inches of water, the unmanured plots would lose approximately 50 pounds of sulphur trioxide annually per acre, while those re- ceiving fertilizers containing variable amounts of sulphates would lose from 85 to 220 pounds annually per acre. It will be seen from the above data that the loss of sulphur- trioxide by drainage is considerable and in the case of the unmanured plots it is nearly treble the amount brought by rain to an acre sur- face from the atmosphere. While these figures may not be ap- plicable to all climates and all soils, nevertheless it would be conservative to state that the loss by drainage at least equals and probably exceeds the amount brought to the soil from the atmosphere in the humid regions of America. General Discussion. The general question raised by the data presented above is one of great importance. The fact that common crops remove from the soil considerable quantities of the element sulphur, while the compensating factor of supply from the atmosphere is very probably offset by the losses which the land sustains by .17 The Soil, p. 200. Sulphur Requirements of Farm Crops 17 drainage, makes it apparent that for the maintenance of a per- manent supply of sulphur in the soil, this element must be added systematically either as a constituent of commercial fertilizers, or with the farm manure. The supply of sulphur in soils is low, the surface eight inches of a normal soil containing sulphur trioxide sufficient for about 100 crops of barley, and this would suppose that the crop could produce normal yields with a steadily decreasing supply of sulphur. The upward movement of water within the soil occa- sioned by surface evaporation, would possibly aid in bringing sulphates from the lower soil areas, but this could not continue indefinitely. The sulphur trioxide content of sub-soils accord- ing to the analyses made by Hilgard is no greater than that of the surface soil. Many factors are operative in the maximum production of crops ; ample supplies of essential elements, proper biological agents, absence of toxic substances, and proper physi- cal environment are all important soil factors. So many fac- tors are operative that an optimum condition is probably seldom obtained. For this reason the disturbance of the equilibrium in the soil by the addition of a single agent is often followed by better crop production. Whether it affects the biological agents, improves the sanitary conditions, alters the physical status of • the soil, or acts by furnishing an ample supply of the elements essential for plant growth is always difficult to answer, but that there must be maintained an ample supply of the essential ele- ments. is one of the first principles of plant production. Most normal soils contain an abundance of the essential ele- ments, potassium, iron, magnesium and calcium. Nitrogen, phosphorus and sulphur alone are percentagely low and in a class by themselves. The former two are today recognized as valuable essentials of commercial fertilizers and manures, and field experiments in many instances have shown the utility of the application of potash salts in available form. When calcium is added it is usually as a carbonate and more for the purpose of maintaining a neutral or slightly alkaline reaction in the soil solution than as a source of needed calcium for plant growth. The use of sulphates in fertilizers has been unconsciously practiced for many years. The acid phosphate of the fertilizer industry is a product containing a large proportion of calcium sulphate. Whether the beneficial results accruing from its ap- plication are to be attributed alone to the phosphorus it supplies 18 Wisconsin Experiment Station or whether they are twofold and due to both the phosphorus and sulphur contained in this material are questions raised by this investigation. It is not impossible that the superior results sometimes obtained in field practice with acid phosphates over other phosphates, such as Thomas Slag, or ground rock phosph- ates, are not due entirely to a difference in solubilities, but to the additional sulphur supplied by the former material. The supe- rior results so often noted from the use of potash salts when fur- nished as sulphates rather than chlorides, may rest in part upon the sulphur content of the material added. It was once common practice to use gypsum as a fertilizer. The beneficial results often resulting from the use of this ma- terial, have been explained on the basis of its action as a stimu- lant. Boussingault showed that its application increased the amount of potassium taken up by the plant. This was accom- plished through a double decomposition of potassium silicates with liberation of potassium sulphate. Consequently its action was said to be indirect and it was classed as a stimulant for the reason that while it often produced increased plant growth, it furnished no necessary plant elements. While the above reaction may be entirely true, the explanation appears to be only partial. If the idea here presented, that sulphur may become a limiting element in crop production, is true, then the beneficial results • accruing from the use of gypsum may result in part from its sulphur content. In the 25 years of carefully recorded plot fertilization experi- ments by the Pennsylvania Experiment Station, 18 the evidence for the necessity of occasional sulphur applications is negative. While these experiments were not planned to answer the ques- tion of sulphur requirements, nevertheless they are the only experiments available, so far as the authors are aware, which shed any light at all on the question involved. On plots 9 and 17 the treatment consisted of dried blood, muriate of potash and dissolved bone black, which probably contained calcium sulphate sufficient to furnish from 30 to 40 pounds of sulphur trioxide per acre. On plots 12 and 35 the treatment was dried blood, muriate of potash and ground bone. Identical quantities of phosphorus pentoxide and potassium oxide were added in the 38 Rep. Pa. Expt. Sta., 1907-1908, p. 80. Sulphur Requirements of Farm Crops 19 two treatments, with 24 pounds of nitrogen in the first mixture and 30 pounds in the latter. Unless the muriate of potash and dried blood used contained sulphates sufficient 19 for all crop needs, then the only difference in the kinds of essential elements added was in the greater amount of sulphur contained in the first mixture ; yet the total annual yield in crops over a period of 25 years was slightly in favor cf the second treatment. Nevertheless, while an annual removal of 15 pounds of sulphur trioxide per acre by the crops grown in the above experiment, may not as a single factor be sufficient to reduce the productivity of the soil receiving suffi- cient nitrogen, phosphorus and potassium in 25 years, it appears probable that this could not continue indefinitely. Our own data on the partial depletion of sulphur in soils unmanured but cropped for 50 to 60 years is evidence in support of this view. Unfortunately none of the experimental plots at Rothamsted, England, have been deliberately planned with respect to the effect of fertilizers with and without applications of sulphur on long continued cropping. We are, however, through the courtesy of Director Hall, in possession of materials from that station the analysis of which may throw additional light on the question involved. Such data will be reported later. Careful experimentation and practical agriculture must de- cide in what form sulphur, when needed, should be added to the soil. Economy and safety are the factors involved. The prin- cipal sulphates normal to the soil are those of calcium, mag- nesium and potassium. No harm can result from a judicious use of any of these. Sulphur can be added to the soil either as land plaster; with acid-phosphate, in which it exists as cal- cium sulphate ; or as a sulphate of potassium or ammonium. All these materials are now offered by the trade. A ton of land plaster contains about 900 pounds of sulphur trioxide and a ton of acid-phosphate will carry from 200 to 300 pounds of sulphur trioxide, in addition to the phosphorus pentoxide it contains. A ton of high grade sulphate of potassium will contain about 900 pounds of sulphur tri oxide besides about 1,000 pounds of potas- sium oxide. A ton of ammonium sulphate will contain about is From some of our own analyses of dried blood and muriate of pot- ash the amounts of these materials used bi-annually in the Pennsyl- vania Experiments could have furnished about 14 pounds of sulphur’ trioxide per acre. 20 Wisconsin Experiment Station 1,000 pounds of sulphur trioxide besides its large amount of nitrogen. Of course, under systems of stock farming where the crops and purchased feeds are fed and the manure saved, the sulphur will find its way back to the land, but whether even then the- losses by drainage and in the- practical handling of the manure must not be met by additional applications of sulphates, are questions still to be determined. In systems of grain farming it appears, from the data here presented, that some form of sulphate must be used systematically in the fertilizer treat- ment of the soil for the purpose of maintaining therein a per- manent supply of the element sulphur. Conclusions. 1. The sulphur content of a number of our common farm products has been determined and in agreement with other in- vestigations the quantity is much larger than found by Wolff in the ash from such products. 2. The amount of sulphur trioxide removed by crops is con- siderable, being equal in the case of average crops of cereal grains and straws to about two-thirds of the phosphorus pent- oxide removed by these crops; the grasses of mixed meadow hay remove quite as much sulphur as phosphorus, while the legume hays may approach, and in the case of alfalfa, even exceed in this respect. Members of the Cruciferae, as the cabbage and turnip, are heavy sulphur-using crops and may remove two to three times as much sulphur trioxide as phosphorus pentoxide. An average acre crop of cabbage will remove about 100 pounds of sulphur trioxide. 3. Normal soils are relatively poor in total sulphur trioxide; a limited number of analyses showed a percentage content of from 0.033 to 0.140; most, of them contained less than 0.10 per cent. An acre foot will contain from 1,000 to 3,000 pounds of total sulphur trioxide. About the same quantity of phosphorus pentoxide will be found in an acre foot of normal soil. These results for sulphur trioxide are based on analyses made by the method of fusion with sodium peroxide. Determinations by extracting with hydrochloric acid or with nitric acid and bro- mine will not give the total sulphur content of soils. 4. Soils cropped for 50 to 60 years and either unmanured or Sulphur Requirements of Farm Crops 21 receiving but slight applications during that period have lost on the average 40 per cent of the sulphur trioxide originally present as determined by comparison with virgin soils. 5. Where farm manure has been applied in regular and fairly liberal quantities the sulphur content of the soil has been main- tained and even increased. 6. The total sulphur trioxide precipitated at Madison, Wis., with the rain amounted in the five months of June to October, 1910, inclusive, to 11.7 pounds per acre. The annual amount may tentatively be placed at from 15 to 20 pounds. 7. The losses of sulphur trioxide by drainage, based on the analysis of the drainage waters at Rothamsted, England, and on a yearly drainage of 10 inches, would amount to about 50 pounds per acre' yearly. 8. Even with much less loss by drainage it does not appear that the atmosphere can serve as a complete compensating fac- tor for losses of sulphur trioxide which soils sustain through both cropping and drainage. The partial depletion of the sul- phur of the soil by continued cropping without adequate fertil- ization is evidence in support of this view. 9. From the data here presented it appears that for perma- nent and increased production of farm crops such systems of fer- tilization must be inaugurated as will supply to the soil from time to time, in addition to the elements now T recognized as generally necessary, — namely, nitrogen and phosphorus, — a sufficient quantity of sulphur to meet the losses sustained by cropping and drainage. 10. Such sources of sulphur are farm manures ; the trade fer- tilizers, such as super-phosphate, ammonium sulphate and sul- phate of potassium; and the so-called soil stimulant, gypsum or calcium sulphate. In a problem of such large significance the authors realize the desirability of extreme caution and conservatism in presenting the views outlined. No attention, so far as we are aware, has been directed to this problem in America. It is hoped that the thesis here presented may be made the subject for further re- search by chemists and agronomists and the relative importance and necessity for sulphur in systems of fertilization finally established. f jfT Experiments on Spore Germination and Infection in Certain Species of Oomycetes 1 I. E. MELHUS INTRODUCTION Some two years ago special work was undertaken upon Cysto- pus and other Oomycetes aiming to learn more as to the methods of spore germination, zoospore formation, methods of infection, and as to the possible existence of so-called “physiological spe- cies” in the genus Cystopus. The parasite Cyst opus candidus was common on the garden crop of radish, Eaphanus sativus, and attempts were made at the outset to produce infection by transferring spores from this to radishes growing in the green- house. Although repeated trials were made in connection with these early studies, only meager and irregular infections resulted. This suggested that some variable factors of unknown nature were present in the greenhouse trials. Difficulty was also en- countered at the outset in securing uniform results in spore germination by the methods described by earlier workers. Thus it soon became evident that some more specialized methods were necessary in order to secure the germination of the conidia and host infection in abundance and with a satisfactory degree of certainty. We were thus led to attempt to determine the relations of * various conditions to spore germination and infection with this fungus, not only host relations, but also relations of the age 1 The author wishes to express to Dr. R. A. Harper and Dr. L. R. Jones his sincere appreciation for the kind criticism and keen interest shown during the progress of tbi$ work and preparation of the mamj. script, 26 Wisconsin Experiment