Classification of the Decapod Crustaceans. 463 



base of the mastigobranch and appears as an outgrowtli from 

 it, suggestin;^ strongly that the simiLir process on the masti- 

 gobranch of the third inaxilliped of many crabs lias the same 

 origin and that the two branches of the forked " epipodite " 

 of some PcnaiidiB represent the setobranch and mastigo- 

 branch respectively. Of course there are also cases in the 

 lower Penseidas and elsewhere where the podobranch and 

 both arthrobranchs are found together. I would suggest, 

 therefore, that in the primitive Decapoda the epipodite divided 

 not into two but into four structures — (a) the mastigobranch, 

 (/S) the setobranch, (7) the podobranch, (8) the anterior 

 arthrobranch — ^just as in the Lophogastrida? the proepipodite 

 has sometimes as many as four branches. At the same time 

 it must be remembered that the connexion of the anterior 

 arthrobranch with the mastigobranch is not a proved fact, as 

 is that of the posterior arthrobranch with the pleurobranch. 

 It seems quite possible that the ancestors of the Decapoda 

 bore not two but three rows of epipodial outgrowths on their 

 limbs, and that the anterior arthrobranchs represent the 

 middle of these three rows. Besides the " epipodite " and 

 " proepipodite," Branchipus bears on the outer side of its 

 thoracic limbs a third outgrowth of somewhat different form. 

 This has been doubtfully claimed as the exopodite, but may 

 quite possibly represent the mastigobranch. 



(3) Whereas the Reptantia (Eryonidea &c.) appear in the 

 Trias, the geological record shows no trace of Caridea till late 

 Jurassic times. This group, in fact, is a late and somewhat 

 specialized offshoot from the Penseid stem. The lower 

 Keptantia have, perhaps, evolved further than the lower 

 Caridea, but they are still in some respects more primitive 

 and they took origin much earlier. Boas's arrangement is 

 therefore justified. The Natantia are as natural a group 

 as the Reptantia, and into these two suborders the order must 

 be divided. 



IIJ. 



In considering the subdivision of the Natantia it will be 

 evident from what has been said that the Penseidea and the 

 Caridea must stand as two tribes of tlie suborder. To these, 

 however, must be added a third whose position needs some 

 examination. The little family kStenopidie was placed by 

 Boas with the Penaiidea, which it resembles in its three chelate 

 legs and in other respects ; but other authorities have very 



• It is quite possible that the trichobranchiate nature of the gills of 

 the lower Ueptautia is another priuiiiive feature lost by the Penseidea 

 and Caridea. 



