340 DISCOVERY REPORTS 



though within the large group so formed it is possible to distinguish subgroups which 

 may eventually prove to correspond to valid systematic divisions. 



So far as our knowledge goes the larvae of Upogebiidae and Laomediidae have well 

 marked characters common to all the species known, and in both families very markedly 

 different from those of the Axiid-Callianassid group. They resemble each other in 

 some important respects, so that two groups of larvae may be defined as follows : 



HOMARINE GROUP 



Axiidae and Callianassidae 



(i) Rostrum long, flattened, with serrated edges. 



(2) Abdominal somites usually with dorsal spines. 



(3) Telson with large median spine at all stages. 



(4) Exopods on leg 4 or legs 4 and 5, with rare exceptions when development is abbreviated. 



(5) Endopod of maxillipede 3 functional and normally placed. (With one doubtful exception, 



P- 33°-) 



Anomuran group 

 Upogebiidae and Laomediidae 



(i) Rostrum small and round. 



(2) Abdomen without dorsal spines. (For exception, see p. 332.) 



(3) Median spine on telson small or absent; always absent in stage i. 



(4) Exopods on legs 3 or legs 3 and 4, never on leg 5. 



(5) Endopod of maxillipede 3 rudimentary and seated at base of basipod. 



The last character is that upon which the greatest stress has been laid as evidence 

 of relation to Anomura, but it is disputable. In Homariis, and also in some Caridea, 

 there is a tendency for the exopod to be carried on an extension of the basipod, and 

 the condition found in these Thalassinids might be regarded as simply a further 

 development of this modification. On the other hand, there is no parallel, outside the 

 Thalassinidea and Anomura,^ for the rudimentary condition of the endopod, or for 

 its position quite at the base of the segment. It seems there is not only extension of 

 the basipod, but also a shifting of the endopod, and the position is so peculiar that it 

 must have phylogenetic significance. 



A difference between the two groups which is not easily defined, but is quite evident, 

 concerns the whole form of maxillipedes i and 2. In the Homarine group coxa and basis 

 of maxillipede i are broad, and in both appendages the endopod is weak and armed 

 with rather delicate setae. In the Anomuran group coxa and basis are slender, and 

 the latter elongated, while the endopod is stronger and armed with stiff setae which 

 may be coarsely plumose. The structure is exactly the same in Anomura such as Galathea 

 and Eupogurus. 



1 In Stenopus maxillipede 3 is normal, but the legs have the unformed endopods seated low down as in 

 Anomura. 



