42 
In the co-operative action, one has suggested it as one source of 
safety. But further, if one is correct in thinking that one undoubted 
fact may be possibly extended by further observation, the bright 
colour may be a further source of safety. 
This fact is that some enemies of insects, e.g., wasps, have been 
observed to select as their prey the 2 s, to the exclusion of the 3 , 
except at times when copulated ; and they, with this exception, with 
any sort of supply of $ s, pass over all 3 s. If this preference tends 
at all to hold good for other insects, and this, presumably, preference 
for eggs extends at all to sexually dimorphic insects, the brilliance of 
the 3 s may definitely aid in the saving of them from foes in search 
of eggs for breakfast, and one more than suspects that this is the case 
with some such foes as crickets, grasshoppers, dragonflies, at any rate, 
at some times. 
This supposition is not contradicted by such things as the marked 
cases like E. cardamines, where, except in breeding, 2 s are as com¬ 
pared with 3 s immeasurably “ rarer,” i.e., are less often recognized 
in a wild state. 
This scheme, of the 2 tending to the obscure, and to hiding, and 
of the 3 ever tending to obtrude its non-femaleness, falls into line 
with much that is known of mimicry and polymorphism. Where 
mimetic phases are adopted by the imagines, if the 2 and 3 do not 
alike mimic, which is very commonly the case, the 2 s are generally 
those that mimic one, two, or more other species for protection. 
Thus in the case of Hgpolimna .s bolina the 3 is fairly constant, while 
through Australia to India, and many other countries to Mauritius, 
the $ s have many forms in one country even, as Australia, and 
various forms in different countries. 
The mimicry of Danaids by Hgpolimnas misippus $ , in a single 
form, so entirely different from any 2 of H. bolina, while 3 s are not 
so greatly unlike, points to special development in the 2 , the 3 
merely needing to be quite different from the 2 s. The 2 of both in 
North Queensland needed especial attention to secure series, whereas 
the 3 s flaunted commonly. These facts do not seem strongly in 
opposition to the idea of 2 -hunting enemies. 
This idea, that the bright colours of such 3 s as of L. bellargus, 
E. cardamines, or the duller colours, as in E . ianira of the 3 s may 
be to some extent a warning colour to tell the foes that they are not 
2 s, does meet one supposed difficulty, which really includes two 
questions. This difficulty concerns the apparent enormous increase 
of dangers accruing from brighter colours of 3 s which are not dis¬ 
tasteful to the surrounding enemies. This difficulty is, however, 
perhaps considerably reduced by its separation into the two questions 
which one’s personal beliefs answer in the negative : (i.) Is the pre¬ 
vailing idea of greater prevalence of 3 s as compared with 2 s 
correct ? (ii.) Is it correct to consider, as is largely done, that the 
life of the 3 is considerably of less importance than that of the 2 
in the matter of continual preservation from enemies? 
The converse of both these ideas is supported by some evidence. 
Thus a proposition is advanced that the number of 2 s, on the whole, 
is not greatly, if at all, generally inferior to that of 3 s., Indeed, 
in some cases, in whole families the 2 s have been known to be 
xix. 
