Jttly 1, 1910] 



SCIENCE 



29 



that the average size of any lot of one species 

 of fish is very largely affected by the character 

 of the water from which the lot was obtained. 

 What the particular factors in any body of 

 water are that afPect this size is practically 

 unknown. A very interesting case of this 

 kind is found, however, in the common yellow 

 perch of our lakes here at Madison. Lake 

 Mendota, the largest lake, contains more 

 perch per unit of area than any other body of 

 water I have ever examined. But they are all 

 small. Lake Monona, a smaller and somewhat 

 shallower body of water, but freely connected 

 to Mendota, which flows into it, contains a very 

 much smaller number of this species, but they 

 will average twice as large in size. The ex- 

 planation, so it seems to me, is fairly clear: 

 Lake Mendota (for reasons now unknown, al- 

 though a greater relative plankton content 

 probably plays a part), is more favorable for 

 the hatching of perch spawn and the adequate 

 nourishment of the young perch, as well as 

 for their protection, or freedom from enemies. 

 The absence of Stizostedion may be a factor 

 here. As a consequence a relatively large 

 proportion of the eggs laid develop to a stage 

 where the perch need larger organisms for 

 food. This very abundance of young perch 

 brings about a struggle for the food supply, 

 a struggle which results not so much in the 

 extinction of the weaker individuals, but in a 

 reduction of the amount of food obtained by 

 each individual and hence a reduction in the 

 rate of growth. Such a state of semi-starva- 

 tion probably has little or no effect on the 

 fecundity of the individuals concerned, for 

 as we now know, the essential reproductive 

 organs are about the last parts of the body to 

 be affected by starvation.' 



In Lake Monona, on the other hand, we may 

 suppose the conditions for the development of 

 perch eggs, and for the proper nourishment 

 of young perch and their inrmunity from 

 enemies to be much less favorable. Hence a 

 much smaller number, as compared with the 

 number of eggs laid, would reach the stage 



' Stoppenbrink, Zeitschr. tciss. Zool., Bd., 79, 

 p. 496; Schultz, Archiv f. Entw. Mechanik, Bd. 

 lb, p. 555. 



when larger food is taken. The much smaller 

 number struggling for the supply of this 

 larger food, would allow each individual a 

 much larger share, and hence a much more 

 rapid rate of growth, which would of course 

 finally result in a much larger average size. 

 That great differences in size among individ- 

 uals of the same age may be produced thus 

 has been well shown in star fishes, where the 

 disproportion is sometimes startlingly great.' 

 That such great differences may be possible 

 demands great resistance on part of the ani- 

 mals against partial starvation, and this we 

 know to exist in fishes as in most other poikilo- 

 thermous vertebrates. 



Similar observations among fishes, more 

 particularly carnivorous fishes, can be made 

 by any one. We believe that in the majority 

 of cases a lake yielding regularly large indi- 

 viduals of a species (large-mouthed black 

 bass for instance) will not yield the species 

 in great numbers, while one yielding many 

 individuals will rarely yield any above aver- 

 age size. This is corroborated in a general 

 way in our collections of Gasterosteidse from 

 Wisconsin, and is very apparent indeed to any 

 one actually collecting them. The character 

 of size, therefore, can be set aside as of no 

 value whatsoever. 



The relation of body depth to total length 

 has already been disposed of. To reach some 

 conclusion concerning the dorsal and anal fin 

 formula, I selected an individual of approxi- 

 mately the size of Agassiz's largest, eleven 

 sixteenths of an inch. I measured to the base 

 of the tail in order to give his description the 

 benefit of the doubt. Even so, it must be 

 admitted that it is a very small specimen to 

 work with. By means of a binocular magni- 

 fying sixty-five diameters, it is comparatively 

 simple to erect the dorsal and anal fins, and 

 count the presence of ten soft rays in each. 

 But Agassiz had no binocular, nor had he the 

 better preserved specimens which result from 

 killing in formalin. Trying to examine this 

 specimen by the ordinary microscope imme- 

 diately shows how difficult it would have been 

 thus to determine the right number of soft 



° Mead, Am. Naturalist, Vol. 34, p. 17. 



