MARSUPIAL FISHES. 67 



the mainmalia, and in some genera of sbarKs; an 

 observation which is as old as Aristotle, and wnich 

 Roiidelct represented in one of his j)lates. Stenonis, 

 in the year 1()73, described the onibryo of Gahus 

 Jcnvis as connected by means of the })lacenta ; and 

 Cnviersays briefly, that in the Carcharias the yolk- 

 sac is attached as firndy as a jdacenta. But on 

 this subject we must not enlarge ; and only repeat, 

 that very ani))lc details on these curious points will 

 be found in tlic pa^cs already referred to. 



Many of those who read these pages are familiar 

 with the fact, tliat there is one group of the mam- 

 malia in wliich the ovo-viviparoiis mode of deve- 

 lopment, as frequently stated, exists. — (Nat. Lib. 

 Mam. xi. 69.) We allude to the Marsupiata, includ- 

 ing the Kangaroos, Opossums, and other families. 

 Tn connexion with this subject, it is interesting to 

 know that there are true Marsuplalia in tlie class 

 of fishes; that is fish with a marsupium — a purse 

 or bag for the safe custody of their young, first in 

 th(! state of ova, and subsequently in that of fry, 

 which, from their premature extrusion, are altoge- 

 ther unable to take care of themselves. The ana- 

 logy so far is very striking. But a marked differ- 

 ence exists in this respect ; that whereas in the 

 mammalia the marsu])ium is in the female, in the 

 fishes it is found in the males. This provision 

 is met with in various species of the Synynathi^ 

 or Pipe-fishes, and also in the short round /////- 

 pocampus^ frequently called the Sea-horse. To 



