S H 
 
 T6 
 
 FISH-CULTURAL PRACTICES IN THE 
 UNITED STATES BUREAU OF FISHERIES 
 
 From BULIvETIN OF THE BUREAU OF FISHERIES, Volume XXVIII, 1908 
 
 Proceedings of the Fourth International Fishery Congress : 
 
 Washington, igo8 
 
 WASHINGTON ::;::: GOVERNMENT PRINTING OFFICE :::::: 1910 
 
Class :^B \ 5 \ 
 
 Book _— L^ 
 
FISH-CULTURAL PRACTICES IN THE 
 UNITED STATES BUREAU OF FISHERIES 
 
 From BULLETIN OF THE BUREAU OF FISHERIES, Volume XXVIII, 1908 
 Proceedings of the Fourth International Fishery Congress : : Washington, igo8 
 
 -\;\ 
 
 WASHINGTON :::::: GOVERNMENT PRINTING OFFICE 
 
 1910 
 
5^ 
 
 s^^ 
 
 BUREAU OF nSHERIES DOCUMENT NO. 675 
 Issued April, 1910 
 
 
 016i OS >^!l' 
 
 ■ <j a c V 
 
a 
 
 FISH-CULTURAL PRACTICES IN THE UNITED STATES 
 BUREAU OF FISHERIES 
 
 By John W. Titcomb 
 
 Assistant in Charge of the Division of Fish Culture 
 
 Address before the Fourth International Fishery Congress 
 held at Washington, U. S. A., September 22 to 26, 1 908 
 
 697 
 
CONTENTS. 
 
 ' Page. 
 
 Resources of the Bureau ^99 
 
 Extent and general methods of the work 699 
 
 River fishes of the east coast 702 
 
 Shad 702 
 
 White perch 704 
 
 Yellow perch 705 
 
 Striped bass 7o6 
 
 Fishes of the Great Lakes 707 
 
 Whitefish • 707 
 
 Pike perch 708 
 
 Lake trout 7'° 
 
 The brook trouts, chars, and eastern salmons 712 
 
 Spawntaking in Colorado 7I3 
 
 Rearing methods 7^4 
 
 Special devices applied at trout stations 716 
 
 Atlantic salmon 7i8 
 
 Pond culture 719 
 
 The ponds 7i9 
 
 Food for the adult fishes 720 
 
 Artificial nests 720 
 
 Number of brood fish 7^1 
 
 Collecting the young iish -- 722 
 
 Rescue of fishes from overflowed lands 724 
 
 The Pacific salmons 725 
 
 Barricades and traps to intercept spawning runs 726 
 
 At Baird, Cal 726 
 
 At Battle Creek, Cal 728 
 
 At Birdsview, Wash 728 
 
 Taking and hatching the eggs 732 
 
 Seining operations and spawntaking at California stations 732 
 
 Local practices in different regions 735 
 
 Marine fish culture 73^ 
 
 Cod, pollock, and flatfish 73^ 
 
 Lobsters 737 
 
 Measuring and counting the eggs and fry 739 
 
 Transportation of eggs 74' 
 
 Usual style of. packing case 742 
 
 Adaptations and variations of method 742 
 
 Argentine case 747 
 
 German-Chile case 750 
 
 Transportation of fish 752 
 
 Distribution and planting of commercial fishes 755 
 
 Distribution of game fishes 75^ 
 
 698 
 
Bn.. r. S. n. F., 1908. 
 
 Plate LXXVII. 
 
 iMC. I.— Bi-ook ti-.nU eggs cm Ivay, just bcginniiig t.i lintcli. 
 Fig. 2.— Brook trout fry in trough, sac stage-. 
 Fig. 3.— Brook trout fry, sac nearly absorbed, ready to feed. 
 Figures about natural size. See also figure 12, plate Lxxxii. 
 
FISH-CULTURAL PRACTICES IN THE UNITED STATES BUREAU 
 
 OF FISHERIES." 
 
 By JOHN W. TITCOMB, 
 Assistant in Charge oj the Division of Fish Culture. 
 
 RESOURCES OF THE BUREAU. 
 
 The fish-cultural work of the United States Bureau of Fisheries, adminis- 
 tered from headquarters at Washington through the Division of Fish Culture, 
 is conducted by means of numerous hatching stations located in various parts 
 of the country, including Alaska. There are thirty-two such stations recognized 
 for administrative purposes. Auxiliary to these are numerous field stations, 
 some of them equipped as hatcheries and in operation throughout the year, 
 others used only during the spawning season of the fishes they are concerned 
 with. The land owned by the Bureau at its fish-cultural stations has an aggre- 
 gate area of 12,000 acres, valued at $225,000. The improvements on this land 
 in buildings, ponds, and special equipment represent an investment of about 
 $1,000,000. The Division of Fish Culture has 225 civil-service employees, whose 
 salaries aggregate approximately $200,000; while there is an annual appropria- 
 tion of $275,000 for the maintenance and operation of stations, including the 
 employment of temporary laborers and assistants in fish-cultural work. 
 
 It will be my endeavor to allude to some of the more important phases of 
 the fish-cultural work which these resources make possible, and to explain more 
 fully than would otherwise be permissible features of the work in which recent 
 changes in methods or equipment have not heretofore been so fully illustrated 
 or described. 
 
 EXTENT AND GENERAL METHODS OF THE WORK. 
 
 During the fiscal year 1908 the Bureau distributed fish and eggs to the 
 number of 2,871,456,280. The table accompanying gives an idea of the extent 
 of the work for a series of years. 
 
 "This paper as read before the Inter^.^tional Fishery Congress was illustrated by lantern slides, to 
 supply the absence of which in publication the text has been amplified. 
 
 699 
 
700 
 
 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 Aggregate of Adult Fishes, Yearlings, Fry, and Eggs Distributed by the Bureau of Fisheries 
 
 FROM 1872 TO 1908. 
 
 Species. 
 
 1872-1897- 
 
 1898-1908. 
 
 Shad 
 
 1.379.247.350 
 
 J68.073. 769 
 
 70.499. 66s 
 
 1.579.923.409 
 
 7.212,403 
 
 796.027.818 
 
 520. 223,000 
 
 94.522.019 
 
 47.446,864 
 
 387,989. 712 
 
 1.381.209.594 
 
 1.050. 738.390 
 
 S06. 781.98s 
 
 3.889.418, I3S 
 
 34, 702, 267 
 
 5.056.995, 274 
 
 2, 241, 063, 000 
 
 I. 921, 625. 000 
 
 7. 217.024 
 
 I. 182,471.440 
 
 
 Trouts _ 
 
 
 
 Perches - - 
 
 
 
 
 
 
 
 Total 
 
 5. 051, 166, 009 
 
 17, 272. 222, 109 
 
 
 The first column represents the output of the United States Fish Commission 
 from its inception to the beginning of the present Commissioner's administration, 
 a period of twenty-six years. The second column represents the output for the 
 subsequent period of eleven years. The total operating expenses of the entire 
 eleven years, 1898 to 1908, did not exceed those of the preceding twenty-six 
 years, while the output has been more than trebled. About 50 species of fishes 
 are now handled. 
 
 The results of fish culture, as shown by numerous replenished waters and 
 by actual returns in fish, might easily be made the subject of a lengthy discourse, 
 but for present purposes will be alluded to only incidentally. A marked evi- 
 dence of success may be noted in the constantly increasing demand for young 
 fish to plant. Notwithstanding the fact that the bureau, by the increase of its 
 facilities and by progressive methods, has steadily increased its output, the 
 demand of the pubHc for fish has increased therewith until in some lines of 
 work, notably the production of the basses, crappies, other sunfishes, and the 
 catfishes, it is greater than can be met with present means. 
 
 It is a point to be emphasized that the fish-cultural work of the Bureau is 
 of two classes with respect to its economy. Many of the most valuable food 
 fishes, being in their prime for market purposes just prior to the spawning 
 season, are most extensively captured at the very time they should be spared 
 for the perpetuation of their kind. Whenever possible, the Bureau secures 
 the eggs of these fish from the fishermen. Fully 96 per cent of all the 
 eggs collected and hatched by the Bureau are taken and fertilized from fishes 
 destined to the market, and this without detracting from the value or edible 
 qualities of the parent fish. 
 
 The collection of fish eggs under these conditions is to be distinguished from 
 the work which utilizes the eggs of fish that have reached their spawning grounds 
 and which it is customary to capture for the express purpose of obtaining and 
 fertilizing their eggs. The latter is fish culture in the usual sense — extensive 
 
FISH-CULTURAL PRACTICES IN^ THE BUREAU OF FISHERIES. 701 
 
 increase of the numbers of young by protecting the eggs and providing the 
 most favorable conditions for hatching. The other is fish culture also, but in 
 addition to conservation of a resource that would otherwise be unavailed of. 
 
 Some of the freshwater species, valued chiefly as game fishes although 
 marketed also, are cultivated by confining them imder conditions which will 
 secure the maximum reproduction by natural processes. Practically all of the 
 important commercial fishes, however, can be propagated, and much more 
 ^ numerously, by stripping them of eggs and milt by hand and incubating the 
 fertilized eggs in hatcheries. It is with these that the Bureau is most largely 
 concerned, their numbers being nearly 98 per cent of the entire output of the 
 hatcheries. 
 
 There are some variations in the methods of spawntaking, according to species, 
 but in general the operation consists in expelling the eggs by a gentle pressure 
 of the thumb and forefinger along the walls of the abdomen, the strokes being 
 continued until all ripe eggs have been secured. The fish is usually grasped 
 near the head, and to hold it firmly may be pressed against the body of the 
 spawntaker. The receptacle into which the eggs are expelled is usually a 
 6-quart milk pan which has been dipped into the water and then emptied, thus 
 leaving it slightly moist. Other forms of receptacles, such as marbleized or 
 porcelain-lined pans or wooden vessels are sometimes used where the eggs are 
 especially adhesive. The milt is obtained by the same process as the eggs, and 
 applied to the latter in the pan used to receive them from the fish. 
 
 The hatching processes are, generally speaking, of three classes with respect 
 to equipment, determined primarily by the specific gravity of the eggs. Heavy 
 eggs, such as those of the trouts, salmons, and the grayling, are incubated in 
 wire-bottom trays or wire baskets set in troughs of running water. The mesh 
 of the wire is of size to suit the size of egg and to permit the young fish as they 
 hatch to drop through into the trough. The troughs are usually plain open 
 boxes varying in length from 1 2 to 16 feet and in depth from 4 to 1 2 inches, to 
 suit conditions. An arbitrary width of 14 inches, inside measure, has been 
 adopted, uniformity in width being desirable for economy in interior equipment. 
 For handling large quantities of eggs the troughs are frequently provided with 
 either permanent or removable partitions to regulate the direction of the current 
 of water through the eggs. Thus they may be converted into so-called William- 
 son and Clark- Williamson types of troughs. 
 
 Semibuoyant eggs, such as those of the whitefish, pike perch, shad, yellow 
 perch, and white perch, are usually hatched in glass jars. The styles of jar are 
 in general two — closed top and open top, the McDonald Universal hatching jar 
 being of the former pattern, the Chase, McDonald open-top, and Downing being 
 typically the latter. 
 
702 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 The principles of the McDonald Universal (patented) jar are familiar." 
 By substituting foi the closed top a screw-cap rim to which has been soldered 
 a pitcher-like spout, the Universal or automatic jar may be converted into an 
 open-top jar and thus is the preferable equipment at hatcheries where both 
 closed and open-top jars are required. As an open-top jar it is operated the 
 same as the Chase and Downing jars. At hatcheries where only open-top jars 
 are used the Downing pattern is preferred. 
 
 RIVER FISHES OF THE EAST COAST. 
 SHAD. 
 
 The most important fish of the east-coast streams, the shad, is the especial 
 object of three hatcheries — at Havre de Grace, Md., on the Susquehanna River; 
 at Bryans Point, Md., on the Potomac; and at Edenton, N. C, on Albemarle 
 Sound. The steamer Fish Hawk is also equipped as a hatchery, and utilized 
 at such points as may be advantageous. 
 
 As all of the eggs for the hatcheries are obtained from market fish, the shad 
 work is primarily conservation. The exhaustive fishing at the mouths of the 
 rivers leaves, moreover, so few fish to reach the spawning grounds that the 
 fishery is now, in the northern streams, entirely dependent upon the hatcheries, 
 which are themselves interfered with by the scarcity of ripe fish. Recent legis- 
 lation in North Carolina has widely restricted fishing so that there has been a 
 notable improvement of conditions in that region, and much larger collection of 
 eggs at Edenton. 
 
 Curiously enough shad are seldom caught in ripe condition during daylight 
 until late in the afternoon. Thus the fishermen's catch of the late morning or 
 early afternoon is not available for the rescue work of the spawn taker. But 
 on the approach of evening during the spawning season of the shad, the Bureau's 
 agents may be found leaving their camps to embark preparatory to being dis- 
 tributed on the various fishing boats or to fishing shores where shad in ripe 
 spawning condition are to be had. 
 
 In the collection of shad eggs the fishermen often personally manipulate 
 the ripe fish for eggs and milt, and it is customary for the Bureau to provide all 
 such men with the usual spawntaker's equipment of pans, buckets, etc. It is 
 always necessary to employ a force of experienced men to go among the fisher- 
 men, to see that the eggs of all ripe fish are saved, fertilized, and properly cared 
 for until they reach the hatchery. It may not be amiss to say that spawntakers 
 should also be experienced boatmen, not only as a matter of safety to them- 
 selves, but because the fishermen are averse to allowing inexperienced men in 
 their boats. 
 
 "■ Manual of Fish Culture, revised edition, 1900, p. 138. Published by U. S. Bureau of Fisheries. 
 
BuL. U. S. B. F., 1908. 
 
 PI.ATE LXXVIII. 
 
 iJ^'\ 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 703 
 
 The treatment of eggs of the shad after the milt has been applied is in 
 substance as follows: 
 
 About half a pint of water from the river is dipped into the pan, which is 
 then given a slow rotary motion until the milt is thoroughly mixed with the eggs. 
 This pan with its contents is then set aside and another is used in a repetition 
 of the process. After all the stripping has been done the milt is washed from each 
 pan of eggs by dipping water from the river and pouring it off, repeating until 
 the milky color of the water disappears. The pans of eggs, with about i pint 
 of water in each, are then set aside, and after fifteen or twenty minutes it will be 
 noticed that the eggs are absorbing the water. A little more water must then 
 be added from time to time until they have fully expanded, which usually 
 requires from forty to sixty minutes, but varies somewhat with the water tem- 
 perature. When fully water-hardened, the eggs feel like shot to the touch of 
 the fingers, and now, less sensitive to concussion, are ready to be transferred 
 to the buckets in which it is customary to convey them to the hatchery. 
 
 En route to the hatchery it may be necessary every twenty or thirty min- 
 utes to replace the water on the eggs with water from the river in order to 
 maintain an even temperature, the air on very cool nights affecting the water 
 in the buckets to such degree as to injure the eggs. 
 
 The first thirty-six to forty-eight hours after arrival at the hatchery is the 
 period of greatest mortality to the eggs, and until this has passed they are kept 
 in open-top jars or in McDonald jars without tops, the water flowing through 
 them being allowed to waste over the tops of the jars. The dead eggs being 
 lighter than the live ones work toward the surface and are easily removed with 
 a siphon of ,' 2-inch rubber tubing. The good eggs are then measured (by means 
 of a device which will be described later) and their number is accredited to the 
 fisherman from whom they were obtained. They are then put up for hatching in 
 the McDonald Universal (i. e., closed-top) hatching jars, which are arranged on 
 specially constructed tables in connection with rectangular glass aquaria or 
 receiving tanks and subjected to a current of about 4 or 5 pints of water per 
 minute at 8 pounds pressure, which is sufficient slightly to elevate the eggs 
 from the bottom of the jar, thus giving the entire mass a slow revolving or boil- 
 ing motion. The number of eggs to each jar is from 85,000 to 100,000. As the 
 young fish emerge from the eggs they rise toward the surface, where they come 
 in contact with the suction outlet tube of the jar and pass through it with the 
 waste water to the collecting tanks. From this they may be removed to the 
 distributing cans by means of a ^2-inch rubber siphon. Another method of 
 removing them is to dip them from the aquaria after siphoning off a portion of 
 the water. By this method, emptying an equal number of dippers of fry into 
 each can, it is possible to quite equally distribute the total number to be 
 removed. 
 
704 BULI.ETIN OF THE BUREAU OF FISHERIES. 
 
 WHITE PERCH. 
 
 With the decadence of the valuable shad fishery there has arisen a demand 
 for the artificial propagation of the white perch {Morone americana) , and this 
 work has been extensively conducted at the mouth of the Susquehanna River 
 during the past four seasons in connection with the hatching of shad. Like 
 shad culture, the propagation of white perch is purely the conservation of eggs 
 which would otherwise go to market in the parent fish. 
 
 The spawning season in the latitude of the Bureau's station at Havre de 
 Grace, Md., is from the middle of April to the latter part of May, the time varying 
 with the character of season. The eggs are taken and fertilized by the usual 
 dry method, the work being performed ordinarily by the fishermen. Owing to 
 the adhesiveness of the eggs it is preferable to strip them into marbleized or 
 porcelain-lined pans. 
 
 As with shad and other spring spawning fishes, the white perch spawns on 
 a rising temperature, ripe fish being taken first when the temperature is about 
 47°. However, the eggs seem to produce better results when hatched in water 
 of a higher temperature. At about 60° F. they hatch in forty-eight to fifty- 
 two hours, while at 68° to 70° they hatch in twenty to twenty-four hours. 
 They are sensitive to sudden drops in temperature, and although at 58° or 60° 
 about 75 per cent of the eggs produce vigorous fry, at 50° or lower few if any 
 eggs or fry survive. In one case eggs taken at 56° and held for twenty-four 
 hours, after which the temperature dropped to 46°, finally hatched at a temper- 
 ature of 54°, producing a fair percentage of fry. Observations thus far made 
 indicate that eggs taken at mid-day or soon after mid-day produce better 
 results than those taken earlier in the morning, although the reason has not 
 been ascertained. 
 
 After being held for from six to twelve hours in jars without tops, white 
 perch eggs are incubated and hatched in McDonald automatic jars on shad 
 tables in the same manner as shad eggs. As they are especially liable to fungus, 
 a good circulation is particularly important, and it is inadvisable to carry more 
 than one-fourth capacity to a jar, or from 800,000 to 1,000,000. The eggs are 
 comparatively heavy, and for this additional reason require more water circu- 
 lation than is needed for shad, or 2 X to 3 quarts per minute. It is customary 
 to reduce the flow about one-half during the last few hours before hatching, in 
 order that fungussed masses and attached good eggs may not be carried out 
 into the receiving tank. 
 
 When undeveloped the eggs are very white and hard, and it is difficult for 
 the novice to tell the live ones from the dead. vSome that apparently are dead 
 suddenly ej'e and hatch, while, on the other hand, some apparently of good 
 quality frequently fungus and prove a total failure. By examining in a glass 
 
 \. 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 705 
 
 tube, with which a few at a time may readily be removed from the jars, it is 
 usually possible to distinguish the live eggs at any stage of development. 
 
