NO. II LARVAE OF THE ELATERID BEETLES GLEN 3I 



The subtiasale is a useful supplementary feature, but it is seriously 

 eroded too frequently to be generally reliable. 



Mandibles of distinct types provide splendid separating characters. 

 Erosion dulls and shortens the points, but only minute carinae and 

 denticles are likely to be completely worn away. 



Antennae are of limited value. Sometimes the shape of the seg- 

 ments or their relative length is characteristic, but more often it is 

 the number of "sensory" appendices on the second segment. This 

 character is decidedly valuable, but as with many other sensory struc- 

 tures a knowledge of normal variation is essential to its proper use. 



Eyes present or eyes absent is an excellent supplementary charac- 

 ter, but the pigmented ommatidia may be displaced, especially in 

 prepupal and premoult specimens. However, in species that have eyes 

 the epicranial plate is not pigmented in the region normally occupied 

 by the eye, and the presence of a clear spot at the appropriate location 

 may be taken as evidence of the presence of eyes. 



Spiracles are difficult to examine or to use except in their grosser 

 aspects. Marked differences in size, shape, or position in the segment 

 offer good characters, but the writer has found it impractical to 

 attempt distinction on the basis of the number of transverse trabeculae. 

 In fact, the basic structure of the bifore spiracle seems to be a con- 

 troversial matter requiring further careful research. 



The power to appraise descriptions and to use them is materially 

 increased if one fully appreciates the main factors concerned in the 

 making of adequate larval descriptions, namely, suitable material, 

 detailed examination, and comparison with closely related species. 



Suitable material must include larval exuviae of available reared 

 adults, as proof of identification, but whole larvae are necessary for 

 proper morphological examination. Several specimens of each species 

 should be available so that individual variation can be appraised. The 

 structures that vary and the degree of variation differ from species 

 to species and cannot be predicted. 



Detailed examination is essential. Experience has shown that larvae 

 of closely related species are morphologically very similar. Specific 

 characters frequently are inconspicuous and are entirely overlooked 

 or inadequately described in a superficial study. 



Wherever possible the study should be comparative and the com- 

 parison must include the closest allies known. This point is illustrated 

 by the Lndius larvae occurring in Saskatchewan, where nine species 

 are recognized in the larval stage. These belong to eight different 



