No.im. BREEDING HABITS AND EGG OF PIPEFISH— GVDGER. 457 



(3) Females of these two species without pouches were found. 



(4) Males of S. hncculentus have pouches; females never do. 



(5) Males only of Scyj>}iieiis feres possess the pouch. 



Thus was the work of Rathke corrected in part, confirmed in part, 

 and wholly cleared up. It may be well to say here that, in hundreds 

 of pipefishes at Beaufort, males without and females with pouches 

 have never been found by the writer. 



In 1874, Dufosse described how sea-horses under his ol)servation in 

 1854 held themselves tightly together by their twisted tails. Observa- 

 tions made in the year of publication showed that, while thus held, 

 the female passes the eggs into the pouch of the male. Dufosse noted 

 that at tliis time the pouch possesses many thick folds, which secrete 

 a mucus for the nourishment of the young. He seemed to have been 

 wholly ignorant of the work of his compatriot, Lafont. 



In May of the same year Fanzago, working in the Zoological Station 

 at Naples, independently made the same disco ver.3'. He writes that 

 the sea-horses make use of their prehensile tails as an aid in the act of 

 coition. A few eggs only, perhaps just one, are passed at a time, 

 hence the coition must be repeated. The male apparentl}^ is passive 

 and invites the female to introduce the oviduct into the mouth of the 

 pouch. Contact is short and is repeated five or more times in a short 

 while. As will be seen later, in S. florid x there is a sexual embrace 

 in which both animals are active. 



A. H. Malm, in his inaugural dissertation at Lund, in 1874, finds 

 no continuous fin-fold in S. typhh\ but states that the tail is at first 

 protocercal, secondh^ heterocercal, and finall}^ homocercal by resorp- 

 tion of the end of the notochord. Malm agrees with Eckstroem that 

 the transfer takes place in deep water, and thinks with Kroyer that 

 fertilization takes place after transfer. He found a young male 90 

 nmi. (8.() inches) long with a pouch, and another 140 mm. (5.0 inches) 

 long with eggs. It is noteworthy that Malm concludes that the 

 "slime'' in the pouch is identical with that on the body, Ijut, pro- 

 tected l)y the pouch, it is not washed away; thus in a sense he antici- 

 pates both Huot and Cohn, but he does not think that it is used for food. 



At Kiel, Heincke (1880) found that in. S. typlile the females are 

 larger and more numerous. Both these points hold good for the pipe- 

 fishes of Beaufort, the proportionate numl)ers being about three males 

 to every seven females. In ^S. typhJe the pouch is not filled at one 

 time, but there may be several ti-ansfers extending over several days. 

 This is true of A', florldiv^ sometimes eggs of three difi'erent stages 

 being found in the same pouch. For the period of gestation, Heincke, 

 not knowing the ages of the eggs at the beginning, fixes a mininuun 

 period of fourteen daj's. As will be seen later, the period for 8. 

 florldie seems to ])e ten days. Breeding in -S', fy/>/ile takes place from 

 May to August; the pouch is not resor])ed and the young do not go 