 The eggs average 29 to the linear inch when first expelled and 28 to the 
 inch after being water hardened. They are usually measured into the jars 
 with a I -quart apothecary's graduate, and are computed at 1,600,000 to the 
 quart. 
 
 To guard against the escape of very small fry from the receiving aquarium 
 the usual waste water outlet is closed, and the water is carried off by means of 
 one or more siphons, the suction ends being provided with wire cages covered 
 with two or three thicknesses of cheese cloth. To regulate the level of the 
 water in the receiving tank the outlet ends of the siphons empty into an adjacent 
 receptacle, usually a hatching jar, the rim of which is at the desired water level. 
 
 YELLOW PERCH. 
 
 It is feasible, and under some conditions desirable, to expel and fertilize 
 yellow perch eggs artificially, but under proper conditions practically all natu- 
 rally deposited eggs are fertilized. For this reason the Bureau's work largely 
 consists in the incubation of the naturally deposited eggs obtained from fish 
 confined in crates for the purpose, and to some extent by collection of natural 
 spawn in marshes where left by receding waters. 
 
 It has been found that the eggs can be hatched in almost any kind of con- 
 tainer through which water is flowing — open troughs, open jars, shad aquaria, 
 etc. — but careful experiments indicate that the McDonald open-top and the 
 Downing jar are preferable to other kinds of equipment, the Downing jar pos- 
 sessing the advantage of greater capacity. It will hold 130,000 to 150,000 
 eggs, and when thus incubated the eggs are subjected to a water circulation of 
 about I pint per minute. 
 
 As little has been published on the methods of culture of this fish, the data 
 of the superintendent in charge of the yellow perch operations at Bryans Point, 
 Md., will be given in some detail: 
 
 Here the fish are purchased from commercial fishermen, are held in crates 
 having comparatively tight bottoms, and allowed to spawn naturally. The 
 crates are 8 feet long, 4 feet wide, and 2>^ to 3 feet deep, and are placed in the 
 mouth of a small creek tributary to the Potomac reached by tide water from 
 the river. A small number of fish are also kept in a large tank at the hatchery, 
 through which there is a constant flow of river water in connection with the 
 regular station supply. 
 
 The collecting of fish begins about the middle of March or when the water 
 temperature ranges from 34° to 37°. The water temperature in which spawn- 
 ing occurs ranges from 42° to 46°. In 1908 the first eggs were obtained March 
 
 B. B. F. 1908—45 
 
7o6 BULI^ETIN OF THE BUREAU OF FISHERIES. 
 
 21 in a water temperature of 46°. Practically all the fish had finished spawning 
 on April i, at which time the water temperature was 58°. The spawn is taken 
 daily from the crates and transferred to the hatchery in buckets. 
 
 The eggs are comparatively heavy, and are slightly adhesive when first 
 deposited. The period of incubation in a water temperature of 47° to 54° is 
 ten to twenty days. The absorption of the yolk sac requires a period of eight 
 days in a mean temperature of 54°. On the Potomac River the average number 
 of eggs from fish of one-half pound weight is about 15,000. 
 
 When measured into the jars the eggs are computed at 100,000 to the 
 quart. Actual measurements of 10 eggs average 1.662 millimeters or 0.065 
 inch in diameter, but on account of the peculiar gelatinous envelope of the 
 egg, these individual egg measurements are of no use whatever in estimating 
 the number in a given volume. 
 
 STRIPED BASS. 
 
 At Weldon, N. C, is a field station for the propagation of the striped bass 
 {Roccus lineatus). It has been operated for a number of years with rather 
 negative results as to the number of eggs collected, but with experience of much 
 value as suggesting lines of improvement in methods of manipulating ripe fish 
 and the eggs. 
 
 The chief obstacles to successful work in the propagation of this fish are 
 the difficulties of obtaining ripe spawning females in numbers sufficient to pro- 
 duce large results, and of obtaining a ripe male at the time a female is available. 
 Penning the fish has not proved successful. The delicate nature of the eggs 
 and the limited period of incubation — 1% days — are reasons for believing that 
 they can not be successfully transported from the source of supply. 
 
 The McDonald automatic hatching jars have been somewhat successfully 
 used at Weldon, but the sac fry of the striped bass are more tender than those 
 of any other fish and it was found that many were injured in their passage to 
 the aquarium through the rubber tube of the closed top. The style of jar was 
 then changed to open top, and has given highly satisfactory results. An 
 elongation of the pitcher mouth by means of a trough of canvas delivers the 
 fry from the jar to the aquarium without friction and concussion. Owing to 
 the buoyancy of the eggs less than one quart of water per minute is supplied to 
 each jar and at time of hatching only about one pint. When water hardened 
 the eggs are computed at 35,000 to the liquid quart and it is customary to 
 incubate 80,000 in each McDonald jar. 
 
 In the latitude of Weldon, striped bass spawn in a temperature of 70° to 
 77°, the season being from the middle of April to early in June. 
 
Bui,. U. vS. B. F., 190S. 
 
 Plate LXXIX. 
 
 Fig. 5. — Spawiitaking operatious oti the Detroit River, Michigan. Whitefish caught in commercial fishing are held 
 in crates and pounds until ripe. The crates are provided with false bottoms, wliich may be raised for easier access 
 to the fish. 
 
 Fig. 6.— Two men with dip nets are lifting Iisli In^ni crates (male whitcnsli m i.iie crate, fL-niaks in anntherj into the 
 tubs at the spawntaker's right. The pan of eggs is passed to the man at the extreme left, who " washes them up " 
 before they are put into the neighboring tub for water hardening and removal to the hatcher>'. (Detroit River. 
 Michigan.) 
 
FISH-CUI<TURAL PRACTICES IN THE BUREAU OF FISHERIES. 707 
 
 FISHES OF THE GREAT LAKES. 
 
 For the maintenance of the fisheries of the Great Lakes the Bureau operates 
 hatcheries at Put-in Bay, Ohio, Northville, Mich., Duluth, Minn., and Cape 
 Vincent, N. Y. At Cape Vincent the work consists largely in hatching and 
 distributing the product of eggs received from other hatcheries, Lake Ontario 
 not being a fruitful field for the collection of eggs. 
 
 WHITEFISH. 
 
 At Put-in Bay station in the fall of 1907 the collections of whitefish eggs 
 reached a total of 336,000,000, the largest on record for any station." In this 
 locality the eggs are in large part obtained by direct purchase from the fisher- 
 men, who have expelled and fertilized them as they removed the fish from the 
 nets. In addition it is customary to pen 8,000 to 10,000 fish in crates at points 
 convenient to the base of operations. These fish are obtained from fishermen 
 at a nominal price for use until after spawning, when they are returned to the 
 fishermen to be marketed. As it has not been practicable to confine the fish 
 successfully for a very long time, this procedure is not undertaken until they 
 are nearly ripe, the total period of confinement from the begiiming of the collec- 
 tion to the close being about three weeks. 
 
 The crates used for penning the fish are 16 feet long by 8 feet wide by 6 
 feet deep, and have a partition dividing each crate into two compartments 8 
 feet square by 6 feet deep. Each compartment is provided with a false bot- 
 tom, which may be raised at will while the fish are being manipulated, and can 
 be shoved down against the stationary bottom aftenvards. For convenience in 
 handling the crates are made "knock down," and thus can be easily removed 
 from the water and used year after year. Wliile in the water they are held in 
 position by floats 52 feet long, the crates placed between them endwise, five crates 
 forming a raft. Only 12 inches of the crates extends above the siu^ace of the 
 water. The rafts are held in place by stakes driven into the bottom of the bay. 
 Of the total number of eggs collected about 25 per cent are taken from fish con- 
 fined in crates. 
 
 Another important field station for the collection of whitefish eggs is in the 
 Detroit River, where operations are conducted from the Northville (Mich.) sta- 
 tion. Here the eggs are derived from fish caught for fish-cultural purposes by 
 market fishermen who operate under permits issued by the Bureau. The fisher- 
 men are willing to incur all expenses of the collections, as well as the losses from 
 penning, for the privilege of disposing of all the marketable fish after the Bureau 
 has taken the eggs. Penning stations are located accessible to the seining 
 
 o Since this writing another spawning season has passed with a record for the Put-in Bay station 
 of over 373,000,000 whitefish eggs. ' 
 
7o8 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 grounds, and all fish are transferred to the crates before being spawned. Spawn- 
 taking operations are conducted daily throughout the entire period while the 
 fish are penned. About 75 per cent of all the females confined in crates yield 
 good eggs; the remainder either cast their eggs in the crates, become plugged, 
 or fail to ripen. 
 
 At the Put-in Bay station the Downing jar, an open-top glass vessel with 
 pitcher lip, devised by the superintendent, has supplanted other forms of hatch- 
 ing jars. In each of these jars 4^^ quarts of green eggs to 5X quarts of eyed eggs 
 are subjected to a flow of 4 quarts of water per minute. The jars are arranged 
 in so-called batteries, which are described in the Manual of Fish Culture else- 
 where cited. The batteries at Put-in Bay are of the type known as single 
 batteries, each having an independent water supply. They have some new 
 features, notably, for economy in arrangement, the alternation in position of the 
 jars in vertical rows, thus making it possible to bring the troughs more closely 
 together. In the batteries at this station there are 1,692 jars and the flow of 
 water through them is so economically arranged that only 330 gallons per 
 minute is required when all are in operation. This type of battery is the one 
 commonly used at the Bureau's stations, called single battery to distinguish it 
 from the double battery at Detroit." The Detroit hatchery is equipped with 
 1,487 Chase and Downing jars, and a total of 441 gallons of water is required 
 when the entire battery is in operation. 
 
 At Cape Vincent station the battery is a single tier, but the arrangement of 
 the jars is not so compact and the tiers are not so well set up as to economy in 
 water supply, 492 jars requiring a total volume per minute of 123 gallons. At 
 Swanton station a single battery of 606 jars requires a total volume of 227 
 gallons per minute. The batteries at Detroit and Put-in Bay are constructed 
 entirely of wood; the battery at Cape Vincent was originally of wood, but as 
 the troughs began to decay galvanized iron was substituted without remov- 
 ing the original stand on which the wooden troughs rested. At Swanton the 
 supply troughs in the battery are also of galvanized iron, the first cost of 
 which is more than for wood, but taking into consideration durability of material, 
 may be considered more economical. 
 
 PIKE PERCH. 
 
 By methods similar to those pursued in the conservation of whitefish eggs, 
 pike perch eggs also are extensively collected, the most important field, as with 
 the whitefish, being within a 40-mile radius of the Put-in Bay station. Here and 
 at other points on the Great Lakes the eggs are all obtained from ripe fish as 
 caught, the penning of the pike perch by methods described for the. whitefish 
 having proved not feasible in the sheltered bays where such work might other- 
 
 <J Manual of Fish Culture, p. 117. 
 
FISH-CULTUR.\L PRACTICES IN THE BUREAU OF FISHERIES. 709 
 
 wise be possible. The method of spawntaking at Put-in Bay is in general as 
 follows : 
 
 On account of the adhesiveness of the eggs a wooden or fiber receptacle is 
 used instead of the usual tin spawning pan. A liberal quantity of milt is applied, 
 the mass is thoroughly stirred with a feather or with the bare fingers, and a 
 little water is added. Then pan and contents are lowered into a keg, containing 
 about 2 gallons of water, and the pan is carefully emptied. This process is 
 repeated until the keg is about one-third full. The eggs are now left undis- 
 turbed until spawning operations are ended — not over two or three hours ; then 
 water is added, a little at a time, until the keg is nearly full. Some of the 
 water is then poured off and more added, this process being continued until all 
 the milt is washed off and the eggs are thoroughly water hardened. 
 
 The chief precaution during the water-hardening is this constant pouring 
 off and addition of water. Care must be taken in pouring the water to have 
 it fall not directly on the eggs but against the side of the keg. During this 
 period the eggs must not be stirred, for the reason, it is said, that such motion 
 tends to rupture them. It is also important to avoid exposure of the eggs to 
 the air, since because of their adhesiveness they will form in a nearly solid mass. 
 When they are sufficiently water-hardened they may be separated by gently 
 loosening them with the bare hand. After separating them it is necessary to 
 change the water frequently until they can be transferred to hatching jars. 
 
 The fact that pike perch spawn in the spring, while fall is the spawning 
 season of the whitefish, makes it possible to utilize the same batteries and 
 other hatching equipment for both species. 
 
 In order to overcome the adhesiveness of pike perch eggs and prevent 
 their forming unwieldy masses, it was formerly considered necessary to use 
 muck or starch in the water into which the eggs are placed immediately after 
 fertilization. The use of muck has now been entirely abandoned and neither 
 starch nor muck is used at Put-in Bay. 
 
 At Swanton, Vt., the pike perch is regarded as of more value as a game 
 fish than for the maintenance of a commercial fishery. The hatchery is stocked 
 with eggs taken from fish captured purely for fish-culture purposes from the 
 waters of Missisquoi Bay and River. The conditions under which the eggs are 
 taken are more favorable than those existing on the Great Lakes, it being possi- 
 ble to pen the unripe fish for a week or more while maturing, the penning crates 
 being located in a current of water in the river. Delivery of the eggs at the 
 hatchery is also more expeditious. 
 
 In the manipulation of eggs here the methods are somewhat different from 
 those just described. After impregnation in the usual manner a little water 
 is added and the eggs are left for five to eight minutes ; then more water is added 
 and the eggs are withdrawn carefully into half a bucket of water into which two 
 
yiO BULLETIN OF THE BUREAU OF FISHERIES. 
 
 tablespoonfuls of ordinary corn starch have been thoroughly stirred. Here 
 they are allowed to remain ten minutes. The bucket is then filled with clear 
 water and the washing process begins, the water being replaced until only clear 
 water remains with the eggs. They are then stirred continuously with the 
 hands for a period of forty-five minutes, the bucket being at the beginning 
 about one-third full of clear water, to which more is added during this time. 
 At the end of the continuous stirring the worst period of adhesion is over and 
 from that time on the eggs may be stirred and fresh water added every hour 
 until they reach the hatchery. Here they are held in a tub of running water 
 overnight, then screened, measured, and placed in jars on the batteries. 
 
 It will be noted that the two methods of manipulating the eggs are quite 
 at variance. There are so many factors to be considered that it has not yet 
 been decided which procedure is best. 
 
 Pike perch eggs require more care than do the eggs of any other species 
 handled by the B ureau. When received they are usually massed together in lumps 
 and must first be separated with the bare hands and passed through a screen of 
 soft bobbinet Isefore being placed in the jars. While in the jars they require 
 constant attention and must be frequently separated, it often being necessary 
 to take down individual jars several times and pass the contents through the 
 bobbinet screen. 
 
 Although pike perch are found in ripe condition in water ranging from 38° 
 to 60°, more eggs are taken in a temperature ranging from 38° to 50° than 
 above 50°, and the higher temperature seems to be most favorable for hatching 
 the maximum number of fry. Unfortunately, however, the water of the hatch- 
 eries is usually of the colder temperature in which the fish spawn, and a high 
 percentage of fry has therefore been unattainable. An average production- in 
 fry of 50 per cent of the eggs taken may be regarded as very good. 
 
 LAKE TROUT. 
 
 The station at Northville, Mich., with its several other lines of fish culture, 
 is also the principal center of the lake trout work, and has a record of over 
 58,000,000 such eggs in one season." As the spawning season is short, however, 
 and is at a period of the year when the fishermen, on account of rough weather, 
 often can not set or take up their nets at will, the collection of eggs must neces- 
 sarily vary in quantity and quality from year to year according to weather 
 conditions. 
 
 The data for one season show the average weight per fish to have been 7.4 
 pounds; that 66 per cent of the catch of females yielded eggs; and that the eggs 
 averaged two-thirds of a fluid quart per fish. 
 
 "Since this writing another spawning season has yielded 71,000,000 eggs at the Northville station. 
 
BuL. U. vS. B. F., 1908. 
 
 Plate LXXX. 
 
 Fig. 7. — Interior of one wing ol Iiatcln.'r\- at Put-in Hay, Ohio, showing four batteries and the tanks 1 left foreground) 
 into which the fry are carried by the flow from the jars. 
 
 Fig. S. — Downing jars set up for vise on ■batten,-" 
 at Put-in Bay station. Ohio. The troughs into 
 which the water flows from the pitcher mouth of 
 the jars ser\-e also to supply the jars in the next 
 tier, by means of the wooden faucets, and there is 
 a small overflow at alternate ends of each trough. 
 This is the common equipment for hatching wlnte- 
 fish and pike perch. 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 
 
 711 
 
 The hatchery equipment for lake trout at Northville is the common form 
 of wire tray, stacked in troughs of the Clark and Clark-Williamson types." 
 The collections in this region have increased so greatly in recent years, however, 
 that they have outgrown the capacity of the hatchery, and it has been necessary 
 to deepen some of the Clark- Williamson troughs to accommodate more trays 
 during the eying period of the eggs. The deeper troughs are 15 feet long and 
 2)14. feet wide, with a division through the center the entire length; the width 
 therefore is that of a pair of troughs having a common bottom. The outside 
 depth is 18 inches. Each of the deep troughs contains, besides the bulkhead, 
 15 compartments 19 inches by 10 inches by i6J^ inches deep, with a capacity 
 of 16 trays i8>^ inches long by c)yi inches wide, on each of which may be 
 placed, if crowded, 10,000 eggs. The total maximum capacity of each pair of 
 
 Lonffifudino/ Secfion 
 
 Fig. I.— Clark-Williamson trough. 
 
 troughs is then 4,800,000. For best results, 8,000 eggs to the tray, or a total of 
 3,840,000 to each pair of troughs, is a proper number. 
 
 During incubation the eggs seem to do equally well in either up or down 
 current of the Clark- Williamson troughs ; at other stations where eggs are incu- 
 bated in stacks of trays the Williamson type of trough is used. 
 
 As lake trout eggs are taken from fish that have been caught in nets hauled 
 into fishing tugs by steam power, often during rough weather and frequently 
 after the nets have been inaccessible for several days, the percentage of good 
 eggs is not equal to that secured from most of the species manipulated. Con- 
 sequently a large force of young women, who are more deft with their fingers 
 than are men, are temporarily employed to pick over the eggs. A shallow trough 
 with water flowing through it is provided for this work; in this trough, standing 
 
 1 For full description of these troughs see Manual of Fish Culture, p. 97-99. 
 
712 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 above the water on four short legs, are wire baskets, one for each egg picker, 
 into which to throw the dead eggs. The farther side of each basket, to the 
 picker's right, has a high back to stop the eggs as thrown from the tweezers, 
 thus making it unnecessary for the eyes of the picker to follow each egg, and 
 thereby facilitating the entire operation. 
 
 THE BROOK TROUTS. CHARS. AND EASTERN SALMONS. 
 
