ne ales 
of six well defined forms which occur regularly, and a seventh of 
which only two specimens are known and which may be regarded 
as an accidental colour monstrosity. Individuals which stand inter- 
mediate between any two forms are rare, but do occur. Systema- 
tists, until recently, have considered these Butterflies as being eight 
distinct species. The extremes are a variety with a broad white 
band : f. platydesma KR. and J. (1906), and a variety without any 
band : f. pomponius HOPFF. (1866). The four other principal forms 
have a white band or patch on the forewing, the hindwing also 
bearing a white central area or being without it. The form which 
probably most nearly resembles the ancestral one is f. /ysithous 
(fig. 31@). It has the widest distribution, and the pattern of all the 
other forms can casily be derived from that of f. lysithous. The six 
forms, however, cannot possibly be regarded as six phyletic stages 
of one straight line of evolution; they represent several lines 
of development. 
All these varieties bear a remarkable likeness to Aristolochia- 
Papilios of the same country. In fact, the whole group of Papilios 
to which /ysithous belongs was always classified with the Aristo- 
lochia-Papilios, until an ardent believer in mimicry, E. HAASE, 
discovered that the resemblance was merely superficial. We have 
here in P. /ysithous a similarity between a single polymorphic spe- 
cies on the one hand and a series of distinct species as « models » 
on the other. The variability of P. lysithous is yet more complicated 
than it appears from these bare statements. There is a geographical 
element in the resemblance. The six main forms do not all occur 
together; only three of them may be expected to be met with in 
the same district. The forms with broad band and large submar- 
ginal spots are confined to the northern provinces of South-Eastern 
Brazil ( from Bahia to Rio), and the black f. pompontus (pl. XXIV, 
fig. 330) as well as f. rurik (pl. XXIV, fig. 32a) are found arder 
south; but the exact limits of ista are not yet known. Now, 
it is significant that the distribution of at least two of the « models » 
has somewhat the same limits as that of the « mimics ». P. lysithous 
f. platydesma flies in the province of Rio, which is also the habitat 
of its « model » P. ascanius, and P. lysithous f. pomponius occurs 
in Paraná and further south, where is also found its model P. per- 
rhebus (cf. pl. XXIV, fig. 31-33). 
The majority of species to which 2. /ysithous belongs exhibit a 
more or less close resemblance to Aristolochia-Papilios, many of 
