I90 SMITHSONIAN MISCELLANEOUS COLLECTIONS [vol. 50 



to cause the whole mass of eggs to rock visibly backwards and for- 

 wards. This action was done at brief intervals and from this time 

 onwards. Later, when the eggs were hatching, it was redoubled, 

 and great activity was shown with the fins. The movement thus 

 created in the water very probably helped the escape of the larval 

 fishes from the eggs. At this time the 'pumping' or 'blowing' action 

 was at the rate of fifteen or sixteen in ten seconds, and in the pauses 

 the fins were kept vigorously at work." 



For more than a month this guardianship of the eggs was kept 

 up and for most of the time no food whatever was taken. If any 

 was dropped near him or on the eggs, such as mussels, he would 

 remove them ; a little crab dropped on the eggs was, after some 

 effort, caught and carried away to be dropped instead of eaten, as 

 already told. This vigilance and restraint at length changed his 

 appearance. He "lost his brightness" and became dingy ; naturally 

 he also became thin and was infected with ecto-parasites (Caligi) 

 and appeared sometimes exhausted by his onerous task and pro- 

 longed fast. But "that this was not entirely due to those causes was 

 shown when the supply of water was increased, and when it was 

 directed to his corner. After a refreshment of this kind he movea 

 about with vigor, energetically spouting water on the eggs and 

 fanning them with his fins." Not until the 26th of April was the 

 watchful male persuaded to take a mussel, but after that "on some 

 days he ate as many as five ; any excess he carried off and ejected, as 

 before ; and at the beginning of May he was as alert, active, and 

 pugnacious as ever." Once persuaded, he occasionally accepted food 

 several times afterwards. 



At last, on the 5th of May, "42 days after the eggs were deposited 

 and fertilized," larvae began to issue from the eggs ; for the first 

 few days "the tadpole-like larval lumpsuckers were found in small 

 numbers in the overflow-filter every morning, and they slowly in- 

 creased in numbers. They were very active, swimming with great 

 rapidity by a lashing movement of the tail, a large yolk containing 

 an oil globule at the right side being conspicuous. Up to the 22d 

 of May, or almost exactly two months from the time the eggs had 

 been spawned, and seventeen days after they had begun to hatch, 

 the conditions described continued. The young lumpsuckers Merc 

 appearing in greater numbers, but still not in such abundance as one 

 might have expected. The largest number was about two or three 

 hundred a day. They were also to be seen adhering to the glass 

 front of the tank, and numbers were thus accounted for. None were 



