270 DR. G. H. FOWLER — BISCATAN PLANKTON : 



representecl in the female Stage I. of magna. Provisionally, therefore, I regard magna 

 as exhibiting an early phase of a tendency to split into two races. 



When this investigation was begun, it was of course not realized at first to what the 

 measurements of the specimens by sexes would lead. Unfortunately therefore, owing 

 to the probability tliat one or two other species were included with magna in the 

 A larvae (p. 2i<5), and because my object was merely identification of species with a 

 view to questions of distribution, it did not at that time seem worth while to carry the 

 measurements lower than I'O mm., and not even all specimens of that length were carefully 

 recorded, as was the case with the greater lengths. It could not tlius early be foreseen 

 that one was thus arbitrarily cutting off the shorter members of Stage III., with the 

 result that the mean calculated ou the recorded specimens is necessarily too high. This 

 seems to be shown, at any rate in the females, by dividing downwards, instead of the 

 method of multiplying upwards adopted in other species. By dividing the mean for 

 Stage I. by that for Stage II., 3-36-l-18S=l-78 the growth-factor, and l-88^1-78 = l-05 

 the probal)le mean for Stage III., whereas my mean is 1'14. As the meaning of the 

 double-humped curve of the males is obscure, one does not know how to handle 

 the figures of Stage I. ; treated as in the females, the observed mean of Stage I. gives a 

 probable mean of 1'07 for Stage III., and the ordinate of the lower hump gives a 

 probable mean of 1'12. 



It may be possible to sex and measure these smaller specimens later, but at the 

 moment it would delay publication for some months. 



There are other features noticeable in these curves which I must leave to better 

 mathematicians than myself to explain. 



The a LABV.-E. 



Only about a quarter of these larvse were measured, for the reasons already given. 

 They yielded 25 at 0-9 mm., 209 at 08, 76 at 7, 59 at 0-5, 3 at 0-4 mm. The absence 

 of any specimens at 0'6 seems to cut off those at 09, O'S, and 0'7 as Stage IV., at 0'5 

 and 0"4! as Stage V. Those at Stage IV. form a good curve of frequency (plotted on the 

 diagram for magna), but any difference which may exist between males and females at 

 this stage is of course obscured. 



A few specimens of Stage IV. were sexecl and measured to a second place of decimals ; but against 

 this is not only the uncertainty of the specific determination, but the difficulty of absolute determinatioa 

 of the sex. In some specimens the rudimeut of the penis is unmistakable, in others its absence ; but iu 

 a large number one could not trust one's own decision. So far as it went, the measurement seemed to 

 show a mean of 0'79 for females and of 0"83 tor males, which may or may not have been true. 



CONCHCECIA PROCERA, Miiller. 



As already mentioned (p. 2i6), Stage I. apjiears to be lacking in the ' Research ' 

 collections, and is perhaps the form named decipiens by Miiller. 

 Tiie following numbers Avere observed : — 



