120 
The Respiratory System of Mallopliaga 
out any cell boundaries, and the membrane is of a thickness hardly 
measurable except where a nucleus occurs. In specimens that have 
been heated in potash till the soft parts were destroyed, and dehydrated 
in glacial acetic acid, the latter causes the taenidia to swell and burst 
the chitinous lining. They then tend to uncoil, and their arrangement 
may be observed. I have noticed in a dissected preparation of 
Doco'phoroides that the main trunk is composed, not of a single thread 
wound spirally, but of a band of five. The taenidia are continuous 
throughout several complete turns, but it is easy to observe their 
tapering ends in the larger forms. 
I have made no mention of the mechanism of respiration, as I have 
not made any direct observations upon it. The usual dorsiventral 
depressor muscles exist, and it is probable that air is expelled from the 
tracheal system by the contraction of these, and re-enters on their 
relaxation. 
Comparative Notes and Theoretical Considerations. 
I have made some comparison of the respiratory system of Mallo- 
phaga with that of the Anoplm-a and the wingless Copeognatha, the 
two groups with which a gradually accumulating body of evidence 
indicates their close affinity. Mjoberg (1910, p. 222) has already 
pointed out the general agreement between Mallophaga and Anoplura. 
The latter have fourteen pairs of stigmata in all the forms I have 
examined. Of these the thoracic stigmata usually overlie the meso- 
thoracic legs, but are often situated between the two anterior pairs. 
In any case the imperfectly divided thorax of the lice prevents much 
significance being attached to this difference, if difference it be. The 
abdominal stigmata are typically on segments 3-8, though in some cases 
reduction of the first segment makes it appear that the first abdominal 
spiracles open upon the second and not the third segment. The structure 
of the stigmata closely corresponds with that of the simpler Mallophagan 
type. The tracheal vessels show the same general distribution in two 
main trunks, with stigmatic branches, thoracic triangles, a posterior 
commissure, and in some cases at least (e.g. Haematopinus equi and 
Polyplax) with neural commissures anteriorly. This agreement is so 
strong that it is difficult to believe it can have any other cause than 
common derivation. 
With regard to the Psocids, I have examined some undetermined 
wingless Atropidae, and find a broad agreement, though details differ. 
