ATKINS^ THE SALMON AND ITS ARTIFICIAL CULTURE. 255 



ting a large niiiuber of salmou pass up the maiu brook. Ou the morn- 

 ing of the 10th I was ai)priseil of their presence in a trout brook, to 

 which they had ascended over rapids that would have been quite inac- 

 cessible had it not been for the freshet. Repairing immediately to the 

 spot, with an assistant, I found the bottom of the brook, in every avail- 

 able spot, completely filled with their ridds, and salmon of both sexes, 

 spent or full, occupying all the pools. We -caught and examined nine- 

 teen female fish, of which six were wholly, and nine partly, spent. If 

 these fish had begun to spawn in this spot, six of them had completed 

 the whole i)rocess within forty-eight hours, or sixty at most. Tliere is 

 nothing conclusive about it, since it is by no means impossible that the 

 females found spent on the 10th had begun their spawning before the 

 storm of the 7th at some point on the shore of the pond, and feeling tiie 

 sudden rise of water, had left the ground where they had already laid a 

 portion of their eggs, and had traveled a mile or two up a small brook 

 in search of a place to deposit the remainder. 



The ratio of fecundation was ascertained in from twenty to thirty-five 

 days after the eggs were taken. The germ begins, some days before 

 that, to spread over the surface of the yolk, which it in the end com- 

 pletely envelops. During this process its advancing margin appears to 

 carry along with it a row of colored oil globules, which form a distinct 

 ring on the surface of the yolk. At first it enlarges as it advances, until 

 it has passed half way round the yolk, when it is at its largest size ; 

 from that moment it grows smaller, until it finally closes up. In spring- 

 water this phase of development begins at the tenth or twelfth day, and 

 is completed in eight or ten days. During its progress the plainly 

 marked ring of oil globules afitbrds an easy means of distinguishing a 

 fecund from an unfecund egg, since in the latter no such expansion 

 of the germ occurs. A very strong light should be thrown through the 

 eggs to see their condition clearly. This is best accomplished by plac- 

 ing them in a shallow metal box whose bottom is perforated with round 

 holes almost, but not quite, as large as the eggs, placing the latter over 

 the holes, and holding them so that the light shines througli them from 

 beneath. The box commonly used at Bucksport contains forty eggs, 

 and if one of them proves infecund, it is held to indicate a rate of 

 fecundation equal to 97^ per cent. If all of the forty are fecund, the 

 rate is 100 per cent. A sample of forty from each lot is thus tested, 

 and as there were in 1872 two hundred and fifty-seven lots, the average 

 result is believed to be very accurate. 



The eggs taken from full fish at the time of first handling them, and 

 treated in the ordinar3' way, were, in the majority of cases, fully 

 fecundated, and the average rate of fecundation obtained by the ordinary 

 way was 98 per cent. There were, however, thirty eight lots of this sort, 

 in which the test applied gave a proportion of in fecundated eggs. In the 

 most of these cases the rate of fecundation indicated was 97^ i)er cent. ; 

 in only three of them did it fall below 90 per cent., being 87^ in two 



