128 



SCIENCE-GOSSIP. 



this form (figs. i6 a-d). On comparing it with 

 Gorilla erronea, it is at once noticeable that it has 

 great affinity with that species; it is, in all 

 probability, only a well-marked variety of it, and 

 as it is known only from the single specimen it 

 would certainly be imprudent to accord it higher 

 than varietal rank. It possesses the same number 

 of plates, but the shell is much smaller and 

 thinner, and the palatal plates are much shorter 

 and placed much nearer the mouth of the 

 shell. The outer terminations of the parietal 

 plates and the whole of the three upper palatal 

 plates are visible from the aperture (see fig. i6a) ; 

 palatal plate No. 3, which in Corilla erronea is 

 nearly horizontal, is here strongly oblique and 

 ascending, while No. 4 reaches nearly to the 

 peristome (see fig. i6c). The form must be known 

 provisionally as Corilla erronea, var. erronella. 



All the known forms of Corilla fall naturally into 





hwniitt/it 



Diagram or Relationship of Corilla. 



the three groups of C. charpentieri, C. anax and C. 

 erronea, already indicated. The first group, that of 

 Corilla charpentieri (including the var. hinidunen- 

 sis), is without internal plates ; the second, that of 

 Corilla anax (including the two forms of C. 

 beddomeaej, has oblique palatal plates ; and the 

 third, that of Corilla erronea (including the rest of 

 the genus), has horizontal palatal plates. It will 

 be remembered that the structure of the armature 

 in young shells differs in a remarkable degree from 

 that found in full-grown specimens. I have 

 pointed out that in the former case the plates are 

 invariably broad and obliquely slanting upwards 

 (see ante pp.90 and 91), while in the latter case they 

 are, in some species at least, narrow and horizontal. 

 From what we know of the retention of ancestral 

 characters in young individuals , as explained by 

 Mr. Darwin ("Origin of Species," sixth edition, 

 p. 388), it may safely be assumed that the 

 immature form of plates found in the young shells 



represents the form of plates which were possessed 

 by the progenitors from which the existing armed 

 members of the genus have sprung. Consequently, 

 those species which have to some extent retained 

 such characters in the adult state {i.e. Corilla anax 

 and C. beddomeae) are the older forms ; while those 

 species which diverged most in the adult state 

 {i.e. the group of Corilla erronea) are of more 

 recent origin. Assuming that the prototype which 

 gave rise to the armed forms was devoid of arma- 

 ture, the Corilla charpentieri group would represent 

 the oldest forms of all, while Corilla beddomeae and 

 C. anax would come next in the line of descent in 

 one direction ; C. humberti still later, but in another 

 direction ; next C. odontophora, C. erronea and C. 

 rivolii would appear to have branched off in 

 separate directions ; and lastly, C. erronella and C. 

 fryae have diverged from the common stock. As it 

 is extremely difficult to indicate the true relation- 

 ship between any given group of species in a linear 

 arrangement, I have attempted to overcome this 

 difficulty in the accompanying diagram. It will, 

 of course, be understood that this has reference to 

 conchological characters only. 



I append a key to the species of Corilla which I 

 venture to hope will prove serviceable : 



A. Shell without internal folds. 



a. Shell large, diameter 29 mm. charpentieri. 



b. Shell smaller, diameter 22 mm. 



v. hinidunensis . 



B. Shell with internal folds, 

 a. Palatal folds oblique. 



a. Two parietal folds .... beddomeae. 



13. Three parietal folds aiiax. 



h. Palatal folds horizontal. 



a. One parietal fold humberti. 



j3. Two parietal folds .... odontophora. 

 y. Three parietal folds. 



* Shell elliptic, palatal folds short, 

 second scarcely curved. 



t Lip much reflected rivolii. 



tt Lip little reflected. 



1. Third palatal fold, 



almost horizontal . . erronea. 



2, Folds very short, nearer 



aperture, third palatal fold 

 very oblique, ascending 



V. erronella. 

 ** Shell rounded, palatal 



folds longer, second 

 much curved . ., . fryae. 



In concluding the consideration of the genus 

 Corilla, I take the opportunity of expressing my 

 grateful thanks to Mr. John Ponsonby, Colonel 

 Beddome and Dr. Brot, for their kind and liberal 

 assistance with specimens and information, and 

 to Lieut. -Colonel Godwin-Austin for valuable 

 information and suggestions. 



(To be continued.) 