 For brook trout eggs {Salvelinus fontinalis) the Bureau depends largely 
 upon commercial trout raisers, eyed eggs being obtained from them at lower 
 cost than it is possible to collect from wild fish at most places or from brood 
 fish maintained only for their eggs. About 8,000,000 eggs are annually pur- 
 chased from ten to eleven dealers. For the purpose of making a just com- 
 parison as to quality and final cost of fish produced from each purchased lot 
 the eggs received from each dealer are distributed to several hatcheries, that all 
 may be alike subject to different conditions of quality and temperature of the 
 water supply. 
 
 At some stations, however, eggs from wild trout are more satisfactory. It 
 has been found that eggs from the domesticated fish hatched and reared in 
 spring water which is not subject to seasonal variations do not produce good 
 results where the temperature of the water supplying the hatchery is below 35° 
 or is subject to variations of several degrees. Vermont and Colorado are the 
 only states in which eggs of the wild brook trout are collected in sufficient 
 numbers to stock the Bureau's hatcheries in those states as well as to have a 
 surplus for distribution to other hatcheries. It is interesting to note, further, 
 that in Colorado, where the eastern brook trout is an introduced species, the 
 eggs can be collected in greater numbers and at less cost than in any other state. 
 
 In Vermont" eggs are obtained from trout inhabiting artificial lakes on 
 private preserves. During September and October principally, but in some 
 localities beginning sometimes as early as July and continuing into Novem- 
 ber, the fish ascend the streams in large schools on each rise of water. The 
 fish culturist has only to provide suitable racks and traps in anticipation of the 
 period of migration, constructing them in the streams that have been dammed 
 to make the lakes. The fish are dipped from the trap into adjacent pens above 
 the rack, the pens being kept covered to guard against the escape of the fish m 
 case of a possible flood. 
 
 A field station of this character is sometimes managed by one man, who 
 constructs the trap, rack, and pens, cares for and strips the fish, and then cares 
 for the eggs, which are incubated until eyed in stacks of trays in the William- 
 son type of troughs, then are packed and shipped to the central station at St. 
 Johnsbury. 
 
 »Titcomb,*J. W.: Wild trout spawn; methods of collection and utility. Proceedings of the 
 American Fisheries Society for 1897, p. 73-86. 
 
BuL. U. S. B. F., 1908. 
 
 Plate I.XXXI. 
 
 Fig. 9.— Capturing spawning blackspotted trout in small tributary of Grand Rlcsa Lake, Colorado. 
 
 Fig. 10. — Trap fur o.tiiiiiniiL,- ■-pa\\-niii- r.tinliow trout at niuiUh <>i priii 
 
 Colorado. 
 
 ■ il triliut,i:\- 1.. I. ;kr -^t. Chn-lnbal, 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 
 
 713 
 
 The construction of the racks under the many varying conditions which are 
 wont to prevail requires good judgment and extreme care. Such barriers must 
 be made in anticipation of and providing for floods and must be fish-tight. One 
 small hole large enough for the entrance of one fish may result in the escape of 
 the entire lot. 
 
 The eying stations adjacent to these collecting stations are small inexpensive 
 structures, a shanty 12 by 16 feet being adequate to eye a million eggs. 
 
 To compensate for the eggs taken from these waters, about 25 per cent of 
 the fish produced therefrom are returned to them, this being an ample propor- 
 tion to keep them well stocked. The parent fish are always returned to the 
 waters from which they were captured. 
 
 Eggs of the landlocked salmon {Salmo sebago) are collected by methods 
 quite similar to those pursued in the brook trout work. Although the range of 
 this species has been extended, the field of egg-collecting operations is almost 
 exclusively the native habitat in Maine. 
 
 Perhaps the most extensive trout-culture operations in the world are con- 
 ducted from the station at Leadville, Colo., in connection with which are field 
 stations for the collection of eggs of wild trout of three species. The output of 
 the lycadville station for 1908 was as follows: 
 
 Species. 
 
 Eggs. 
 
 Fry. 
 
 Fingerlings. 
 yearlings, 
 and adults. 
 
 Landlocked salmon 
 
 
 
 8. 400 
 144- 000 
 210, 000 
 380. 500 
 
 
 45. 000 
 I. 404. 100 
 1.833,900 
 
 100. 000 
 
 3, 736,000 
 
 I, 905. 000 
 
 50.000 
 
 
 Brook trout 
 
 Grayling ._ . .. _. ._ _ __ 
 
 
 
 
 SPAWNTAKING IN COLORADO. 
 
 To see the methods of work in Colorado it may be well to follow the spawn- 
 takers as they leave their camp on one of the Grand Mesa Lakes for a day's 
 work with the native trout of the Rocky Mountains {Salmo darki) at Big Island 
 Lake, 10,000 feet above the sea level. 
 
 Each spawntaker is provided with a neck yoke and two lo-quart buckets 
 in which to bring in the results of the day's operations. The fish have assembled 
 in great numbers around the outlet of the lake, where as many as can be con- 
 veniently handled are caught at each haul of the seine and the ripe ones imme- 
 diately stripped. Work at this point may continue all day or it may be advisable 
 after a time to seek other spawning grounds, perhaps at the mouths of small 
 inlets where the water from melting snow is flowing into the lake. 
 
 A most interesting phenomenon in connection with this work is the run of 
 trout around the island from which the lake derives its name. Every two or 
 
714 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 three years, and possibly more often, at some period during the spawning season 
 there is a procession of fish in twos, threes, and fours around this island. They 
 follow the indentations of the shore line closely. There is no apparent break in 
 the procession, the line being visible from any view point on the shore. It 
 usually continues, moreover, for several days. A 200-foot Baird collecting seine 
 run from the shore line of the island to form an obtuse angle intercepting the 
 run for ten minutes will be full of fish. The spawntakers, standing about the 
 bag of the seine, in two or three feet of water, proceed to strip the fish in the seine 
 while the procession closes in, the line of trout winding in and out about the 
 legs of the men and apparently in as large numbers as before. 
 
 The spawntakers with their full pails of spawn proceed to a station near 
 their headquarters, where all eggs are spread on trays and the latter are stacked 
 closely in Williamson troughs supplied with water from an adjacent lake. Here 
 the eggs are eyed preparatory to shipping a portion of them to the Leadville 
 and other stations. Some are hatched at the field station to replenish the 
 waters from which they were collected, or other waters in the vicinity. 
 
 REARING METHODS. 
 
 At most stations a portion of the fry are reared to fingerlings, and at some 
 stations it has been found advisable to carry brood fish, both of brook trout 
 and rainbow {Salmo irideus). 
 
 The latter, a native of the streams on the Pacific coast, has been domesti- 
 cated and successfully propagated at stations in Missouri, Iowa, Virginia, West 
 Virginia, and Tennessee. At stations farther north whose minimum water tem- 
 perature is usually lower and subject to extreme changes, it has been cultivated 
 with varying, but on the whole rather negative, results. It has been successfully 
 acclimatized in some of the more northerly states, notably in Michigan; but it 
 does not thrive in waters subject to extremely low temperatures during the 
 winter months, and in New York and the New England States has proved in 
 most streams a failure. 
 
 The domesticated brood fish of either species are usually the product of 
 eggs collected from wild fish, and are reared in the usual manner. The young 
 fish may be confined for the first four or five months, or until 3 to 5 inches in 
 length, in the hatching troughs or in a battery of outdoor rearing troughs of 
 dimensions and in other respects quite similar to the indoor troughs, about 12 
 feet long by 14 inches wide. Care must be taken, however, to guard against 
 overcrowding as the alevins increase in growth. The actual number of young 
 fish of a given age which can be successfully carried is dependent upon the 
 quality of the water supply, temperature being an important factor, not only as 
 to the number for a given space, but also as to their rapidity in growth. At 
 the White Sulphur Springs (W. Va.) station, with a supply per trough of 10 
 
BuL. U. S. B. F., 1908. 
 
 Plate LXXXII. 
 
 -l-iuld hatchciy at Graiul :Mcsa l,akub, Colorado. Blackspotted trout eggs Uj thu iiiimbL-r uf 7,wo,ooo 
 in one season have been developed here to the eyed stage with only a normal loss. 
 
 Fig. 12. — Tray of trout eggs in hatching trough. The fry fall or work through the rectangular niesh of the tray bottom 
 into the trough, where they are visible at the left of this picture. Tray is wedged at proper depth in trough during 
 incubation, and, for convenience in removing dead eggs from time to time, may be floated by releasing the wedges. 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 715 
 
 gallons of water per minute at a temperature of about 50°, it is customary to 
 hatch and hold in each trough 50,000 sac fry, 25,000 advanced fry, 12,500 i}^- 
 inch fingerlings, 4,000 i^^-inch fingerlings, 2,000 i^-inch fingerlings, and 1,000 
 fish 2 to 3 inches in length. Much larger numbers are often carried under simi- 
 lar conditions without serious loss, though often with the result that the fish 
 prove weak in transportation. 
 
 At stations where the facilities permit, a congested condition of the hatch- 
 ing troughs is avoided by transferring some of the fish to outdoor ponds as soon 
 as they have learned to take food readily, or, if weather conditions are suitable, 
 after being fed two or three weeks. In rearing fingerlings for four to six months, 
 concrete ponds 18 to 25 feet in length by 5 to 6 feet in width and 2^4 feet deep, 
 with a fall of 8 to 10 inches in the bottom for drainage, give good results at the 
 Manchester (Iowa) station. The stock ponds at Manchester, 76 feet long, 17 
 feet wide, and 3 feet deep, supplied with 40 gallons of water per minute 
 at a temperature of 50° to 60°, have a capacity for 3,500 rainbow trout 2 years 
 of age; 1,800, 3 years old; 1,000, 4 years old, and 900, 5 years old. This 
 trough and pond system is typical for rearing any species of brook trout, as 
 well as lake trout and Atlantic and landlocked salmons, for three or four 
 months, which is as long as it is customary to hold young fish intended for dis- 
 tribution. Brood fish may be obtained by selection from the fingerlings 
 intended for distribution, which as they develop are transferred to stock ponds. 
 
 As soon as the fry swim up looking for food they are fed several times a 
 day an emulsion of finely ground liver. This diet is continued as the young fish 
 develop, with the difference that the liver is less finely ground and is given less 
 frequently — two or three times a day being sufficient when the fish have attained 
 a length of 2 or 3 inches. The kind of liver used varies at different stations, 
 that of sheep, beeves, and hogs being extensively used, and the relative value 
 of each being in the order named. The food for the larger fish consists of the 
 liver, lungs, and hearts of the animals already mentioned. 
 
 At Manchester, Iowa, it has been found advantageous from an economical 
 standpoint to mix the animal food, after it has been ground, with a mush made 
 by cooking wheat middlings or shorts, to which a moderate amount of salt is 
 usually added. After the mush has been thoroughly cooled the animal matter, 
 uncooked, is stirred into it in the following proportions: For fingerlings, i part 
 animal matter and 2 parts mush; for adults, i part animal matter and 3 parts 
 mush. Twenty gallons of boiling water and 50 pounds of wheat middlings will 
 make about 202 pounds of mush. 
 
 The provision of food for domesticated fish is one of the greatest problems 
 the fish culturist encounters, and has been the subject of considerable experi- 
 mentation. The intensive production of natural food has not received in the 
 United States so much attention as in Europe nor so much as the subject 
 
7l6 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 deserves. The American method of raising trout precludes the economic use of 
 natural live food, although unquestionably the edible qualities of the fish might 
 be much improved thereby. 
 
 For several years the roe of the branch herring {Pomolobus pseudoharengus) 
 has been used at several stations as a substitute for liver in feeding fry and 
 fingerlings. It is purchased from cannery men, who preserve in tins large quan- 
 tities of it for human consumption, and consequently are able to sell it at a price 
 close to or less than the cost of liver, due allowance being made for the waste 
 in liver and the labor involved in its preparation as a fish food. The herring 
 roe has the advantage of always being on hand in condition ready to feed and is 
 therefore especially desirable for isolated stations. It is not customary to feed 
 young fishes on this food after they are 2 or 3 months old. 
 
 At Craig Brook, Me., fly larvae," used for a number of years in rearing 
 trout and salmon to fingerlings, proved to be a very satisfactory food and the 
 fish attained a more rapid growth than when fed on liver and other dead material. 
 From a financial standpoint it was not as economical as freshly prepared liver 
 or other animal foods, and this, coupled with the objectionable odor attendant 
 on its preparation, had much to do with its discontinuance. 
 
 It is possible, however, to utilize animal refuse advantageously for the pro- 
 duction of fly larvae by means of a contrivance in which the material on which 
 flies have deposited their eggs may be suspended over the water. This con- 
 trivance consists of a wooden frame, like a box without top or bottom, placed 
 on floats and having an air-tight cover to prevent the escape of foul odors. 
 Within the frame are two trays, the bottoms of which are made of coarse 
 wire cloth (odds and ends of old hatching trays). Excelsior or straw is placed 
 on the trays, and waste meat or other animal refuse is placed on top of this 
 material. As the larvae hatch they work through the excelsior, cleaning them- 
 selves thereby, and drop into the water. Two small trays are preferable to one 
 of larger size, that the meat may alternately be renewed, thus insuring a more 
 constant supply of larvse. The noxious animals killed in the protection of fish 
 may advantageously be disposed of in these trays. 
 
 SPECIAL DEVICES APPLIED AT TROUT STATIONS. 
 
 Study and experience in recent years have revealed abnormal aeration as 
 a condition existing in various water supplies at trout-culture stations, with 
 consequent mortality among the fish. Of the methods used to correct such 
 abnormalities, that devised by the superintendent of the station at White 
 Sulphur Springs, W. Va., Mr. R. K. Robinson, has proved the most efficient. 
 
 » Atkins, Chas. G. : The live food problem, American Fisheries Society, 1903; also Food for young 
 salmonoids, Proceedings Fourth International Fishery Congress, Bulletin Bureau of Fisheries, vol. 
 xxviii, 1908, p. 839-851. 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 717 
 
 It consists of a series of ordinary milk pans held in a frame, one above another, 
 in numbers to suit conditions, the bottoms of the pans perforated with a nail or 
 other pointed instrument which will leave a ragged edge to each perforation on 
 the underside of the pan. 
 
 This apparatus is set up at the head of the trough in such position that the 
 supply pipe empties into the topmost pan, and the water must pass through 
 the series before reaching the trough. By this separation into fine streams the 
 water is thoroughly exposed to the air, thus rectifying any abnormality of air 
 content. 
 
 At nearly all trout hatcheries it has been customary to place horizontally 
 below the supply pipe at the head of each trough a screen consisting of a light 
 frame, bottomed with wire cloth or perforated metal. This is designed not only 
 to break the force of the stream entering the trough but to aerate or deaerate 
 the water and at the same time catch foreign substances and animal life — the 
 latter at times being quite objectionable. Such screens, however, have almost 
 invariably caused the water to spatter over the sides of the trough, resulting in 
 constantly wet surroundings. To overcome this objectionable feature a conical 
 perforated screen has been devised by Mr. M. E- Merrill, of the St. Johnsbury, 
 Vt., station. When the screen is in place the current of water falls directly 
 on the apex of the cone, and thus is spread over the entire perforated surface, 
 accomplishing the objects of all other styles of head screens, and avoiding the 
 spattering of water over the sides. 
 
 A device for assorting young salmon and trout was introduced in the 
 Bureau's operations by Mr. J. P. Snyder, an employee at one of the stations. 
 It consists of a series of screens by means of which to separate the fingerlings 
 into sizes. 
 
 In length the screens are slightly less than the width of the troughs, to 
 facilitate sliding them along. There should be two screens for each size of 
 mesh. In use one screen is carefully inserted at the foot of the trough close to 
 the tail screen, due precaution being taken that no fish are pinched and that 
 none are left between the foot screens and the end of the trough; midway of 
 the trough a screen should be securely fastened in a vertical position by wedging. 
 After the first screen is in position a similar one is inserted at the head of the 
 trough and then moved along toward the center. 
 
 As the two screens are brought closer together the fish between them 
 become frightened, and all that are small enough escape through the mesh of 
 the screens. The distance of the second screen from the first should depend 
 upon the number and size of the fish in the troughs, and also upon the number 
 that escape through the screen. The second screen should also be fastened by 
 wedging. Then the hand of the attendant is moved about among the fish 
 between the screens to guard against any small ones finding a hiding place. In 
 
71 8 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 their efforts to escape some of the fish will be hung in the wire cloth, but it will 
 be noticed that every trout which gets its head through the screen can pass or 
 be assisted through without injury. The few which are caught in the mesh 
 should be assisted by grasping the tail and pushing them. It would be well 
 to refrain from feeding for about twenty-four hours before assorting the fish. 
 
 By using two or three sets of screens in different troughs at the same time 
 one man can assort many thousands of fish in a day, and the sizes will be much 
 more uniform than when assorted with a scaff net. Wire cloth, 6 bars to the 
 inch each way, painted with asphaltum varnish, will permit all brook trout 
 under i inch in length to pass through. By varying the mesh of screens brook 
 trout may be assorted into six uniform sizes as follows: 
 
 Number of bars to the inch : Size of fish. 
 
 6 All under i inch. 
 
 5 All between i inch and i J^ inches. 
 
 "■ 4 All between I J-2 and 2 inches. 
 
 3 All between 2 and 2 H inches. 
 
 2/5 All between 2}'s and 3 inches. 
 
 2 All between 3 and 31$ inches. 
 
 The frames of these screens are made of half-inch wooden strips grooved 
 and tongued at the ends. These frames are one-eighth inch less in length than 
 the inside width of the troughs and in height equal the depth of the troughs, 
 being rectangular in form. They are covered on one side with wire cloth held 
 firmly by copper tacks, both the wire cloth and the frames being painted with 
 asphaltum varnish previous to tacking the wire on the frames. This not only 
 helps to preserve the wood and keep the wire from rusting, but smooths the 
 latter so that there are no rough surfaces or projections to injure the fish as 
 they work their way through. 
 
 ATLANTIC SALMON. 
 
 Another important branch of fish culture is conducted at the Craig Brook 
 station, near the Penobscot River, not far from Bucksport, Me. While not 
 restricted in its work to this one species of fish, the primary object of this hatch- 
 ery is the propagation of the Atlantic salmon. The decadence of this important 
 fishery on the North Atlantic coast, due to the ruthless but natural progress 
 of civilization, is too well understood to call for an explanation here. Suffice 
 it to say that to-day the Bureau is maintaining a commercial fishery for the 
 Atlantic salmon on the Penobscot River purely by artificial propagation. It is 
 the only river in the United States where this once abundant salmon is now 
 found in sufficient numbers to support a fishery or to warrant its artificial 
 culture, and here, with the natural conditions so changed, it is with no little 
 difficulty that the extinction of the species is prevented. 
 
BuL. U. S. B. F., 1908. 
 
 Plate LXXXIII. 
 
 
 1 
 
 1 1 
 
 pPi 
 
 Si 
 
 K« 
 
 P"^^ / 
 
 ^^J^ 
 
 a^Z^^I^S 
 
 ■ ,_ \_ ..:^^ 
 
 ""^ll^ 
 
 ^^^^^ 
 
 
 ^^^^ r ^^^1 
 
 K;^ 
 
 ^Bt^3E^^^^^^C^^^^^^^^^gr~"_ niiT." 
 