Station. and maturity of the spores, moisture, food, chemical stimuli, light and temperature. Other species were subsequently tested. The results secured are of such definiteness and breadth of applica- bility as to justify their publication, although much remains to be done upon the problems as originally defined. Before discussing my own work and conclusions, a brief review will be made of the results of previous studies upon these and closely related mat- ters. Review of Earlier Work The germination of the asexual spores of Cystopus was first observed over a century ago by Prevost (1807 :29). He studied the species commonly parasitic on crucifiers and purslane respectively, then known as TJredo Candida and TJredo portulacae. His description is clear and interesting. He states that the spores germinated one or two hours after immersion, some- times within 40 or 45 minutes, owing in all probability to differ- ences in temperature, which, during the observation, fluctuated between 12° and 16° R. (equivalent to 15° and 20 d C.). In the process of germination the spores absorb water and become bottle shaped; soon a globule (zoospore) is seen on the outside and this is immediately followed by several others, sometimes as many as six more. He states that these globules instantly reunite into a mass which moves as a unit by rolling about in the water. The globules, as a rule, separate from one another in a very short time ; sometimes, however, two or even three globules may remain attached together, either immediately touching or as if joined by a string. Those globules which separate from one another, and they are by far the greater number, are sometimes a little angular and possibly a little hollowed or pushed in at one side. They swim about in the same way as when united in mass. Soon the movement of the globules ceases and they become fixed at the surface of the water or at the edge of the drop. He observed that the swarmspores developed germ tubes, regarding which however he gives little detail. Prevost likewise studied the Cystopus on salsify and Amaranthus and found the last two # forms much more difficult to germinate. Tulasne (1854:77) states that he germinated the spores of tiredo portulacae and TJredo Candida but was unable to get them to form swarmspores in the manner described by Prevost, find- ing ofily the germination by a tube. Experiments on Spore Germination. 27 Hoffman (1859 : 210) was also unable to confirm Prevost as to the formation of swarm spores in Cystopus. He describes the germination of the spores by tubes in the same manner as described by Tulasne. DeBary (I860: 236) studied the method of germination in Cystopus cubicus and Cystopus candidus and found the germina- tion was by zoospores, as described by Prevost, but that germina- tion might take place at any temperature between 5° and 25°, C. He emphasized for the first time that the spores are really sporangia producing from five to eight zoospores in Cysto- pus candidus and from eight to twelve in Cystopus cubicus. DeBary describes the changes in the sporangium very clearly. The spores absorb water and swell when sown in a drop of water. On one side an obtuse papilla is developed and vacuoles form in the granular protoplasm which disappear later. At this stage of development, fine lines of demarcation divide the protoplasm into five to eight polyhedral portions leaving at the center a small feebly colored vacuole. When the division of the content of the sporangium is complete, the papilla swells, opens and the zoo- spores are pushed to the outside one by one, showing no sign of movement. Once outside the sporangium they become lenticular in form and group themselves before the opening of the sporan- gium in a spherical mass. Yery soon the swarm spores begin to move, vibratile cilia appear and the globular mass is set oscillat- ing. The zoospores ultimately become free and swim away singly. The motile spores are' plano-convex or slightly concavo-convex having a small disk-like vacuole on one side. Attached near the vacuole are two cilia, a short one in front and a long one behind, both on the same side. In from one and a half to three hours after being placed in water the escape of the zoospores begins. They will develop either from sporangia freshly formed or from those which have been kept as- long as six weeks. DeBary (1863:14) found the conidia of Cystopus germ- inating on the leaves of the host plants. Zoospores were found in the drops of water on the leaves. Infection experiments with Cystopus candidus were made on various hosts. In the case of Lepidium sativum the zoospores readily entered the stomata of both leaves and cotyledons but produced infection only in the latter. Various species of Brassica showed the same tendency, though not to so marked an extent. 28 Wisconsin Experiment Station. Farlow (1875:319) studied the germination of the conidia of Phytophthora infestans and observed that sometimes the con- tents of the spore discharged in one mass, and from this mass zoospores are produced as before. He believes that the produc- tion of zoospores is favored by darkness, whereas germination by a germ tube takes place more frequently in the light. He states, however, that he has repeatedly sown spores in watch glasses and both methods of germination resulted. The germinating power of the spores was retained for several weeks, but they did not germinate after a winter’s exposure. Farlow (1876 :419) also describes the germination of the conidia of Plasmopara viticola. During the month of October when the disease is most prevalent, he found that in one and one-fourth hours zoospores were formed and began to make their way to the outside of the sporangium. The conidia might also germinate directly, i. e., by germ tubes. Darkness was more favorable for the germination of the conidia, whether directly or indirectly. He found the zoospores to swim about for fifteen to twenty min- utes, after which the motion gradually became slower until they finally came to rest. In another quarter of an hour an outgrowth appeared on one side which rapidly developed. Scribner (1886:10) states that temperature exercises a consid- erable influence over the germination of Plasmopara viticola from the grape. The most favorable temperature lay between 25° and 35° C. At a lower temperature, germination took place more slowly, but the temperature could be reduced to zero without destroying the vitality of the conidia. Under exceptional cir- cumstances, Scribner states, another form of germination might occur in which a conidium may push out a tube. This method, he reports as undoubtedly rare. The question of spore germination and physiological species in the genus Cystopus has quite recently been studied in consid- erable detail and with very interesting results by Eberhardt (1904:622). He began his work in June 1902 and continued it until the fall of 1903. Cystopus candidus on various crucifer- ous plants were used except in one experiment where he used Cystopus cubicus on Tragopogon pratensis. Eberhardt (1903:655) found, as we have, considerable diffi- culty in germinating the conidia of Cystopus. He tried the dif- ferent methods used by DeBary and Zalewski, but obtained only a low germination. To solve the question of spore germination, Experiments on Spore Germination. 29 Eberhardt turned his attention to methods of properly maturing me comma, inuring ail die montii of iviay ana xhe beginning of June Cyst opus cantiidus was available on Capsella, twelve experiments were inaae at as many different times of which tlie following may be taken as typical. (Jonidial pustules not opened, were collected from time to time. The contents of die pustules Were placed in a vial containing a small amount of rain water. This was called vial No. 1. In vial No. 2, containing ram water was placed the spores shaken from open pustules, in still another test a shoot was taken which bore many unopened pustules. It was wrapped in a moist cloth to be kept for furtner observations. In vial No. 1 the conidia remained in chains, the protoplasm became granular and later bacteria developed decomposing the conidia. Vial No. 1 may show a small number of conidia ger- minated, but the larger part disorganized. In vial No. 2, containing spores shaken from open pustules one third of the conidia formed zoospores while the remainder decomposed. ATien taking spores from pustules just opened, more than one half germinated. It is permissible to think that the former were relatively old and that they had lost their capacity of germination whereas the later conidia were taken from pustules just opened and gave a germination of more than one half of the spores. The third test, in which the shoot containing pustules was wrapped in a moist cloth, gives further evidence in support of the above interpretation. After the pustules had been kept moist for one day a microscopic examination of three of the pustules showed the following conditions on being placed in a drop of water. The young pustule was found to contain the zoosporangia in chains which gave no germination. The second showed a small fraction of the spores empty or bottle- necked, ready to germinate. The other conidia were decom- posing. The third presented all the conidia disassociated and about one third produced zoospores. The shoot was kept fur- ther until almost all of the pustules had opened liberating the spores that germinated in fairly large numbers. This same shoot was kept two weeks longer when it had wilted and dried up in places. The conidia that fell from it gave no germination. These experiments show us that the zoosporangia are the organs of immediate infection, requiring for germination to be collected at the time when the pustules open. Infection occurs 30 Wisconsin Experiment Station. when these spores fall upon the young host plants humid from rain or dew. In the numerous cultures made for infection purposes none were kept after the third day if germination had not already occured. It is well here to notice the variation in time required, for germination of conidia. A few notes from Eberhardt’s experiments will suffice. Eberhardt used “room temperature” in every case except in one.. test made on the twenty-ninth of March. The exact temperatures are given in only two cases, however. In the first it varied from 11° to 17° C. In the experiment referred to, on March 29, the temperature varied from 2° to 8° C. The latter condition was obtained by placing the culture on a win- dow sill where the influence of the out-door temperature had some effect. His germination experiments that are described were carried on from March 28 to August 9, 1902, on five different dates, all resulting in germination in from three to forty hours. There was also another difficulty greater than that of the selection of the conidia, it was the choice of the proper time to inoculate the host plant. The task of growing and caring for the young plants, the delicacy of the material of infection and the considerable space required for growing the crucifers has made it impossible to often repeat the same infections. In view of the fact that the following statements of the author are not clear we quote directly. “Un facteur important etait la coincidence qui doit tou jours exister entre la recolte du parasite et l’etat de germination propice de la plante a infecter. C’est pour ces raisons que nous ne pouvons poser en ce moment aucune regie basee sur l’experience, relative a 1 ’optimum de receptivite du parasite par le vegetal nourricier. Mais ce que nous pouvous certainement avancer, c’est qu’il ne suffit pas dans toutes les Crucifers d ’avoir des cotyledons bien etales. La ques- tion de 1 ’optimum de receptivite demande plusieurs annees de recherches. Comme nos infections tendaient plutot a prouver 1 ’unite d’espece, ce n’est que tres tard que nous nous sommes aperu, apres des insucces nonib reux, combien l’etat du jeune plant influe sur la reussite de l’experience. Ainsi nous avons vu que certains cotyledons, sortant de la game refusent 1 ’entree du parasite, tandis que lorsqu ’ils sont bien etales, ils sont sus- ceptibles d ’infections. II nous a semble que plusieurs de nos especes qui avaient ete infectees au moment ou les cotyledons Experiments on Spore Germination. 31 etaient fanes penvent recevoir Fendophyte par le jeune bourgeon foliaire deroulant ses feuilles. Mais nous ne pouvons encore rien affirmer a ce propos. Au reste, De Bary lui-meme indique que Heliophila crithmifolia est apte a etre infectee par les jeunes feuilles. ’ ’ Comparatively speaking, but very little work has been done on * the question of physiological species in the Oomycetes. Eber- hardt (1904:714) has investigated this problem in Cystopus can- didus occurring on various crucifers. Plants were inoculated by germinating the conidia and placing the liquid containing the zoospores on the lower side ’of the cotyledons of the plants to be infected or simply dipping the cotyledons to be infected in the water containing the zoospores. The seedlings were grown in large flower pots and small bunches were removed as needed for infection. His infection experiments were carried on both, out of doors and in an ordinary laboratory. The conidia of five different species were used as material for infecting other cruciferous plants. It is further interesting to note that Eberhardt made inocula- tion experiments with the oospores of Cystopus candidus from Lepidium sativum. In order to infect the plants of Lepidium and Capsella with the oospores from Lepidium sativum, the parts of the host containing the oospores were placed in small bags and hung out of doors in the open air during the winter months. In March and April the oospore material was distributed over the surface of pots where it might decay and liberate the oospores. Two of these pots were then seeded with Lepidium sativum and Capsella Bursa-pastoris and the young seedlings became infected. It should be noted in this connection that Eberhardt used no con- trols in this series of experiments. His results may be summarized as follows: With the conidia of Cystopus candidus from Capsella Bursa-pastoris he infected Capsella Bursa-pastoris. Lepidium sativum , Iberis amara, and Arabis alpina. Conidia from Capsella Heegeri infected Capsella Heegeri, Capsella Bursa-pastoris and Lepidum sativum. Conidia from Lepidium sativum infected Lepidium sativum and Capsella Bursa-pastoris. Spores from Brassica rapa infected Brassica rapa, Brassica oleracea (varieties: botrytis, capitata, and congylodes), Brassica