no. 1480. S( 'II í/o/ 'O I) ( 'R USTA ( 'E, 1 XS—ORTM. ! NN. 33 



times, chiefly in old individuals, they are actually wanting. Their 

 presence or absence cannot constitute a specific character. 



(3) The supraorbital spine is sometimes distinctly visible, some- 

 times entirely obsolete. If present, it is always marked by its 

 position. Even when developed, it is so small that its presence or 

 absence cannot be of specific value. 



(1) What Wood-Mason means by "roof -shaped" upper lateral 

 keels, I cannot imagine. 



(5) The oblique dorsal keels are also found in Sars's species; they 

 are slightly indicated in his fig. 2 (chiefly the posterior pair, which 

 is most important). In poorly preserved, flabby specimens, which 

 have undergone much rough handling, they are sometimes indistinct. 

 They are present in all my individuals, and hence this character can 

 not be accepted as constituting a difference between bengalensis and 

 calcarata. 



(6) The degree of emargination or truncation of the antennal scale 

 offers variations, as is already indicated in Sars's figures (Plate IV, figs. 

 2, 4, 5). I have called attention to this fact in connection with the 

 Hawaiian material a , wh^n is further confirmed by the present material. 

 A real emargination {i. e., a concave marginal line) is comparatively 

 rare; generally there is a truncation, with the marginal line between 

 tip and first tooth straight. 



(7) The description of the epimeral plate given by Wood-Mason 

 corresponds exactly to what we see in my figs. 2a to 2f, with the 

 exception that the inner tip of each epimeral lappet is sharp, not 

 blunt. In younger specimens, however, it is blunt (see my figs. 2b 

 and 2c). Thus this character agrees well with the assumption that 

 G. bengalensis is an older and larger G. calcarata. 



Thus of the seven characters given by Wood-Mason for G. bengalen- 

 sis, six are not actual differences, and one, the fourth, is unintelligible. 

 The only real difference from Sars's description and figure is found in 

 the epimeral plate of the sixth abdominal segment; but this organ, as 

 shown, changes its form with age, and G. bengalensis is a rather large 

 individual (91 mm.). Specimens from my material of the same size 

 present an epimeral plate (see fig. 2c) closely corresponding to Wood- 

 Mason's description. 



Sars had two specimens of this species; the large one was 98 mm., 

 and to it belong the figures of the whole animal (slightly enlarged, 

 Plate IV, figs. 1, 2). The carapace of the smaller one (68 mm.) is 

 figured in his fig. 3. Sars does not say from which individual the 

 other figures are taken, but it seems from the latter. Then his figure 

 of the epimeral plate (fig. 6) belongs to this smaller individual. The 

 same plate of an individual of the same size (68 mm.) is figured in my 



«Bull. U. S. Fish Comm. for 1903, 1905, p. 989. 

 Proc. N. M. vol. xxxi— 06 3 