 3t 
 
 ^■■■fei 
 
 ^^^^^^ 
 
 
 
 ^HHi^H^k^ 
 
 Fig. 13— Series of covered trout-hatching troughs at White Sulphur Springs station (West Virginia) 
 equippt.(i with thr R..biiisnn pan aerator. (See p. 716.) 
 
 Fig. 14. — Trout-hatching tninyhs wiih Merrill aerating coni-, (Suf p. 717.) shows also tray basket <'t a I'.ittern 
 used at some trout stations, much deeper than ordinary tray and intended to hold fry after hatching. l^eft 
 trough with tray removed to show shower of water produced by aerating cone. 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 7 19 
 
 The operations at this hatchery, fully described in the Manual of Fish 
 Culture published Ijy the Bureau of Fisheries in 1900, have undergone slight 
 change of method and need not be dwelt upon here. The source of egg supply 
 is the catch of the fishermen's weirs usually during the month of June, the fish 
 being purchased and towed in live cars to the station, where they are trans- 
 ferred to inclosures and there retained until the spawning season, in October 
 and November. When ripe they are stripped and the eggs placed upon wire 
 trays, which are stacked in troughs and carefully tended until the early spring, 
 when the eggs hatch. The young fish are distributed for the most part as fry, 
 but a considerable number are reared to the fingerling stage. 
 
 POND CULTURE. 
 
 Pond culture in the United States is applied only to nest-building fishes, 
 such as the basses, sunfishes, and the common catfish {Ameiurus nebulosus). 
 These species do not submit to manipulation for taking and fertilizing their eggs, 
 but fortunately a very large percentage of the eggs are fertilized when the 
 spawning functions are permitted to occur naturally, and the parent fish care 
 for and protect the young until the latter are free swimmers. The cultivation 
 of these fishes, therefore, consists in providing ponds which shall give to the 
 maximum number of breeding fish and their young all the essential conditions 
 of a natural environment, while at the same time protecting them from their 
 enemies and holding them under control. 
 
 THE PONDS. 
 
 Economy in construction usually dictates the shape and area of the ponds, 
 but an independent water supply and drainage to each is desirable. For con- 
 venience of the fish culturist the area usually ranges from one-fourth to one acre, 
 although some ponds of larger size are desirable. It was formerly considered 
 essential to have at least one-fourth the area of the breeding pond not exceeding 
 I foot in depth, but it has been found that the deepening of the shallower por- 
 tions to a minimum depth of from i foot to i yi feet has largely increased the 
 productive area. 
 
 The presence of aquatic plants in fish ponds is a prime essential. The 
 young of the nest-building fishes do not accept artificial food, and must there- 
 fore have their natural diet of minute animal life, the abundance of which is 
 dependent to a large extent upon the character and abundance of plant growth. 
 Plants are also important as oxygenators of the water and afford shelter and 
 shade for the fish. The selection and control of aquatic vegetation, therefore, 
 is a matter to which the fish culturist must give much attention, and experience 
 at the various stations indicates that it oft'ers a direct means by which the output 
 
720 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 of the ponds may be increased. The subject has not been sufficiently studied, 
 but observations so far made suggest various practical possibilities of much 
 interest." 
 
 The supposed loss of young fish by the voracity of their parents induced 
 the practice of partitioning the ponds in such a manner as to confine the adults 
 in one portion while permitting the young to escape through the partitions to 
 safety. It has been found, however, that the loss from cannibalism is due 
 chiefly to the young fish themselves, and accordingly they are separated from 
 their parents or not, merely as a matter of convenience. The principal pre- 
 caution against cannibalism is, instead, the provision of an abundant food 
 supply, to divert the fish from each other. 
 
 FOOD FOR THE ADULT FISHES. 
 
 Food for the adult fishes is largely a matter of local conditions and con- 
 venience. Chopped fish is extensively used at some stations, and crawfish, so 
 abundant in some localities, when chopped make admirable food for the adult 
 stock. The basses, although not appearing to care for pollywogs as naturally 
 present in the ponds, will devour frog tadpoles voraciously if the latter are seined 
 out and thrown back by the fish culturist, but they absolutely refuse toad 
 pollywogs when similarly served to them. Minnows are a good food, but should 
 not be introduced into the ponds near the spawning season, as they eat not 
 only the small forms of life upon which the fry depend but often eat the fry as 
 well. Dead minnows thrown into the water one at a time are greedily taken 
 by the adult basses. 
 
 Adult bass may also be advantageously fed on strips of beef liver about 2 
 to 3 inches long and from one-half to one-fourth inch in width or thickness; 
 and prepared food consisting of ground liver or other animal substance mixed 
 with a mush of cooked shorts, com meal, or middlings has been employed in a 
 rather limited way. It is worthy of note that for this prepared food to be 
 attractive to bass it must ordinarily contain at least two-thirds of the animal 
 substance, whereas prepared food containing only 10 per cent of the animal 
 material is taken with avidity by trout. 
 
 It has been quite conclusively demonstrated that one of the principal 
 causes of loss among brood fish is overfeeding, resulting in a fatty degeneration. 
 This loss has been largely overcome by reducing the food supply and at the 
 same time varying the kind of food furnished. 
 
 ARTIFICIAL NESTS. 
 
 In the cultivation of the small-mouth bass, and to some extent the other 
 species, it has been found profitable to provide artificial nests. These are of 
 
 " Titcomb, J. W. : Aquatic plants in pond culture. Bureau of Fisheries Document 643. 1909. 
 
BuL. U. vS. B. F., 1 90S. 
 
 Plate LXXXIV, 
 
 ^^.. 
 
 
 ''I 
 
 > " ^m^ 
 
 Fig. 15. — Placing cheese-cloth retainers over nests containing bass fn.- about ready to -^wiiu up. Most of the nests are 
 seen covered with retainers. When a nest is occupied it is numbered and complete record kept of its different 
 stages. The retainers are to confine the fry for convenience in transferring to other ponds to make room for 
 following broods. (Mammoth Spring station. Arkansas.) 
 
 JO. — Pond drawn down und ne>t buxes phiLt. i 1-: 
 150 breeding smallmouth black ba: 
 
 .', i;:il_;. .Vli-ra of pond, o,,.. . .-,....:;. i.i 
 
 (^Mammoth Spring station, Arkansas. j 
 
 i.:5L? Ill pond; 
 
r 
 
 I 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 72 1 
 
 various design, but embody in general some sort of a container for the gravel 
 the fish require, together with a shield or screen on two or three sides." The 
 primary use of screens was for the purpose of shielding the fish from view of 
 passers-by, but the practice resulted in the discovery that the fish will accept 
 shielded nests at more frequent intervals than when visible to each other. It 
 is therefore of first importance that in the placing of nests the screens be arranged 
 to meet this condition. 
 
 Wlien shielded artificial nests are not provided it is customary to deposit 
 here and there in the ponds mounds of coarse gravel, 18 inches to 2 feet in 
 diameter and about 6 inches in height, that the breeders may select and prepare 
 their nests with these. 
 
 Large- mouth black bass, though they sometimes accept gravel mounds as 
 nests, naturally seek a weedy bottom, devoid of gravel. Peat-like sods or 
 similar substances put into the pond prove acceptable to this species as nesting 
 material. 
 
 The superintendent of the Cold Springs, Ga., station endeavors to imitate 
 nature by providing "homes" for all adult fishes whether spawning or not. 
 For the large-mouth black bass and rock bass, boards 3 feet to 10 feet in length 
 are laid flat under the water so that by the natural contour of the bottom spaces 
 5 inches to 8 inches deep are formed under the boards. Catfish {Amciurus nebu- 
 losus), which prefer to dig their own nests, are provided with boards either laid 
 flat on the pond bottom or where there will be under the center of the board a de- 
 pression an inch or two in depth. Where necessary, the boards are fastened by 
 stakes at both ends, but when placed along the bank, where conditions are favor- 
 able for such a course, one end of the board may be driven a short distance into 
 the embankment, while the other end is staked. After having been submerged 
 a month or so the boards will remain in place without fastening. With proper 
 precautions against projections these shelters do not materially interfere with 
 seining operations. 
 
 For the crappie, roily water seems to be essential during the spawning 
 season. At stations where there are no naturally roily ponds it has sometimes 
 been found desirable to introduce a few carp, which roil the water in rooting 
 around the bottom of the pond and do not seem to disturb the crappie. 
 
 NUMBER OF BROOD FISH. 
 
 The desirability of maintaining a maximum number of brood fish to a given 
 pond area has led to a comparison of experience at the different stations in an 
 effort to arrive at some approximate average for a working basis. Conclusive 
 
 "One form of nest in use is the design of A. E. Fuller, described in his paper entitled "New and 
 improved devices for fish culturists," Proceedings Fourth International Fishery Congress, Bulletin 
 U. S. Bureau of Fisheries, vol. xxviii, igo8, p. 991-1000, pi. civ-cvi. 
 
 B. B. F. 1908—46 
 
722 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 determination can not be made, of course, owing to the various factors of 
 quality and temperature of water, abundance of vegetation and of natural 
 food, etc., but reports from the several localities are of interest. 
 
 At San Marcos, Tex., the best results are obtained with 24 to 30 large- 
 mouth bass to the half acre; of the smaller fishes — bream, rock bass, crappie — 
 60 to the half acre. 
 
 A i-acre pond at Mammoth Spring, Ark., supports 100 small-mouth black 
 bass, and an average of 2,000 fry to each productive nest has been obtained, 
 the maximum number from one nest being 5,200. 
 
 At Cold Springs, Ga., 100 adult large-mouth black bass in a pond of three- 
 fifths of an acre in area proved too many, and the number had to be reduced 
 to 60 or 70 for satisfactory results. In general, 50 to 75 brood bass to three- 
 fourths of an acre to an acre have been found the best number at this station. 
 Of catfish {Ameiurus nebulosus niannoratus) , 100 to the acre have l)een found 
 satisfactory. 
 
 Ponds at Wytheville, Va., accommodate 75 pairs of large-mouth black 
 bass to the acre with good results; of rock bass, 300 fish to the acre. 
 
 At Northville, Mich., in a pond three-fifths of an acre in area it has been 
 observed that most satisfactory results were obtained with small-mouth bass 
 to the number of 29 females and 23 males, allowance being made for the occa- 
 sional polygamous tendency of the male. 
 
 At White Sulphur Springs, W. Va., 36 pairs of small-mouth black bass is 
 considered the number for a i-acre pond. 
 
 The maintenance of an abnormally large number of brood fish to a given 
 area results in more or less loss according to species, the mortality among the 
 brood fish of the small-mouth black bass being greater than with the large-mouth 
 bass and other pond fishes. The replenishing of the stock is most advanta- 
 geously accomplished by securing wild fish, preferably in the spring of the 
 year, and they may be advantageously transferred up to within two or three 
 weeks of the spawning season. 
 
 COLLECTING THE YOUNG FISH. 
 
 It is often desirable to remove the surplus fry from a pond before they have 
 left their nests, and there is now in use for this purpose a combination fry trap 
 and retainer which is placed over the nests, taking advantage of the fact that 
 the fry rise vertically." This trap has proved of practical value in fish-cultural 
 ponds for small-mouth bass, and, where the nest area was not too great, for 
 large-mouth also. It has been used likewise in collecting small-mouth bass 
 fry from natural lakes, and is believed to be applicable to fry of other nest- 
 building fishes than the basses. 
 
 "See Fuller, op. cit. 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 723 
 
 Soon after the yolk sac has been absorbed, or after the fry have been feeding 
 for two or three weeks, a portion of them are removed from the ponds and dis- 
 tributed to the waters they are to stock. The first crop may often be obtained 
 by seining around the edges of the pond without the prehminary clearing away 
 of vegetation, and for this purpose a novel casting seine has come into use at 
 Northville, Mich. The web is rigged upon two long bamboo poles, so that the 
 device may be operated entirely from shore, without roiling the water or unduly 
 disturbing the fish." 
 
 After the young fish have sought the deeper portions of the ponds, pre- 
 liminary to drawing off' the water to effect their capture, it is necessary to remove 
 the aquatic vegetation, a process of much labor and expense, consisting in 
 general of mowing under water, and carrying away the foliage by means of 
 pitchforks and boats. Various methods and devices for this purpose have been 
 evolved at the different stations, as described elsewhere.'' 
 
 Ordinarily the assorting of young pond fishes by size is accomplished by 
 hand manipulation with a scaff net. To some extent, however, the separation 
 may be accomplished by regulating the size of the mesh in the nets used to 
 effect their capture. The superintendent of the San Marcos, Tex., station 
 suggests having an ample bag to the dip net in which quite a large lot of fish 
 may be taken from the tub or other retainer, then passing the net gently to 
 and fro in the water to allow the fry and smaller fish to escape, while the larger 
 ones are retained. This method is principally used for assorting black bass, as 
 it frequently happens there are schools of both fry and fingerlings in the ponds 
 at the same time. Nets of one-fourth inch square mesh will permit the escape 
 of all fish up to I j/2 inches ; one-half inch square mesh will permit the escape of 
 all fish tinder 3 inches in length. 
 
 At the Cold Springs, Ga., station the superintendent uses a box 3 or 4 feet 
 long by I foot wide and i foot deep, and water-tight to a depth of 2 inches. 
 Above this 2 inches one side is covered with wire cloth instead of being closed in, 
 the size of wire mesh being regulated for known sizes of fish. The box is par- 
 tially submerged in the pond in which it is intended to place the smaller fish of 
 a lot to be assorted. The young fish as caught are placed in the box, and then are 
 left undisturbed for an hour or two. At the end of this time the smaller fish 
 will have escaped from the box through the screen into the pond, when the 
 box with the larger fish remaining in it may be transferred to another pond and 
 emptied, or the contents may be poured into a suitable receptacle for transpor- 
 tation by tipping the box toward its solid side. Square-meshed galvanized 
 cloth is used for the screen, and if the fish are given plenty of time to separate 
 none of them are gilled, hung, or otherwise injured. 
 
 o Fuller, op. cit. 
 
 6 Titcomb, J. W. : Aquatic plants in pond culture, Bureau of Fisheries Document 643. 1909. 
 
724 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 RESCUE OF FISHES FROM OVERFLOWED LANDS. 
 
 In the upper Mississippi and Illinois rivers there is an annual spring flood 
 period caused by the melting of the snow in the northern forests and freshets 
 in the local tributaries after heavy rains. The period begins with the approach 
 of warm weather, usually about March 15, and continues until about June i, 
 when the crest of the high water has been reached. Soon after this date the 
 water begins slowly to recede, and usually by July 1 5 the river has reached its 
 normal stage. 
 
 Between the extreme low and high water marks there is a variation of 1 2 
 or 1 5 feet. There is, of course, a variation in the extremes of the water level in 
 different seasons, but seldom, if ever, does'the water fail to rise high enough to 
 flood the lowlands. The adult fishes are thus permitted to enter the overflow 
 basins and bayous, and invariably do so during the spawning season. After 
 spawning most of the adult fish escape to the river before the water has receded 
 sufficiently to cause them to be hemmed in, but immense numbers of theif 
 progeny are left in the lakes and bayous where they were hatched. These waters 
 gradually dry up, become choked with vegetation, or overheated and unfit for 
 fish life; some of the larger and deeper lakes and bayous, although cut off from 
 the main river, may contain water the year around, but on account of the 
 seepage and evaporation during the summer the depth of water in them de- 
 creases to such an extent that they freeze solid during the winter months. 
 Sometimes the lakes from which fishes are rescued are in the hollows on farm 
 lands, where in dry seasons crops are cultivated. Thus it will be seen that the 
 fish imprisoned in overflow waters are doomed to destruction in one way or 
 another. 
 
 One branch of the Bureau's operations is annually to rescue large numbers of 
 these fishes. At present the work is confined to waters convenient of access — 
 namely, the overflow lakes and bayous on the low islands in the rivers and on 
 the adjacent mainland. Many of the fishes are returned to the rivers. Another 
 portion of the more desirable species is distributed in various other waters, 
 often far from the source of supply. 
 
 It has been found, however, that the fishes rescued from these warm waters 
 do not bear transportation long distances without heavy losses if immediately 
 started upon their journey. Therefore a hardening process is resorted to, which 
 consists in holding the fish in large tanks flowing through which are streams of 
 clear cool water. To facilitate the work the Bureau has a number of field 
 stations — one on the Illinois River and three on the Mississippi — convenient of 
 access to the railroad, and each equipped for holding one or more carloads of 
 fishes for several days, or until they have become sufficiently hardened to bear 
 transportation bv cars. Adjunct to these stations are vessels, launches, and 
 boats of various types suited to the work. 
 
BuL. U. S. B. F., 1908. 
 
 Plate LXXXV. 
 
 Fig. 17. — Both trout and bass are cultivated at many of the stations. This view of the station at Manchester, Iowa, 
 shows stock ponds in foreground, then the smaller nursery ponds, all of these for trout and built of cement. 
 Beyond, in front of the hatchery building, is a bass pond, with earth bottom and sides. 
 
 KiG. iS. — Preparing a shipment of sniallmouth bass. Tanks used for hardening the young fish prior to transportation. 
 The fingerlings are held for 12 hours in cold s])ring water, then dipped into tubs, counted into pails, and transferred 
 to the transportation cans. These are then placed under a -^-inch stream of spring water and held until train time. 
 For an early morning shipment the fish are "canned" the evening before. (Mammoth Spring station, Arkansas.) 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 
 
 725 
 
 By this means not only is a conservation effected but the Bureau is enabled 
 
 to meet the great demand of applicants for the basses, sunfishes, and perches at 
 
 a less sum than it would cost to produce them at a station maintained especially 
 
 for their propagation. 
 
 THE PACIFIC SALMONS. 
 
 On the Pacific coast the Bureau has six permanent stations, including two in 
 Alaska, all of them maintained primarily for the propagation of the Pacific 
 salmons. Subsidiary to these there are six important field stations and other 
 smaller ones where salmon eggs are collected and hatched. An idea of the 
 extent of the work may be obtained from a statement of the output of these 
 stations during the fiscal year 1908. 
 
 Output of the Pacific Salmons in igo8. 
 
 Stations. 
 
 Species. 
 
 Eggs. 
 
 Fry and 
 fingerlings. 
 
 Alaska 
 
 
 
 61.369. 000 
 4.780.855 
 
 
 Chinook - 
 
 64. 990. 550 
 3.S30.00O 
 
 
 
 
 Silver _ 
 
 215. 932 
 
 498. 309 
 
 13. 262. 714 
 
 8 5M.30S 
 
 6, 764. 762 
 
 
 Chinook 
 
 
 
 Silver 
 
 296. 000 
 
 75 000 
 
 502. 000 
 
 
 
 
 
 
 
 The eggs shown in this table were transferred to state fish hatcheries and 
 other places for incubation. In California, as will be noted, a very large propor- 
 tion of the eggs taken are so distributed. 
 