nigra, Sinapis arvensis , and Diplotaxis tenuifolia. The widest range of infec- 32 Wisconsin Experiment Station. tion was obtained with the conidia from Arabis alpina which infected Arabis alpina , Arabis hirsuta, Arabis II alter i, Arabis turritis, Lepidium sativum, Iberis amara, Cardamine pra- tensis, Cardamine amara, Capsella Bursa-past ons, and Sene- biera eoronopus. With the oospores from Lepidium sativum he infected Lepidium sativum and Capsella Bursa-past oris. It was impossible to infect any of the Cruciferae with conidia from Tragopogon pratensis, but quite easy to infect Scorzonera Jus panica. From these data Eberhardt believes that there are no biological species in the species, Cyst opus candidus. It should be noted, however, that the above conclusions are based upon only one trial in some cases with each species or variety of plant. Eberhardt states in this connection that the laborious task of growing, inoculating and recording results, together with the fact that the ground used for growing the plants could not be had during the next year, made it impossible to duplicate any of the series of experiments performed. The fact that Eberhardt ? s work was incomplete and not fully convincing, and that it is especially important that each group of parasitic fungi be fully understood as to the existence of physiological species, led me to study this problem. EXPERIMENTAL STUDIES IN SPORE GERMINATION Method As has been previously stated, the major portion of the experi- ments recorded in this paper were carried on with the common white rust, Cyst opus candidus, as it occurs on various garden plants. The culture work was all carried on either in the green- house or in an ordinary laboratory. Both distilled and tap water were used to germinate spores. The tap water in this case is un- filtered water drawn directly from Lake Mendota. The conidia were gathered from infected plants growing out of doors until frost, when they were taken from infected plants in the green house and were sown the same day as gathered. The spores were sown in a drop of water placed near the center of a clean slide. It was always difficult to make the conidia sink, but by stirring the drop with a scalpel, a large per cent could be finally made to settle Experiments on Spore Germination. 33 to the bottom. The aim was always to add only a moderate num- ber of spores to each drop, since too many spores make the water opaque and difficult to examine. Care was taken to obtain fresh spores in every case, which were stirred into the drop of tap water. Sometimes small pieces of leaves bearing pustules were dropped into water on a slide. The first experiments were made at room temperature, but the irregularity of the results and the small per cent of the total which germinated even in the most favorable cases led to experi- ments with low temperatures. With this idea in mind, the slides sown with spores were placed in an ice box of the usual construc- tion. A Richard’s self -registering thermometer was also placed in the ice box so that a complete record could be had of the temperatures to which the spores were exposed. In the case of the controls at room temperature, the stands holding the slides sown with spores were placed on a wet earthen plate and a small bell glass placed over them to prevent rapid evaporation. The effect of darkness was also tested by placing similar cultures in th*e dark room. The temperature of the dark room was noted when the experiment was started and stopped and the average taken. The cultures at room temperature, not in the dark room, were kept in a greenhouse where another self-registering ther- mometer recorded the temperature. Here also the temperature was noted when the experiment was started and stopped. For convenience in tabulating, the average of the two extremes was taken as the prevailing temperature. Cultures in the ordinary Yan Tieghem cell were also used. The cell was partially filled with tap-water and a hanging drop made containing the spores. Vaseline was used to prever' L evaporation and to hold the cover glass in place. The cells were laid on a stand as described above for the slides. Watch crystals were used when it was desired to secure large quantities of spores in differ- ent stages of germination. Results of Germination Experiments My earliest experiments in germination of the spores had the double aim of providing material for the cytological study of the processes of nuclear and zoospore formation in the germination of the conidia and of obtaining a reliable method for germination of the conidia for infection experiments in determining whether 34 Wisconsin Experiment Station. physiological species are to be found in the group. As noted earlier, it was found difficult to obtain germination and a long series of one hundred experiments was made, testing the general questions as to the effect of age and maturity of the spores; the weather conditions under which they were collected ; possible in- fluence of rain water, dew, tap water and distilled water; and the effect of gathering the spores at different times of the day. Attention was turned to the possible effect of artificial media and a number of attempts were made to germinate the spores of Cystopus in artificial media, these experiments being made in a greenhouse the temperature of which varied from about 33° C. in the day time to 22° C. at night. First ordinary nutrient agar was tried (3 gr. meat extract, 10 gr. agar, 3 gr. salt, 1000 cc. water). The conidia were sown on this agar in petri dishes and kept under observation for twentyfour hours, but no germination resulted in any of the ten experiments tried. Beef bouillon (3 gr. meat extract, 3 gr. salt, 1000 cc. water) was tried in two experi- ments, but gave no germination. Following this, some special media were tested. Ten experiments were tried with lima bean agar, as prepared by Clinton (1908:904) for growing Phytoph- thora, and eight experiments with his pumpkin agar ; none of the cultures showed any signs of germination at the end of twenty- four hours. Various other forms of artificial media were tested, including mustard leaf decoction, four trials; corn meal agar, five trials ; a two per cent sugar solution, six trials. The cultures were kept under observation in each case for twenty-four hours. In no case did germination result either by germ tubes or by zoospores. The above experiments were made during July and August, 1909. The conidia used were from Cystopus candidus, C. bliti, C. cubicus and C. portulacae. They were collected at various times of the day ranging from seven o’clock in the morn- ing until seven in the evening, the great majority being gathered about eight o ’clock in the jnorning. The infected leaves were cut off from the host plants and carried to the laboratory, where the material was used immediately. A number of times the infected leaves were immediately placed in a damp chamber after they were removed from the host plant. A large number of tests were made with both young and old conidia, before and after the epi- dermis of the pustule had ruptured. These conidia were, of course, also tested in water on a slide or in a hanging drop in a so-called Van Tieghem cell. The slides Experiments on Spore Germination. 35 were placed on a small stand on a wet earthen plate under a bell jar in the greenhouse. In all, fifty-four trials were thus made in water and in no case was germination observed. Each experiment lasted for twenty-four hours and several observations were made during that time. The effect of chilling was next tried, both on various nutrient media and in tap water. No germination was obtained in this way when using nutrient media but with water, germination was secured. Thus, when four slides were prepared as before and placed on a metal stand in an ice box, the conidia had germinated by the production of zoospores in the course of 1% hours. A large number of cultures were subsequently made by this method to secure material for cytological study and with controls kept at room temperature to show the exact value of the chilling in influ- encing germination. The results are strikingly uniform and in strong contrast with those obtained before without chilling. Summary of Table I The first experiment with chilling in germinating the conidia of Cystopus was made on August 10, 1909. From that date up until April 9, 1910, experiments were carried on as indicated in the table. In all, 197 experiments were made, giving germina- tion in 147 cases. From the number of those which showed no germination should undoubtedly be subtracted the results ob- tained on August 12 and 15 with conidia of Cystopus portulacae. The plant from which the conidia were obtained had been dug up, potted and taken to the green house July 25. It died in the course of ten days and the vitality of the conidia may have been reduced by its condition on August 2. If these experiments are omitted the number of failures to germinate is reduced by ten and we should have about 78 per cent of the experiments showing germi- nation and about 22 per cent negative. As shown in the table the conidia of four different species of Cystopus were used in these experiments, including Cystopus candidus 78 trials, Cystopus cubicus 60, Cystopus bliti 45, and Cystopus portulacae 14. Until October 19 the conidia were taken from live plants growing out of doors and used immediately. The terms “old” and “young” as used in the table indicate only the approximate age of the conidia. The conidia "were called young, although the epidermis of the pustules had ruptured, as long as 36 Wisconsin Experiment Station, Table I. — Effect of Lowering Temperature on the Germination of the Conidia of Certain Species of Cystopus. Date. No. of cultures. Species tested. Age of spores. Period friger Hrs. OF RE- ATION. Temp. l Re- 1 suit. Out-door tempera- ture. Minimum. Aug'. 10. . 4 Cystopus bliti Young . . 1.5 21 + 17 Aug. 11 .. 4 C. candidus Young . . 2 21 + 15 Aug-. 12 : 2 C. candidus Young . . 1 21 + 19 Aug. 12 . . 2 C. bliti Young . . 1 21 + 19 Aug. 12.. 2 C. candidus Young . . 1.5 21 + 19 Aug. 12.. 2 C. bliti Young . . 2 21 + 19 Aug. 12 2 C. bliti Old 2 21 + 19 Aug. 12. . 4 C. portulacae Young . . 2.5 21 19 Aug. 13. . 4 C. bliti Young . . 2.66 20 + 18 6 Old 3.5 14 2 i Aug. 16 2 + Young . . 1.5 18 Aug, 16.: 3 C. candidus Young . . 2 is f 18 Aug. 18.. 8 + C. candidus Young . . 2 18 + 17 Aug. 20 3+ Young . . 2.5 + 19 Aug. 24. . 4 C. candidus Young . . 11.75 18 12 Aug. 23.. 4+ C. candidus Old 2 17 17 Aug. 23.. 4 + C, candidus Young . . 24 17 + 17 Aug. 23.. 5 C. bliti Young . . 2.5 17 + 17 Aug. 23.. 4 C. bliti Young . . 4.5 17 + 17 Sept. 25. . 2 C. candidus Young . . 24 7 6 Sept, 27.. 3 C. candidus Young . . 7 6 + 6 Sept. 27.. 1 * C. candidus Young . . 7 6 + 6 Sept. 27.. 1 * C. candidus Youug . . 24 6 + 6 Oct. 1 ... 4* C. cubicus Young . . 36.5 11 + 6 Oct. 1 ... 3* C. cubicus Young . . 28 10 6 Oct, 28... 2 C. candidus Young . . 4.5 8 8 Oct, 28... 1 * C. candidus Young . . 4.5 8 + 8 Sept. 23.. 2 * C. cubicus Young . . 5.75 8 + 8 Sept. 23.. 1 + C. bliti Young . . 4.5 8 + 8 Oct. 4... 1 + C. bliti Young . . 3 8 + 6 Oct, 4... 1 * C. bliti Young . . 3 8 + 6 Oct. 4... ,3* C. cubicus Young . . 4 9 + 6 Oct. 9. . . 2 * C. candidus Young . . 3.5 10 + 13 Oct. 9.. Z* C. candidus Young . . 16 10 + 13 Oct. 14... 2 + C. bliti Young . . 15 13 + — 2 Oct. 14... 3* C. bliti Young . . 2.66 12 + 2 Oct, 28. . . 2 C. candidus Young . . 7.5 12 + — 4 Oct. 8 ... 2 + C. bliti Young . . 6 11 11 Oct, 19... 4 C. cubicus Young . . 6.25 10 __ 1 Oct. 10. . . 3 C enhiens .... Young . . 3 10 + ( >,‘t. 10 . . . 3 C bliti Young . . 3 10 .T an . 24. 4 C. candidus Young . . 47 11 Jan. 25. 3 C. candidus. . . . Young . . 10 10 4 Jan. 26. 4 C. c.a.ndidns Young . . 22 11 .1 an. 26. 3 C rand id ns Young . . 12 11 -+- Aug. 25" 1 C. bliti Young . . 5.5 18 4 20 Aug. 28. . 4 C. candidus Young . . 6 14 4 16 Aug. 30.. 4 C. portulacae Young . . 3 15 4 11 ' Aug. 31.. 3+ C. bliti Young . . 4.25 11 4 8 Aug. 31.. 1 C. cubicus Young . . 18.5 11 4 8 Sept, 2.. 1 + C. cubicus Young . . 2.12 10 — 7 Sept. 4.. 4 C. cubicus Young . . 10.75 10 4- 12 Sept. 4.. o C. bliti Old 2.5 -10 + 12 Sept. 4.. 1 C. candidus Young . . 2.5 10 12 Sept. 6 .. 1 C. cubicus . Old 5 25 8 4 9 Sept. 6 .. 1 C. cubicus Old.. .. 5.25 8 9 Sept. 7 . . 2 C. cubicus Old 6.5 9 4 9 Sept. 10 .. 4+ C. cubicus Young . . 4.25 11 + 13 Sept. 10 .. 1 O. candidus Young . . 7 11 4 13 Sept. 13. . 3 C. cubicus Young . . 2.33 8 4 18 Sept. 14. . 4 O. cubicus Young . . 2.75 8 4 It Sept, 14. . 4 C. cubicus Young . . 3 9 4 1 + Sept. 16. . 1 * O. Cubicus Young . . 4.87 10 4 9 Sept. 20. . 1 * C. cubicus Young . . 5.5 4 15 Sept. 21 . . 4+ C. candidus ! Young . . 9.25 12 i 4 16 * Experiments carried on in watch crystals. t Pieces of leaves, on which there vere pastilles, were laid on the slide. Experiments on Spore Germination. 37 Table I. Continued. — Effect of Lowering Temperature on the Germination of the Conidia of Certain Species of Cystopus. Date. No. of cultures. Species tested. Age of spores. Period frigei Hrs. > OF RE- LATION. Temp. Re- sult. Out-door tempera- ture. Minimum. Sept. 21.. 1 C. candidus Young . . 23 14 + 16 Sept. 21 . . 1* C. candidus Young 23 14 + 16 Sept. 21.. 1 C. cubicus Young .. 4.87 12 + 16 Sept. 21 . . 3 C. blip Young . . 29 1.4 + 16 Sept. 22.. 4* C. cubicus Young . . 6 11 + n Sept. 24. . 3* C. cubicus Young . . 3.75 8 + 7 Sept. 25. . 4 C. cubicus Old 23 y Q Sept. 25. . 2* C. cubicus ^ Young . . 