81 
A Nymphalid, then, is a Pierid that has got rid of the girth; and 
here it is of much interest to note that the Papilionid made an effort in 
the same direction, but only succeeded in traversing a short part of the 
distance, evolving the remarkable form we find in Thais. 
The account given of the method of suspension of the pupa of Thais 
in Scudder’s monumental work is only at second-hand, and appears to 
me to be erroneous. The existence of a double set of hooks on the 
head-prominence of this pupa is so unusual and extraordinary a pheno¬ 
menon, that one accepted without hesitation Scudder’s statement that 
this is entangled in a special silken pad. Subsequently, however, the 
examination of a consignment of pupae of Thais led me to entertain 
great doubts as to its accuracy, and to think that the nose-hooks and 
their use, extraordinary as they still are, might yet be so explained as 
to permit the ranging of the pupal structures and habit with those of 
the Papilionids, and might enable us to understand how their develop¬ 
ment was brought about. 
According to the account given in Scudder’s work, the pupa has 
three points of attachment—a pad for the tail, a girth for the body, and 
a pad for the nose. Now we know that in Papilio, Pieris, etc., there 
are only two—one for the tail, and the girth—whilst the line of 
evolution is in the direction of a loss of the girth, and thus of a 
diminution of the number of points of attachment (Nymphalids). 
These two points themselves are, no doubt, the result of a gradual 
modification of the cocoon of the Hesperids ; it seems to me, therefore, 
most improbable that a third point of attachment should be evolved. 
An examination of the pupae referred to, shows that the nose-silk is 
not a new structure but is the girth slipped forwards; also that the 
slipping forward does not occur, in many cases at all events, until the 
girth has had time to impress its marking across the wing-covers. It 
must occur whilst the pupa is still soft and capable of some activity, 
but I have not seen how it is accomplished. The peculiar curved 
attitude of the larva3 of Euchloe and Colias, after suspension but before 
pupation, may afford a hint as to how the shift originated. When the 
girth did so slip forwards, one of two things would happen; either the 
pupa would swing free, thus leading to the Nynrphalid type; or the 
girth might catch on the projecting nose-horn, and in time lead to the 
structure and habit that characterise Thais. I may note that the pupa 
of Thais has both the 5th and 6th abdominal segments free and with 
power of movement in all directions; so far, therefore, it is clearly a 
Papilionid rather than a Parnassid or a Pierid. The nose-thread is a 
loop, exactly such as the girth slipped forwards would be; it passes 
forwards from its points of origin to the nose, whilst there is a groove 
marked on the wing-covers, that is exactly in a line from the same 
points of origin, backwards (fig. 3). A close examination of a number 
of specimens shows no trace of an}'' other girth than this nose-loop, 
either on the pupa itself or in the form of any attachment to the silk pad, 
which is smooth and finished throughout, and clearly has not had any 
other loop torn from it or removed. That the pupa is tolerably active 
when the slipping forward takes place is evidenced by the nose having 
twisted itself into a second loop of the girth round the hooks in many 
instances. 
