72 T-a'ciify-sci'ciifh Annual Meeting 



The practical application of this principle is as follows: If 

 it proposed to test the result of re-stocking a locality in which a 

 species of fish has become reduced in numbers, it is necessary to 

 first determine the "curve of distribution" from fish native to 

 the locality. This curve may be based on any measurable char- 

 acter, such as the number of fin-rays, the number of scale-rows, 

 or the number of vertabrae. When this has been done, it is 

 then necessary to .determine the "curve of distribution" for the 

 same structural character of fishes of the same species, but 

 abundantly found in another locality, from which locality the 

 "brood fish" are to be taken. After the "planted fish" have had 

 time to mature, new curves should be plotted for the first local- 

 ity. If these curves are practically the same as those originally 

 made, it is reasonnable to conclude that re-stocking has been 

 ineffectual. If, however, the curve of the original localitv be- 

 comes modified and approaches that of the second locality (that 

 is, the locality from which the brood fish were taken), it is rea- 

 sonable to conclude that the influence of the fish new to the 

 locality has been felt, and that the re-stocking has been effectual. 



The following objections may be raised to the method just 

 given : 



1. It may be that due to the small number of specimens, 

 the curve represented on the first diagram is not really charac- 

 teristic of the Woods Hole specimens. — To test this source of 

 possible error, three separate groups of flatfish were examined, 

 all from the same locality, and each group containing one hun- 

 dred speciments. The resulting curves were strikingly alike. 

 (Of course it would be much more satisfactory to base all the 

 curves on the enumeration of one thousand rather than one hun- 



• dred specimens, but even one hundred specimens evidently yield 

 fairly definite results, though, to be sure, the curves are some- 

 what uneven.) 



2. It may be that the variation in the position of the curves 

 on the two diagrams is a result of age — i. e., the fishes from 

 Woods Hole average a larger number of fin-rays simply be- 

 cause they are somewhat older. This possible increase on the 

 part of the older specimens, if present, can readily be detected 

 by simply comparing the average number of fin-rays of the 

 younger with the average number of fin-rays of the older fish. 

 Fifty-three' young, less than ten inches in length, have a math- 

 ematical average of 66 dorsal fin-rays ; forty-seven older fishes 

 from the same locality, all over ten inches in length, average 



