VARIATION EST SIZE OF EGGS. 25 



undoubtedly differ in size. The following figures -which will be more fully entered 

 upon when considering the Tarions species, are here given from measurements 

 made at Howietoun. Diameter of eggs of grilse 0'20 to 0-22 of an inch ; of a 

 16 lb. salmon, 0-24 in.; Lochleven trout at Howietoun, 2 and 3-year-old, 0-17 in. ; 

 6-year-old, 0'18 to 0-19 in. 5 8-year-old, 0-20 to 0-24 in. American char, 2-year- 

 old, 0-14 in. ; 3-year-old, 0-17 in. ; 4-year-old, O'lS in. 



The above are only some out of the many measurements which I have made, 

 and show very conclusively that the diameter of the eggs in salmonoids increases, 

 at least up to a certain period, with the age and probably the condition of the 

 parents ; thus the ova of grilse are inferior in size to those of the salmon. And in 

 large trout as in the older Lochlevens there is an augmentation in the thickness 

 of the shell, a rather important question to the fish-culturist, because such shells 

 require a strong embryo in order to readily hatch ; this, however, may be obtained 

 by giving an increased supply of water during their eyed period. It was also 

 found, as a rule, that treat eggs obtained latest in the season were of rather 

 larger size than such as had been taken in the earlier months, while in nature 

 these larger eggs generally produce more rapidly growing young and thus to a 

 certain extent counterbalance any bad effects from late hatching. 



Next is the consideration of whether any difference is perceptible in the eggs 

 of two parents of the same age that have been kept under identical conditions ? 

 Among the young salmon reared at Howietoun from eggs obtained from the same 

 parents, and reared in the same pond, some taken on December 1st, and agaia on 

 the 13th, 1884, averaged 0-22 of an inch; some from similar fish on December 9th 

 averaged 0'20 of an inch, while others taken the same day were 0"22 of an inch. 

 The same phenomenon has been observed among the Lochleven trout, wherein, 

 although the diameter of the eggs of fish at eight years of age was generally 

 0-24 of an inch, in some it was 0'22, in a few 0-21, and in three or four 0'20 ; also 

 in six-year-olds it has been observed to vary from 0-18 to 0-19 of an inch. That 

 such variations might theoretically be anticipated most of us must admit, for 

 whatever causes deficiency of nutrition might occasion diminution in the size of 

 the ova. Also the size of the parent ought possibly to be taken into account, for 

 disease, crowding them when young, and other causes may dwarf these fishes, and 

 such would probably give comparatively small eggs. 



Can we detect any variation in the size of the eggs given by a single fish ? I 

 have already adduced instances of this among salmon or rather grilse and Loch- 

 leven trout. In November, 1884, I placed 80 unimpregnated eggs of one of these 

 fish in a tub of water, wherein they were kept for a week, the water being 



fishes, as the size of the ova is not only invariably the same in individuals of whatever size, but, 

 as far as our experience reaches, is even often characteristic of the species of a genus." This 

 opinion, slightly modified, Dr. Gunther still apparently held in 1880, for he asserted (Introduction 

 to the Study of Fishes, p. 159) that " the ova of Teleostean fishes are extremely variable in size, 

 quite independently of the size of the parent species. The ova of large and small individuals of 

 the same species, of course, do not difier in size." 



On the other hand, Mr. B. Elanohard in 1866 (Poissons des Eaux Douees de la France, p. 461) 

 observed that the ova " of the grilse are always sensibly smaller than those of the adult Salmon." 

 Livingston-Stone, 1877 {Domesticated Trout, 3rd edition), remarked that in American Trout 

 (Salmo fontinalis) that reside in spring- water, which is equivalent to a diminished supply of food, 

 smaller eggs are developed than in such as reside in brooks. On the MoCloud river (Report of 

 United States Fish Commission) ' ' it was noted, in 1878, that the parent Salmon were unusually small, 

 their average weight being under 8 lb." This small size was stated to be undoubtedly caused, 

 in whole or in part, by the fishery at the canneries of the Sacramento, where the 8-inoh meshes of 

 the innumerable drift-nets stopped all the large Salmon, but let aU the small ones through. The 

 eggs when taken proved to be at least one-third smaller than those of most previous years, and 

 the average number of eggs to the fish was about 3500 against 4200 in the previous year. 

 Livingston-Stone in 1882 (Bulletin of the United States Fish Commission, ii, p. 11) writing respect- 

 ing the eggs of Salmo fontinalis, observed that those from the small fish of the elevated rivulets 

 are not so large as those of the finer breeds from warmer streams, concluding that the ova of these 

 last are fully twice as large as those found in the race residing in mountain rivulets. Also that 

 " large eggs are the result of keeping the breeders in water that warms up in the spring and 

 summer. It is true, if it becomes too warm, say above 70°, it is injurious ; but water that stands 

 at 65° in the summer will make larger eggs than water at 55°, and very cold spring water will 

 always develop small eggs." (Page 183.) 