3.25 7 + 6 ^Experiments carried on in watch crystals. a considerable number of spores remained in the pustules; and old after the pustules were nearly empty. The conidia were either taken out of the pustule and placed in a drop of water or small pieces of leaves with pustules were laid in a drop of water. If the pustules on the pieces of leaves were not already open when they were laid in the drop of water the epidermis was broken with a needle. An ordinary ice box was used and in it was kept a self regis- tering thermometer. By referring to the table it can be seen that the temperatures from August 10 to August 25 ranged from 15° to 21° C. The temperature was usually above 18° C. The ice box was kept in a rather warm room adjoining the green house, and it was also used for other purposes so that the doors were opened and closed quite often. The temperature curve was very irregu- lar. There were fluctuations of 10° C. in five hours in some cases although usually it was less. Since the fluctuations were too numerous to explain in connection with each test, the average of me maximum and minimum temperature Has Deen taRen as the prevailing temperature and is that recorded in the table. This, in some experiments and especially in those before August 13, does not give the correct temperature conditions. In the tests after August 13, there was much less variation and the average of the two extremes is much nearer the prevailing temperature condition. The exact temperatures during two tests are given in detail to show more clearly the existing conditions. For example, the temperature varied as follows during the experiment on August 25 : The test was started at 9 :15 a. m. with a temperature of 20° C. The temperature remained constant until 10 o’clock. Thirty minutes later the temperature was 21° C. At 11 o’clock 38 Wisconsin Experiment Station. it was 19° C. where it remained until 12 o’clock noon. At 1 o’clock it was down to 14° C.; at 2 o’clock it was 13° C. and at 2 :45 p. m. it again rose to 14° C. During 5 1-2 hours, the tem- perature varied 8° C. After August 25, the temperature of the ice box was much more constant and often fluctuated only one degree during the time of an experiment. The period of refrig- eration on September 2 started at 9 :30 a. m. with a temperature of 10%° C. and stopped at 2 :35 p. m. with a temperature of 11° C. During the five hours of refrigeration, the temperature only varied y 2 degree. It is to be noted also that the temperature grew gradually lower until October 9. This was due to the fact that before this time, no artificial heat of any consequence was used in the green house ; while after this date it was heated. For comparison with the temperature in the ice box, the minimum out door temperature is given for each day on which an experiment was made, as pub- lished by the local weather bureau of Madison in their monthly meteorological summary. It will be seen that the temperatures in the ice box varied two degrees or less from the minima out doors in 72 per cent of the tests until about October 14. This suggests that germination can readily take place at temperatures equal to or varying two degrees or less from the minima for the outdoors. The length of time required for germination varied from one to thirtysix hours. The one hour period required for germination was on August 12 and the thirty-six hour period was on October 10. All of the experiments that required an unusual length of time for germination were examined from one to six times before the final observation was made. Water was added to replace the amount that evaporated. It should be said, however, that the usual period required for germination in the majority of the cases was less than six hours. From August 10 to 31 the average length of the period of refrigeration necessary to produce germi- nation was about 3y 2 hours. The longest period was 18 y 2 and the shortest period, one hour. During the month of September experiments were made on twentyone different days, one more than in August, and the average length of the period of refrigeration was about 7y 2 hours; here the longest period was twentynine hours and the shortest period two hours and five minutes. In October, experi- ments were made on nine different days. The average of the periods of refrigeration used where germination resulted was nine Experiments on Spore Germination. 39 hours, the longest period being S6y 2 hours, and the shortest two hours and forty minutes. From these facts it is strongly sug- gested that the period of refrigeration is longer in the fall than in the summer as has already been pointed out by Zalewski (1883:215). No germination experiments were carried on from October 28 to January 24, 1910. However, on January 24, 25, 26, 1910, fourteen trials were made in which six germinations occurred. The time required was ten hours in the trials on Janu- ary 24 and twelve houTs in the three successive tests on the follow- ing days. Although only a small number of tests were made dur- ing the month of January, it was quite evident that the conidia responded differently at this time than in the summer. In the tests made in January the zoospores lost their motility in less than one hour and developed long germ tubes. In none of the tests made before that time had germ tubes been seen. The different behavior of the conidia in the late fall and winter as compared with spring and summer, are attributed to the loss of vitality of the host and fungus or to the improper maturing of the spores. Table II. — Effect of Lowering Temperature in Germination of Conidia of Certain Species of Cystopus, With Controls at Room Temperature. Date. Species tested. Experiment at Low Tem- perature. 1 Controls at Room Tem- perature. No. cul- tures. Period of re- frigeration. Re- sults. No. cul- tures. Time con- tinued. Re- sults. Hrs. Temp. Hrs. Temp. Aug 1 . 11. . Cystopus bliti. . 4 1.5 21 + 4 1.5 25 Aug. 18. . C. bliti 8 2 18 + 8 2 ’ 27 Au g. 20.. C. bliti 3 2.5 19 + 3 2.5 27 Aug. 21.. C. bliti 4 10 18 4 10 25.5 Aug. 23. . C. bliti 9 7.3 17 + ' 9 7.3 27 Aug. 16. . C. candidus 5 1.6 18 + 5 2 28.5 1 A Sep. 15.. C. bliti 2 5.5 10 + 2 5.5 22 Sep. 25.. C. candidus 2 24 8 2 24 22 Oct. .9.. C. candidus 2* 5.5 11 + 2 * 23.75 22 Oct. 9.. C. candidus 2 28 10 + 2 28.5 21 1 = 1 | p* C. bliti 4 2.6 20 + 4 14 27 — *Conidia placed on watch crystals, instead of slides. Summary of Table II As noted, no control experiments were kept in connection with the trials reported in Table I. The conidia were germinated as material for cytological study which will be reported on later. A second set of similar experiments (Table II) with controls, was 40 Wisconsin Experiment Station. carried on to demonstrate beyond question that chilling is neces sary for abundant germination. In this series of forty five cul- tures subjected to low temperatures, six failed to germinate. That is, about 85 per cent of the tests gave germination. In the controls, no germination was observed. Thq temperature during the period of refrigeration was very high for ice-box temperature in most of the experiments, due to the conditions explained in the summary of Table I. In all of these experiments the conidia were placed in tap water on the slides. In order to prevent too rapid evaporation in the trials at room temperature, the slides were laid on a metal stand placed on a wet plate under a small *bell. jar. The final observa- tions were made in both sets of experiments at the same time and the results recorded. The average room temperature at which the experiments were made varied from 21 to 28.° C., as can be seen in Table II ; while the temperature during the period of refrig- eration varied from 8° to 21° C. Otherwise the conditions were the same in both sets of experiments. These results show that a slight lowering of the temperature stimulates the conidia of Cystopus to germinate with the production of zoospores. Table 111.— Effect of Light on the Germination of the Conidia of Certain Species of Cystopus. Controls Chilled. Culture at High Temper- atures. Period of Period in Period No. refrig’ra’n. dark room. in light. Date. of cul- Species tested. c/5 • c/5 Ji c/5 a C/3 tures |. 3 s c n 2 I C/3 3 S 9 C/3 1 5 H (S & H M K H Aug. 11. 4 Cv r stopus Candidas : 2 21 -f , 8 75 26 Aug. 13. . 4 c‘. bliti 2.66 20 + U. 27 14. 27 — Aug. 20. . 3* C. bliti 2.5 18 4- 1.5 28 — 2.5 27 — Aug. 23. . 4* C. bliti 9. 17 + 4. 27 — 7.33 27 — Sept. 15.. 2 C. bliti 5.5 10 + 5.5 27 — 5.5 22 — Sept. 25.. 2 C. candidus 24. 8 24. 28 — 24. 22 — Oct. 9. . . . 24 C. candidus 5.5 11 + 6. 27.5 — 23.75 22 * Pieces of leaves on which there were pustules, were laid od the slide, •('Experiments carried on in watch crystals. Summary of Table III The possible effect of light on the germination of the conidia of Cystopus was also tested. Controls were kept to ascertain the viability of the conidia and were chilled as described above. Twenty-one cultures were exposed to light and seventeen were Experiments on Spore (termination. 41 kept in darkness, in each case without chilling. All failed to germinate. The controls all germinated except two. These results are shown in Table III. All the experiments were carried on simultaneously and all the conditions were the same except that no bell jars were used to cover the controls while bell jars were used in the experiments in the light and in the dark. The conidia were taken from freshly matured pustules and placed in a drop of water on slides that were well cleaned. One series of cultures was kept in the diffused light of an ordinary laboratory where no direct sunlight fell upon them; the other series was kept in a dark room. The high temperatures, from August 11 to 25, have been explained in connection with Table I. The question might naturally be raised as to whether germination of the conidia in the ice box were not due to the dark moisture saturated atmosphere of the ice box rather than to the low tem- perature. These experiments answer this question. The series of seventeen cultures kept in the dark room were in a saturated atmosphere the same as the controls in the ice box. The only dif- ference was in the temperature which varied from 26° to 28° C. in the dark room and from 8° to 21° C. in the ice box. The ice box cultures all germinated except two ; while none of the seven- teen cultures in the dark room germinated. From these results it is clear that light is not a determining factor in germination. It is also clear that a saturated atmosphere at high temperatures will not cause the germination of the conidia of Cystopus. Table I Y.— Effect of Using Still Lower Temperatures in Germina- tion of Conidia of Certain Species of Cystopus. Date. No. of cul- tures Species tested. Age of spores . Period of refriger- ation. Re- sults. Out-door tempera- ture. minimum. Hours. Temp. Sept. 9.. 1 Cystopus bliti Young 1.16 § 15 Sept. 9.. 2 C. cubicus Young 8 § + 15 Sept. 9.. 1 C. candidus Young 8 § + 15 Sept. 13.. 8* C. cubicus Young,. . . 2.5 —1 — 18 Sept. 8.. 2 C. cubicus Young.. . . 4.5 § -+- 12 Sept. 4.. 1+ C. bliti Young 19.5 § + 14 Sept. 27.. 1 + C. cubicus Young... . 27 § — 6 Sept. 27.. It C. cubicus Young... . 27 § + 6 Oct. 4.. 1 C. cubicus Old 10.75 § 4- 6 Oct. 4.. 1 C. bliti Young 10.75 § + 6 Oct. 12.. 3 C. bliti* Young.,. . 3 10 — -4 Oct. 12.. 3 C. cubicus? Young 3 10 + —4 Oct 16 2 O e.iihimist Old 3 12 + 0 Oct. 16.. 4 C. bliti? Old 3 12 + 0 *This experiment was made with a Van Tieghem cell and hanging drop. tPiec^s of leaves with pustules in water on watch crystal. ^Conidia taken from frozen leaves. §The slides in these experiments were laid on a block of ice in the ice box. 42 Wisconsin Experiment Station. Summary of Table IV ihe effects of lower temperatures than those ordinarily obtained in the ice box were also tried. The results of these experiments are shown in Table IV. The slides were laid on blocks of ice, except in the case of the experiments on September 13 and October 12 and 16, which are further described below. Twelve cultures were made. In nine, the spores germinated while in three, there was no germination. These results are somewhat at variance with DeBary, who found the minimum to be 5° C. There can be no doubt from the above results that the minimum for germination is very near 0° C. To test the effect of still lower temperatures, three Van salt giving a temperature of 1° C. At the end of thirty minutes no germination had occurred. These slides were allowed to remain in the laboratory at 28° C. for ten hours after being re- moved from the freezing mixture, but no germination resulted. This indicates that a change from high to low and then back to high does not lead to germination. The effect of frost on the eonidia of Cystopus outdoors was also tested. The eonidia were collected October 12 and 16 from frozen leaves which had been allowed to thaw out in the laboratory. Twelve cultures were kept at 10° to 12° C. for three hours. Nine of the twelve cultures germinated and three did not, indicating that the eonidia were not killed by a frost. Summary of Table V In view of the fact that germination was obtained in the ice box at temperatures above 20° C. in some cases, it was thought advisable to make a further study of the relation of temperature to spore germination at room temperatures. During the latter part of March and first part of April, 1910, further experiments were made to determine whether the eonidia of Cystopus would germinate at room temperatures. As previously described, the experiments with cultures at green house temperatures in the summer of 1909 had given oply a low percentage of germination. In fact germination was observed in only one or two cases. In the new series of experiments, seventythree cultures were made from March 27 to April 8, at temperatures varying from 17%° to 25° C. Controls were kept at ice box temperatures in cases Table V — Further Experiments on Germination of Conidia of Cystopus at Room Temperature From March 27 to April 8. Experiments on Spore Germination. 43 Result. * * -H-+ 1 1 1 1 1 +++++++++++ 1 +++*!+* * * *■ 1 l++x l rv c3 : : : : ^ : : : : : : >, : ; ^ ^ ’. : - . >s ; : ; ; ^ ; ; ; w II! :*o ; : : : : i'o-' i :t 3 : i i'S : : . : . .rQ • • • CM CM CM CO irt lO co CO CO CO* CM CM CM CM UO HO m ^ CM CM - 5 & ^ ^ ooaZj.ZZ x £ x x x ix/Zxx x ^xxx x x x xxl?Z x x £ x : o-S r a'cs 0^3 lO iO lO UO ^^COCOC5CMCM05CMCMCMCMCMCMi-Ii It-itHi-iCMCMCMCMCMCMi IH CO CO CO CO l>- 1>» t— l>- irs 10 10 >0 10 10 a) ^ co ha 55 &j oiciO!MeaocofO'^'^'^-^vftif5ioif5imraimf5coa5cococceciftift>i5ir5 si»*«!»iiiima!mi/iiB ODSPS33333S ^ r t3 r cn3T3rc r O r o r b r c r c; r b r o , c! r w , a r a r o r c' E^SSScdSScoSeScScdcOcdcScOcdcd CCOOOOOOOOOOOOOOOOOO •yjtflcfitwt/itn'yjasMi/itnt/lajMMcfitncntKQD 33:3:3:33322322233135:332 ^T3232! r 3 r O r 3 r 3’25 r 3 r 3 r O r 3 r 3 r O , 3;3 r 3;3 r 3 ,FH *” "| jfi ,rH ,rH 'S 'S S S 'S 'S ^3 S S S £ scs C'O ccaaccfl cdcdcdcOcdcdcdcdcdcdcdcO OOOOOOOOOOOO OOOOOOOQOOOOOOOOOOOOOOOOOOOOOOOO i«(Mt)(NN(NNNMNNWO!NWNNHHHHNWNN' NNNNOOOOOOOOrtrtHHNN«NNWNN^' v50©rc<3im®oooooo.— '—1 Hours. « ^ — S8g Source of conidia. Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Ne Plus Ultra Variety of radish. ' White Icycle 1 Triumph Triumph China Rose Winter Mixed NePlus Ultra NePlus Ultra NePlus Ultra | NePlus Ultra Olds’ Golden Globe Olds’ Twenty Day Mixed NePlus Ultra NePlus Ultra NePlus Ultra NePlus Ultra NePlus Ultra NePlus Ultra Mixed Mixed First and Best NePlus Ultra NePlus Ultra Date. Oct. 29... Dec. 21... Jan. 3 Jan. 3 Jan. 11.. . Nov. 8... Nov. 8 ..., Dec. 3..., Dec. 8 Jan. 10... . Dec. 13.... Jan. 11.... Dec. 3 Dec. 8 Feb. 22... Feb. 22... Feb. 22... Feb. 22... Jan. 12.. . . Jan. 12 Jan. 12., . . Dec. 28.... Dec. 28... . No. of exp. Experiments on Spore Germination. 