442 REPOET OF COMMISSIONER OF FISH AND FISHERIES. 

 Table of distribution of shad and eels. 



21. — MODE OF ESTIMATINQ NinMBERS OF EGGS AND FISH. 



The estimation of the numbers of sliad-eggs and of the young fry is a 

 rather difficult matter to accomplish satisfactorily or with even approxi- 

 mate defiiiiteuess. The standard made use of on the Connecticut River 

 is, without doubt, very much too high. Mr. 0. C. Smith, who is a most faith- 

 ful and successful breeder, has adopted the exaggerated estimates of his 

 predecessors, and in the earlier years of his work, when the restoration 

 of the shad by artificial propagation was an experiment, the contraction 

 of the round numbers that had appeared in the reports of previous years 

 might have discouraged the people and the legislators, and the support 

 that was so necessary to this initiation of the work might have been with- 

 drawn. But now that the experiment has proven itself so thoroughly 

 and evidently a success, and the river is again teeming with shad, it 

 makes very little difference to the citizens of the State whether it has 

 been accomplished by placing 10,000,000 or 90,000,000 of young shad in 

 the waters, and a system of measurements and counts should be carried 

 out so that some more definite knowledge of the number of eggs handled 

 may be attained. One of the commissioners of the State, in 1870, feel- 

 ing dissatisfied with the estimates afforded, attempted to induce more 

 careful modes of numbering the eggs and fish, and in the following year 

 succeeded in having them modified to a certain extent. The numbers 

 of fish we carried in the six cans were, by these standards, millions, 

 while the most careful means of numbering we could employ did not 

 place them over from twelve to twenty thousand to the can. 



In estimating the numbers of fish in the cans, we had, as our sole reli- 

 ance for accuracy, only their aj^parent thickness in the water, which, 

 after considerable experience, afforded an approximate estimate within 

 a margin of a few thousands of the real number. 



