288 



BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



The general structure of the genus is now so well known that all 

 that is needed here is a discussion of characters which may be ex- 

 pected to be of specific importance. The series, arranged according 

 to size, gives the following data : 



i Too crumpled to measure. 



Vanhoffen (1911) has recorded the following data for the Valdivia 

 specimens : 



If we consider the specimens not only as individuals, but quadrant 

 by quadrant, we shall obtain a fairly complete survey, because 3 

 of the Albatross series are sexually mature. The number of tentacles 

 increases somewhat irregularly with growth, six interradials being 

 the observed maximum. If this number were present in all four 

 quadrants of any one individual the total would be 28, radial and 

 interradial. But since this number is not reached in either of the 

 large Valdivia specimens, it is doubtful whether it is normally at- 

 tained. Only in three quadrants did I find young tentacles — one in 

 each. And these, as it happened, were not in the smaller, but in the 

 two largest individuals. All the large tentacles, as noted by Van- 

 hoffen (1911) end in oval terminal nematocyst swellings. 



Manubrium and gonads. — The gonads are purely interradial. In 

 the smaller specimens the manubrium is very short; in the two 

 largest, which are apparently mature, it is one-third as long as the 

 bell cavity. One is a female, with large eggs, which drop from the 

 sexual masses at a touch. In one specimen the manubrium is torn 

 off; in another, 7 by 7 mm., there are no gonads; in all the others its 

 walls are thrown into transverse folds (pi. 40, fig. 4). These, how- 

 ever, are not definite plications, such as are seen in Calycopsis, but 

 are so irregular and vary so in number on different sides of a given 

 manubrium that it is a question whether they are anything but evi- 



