460 Rev. Canon Norman's Revision 



(Hyatt), and the shell itself is reduced in importance, narrow, 

 lanceolate or spathulate, and its structure corneous, or some- 

 times altogether absent ; and from these we pass to the 

 Octopoda, where, as one of many wide differences of structure, 

 the shell is rudimentary or wholly absent*. 



It must not be supposed that I have been intimating in the 

 preceding paragraph that the several groups of Decapoda have 

 been derived from each other. That, I take it, certainly is 

 not the case. They appear all to be derivatives in different 

 directions from the ancient Belemnites. 



No linear arrangement can ever adequately and fully demon- 

 strate the varied alliances of groups. Such groups have 

 diverged in many directions from earlier types, and while 

 differing more and more widely in the especial characters in 

 which their divergence is evidenced, will nevertheless be, so 

 to speak, laterally bound together by the retention of many 

 points in common during a shorter or longer period in their 

 onward course of successive modifications. Again, the law of 

 recurrence must be supposed to be not unlikely to come in. 

 Organs which have been modified or to a greater or less extent 

 suppressed under certain conditions of life and environment, 

 when those conditions are partially or wholly reversed may be 

 expected to revert more or less to their original condition 

 rather than that they should undergo change in a new direc- 

 tion, although such new direction might equally subserve the 

 same purpose. Again, the very same modifications which 

 have taken place in a line of divergence which we will call A, 

 may supervene at a much later period in another line of diver- 

 gence B ; for B had at an earlier time been undergoing modi- 

 fication in other parts of its structure than those at that same 

 period followed by A, but when ultimately somewhat similar 

 modifications having taken place in the same organ, which 

 had long before diverged from the original type in A, the 

 distant descendants of B may in this respect appear to. us 

 actually more nearly related to A than were its progenitors. 



The Oigopsida preserve in their hooked acetabula (or 

 suckers), in the indications of a phragmocone at the extremity 

 of the internal shell (Ommatostrejjhes) , and general form of 

 body more of the characters of the ancient Belemnites than 

 any other existing group. In Belemnoteuthis antiqua, Pearce, 

 of the Oxford Clay, we seem to recognize a form which may 

 represent a connecting-link ; the arms are furnished with 



* The beautiful egg-case of the genus Argonauta is riot a true shell. 

 It does uot take its origin in a shell-gland, but is a secondary product of 

 the dorsal arms, which are greatly expanded and turned back over the 

 mantle. 