55 earthen plate under a bell jar and placed in the ice box for a period varying from six to twentyfour hours. When the stock cultures were removed from the ice box they were placed on a bench in the green house and in from seven to twelve days the fungus made its appearance. The controls used in each experi- ment were treated in the same manner as the stock cultures except that they were not chilled but placed directly on a bench in the green house. The results were photographed when the pustules were fully developed and about to burst. The extent of infection secured by the chilling method can best be seen by referring to the plates. Plates I and II show two stock cultures of radish seedlings that were inoculated when the seedlings were ten days old. In seven days, the cultures began to show infection. Four days later, or eleven days after inocu- lation, the conidial pustules were fully developed and the results photographed to show the difference in the extent of infection between cultures chilled and not chilled. Pustules developed on both the upper and lower side of the cotyledons but only on the lower side of the first true leaves, which can be seen in the plates. November 8, two more stock cultures were inoculated and became infected November 15. The results were photo- graphed November 17, two days later (Plates III and IV). Here, 'again, it will be noted that chilling produces the more abundant infection. These two stock cultures were kept as were many others as a source of conidia for future inoculation. November 29, or twentyone days after inoculation, the stock cultures shown in Plates III and IV were photographed again to show the further development of the fungus as well as the effects on the host plants (Plates V and VI). The radish seedlings in the chilled culture were being killed rapidly by- the fungus while in the control (Plate VI) the plants are healthy with no marked increase in the amount of white pustules. It was very evident in this experiment as w r ell as in all the others that more abundant infection had occurred in the cultures that were chilled. The same striking results were secured when the radish seedlings were grown in small pots (Plate VII). In these experiments, radish seedlings were grown in three-inch pots and treated as described above. The effect of chilling can be readily seen. The results shown in Plate VII suggest that an extensive infec- tion is not dependent upon the crowded condition of seedlings in the stock cultures, but upon chilling. Not only do the coty- 56 Wisconsin Experiment Station. ledons become heavily infected by the chilling' method, but also the leaves. The curled and hypertrophied leaves can be seen in Plate VIII. The leaves are as readily infected as the cotyledons of the radish. In Lepidium sativum only the cotyledons can become infected according to DeBary (1863; 27). Twenty-four tests with a control for each trial were made, be- tween October 29, 1909, and February 22, 1910. Experiments have been continued since the last named date with the same striking results. Plates I and X show stock cultures made from October 29, 1909, to June 6, 1910, and photographed when the fungus was well developed. In every case in which the stock cultures were chilled, a heavy infection resulted; while the con- trols, except in one case, showed but little development of the white rust. The control in experiment No. 4 on January 11, showed an exceptional development of infection. It was as heavily infected as the stock culture that was chilled which again shows that the conditions favorable for infection do occasionally occur in the green house, as has already been pointed out. It might naturally be asked whether chilling has the same effect in the spring and summer as in the fall and winter. Experi- ments have been carried on continuously since February 22 (although the results are not tabulated) until June 6,, 1910, with the same marked results as were obtained during the fall and winter of 1909. The above data lead me to conclude that chilling strongly favors the infection of radish plants with Cys- topus candidus. THE RELATIVE SUSCEPTIBILITY OF COTYLEDONS AND LEAVES TO CYSTOPUS CANDIDUS In the preceding experiments it was quite impossible to deter- mine whether any marked degree of difference of susceptibility existed between cotyledons and leaves of the radish. DeBary found that it was only the cotyledons that showed any marked degree of susceptibility. In order to test this point for the coty- ledons and leaves of radish, shepherd’s purse, white mustard, and garden cress, the following experiments were planned. All the plants utilized in this series of experiments were grown in the green house and were in all cases free from infection at the putset. Twelve three-inch pot§ were seeded with radish on Experiments on Spore Germination. 57 October 15, each pot containing only two plants. On November 11 the plants had lost their cotyledons and the first leaves were two to three inches long. At this time the twentyfour plants were inoculated and chilled and on November 23, twelve of the plants showed infection. On the following day twenty of the twentyfour plants inoculated were infected. There was not the slightest evidence that the infection was abnormal as the pustules were abundant on each leaf and the leaves were becoming badly distorted. On September 15, five plants of Capsella that had grown from seed collected outdoors and planted in the green house were in blossom. At this time the plants were inoculated with conidia from Capsella and September 26, three of the plants were infected on both stem and leaves. On the following day the leaves and young fruits of the two remaining plants also became infected. A stock culture of at least fifty white mustard plants that were planted September 1 in the green house were inoculated September 23 when the plants were six inches tall. They were healthy and the cotyledons had fallen off. On the last named date, the culture was inoculated with spores from the white mustard and chilled. On October 12, at least 75 per cent of the Laves on the plants were infected. I have never collected Cystopus on garden cress so it was im- possible to test this host with the Cystopus growing on it out- doors ; but I have inoculated garden cress with spores from Cap- sella. Three three-inch pots of garden cress containing eighteen plants that were four inches high and without cotyledons were inoculated October 15. Infection resulted October 27. Twelve of the plants became infected. From the above results it is clear that the leaves as well as the cotyledons of radish, shep- herd’s purse, white mustard, and garden cress are readily in- fected with Cystopus candidus. SUSCEPTIBILITY OF DIFFERENT VARIETIES AND SPECIES OF RAPHANUS TO CYSTOPUS CANDIDUS One object in developing a method of germinating the conidia of Cystopus with certainty and in abundance was to provide means for attacking the question as to the existence of so-called physiological species in the genus. It seemed desirable also to test the relatiye susceptibility of different varieties qf radish 58 Wisconsin Experiment Station. to Cystopus. For this latter purpose twentytwo varieties of radish were grown and inoculated. In these experiments on different varieties of radish, three-inch pots were used and about ten seeds of the same variety were planted in each of the three pots. When the plants were eight to twelve days old all of the plants in each pot except three of the healthiest were pulled out. At this time each plant was about 1% inches high and had two well developed cotyledons but no true leaves. These plants were then inoculated. In testing each variety, at least nine plants were inoculated. In Table IX are given the results ob- tained from inoculating twentytwo varieties of radish with conidia taken from the varieties Ne Plus Ultra, White Icicle, or Crimson Giant. In every experiment except those started November 9 and November 12, Ne Plus Ultra was used as the source of conidia. In all, ninetyseven inoculations were made and in ninetyfive cases, infection appeared. Controls of Ne Plus Ultra of the same age and under the same conditions were kept in every experiment and always showed abundant infection. 97% per cent of the cotyledons inoculated became infected. This would suggest very little immunity for any of the twenty- two varieties of radish tested, to Cystopus Candidas. It is quite evident that the same form of Cystopus Candidas can grow on all the varieties of radish. Tests were next made to determine whether or not a different species of Raphanus (Raphanus caudatus) could be infected with conidia of Cystopus from Raphanus sativus. Nine plants grow- ing in three different pots were inoculated and all became in- fected showing that Cystopus Candidas on Raphanus sativus is not limited to this species, but can also infect Raphanus cauda- tus. Many tests have been made with the above species since this experiment with similar results. Table IX. — Relative Susceptibility of Different Varieties of Baphanus sativus to Cystopus candidus from: Baphanus sativus, Variety Ne Plus Ultra and Others. Experiments on Spore Germination. 59 O o acsaflgflSEJcscflcccc Dead. OrHOOOOOOO'MOrHr^O'^OOrH^rH(MOOOTH(MOO(MOOTHOOOOOOO xn H 3 d Not 1 inf. HOHHOlMO^MHNOHWNOlMCNHON^OOWOJ^^OHHOHfOWnHO q CO H P3 Inf. Date. in«M/!i(5U5inHOO)OOOOOWaiffittNOwHWW^ffl!OBNHOHOHHOHOifi r- rH (M NWNWNTHrtrtr^HWWMNNMWWMN CO CO CO CO > > > > > > > > > > > >>>>>>>> > t>" > > > > > >■' >' > o> 6 > 6 © 1> d > o' 00000000000000000000000000000©0©0®©0®0© •OJ'-O-O'-O- ■ oo to to iiO t • 03 CO CO Lft CO CO CO < NNNNNNNIMN 1 ' — 1 ’ — 1 ' — 1 ' — 1 ^ ■ — 1 ’ — 1 t — ? — l>- t- I- Cl 7> I °pS lO K 5 UO KO IfJ lf 3 1 ft ift UO K 5 iOir 5 ifiiOiOiOK 5 ifl« 5 ii 5 iflinK 5 (■—l WMWKNOl WW cj >i o © cococococococOKOiouoiouoiomioioiotfouoifOioioicsioioxoiouoioiciiOinioioioiOKOiom O-fa 3 ^ . 5 o © o 2 2 ^ a.S~ C5CiC53;OJOiG5 ci d Q c 3 5 > O O --H gj © © ' ||g| © >> ■ a ; a ’3 :a s v s Sfe-Sfe-S 7 ^ fejs o ^CPiSS^I^gbo^ o §2 * s * ■ — 1 GMix! ©_, © «s s qs © 2 a q’g 5 ©tj. 2 1, ^CO^CKm £.0 £ _ „ „ .fa.. » q fei£i£i2 bu m ,53 q e <-, « ••= §2 b .5 © ©£ g-c ©S 22 -C 2 S-C § oS « © © 2 ©.a Krt o.fa V.fa O.^ o ~ 2 | g | © | O ~eh 2 h m t . c3 © ^ t>, © © 2 ^'Vjq'qx: aQfcfc 0 £W 00 bfc& 5 tfi c 3 * S 3 O . r&m . c _^, .! 43 to 43 m q « 03 © 03 © ' 4 J 4 J - > > >>>•>>■>>> > > > >• > t> >>>>>>>■>>>>> >. > ©© 0 © 00 o 0000000000000000000000000000000 {OWMMmMM^CJNHHNNNMeOCCWWMM'tM- IHHHNWi 60 Wisconsin Experiment Station. SUSCEPTIBILITY OF OTHER CRUCIFERS TO CYSTO- PUS CANDIDUS FROM THE RADISH Tests were made to learn if the spores of Cyst opus candidus from the radish can infect turnips (Brassica rapa). The tur- nip plants used in this series of experiments were grown in the same way as the radish plants, as explained in connection with Table VII. Three healthy plants were allowed to grow in each pot and these were inoculated by spraying them with the spores of Cyst opus candidus from the radish. The spores were sprayed on with an atomizer and the plants chilled in the usual manner. As shown in Table X, ten varieties of turnips (Brassica rapa ) were inoculated with conidia of Cystopus candidus from the radish, variety Ne Plus Ultra. In each variety, at least nine plants were inoculated, while in the case of Snowball, eighteen plants were tested. This would mean that eighteen or more cotyledons were inoculated. These experiments extended from November 13 to January 28, and at no time did any infection result, although in every trial the controls were infected. It was impossible to infect any of the ten varieties of turnips with Cystopus candidus from Raphanus sativus , variety Ne Plus Ultra. These results suggest that there may be a physiological species of Cystopus candidus occurring on radish and turnip respectively. Table X. — Relative Susceptibility of Different Varieties Brassica rapa to Cystopus candidus from Raphanus sativus , Variety Ne Plus Ultra. No. of cult, tested. Date. Host Plant. Source of conidia. I No. pi. inoc. , No. controls inoc. Period refrig- era- tion . Results Date. Inf. coty. 1 1 Controls. || i 3 O 1 © 3 Nov. 30. Yellow aberdeen Ne Plus TTltra. 9 5 4 16 Dec. 17,. 0 Tnf . 3 Nov. 30. Olds’ heavy copper Ne Plus Ultra. 9 5 4 16 Dec. 17.. 0 Inf. 3 Dec. 4.. Golden ball Ne Plus TTltra. 9 5 5 17 Dec. 22.. 0 Tnf. 3 Dec. 6.. White flat dutch Ne Plus TTltra. 9 5 8 11 Dec. 17.. 0 Inf. 3 Dec. 28. Early white milan Ne Plus Ultra. 9 5 17 9 Jan. 13.. 0 Inf. 3 Dec. 28. Cow horn Ne Plus TTltra. 9 5 17 9 Jan. 13.. 0 Inf. 3 Dec. 28. Extra early purple top milan. Ne Plus TTltra. 9 5 17 9 Jan. 13.. 0 Tnf. 3 Dec. 28. Purple top white globe Ne Plus TTltra. 9 5 17 9 Jan. 13.. 0 Tnf. 3 Dec. 28. White egg Ne Plus TTltra. 9 5 17 9 Jan. 13.. 0 Inf. 3 Jan. 10. Snowball, Ne Plus TTltra. 9 5 23 10 Jan. 22.. 0 Tnf. 3 Jan. 22. Snowball Ne Plus TTltra. 9 5 26.5 12 Feb. 15.. 0 Tnf. 2 .T an. 28. Early white milan Ne Plus TTltra. 6 5 26.5 12 Feb. 15.. 0 Tnf. 1 Jan. 2S. Golden ball Ne Plus Ultra. 3 5 26.5 12 Feb. 15.. 0 Inf. No. cult, tested. Experiments on Spore Germination. 61 Summary of Table XI An attempt was made to infect three different varieties of rutabaga {Brassica campestris) , namely: Olds’ Improved Purple Top, New Necklace, Olds’ Large White, with the conidia from the radish, variety Ne Plus Ultra. Three seedlings were grown in two-inch pots in the same manner as described in connection with Table IX. Controls were kept in every experiment and always became infected. Final observations were made in the case of the different varieties of rutabaga several days after the controls had become infected. This was done so as to exclude any possibility of overlooking the disease, which possibly might take longer to develop on the rutabaga. The age of the seedlings inoculated varied from eleven to twentynine days. The plants had one to five leaves. Both coty- ledons and leaves were inoculated. In all, sixteen separate trials were made. In each trial, three plants were inoculated, making a total of fortyeight plants inoculated, and in no case did infec- tion result. These experiments extended from September 25, 1909, to January 28, 1910. Plants of different ages were tested. It is quite evident that Cystopus candidus occurring on Rapha- vus sativus, variety Ne Plus Ultra, will not grow on Brassica campestris , variety Olds’ Improved Purple Top, New Necklace and Olds’ Long White. Table XL— Relative Susceptibility of Different Varieties Brassica cam- pestris to Cystopus candidus from Raphanus sativus Variety Ne Plus Ultra. 