BORRADAILE— ON THE PONTONIIN^ 351 



of this in another group of Crustacea* and have remarked on the appearance it presents 

 of the kaleidoscopic distribution in several ways of a set of pairs of characters, and called 

 attention to its suggestion of Mendelism. A simple case of this sort occurs among 

 the members of the subgenus Falciger, where it is hard to decide whether the primary- 

 grouping of the species should be on the ground of the presence and absence of the 

 supraorbital spine or on the shape of the rostrum. Many other such cases may be 

 detected in the keys which are given later in this paper. Another form of the same 

 difficulty arises in the sporadic distribution of a character in circumstances which make 

 it impossible to trace direct connection by descent between the species in which it occurs, 

 except by means of other species from which it must have been absent. This is the case, 

 for instance, with the supraorbital spine throughout the subfamily, and with the denti- 

 culation of the fingers of the first leg in several Periclimenes. 



In earlier days these phenomena would have been even harder of explanation than 

 they are at present, when convergence is receiving more attention than formerly, and 

 Mendelism has risen over the zoological horizon. In many cases close examination 

 of features which at first appear to be due to repetition reveals, either between the fully 

 evolved organs or between stages that lead up to them, differences which show that we 

 are dealing with convergence. In others, the suggestion of Mendelism is probably correct, 

 the feature in question having been suppressed over a series of generations by the absence 

 of some factor necessary for its development or the presence of some hostile factor. 

 It may, indeed, be doubted whether much of what is known as convergence in organisms 

 be not rather due in this way to Mendelism. A further question, however, arises here. 

 Is the sporadic repetition of such a feature, for instance, as the supraorbital spine, 

 in apparently identical reincarnation, due to the releasing of the same " factor," which, 

 once lost, cannot be regained, or is it brought about by a less organized tendency of 

 the mechanism of development to fall into an identical condition from time to time ? 

 Far though we be from answering them, these questions are not without interest for 

 others than the systematist. But it is doubtful whether they would be asked if 

 phylogenetic speculation were regarded as wholly a waste of time. 



Another problem which is continually suggesting itself concerns the mode in which 

 special features arise and disappear. This has been generally assumed to take place 

 by the gradual elaboration of a rudiment into the highly developed organ, and its loss, 

 when it has outlasted its usefulness, by a series of vestigial stages. Now in some cases 

 it is certainly possible to trace a set of incomplete forms of an organ, and it is justifiable 

 to regard these as either rudimentary or vestigial, though it is often hard or impossible 

 to choose one or other of those alternatives. As instances of such phenomena within 

 species there may be quoted the various forms of armature of the finger of the great 

 chela in Harpiliopsis depressus, where stages in the perfection of a cutting flange 

 may be seen, and the varying degrees of inequality of the legs of the second pair in 

 Coralliocaris graminea. It is easy to imagine that in either of these cases one form 

 of the organ might come to persist to the exclusion of the others. Within the limits 

 of a genus, the same thing may be seen in regard to the remarkable protuberance 

 * On the Genera of the DromiidEe, Ann. Mag. Nat. Hist. (7), xi. p. 303 (1903). 