 Observations at various field stations indicate that a large percentage of 
 salmon eggs deposited naturally are fertilized but for various reasons only a 
 small percentage hatch. Modern fish-culture methods permit of a much higher 
 percentage of impregnation than under natural conditions, it being possible to 
 actually hatch and distribute as fry more than 95 per cent of all eggs collected. 
 So long, therefore, as a proper number of salmon are permitted to escape the 
 various fishing devices in their ascent to the natural spawning grounds, and it is 
 possible to capture them for the purpose of obtaining and impregnating their 
 eggs, perpetuation of the salmon fishery is assured. 
 
 In the culture of the Pacific salmon it is impossible to save eggs from the 
 commercial catch, because the latter is made before the fish are ripe, and to 
 retain them until ripe is not feasible. By the time they have ascended the 
 rivers to the spawning grounds and are in condition for the fish culturist the 
 flesh has so deteriorated in quality that they are unfit for market in any form. 
 The Bureau must therefore itself capture the fish it requires, and this is usually 
 done by the construction of barricades to intercept the run at the most suitable 
 point below the spawning grounds. 
 
726 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 BARRICADES AND TRAPS TO INTERCEPT SPAWNING RUNS. 
 
 That successful work at the salmon stations depends largely upon stable 
 and suitable barricades, or racks, as they are more commonly called, may be 
 instanced by the results of the work at Battle Creek, Cal., in 1903 and 1904. 
 At the height of the season in 1903 a freshet carried away several sections 
 of the rack grating, permitting the fish to escape upstream." As a consequence 
 only about 27,000,000 eggs were secured. The following season no flood occurred 
 at Battle Creek until near the end of the spawning season, and the collections 
 that year numbered over 57,000,000 eggs. 
 
 As all of the streams are subject to freshets, the water in some instances 
 rising over 20 feet, the racks must be firmly built, and their successful operation 
 depends not only upon ingenuity in construction but the care that is taken to 
 guard against their becoming clogged with leaves and other debris in times of 
 flood — work which at times is exceedingly hazardous. Methods in the construc- 
 tion of racks vary with local conditions, as do also the methods of capturing the 
 salmon thus intercepted. 
 
 At the Mill Creek station, in lieu of a main or upper rack, the Bureau is able 
 to take advantage of a mill dam 12 feet high, which effectively stops the passage 
 of salmon. Half a mile below this dam a retaining rack with the usual traps 
 prevents the fish from dropping down stream. Seining and spawning operations 
 are conducted on the streams between the dam and the retaining rack. 
 
 At Baird, Cal. — At the Baird station on the McCloud River in California 
 are two racks or barriers between which is formed a pool 400 feet in length. The 
 upper rack intercepts the further passage of salmon, and the lower or retaining 
 rack gives the fish free entrance to the pool, but effectually prevents their return. 
 The upper rack reaches across the river, a distance of 250 feet, and is primarily 
 supported by 10 concrete piers averaging 8 feet in height and extending 5 feet 
 above low-water mark. The piers are prdperly fastened to the bed rock of the 
 river bottom by means of heavy iron bolts. They have a flat top 4 feet wide 
 and 6 feet long, and from top to bottom is a beveled nose extending upstream 
 at an angle of 60 degrees, making them 4 feet wide and 10 feet long at the bottom. 
 On either bank a small crib pier filled with rock supports the shore ends of two 
 10 by 10 inch stringers laid parallel from shore to shore across the tops of the 
 piers. A 2 -foot walk is built between the stringers and the whole is securely 
 wired to eyebolts built in the pier tops. 
 
 Across the river bottom, against the nose of the piers, is a lo-inch sill. At 
 intervals of 3 feet poles 4 inches in diameter extend at an angle of 60° from the 
 sill at the bottom to the stringer at the top, and are securely fastened to the 
 
 2 At Battle Creek the low-water mark is lo feet below the top of the stringers on the rack, and 
 during a recent flood the water was 12 feet above the top, making a 22-foot rise. 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 727 
 
 latter by large spikes. Against these poles or inclined uprights rest the gratings 
 of the rack, which for the sake of convenience in handling are built in sections 
 6 feet wide and from 6 to 10 feet long to suit the varying depth of water. The 
 gratings are made of i X '^Y 3H inch slats of dressed lumber set i }4 inches apart, 
 their thin edge facing the current, the edge being convex to facilitate cleaning, 
 and permit the passage of leaves. The ends of the gratings are nailed between 
 two pieces of i^4 by 4 inch material, notched into the slats to make a flush 
 surface. The space between the slats is gaged by nailing on i X by 4 inch blocks 
 to each end. The longer gratings are braced with two strips i>^ by 4 inches 
 nailed on 3 feet from the bottom. 
 
 In the upper rack is placed a trap 10 feet square with vertical slat sides 
 similar to the rack gratings and having a solid board bottom. The narrow 
 opening which allows the fish to enter is so constructed as to reduce to a mini- 
 mum their chance of escape. The trap is primarily used for observing the 
 general condition of the fish in the pool prior to the beginning of seining or 
 spawning operations. 
 
 The retaining rack is at the lower end of the pool, where the stream narrows 
 to about 190 feet. It is supported on 6 stone-ballasted crib piers with sides 14 
 feet long, made by spiking together logs 8 to 12 inches in diameter until the 
 required height is reached. The piers are built on shore, floated into place, and 
 filled with rock. Across the upstream end of each pier are two 10 by 10 inch 
 stringers laid parallel and supporting a board walk, as in the upper rack. 
 Two small temporary piers are also built, to support the shore ends of the rack. 
 Gratings having 2-inch interstices are placed across the stream, similar to those 
 in the upper rack, with the exception that 5 openings 2 feet wide are left between 
 the piers nearest the center of the stream. These openings are covered by the 
 usual traps, which extend upstream into the pool g^i feet. The traps are 4 feet 
 in height and 6 feet in width at the entrance, being shaped to fit the slant of the 
 gratings. The sides are of i 'X by 4 inch material spaced 2 inches apart, and 
 with the broad edge toward the current. Braces are placed across the top, and 
 at the apex of the trap is an opening 3 inches in width from the surface of the 
 water to the bottom. The salmon pass into the pool through this opening and 
 rarely, if ever, find their way out. 
 
 Before the installation of the retaining rack, some ten years ago, many eggs 
 could not be collected by reason of the loss of fish from their running back down- 
 stream. This violation of the natural instinct of salmon to work ever upstream 
 was due to fright resulting from the continual sweeping of the seine just below 
 the upper rack. In the early days Indians were engaged to walk on either shore 
 for a mile or so below the rack and beat the water with brush in an endeavor to 
 drive the fish up to the seining ground. Since the installation of the retaining 
 rack such measures have been entirely unnecessary. 
 
728 BULLETIN OF THE BUREAU OF FISHEiaES. 
 
 At Battle Creek, Cal. — The main or 'upper rack at Battle Creek, Cal., is 
 constructed on a comparatively soft and shifty river bottom, and is supported 
 by piling, instead of by the log cribs anchored with rocks, more generally used. 
 There are 12 bents of piling, each bent consisting of 3 piles driven firmly and 
 braced with heavy timbers. The 3 piles comprising a bent are driven parallel 
 with the current, the front one standing some 2 feet above low-water mark and 
 the others about 8 feet above. The front and rear piles are placed about 10 
 feet apart. On these bents of piling and reaching across the stream are placed 
 three 12 by 12 inch stringers, against which are secured 4 by 4 inch slanting 
 supports, about 6 feet apart, the lower ends of which rest on a mud sill placed 
 in the bottom of the stream. Stringers and these supports are so placed that 
 the face of the rack will meet the current of the stream at an angle of about 60°. 
 The gratings of the rack are built in sections of varying length but of a uniform 
 width of 5^ feet. The slats for these gratings are of dressed lumber i by 3 
 inches, the sides set parallel with the current, the upstream edge convex. At 
 either side of the stream, in the shallower water, single sections of gratings 
 about ID feet long extend from the bottom to the top stringer. In the deeper 
 portions of the stream two 6 or 8 foot sections of the gratings are used, one above 
 the other, with an opening between the upper and lower sections for convenience 
 during the lower stages of water in the removal, with rakes and hooks, of 
 rubbish and trash drifting downstream. When the water rises the closing of 
 this aperture is easily accomplished by knocking out the blocking between the 
 two, thus permitting the upper rack to slide down flush with the upper edge of 
 the lower section. The length of the rack from shore to shore is about 300 feet, 
 its vertical height above low-water mark being about 8 feet. During low water 
 the front is submerged to a depth of from 18 inches to 2 feet, but there are holes 
 considerably deeper behind the rack. A walk 2 feet wide is built on the top of 
 the rack. A half mile below the barrier is a retaining rack quite similar to the 
 one described for the Baird station. 
 
 At Battle Creek the racks are usually installed during September in time 
 to intercept the fall run of salmon, and unless carried out by high water, remain 
 until the close of the work in December. Gratings, stringers, etc., are then 
 removed and stored for use another season. 
 
 At Birdsview, Wash. — A permanent barrier at the Birdsview station, an 
 auxiliary of the Baker Lake station, in Washington, is of novel construction and 
 calls for more than passing notice. This barrier is located in a portion of Phin- 
 ney Creek where formerly there was a dam built for the purpose of obstructing 
 the passage of steel-head trout. When the dam washed out, a new channel 
 formed and the river bed was very much broadened. 
 
 The first step in the construction of the new barrier was the laying of four 
 heavy log stringers across this new channel from the abutment on the north to 
 
BuL. U. S. B. F., 190S. 
 
 Plate LXXXVI. 
 
 Fig. 19. 
 
 -Main rack across Battle Creek near the Battle Creek station, California. Upper sections of rack raised to 
 facilitate disposal of leaves and other debris. 
 
 Fig. 20.— Seining spawning 
 
 almon on the McCloud River, California, at the Baird station, 
 replaced the hand windlass. 
 
 Steam power has now 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 729 
 
 the new bank on the south side of the stream. The logs were let down through 
 the dam foundation to low-water level on the north side and the deep charmel 
 under them on the south side was filled with brush and gravel. The logs were 
 spotted down to form a practically level bed, reaching the width of the stream. 
 Heavy piles were then driven Ijehind each stringer to fonn alternate single and 
 double rows extending up and down stream. The log stringers were next 
 planked over, forming a platform 18 feet wide, similar to a regular dam apron, 
 extending from the north abutment to the final row of piles on the south side, 
 a distance of about 140 feet. 
 
 By planking the sides of the single rows of piles and all around the double 
 rows and filling the space with rocks, piers 4 feet high and approximately 2 feet 
 and 4 feet wide were formed. Through each pier at the bottom, behind the 
 upstream pile, openings i foot square were left, connecting the spaces between 
 the piers. These spaces, 12 in number, are approximately 8 feet wide and are 
 filled by swinging gates hinged to a 3 by 12 inch timber, spiked securely to the 
 piers on either side and forming a dam or flashboard across the space above. 
 By the insertion of other flashboards above this one a tight dam 4 feet high 
 can be quickly formed at any time. The utility of this feature will be explained 
 elsewhere. 
 
 The gates are made of i by 4 inch fir set on edge and nailed to 2 by 4 inch 
 joist, being strengthened by 2-inch blocks set between the rack bars and nailed 
 to them and the joists. These blocks thus determine the width of the inter- 
 stices in the gates. At the upper end of each gate an auger hole is bored 
 through the bars and blocks, to accommodate a 2-inch iron pipe, which passes 
 through the entire upper end of the gates. Ringbolts clasp these pipes and 
 are fastened to the 3 by 12 inch timber forming the flashboard, acting as hinges 
 upon which the gates swing. At the lower end of each gate a wide board, 
 i}4 hy 16 inches, is secured by means of braces, forming an angle of 45° with 
 the lower end of the gate. 
 
 At any ordinary stage of the stream the downstream ends of 'the gates 
 rest on supports which hold them a foot or more higher than the upper ends, 
 the water passing down through them to the floor of the apron, where it runs 
 away. The fish working up under the gates to the dam board find the cross 
 passages through the front end of the piers and finally reach the trap. It was 
 expected that during freshets the current acting on the flashboard would 
 always keep the lower ends of the gates above the surface of the water, and 
 up to a certain point this expectation was realized, but at very high stages of 
 the stream the large quantity of gravel in the water soon clogs and sinks the 
 gates. As the gates are only two-thirds the length of the apron, however, and 
 rise toward the lower end, the water shoots over them with such force that it 
 is projected some distance below the end of the apron, and fish attempting to 
 
730 
 
 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 u-S 
 
 a: >-; !< Kj" § 
 
 ^ 
 
 n 
 
 T3 
 
 C 
 
 ^ 
 
 
 
 K 
 
 
 UJ 
 
 « 
 
 
 a 
 
 0) 
 
 a 
 
 
 <L< 
 
 j: 
 
 c 
 
 i" 
 
 0; 
 
 c 
 
 M 
 
 > 
 
 
 
 c 
 
 J 
 
 •«-< 
 
 
 
 1 
 
 
 
 ^) 
 
 & 
 
 
 R 
 
 VI 
 
 •0 
 
 
 
 S 
 
 rt 
 
 rt 
 
 •o 
 
 m 
 
 
 
 cc 
 
 cq Q ui u; {£ 
 
 J5 
 
 
FISH-CULTURAL, PRACTICES IN THE BUREAU OF FISHERIES. 
 
 731 
 
 ¥ 
 
 
 
 
 ^ 
 
 ci 
 
 
 
 
 t 
 
 
 
 
 
 rt 
 
 
 il 
 
 
 
 
 '^ 
 
 
 
 
 
 w 
 
 V 
 
 B 
 
 I1 
 
 
 1 
 
 a 
 
 r. 
 
 
 
 bd 
 
 c 
 
 & 
 
 c 
 
 .- 
 
 w 
 
 y= 
 
732 BULLETIxV OK THE BUREAU OF FISHERIES. 
 
 scale the obstruction fall far. short of the ends of the gates. The barrier has 
 been watched many times when fish were jumping and when the largest drift 
 ran clear, and none has ever been seen to pass it. 
 
 By means of the dam boards entire control of the current can be had 
 during ordinary stages of water and any desired quantity sent to any section of 
 the barrier. Thus a strong current can be maintained through the trap section, 
 leading the fish to it, and when it is desired to remove the fish from the trap 
 the water can practically all be turned to some other section of the barrier. 
 
 One of the greatest difficulties in maintaining traps in the streams in this 
 section is due to the tremendous quantities of gravel carried in the water during 
 freshets, a sufficient amount being frequently deposited in front of a trap at 
 such times to change the course of the stream. With the present form of barrier 
 no trouble is experienced from this source, the insertion of the dam boards and 
 the opening of one space at a time quickly clearing away the accumulated gravel. 
 
 The ninth and tenth piers were continued upstream by driving three addi- 
 tional piles above each. The piers form the sides of the trap. Its floor is a 
 plank bottom, similar in construction to the apron, and the front is barred by 
 i%'-inch pickets placed i^ inches apart, the fish entering by the usual upstream 
 V of pickets. To protect the trap from high water the two piers between 
 which it is located were carried to a height of 8 feet. When it is desired 
 to fish the trap, the gate at its head is closed and entrance is made from below 
 by means of a door in the north side of the V. 
 
 The upper end of the fishway of the old dam was left in place, the narrow 
 passage between it and the new trap protecting the spaces at the south end of 
 the barrier from the current and from drift. These spaces have been racked 
 above and below to form commodious pens for males and unripe females. The 
 south end of the barrier is protected by a substantial abutment. 
 
 The maintenance of racks in Phinney Creek has been a very heavy item of 
 expense in past years, and the trap was frequently carried away by freshets just 
 at the height of the season, allowing large numbers of fish to escape and 
 considerably reducing the season's take of eggs. It is believed the new barrier 
 will stand any possible test that may be put upon it and will fish successfully 
 in almost any stage of water. The design is to be credited to Mr. A. H. 
 Dinsmore, superintendent of the station. 
 
 TAKING AND HATCHING THE EGGS. 
 
 Seining operations and spawntaking at California stations. — At the Baird 
 station all the salmon are caught by seine with the exception of a very few taken 
 in the trap in the upper rack. Owing to the swift current and the formation 
 of the river banks the seine is always landed at one place, the rope attached to 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 733 
 
 the upstream end being hauled by steam engine and the lower or down- 
 stream rope by a whim operated by horsepower. 
 
 When a haul is to be made the seine is carried by boat upstream about 50 
 yards from the landing before it is paid out from the shore. The downstream 
 rope is paid out automatically as the boat moves and is passed through a snatch 
 block to keep the current from closing the ends of the seine until opposite the 
 landing place, when both ends are drawn in at the same time. As soon as the 
 seine has been laid out above the pool the engine is started, drawing the seine 
 downstream toward the landing. When the last of the seine is in the water the 
 whim is started and the lower end is also hauled down and in. A landing having 
 been made, the ends of the seine are secured to pieces of rope fastened to pins 
 driven into the ground. Hooks on loose ends of these ropes are slipped behind 
 the corks of the seine. A trestle is placed under the rope to hold the cork line 
 high above the water and thus prevent the salmon from leaping over it. The 
 lead line is pulled upon the bank and thrown ovei iron pins to keep it from 
 slipping back into the deep water. 
 
 The ends having thus been fastened, all the men are free to handle the fish, 
 which must be quickly removed to prevent injury from crowding, this being 
 especially true of the summer run when the water is warm. The men stand in 
 the water and lift the fish from the net one by one until all have been looked 
 over, handling them with the aid of heavy woolen gloves, and grasping them 
 by the tail with the left hand while the right is placed just behind the pectoral 
 fins. A gentle pressure of the fingers at this point forces out a few eggs if the 
 fish is ripe; if not ripe it is thrown over the seine into the river. When a ripe 
 female is found it is carried to the pens beneath the spawning platform a few 
 yards distant, care being taken to hold it vent side up to prevent loss of eggs. 
 The males are handled in a similar manner, the ripe ones being placed in pens 
 near the females. 
 
 The seining crew are likewise the spawntakers, and after two hauls in the 
 morning they strip the fish caught the day before. The ripe salmon of each day's 
 catch are thus held over as a precaution against any possible immaturity of 
 the eggs. 
 
 On the spawning platform is a frame 6 or 8 inches wide, with sides con- 
 verging toward the bottom and open at one end, into which a spawning pan 
 can be slipped. The pan is rectangular in shape, about 6 inches wide and 14 
 inches long, with slanting sides and flaring ends. This shape is preferable to the 
 round pan because of facility in washing the eggs, it being possible to dip a 
 thin stream of water into the pan the entire length of one side. The frame 
 holds the pan secure and at the same time its slanting sides assist in guiding 
 the eggs into the pan. 
 
 After washing, the eggs are placed in 6-gallon spawn buckets which have 
 previously been filled with water. These buckets are made of galvanized iron, 
 with a wire-cloth strainer inserted near the rim to permit the escape of surplus 
 
734 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 water as eggs are added; they are painted inside with asphaltum, and are 
 provided with covers. With the eggs in them they are placed on a platform 
 which is built under water at sufficient depth to submerge about two-thirds 
 of the body of the bucket, and thus maintain a proper temperature durin- 
 water hardening. It is an essential feature of this platform that it be entirely 
 independent of the spawning platform in construction. The period between 
 the washing and hardening of the eggs is a critical one, and if they are jarred 
 or disturbed there will be consequent greater loss in the hatchery. 
 
