1903.] medusjE erom British Columbia and Alaska. 185 



and sinks, with its tentacles fi^lly expanded, until it reaches the 

 bottom, or another piece of sea- weed where it attaches itself, and 

 after remaining suspended a little while, repeats the same opera- 

 tion." Agassiz has emphasised this habit of turning over in 

 the specific name vertens. This power of attachment is not, 

 however, due, as Agassiz states, to the lasso-cells, but to a definite 

 structure — an adhesive pad, an enlarged view of which is shown 

 in PL XXII. fig. 3, which is situated near the ends of the 

 tentacles, and acts like a sucker, which is sufiiciently strong to 

 tear the tentacles without loosening its hold. 



This power of attachment must be of great service to the 

 animal, for it prevents its being carried away by any strviggling 

 animal it may capture. It is quite probable this Medusa often 

 captures animals fully as large as itself, as it readily tiies to 

 digest pieces of meat, almost twice its size, which are di'opped 

 into the bell-jars in which they are kept during captivity. These 

 Medusa seem remarkably hardy and remain alive in small jars 

 of sea- water without change of water for several days. 



2. GoNiONEMUS AGASsizii, sp. nov. (Plate XXI. figs. 1, 2, & 3, 

 and Plate XXII. fig. 3.) 



Speci/iG description. — The bell is 9 by 17 mm,, or a little taller 

 than a hemisphere. The subumbrellar surface dips down in the 

 centre formimg a gastral cone. 



The velum is rather broad, strong, and well developed. 



The tentacles are shorter, more numerous than in any other 

 species of this genus described. They number eighty in specimens 

 of the above size. They are inserted into the bell-margin at 

 difierent levels, giving the appearance of one row above another. 

 In a moderately conti'acted condition they are not more than 

 two-thirds the diameter of the bell in length. They are much 

 thicker at the base, and taper more gradually towards the tips, than 

 in any of the species already known. Each is provided with 

 a small glandular attachment-pad some distance from the end 

 (PI. XXII. fig. 3). As already remarked, the proximal end of 

 the tentacle is slightly enlarged just before entering the margin of 

 the bell as in other species. This end of the tentacle is a tapering 

 root connected with the circular canal. Just at the outer margin 

 of the velum, under the circular canal, is a small ridge or welt of 

 ectoderm filled with urticating-organs. It is reddish-brown in 

 colour, the same as the papillae below. 



There are papillse under the circular canal on the bell-margin, 

 just below the insertion of the tentacles, containing diverticula of 

 this canal. Their outer layer is composed of the same tissue as 

 that composing the nettle-ridge, and of a similar brown colour. 

 There are not so many otocysts as tentacles, though there is no 

 definite regularity. The pigment-spots at the bases of the tentacles 

 are not visible in the pi'eserved condition. 



