132 



MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



while the two rami are more as in Petalophthalmus, though broader. The other biramous abdominal 

 appendages in the Gampsonychidte are truly schizopodal. 



Fig. 3o.— Telson of Petii- 

 lophthlamuscf. 



no. 3.— Petalolplithalmns armatascT. 



Fig, 4. — PetalopMhalmns armatns 9 . TMs and Kg. 3 after TV. Snhm. 



Fig. 4(1. — Second gna- 

 thopod of Petalopbtlialmua 

 9. I, breeding lamella. 



F I G. 46. — Third 



{>ereiopod of Peta- 

 ophthalmus 9 • 



Classifying the Schizopoda by the carapace, modifying Willemoes-Suhm's table by throwing 

 out the Nebaliadse and substituting the Gampsonychidae, there would seem to be three groups, as 

 follows : 



I. Carapace absent (Gampsonychidae). 



II. Carapace free, varying in size (Gnathophansia, Petalopthalmiis and Chaiaraspis). 



III. Carapace fastened to tlie tliorax (Mysis, Lopliogaster and Euphausia). 



But I should agree with Willemoe.s-Suhm that this is not a natural genealogical classification, 

 and throwing out the Nebaliadse, which, as we have endeavored to show, belong to a distinct order 

 of Cru.stacea, the families of Schizopods may be enumerated thus (after adding the Gampsonychidse 

 to von Suhm's table), all having seven abdominal segments : 



Carapace absent, sis pairs of tboracic legs I Gampsonychidae. 



Carapace well developed, six pairs of tlioracic legs II Mysidfe. 



Carapace well developed, eight pairs of tboracic legs Ill Eupbansiidas. 



Carapace well developed, four pairs of tboracic logs IV Cbalaraspidae. 



Carapace well developed, seven pairs of thoracic legs V Lophogastridoe. 



When we compare the Gampsonychidse with the Syncarida (Acanthotelson), we see that both 

 groups have the same number of body-segments, and that both lack a carapace; and thus, while 

 the Gamp.sonychidae are the ancestors of living Schizopods, the group as a whole probably de- 



