NO. 1431. BREEDING HABITS AND EGG OF PIPEFISH— GUDGEB. 453 



armature grows downward to help form the sides of this pouch (so in 

 S. floridir), and it is hard to understand how this can undergo the 

 changes alcove noted. 



Rathke thinks that since the anus in his fishes (aS'. rariegatm, huecu- 

 lentus^ and argentatus) is inclosed in the upper end of the pouch, the 

 eggs glide out of the oviduct and into the pouch accompanied by an 

 albuminous fluid, which on contact with the water cements the lips of 

 the pouch together. He finds that the interior of the pouch is like a 

 '■'.schlelmhaut,-^ and that finally, through the great development of 

 thet-apillaries, it becomes ""like an inflamed mucous membrane." 



In the ovaries, lying in an albuminous fluid, he finds large tvh/te 

 cells, which when put into water become tightly stretched. In some 

 individuals with cells, like the above, free in the lumen of the ovary, 

 he finds not the least trace of a pouch; others have the skin under the 

 tail very much thickened into angles at the outside, and others have 

 broad folds. Hence he concludes that the ripening of the eggs and 

 the formation of the pouch keep pace with one another. 



Rathke thinks Eckstroem's discoveries need confirmation, since no 

 other fish in the world possesses such a peculiar testis. He positively 

 affirms that, even if his opponent be correct, the females at the breed- 

 ing season possess the rudiments of a pouch. His great objections to 

 Eckstroem's discover}^ are (1) that the fishes have no organs to hold 

 themselves together during the transfer; (2) that he can not conceive 

 how the skin folds can open for the reception of eggs and close again, 

 nor how the brood cavity can become filled with eggs to the very end. 

 My own discoveries make these points clear. 



Rathke confirms the Swedish naturalist that, in addition to the yolk, 

 the liquid filling the brood pouch serves as nourishment for the 

 embryos, and thinks that they absorb it through both skin and mouth. 

 His description of the development of the larva^ is very full and cor- 

 rect. Noteworthy is his discovery that at first the entire operculum 

 is free and that it begins to grow fast to the other parts in the antero- 

 ventral region and the closing proceeds posteriori}^ and dorsal ly. 



In 1838, Valentin (reference from Marcusen not verified) described 

 females bearing pouches, thus confirming Rathke. In the same year 

 Fries, without entering into the controversy, accepted Eckstroem's 

 results. He put a male SyngnatJms luiiihricifornds having eggs, with 

 young outlined (48 to 60 hours old, probably) and cemented onto the 

 bell}^ into an aquarium, and on the ninth day thereafter some young 

 were hatched and on the next day the others. These lived seven days, 

 and in that time nearly doubled tljleir length. 



The adult fish has neither pectorals nor caudal and the rounded tail 

 is prehensile, the body is densely pigmented, and the operculum is 

 bound down to the shoulder girdle, leaving only a small dorsal open- 

 ing. Fries, however, figures and describes the newly hatched young, 



