20 



is not quite correct. In all 

 my specimens the scaly mark- 

 ings of the surface are more 

 rounded than pointed, and 

 the elevated band on the an- 

 terior margin of each joint 

 is dentate, not serrated. 

 (Fig- 7-) 



Two almost complete spec- 

 imens measure eight inches 

 in length by 2^ inches the xk Fig. 6 



greatest width; carapace semicircular, anteriorly and laterally mar- 

 gined by a slightly elevated rim; length to width as two to three; 

 eyes lunate, convex; no markings vis- 

 ible. The width across the base of 

 the carapace is smaller than across 

 the thorax. The first five segments 

 have their lateral margins rounded, 

 the others extend m angles at the 

 lateral edges. The last four ab- 

 dominal joints are longer than the others, which are six or seven 

 times as wide as long. The tail is a strong, triangular spine, serrated 

 at the angles. One specimen shows the rows of little scale-like ele- 

 vations on the back very plain, six rows on the first three joints, 

 four rows on the next four, and two rows on the remaining abdomi- 

 nal joints. The other scaly surface markings are seen on the back 

 part of the carapace and all over the body. The swimming-feet in 

 these specimens agree with the above quoted figure. The postoral 

 plate is oval, the greatest width above the middle. The anterior 

 feet are missing or broken in all my specimens. 



A comparison of the number of species found last summer indi- 

 cate that E. pachycheirus was the most abundant. 



6. The central thoracic appendage of Dolichopterus macrockeirus 

 is very plain in the specimen in the museum. Extending to the 

 fifth joint of the body, it shows the triangular prolongation of the 

 first joint adjacent to it over the next joint, and terminates in 



F 