 In the spawntaking operations five or six male fish are first dipped from the 
 pen and dropped on the floor, where they are allowed to lie until they stop struct 
 ghng and may be more easily handled. Then several females are'dipped ou^t 
 killed by a blow on the head with a piece of iron piping, and laid in the dead 
 box. The bottom of this box is inclined and has a narrow sht at the lower 
 end, through which any eggs that may escape from the fish will fall into a pan 
 underneath, where they will be fertilized by occasional applications of milt 
 The dipping and killing continues until all the fish are stripped, males always 
 being kept on the floor in order to have several ahead of the spawntakers 
 After being stripped the best-appearing males are returned to the pens, where 
 they usually recuperate and are again used to supply milt. 
 
 The female fish is grasped by forcing the fingers through the gills and the 
 thumb into the mouth, the hand being protected by a stout leather glove The 
 man who does this, the headholder, stands, holding the fish vertically. The 
 spawntaker, in a kneeling position, seizes the fish by the tail, bending its body 
 over the pan, while with a sweeping motion he makes an incision in the thin 
 side of the belly beginning near the pectoral fins and extending to the anal 
 The incision is usually made with a pocketknife, the blade being held between 
 the thumb and index finger within one-half to three-fourths inch of its extreme 
 point to prevent cutting too deep. Most of the eggs follow the knife and fall 
 into the pan, the remaining ripe eggs being released by running the fingers into 
 the body cavity. As soon as the eggs begin to flow into the pan milt is forced 
 over them and they are stirred by the spawntaker with a few movements of the 
 hand. They are then passed to the washer. The milt of one male may serve 
 for several pans of eggs, but it often happens early in the season that several 
 males are required to one pan. 
 
 As soon as the washer receives the pan of eggs he dips the edge of it in the 
 river, the inflowing water causing the eggs and milt to boil up. The eggs at 
 once settle back and the milt is poured off over the side of the pan. Ordinarily 
 two such dips suffice to clean the eggs, after which they are poured into spawn 
 buckets, and at once settle to the bottom, the surplus water escaping through 
 the wire-cloth strainer near the top. Spawning operations usually consume 
 about an hour. After water hardening the eggs are carried to the hatchery, 
 about 200 yards distant, and turned over to the hatchery crew. 
 
BUL. U. S. B. F., iqoS. 
 
 Plate LXXXVII. 
 
 Fig. 21. — Spawiitaking operations. Baird. Cal. The fish (chinook salmon) are dipped Irom the pen, killed by a blow 
 on the head, and passed to the spawntakers. The eggs are taken by opening the abdomen, and the stream of eggs 
 may be seen in the picture following the hand making the incision. 
 
 Fig. 22.— One method of stripping steelhead trout. Since this fish normally does not die after spawning, 
 the ripe fish are not killed as are the salmon, and they are so large, and so powerful in their strug- 
 gles, that the strait-jacket here shown is sometimes resorted to. 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 735 
 
 The spawning operations at Battle Creek and Mill Creek are practically the 
 same as those at Baird, and the seining differs only in minor details to meet 
 local conditions. 
 
 The method of taking eggs by incision has many advantages over the older 
 practice of expelling by hand, chief among them being the larger number of good 
 eggs secured and the higher percentage of impregnation. There is also a great 
 saving of time and labor, and, as all of the Pacific salmons die after once spawn- 
 ing, the killing of them in the process of taking the eggs merely hastens the 
 end, and constitutes no loss of adult fish. 
 
 The eggs of the Pacific salmons are comparatively large, ranging in diameter 
 from one-fourth inch for chinook to one-fifth inch for sockeye. Fortunately, 
 however, these large eggs do not require the same treatment and care as do those 
 of the other Salmonidae, but may be placed 12 to 20 deep in wire-cloth baskets, 
 30,000 to 50,000 to the basket, water temperature being an important factor in 
 deciding the quantity. The baskets, rectangular in shape, conform in width 
 to the troughs; the latter are of the Williamson type (up-current), the flow of 
 water to each trough varying with local conditions at the different hatcheries 
 from 8 to 20 gallons per minute. 
 
 Local practices in different regions. — Experience seems to indicate, with 
 regard to the sockeye, or red salmon, that there is an advantage in bleeding 
 the fish before stripping, thus obviating a flow of blood with the eggs when the 
 incision is made. It is therefore now customary at sockeye stations to decapi- 
 tate or else cut off the tails of the female fish as the first step in the spawn- 
 taking process. At Baker Lake, Washington, the practice is as follows: 
 
 The fish, which have been brought to the pens usually the da}- previous, 
 are dipped out one by one, decapitated, and dropped upon a draining rack, 
 where water is thrown over them to cleanse them for handling by the spawn- 
 taker. The latter impales the fish on a short spike conveniently located to hold 
 them over the pan while he makes the incision and removes the eggs. Two men 
 are occupied in the work so far. A third fertilizes and washes the eggs, then 
 conveys them to the hatchery. A million eggs may be secured in this manner 
 in one forenoon b)' one such crew. 
 
 Although not suitable for canning or for the market, the sockeye at this period 
 is edible and many are taken by local residents for food. Indians camp at 
 some of the stations and preserve large numbers of the salmon. The majority 
 of those killed, however, go to waste. At the Battle Creek station in one season 
 it is not an unusual thing to btiry during the spawning season 15,000 to 20,000 
 pounds of fish. 
 
 A convenient and economical method of separating dead eggs of salmon 
 from living ones is the use of a salt solution." If a basket of eggs is emptied 
 
 "O'Malley, Henry: Salt solution as an aid to fish culture. Transactions of the American Fish- 
 eries Society for 1905, p. 49. 
 
736 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 into the solution, the unfertihzed and dead eggs rise to the surface, where they 
 may be quickly skimmed off. The method is not applicable in the manipulation 
 of small quantities of eggs, but is very useful when, for some cause, there is an 
 abnormal loss. It has not been successfully extended to the eggs of species 
 other than the chinook, silver, and humpback salmons. 
 
 At the Baker Lake station, operated primarily for the propagation of the 
 sockeye salmon, both sockeye and silver salmon are captured by means of a 
 pound net as they enter the lake, but as the fish are not ripe at this stage of 
 their upstream passage it is necessary to hold them until the spawning season. 
 For this purpose a slough or bay at the head of the lake has been inclosed by 
 racks and webbing to make a pound about 20 feet wide and 500 feet long, with 
 an average depth of 6 feet. The entire bottom area is of soft mud ; in it are two 
 hollows of a maximum surface area of 300 feet and 400 feet, respectively, where 
 the water is approximately 12 feet in depth. The water supply, from 800 to 
 2,400 gallons per minute, is derived from several small mountain creeks fed by 
 glaciers and snow, and from springs. 
 
 This has proved an ideal place in which to hold salmon while their eggs are 
 maturing. As many as 8,000 fish have thus been confined for thirty days and 
 6,000 for three months. After remaining in the inclosure for three or four 
 months the fish are as clean and free from abrasion as fish that have not been 
 penned. A noticeable fact, however, is that very few of these salmon ripen 
 until October or November, while in previous years the spawning season has oc- 
 curred during September and before October 15. The difference is attributed 
 to the low temperature of the water in the inclosure, which remains from 6 to 8 
 degrees below that of the natural spawning beds of the lake. (During the sum- 
 mer the average water temperature on the spawning beds is about 56°.) The 
 success of the impounding of the salmon is attributed to the low temperature 
 of the water. 
 
 Silver salmon, as well as sockeye, have been successfully impounded at 
 
 Baker Lake. A few impounded chinook, however — never more than 12 at a 
 
 time — fungussed rapidly, and after three weeks usually died before the eggs 
 
 were taken. 
 
 MARINE FISH CULTURE. 
 
 COD, POLLOCK, AND FLATFISH. 
 
 On the New England coast the Bureau maintains three stations, for the 
 hatching of marine fishes and the lobster. 
 
 Of the fishes the cod is first in importance. Its cultivation consists prin- 
 cipally in hatching eggs obtained from market fish, the fish being stripped either 
 by the fishermen or by the Bureau's spawntakers. The latter are daily distrib- 
 uted among the fishing vessels for the double purpose of stripping ripe fish as 
 fast as hauled aboard and of collecting and caring for all eggs the fishermen may 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 737 
 
 have taken. Coming as it does during the winter months, from early December 
 to March on the Massachusetts coast and extending through March and April on 
 the Maine coast, the spawning season is a trying one for the spawntakers, who 
 must share many of the dangers and hardships of the fishermen. 
 
 At the Woods Hole station advantage has been taken of the presence of a 
 large salt-water reservoir under the hatchery to test the Norwegian method of 
 obtaining eggs. By this process adult fish are penned and allowed to spawn 
 naturally. The eggs float to the outlet, where they are caught in a receptacle 
 placed for the purpose, and thence are transferred to the hatching boxes. It 
 has been demonstrated that a larger percentage of fertilized eggs per fish can be 
 obtained thus than by the former method of stripping penned fish. The increase 
 is not due to a higher percentage in fertilization, but to the fact that in the case 
 of the penned fish there is an almost unavoidable loss of eggs extruded in the 
 crates. Both of these methods are an improvement over nature, in that the eggs 
 are protected from the time of extrusion until they have hatched and the surviving 
 parent fish are returned to the ocean. This is fish culture in the usual sense and 
 not purely conservation of an otherwise waste product, as is the collection of 
 eggs from market fish. 
 
 Prior to the spawning season for cod, or from the middle of October to the 
 latter part of December, spawntakers are distributed among the pollock fisher- 
 men, and from the eggs thus collected many millions of pollock fry are hatched 
 and distributed annually. 
 
 The cultivation of flatfish is conducted on a more extensive scale than any 
 other marine fish-culture work. The adult fish are taken from the fyke nets in 
 which captured, usually the Bureau's own, directly to the hatcheries, where they 
 are placed in tanks and held until they have spawned, the eggs being removed 
 daily from the tanks to the hatching apparatus. It is possible to spawn flatfish 
 artificially, and eggs are sometimes obtained in that way. 
 
 LOBSTERS. 
 
 Lobster culture also, as conducted by the Bureau, effects a saving of an other- 
 wise lost resource, berried lobsters purchased from the fishermen furnishing eggs 
 for the hatchery and being later retvu-ned to the ocean. At the Boothbay Har- 
 bor, Me., station the parent lobsters are held until the eggs are ripe in a pound 
 similar to those used by the lobster men ; and as it has been found that eggs taken 
 late in the fall or early winter do not hatch successfully, one such pound is utilized 
 to hold some 10,000 or 12,000 lobsters throughout the winter. The losses of 
 lobsters during the period of confinement are only normal, and the quantity and 
 quality of the eggs are superior to those obtained from freshly caught stock. 
 It is noticeable also that the eggs of the impounded stock hatch almost simul- 
 taneously and somewhat earlier than those from freshly caught lobsters, undoubt- 
 
 B. B. F. 1908 — 47 
 
738 BULIvETIN OF THE BUREAU OF FISHERIES. 
 
 edly bee use of the warmer, shallow, inshore water. By the middle of April the 
 eggs are sufficiently well advanced to be removed from the parent and put up 
 in the hatching jars; and as at this time the lobsters become quite active in a 
 rising water temperature it is quite important that they then be removed to 
 avoid serious losses from mutilation. 
 
 The removal of impounded lobsters is effected at low tide, the flood gates 
 being opened and a portion of the water drawn off with care to retain enough 
 water to protect the lobsters from exposure. Men in dories then go about the 
 shallow portions of the pound picking up the lobsters on ordinary clam forks or 
 hoes, the sharp teeth of which have been blunted. It was formerly the custom 
 to use a drag seine for gathering the lobsters, but taken in such quantities they 
 mutilate one another and it has been found preferable to remove them by hand. 
 After a portion of the stock has been removed the water is drawn still lower, 
 until finally only a small area of the pound is flooded, and the remaining lobsters 
 are removed. 
 
 As it has not been possible to transport lobster eggs successfully when 
 detached, the berried females are always taken to the hatchery to be stripped, 
 the transfer being made in the wells of fishing smacks or the Bureau's vessels. 
 From this time to the close of the season in July berried lobsters are collected 
 from the fishermen and transferred to the hatchery to be stripped. Imme- 
 diately after the close of the season the collection of fresh berried lobsters for 
 stocking the pound is begun and continued into November. 
 
 It is unquestionable that the impounding of lobsters as practiced in Maine 
 is superior to any other present method of holding the adults for a length of 
 time. The character of the Maine coast, with its numerous natural inlets and 
 its unusual rise and fall of tide, affords especial advantages for the use of pounds. 
 When these conditions do not exist, however, recourse can be had to cars, 
 although data thus far obtainable fully demonstrate that a larger number of 
 lobsters can be held for a longer time and with a smaller percentage of loss in 
 pounds than in cars. 
 
 Contrary to the custom of the pound fishermen, the ice on the surface of 
 the pound operated by the Bureau of Fisheries is removed from time to time 
 during the winter, with the result that a much larger percentage of lobsters is 
 found in the spring. It has been observed that the rising and falling of ice 
 with the tide frequently crushes lobsters that happen to be in the shallow water 
 near the edges of the pound, and removal of the ice at intervals obviates this 
 difficulty. 
 
 Experiments have been made as to the poimdkeepers' practice of inserting 
 wooden plugs in the claws of penned lobsters to prevent their mutilating one 
 another. For lobsters intended for market the procedure seems suitable, 
 though it results in an unsightly discoloration of the muscles. There seemed 
 
BuL. U. S. B. F., 1908. 
 
 PI.ATE LXXXVIII. 
 
 HiG. 23. — Eiliiipiiient of IMcDuiiakl ;iiitonicitic tidal buxtr>. fur hatching cod, liootlibay Harliur. ^Ic. Shuw:^ buxcb lifted 
 out of troughs and bottom upward on farther tables. The bottom is of scrim, and by means of cleats is held i^ 
 inches above the bottom of the trough at the center. By an arrangement of partitions at the head of the trough 
 the eggs receive the supply of water through the scrim bottoms of the boxes, also through a small hole in one end 
 of the box. The distinctive feature of the apparatus, whence it is called " tidal," is the automatic siphon outflow, 
 by means of which the water is alternately drawn down and replenished. Staudpipe with siphon cap is shown in 
 near troughs; waste trough below. 
 
 Fig. 24.— Beri-ied lobsters, taken fn.in pMund .ii l;. m ,iii!,,,y ILirh.i. -tUiuu iMaine). in course of transfer to wells of the 
 steamer which is to couvey them to the hatchery for stripping. 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 739 
 
 to be a question, however, whether it would be feasible with the impounded 
 stock of the Bureau destined for liberation in the open waters after removal of 
 their eggs. The experiments showed that out of 2,110 lobsters removed from 
 the pound, all of which had been plugged when put in, 742 had lost both plugs, 
 563 had lost one plug, and 605 retained both. The warm weather when they 
 are first confined, however, is their most active period, when the plugs do most 
 good. Mutilation is thus prevented and the plugs apparently work no per- 
 manent injury to the lobsters. 
 
 It is important in the impounding of lobsters to take precautions, so far as 
 possible, for the exclusion of eels, which have an especial liking for the eggs and 
 will strip a female lobster in a very short time. Even with all precautions it 
 seems impossible to exclude eels entirely ; it is probable that many enter when 
 small and grow up in the pound. 
 
 The rearing of lobster fry to the fourth molt, as practiced by the Rhode 
 Island Fish Commission and so admirably set forth in a paper read at the con- 
 gress," has not as yet been taken up by the Bureau, but is doubtless feasible at 
 the Boothbay Harbor station. Before the first experiments in this direction at 
 Boothbay it was thought that owing to the lower temperature of the water in 
 this more northern latitude the periods of molting would be prolonged and the 
 feeding and care of the fry consequently attended with abnormal losses. It has 
 since been found that this difficulty can probably be met by installing the 
 rearing plant in a lobster pound, where the temperature is higher and more even 
 than in the open waters. The Bureau therefore hopes to enter upon this under- 
 taking in the near future, for the purpose of rearing a portion of the lobster out- 
 put. To attempt to rear to the fourth molt the entire product of the Boothbay 
 station would involve an expenditure far beyond present financial resources. 
 
 MEASURING AND COUNTING HSH EGGS AND FRY. 
 
 Immediately after water hardening, for a short period varying with the 
 species and water temperature, the careful handling of fish eggs is not injurious. 
 During this period their numbers may be very definitely ascertained by the use 
 of any receptacle suitable for a measure, the capacity of the receptacle having 
 first been ascertained by counting the whole or a fractional part of its contents. 
 
 For eggs of the trouts and those of smaller size an apothecary's graduate or 
 the ordinary graduated quart or pint measure is commonly used. For large 
 quantities the long-handled dipper used in transferring them to the hatching 
 apparatus may be advantageously utilized. As many eggs as possible are 
 poured into the measure, nearly all of the water being forced out over the rim. 
 
 "Mead, A. D.: A method of lobster culture. Proceedings Fourth International Fishery Congress, 
 Bulletin of the Bureau of Fisheries, vol. x.xviii, 1908, p. 219-240, pi. vii-xi. 
 
740 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 Unless the eggs are to be transferred to a hatchery beyond the jurisdiction 
 of the shipper, the eggs may not be measured until a more convenient time, it 
 being possible from long familiarity with the capacity of the apparatus in actual 
 use to estimate quite accurately the number of eggs on hand at any time. Pro- 
 viding they are spread uniformly, the number of eggs to a square inch is a fairly 
 accurate basis for ascertaining how many eggs are on each tray. Some fish 
 culturists prefer to ascertain the actual number of eggs on hand by weighing 
 them after having determined by actual count the basis for such calculations. 
 
 These methods are especially applicable to the heavy eggs of the Salmonidge, 
 and may be employed not only after water hardening but also at any stage of 
 incubation after the eggs are eyed up to a day or so before hatching, at which 
 last stage a measurement closely approximates the number of fry that will be 
 in the subsequent hatch. 
 
 Eggs hatched in jars are usually measured by means of a graduated scale 
 in the form of a square made of wood, the units indicated on the long leg of the 
 square. The square is adjusted to the jar as shown in figure i. The scale reads 
 from the bottom line upward, the first or bottom line being at a height corre- 
 sponding to the level attained in a jar by a measured half pint of water, and 
 each line represents the number of eggs of a given species as established by actual 
 count from a measured half pint. The dead eggs will have been from time to 
 time siphoned oif and when the remainder are fully developed or about to hatch 
 the scale is applied to each jar and a very careful measurement is made to ascer- 
 tain their number. The number of fry available for distribution will be approxi- 
 mately the number of eyed eggs in the jars just before hatching, as the mortality 
 after this stage of development is usually inappreciable. 
 
 A novel method of obtaining the number of eggs in a given lot has lately 
 come into use, and as the work can be done without counting a large number 
 of eggs, has proved especially valuable in dealing with eggs of small size at field 
 stations where no measurements have been established. This method, devised 
 by Mr. H. von Bayer, architect and engineer of the Btueau, employs a gauge 
 which quickly gives the diameter of the eggs, knowing which it is possible by 
 reference to a diagram to determine at once the number of eggs to the quart." 
 
