136 Mr. W. M. Maskell on some 
have been ‘ yellow” and what “green.” The fringes, again, 
are sometimes white and sometimes pink; and authors have 
used these as separating characters. But I have found 
species in which the fringe varied in colour according to the 
bark of the twig they were on; so that this also is fallacious. 
Taking the insects themselves, it is also very difficult to 
detect differences. All of them are of a more or less sub- 
globular, elliptical, slightly tapering form. None of them | 
exhibit any feet. All (with the exception of P. sardoa) have 
quite atrophied antenne. All have precisely similar rostra 
and sete. But it occurred to me that possibly some clue 
might be obtained from the arrangement of the peculiar 
figure-of-eight spinneret-orifices, pores, or glands (or whatever 
they are) which are common to all the species, but which do 
show, when closely examined, certain differences. 
With this object I have studied with some care a number 
of prepared specimens, with the result that I find them as 
follows :— 
One marginal row of 8's; one row of simple orifices; no 
Dorsal sae TUDEs,..2... olen sites oa saree ae er ae Jimbriata. 
Do. Do. Do. quereicola. 
Do. Do. Do. pustulans. 
Do. Do. Do. epacridis. 
One marginal row of 8's; one row of simple orifices (not 
numerous): many dorsal ibes.5 0055 ok e's co oe styphelia. 
One marginal row of 8’s; one row of simple orifices; scattered 
8’s:on dorsum (not numerous) $51. aos Sealine sue oe bambuse. 
One marginal row of 8’s; one row of simple orifices ; scattered 
S sion dorsum (NuUMeroUs)i 2k, e.  arkeacenGh o ope eee oncrdii. 
Two marginal rows of 8's; scattered (numerous) simple orifices, ventruosa. 
According to this classification épacridis, pustulans, and 
quercicola would be varieties of fimbriata ; styphelie would be 
close to them, but removed by the dorsal tubes which spring 
from simple orifices ; oncid’? would be a variety of bambuse ; 
and ventruosa would stand alone. As for costae, sardoa, and 
massalongiana, I imagine that they will be found hereafter to 
be very close to fimbriata. 
I fully acknowledge that there are many difficulties in the 
way of a satisfactory arrangement of this very peculiar genus ; 
and I admit that the external appearance and colours of the 
tests do vary somewhat, e.g. as between epacridis and querci- 
cola or oncidit and bambuse. Yet it is, to me, undoubted 
that the real basis of all study of Coccide ought to be the 
study of the characters of the insects themselves, much more 
than that of their external coverings. I have already (in my 
paper of 1893, N. Z. Transactions, vol. xxvi.) observed how 
