298 BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



so than in levuka. All these agree in having eight or more tentacles. 

 Modesta Hartlaub is probably also a variety of levuka; but the single 

 specimen was immature. Vanhoffen (1912, p. 21), it is true, states 

 that it does not even belong to the genus Eulima. But I can find 

 nothing in Hartlaub's account (14 tentacles, many marginal bulbs 

 flanked by cirri, 8 otocysts; peduncle present; gonads on subum- 

 brella) to warrant this view. 



Curva and orientalis Browne have only four tentacles and differ 

 from each other only in the number of gonads. Australis Mayer 

 is not separable from curva, the two agreeing in the number of 

 tentacles and form of the tentacular bulbs, extent of gonads, and 

 general form. In short, all three probably belong to one species. 

 According to Vanhoffen (1912) they are indistinguishable from the 

 Atlantic E. mira. He also classes as mira the small specimens, with 

 four tentacles from the west coast of Mexico described by me (1909a) 

 as levuka. But as they were very young, and since levuka must also 

 pass through a four-tentacle stage, it is equally possible that they be- 

 longed to that species. 



Vanhoffen classes levuka as a synonym of the Atlantic gentiana on 

 the strength of its numerous tentacles and long peduncular gonads, 

 and records specimens of this type from Amoy and Hongkong 

 under the latter name. But although, as I have pointed out (1909a), 

 levuka and gentiana are closely allied, it is by no means certain that 

 their true relationship would be best represented by uniting the two, 

 and maintaining the combined species as distinct from gegenbauri, 

 because it is not unlikely that Haeckel's original gentiana was an 

 abnormal gegenbauri (Vanhoffen 1912, p. 23). At any rate, it had 

 no subumbrellar gonads, which is seldom the case in adults either of 

 gegenbauri or of levuka. 



The only thing separating levuka from gegenbauri is its long 

 peduncular gonads. But so far as the various published accounts of 

 the latter go (Haeckel, 1879; Vanhoffen, 1913; Apstein, 1913; Neppi 

 and Stiasny, 1913), this difference seems to be constant. Further- 

 more, as Vanhoffen points out, no Eutima of the gentiana (or levuka) 

 type has been recorded from the Atlantic since 1879. It is true that 

 E. variabilis McCrady (1857), figured by Brooks (1886), resembles 

 it in having 12 to 16 tentacles and long peduncular gonads. But it 

 differs in general form (short peduncle), and in the large number 

 (10 to 12) of otoliths in each otocyst, from all other Eutimas, though 

 the number of otocysts (eight) places it in that genus. 



Until the range of variation of gegenbauri and levuka is better 

 known, it is wisest to retain the latter as a distinct species, though 

 with the reservation that the two may finally be united. 



