400 



TORREY 



The Salem specimens exhibit three subtypes which correspond to 

 those manifest in the Newport specimens, the proportions of individ- 

 uals only being slightly different. I have arranged the Oakland speci- 

 mens in two columns, the first enumerating them exactly as they are, 

 the second on the assumption that half pairs represent incomplete full 

 pairs. The only change which this latter method of enumeration has 

 caused is a reversal of the relative importance of the forms with four 

 and with seven pairs of non-directives respectively. In both columns 

 the hexamerous is the dominant and only clearly inarked type. 



According to the first Oakland column, 53 polyps, or 21 percent, 

 are hexamerous, in a total of 354 ; according to the second column, 

 64, or 35 percent, are hexamerous. Of these, 39, or 15 percent, are 

 regular. Among the Salem polyps 31, or 33 percent, are hexamerous 

 and approximately regular. Among the Newport polyps 30, or 26 

 percent, are hexamerous ; how many of these are also regular is not 

 stated. 



TABLE III. DIGLYPHIC TYPE OF POLYPS. 



Table III exhibits the variation not only in the number but also in 

 the arrangement of the mesenteries of a limited number of diglyphic 

 polyps. There is great discrepancy between the number of regular 

 diglyphic polyps from Newport, on the one hand, and from Salem and 

 Oakland on the other. I can offer no satisfactory explanation for this, 

 since the methods of Dr. Parker's collector seem to have been similar 

 to my own. 



With the foregoing record before us, we may proceed with the dis- 

 cussion. 



A complete explanation of variation in Mett-idhim must account, 

 among other things, for the variation 



(a) in the number of perfect mesenteries, 



((5) in the number of siphonoglyphs, 



(c) in the relative arrangement of siphonoglyphs and mesenteries. 



(a) There is no direct evidence upon which to formulate the true 

 numerical ratio of sexual to non-sexual individuals. Hence it is im- 

 possible to state whether or not all departures from the assumed regular 

 diglyphic hexamerous type are non-sexually produced. It is probable 



