308 Universitij of California Publications in Zoology [Vol. 11 



lata is stated to have but four gonads suspended from each of 

 the four radial canals, and thirty-six tentacles, also equal in size 

 and attached in a single row. The third species given is P. 

 campanulata. This form has a much greater number of lateral 

 branches to the radial canal, a large number of tentacles, varying 

 from fifty to one hundred and fifty, arranged in four rows over 

 one another on the rim. These are divided into eight large upper 

 and outer (four radial and four inter-radial) ; sixteen shorter, 

 within and below the outer; forty-eight still shorter and inner; 

 and fifty to one hundred to numerous, very short, and placed on 

 the extreme lower edge of the rim. Murbaeh and Shearer (1903) 

 give P. minuta as a distinct species, based principally on the 

 size. Our material leads us to regard these three species as growth 

 forms of one species, P. penicillafa. 



The specimens upon which the present studies were made 

 were obtained at different times from December to the middle 

 of April at the Oakland pier. San Francisco Baj-. 



METHODS 



The preparation of slides for the study of the structure of the 

 ocelli presents many difficulties. For his willing help in aiding 

 me in this difficult technique I wish to express my appreciation 

 to Dr. J, A. Long and also to Professor C. A. Kofoid, under 

 whose direction this study was luidertaken, for his kindly aid 

 and criticism. 



One of the fir.st difficulties encountered is that of infiltration. 

 The animal is of such a jelly-like consistency and contains such 

 a high percentage of water that dehydration without undue hard- 

 ening becomes a ])roblem. After repeated trials it was found 

 that small pieces of the margin of the bell bearing the bases of 

 the tentacles, from animals fixed in Zenker's or Bouin's fliiid 

 and preserved in seventy per cent alcohol, could be satisfactorily 

 dehydrated by leaving them in ninety per cent and then in abso- 

 lute alcohol for from eight to ten hours each. For infiltration 

 the best results were obtained by transferring the tissue directly 

 from the one hundred per cent alcohol to chloroform, to which 



I 



