290 



BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



that in the simulans form many, or all, of the centripetal canals 

 finally unite with the cruciform base of the manubrium, a union 

 taking place in specimens in which only 12 canals (radial and sub-* 

 radial), are developed, as the present example shows. But should 

 new canals be later interpolated in such a way that they alternate 

 more or less regularly with the preexisting ones, the condition 

 observed by Vanhoffen in the Valdivia specimen credited by him 

 to typa but christened valdiviae by Hartlaub (1913) — namely, 24 

 canals which unite with the stomach and 36 blind ones alternating 

 somewhat irregularly with them — might result. In my Gulf Stream 

 specimens (1909&), on the other hand, all the canals were blind even 

 in a specimen with mature ova, 37 mm. high, with 21 canals in all. 

 In Fewkes's type specimen, which I have recently examined (1913, 

 p. 21), every centripetal canal which is well enough preserved to 

 show its termination is blind. Moreover, as I have pointed out 

 (1913), his specimen and all of mine from the Gulf Stream, includ- 

 ing two series collected by the Grampus (19096, 1915a), though taken 

 more than 30 years apart, show the apical depression of the sub- 

 umbrella which I have described and figured elsewhere (Bigelow 

 19095). But no such character has been observed in the simulans 

 form. 



If it were the Grampus series alone which showed it, it might well 

 be credited to individual or to swarm variation. But when we find 

 it present in all examples from the Gulf Stream, though taken so 

 many years apart, it is a reasonable conclusion that we have here 

 evidence of a distinct local race. And if this be true, it deserves 

 recognition in nomenclature, at least as a geographic variety. 



Large series may show that the centripetal canals in the Atlantic 

 form are not permanently blind, as they now seem to be, and that the 

 apical depression is not of importance. But for the present it is 

 wisest to refer the Philippine specimen to var. simulans of typa. 



CALYCOPSIS GEOMETRICA (Maas) Bigelow. 



Plate 40, figs. 5-7 ; plate 41, fig. 2. 



Sibogita geometrical Maas, 1905, p. 17, pi. 3, figs. 16-18. 

 Calycopsis geometrica Bigelow, 1918, p. 377. 



Calycopsis geometrica — material examined. 



