42 BULLETIN OF THE 



ovate, the entire margin ciliated, and with the tooth of flie outer margin small 

 and far back from the tip. The other parts of the antenna are nearly as in 

 Crangon. Tlie oral appendages are very similar to those of Cranyon and the 

 closely allied genera, and the labrum, nietastonia, mandibles, maxillae, and 

 second pair of maxillipeds are very nearly as in Crangon vulgaris. In the first 

 maxilliped the proximal lobe of the endopod projects inward much more prom- 

 inently and the distal lobe is longer than in Crangon. There are two arthro- 

 branchia; at the base of the external maxilliped, as in Sahinea, and the stout 

 endopod is composed of three segments as in Crangonina-, but the two distal 

 segments are very short and the terminal one acute and spined. 



The legs of the first pair are symmetrical and about as large as the external 

 maxillipeds ; the propodus is short and tapers distally, and the dactylus is small, 

 slender, and capable of flexion against the inner side of the propodus. The 

 legs of the second pair are elongated, slender throughout, and, in all the speci- 

 mens examined, slightly unsymmetrical in length ; the carpi are long and 

 muiltarticulate ; and the chela; small. The last three pairs of legs are slender 

 and nearly alike. 



The number and arrangement of the branchiae differ from ell the Crangonidse 

 known to me. In R. sculpta, the second species here described, there are 

 epipods on the bases of the first and second maxillipeds and two arthro- 

 branchia; at the base of each external maxilliped, one arthrobranchia for each 

 of the thoracic legs except the last pair, and a pleurobranchia for each side of 

 the last five thoracic somites, — making two epipods, six arthrobranchia;, and 

 five pleurobranchise each side, as indicated in the following formula : — 



ll+(2) 



The abdomen is sculptured and spined to correspond with the carapax, and 

 the exoskeleton is througlumt very thick and massive. 



In the three species here described, the hinges at the last three articulations of 

 the abdominal somites — that is, at the articulation of the fiftli with the fourth, 

 the sixth with the fifth, and of the telson with the sixth — present a peculiar 

 modification by which the liinge is very much strengthened and is at the same 

 time apparently made capable of being clamped or locked so as to hold the 

 terminal somites firmly extended. In addition to the ordinary hinge, at each 

 of these articulations, there is a process arising from the anterior somite just 

 below the hinge and curved backward and upward concentrically with the 

 hinge, and this process fits accurately and is slightly overla]>ped along its edges 

 by a siijiilarly curved groove in the posterior somite. When the abdomen is 

 completely flexed the ends of these curved processes project dorsally consider- 



