[33] 



MEDUSA FROM THE GULF STREAM. 



959 



Family ^QUORID^ Eschscholtz, 1829. 

 PoLYCANNA Haeckel, 1879. 



lu studying- the characters of the ^quorid^ found at Newport a pe- 

 culiar structure in Polycanna {Zygodactyla) Grcenlandica (?) (Ag.) has 

 presented itself, which, as far as I am aware, has not been observed by 

 others. This character easily distinguished this species, and may even 

 be found to be of generic worth.* On the subumbral side of the um- 

 brella we iiud between: the radical tubes small rounded i)rominences ar- 

 ranged in rows, a single row between each x^air of tubes, which cor- 

 respond in number with the number of ovarian frills on the chymiferous 

 tubes. The number of these projections varies with the size of the 

 specimen. At Newport we have two species of Polycanna {Zygodactyla 

 Ag.), one of which has these projections, while a second is destitute of 

 the same. 



Among the ^Equoridse collected by the Albatross there are two 

 species of Mesonema,\ M. cyaneum Hackel, and-Jf, Bairdii, sp. nov. 

 There is one species of Polycanna, P. Americana^ sp. nov. 



In the two species of Mesonema here considered the number of tentacles 

 is less than the number of chymiferous tubes. They are very close to 

 each other structurally, and may beeventually placed in the same species. 



PolyGa7ina Americanah-da ovaries like those of Po/T/crtwwa {Zyyodactyla) 

 crassa (A. Ag.). 



POLYCAWNA AMERICANA, Sp. nov. 



There are three specimens closely related to P. Gennamca, H;Bck., and 

 P. Italica, H'«ck., one of the subgenus, Crematostoma, to which is given 

 the name P. Americana.! 



Specimen examined. 



The bell is slightly rounded in vertical profile, discoid, with thin 

 gelatinous walls. Diameter of the disk 45'"'". Tentacles very long, 



^As these structures are not mentioned in the description of Polycanna {Z\j<jodachj\a) 

 Grcenlandica by Agassiz, I am unable to say whether this is the same si)ecics as liis or 

 not. As the tentacles are in a single row in this animal, it cannot, according to 

 Haeckel, be placed in the Zygodactyla of Brandt. 



1 1 have here used the term in the restricted sense adopted by Hajckel. 



tin the preliminaiy list the generic name Zycjodactyln is used for tliis mtdusa. In 

 Brandt's Zv^o(Za<,'<//7«. the tentacles are " biseriata." In Poi//cart«a the tentacles arfj 

 *'uniseriata." P, Grcenlandica has the tentacles in a single series. 