3 o o Pertop OF Re- FRI- 3 3 3 3 3. 1. 2 . 1 . Date. Host plant. Source of conidia. % • O o o ■2; GERA- TION. a g S K H Dec. 4.. .Tan. 10. Jan. 10. Jan. 22. Jan. 22. Jan. 28. Sept. 23. Sept. 25. Olds’ improved purple top. New neckless Olds’ large white Olds’ large white New neckless Olds' large white Olds’ large white Olds’ large white Ne Plus Ultra. Ne Plus Ultra. Ne Plus Ultra. Ne Plus Ultra Ne Plus Ultra. Ne Plus Ultra. Ne Plus Ultra. Ne Plus Ultra. 9 9 9 9 9 3 6 3 5 5 5 24 5 24 5 26.5 5 24.5 5 20 5 9.5 5 3 17 10 10 12 12 12 12 9 Results. Controls. 5 -P ,g "H o a o '" 1 (O^HIMOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO CNNONNNOO H- tD "O ‘-D tD iO CO *.D 50 (O PO -* 35 t- t * - S 2 g g 2 s £ 2 g S 2 S S S S S 2 2 S 2 SS33333333SS33333333 < “=3888888833. OOOOOOOOOOOOOOOOOOOO p a p p P JsssJsss^lsssIsssss OGCBC/JX^CCCCCC* S®cB®SaQac!*®oQ® • ^ s-d *111 SS3S38S88aS8asa%S888* , S®S8888888aa8a IIIIIIIIIIIiI!lllllllSS|l i ppppBppppp2p£pBppBpBBpp2p;~ 11 a ,r o ap w m SSWigag „ .2+>,s5a®>?a2®g p£ 322 *2 § * «2 ©2 3 cs-o a; c3 — t. 0 ) 0 / W £ P 3 Sh Q* S 1. ■sl E« 2X3 of IS !| . « & rr o 22§ 8^^ § Ssl g 32* § 822 §2 §2 32 o2 || OOfe«HE-iaQpOOD.COCSfi«OOHJOJOKOWWffl5 “JD *0 s© CD *© S^S88SSSS8SSSSSSgJJJasgs' b b b b f- S-‘ «' U t.' in' >s >5 >. it bi tl L SiSdcsojQaaqipiftOQaaaaaQaQartacJspaQ Experiments on Spore Germination. 65 other series of experiments was made from April 13 to August 24, 1910. The fifteen varities of cabbage were all tested again as well as Kale, Kohlrabi, Brussel’s Sprouts, and Cauliflower as shown in Table XII. From twenty to sixty plants were inocu- ated and tested as described above. No infections were obtained in this series except on one plant of Brussel’s Sprouts. One plant became infected and showed five pustules on its twp cotyle- dons, which again indicates only very slight susceptibility in all the varieties of Brassica oleracea tested. Summary of Table XIII The mustards, both Brassica nigm and Brassica alba were next inoculated with spores of Cystopus Candidas from the radish. The seedlings were grown in three-inch pots in the same way as described for the experiments on the turnips, and inoculated by spraying the spores on with an atomizer. The plants were placed under a bell jar on a wet earthen plate as previously described and placed in the ice box for various periods of time, varying from 3 to 23% hours. Six cultures of three plants each were tested. In all, eighteen plants of black mustard were inoculated with Cystopus Candidas from the radish, variety Ne Plus Ultra, None of the eighteen plants became infected. A variety of white mustard, New White Chinese, was tested eight times. Each time using three plants in the same manner as in the case of the black mustard. Again no infections were secured. Further thirteen tests of three plants each or a total of thirtynine plants of white mustard were inoculated with Cystopus Candidas from the radish, variety Ne Plus Ultra. In five of the trials infection occurred; three coty- ledons and six leaves became infected. The infections secured on the white mustard were not as vigorous as those secured on the controls in the same experiment. The susceptibility of white mustard was further tested in the spring of 1910 on April 11 and May 6. On April 11, three seven-inch pots containing both radish and white mustard growing together, were inoculated when they were nine days old. Infection occurred in seven to eight days. The amount of infection varied from 10 to 40 per cent of the cotyledons. On May 6, eight tests were made in which the seedlings were grown in three-inch pots. The number pf plants per pot varied from seven to forty and when the plants Table XIII. — Susceptibility of Brassica alba and Brassica nigra to Cystopus candidus from Baphanus sativus, Variety Ne Plus Ultra. 66 Wisconsin Experiment Station. og 05 G 22 22 22 S 2 ECSCEEGEEEEEEEEGECSEEEECG o -3 ■c 8 (25 £ OOOOOOOOOOOOOOOiMrirHiMeCO oo'o'^o^o 00^5 '<* co to to to to to t-t-cccoiniftcitooooooodoaoooooooooooaoooooas — — — — — — — 55 02 55’ -< ’ - '’' J ~ 0 *-‘' H '" ,< ’ -fCOOOOOOO<: ' 1 s ' 1 ^ t— in in m 3 , . • g 23 T3 o x» o ® .42 23 -d o a o © !^ 4 H ^ b3 iniflifliflwimoiflininiflioifl^icunioiniflKnomoopooooooo csrocococococoaiastotorocotorocococoeococoMoot-ao ° o3 P !§ SSSSSSIJSSSSSSS 3 3 3 3 3 3 Cl a. CU ^ Ph Ph CU Ph Ph Ph Cl Cl CL 0- CL Cl 52 P-i CL CL Cl £ ” — — — .2 -2 .22 42 .22 .2 © a) ® © ® © ® © ® a) © B’B. •--I r 23 r O f G) sss llllllllllllllllilllllllllllllll 42— ©42 ©42 ©42 © ©fHXj'^xjSSxjSSSSSSSSSlSSSSSS CQ«£Cfi£CP^CC^^£^£ S 5 &S 888 S®®§S£§ga!i 38 S£-gi 8 £ 2 ®®® ; 0 ' 0 ' 0 ® 5 :; 3 ;S o o o o o o o 7777777 QQ 6 --_S (rjHHHHHrtCOrtNH'VHr.MHHHHHF-, Experiments on Spore Germination. 67 were eleven days old they were inoculated with conidia from the radish. All of the experiments showed infection on May 18. From 35 to 50 per cent of the total number of cotyledons were infected as shown in Table XIII. Only a small number of leaves became infected due to the fact that not many of the plants had leaves when they were inoculated. The results in these experiments show clearly that the white mustard is sus- ceptible to the cystopus that occurs on the common radish. Summary of Table XIV During the spring and summer of 1910, various other hosts of Cystopus were grown from seed in the green house and inocul- ated as already described. The plants were grown in three inch pots and were vigorous and healthy when inoculated. In every species, at least two separate tests were made. The number of plants in each pot varied from one to twenty. If the plants to be inoculated were large before inoculated, enough were removed to allow the remainder to develop normally. Twenty-six plants cf Cap sella Bursa-past oris, nineteen plants of Sisymbrium offici- nale, twenty plants of S. altissimum , forty plants of Lepidium sativum, forty plants of Nasturtium officinale, sixty plants of Brassica nigra, and thirty plants of Iberis coronata were inocu- lated and no infections were observed. Inoculations on radish were used as controls in every case and infections were always obtained. Further tests must be made with these plants before positive statements can be made as to their susceptibility to the spores from Cystopus from the radish. It should be noted that a larger percentage of infection was obtained in the spring than in the fall. Thirteen tests were made in the fall of 1909 from October 20 to December 28 in which thirtynine plants were inoculated and only three cotyle- dons and six leaves became infected. While in the tests in the spring from 10 to 50 per cent of the cotyledons became infected. The difference in extent of infection was possibly due to a differ- ence in the host plants. The white mustard seedlings grown in the fall were not as vigorous as those obtained in the spring. No. cult, tested. Wisconsin Experiment Station. 08 Table XIV. — Susceptibility op Other Crucifers to From Radish. Cystopus candidus Date. 3 June 9. 3 I June 9. May 23, Apr. 29. Apr. 29. Apr. 2 t. May 6. Au?. 21. Apr. 30. Host Plant. Capsella Bursa-pastorls Sisymbrium officinale... Sisymorium officinale. . . Lepidium satiyum Brassica nigra Capsella Bursa-pastoris' Nasturium officinale.... iberis umbellatta Sisymbrium altissimum. Source of Conidia. §5 \l o Ne Plus Ne Plus Ne Plus Ne Plus Ne Plus Ne Plus Ne Plus Ne Pius Ne Pius Ultra. Ultra. Ultra Ultra. Uitra. Ultra. Ultra. Ultra. Ultra. 20 20 200 12 12 12 10 20) 20 Period of refri- ger- ation. Results. Con- trols Inf. Hours. Temper- ature. Date. Coty Inf. 8 14 June 24. 0 I Inf. 8 14 June 24. o 1 Inf. 22.5 17 June 8. 0 Inf. 27.5 9 May 25. 0 Inf. 27.5 9 May 25. 0 Inf. 27.5 9 May 25. 0 Inf. 5 10 May 21. 0 Inf. 14 12 Sep. 9. 0 Inf. 6 13 May 13. 0 Inf. DISCUSSION AND CONCLUSION Germination of the Conidia These studies with various species of Cystopus and other Oomycetes have shown that germination of the conidia is con- trolled by certain factors of which temperature is the most im- portant. Prevost (1807:33) in his studies made over a century ago states that at a temperature of 12° to 16° R. Cys- top us candidus spores sometimes germinated in 40 to 45 minutes, whereas they ordinarily required from one to two hours. DeBary (1863:14) found the conidia of Cystopus to germinate at tem- peratures ranging from 5° to 25° C. My own results have led me to conclude that chilling the spores to a relatively low tem- perature is necessary fo,r the most vigorous germination. In our preliminary series of over one hundred cultures of Cystopus spores at green-house temperature, 22 to° 33° C. during the sum- mer, only a very low percentage of germination occurred. In none of the cultures were zoospores observed and the only indi- cation of germination was the presence of an occasional empty sporange which may possibly have been empty before the spores were brought into the laboratory. In a later series (Table II) during the summer, including fortyfive cultures chilled with controls at room temperature, it Experiments on Spore Germination. 69 was found that 85 per cent of the chilled cultures germinated, whereas none of the cultures kept at higher temperatures germ- inated. In a third series (Table V) of seventythree cultures during the following spring where about one third were held at room temperature (i. e. above 20° C.) scanty germination occurred in 48% of the cultures, whereas 69% of the cultures which were kept at a temperature below 20° C. showed abundant germination. This last series shows that germination may occur at room temperatures and above, as has already been pointed out by DeBary (1863:14), Zalewski (1883:215), Biisgen (1882:22), and Eberhardt (1904:614). But it was also clear that the percentage of germination was much increased by using lower temperatures as was further shown in the behavior of the stock cultures described' in connection with Table VIII. The chilled stock cultures became heavily infected wdiile the controls not chilled showed only a low percentage of infection as is conclusively shown by referring to plates I and II. Two things are clearly indicated from our results. First, and most important.; temperature exercises a marked influence upon germination ; second, this influence w r as more marked with spores obtained in the late summer and autumn than with those de- veloped in the spring. We interpret this latter difference as probably due to the greater vigor of host and fungus during the spring months. Be this as it may, it was at all seasons evTdent that comparatively low temperatures were necessary to induce strong normal germination. Various media were used for germination trials with spores of Cystopus ; such as rain water, distilled water, tap water, ex- tracts of the host, sugar solutions and certain nutrient media. No germination was obtained in any medium except water and no marked difference was noted in the percentage of germina- tion obtained whether it was rain water, distilled water or tap water from Lake Mendota. In all of the subsequent tests this tap water was used. Where spores were immersed in a drop of water, the condition of the surrounding atmosphere, whether saturated or dry, and the amount of evaporation had no effect on the percentage of germination. Other physical changes in the culture containing the spores, such as diffusion of the drop and changes in surface tension, were of no consequence. No attempt has been made to determine with exactness the optimum temperature for germination. Indeed, the variations 70 Wisconsin Experiment Station. associated with seasonal and host conditions as just noted and with maturity of the spores may preclude perfectly definite con- clusions upon this point. It was at least strongly suggested by the large number of experiments made in the laboratory and from the observations made out doors that the optimum for normal spores produced under the best conditions was about 10° C. Results as to the maximum temperature of germination tend to substantiate DeBary, who found, as previously noted, that the maximum temperature was 25° C. In a series of over one hundred cultures carried on in the greenhouse during the months of July and August of 1909 at temperatures varying from 22° to 33° no germination was obtained. Again in' a later series (Table V) in the spring of 1910 scanty germination was secured at 25° C. Although these experiments were not planned espe- cially to test this point, yet they show that the maximum tempe- rature of germination is about 25° C. DeBary (1860:236) concluded that the minimum temperature for germination was 5° C. My results show that the conidia of Cystopus will germinate at temperatures below 5° C. Twelve cultures were made and laid on blocks of ice in the ice chamber of the ice box and nine of the cultures germinated in 4% to 27 hours. The temperature of the cultures was between 0° and 10° C. In view of these facts it is quite clear that the minimum is very near zero. Not only do the conidia germinate at low temperatures in the laboratory but also out of doors. Observations were made on seven different days between 5 and 9 o’clock A. M. in the au- tumn of 1910, and in every case except one, zoospores were found on the infected leaves of both radish and salsify. The minimum early morning temperatures on the days when obser- vations were made, varied from 5 to 11 2-3° C. as is shown in Table VI. DeBary also records having found the motile zoo- spores in the morning dew. Since the motile zoospores were thus found in the morning dew by both DeBary and myself, it sug- gests that the conidia germinate in the coolest part of the day when moisture is at hand. According to Salisbury (1908:556) it is a well established fact that the surface of the earth is the coolest at about sunrise, a condition that leads to the formation of dew and thus moisture and low temperature naturally asso- Experiments on Spore Germination. 