THE LARVit OF XANTHOPASTIS TIMAIS CRAMER 21 



structural, and would be thought to clearly indicate another species, if not 

 another genus. 



I have lately had the opporunity to observe the Mexican form in larvae 

 from Misantla, State of Vera Cruz, bred by Mr. W. Gugelmaim. They 

 resemble the Florida form quite closely, being rather more generalized. 

 They are smooth, without prominent tubercles, head, cervical shield and 

 anal segment as in the Florida form, but the segments with a row of spots 

 on anterior border, the lower two spots on each side partly joined, but 

 not joined subventrally to' the broad band on posterior border. At the 

 extremities the bands are broken into spots, on joints 2 to 5 and 1 1 to 

 1 3. This is only slightly indicated in the Florida form, where the ante- 

 rior bands show an irregular outline. The difference is, then, an advance 

 in the Florida form over the Mexican one in the loss of the subdorsal spot 

 out of the anterior row and in the strengthening of the posterior band, 

 widened and straightened and fused subventrally with the lower spot of 

 the anterior row of the following segment. Thus the Mexican form 

 agrees essentially with the Florida one, differing in characters which may, 

 somewhat violently perhaps, be considered varietal and not specific. 

 What becomes of the pattern in the rest of the vast range of the species 

 cannot be conjectured. There is no local variation indicated in the mate- 

 rial before me. The four Florida larvae are alike, as axe the two Mexi- 

 can ones. The species ranges from Maine to Argentina, and if the lar- 

 vae have chcinged as much from Mexico to Florida as the specimens 

 show, it is probable that other changes occur m the much greater distance 

 covered by the range to the end of the continent of South America. 

 However, I think that we eire entitled to assume that there is not a rad- 

 ical difference, because the territory is continuous. 



To return to the Antillean form, the conditions aure different. It is to 

 be supposed that the large tubercles are a primitive character. The spots 

 were probably formed by the breaking up of longitudinal lines, which 

 later form the transverse bands by lateral fusion. The Antillean larva, 

 then, is in a generalized condition in both respects. It still has the large 

 tubercles ; its longitudinal lines are well broken into spots, which have 

 become rounded, while the first step in forming the posterior band has 

 been taken in the enlargement of those spots. But it has gone no far- 

 ther, and there is a vast interval between it cind the continental larva. 



The conclusion seems irresistible that there are two species represented. 

 I have examined series of adults carefully, but can see no difference in 



