OF THE LARVA OF PYG^RA BUCEPHALA. 185 



No naturalist had previously undertaken this task, though Lypnet has mentioned here 

 and there certain valuations observed by him. 



It Avould have taken a great deal of time, and seemed to me scarcely worth while, to 

 make this comparison in all the segments of the body ; and I have therefore confined 

 myself to the third, fourth, fifth and sixth, thus including a thoracic somite, two abdo- 

 minal without prologs, and one with these appendages. 



These four segments I compared carefully in nine different specimens, five of which 

 were fed upon Oak and foiu- upon Beech. But in order to determine a few minoi- 

 points, I dissected perhaps as many more specimens in a cursory manner, which how- 

 ever confirm me in the opinion that the larger muscles scarcely vary at all. Indeed 

 the only mutation observed in them at all worthy of notice was that in one case the 

 muscles of the thu'd abdominal segment, 16, 17 and 18, were not attached to the skin 

 in front, but were continuous with the corresponding muscles of the preceding segment. 



The diflerences which do exist are not exactly those which I expected to find ; whilst 

 some things which I rather thought might vary, remain on the contrary always the 

 same. 



Let us take for example Nos. 4 and 5. In the first abdominal segment 4 passes on the 

 inner side of 5 ; the contrary, however, might apparently have been the case, without 

 alteration of the places of insertion, or of the mode of action of the two muscles ; and I 

 thought it probable, therefore, that 5 might sometimes lie on the inner side of 4. 

 But although six segments in at least fifteen specimens were examined, making in all 

 ninety instances, the arrangement was invariably the same in every instance. 



This remark is applicable to several other muscles, for instance G and 8, 7 and 8, and 

 23 and 18 in the thu-d abdominal segment ; and, indeed, I never observed a case in which 

 a muscle varied in this respect in its relation to another. 



The muscles 4 and 5, however, offer a different arrangement in the thorax to that 

 which is so constant ia the abdomen. Here, if they are rightly identified, which perhaps 

 may be doubted, though I believe that I am correct, 5 is completely double and lies on 

 the inner side of 4. 



This altered arrangement appears to be necessitated by the altered position of the 

 fore end of 4, and by the presence of 60 and 61. 



Again, 8, which in the following segments lies on the inner side of 11, in the fii-st 

 thoracic segment has shifted its position and passes on the outer side. 



A careful examination of the Plates Avill show a few more instances of this fact, which, 

 however, are not very frequent. 



The relative sizes of the different muscles appeared to vary very little in different 

 specimens, except indeed in those muscles which might be termed the variable 

 muscles. 



Such were especially 12, 13, 14 and 15 ; 28, 29 and 30 ; 56 ; 59 and 60 ; and in a 

 lesser degree, 31, 32, 33 and 34 ; 57 and 58 ; 11 ; 23 and 24 in the third thoracic segment, 

 8 in the third thoracic segment, and 40, 41 and 42. 



If we suppose, as seems probable, that muscles composing the series 12, 13, 14 and 15 

 usually act together, and that the same is the case with the other thi-ee series, 28, 29, 30 ; 



