98 
Psyche 
[Sept. 
on Amaranthus and Salsola, in considerable numbers. 
Under these plants were scattered patches of Chamcesa- 
racha, a low solanaceons plant, which had hispinosa feed- 
ing on them. The species, as mentioned before, differ 
only in the male secondary sexnal characters and several 
specimens were collected from both hosts before the dif- 
ferences were discovered. Then careful collecting by host 
plant revealed that out of 87 maculata and 63 hispinosa 
males not a single one was on the wrong plant. It can be 
safely surmised that the females show the same selection. 
The question of expressing the known relationships in 
the taxonomy of the group is one to which the author has 
given considerable thought and which he has discussed 
with his colleagues in some detail. Dr. George Horn 
would probably have left at least the three very similar 
species as one, readily identifiable, species. This ap- 
proach is particularly attractive to one who attempts to 
determine numerous museum specimens but is becoming 
increasingly indefensible as we attempt to apply the 
taxonomist ’s results to problems in the field. 
Assuming that we attach names to all three forms, we 
still have at least two possible techniques, each with some 
merit. Using the extreme similarity of the three as a 
criterion, we can place them all in one species, with the 
typical and two other subspecies, with supposed geo- 
graphical replacement. This view would be strengthened 
if hispinosa were found to have a wide range in northern 
Mexico and maculata not. It serves to point out the ex- 
tremely great similarity of the three. It is weakened by 
the lack of evidence that the three hybridize where they 
meet, as in southeastern Arizona. We expect subspecies 
to be populations which have differentiated slightly be- 
hind barriers but which have not gone so far in differentia- 
tion that they cannot interbreed wherever they come to- 
gether geographically. 
The alternative method, and the one which the present 
author favors, is to call each a species. From the avail- 
able evidence we have three geographically isolated 
species which show no tendency to interbreed. Where the 
ranges of maculata and normalis meet, along the front of 
