1903.] MEDUS.E FROM BRITISH COLUMBIA AND ALASKA. 177 



although he thinks they may possibly be different ; in this event 

 he proposes that it should be called after Eschscholtz. Blainville 

 (4) gives a description and coloured figure of the Melicertum 

 penicillatum of Eschscholtz, but these are copied directly from 

 Eschscholtz without further additions. 



So far as can be judged, the species figured by Eschscholtz was 

 not based on immature specimens, considering the number and 

 length of the gonads, the tentacles, and the height of bell. From 

 the papers of Agassiz and Fewkes we are familiar with the 

 young stages of the Pacific species of Polyorckis ; these are quite 

 difiei-ent in essential points from Eschscholtz's di-awing of the form 

 which he took off the coast of Galifoi-nia (11, fig. 4, pi. viii.). The 

 eight long tentacles, four radial and four interradial, are out of all 

 comparison with the four rather large tentacles at the ends of 

 the radial canals in the young stages figured by Fewkes. It is 

 to Agassiz (3) that we owe the present name of Polyorchis, 

 Haeckel (18) retaining this name and placing it under the Lepto- 

 medusfe, family Oannotidse. 



Of the three species already known, by far the most interesting 

 is P. camjjanulatus, originally described as Medusa ea'mpanulata, 

 by Eysenhardt and Ohamisso (12). Here the bell is much lower 

 and more conical than in the other species. It is eight-sided, and 

 the sides meet in angles. These characters, combined with the 

 position and structure of the gonads as found in all Polyorchids, 

 are remarkably like what is found in the Aglauridee. For instance, 

 the long finger-shaped gonads of Aglantha in position and structure 

 are very strikingly similar to the gonads in Polyorchis, although 

 they are more numerous. While there are never more than four 

 radial canals in Polyorchis, as compared with the eight of the 

 Aglaiu-idfe, no great significance can always be attached to the 

 number of radial canals. 



The possession of free " Horkolbchen " by the Aglauridfe would 

 seem to separate them definitely from Polyorchism although there 

 are Leptomedusse, such as I^aodice, which possess the true endo- 

 dermal sense- clubs of the Ti'achomedusse. The most distinctive 

 character between these two groups is the possession by Polyorchis 

 of diverticula on the radial canals, but these undergo marked 

 change dm-ing the growth of the animal. Hardly distinguishable 

 in the young, they become apparent as the animal increases in^age ; 

 this points to their being a recent acquisition in the evolution 

 of the race, probably within the limits of this particular group. 



Of the three species of Polyorchis at present known, two are 

 from the Pacific, the third from the Adriatic. It has already 

 been mentioned that Agassiz (3) found P. penicillata in the 

 region of Puget Sound, while Fewkes (14) found it as far 

 south as Santa Oruz on the southern coast of California ; it would 

 seem to be, therefore, one of the few Medusa-forms common to 

 both the northern and southei-n fauna of the West Coast of 

 N. America. 



Proc. Zool. See— 1903, Yol. IL No. XII. 12 