 In the keeping of accurate hatching records it is important that the basis 
 for computations be ascertained immediately before any general measuring 
 methods are applied because it is well recognized that eggs of most species vary 
 in diameter with stock from different waters and that eggs from any given col- 
 lecting station vary at different periods of the spawning season, those taken at 
 the height of the season being larger and more uniform than those taken earlier 
 or toward the close of the season. For instance, brook trout eggs taken at a 
 
 " von Bayer, H. ; A method of measuring fish eggs Proceedings Fourth International Fishery 
 Congress, Bulletin of the Bureau of Fisheries, vol. xxviii, 1908, p. 1009-1014. 
 
FISH-CUI<TURAL PRACTICES IN THE BUREAU OF FISHERIES. 741 
 
 particular station may run 250 to the fluid ounce during the height of the season, 
 while the first take may have run 300 to the ounce and the last eggs of the 
 season may average 400 or more to the ounce. These variations make it neces- 
 sary frequently to establish a new measure for ascertaining the actual number 
 of eggs. There is also to be taken into consideration the fact that eggs of 
 almost all fishes increase in size from 4 per cent to 1 5 per cent according to the 
 species, from the time they are water-hardened up to the time they are about 
 to hatch. 
 
 This fact, coupled with the fact that eggs of the same species vary in size 
 at different sources of supply and periodically at the same source of supply, is 
 a point in favor of the von Bayer method of computing the numbers of eggs of 
 small diameter. 
 
 Sac-absorbed fry and advanced fry of the trouts, landlocked salmon, etc., 
 may be measured in the same manner as are the eggs — in an apothecary's grad- 
 uate or other container, straight vertical sides being preferable to the flaring 
 sides of the ordinary glass graduate. The ordinary graduated half pint or pint 
 cup used by cooks is a very convenient measure. The fry are poured in until 
 the measure is overflowing with them to the exclusion of practically all the water, 
 the filling and emptying being done quickly. Actual count of the number in 
 one measure establishes the basis for computation. The growth during this 
 period being very rapid, however, a new unit must be determined daily. 
 
 The numbers of fingerlings are ascertained by actual count of each lot as 
 dipped a few at a time from trough to transportation can or other receptacle 
 by means of a small hand net of tightly stretched bobbinet. 
 
 TRANSPORTATION OF EGGS. 
 
 To equalize and facilitate the work of the hatcheries it is customary to 
 transport, sometimes to considerable distances, both green and eyed eggs from 
 one station to another, thus effecting the distribution of fish through the dis- 
 tribution of eggs. Several auxiliary stations are maintained on the Great Lakes 
 solely for hatching eggs received from Northville, it having been found economy 
 to transfer to them as eyed eggs the portion of the Northville station output des- 
 tined for distribution to waters in those localities. Both green and eyed eggs 
 are also shipped to state hatcheries and the latter to foreign countries. 
 
 The methods of conveying green eggs from the field where collected vary 
 to suit conditions. The stations at which eggs of commercial fishes are hatched 
 in large numbers are usually located conveniently to the source of supply so 
 that it is possible to carry the freshly fertilized eggs to them in the pans, buckets, 
 or other receptacles which constitute the equipment of the spawn taker. It 
 often happens, however, that the eggs must be held in the field or be in transit 
 for two or more days, and in such cases a packing case is employed. 
 
742 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 USUAL STYLE OF PACKING CASE. 
 
 For ordinary purposes a packing case consists of a wooden box which will 
 accommodate a stack of trays and an ice hopper, with 2 to 4 inches of insulation 
 or packing on all sides and under the tray stacks. The frames of the trays are 
 made of light, soft wood, usually white pine, ^ inch by % inch or Js inch by 
 % inch, over which is tightly stretched a bottom of canton flannel, nap side up, 
 or of heavy cheese cloth, with perforations in the cloth to facilitate the passage 
 of water. 
 
 For long-distance shipments it is customary to make the bottoms of the 
 trays of wire cloth painted with turpentine asphaltum, over which canton 
 flannel or cheese cloth may be spread before putting the eggs upon them, and 
 a thin layer of moss under the cheese cloth in addition is advocated by some 
 fish culturists. The soft spongy bed of moss prevents concussion in handling 
 and retains moisture, while at the same time it allows a sufficient circulation 
 of oxygen and free passage for water from melting ice, etc. For very long or 
 warm-weather shipments it is sometimes advisable to use a case with double 
 sides, with insulation between, the space between the inner case and stack of 
 trays being filled with ice. 
 
 The ice hopper is about 3 or 4 inches deep, of the same length and width as 
 the tray frames, and rests upon the top of the stack. Its bottom is perforated 
 to allow a drip from the ice through the trays and thus keep the eggs constantly 
 moist and cool. Double cases, the length twice the width, arranged for two 
 stacks of trays side by side with a partition between them, are sometimes used 
 in the Pacific salmon work. 
 
 The fish culturist often works in isolated places and must use the material 
 which is most accessible and economical. Moss is therefore very generally used 
 for filling the space between the stack of trays and the packing case; mineral 
 wool, leaves, sawdust, and shavings also well serve the same purpose, though 
 with ice in contact with the inside lining of the outside case mineral wool is 
 objected to because when damp it has a tendency to sag. Any of these materials 
 may be used for insulation in the long-distance cases. Cork board insulation, 
 also, is very efficient and of lighter weight than the others, but it has not been 
 tested so fully as have shavings, at present the material most popular with 
 caretakers. 
 
 The cover to the case may be screwed on, but for shipments requiring the 
 renewal of ice it is customary to provide a hinged cover fastened with hasp and 
 staple. 
 
 ADAPTATIONS AND VARIATIONS OF METHOD. 
 
 Eyed eggs of the Atlantic and Pacific salmon and of the steelhead trout 
 have all been successfully shipped in the ordinary case, but the method of packing 
 eggs of the Atlantic salmon at the Craig Brook (Me.) station has the special 
 
Bt'L. U. S. B. F., 1908. 
 
 Plate LXXXIX. 
 
 Fi(^.. 25. — Box of trout egps just opened. Showintj ice hopper at left, stack of travs which have been taken out of tlie 
 luoss in center of box. and one tray with covering of mosquito net and moss removed. Tliis is the common method 
 of packing for ordinary shipments. (Described on p. 742.) 
 
 f :i;- 
 
 ^.■:i\ 
 
 iifWP 
 
 Fig. 26.— Tray of trout eggs with mosquito net and moss in which packed. 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 743 
 
 advantage of making a comparatively light package — a factor of great economic 
 importance in transportation. The outside case may be an ordinary box of 
 suitable dimensions. In it are packed, surrounded by moss, several boxes made 
 of ^-inch boards, and usually 12 inches wide by 15 inches long by 3K inches 
 deep, each box containing a mass of 10,000 to 20,000 eggs in mosquito netting, 
 with moss around all sides. No ice is used, care being taken that the packing 
 be done in a temperature below 50°, that all packing material be kept in a 
 place slightly below freezing point, and that the moss in which the eggs are 
 packed be sprinkled with snow. This method of packing is an economical one 
 for shipments of eggs of Salmonidae during cold weather, but can not advan- 
 tageously be used for eggs of spring spawning fishes unless there is available a 
 cold-storage room in which to do the packing. Recently the superintendent of 
 the Baker Lake (Wash.) station, who has had occasion to ship eggs of steelhead 
 trout and Pacific salmon in warm weather, has packed them in light cases with 
 alternate layers of moss, and then placed two tiers of these thin cases side by 
 side in an outer case with a large hopper of ice over the whole, the drip passing 
 down between the two tiers of inner cases. The chief advantage of this case 
 for long-distance shipments is in the fact that less ice is required than in other 
 forms of cases using ice, with a consequent saving in transportation charges. It 
 can also be used in warm as well as cold weather. It is believed it will be 
 economy to extend the use of this case in packing eggs of other species of 
 Salmonidae. 
 
 Green eggs of the brook trout and chars are carried in spawning pans or 
 buckets, the spawntakers sometimes, by the use of a neck yoke, carrying two 
 pails of trout eggs several miles. It is possible to ship the green eggs a half 
 day's journey without serious loss, but it is preferable to eye them at places 
 convenient to the traps where the parent fish are caught, after which they are 
 packed by the ordinary method. 
 
 For grayling eggs cheese cloth is used for the bottoms of the trays instead 
 of canton flannel and it is preferred by some in packing other kinds of eggs 
 because it permits of a better circulation of air and is not so apt to hold water. 
 No moss is used on the trays over the eggs, but only mosquito netting, as the 
 eggs will not stand pressure. Both the hopper and the chambers around the 
 tray stack are kept filled with ice, thus maintaining in warm weather a tem- 
 perature of about 40° F. 
 
 In transferring by messenger large numbers of eggs, whether green or eyed, 
 of any species, it is customary to omit the packing on and around the trays, ice 
 being used to regulate the temperature. 
 
 Eggs of the shad and other species of which the period of incubation is but 
 a few days are usually shipped within forty-eight hours after being collected. 
 Shad eggs are seldom shipped for more than a few hours' travel. For this 
 
744 
 
 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 ^ 
 
 ^Xxic/-'<^^ 
 
 \I 
 
 
 /;;^v5.^:^/ ^-^^^c'^^^^^v^.^ :-:->-:;•/''; ^•■ ' 
 
 / 
 
 .K 
 
 '.VvM^./A 
 
 
 Xi^tycTeoF' 
 
 '4 
 
 Fig. 7- — Atkins-Dinsmore shipping case. Longitudinal section 
 
 -z;^ 
 
 »3- 
 
 -J^ 
 
 X 
 
 ^ 
 
 1 
 
 ■.s^ 
 
 «^1 
 
 \ 
 
 A 
 
 lyi^ 
 
 
 
 F^ 
 
 V 
 
 fi 
 
 Fig. 8. — Atkins-Dinsmore shipping case. Plan. 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 
 
 745 
 
 Aaja 
 
 i-l. •-, 
 
 
 
 :i, 
 
 ' } — ^^ — " ||h .-'3- 
 
 
 J V 
 
 :£ 
 
 -t3 — rr 
 
 
 iht'n^fs 
 
 purpose they are laid in cheese cloth on wire-bottom trays, between wooden 
 frames also covered with wire cloth, strapped together, and shipped without 
 further packing. 
 
 Experiments at Havre de Grace, Md., seem to demonstrate that it is prac- 
 ticable to pack eggs of the white perch in the ordinary trout egg case with ice 
 hopper and ship them on journeys of thirty-six hours' duration without apparent 
 injury. Two lots of green eggs taken at a temperature of 56° and 70°, respec- 
 tively, and held in trout egg cases for twenty-four to twenty-six hours, hatched 
 as well as eggs placed directly in the hatching jars. In the first experiment 
 there was no change in temperature, but in the second experiment there was 
 a fall from 70° to 64°. White perch eggs to the number of 3,000,000 have 
 in several instances been .^i"*" 
 
 shipped by express from 
 Havre de Grace, Md., to 
 Washington, D. C, a half 
 day's travel, in four McDon- 
 ald jars packed in sawdust, 
 ice being used in the packing 
 when the air temperature 
 seemed to require. The 
 jars were equipped with 
 the usual glass tubes, which 
 extended above the packing, 
 but whether this provision 
 for aeration was necessary 
 has not been tested. To in- 
 sure proper aeration, how- 
 ever, it would seem advis- 
 able, with present knowl- 
 edge of the subject, not to ship large numbers of white perch eggs in 
 water for travel of four or more hours without a caretaker. 
 
 Attempts to transport yellow perch eggs on trays have not given satisfac- 
 tory results, but it is apparently possible to carry them successfully almost any 
 reasonable distance in the ordinary transportation cans, i to 2 gallons of eggs 
 to 8 gallons of water, the proportion varying with the distance to be traveled, 
 and care being taken to aerate and temper the water. 
 
 Green pike perch eggs may be carried from near-by collecting grounds to 
 the hatchery in tubs or transportation cans, care being taken to renew the 
 water frequently, to keep it well aerated, and of a proper temperature. Ice 
 must be prevented from coming in contact ,vith the eggs, because, unlike most 
 
 Fig. 9. 
 
 -Alkius-Diusmure shipping case 
 
 Cross section. 
 
746 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 other eggs, they are very sensitive to such exposure. Half-barrel fish kegs or 
 kits with one end knocked out and iron handles attached make very good and 
 economical vessels in which to transport pike perch eggs. It is customary to 
 insert a wire-cloth drain in the top of the kit on one side and the kits are 
 asphalted inside. Canvas is thrown over the top to serve as a cover. When 
 the eggs are held over night in these kits it is expected to supply them with 
 running water whenever possible to do so ; otherwise the water must be period- 
 ically renewed or aerated. At some of the field stations a pipe arranged with 
 pet cocks and rubber tubing for supplying water to a number of kits or trans- 
 portation cans saves much labor in the matter of aerating eggs which must be 
 held for a day or more. 
 
 Pike perch eggs collected at a distance from the hatchery are conveyed 
 thereto in the usual transportation case on trays, a small amount of ice being 
 placed on the top tray, which is substituted for the usual ice hopper; an inch 
 space around the sides of the stack is also filled with ice. Ice over the stack 
 must be used sparingly or the green eggs may be injured by the cold water 
 where it trickles upon them. For large shipments it is customary to have a 
 caretaker accompany the packages to regulate the temperature, etc., the use 
 of moss on the trays as well as the insulation material then being omitted. 
 
 Since eyed pike perch eggs are usually shipped during the month of May, 
 for safety ice is used around the stack of trays as well as in the hopper, even 
 on a two days' journey. 
 
 Whitefish eggs are transported from the field of collection by both of the 
 methods employed in the transfer of pike perch eggs, although they do not 
 require quite so much attention. Eyed whitefish eggs are packed on trays in 
 the ordinary way. 
 
 The fields for the collection of lake trout eggs being widely distributed, it 
 sometimes happens that green eggs are held in transportation cases for several 
 days before their arrival at the hatchery. They carry as well laid directly on 
 the wire-cloth bottoms as on a layer of cheese cloth. No packing is necessary 
 if they are in the care of an attendant, since the latter can regulate the temper- 
 ature by the use of ice and with water of the proper temperature, the trays 
 being removed and sprinkled at least once in twenty-four hours. 
 
 In packing lake trout eggs for short distances, say up to a thousand miles, 
 the ice hopper is omitted. Mosquito netting and moss are put on in the usual 
 manner and the top of the moss is slightly frosted before the trays are stacked 
 on the baseboard. A light piece of lumber is used in place of the ice hopper 
 and fine shavings are solidly packed under, above, and on all sides of the stack 
 of trays. For the longer shipments the ordinary ice hopper is used, fine shav- 
 ings being placed around the stack of trays for insulation. 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 747 
 
 In the handling of eggs of the cod and other marine fishes a so-called kettle, 
 oval and with a concave top, is used to retain them on board ship in a choppy 
 sea. The water in which they are kept must be frequently changed or aerated. 
 The eggs are shipped to the hatchery in large fruit or butter jars, rockweed or 
 moss, together with ice or snow, being used in packing them. It is regarded 
 as impracticable to ship eggs of marine fishes for travel of more than two or 
 three days. 
 
 ARGENTINE CASE. 
 
 While the various methods above described have all been successfully 
 employed in the transportation of eggs across the United States and also to 
 Europe without an attendant, shipments of eggs to points south of the equator, 
 usually leaving this country in winter and arriving at their destination in sum- 
 mer, have called for more than usual attention to the methods of packing them, 
 and a caretaker is quite essential. 
 
 A highly efficient form of shipping case has been developed during the past 
 few years for the transportation of eggs of the Salmonidee from this country 
 to Argentina. It is 3 feet 6 inches long, 2 feet wide, and not exceeding 30 
 inches high, outside measurement, and is constructed of selected tongued and 
 grooved lumber. It has double walls, with bottom and top common to both, 
 the 2-inch space between the walls being filled with nonconducting material, 
 preferably tightly packed shavings. Between the inner wall and the stack of 
 trays is a 2i'4^-inch space for ice, separated from the trays by perforated zinc. 
 Between the latter and the trays, in a 3^-inch space, are the vertical supports of 
 the zinc, viz, double corner supports, one being ^ by iX inches, the other 
 being % by i inch; two intermediate supports of ^ by i inch material, which 
 are provided on either side of the case and one at each end; and cross braces 
 of X by I inch material, which extend from the uprights to the inner walls of 
 the case. 
 
 The ice hopper, 3 inches in depth, and having the same outside dimen- 
 sions as the trays, rests upon the latter and fills the space between the upper- 
 most tray and the top of the case. It has a perforated zinc bottom, and, to 
 facilitate handling, cleats of small ropes are attached to it. The top of the 
 case is insulated with a 2 -inch thickness of nonconductor covered with sheet 
 zinc, this insulation fitting closely into the chest when closed, and thus covering 
 not only the ice hopper but the ice spaces around the sides as well. In the 
 bottom grooves lead to a %'-inch drain hole, which is provided with a cork. Two 
 cleats li, by 3 inches are attached lengthwise to the bottom on the outside. 
 
 The trays are one-half inch deep, 27 inches long, and 9 inches wide inside 
 measurements, the frames being of yi by X inch material. The bottom of each 
 
748 
 
 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 tray is covered with wire cloth no. 25 gauge, about 12 meshes to the inch, 
 stretched tightly to prevent sagging and consequent uneven distribution of the 
 drip water. A narrow binding of cloth is tacked around the bottom of each 
 tray to prevent the wire edge from catching on the mosquito net covering of 
 the tray beneath. On the inside ends of the traj^s are fastened short lifting 
 cleats, and wedges hold the trays securely in place. The bottom tray rests on 
 three X"inch cleats extending lengthwise of the case, one at either side and the 
 other in the middle. It is important to have the trays of uniform size, that 
 they may be interchangeable. 
 
 The trays and interior of the case are coated with asphaltum. To facili- 
 tate opening from either side, four hasps are used, two on each side of the case. 
 Two rope handles side by side are placed on each end of the case, with a cleat 
 of three -fourth inch material just above the holes for each handle. 
 
 /■/ra /a trji^ff 
 
 Fig. 10. — Argentine shipping case. Section. 
 
 Eggs selected for shipping should barely show the eye spots without the 
 aid of a glass. In packing, a layer of damp moss is spread one-fourth of an 
 inch deep as evenly as possible over the tray bottom, and upon this is placed 
 a covering of mosquito net or bobbinet. The eggs are laid upon the netting 
 one or two layers deep, spread to within one-half inch of the tray frame and 
 covered with another piece of netting to keep them separate from the moss, 
 which is sprinkled in a light layer over it, filling the tray. The netting is cut 
 large enough to extend over the outer edges of the tray, so that the eggs may 
 not be disturbed when a tray is lifted for examination. 
 
 On shipboard, as the greater part of the journey is made, the cases of eggs 
 are kept in one of the fruit or cold storage rooms having a temperature of about 
 38° F. To this room the attendant has access, and it is his duty daily to moisten 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 
 
 749 
 
 the eggs by pouring through the ice hopper water of the same temperature as 
 the eggs, 34° to 35°. The ice compartments are frequently replenished and the 
 eggs are picked over whenever necessary. 
 
