190 Report S.A.A. Advancement of Science. 



a feature as in the extremes of tentoria, and others in which the 

 flatness is as complete as in extremes of oculifera. The condition 

 seems more constant in trimeni, a moderately conical shape being the 

 rule. 



The form of the shields can probably be considered as one of 

 the most variable of the taxonomic characters. All we are justified 

 in saying is that the conical condition is usually emphasised in the 

 tentoria sub-group, the flattened condition is more general in the 

 oculifera sub-group, while between these extremes are all transitional 

 stages, often in the same sub-group. As in so many other instances, 

 the tentoria, verreauxii, and oculifera sub-groups include both the 

 extremes and the intermediate stages of the character. 



To a limited extent, Siebenrock has employed the relative sizes 

 of the carapace shields for diagnostic purposes. Where only a few 

 specimens are available, the character might be of some value, but 

 among a large number it is found that little constancy is exhibited. 

 The smaller variations in the relative sizes of the shields are largely 

 determined by the general shape of the shield, whether conical 

 or flat, and by the general shape of the shell, and, to a limited 

 degree, these are dependent upon the adjacent shields. It is manifest 

 that where so many variables are concerned the proportional size of 

 any shield can scarcely be expected to be a fixed constant. 



Both Boulenger and Siebenrock have utilized the relative sizes of 

 the plastral shields, especially as represented by their median suture 

 lines, as taxonomic characters, but in going over a large assemblage 

 of individuals, even from one locality, the feature is found to be 

 so variable as to be practically valueless for indicating relationships. 

 In addition, the gulars and anals usually show sexual differences. 



;. Beak. 



The variation of the beak, whether feebly or strongly hooked, has 

 also served for diagnostic purposes. It is practically always strongly 

 hooked in oculifera, but less consistently so in verreauxii ; in 

 tentoria it is rarely as strongly developed as in those mentioned, but 

 is always feebly produced in trimeni. On the whole, the beak may 

 be considered to have attained fixity of character in some groups, but 

 to be very fluctuating in others, exhibiting all transitions from the 

 feebly to the strongly hooked condition. 



k. Frontal and Prefrontal Shields. 



An examination has been made of the upper head shields in 

 numerous examples of the different sub-groups of geometrica to 

 determine if anv constant character in their relative sizes or number 

 could be established. Frequently there are two somewhat enlarged 

 shields, evidently the prefrontal of other groups, distinguished 

 from the rest, but they are by no means constant in any sub-group, 

 nor is the arrangement of the others adjacent to them. Hence it is 

 manifest that the head shields have not yet attained such constancy 

 as can be utilized for systematic purposes. 



