19-0.] 117 



(lig". 8, h), on tlie otlier luind, is ciuite devoid of erect tilainents, tlionyli 

 — in fresh examples — it may have a marginal fringe of horizontal fila- 

 ments of the same nature. Examination of the structm-al characters of 

 the insects themselves provides a simple explanation of this difference in 

 tlie ovisacs. Eriococctis insiqiiis has a close fringe of marginal spines, 

 hut none on the dorsum ; while E. (jreeni is characterized by the presence 

 of numerous stout sjiines. scattered over Ihe dorsal surface. This clearly 



Fig. 3. — «, profile view of sac of Eriococcns rjreeni : X 10 ; b, ditto, of E. lusff/jiis ; 

 X 10 ; c, ditto, of E. inermiii : X 10 ; d, marg-in of female of E. inermis, 

 showing sunken iiores : X 450. 



indicates that it is the tubular spines that produce the straight, glassy 

 filaments that may be observed on the ovi.sacs of various species of 

 the yenus. 



The ovisac of insignis is usualh^ longer and more slender than that 

 of greeni ; but this character, taken b}^ itself, is unreliable. 



Though the insects ])resumably feed upon grasses, their ovisacs are 

 by no means invariably attached to the food-plant. They are con- 

 structed, almost as frequently, on fallen leaves, pine needles, etc. 

 Eriococciis greeni, in particular, ap))ears to prefer dead material as a 

 foundation for its sacs and sliows a .special predilection for partially 

 decayed bracken foliage, but selects only fronds that have fallen over 

 and are h'ing on the ground. 



Ei'iococcKH inermis Green. 



This species, though hitherto recorded from Surrey only, must be 

 widely distributed. I have found it occurring commonly in the neigh- 



buui'liood of Slirewsburv. 



