X THE VOYAGE OF H.M.S. CHALLENGER. 



it is present on the second pair also ; and these are never present without being attached 

 to a mastigobranchial appendage. This is true of each separate division, both normal 

 and aberrant, with the following exceptions: — viz., Cheramus in the Trichobranchiata, 

 Latreutes and Atya in the Phyllobranchiata. Of these the two former are small 

 specimens, and the mastigobranchia may have been overlooked, and it is present on the 

 second gnathopod in Atya. 



In the genus Stereoniastis there is only one mastigobranchia, and that is attached 

 to the second pair of gnathopoda and is in a rudimentary condition (p. 158, fig. 37) ; 

 there are, however, four podobranchiae attached to the anterior four pairs of pereiopoda, 

 but in this genus they are projected on a stalk and the mastigobranchia has become 

 obsolete and the podobranchia reduced to a degree, which appears to be further advanced 

 than is seen in Pentacheles euthrix, where the niastigobranchiae exist as jjlates of exquisite 

 delicacy. 



In the family Astacidae the majority of the genera are tabulated as having six pairs 

 of podobranchiae and only one mastigobranchia, Cambarus and Astacus having none ; but 

 the fact is that the mastigobranchia in this family is connected with the podobranchial 

 plume throughout the whole of its length in the manner shown in PI. XXVII. fig. 1, 

 pb, and in fig. 1m". This I think may be understood from a knowledge of the fact that 

 in their development the mastigobranchial plate and the podobranchial plume commence 

 in one sac, which afterwards divides by forming a branch that is without branchial 

 filaments, as may be seen in PI. XIIb. fig. 4, g. But whether they be united or distinct 

 from the branchial plume they fulfil the same office, that of separating one set of 

 branchial appendages from another, and sending long serrate hairs between the fila- 

 mentose rods, and thus keeping them free from undue lateral pressure, as may be seen 

 in PI. VII. figs. 1 and 1 bis, and PI. XXVIII. pd.br. 



In many instances, especially where the podobranchiae are not developed, the mastigo- 

 branchise are small ; but though small they can scarcely be considered as rudimentary, 

 seeing that they are developed upon a general plan, and that one of usefulness. In 

 PI. CVII. fig. mb., and PI. CVIII. fig. i", where they are figured as developed with a hook 

 at the extremity, varying in form, they reach only to the extremity of the next suc- 

 ceeding branchia, and sometimes, as in Atya (PI. CXIX. fig. 1), they terminate in a brush 

 of long hairs that penetrate between the plates of the different plumes. 



The arthrobranchiae, or those branchiae attached to the membranous articulation that 

 connects the legs with the body of the animal, are the most abundant and very constant 

 throughout the Macrura. They appear to be present in all the genera alluded to in this 

 Report, with the exception of Pontophilus, Sabinea, Pontocaris, Nika, Paralpheus, 

 Synalpheus, Latreutes, Hippolyte, Spirontocaris, Hetairus, and Pontonia. There is only 

 one arthrobranchia in Alpheus, and that is attached to the second pair of gnathopoda. 



The pleurobranchise, if not the most numerous, are perhaps the most constantly 



