130 Proceedings of the Royal Physical Society. 



hoo series, one regularly decreasing series (2 - 6, T8, 0"6, 0"1) for specimens 

 with 9, 11, 13, 15 tentaculocysts, and another regularly decreasing series 

 (4*2, 3 - 2, 1*0, 0'3) for specimens with 10, 12, 14, 16 tentaculocysts, (It is 

 obvious that there is an insufficient number of individuals with a decrease in 

 the number of tentaculocysts for us to expect a double series to be seen here 

 also.) In this connection, Browne (I.e., p. 100) has the very interesting 

 remark that " if a very slow and gradual change is taking place in the 

 number of tentaculocysts, then the tendency is towards the establishment of 

 a race with ten tentaculocysts," and " if no check occurs and the variation 

 still continues . . . the ultimate result will probably be a race with twelve 

 tentaculocysts." Yet it is not very obvious how his figures led him to this 

 conclusion. Hargitt (I.e., p. 582) considers that Browne's suggestion " seems 

 hardly warranted by the facts as known." 



The measurement of the diameter of the disc was taken in order that 

 the continuity of the series as a whole might be ascertained. When these 

 diameter-breadths, which are summarised in Table I., were arranged in order, 

 there was no indication of a break in the series for individuals whose disc 

 measures from 7 to 27 mm., and therefore we are entitled to assume that we 

 are dealing with individuals of a single race. This, for a single group of 

 immature adults, whose tentaculocysts vary in number, is in accordance with 

 what Browne found for Aurelias at ephyra, immature and adult stages. 

 There is much evidence, as Hargitt points out (I.e., p. 561), in support of the 

 general proposition that " varietal features found in the ephyra persist in the 

 adult, and furthermore, there is no evidence of any selective process involved 

 during these several changes in ontogeny." 



The twenty-two examples I found with a decreased number of sense- 

 organs had altogether 148 instead of the normal 176 tentaculocysts, while 

 138 individuals with more than the normal number had altogether 1517 

 sense-organs instead of 1104. What may be called the "coefficient of 

 productivity " for tentaculocysts thus becomes in the one case - 83 and in 

 the other case 1*37. Looking at this in another way, we find a total of 8385 

 tentaculocysts possessed by 1000 individuals, whereas had there been no 

 variation in this respect there should have been but 8000. Thus, on the 

 average, each of the thousand specimens had 8'385 tentaculocysts, while the 

 abnormal ones, considered by themselves, had an average of 10'4 tentaculo- 

 cysts. This shows how marked the variability is towards an increase in the 

 number of tentaculocysts. Table II. gives some comparative statistics on this 

 point. (It lias to be noted that Hargitt's results are based on an examination 

 of A. fiaviclula.) 



