296 A HISTORY OF RECENT CRUSTACEA 



maxilliped is not of exceptional structure. The second 

 joint is large, usually with a terminal plate having 

 coupling hooks on its inner margin. The seventh joint is 

 wanting, and in the genus Campylaspis the fifth and sixth 

 also, the fourth joint being rudimentary and terminal, and 

 the second joint in this case not carrying an apical plate. 

 All the known forms, however, agree in having in con- 

 nection with the tirst maxillipeds a remarkable respiratory 

 apparatus. The hinder part of this, being certainly at;- 

 tached to the first joint, no doubt represents the epipod, 

 while the front part, though coalesced with the hinder 

 part, may in reality be derived from the second joint, and 

 in that case it will, as Sars supposes, represent the exopod. 

 This latter opinion is sometimes contested, but, however 

 that may be, it is to Sars that belongs the honour of having 

 first clearly made known the manner of action of these 

 structures, as well as of having beautifully delineated the 

 intricate forms which some of them assume. The epipod 

 directed backwards into the branchial cavity finds many 

 parallels, as in the Mysidse, where likewise its oscillations con- 

 tribute to maintain a steadily renewed flowing in of water. 

 But in the Cumacea the epipod is usually furnished with 

 branchial sacs, sometimes in large numbers and spirally 

 arranged. To this there is nothing properly parallel in these 

 appendages elsewhere. The anterior or exopodal portion 

 of the apparatus is even more singular. The pair of these 

 plates on the two maxillipeds are sometimes firmly coalesced 

 together, but in any case they meet and form a tube which 

 can be projected in front of the head and again withdrawn. 

 As long as the animal continues the peculiar twisting 

 movements which in life it so frequently exhibits, the 

 bra,nchial apparatus is completely motionless. But when 

 it is in turn quiescent, the transparent end of the tube is 

 shot out under the pseudo-rostral end of the carapace 

 between the bases of the antennas, while the gill-bearing 

 epipodal plates are set oscillating. When the gill-plates 

 move forward the water is forcibly expelled through the 

 end of the tube. When the plates move back, the slit- 

 like orifice of the tube closes of itself through the elasti- 



