124 ERNEST HILL AND L. G. HAYDON. 



from the figure. If this be the " basal hair " it is not of 

 differential import in our species. 



(2) Frontal or CI y peal hairs. — -In all larvee examined 

 by us three pairs are found, two near to the anterior ex- 

 tremity of the head, and a third a little behind the median 

 pair. The posterior may be very small and require relatively 

 high magnification to detect in the first instance (PI. XXII, 

 «i, a.2). The frontal hairs are of prime specific importance, 

 and it is generally necessary to consider the three pairs 

 together. They exhibit much individual variation, especially 

 the posterior (vide Pis. XV-XXIII, fig. a). In fun est a 

 they may be quite free of branches^ while in pretoriensis 

 they invariably are. The lateral hairs also vary, though to 

 less extent (Pis. XVI, XVIII, XXI, XXIII) ; but it is to be 

 observed that in natal ensis (PI. XXIII, fig. a) on one 

 side is found a fine branched hair, on the opposite a straight 

 smooth bristle. Attention is needed both to structure and 

 to relative size of posterior to anterior median, which, for 

 instance, is much less in pretoriensis than in funesta or 

 cine reus. Relative position of point of origin is not avail- 

 able for comparison, because of the impracticability of 

 arranging a preserved specimen precisely flat in such a 

 manner that exact measurements can be made; it is not 

 possible to secure uniformity of plane without crushing the 

 larva, which is thereby distorted and the hairs frequently 

 broken and detached. 



Thus it happens that in drawings of larvte the proportions, 

 except of palmate hairs, are approximate, although a micro- 

 meter scale was scrupulously applied to every detail as far as 

 circumstances admitted. This more particularly applies to 

 instances in which one or two specimens only were available 

 for drawing. In these instances, however, it so happened 

 that the anterior frontal hairs lie in an uniform plane, and it 

 is mainly the segments of the bodies as to which there is a 

 margin of inaccuracy. Where several specimens are available 

 greater accuracy is attained. Palmate hairs can be detached 

 and mounted absolutely flat, and error entirely avoided. 