71 elate themselves in the environment of the fungus and may well have come to have a correlated influence on its development. Somewhat similar conditions have been reported in the rusts by Jaczewski (1910:21), who found that in the cereal rusts of Russia, both the uredospores and aecidiospores germinated in the morning when the foliage was wet with dew and the tempera- ture was low. The relation. of dew to the asparagus rust has been pointed out by Smith (1904:19) in California. He found that the rust spreads most rapidly when heavy dews are preva- lent. It should be noted, however, that Smith (1904:19) men- tions no temperature factor as especially important. The relation of light to the germination of the spores of Cys- topus was not as marked as it has been reported for Plasmopara and various species of Phytophthora. Farlow (1875:319) con- cluded from his studies with Plasmopara viticola and Phytoph- thora infestans that the conidia germinated better in the dark than in the light. Coleman (1910:59) who has recently studied Phytophthora omnivora states that light is a very important stimulus to germination. I have found in Cystopus that the conidia do not germinate in the light at high temperatures and that they do germinate in the light at low temperatures. My conclusion on the first point is based upon seventeen experiments tabulated in Table III, while the latter conclusion is evident from my outdoor observations (Table VI) and also from laboratory studies not tabulated. I have also incidentally tested Phy- tophthora infestans and Plasmopara viticola as to the relation of light to germination and have found no such marked differ- ence as has been reported by Farlow (1876:419). Zalewski (1883 -.215) concluded that the time of the year had an effect on the time required for the germination of the spores of Cystopus. He found that the time required for germination in the summer was two or three hours, while in the fall it re- quired from one to three days. I have found that during August the average length of the period of refrigeration was 3% hours ; September, 7% ; and October, 9 hours. My results show without doubt that the time of the year has a direct influence on the time required for germination. It may well be due to the different host reaction on the fungus in spring and fall. And again, the different weather conditions of spring and fall, may have a direct influence on Wisconsin Experiment Station. the conidia. The cause of this increase in time required for germination, I do not know, and it cannot be definitely deter- mined, until we are able to absolutely control all of the factors influencing the host and fungus. A more direct comparison of my results with those of Zalewski would be possible if we knew the method he used in Ins germination experiments and the number of tests made. In none of my experiments did the conidia of Cystopus germ- inate by the production of germ tubes as described by Tulasne (1854:77) and Hoffmann (1859:210). Zoospores were always produced when the conidia germinated. It should be noted that not all of the conidia germinated. Some of the spores were dead or immature. Eberhardt (1904:614) considers the proper maturing of the spores as the most important factor in securing germination. As ncted above, his method was to carefully collect infected leaves with unopened pustules, wrap them in moist cloth and place them in a damp chamber until the pustules were about to burst. With these precautions Eberhardt experienced no trouble in germinating the conidia. I have never found it necessary with the method of chilling described to exercise such precaution. To be sure, it is quite necessary in spore germination to have a goodly supply of ripe spores, but in Cystopus, where the spores are produced acropetalously and borne in a pustule, no such precautions were required in order to secure plenty of ripe spores. Had Eberhardt been studying Phytophthora infestans or Plasmopara vit'icola, where the spores are borne on long, much branched conidiophores standing out from the surface of the k af, such a procedure might have been more important. Conidi?« i i om pustules in all stages of development have been used and the conidia readily germinated when chilled. In some of the experiments described above and in many of my preliminary experiments (Seepage 34) not chilled, conidia from pustules just about to open were used without securing germination. In only five cases does Eberhardt (1904:624) record the data of his ex- periments. In two of his tests the temperature varied from 2° to 17° C. In another test the spores were placed in water at 6 o’clock in the afternoon and germination was observed the next morning at 7 o’clock. Under ordinary conditions, the room temperature during the night would be lower than in the day Experiments on Spore Germination. 73 and may well have been between 15° and 20° C., which is sufficient chilling to cause germination. The two remaining tests were made on June 6 and August 8. No record is made of the temperature conditions. On the basis of such data and the microscopical examination of the ccnidia from pustules just opened in which he found the conidia swollen and bottlenecked, Eberhardt concludes that germination depends upon the proper maturing of the conidia. In view of the fact that, in two of the tests made by Eberhardt, the temperature was much below 17° C., and in another the temperature was that of night time, it seems to me probable that temperature may have been a more important factor in Eber- hardt ’s experiments than he realized. Since I conclude that temperature is a controlling factor in spore germination of Cystopus, it is worth while to make a comparison with results obtained in the case of other fungi. In the Myxomycetes, Jahn (1905:489) has found that if the spores are soaked in water for 36 hours and then allowed to dry out, they will germinate in about 30 minutes when again moist- ened. High temperatures for short periods and then normal temperatures tend to hasten the germination of the spores of Reticularia. More recently Kusano (1909:8) has shown that a weak acid solution is possibly the normal stimulus and that a temperature below 20° C. retards germination. It is at once apparent that in the Myxomycetes thus far investigated, there is no correlation of low temperature and spore germination. In certain of the fleshy Basidiomycetes investigated by Duggar (1901 :38) and Miss Ferguson (1902:16) they found that tem- perature changes had only a slight tendency to increase the per- centage of germination. In the rusts or parasitic Basidiomy- cetes, on the other hand, the relation of temperature to germina- tion is more marked, as is evident from the wmrk of Ericksson and Henning (1896:73) and Jaczewski (1910:21). The former found that fresh aecidiospores sown in water at room tempera- ture gave a very low percentage of germination, but when the spores were placed on blocks of ice for a while and then returned to water of room temperature, a much higher percentage of germination was obtained. It should be pointed out in this con- nection, however, that the above investigators made too few ex- periments to draw definite conclusions. Their results arc fuT" 74 Wisconsin Experiment Station. thermore misleading in that such extreme temperatures were used. The germinations obtained by Ericksson and Henning were possibly not due to the temperature of melting ice, but rather to the slightly higher temperatures obtained after the slides were removed from the blocks of melting ice. In similar experiments performed with the spores of Cystopus this has been found to be the case. The observations of Jaczewski (1910:21) further substantiate my conclusions. He found that the aecidio- spores of Puccinia graminis germinated in the morning dew out- doors and at temperatures considerably below normal in the laboratory. The uredospores were found to germinate best also at temperatures slightly below normal (18° C.). It is evident from the above facts that the spores of the sapro- phytic Basidiomycetes and parasitic Basidiomyaetes respond differently to temperature at the time of germination and we would naturally expect that forms so different in habit and en- vironmental relations would respond differently. In the later cases, it is purely an adaptation to environmental conditions in much the same way as I have found it to be in the Oomycetes, although the relation is less marked in the rusts than it has been found to be in Cystopus and various other Oomycetes. In the Ascomycetes there has been no correlation of low tempera- ture and spore germination revealed up to the present time but too little is known of the factors influencing germination in this group to draw any conclusions. In the Fungi Imper- fecti, on the other hand, also little studied as to the .factors influencing spore germination, it has been noted in one instance that low temperature has a direct relation to spore germination. It is very important that parasitic fungi belonging to the two above mentioned groups should be investigated as to tempera- ture relations because of their direct bearing on remedial meas- ures. In order to inoculate various hosts with Cystopus, Eberhardt (1904:625) germinated the spores as described above and placed the water containing the zoospores on the cotyledons of the plants to be infected. In some cases the plants were immersed in water containing zoospores. The method of inoculating plants w T ith the conidia of Cystopus that has been used in my experiments was based on the relation of chilling to germination of the coni- dia. The spores were placed in water and sprayed on the plants Experiments on Spore Germination. 75 with an atomizer, then the plants were covered with a bell jar and placed in the ice box long enough to insure germination. The method I ,have used is more nearly that which occurs in the normal environment of the fungus than that used by Eberhardt. No difference in the susceptibility of the cotyledons and leaves has been noted in any of my infection experiments, although DeBary (1863:24) concluded from his experiments that in Cap- sella and Lepidium the cotyledons only were susceptible to in- fection and that in various species of Brassica, both cotyledons and leaves were susceptible but usually only the cotyedons. Still further tests were made as to the susceptibility of the leaves of the above hosts. Twentyfour radish plants were used, two in each of twelve pots, which had been grown in the greenhouse and had at no time shown any infection. After these had lost their cotyledons they were inoculated. Thirteen days later twenty of the plants, i. e., all but four, showed leaf infection at many points. A pot culture of at least fifty white mustard plants having lost their cotyledons and at no previous time show- ing infection were inoculated and every plant showed infection on at least several of its leaves. Five plants of Capsella in blos- som were inoculated and four of the plants became infected, de- veloping large white pustules on both the stems and young fruits. We have never tested Lepidium sativum with Cystopus from the same host but have succeeded many times in infecting the leaves with Cystopus from Capsella. The same care was exercised in growing the tw T o named hosts free from infection as was noted for radish and white mustard. The details of these experiments are given in an earlier part of this paper. There can be no doubt of the susceptibility of the leaves of the various hosts described. I have noted, however, that the leaves of radish plants about to blossom or in blossom seldom become infected and when infection does occur a marked hypertrophy results. This was evident from stock culture J, which may be taken as typical of the nine described in connection with Table VII. It w T as started September 29, 1909, and became infected October 6. The pustules until March, 1910. Then the stronger of these plants sent up flowering stalks,- bearing scattered leaves and blos- plants sent up owering stalks, bearing scattered leaves and blos- soms. From this time on the fungus development on the basal leaves started to disappear and the upper leaves remained prac- tically free from infection, not only in culture J, but also in the 76 Wisconsin Experiment Station. other eight cultures listed in Table VII. This same condition has been repeatedly observed on plants growing outdoors, and 1 believe that the leaves of the radish plants are not less susceptible at the time flowers are developing than earlier, hut that the decrease in extent of infection is due to less moisture being deposited on the upper leaves. The flowers and young fruits, cn the other hand, may become the seat of systemic infection at this stage of the host plant, in which case oospores are produced. With the method of infection well established my attention was directed to determining the relative susceptibility of the different varieties of radish. Twentytwo varieties were found to be susceptible with no marked degree of variation. Another species, Raphanus caudatus (rat-tail radish), was tested to de- termine whether different species in the same genus were sus- ceptible to the conidia from Raphanus sativus (common radish). It was found that Raphanus caudatus was readily infected. Further tests were made with the conidia of Cyst opus Candidas from radish on species of other genera; Brassica rapa (turnip), B. campestris (rutabaga), B. napus (rape), B. nigra (black mustard), B. oleracea (varieties: cauliflower, kohlrabi, and kale), Capsella Bursa-past oris (shephard’s purse), Lepidium sativum (garden cress), L. virginicum (wild pepper grass), Sisymbrium officinale (hedge mustard), S. altissimum and Iberis umbellata ( candy-tuft). In none of the above cases did infection occur. Infection was secured on both cotyledons and leaves of Brassica alba (white mustard) and on the cotyledons of Brassica oleracea (four varieties of cabbage). My results show that it is possible to inoculate several other crucifers with the spores of Cystcpus obtained from the radish, which tends to preclude the possibility of so called physiological species in accordance with Eberhardt’s conclusions ; yet it may well be that limited specialization exists when further cross inoculations with the spores from other hosts have been made. Eberhardt has already raised the ques- tion as to the existence of a biological form on each of the groups : Lepidium — Capsella — Arabis and Brassica — Sinapis Diplotaxis. My results show further that the spores of Cystopus on the various species of Raphanus are quite limited but it may be that Brassica alba serves as the bridging species. These are questions that can be fully determined only by a large Experiments on Spore Germination. * 77 number of cross inoculations with the spores from various hosts of Cystopus. As has been pointed out, the infections that were secured on B 'i us sic a alba and B. oleracea with the conidia from radish dif- fered in appearance from those usually occurring on the radish. Ihe infection on the radish is vigorous, causing marked hyper- trophy and developing large, white, plump pustules. On the