280 DE. O. H. roWLEE — BISCATAN PLANKTON : 



in the Ostracods, but it seems likely tliat the larger and the longer-lived an animal, 

 the less will be its increase at each moult. It is also probable that (as has been shown 

 in some other marine forms) the rate of growth will not be by any means necessarily 

 the same under laboratory conditions as in a state of nature. 



The data in table 34 are included with others in Herrick's table 25 (p. 97), whore 

 1'15 fits better as a growth-factor than 1'14 ; but even this only contains 246 specimens 

 spread over ten moults. 



In table 28 (p. 99) Herrick gives the sizes of various parts of the skeleton before, and 

 five days after, a moult ; the detailed figures need not be printed here, but show — 

 what, indeed, we knew already — that different parts have each tlieir own growth-rate. 



In dealing with a large form like the lobster, a distinction must be drawn between 

 growth-moults at the young stages, and the moults at the later stages. If a lobster 

 continued to moult at the same brief intervals, and to grow by the same increment as 

 did Heri'ick's larvae, it would be lOh inches long at the end of its first year (instead 

 of 2-3 inches), and in five years would be a dangerous monster of portentous size. 

 We know, as a matter of aquarium observation, that the number of moults per annum 

 and the increment at each moult diminish with age. Unfortunately Herrick's obser- 

 vations do not help us to trace this decrease. It is not easy to measure accurately 

 anything so flexible as a many-jointed lobster, especially after moulting, when the 

 connecting membranes have been stretched and the skeleton has been torn across the 

 middle *. This, however, would probably not matter so much if the observations were 

 numerous instead of being scanty and scattered. Hence it is not sixrprising that the 

 observation in Herrick's tables 20 (82 observations for 11- inches), 32 and 33 (63 obser- 

 vations for 100 mm.), when plotted, failed to show definite groups round medians from 

 which the growth-rate of older stages might be safely inferred. 



Meco7'ds of Carcinus msenas. 



Dr. W. T. Caiman was good enough to call my attention to a paper t which contained 

 measurements of this species by Mr. H. J. Waddington and the late George Brook, 

 and to place at my disposal some of Mr. Waddington's actual specimens. The latter 

 gentleman measured (apparently not more closely as a rule than 0'25 mm.) the greatest 

 breadth of the cast shells of eleven individuals which he had kept in aquaria. The 

 total observations were only 117, to cover a range in growth of 09-62 mm., but seemed 

 to be worth study. When plotted on scale-paper, the observed breadths seemed, at 

 some parts of the scale, to fall into groups round obscure medians ; for example, at 

 9-10, 12-14, 15-17, 18-21 mm. By following out these indications and by working 

 in both directions from them, the following table was obtained. On the immediate 

 left of the vertical line are the means of these vaguely indicated groups ; on the right 



* It would be safer to measure the live lobster between moults, rather than the cast skin ; and in ever}- case 

 to measure also the cephalothorax to the root of the rostrum as well as the total length, carefully noting the sex. 



t H. C. "Williamson: "Larval and Early Toung Stages and Rate of Growth of the Sliore-Crab," Ann. Rep. 

 Fishery Board Scotland, xxi. pt. iii. p. 130. 



