Notes. 453 



of the shell, therefore, must depend greatly, or exclusively, on the de- 

 velopment of the nacreous layer. It is only when the secreting power 

 of the surface of the mantle is extremely small, while that of the narrow 

 edge which secretes the prismatic layer is unusually strong, that the 

 thickness of the shell can be determined by the growth of the prismatic 

 layer. In such exceptional cases the margin of the shell is always thicker 

 than the main part. Any more minute description of these facts seems 

 to me to be here out of place, and unfortunately I cannot refer the 

 reader to other works, as I have not published anything on the subject, 

 nor am I at present likely to do so. 



Nate 102, pcu/e 217. GraefEe says that he has sometimes found a whole 

 family in these galls or cysts. It is not quite clear to me how a family 

 of the Crustaceans could find room in a cavity which is hardly twice the 

 size of the fully-grown HapaUoarcimis. It is possible that young larvae 

 might be found there before their escape, but this is not clearly expressed 

 in the text. 



Nate 103,jr^,'X<7« 221. This crab, living in TracliypliyVAa, a West Indian 

 coral, is extremely like Cryptoeliirus, and perhaps belongs to the same 

 genus ; this can only be determined by future and more exact exami- 

 nation. But the ' cave dwelling ' of this West Indian crab is perfectly 

 unlike that of the Eastern species, which is found from the Ked Sea as 

 far as the Pacific Ocean ; it is not cylindrical, but has one side qiiitc 

 flat, so that its transverse section is almost exactly a half-circle ; the 

 under side of the crab rests against the flat side of the cavity. With 

 regard to the pouch, I have not yet been able to make any inves liga- 

 tions. 



Note 1 M, page 223. The conditions here described will under some 

 circumstances be available for enabling us to form a much more exact 

 estimate of the rate at which a block of coral grows than has hitherto 

 been possible. The data here given, and presently to be worked out 

 more fully, are certainly hardly to be regarded as the result of perfectly 

 exact investigations ; the only really exact observation — carried out, too, 

 in minute detail — is that of Le Conte (in iSiUiman's Journal, series 3, 

 vol. X., 1875), and he found that a coral plateau in Key West (West 

 Indies) exhibited a perfectly regular dependence on the height of the 

 water at different seasons, so that it was always possible to ascertain 

 with great accuracy the rate of growth of the one species of coral ob- 

 served there — a kind of madrepore — which was about 3 J inches in the 

 year. Now, if the period of growth of a specimen of Ci-yptochirm could 

 be exactly determined, the rate of growth of the coral to which it 

 belonged could also be determined with mathematical accuracy, much 

 more exactly than it could be ascertained by direct measurement of the 

 coral itself. If we assume — ^what at present, it is true, cannot be proved, 

 though it is not improbable — that the Cryptoeliirus acquires the first six 

 millimetres of its length in the first year, that would give an average 



