32 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxr. 



are simply pointed and widely separated bj^ a shallow and wide incision. 

 With advancino- ag-e the tips of this plate are more produced (specimen 

 of 55 mm., Station No. 2980, fig. 2/*, and specimen of OS mm., Station 

 No. 2929, fig. 2c'), a slight angulation appears on the inner side of the 

 tips, which are not so widely separated, the incision ))ecoming narrower 

 and longer. Farther on the tips are gradually produced beyond the 

 posterior margin of the segment (specimen of 81 mm., Station No. 

 2919, tig. 2(7, and specimen of 91 mm.. Station No. 4389, tig. 2r), the inner 

 angle develops into a distinct spine, which is shorter than the tip, and 

 the two tips approach each other closely, finall}^ coming in contact at 

 the level of the smaller inner point. The incision becomes long and 

 narrow, slit-like. In the largest specimen at hand (115 mm., Station 

 No. 3HT0, tig. 2f) the two tips approach so closely to each other that 

 the inner point of the left side overlaps that of the right. 



Identity of G. hengalensis with G. caharata. — Wood-Mason gives 

 the following differential characters for his G. hengalensis: 



(1) "Carapace covers the whole of the first and part of tlie second abdominal 

 somite," while in G. calcarata the carapace does not cover the trunk entirely. 



(2) " The antennal, branchiostegal, and postero-inferior spines appear quite smooth 

 to the naked eye, being only obsoletely or microscopically serrated." 



(3) " The supraorbital spine is readily distinguishable by its shape from the rostral 

 denticles." 



(4) "The upper lateral keels are strongly roof-shaped." 



(5) "The oblique subdorsal keels are more pronounced." 



(6) "Antennal scale more Ijroadly emarginate at the apex." 



(7) "The pleural lappets of the last abdominal somite are terminated by two very 

 unequal spines (of which the outer is longer and sharp, and the inner short and 

 blunt), and are separated from one another posteriorly in the mid-ventral line by a 

 long and narrow incision." 



Length of Wood-Mason's specimen (female with a rudimentary 

 brood-pouch): 91 mm. 



Of these characters, the following ma}^ be remarked: 



(1) It depends entirely on the state of preservation how much of 

 the trunk or the abdomen is covered by the carapace. In my speci- 

 mens, there are the following limits: The minimum, when only the 

 trunk is covered, the maximum, when the whole of the first and the 

 anterior part of the second abdominal segment is covered. The latter 

 case corresponds to Wood-Mason's species, ])ut, as it happens, this one 

 is found in a small individual (55 mm. Station No. 2381), which is, in 

 all other respects, and especially in the ventral epimeral plate, a t3'p- 

 ical calcarata. In many of \\\\ specimens, in which the state of pres- 

 ervation permits, they being rather flal)bv, I am able at will to change 

 the degree of covering of the abdomen, b}^ simply pulling out or 

 pushing in the latter. 



(2) The serrations are to mj^ ej'es, which are normal-sighted, 

 always invisible, and I have to use a lens to discover them. Some- 