 ,4^^ 
 
 
 s «». 
 
 It will be seen that the method of caring for the eggs is not novel. The 
 chief improvements in the case are to make it easy for the caretaker to tend 
 the eggs in the crowded quarters of a ship's storage compartments and to facil- 
 itate handling each individual tray. 
 
750 BUI^LETIN OF THE BUREAU OF FISHERIES. 
 
 GERMAN-CHILE CASE. 
 
 Another case used on journeys in which the care of the eggs is the same 
 as above described is the German-Chile case, so called by reason of having first 
 been employed by German fish culturists in shipping trout eggs from Germany 
 to the Chilean Government. It was brought to the attention of fish culturists 
 in this country by Mr. E. A. Tulian, chief of the section of fish culture of the 
 Argentine Government. In 1907 and 1908 this case, somewhat modified, was 
 used by Mr. Tulian in transporting trout and salmon eggs from the United 
 States to Argentina with better results than had been secured with any other 
 form of case. Owing to the absence of moss directly on the eggs, the German- 
 Chile case is especially adapted to the handling of rainbow trout eggs, the mem- 
 brane of which is more delicate than that of any other species of the Salmonidae. 
 From the latest observations it is undoubtedly true that the ideal form of long- 
 distance shipping case for all species, at least when accompanied by an attend- 
 ant, is the one wherein no moss or other substance is placed directlj' on the 
 eggs. 
 
 The German-Chile case is constructed on the same principles as the Argen- 
 tine case above described. The case proper is built of selected lumber and is 
 2()yi inches long, 20 inches wide, and 15 inches high, outside measurements. It 
 is so similar in construction to the Argentine case that, aside from a general 
 description, only the differences between the two will be pointed out. 
 
 The German-Chile case accommodates two stacks of trays 7K t>y 8^ inches, 
 in a double-chambered compartment having walls of unperforated galvanized 
 iron, which is strengthened by a heavy wire around the top edge. A removable 
 metal partition separates the two stacks of trays. The compartment itself is 
 made one-fourth inch larger each way than the tray frames, to allow for swelling 
 and the binding twine which will be placed around the trays. Next to and 
 surrounding the tray compartment is the ice, outside of which is the dry moss 
 or other nonconductor, within wooden walls, the same as in the Argentine case, 
 while resting upon the top of the trays and ice compartments is an ice hopper. 
 For purposes of insulation the ice in the hopper is covered with a cushion filled 
 with dry moss, oilcloth being placed between the cushion and the ice. The 
 metal bottom of the hopper has perforations only over the trays, that the eggs 
 may receive the benefit of all drip water. Small cleats are fixed at either end 
 of the ice hopper to facilitate handling. Under the tray compartment and 
 coextensive with it is a perforated wooden false bottom to the case, between 
 which and the bottom proper is a i-inch air space. A drain hole is provided in 
 the bottom proper. The egg trays are made a trifle deeper than the diameter 
 of the eggs; and the latter are placed on them a single layer deep without any 
 
FISH-CUWURAL PRACTICES IN THE BUREAU OF FISHERIES. 
 
 751 
 
 covering of moss. The tray frames are seven-eighths inch wide, and usuall}" 
 three-sixteenths inch to five-sixteenths inch thick, with a bottom either scrim 
 
 Fig. 12. — German-Chile shipping case. Longitudinal section. 
 
 or canton flannel. Instead of being made into one package for each chamber, 
 the trays are bomid together in fives, with strong twine for binding material, 
 and alternating with each package are double-depth trays, or dummies, filled 
 
 Fig. 13. — German-Chile shipping case. Plan. 
 
 with wet moss. One of these moss-filled trays is placed also in the bottom of 
 each of the two tray chambers. 
 
752 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 TRANSPORTATION OF FISH. 
 
 In distributing young fish from the hatcheries to the waters they are to 
 stock, six special cars are employed. They are equipped with all necessary 
 apparatus for the safe carriage of young and adult fishes and each is provided 
 with a buffet and sleeping accommodations for a crew of five men. The cars 
 are attached to passenger trains, and many of the railroad companies, appre- 
 ciating the benefits arising from stocking waters along their lines, haul the cars 
 gratis; others make special rates for cars and crew. When plants are made off 
 the main railroad lines, the fish are carried in baggage cars in charge of members 
 of the car crew. 
 
 In addition, distributions are made from many stations without the aid of 
 the special cars, the station employees caring for the fish in baggage cars. For 
 this service the railroad companies usually charge regular fare for the attendants 
 but transport the fish and return the empty cans gratis. 
 
 Fry and young fish are usually transported in lo-gallon round-shouldered 
 iron cans, tinned for fresh-water work and galvanized for marine work. With 
 fresh water galvanized iron often proves toxic, and should never be used in the 
 transportation of fresh-water fishes. On the cars it is possible to aerate the 
 water in each fish can by air pumped through a reservoir, from which it is taken 
 through lines of piping along the sides of the car. The piping is equipped with 
 pet cocks, from which the air is carried to the cans in 5 s-inch rubber tubing and 
 forced into the water through liberators made of porous wood, preferably the 
 American linden {Tilia americana) , placed in hard rubber holders. For best 
 results these liberators must not be placed in the water until air pressure is on 
 and must be removed when air pressure is stopped. 
 
 In putting up cod fry and fry of other marine fishes for shipment it is cus- 
 tomary to have several quarts of water in the transportation cans and then 
 carefully dip the fry from the hatching boxes, lowering the dipper to the water 
 in the cans before emptying it. When the box is nearly empty the remaining 
 fry are removed from it to a can by means of a siphon. When transported on 
 a vessel having a conveniently arranged well, linen scrim is securely fastened 
 over the top of each can containing fry and the cans are laid horizontally in the 
 well, the top toward a perforated supply pipe through which water is pumped 
 into the well, thus maintaining a constant current. 
 
 Lobster fry may be carried in scrim- walled containers, or boxes, suspended 
 in the well of a vessel, the motion of the vessel and the constant circulation of 
 water in the well keeping the fry in good condition and preventing their settling in 
 a mass at the bottom. The boxes are made of a framework covered on the four 
 sides and bottom with scrim, which allows a free circulation from all sides. 
 Each box, 42 inches by 29 inches by 29 inches, will hold from 2,000,000 to 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 
 
 753 
 
 3,000,000 fry. It is customary to take from 12 to 15 cans of fry in addition to 
 those taken in the well, the fry in the cans being the first planted. 
 
 On vessels having no wells, aeration for all species of the marine fry, includ- 
 ing the lobsters, is accomplished by siphoning off and dipping the water freely, 
 as will be described for whitefish, pike perch, etc. For tempering the water 
 ice is suspended from the cover of the can in a cylinder. 
 
 Temperature is a most important consideration in the transportation of 
 fishes, and owing to the different conditions under which they are hatched in 
 the various localities, is a feature that requires skilled discretion on the part of 
 the attendant. The general rule is to keep the temperature at least as low as 
 that of the water from which the fish were taken, and lower if the species is not 
 sensitive to changes. 
 
 The maximum number of fishes to be carried most advantageously in a 
 lo-gallon can is another equally important question. The distance to be trav- 
 eled partly prescribes this number, but must be considered also with reference 
 to the temperature; and these factors, interdependent as they are, go to prevent 
 the formulation of any hard and fast rule. In the following table, however, 
 attempt is made to generalize by means of average conditions, and show the 
 number of fishes of specified kind and size as ordinarily transported in a 10- 
 gallon can. The cooler period of the year in which handled accounts for the 
 lower temperature for fry of the trouts as compared with the higher temperature 
 of fingerlings. It also accounts for the lower temperature for landlocked salmon 
 fingerlings as compared with fry. Ordinarily if landlocked salmon, rainbow 
 trout, and brook trout fry from the same source should be distributed in warm 
 weather it would be desirable to reduce the water temperature for the brook 
 trout considerably lower than for the other two species. 
 
 Number op Fishes oV Given Kinds and Sizes to be Transported in a io-Gallon Can under 
 
 Average Conditions. 
 
 
 Fry. 
 
 Advanced fry. 
 
 Fingerlings no. i. 
 
 Fingerlings no. 2. 
 
 Species. 
 
 Number. 
 
 .Tempera- 
 ture. 
 
 Number. 
 
 Tempera- 
 ture. 
 
 Number. 
 
 Tempera- 
 ture. 
 
 Number. 
 
 Tempera- 
 ture. 
 
 Large-mouth and small- 
 mouth black bass 
 
 Rock bass, bream, and 
 
 3. 000 
 
 ' F. 
 60 
 
 1, 000 
 
 " F. 
 
 6! 
 
 400 
 
 400 
 I, 500 
 1.500 
 
 I, 2O0 
 I. 500 
 
 62 
 
 62 
 so 
 48 
 S3 
 
 44 
 
 225 
 
 225 
 500 
 
 500 
 500 
 500 
 
 " F. 
 
 63 
 
 63 
 so 
 so 
 55 
 SO 
 
 
 S.ooo 
 
 S. 000 
 
 5.000 
 
 4. 000 
 
 2. 000 
 
 2, 000 
 
 30, 000 
 
 100, 000 
 
 125. 000 
 
 200, 000 
 
 30. 000 
 
 300.000 
 
 I, 000. 000 
 
 400. 000 
 
 100. 000 
 
 40 
 48 
 48 
 40 
 45 
 48 
 38 
 60 
 50 
 S3 
 68 
 42 
 43 
 
 3. 000 
 3.000 
 1. 500 
 3. 000 
 I, 000 
 I, 000 
 
 45 
 48 
 SO 
 42 
 48 
 48 
 
 Blackspotted trout 
 
 
 
 f ^nHIocWprt «;almnn 
 
 
 
 400 
 
 44 
 
 Whitefisb 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 200 
 
 60 
 
 White perch 
 
 
 
 
 
 
 Shad 
 
 
 
 
 
 
 
 Cod . --- 
 
 
 
 
 
 
 
 Flatfish 
 
 
 
 
 
 
 
 Pollock 
 
 
 
 
 
 
 
 
 S8 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 B. B. F. 1908 — 48 
 
754 
 
 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 Number of Fishes of Given Kinds and Sizes to be Transported in a io-Gallon Can under 
 
 Average Conditions — Continued. 
 
 
 Fingerhngs no. 3. 
 
 Fingerlings no. 4. 
 
 Fingerlings no. 5. 
 
 Fingerlings no. 6. 
 
 Species. 
 
 Number. 
 
 Tempera- 
 ture, 
 
 Number. 
 
 Tempera- 
 ture. 
 
 Number. 
 
 Tempera- 
 ture. 
 
 Number. 
 
 Tempera- 
 ture. 
 
 Large-mouth and small- 
 mouth black bass 
 
 Rock bass, bream, and 
 
 ISO 
 
 150 
 250 
 200 
 150 
 ISO 
 
 ° F. 
 63 
 
 63 
 
 SI 
 55 
 44 
 60 
 
 100 
 
 100 
 125 
 100 
 100 
 100 
 
 <■ F. 
 62 
 
 63 
 
 52 
 
 SS 
 44 
 60 
 
 50 
 50 
 
 61 
 
 61 
 
 30 
 30 
 
 ■> F. 
 
 61 
 
 61 
 
 
 
 
 75 
 
 54 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Note. — The varying usage in the classification of young fish as to size has caused such confusion 
 and difficulty that the Bureau has adopted uniform definitions, as follows: 
 
 Fry = iish up to the time the yolk sac is absorbed and feeding begins. 
 
 Advanced /ry = fish from the end of the fry period until they have reached a length of i inch. 
 
 Fingerlings =&sh between the length of i inch and the yearling stage, the various sizes to be desig- 
 nated as follows: No. 1, a fish i inch in length and up to 2 inches; no. 2, a fish 2 inches in length and 
 up to 3 inches; no. 3, a fish 3 inches in length and up to 4 inches, etc. 
 
 Yearlings=6sh that are i year old, but less than 2 years old from the date of hatching; these may 
 be designated no. i, no. 2, no. 3, etc., after the plan prescribed for fingerlings. 
 
 In the care of fish away from the air circulation of the cars, aeration is 
 usually accomplished by the use of a long-handled dipper, and this method for 
 10 or 15 cans, which is the usual number handled by detached messengers, if 
 energetically operated has no equal. Several forms of aerating devices have 
 their good points, their use is permitted, and efforts are being made to improve 
 upon the present methods of hand aeration in order to lighten the labors of the 
 messengers, who not infrequently care for the fish for two or three days before 
 they arrive at a destination. 
 
 For aeration in cans containing small fry, such as the shad, pike perch, 
 white perch, yellow perch, and whitefish, it is customary to siphon a portion of 
 the water into a pail (the head of the siphon being in a wire cage covered with 
 cheese cloth), aerate it with a dipper, add ice to temper it if necessary, and then 
 pour it back through a large funnel which reaches nearly to the bottom of the 
 can, the lower part of the funnel for about 6 inches being made of perforated 
 tin to break the force of the water. On short trips little, if any, aeration is 
 necessary. 
 
 Cans holding the larger fry and young fish may be aerated by dipping the 
 water up and pouring it back again from a height of 15 or 20 inches. When 
 transportation is by water it is customary to renew the water in the cans en route 
 instead of resorting to hand aeration. With fishes of the Salmonidae family, 
 ice may be placed directly in the water in the cans. 
 
FISH-CULTURAL PRACTICES IN THE BUREAU OF FISHERIES. 755 
 
 DISTRIBUTION AND PLANTING OF THE COMMERCIAL FISHES. 
 
 It being infeasible for various reasons to rear the young of the commercial 
 fishes, practically all are planted as fry and the distributions are usually made 
 by agents of the Bureau. 
 
 Fry of the marine species are often transferred to small boats in order to 
 plant in shallow water or they are carried from the hatchery in launches. In 
 such cases the cans are lowered into the water by two men and there slowly 
 inverted. Wlien planting from a vessel two lines are made fast to the can, one 
 about the top and one about the bottom; by this means the can is lowered 
 over the side and when partially submerged is emptied. 
 
 On the Great Lakes the distributions of lake trout, whitefish, and pike 
 perch are usually made by means of steam vessels. In such instances the water 
 in the cans is renewed as often as necessary by siphoning off a portion and 
 replenishing directly from the lake. The manner of liberating the fish after the 
 point of deposit has been reached varies in practice. In planting lake trout and 
 whitefish the method employed by the superintendent of the Duluth, Minn., 
 station is to pour the water and fry, one can at a time, into a large tub full of 
 water, from which the water and fish are siphoned through a heavy 2 -inch 
 rubber hose attached to a pole or outrigger. Thus they are deposited in the 
 lake about 8 feet from the side of the vessel, very close to if not beneath the 
 surface of the lake, the speed of the vessel being slackened while the planting 
 is in progress. As it is customary to transport the fry on passenger or package 
 freight steamers, where they are necessarily stowed in a limited space and as 
 close to one gangway as possible, the cans may be emptied into a tub sitting 
 firmly on the deck more easily and expeditiously than they could be poured 
 directly into the lake, and the element of danger to the men doing the work is 
 avoided also. 
 
 The superintendent of the Northville station follows a somewhat similar 
 procedure in planting lake trout, pike perch, and whitefish fry. If the deck of 
 the vessel is near the water surface a piece of ordinary tin pipe is attached to a 
 tub and arranged with elbows so as to bring the lower end near the surface of 
 the lake. If the deck of the vessel is high above the water the tub is used with 
 1 5 or 20 feet of 2-inch fire hose instead of the tin pipe, a weight being attached 
 to the lower end. 
 
 From the Put-in Bay station the fry are planted by pouring them over the 
 sides of the vessel as it moves slowly along, and the superintendent of this 
 station, from experiments he has made, believes this method preferable to the 
 use of hose or tin conductor as just described. 
 
756 BULLETIN OF THE BUREAU OF FISHERIES. 
 
 None of the salmon stations on the Pacific coast is of sufficient capacity to 
 hold more than lo per cent of the fry until sac absorption, and in California a 
 large portion of the product is distributed to hatcheries operated by the Cali- 
 fornia Fish Commission. The portion distributed by the Bureau from the 
 Baird station, however, is held tmtil sac absorption. The method of distributing 
 from Baird is as follows: 
 
 The inflow to a trough is shut off, the drain pipe removed, and the water 
 and fry allowed to pass into a double receptacle consisting of a perforated bucket 
 inside a regular 5-gallon spawn bucket, the inner container being about i inch 
 less in diameter and raised an inch from the bottom by wooden blocks. This 
 receptacle has been filled with water, to prevent injury to the fry as they are 
 poured in, and the surplus water escapes through the perforations and over the 
 rim of the outside bucket, leaving the fry in the center. If the trough is some 
 distance from the floor a box is used to elevate the buckets to proper level, and 
 any fry remaining in the trough after the passage of the water are brushed 
 through the opening with a broad, flat paint brush. By this method the fry 
 can be removed from a trough in two minutes with absolutely no loss. From 
 the bucket the fry can be easily poured into a lo-gallon can, but frequently they 
 are carried to the river (about 100 feet from the hatchery) and planted direct 
 from the buckets. So far as possible the plants are made during flood water 
 and always where there is a strong current. By this means few, if any, free 
 swimmers are caught by the ever-present trout, as their natural tendency 
 quickly to scatter is facilitated by swift water. 
 
 DISTRIBUTION OF GAMU FISHES. 
 
 The Bureau does not as a rule attempt to plant the game fishes produced 
 at its hatcheries, but consigns them to individuals, anglers' clubs, protective 
 associations, etc., by whom they are used to stock both public and private 
 waters. It is customary to deliver the fish free of charge to the applicants at 
 the railroad stations nearest the point of deposit. 
 
 The number of fish allotted to individual applicants is, of course, largely 
 determined by the supply available, which depends to great extent upon the 
 difference in methods of hatching applicable to the different species. The area 
 and character of the water to be stocked must also be considered, of course. 
 Moreover, the same water area that would receive a million pike perch fry would 
 perhaps be assigned no more than 200 or 300 black bass 3 or 4 inches long, or four 
 to eight times that many if the bass are planted as fry. The explanation is in 
 the fact that pike perch can be propagated by the hundred million, while black 
 bass, hatched by other methods, or collected from overflowed lands, are pro- 
 
FISH-CULTURAI. PRACTICES IN THE BUREAU OF FISHERIES. 757 
 
 duced only in comparatively small numbers. The Bureau does not attempt 
 to assign any applicant more than a liberal brood stock of the basses or sun- 
 fishes. With brook trout, which are distributed both as fry and fingerlings, 
 assignments of fry are twenty-five to fifty times larger than assignments of 
 fingerlings 3 to 4 inches long. 
 
 Applicants for fish are advised by mail of the approximate date on which 
 the fish will be shipped and later by wire of the hour on which they may be 
 expected to arrive. The advance mail notice also contains the specific instruc- 
 tions for the care of the fish from the time of delivery until they are planted. 
 
LIBRARY OF CONGRESS