194 BULLETIN OF THE UNITED STATES FISH COMMISSION. 



developed. It is really these vascular pinnte wliich liave been exagger- 

 ated in development at the expense of the other portions of the branchial 

 apparatus. 



EARLY DEVELOPMENT. 



From what I have observed of the early stages of development of the 

 pipe-fish, Syngnathus pecMa^ius, and from a study of ova taken from the 

 pouch of a male Hippocamims j^reserved in alcohol, I offer the following 

 approximate account of the early phases of the evolution of the latter, 

 depending upon the former on account of its close relationship for the 

 details not actually observed. This will not permit us to develop more 

 than such points as we are w^arranted to infer from their close affiliation 

 to each other, but even such will be of value. 



The egg of Hippocampus, like that of other teleosts, is constituted of 

 a yelk and germinal material. The former is a rich orange yellow in 

 color; the latter cannot be described, as it has not been seen. In Syng- 

 nathus the yelk is of the same color, and embedded in it superficially 

 and all around it there are deep yellow oil globules. 



The blastoderm of Hippocampus is presumably formed, as in all other 

 known teleosts, by a gradual growth of the germinal disk over the yelk 

 so as to include the latter. The rudiment of the embryo appears at first 

 at the edge of the blastoderm, and develops for some time like other 

 fishes, such as the shad, cod, or stickleback. A segmentation cavity is 

 developed, which probably persists as I have observed in Syngnathus, 

 and a vitelline system of vessels is doubtless also formed as in the 

 latter. 



Up to the time the tail is about to bud out from the caudal swelling 

 at the end of the body of the embryo there is nothing observable which 

 would be considered remarkably different from the type of development 

 exhibited by less modified fishes. The tail of the embryo Lophobranch 

 buds out, and does not develop the prominent dorsal and ventral nata- 

 tory folds so characteristic of the first api)earance of the tail of the 

 embryos of the spiny and soft-rayed forms. It results from this, that 

 the tad is extended backwards, as development proceeds, as a simple 

 cylindrical prolongation of the hind jiortion of the body. There is, after 

 a while, iu Syngnathus a low fold developed where the dorsal and caudal 

 are to appear, but there is nothing like the wide natatory fold apparent, 

 such as we see in the embryos of Alosa, Gadus, and Cyhium of the same 

 age. In Hippocampus there is no caudal in the adult, and we may there- 

 fore expect to find little or no evidence of a caudal fin-fold at any period 

 of its development. 



The yelk-sack, I apprehend, is absorbed in the usual way, there being, 

 iu all probability, no direct connection of the yelk-sack with the intes- 

 tine. The period of incubation in the marsupium, from the fact that 

 development is pretty well advanced when the young leave it, I should 

 think would be not less than twelve to fourteen days. 



The peculiar elongation of the snout probably begins before the yelk- 



