OF WASHINGTON. 371 



above the line of tubercle v. It is evident that the more this tuber 

 cle is moved up the greater is the departure from the primitive 

 condition, and hence the greater the specialization. In the Noc- 

 tuids there is a considerable range of position, varying from a 

 position below the spiracle near tubercle v to one above it, near 

 tubercle iii. It. appears, however, on going over a number of 

 species, that the exact position is a fluctuating one. None of the 

 groups as defined by Prof. Smith are absolutely constant, several 

 generalized larvae appearing in nearly every otherwise specialized 

 group and specialized larvae occasionally appearing in the gen 

 eralized groups, so that the character is not of fixed importance. 

 However, an average of the species of a group shows the general 

 tendency of that group and compares very interestingly with Prof. 

 Smith's results, as I will show. First, to refer to some of the 

 details of arrangement. 



I have mentioned that tubercle iv is not the same on the several 

 segments of the same larva. The differences follow a general 

 rule. Considering the eight abdominal segments with spiracles, 

 on the first segment iv is generally a little higher than on the 

 rest ; on the second, usually lower ; on the third to sixth, usually 

 higher, gradually more and more so posteriorly, till on the sixth 

 segment it is often the highest of all ; on the seventh segment, 

 suddenly lower, often very markedly so, the difference becoming 

 more pronounced in highly specialized larvae where this tubercle 

 seems even to recede while the others advance, so that it may, at 

 the extreme, fuse with tubercle v. On the eighth segment iv is 

 moderate or low, somewhat as on segment 2. The generalized 

 condition of tubercle iv on the seventh abdominal segment (joint 

 1 1) is specially interesting in connection with Dr. Chapman's dis 

 cussion of this segment in Acronycta larvae, since it shows that 

 his "weak segment" is not confined to this group, but extends 

 throughout the Noctuids. Indeed, it extends through the Noto- 

 dontians as well. A few species have specialized this segment, 

 as much so as the others ; I have seen two or three cases, but in 

 one at least of them I can trace a definite cause. I shall refer to 

 this elsewhere. (Hydroecia purpurifascia.} 



I stated above that the Noctuid larvae do not present marked 

 structural characters. To this the Acronyctae and Pantheids are 

 exceptions, as they possess variously modified many-haired tuber 

 cles or warts. It will be remembered that we have generally 

 associated these groups as possessing this larval character in com 

 mon, though Prof. Smith's classification separates them rather 

 widely. It is therefore interesting to note that on the character 

 of tubercle iv they are not closely allied, the Acronyctini being 

 all highly specialized, while the Pantheinaa contain a large pro 

 portion of very generalized species and have none that resemble 

 the Acronyctids in the fusion of tubercles iv and v on the seventh 



