OF WASHINGTON. 149 



NOTES ON THE STRUCTURE OF THE THORAX AND 

 MAXILLAE IN INSECTS. 



By NATHAN BANKS. 

 (PLATE I.) 



When one looks at the complicated structure of the thorax in 

 a fty or a bee, it seems almost impossible to tell much of the 

 origin of the parts. But it is not so difficult to unravel the tangle 

 if we start at the right end. Several years ago while looking at 

 some Chilopods one point in the solution occurred to me. I 

 published upon it at that time, and since then have often been 

 confirmed in the view then expressed that the segments of the 

 thorax are compound. It is generally admitted to day (more so 

 than ten years ago) that the Chilopods are closely related to the 

 ancestors of the winged insects ; and no one, I think, can study 

 them. without arriving at that conclusion. 



If we look at one of the lower and simpler Chilopods, as a 

 Geophilus or Mecistocephalus (Plate I, fig. S), we will notice 

 that the head is followed by a series of segments, subequal in size, 

 each bearing a pair of legs, and (except the first) spiracles. If 

 one examines a more compact Chilopod, as Scolopocryptops, two 

 kinds of segments are seen, the larger with spiracles, the smaller 

 without them. In Lithobius (Plate I, fig. i) this process has 

 gone further, the small segments are smaller, the large are larger, 

 and overlap or even cover the small ones. In Cermatia the dor 

 sal scutas of the small segments have coalesced with the large 

 segments. This, I believe, has happened to all insects ; what we 

 consider a segment of the thorax being really two segments 

 united. There are various external points of proof. One that I 

 previously advanced is that in Machilis there are jointed append 

 ages to the abdomen which appear to represent legs. Similar 

 appendages are attached to the middle and hind coxae ; these, 

 then, may represent the legs to the other segments of the thorax. 

 Three years ago Mr. Walton called attention to the structure ot 

 the coxa, showing that there are present the basal parts of two ap 

 pendages ; the trochantin and epimeron representing basal joints 

 of a leg now no further developed. He reasoned from this that 

 the thorax was compound. He did not then know of my article. 

 I have given (Plate I, fig. 7) a side view of the coxa of a Pa- 

 norpa where one can see that the coxa shows these two sets of 

 basal joints. This appearance is readily noted in most Neurop- 

 tera (very clearly so in Psocidae), and can be made out by a little 

 study on many of the higher insects. 



In my note of ten years ago I stated that I thought the pro- 

 thorax might not be compound, but now I consider that it is 



