REVISION MALACOSOMA HUBNER IN NORTH AMERICA 105 



dorsal black stripes crossed slighdy posteriorly of the middle of each 

 segment by a vertical blue line which may cross the wavy orange line 

 between them unbroken, or which may be interrupted by it. Immedi- 

 ately posterior to the vertical blue line is a second, much fainter, 

 vertical blue line which is nearly always broken by the wavy orange 

 line. On some specimens (fig. 344) the vertical blue lines and the an- 

 terior blue patch may be greatly expanded so that the entire lateral 

 region is blue, but the anterior vertical blue line is still prominent, 

 since it is marked off by black borders anteriorly and posteriorly. 

 Laterally the mesothorax and metathorax are marked by half-moon 

 shaped black spots which are most conspicuous on those specimens 

 having much lateral blue (fig. 344). Dorsal edge of supraspiracular 

 area orange or yellowish, and setting off the subdorsal black stripe. 

 Remainder of supraspiracular and subspiracular areas varying mix- 

 tures of black, blue, orange, and gray. Ventral area mottled black and 

 gray-white, often with a darker median area on each segment. Primary 

 dorsal setae black, often tipped with white or yellow-orange ; secondary 

 dorsal setae orange, except on the eighth abdominal segment which is 

 almost entirely black dorsally and laterally, and bears black primary 

 and secondary setae which contrast with the tufted orange setae on 

 segments seven and nine (see figs. 342-344). Lateral setae orange or 

 white or intermediate shades and conspicuously tufted anteriorly 

 around setal group L2. (All larvae collected in Texas had orange lateral 

 setae as in figures 342 and 343; those from Utah had white lateral 

 setae as in figure 344, and those from New Mexico had either orange 

 or white or intermediately colored setae. In Dyar's original description 

 he notes that the lai'vae had white or orange setae on the sides, so 

 Colorado and New Mexico appear to be a region of transition between 

 these two larval types, although the inflated specimens of Dyar's from 

 Colorado in the United States National Museum are much closer to 

 the Texas form than the Utah form. Very few collections of tigris larvae, 

 however, have been made in the United States, and none have been 

 made in Mexico where tigris occurs as far south as Chiapas, so it is 

 reasonable to expect that other variations will be found which 

 will not agree with this description entirely. When more complete 

 information is available it may be desirable to recognize some popula- 

 tions as subspecies.) 



LARVAL DIAGNOSIS. — Many of the markings of tigris larvae are 

 variable from one geographic area to the next, but some longitudinal 

 orange and black lines are present and the prominent vertical blue 

 line is always present on specimens which have been examined. Possi- 

 bly one of the best characters to identify tigris larvae besides the overall 

 color pattern is the eighth abdominal segment which is mostly black 