white mustard and cabbage this was not the case. No hyper- trophy occurred and the pustules were small, showing none of the signs of vigor evident on the radish. Not only was there a marked difference in the appearance of the fungus pustules on the hosts in question but also in the effect upon them. The fun- gus killed the host tissues very much faster on the white mustard and cabbage than on the radish. A possible explanation of these results would be that the infection of the white mustard and cabbage occurs only in the most vigorous cotyledons; that in these the fungus is able to overcome the host cells and persist in only a few cases and that in such, the host cells when overcome die immediately. In my observations, plants infected with aphids or thrips seem to be quite immune to Cystopus. At no time was I able to get infection on a plant that was badly infected with insects. Reed (1907 :381) also found that it was quite impossible to infect grain seedlings with mildew that were already infected with thrips. This was more evident in the case of wild plants such as Capsella, Lepidium and Sisymbrium than in the case of such cultivated plants as radishes and mustards. The lack of infection can not be attributed to the aphids eating the spores, since some of the plants were fumigated, killing the insects and then inoculated, with similar results. These facts lend support to Cook’s (1911:624) view that plants injured by plant or ani- mal parasites develop an excess of tanin which causes more or less immunity. Not only was it quite impossible to infect plants attacked by insects, but likewise, plants that showed signs of not being vigorous and healthy from other causes. It was also im- possible to infect wild and cultivated seedlings that showed yel- lowing of the cotyledons and first true leaves. This was also true of the more mature plants. If for any reason a stock cul- ture of radishes showed signs of not being healthy and vigorous the extent of infection was at once reduced. As has been 78 Wisconsin Experiment Station. stated earlier, Eberhardt believed that the various hosts of Cystopus do not at all stages of development show the same susceptibility for the fungus. Nowhere does Eberhardt have any data to substantiate this conclusion, nor has he taken into consideration host abnormalities as a factor influencing the question of susceptibility. Prom my results it is at least very evident that Cystopus reacts differently to healthy and sickly plants respectively. It is impossible to infect Capsella Bursa-past oris, Lepidium virginicum, or Sisymbrium officinale when the plants are not vigorous and healthy. Many attempts were made during the fall of 1909 to infect Sisymbrium officinale with Cystopus can- didus from the same host but the infections were very scanty. Out of the fourteen experiments on fiftysix plants, only eight plants became infected. I attribute this to the weakness of the plants that were grown at that time. In every case, it was the largest and healthiest looking plants in the lot which took the disease. Although I have not succeeded in proving entirely to my own satisfaction that the extent of infection is dependent upon the vitality of the host ; yet it seems highly probable that this is the case. Reed (1907 :381) has fully described a similar relation between the host and fungus in the grain mildews. Since there is this evidence in both the mildews and white rusts that sickly hosts do not readily become infected, in testing a species for socalled physiological species, all possible care should be exercised in cases where plants are used as hosts that are at all difficult to grow. Failure to infect may be due to weakness of the host plant. Experiments on Spore Germination. 1 9 SUMMARY The studies outlined in the preceding pages were carried on chiefly with Cystopus candidus as it occurs on the common radish, Baphanus sativus. The leading problems considered are: Con ditions influencing germination of the conidia; conditions in- fluencing infection ; and, the occurrence of so-called physiological species of Cystopus candidus on the various crucifers. Germination of Conidia When the conidia are placed in water they germinate better a strikingly low than at high temperatures. The optimum was not definitely determined, but the results tend to show that it was 10° C. The minimum temperature of germination was very near zero, while the maximum was, as DeBary has shown, about 25° C. ' ; ! | :■ It was found that water is the most favorable medium for germination. No germination was obtained on various nutritive culture media. The time required from the immersion of the conidia to the escape of the zoospores usually varied from two to ten hours. The shortest period in which such germination was observed was 45 minutes. Environmental factors, season and host vitality, seemed to in- fluence the time required for the spores to germinate. It was strongly suggested that the time required in spring and summer is shorter than in the late fall and winter. No difference was observed in the time or percentage of germ- ination which occurred in light as compared with darkness. Spores obtained from leaves after a killing frost germinated. Such factors as evaporation, surface tension, and diffusion of the drop containing the conidia did not influence the percent- age of germination. The conidia germinated as readily in a non- saturated as in a saturated atmosphere. / Conditions of Infection Chilling was also found to have a very marked effect on the degree of infection secured, as can be seen by referring to plates I to X. Ninetyfive per cent of the seedlings chilled became in- 80 Wisconsin Experiment Station. fected while the controls not chilled usually showed less than ‘3 per cent of infection and never more than 15 per cent. This difference in extent of infection. I believe was due to the in- creased percentage of spore germination. It should be noted,, however, that the chilling process may have had some effect on the host, possibly making it more susceptible. This is a point that needs further investigation. the favorable effect of chilling on the conidia of Cystopus is plainly an adaptation to the environment of the fungus. The spread of a fungus by zoospore infection is directly dependent upon the presence of water on the foliage of the host. DeBary found the motile zoospores of Cystopus in the dew drops in the morning on the host plant and I have often made the same ob- servation. The fall in temperature which leads to the deposition of dew and thus provides a medium in which the zoospores may develop serves at the same time as the necessary stimulus to the germination of the conidia. The results obtained suggest that a close relation exists be- tween host vigor and susceptibility in that healthy plants are more susceptible than sickly or abnormal ones. No marked difference in the susceptibility of leaves and cotyle- ; dons of the radish, shepherd’s purse, white mustard and gar- : den cress was observed. . ] So-called Physiological Species i Repeated infection experiments were made using conidia of Cystopus candidus from the common radish, Raphanus sativus. upon this same and other cruciferous hosts to learn whether ; there is any difference in susceptibility. A large number of experiments were made testing the suscep- j tibility of twenty two different varieties of radish, and it was \ found that no marked difference in their susceptibility existed. j It was also readily possible to infect Raphanus candatus with the * conidia from Raphanus sativus which shows that species of the sii me genus are susceptible to the form of Cystopus that occurs on the common radish. Species of crucifers from other genera known to be hosts of the white rust were investigated as to their susceptibility to Cystopus candidus from the common radish. Infection was se- cured on the white mustard, Brassica alba , and cabbage, Bras- 5 Experiments on Spore Germination. 81 sica oleracea. At no time was it possible to infect more than 50 per cent of the cotyledons or leaves of white mustard which were inoculated. With the cabbage, it was even more difficult to secure infection, although fifteen varieties weye tested. Less than 1 per cent of the plants inoculated became infected. No infection could be secured on any of the other crucifers tested. These included turnip, Brassica rapa, ten varieties; black mustard, B. mg\m, rutabaga, B. campestris, three varie- ties; shepherd ’s purse, Capsella Bursa-past oris; garden cress, Lepidium sativum; wild pepper grass, Lepidium virginicum; hedge mustard, two species Sisymbrium officinale and S. altis- simum; candy tuft, Iberis umbellata; water cress, Nasturtium officinale , and wall flower, Cheiranthus cheiri. LITERATURE CITED 1807. Prevost, B. : Memoire sur la Cause immediate de la Carie ou Charbon des Bles, etc., pp. 33-35. 1854. Tulasne, L. R. : 2nd memoire les Uredinees et les Ustil- aginees. Ann. Sci. Nat. Bot. Series IV 1&2:77. 1859. Hoffmann, II.: Ueber Pilzkeimungen. Bot. Ztg. 17:210 1860. DeBary, A. : La formation de zoospores. Ann. Sci. Nat. Bet, series IV 13&1 4:236. 1863. DeBary, A. : Recherches sur le developpement de quel- ques champignons parasites. Ann. Sci. Nat. Bot., series IV. 2 0:14. 1873. Wiesner, J. : Sitzber Akad. Wiss. (Vienna) Math. Phys. Kl. 68, 1. (Abstract from DeBary Morphology of Fungi, p 349.) 1875. Farldw, W. G. : The Potato Rot. Bui. Bussey Inst. Part IV p. 319. 1876. Farlow, W. G. : The American Grape Vine Mildew. Bui. Bussey Inst. Art. I. p. 419. 1882. Busgen, M. : Die Entwicklung der Phycomycetensporan- gien. Jahrb. Wiss. Bot. [Pringsheim] 13:22. 1883. Zalewski, A. : Ueber Sporenabschnurung und Sporenab- fallen bed den Pilzen. Flora 66 :251. J883. Zalewski, A. : Zur Kenntniss der Gattung Cystopus, Bot. Centbl. 15:215. 82 Wisconsin Experiment Station. 1886. Scribner, F. L. : The Downy Mildew. Bot. Diy. U. S. Dept. Agri. Bnl. 2, p. 10. 1893. Viala, P. : Les maladies de la Vigne, p. 93. 1895. Eriksson, J. : Ueber die Fordernng der Pilzensporen- keimnng durch Kalte. Centbl. Bakt. (etc.). 2 Abt. 1 :557. 1896. Eriksson and Henning. : Die Getreideroste. Stockholm. p. 73. 1901. Ludi, R. : Beitrage zur Kenntniss der Chytridiaceen. Hedwigia, 40:1-44. 1901. Duggar, B. M. : Physiological Studies with Reference to the Germination of Certain Fungous Spores. Bot. Gaz. 31:38-66. 1902. Ward, IT. M. : On Relation between Host and Parasite in the Bromes and their Brown Rusts. Ann. Bot. 16:265. 1902. Ferguson, Margaret C. : Germination of the Spores of Agaricus eumpestris and Other Basidiomycetous Fungi. U. S. Dept. Agri. Bur. Plant Indus. Bui. 16:16. 1903-4. Eberhardt, Albert : Zur Biologie von Cystopus candi- dus. Centbl. Bakt., (etc.). Abt. 2. 1 0 :655-656. 1904. Eberhardt, Albert: Contribution a Petude de Cystopus * candidus. Centbl. Bkt. (etc.). Abt. 2. 12:614-631 and 71T-727. : 1904. Clinton, G. P. : Downy Mildew, or Blight of Musk i Melons and Cucumbers. Conn. (New Haven) Agri. j Exp. Sta. Rpt. Part 4. Botanist Rpt. p. 338. 1905. Jahn, E. : Myxomyeetenstudien. Ber. Deutsch. Bot. Gesell. Abt. 2. 23:489. 1906. Reed, Geo. M. : Infection Experiments with Erysiphe > graminis. Wis. Acad, of Sci., Arts and Letters. Part 1, | 15 :135. j 1907. Reed, Geo. M. : Infection Experiments with Erysiphe j Cichoracearum DC., Bui. of Wis. Univ. No. 250,. Sc. ' series 3, No. 2. 3 :381. 1904. Smith, R. E. : The Water Relation of Puccinia Asparagi. Bot. Gaz. 38 :19. Experiments on Spore Germination. 83 1908. Clinton, G. P. : Artificial Cultures of Phytophthora with Especial Reference to Oospores. Conn. (New Haven) Agri. Exp. Sta. Rpt. Part 4. Botanist Rpt. p. 904. : : I ■ , 1 1! »] 1908. Salisbury, R. D. : Physiography, p. 356. 1910. Jaczewski, A. : Studien uber das Verhalten des Schwarz- rostes des Getreides in Russland. Ztschr. Planzen- krank. 20:21. 1911. Cook, M. : Protective Enzymes. Science, n. s, 33:625. 84 Wisconsin Experiment Station. DESCRIPTION OF PLATES The following plates are all . from photographs of radish plants grown in six inch pots, taken from above. In some cases the entire culture is shown with a slight reduction. In the rest only a portion of the culture is shown, but so selected as to be fairly representative. The plates illustrate the advan- tage of chilling in securing optimum spore germination and favorable conditions for infection with Cyst opus candidus. The seedlings were inoculated, covered with bell jars and either- placed in an ice box or kept at room temperature. I. Radish. Var.— Ne Plus Ultra. Sowed May 16. Inoculated May 26. Infected -June 2. Photographed June 6. This culture was chilled. II. Control. Radish. Var.— Ne Plus Ultra. Sowed May 16. Inoculated May 26. Infected June 2. Photographed June 6. This culture was not chilled. III. Radish. Var.— Ne Plus Ultra. Sowed Nov. 1. Inoculat- ed Nov. 8. Infected Nov. 15. Photographed Nov. 17. This eulture was chilled. IV. Control. Radish. Var.— Ne Plus Ultra. Sowed Nov. 1. Inoculated Ncv. 8. Infected Nov. 15. Photographed Nov. 17. This culture was not chilled. V. Same culture as shown in III. Photographed twelve days later, Nov. 29. VI. Same culture as shown in IV. Photographed twelve (days later, Nov. 29. VII. Radish. Var.— Ne Plus Ultra. Sowed Nov. 6. Inoculat- ed Dec. 3. Infected Dec. 12. Photographed Jan. 1, 1910. Cul- tures at right of page chilled ; at left of page controls, not chilled. VIII. Radish. Var. — Triumph. Sowed Dec. 12, 1909. Inoculated Jan. 3, ’10. Infected Jan. 12, 1910. Photographed Jan. 18, 1910. Upper culture chilled. Lower culture control, not chilled. IX. Radish. Var.— Ne Plus Ultra. Sowed May 26, 1910. Inoculated June 6. Inf. June 12. Photographed June 15. This culture was chilled. X. Control-Radish. Var.— Ne Plus Ultra. Sowed May 26,1910. Inoculated June 6. Inf. June 14. Photographed June 15. This culture was not chilled. Experiments on Spore Germination, 85 I.— Radish seedlings inoculated with Cystopus, chilled. (Compare with II.) Experiments on Spore Germination 86 II. — Radish seedlings inoculated with Cystopus, not chilled. (Compare with I.) 4 j ,3 ( . Experiments on Spore Germination. 87 III —Radish seedlings inoculated with Cystopus; chilled. IV.— Radish seedlings inoculated with Cystopus; control not chilled. % % < 3 ^ e % V* Experiments on Spore Germination. 88 V.— Radish seedlings inoculated with Cystopus; chilled. VI.— Radish seedlings inoculated with Cystopus; control not chilled. Experiments on Spore Germination. 89 ATI.— Four small cultures of radish seedlings; two at right of page, chilled; two, controls, at left of page, not chilled. \ \