HYDROMEDUSAE, SIPHONOPHORES, AND CTENOPHORES. 289 



dence of contraction of the manubrium as a whole. The fact that 

 the perradii, as well as the interradii, are more or less folded, though 

 there is no sexual tissue there, is evidence in this direction. 



In at least one adult specimen of anonyma (Bigelow, 1909a) the 

 walls of the manubrium were smooth; but in others which I have 

 examined it was more or less folded (1913, p. 26). 



In two of the smaller specimens the manubrium is sunken into the 

 gelatinous substance (pi. 40, figs. 2,3) in the peculiar manner already- 

 described by Vanhoffen. But this is merely an evidence of a remark- 

 able contraction of the upper part of the bell cavity. There is no 

 longitudinal connection in these cases between the manubrium and 

 the subumbrella or radial canals. 



In the third small example, as in all the larger ones, the manu- 

 brium hangs in its usual position in the bell cavity. It would be 

 interesting to study this peculiar contraction phase on living mate- 

 rial. At any rate, as Vanhoffen (1911) has pointed out, it can have 

 no significance in classification. 



There are no centripetal canals in any of the specimens. 



Color. — On the preserved specimens the manubrium is pale yellow ; 

 otherwise they are colorless. In the Valdivia examples the terminal 

 tentacular knobs are red, and the manubrium (in life?) had four 

 interradial dark brown-violet stripes. 



The genus Heterotiara has previously been taken only in interme- 

 diate hauls, but two of the Philippine specimens are from the surface. 



Genus CALYCOPSIS Fewkes, 1882. 

 Sibogita Maas, 1905. 



It is no longer an open question whether Sibogita can be distin- 

 guished generically from Calycopsis, separated by Vanhoffen (1911) 

 on the assumption that in the former some of the canals are cen- 

 trifugal branches from preexisting canals, whereas all the inter- and 

 adradial canals of the latter are centripetal. If this were true, the 

 distinction would be a valid one; but I have recently (1918) described 

 a species bridging the gap, in that some of its centripetal canals join 

 the radial canals, some the manubrium ; and there discussed the genus 

 as a whole. 



The present collection contains two specimens, one belonging to 

 G. geometrica, the other without question specifically identical with 

 the Medusae from the eastern Pacific, which I described under the 

 name Sibogita simulans, representing a stage in development inter- 

 mediate between the extremes which I then recorded. Vanhoffen 

 (1911) has united simulans with the Gulf Stream species typa. But 

 though the differences between the two are trivial tl^.seem sufficiently 

 constant to distinguish simulans, at least as a variety. It is clear 



