287 [Packard. 



locomotion, nor their jaws to assist in making their way out of their 

 cells, while the liairs are nearly concolorous all over the body, though 

 very faintly shaded with yellowish on the dorsal and lateral portion 5 

 so that the species can be distinguished, as some of the specific charac- 

 ters depending on ornamentation are at this time apparent. We have 

 observed facts indicating three moultings of the skin during the so-called 

 pupa state, in distinction from the larva and imago state, and it is highly 

 probable that there are more. During the larval condition it would 

 be safe to say that there are four distinct moultings, as there are five 

 distinct sizes of larvae. In some of the eggs the larval forms can be 

 indistinctly seen, through the thin walls which we would homoloo-ize 

 with the skin of the insect after birth, for the fertilized egg must be 

 considered as the insect in its inception, in a state equivalent to the 

 larval, or pupal, or perfect state of the insect. The genus Bombus, 

 therefore, may be considered to undergo a series of at least ten moult- 

 ings of the skin, and we are inclined to think farther observations 

 will tend to increase the number. Lubbock* has described twenty in 

 Ephemera, and five have been noticed in several genera, such as 

 Meloe and others. 



The sexes of the larvi^ can be easily distinguished, as the genital 

 armor appears through the transparent skin. 



The specific differences between the larvas of the different species of 

 Bombus are of the slightest possible amount, as they only differ in 

 size, the rings of the body being smooth or rough, and in havino- 

 more or less clearly defined sutures between the pieces composing the 

 head. The eggs of the different species compared presented no 

 appreciable differences. 



In the pupa state, the two sizes of male, female and workers can be 

 more readily appreciated than in the imago state, as the insects can be 

 more easily measured and comparisons made. Corresponding cases of 

 dimorphism in other insects will probably be studied to great advantage 

 when the insects are observed at this period of life. Between the two 

 sizes of the ? in thepupaBofi>om5?<.s'y<:/T<V/«s, there was a difference of 

 .05 inch, and in the S .03 inch. In a number of the worker pupa? of 

 Bombus separatus, there was a difference of .04 inch between the two 

 broods of workers, the more advanced brood being smaller, and not 

 only shorter, but also narrower. 



In this connection, we would present some views relative to a 

 theory of the number of arthromeres composing the head of insects 

 (Jiexapoda), and the number and sequence of their appendages, suo-- 

 gested by studies of the larval forms of hymenoptera, and especially 

 the lower Neuroptera, not omitting insects belonging to other sub- 



* Transactions Linnaean Society. Vol. XXIV. Tart ii. 1863. 



