356 
PACIFIC SCIENCE, VoL XII, October, 1958 
the old genus Dinothrombium actually belong 
in Paratrombium is correct, the probability of 
intergradation in setal types is very great. 
Nearly all of the characters listed below 
show variants which are useful in distinguish- 
ing the two species under consideration here. 
It is to be expected that as more species are 
adequately described even more characters 
and variants can be added to this list. At the 
same time, as other populations of the two 
species involved are studied, the range in 
variation of the characters will become greater. 
The variants of the characters are summarized 
in the form of a formula key. The purpose of 
such a key is not so much the identification of 
species, as to provide a convenient means of 
summarizing the observed ranges in the char- 
acter variants. Of course, species identifica- 
tion can be made simply by selecting certain 
of the more easily interpreted characters and 
determining the particular variant of each 
character which is found in the specimen or 
specimens at hand. In the case of the adult, 
the most easily interpreted characters are 1,3, 
6, 12, and 14. Few larvae have been described 
at the present time so that it is impossible to 
say what will be the most reliable character- 
istics. In the case of the two species described 
here, the form and chaetotaxy of the post- 
scutum, and the position of Si (characters 1, 
4, 10, 11, and 12) are the most easily applied. 
So far as the present data are concerned it 
must be remembered that the measurements 
shown in the tables of distribution of charac- 
ter variants in the case of the larva, are based 
upon the progeny of a single female, and the 
variation of the species as a whole must be 
greater than that shown in the table. In the 
case of the adult only one specimen in each 
case was involved; where a range is indicated, 
this is based on the variation on right and 
left sides of the one specimen. This is ad- 
mittedly undesirable from the standpoint of 
an analysis of variation in the species, but at 
least it is a beginning and shows the probable 
direction which future attempts at the meas- 
urement and expression of interspecific and 
intraspecific variation should take in Para- 
trombium and related genera. 
Characters such as the number of setae on 
each anterolateral lobe of the scutum of the 
adult are likely to be quite variable and to 
show intergradation between similar species. 
Nevertheless, it is obvious that the number of 
setae on the scutum will still be of value in 
separating species which are markedly dis- 
similar with respect to this character. The 
same is true of such characters as the number 
of setae on the genital and paragenital 
sclerites, the number of eupathidia on the 
palpal tarsus, and the proportions of tarsus 
I and tibia I. Characters dealing with the 
positions of setae on the leg segments have 
seldom been used in the differentiation of 
species, nor should they be used critically 
unless some work has been done to assess 
their variability in a series of specimens. A 
study of characters 4, 5, 6, 7, and 8, dealing 
respectively with the positions of Si, e d of 
tarsus I, fi, s 2 , and f 2 , shows that in all of 
these cases the distribution of these setae 
overlaps in the two species. However, the 
mean positions show some variations which 
are certainly significant for the material stud- 
ied, and probably for the species as a whole. 
Thus, while fi and e d are at essentially the 
same level in the two lots of larvae studied, 
Si is considerably more distally placed in P. 
bidactylus , and the range in position of this 
seta in the two species barely overlaps. The 
same tendency toward distal placement of s 2 
and f 2 in P. bidactylus is also seen, although 
the ranges overlap more than in the case of Si. 
In the tables of distribution of character 
variants, those variants given by Feider (1952) 
for Paratrombium divisipilli Feider 1948 (P. d.) 
are included wherever possible [here I use the 
name P. divisipilli as a specific rather than a 
varietal name, since it appears possible that 
P. insidare (Berlese) 1910 and Feider’s form 
may eventually prove to be distinct species]. 
Where superscripts are given, these indicate 
the number of measurements on which a 
given range and mean are based. 
