80 
a Parnassid, is more nearly a Papilionid they have usually been left 
among the Papilioninae. The few forms we have, however, to judge 
from the j)upge which I have been able to examine, are in reality rather 
widely separated from each other, and suggest that they are only the 
remains of a much larger family, of which the intermediate forms have been 
lost. A few words upon the several forms may therefore be interesting. 
1 . Thais (figs. 3-4-5).—Here the head spines of Papilio are either lost 
or modified into the double central knob which carries the hooks. The 
front of the pupa is curiously flattened and hollowed, by the falling in 
of the pro-thorax and front of the meso-thorax. It is to be noted, either 
as showing that Thais is rather a Papilionid than a Parnassid, or as 
showing the line of Parnassid evolution, that a genus usually left with 
Papilio, containing some half-dozen sjDecies, of which dissimilis is perhaps 
the best known, has a very similar form to Thais, though its suspension 
is of the orthodox Papilionid type. 
2. Sericinus. —Here the form and structure are very similar to Thais ; 
but, while the cremastral hooks are lost, the double nose-horns of Pa¬ 
pilio are retained; both the 5tli and 6th abdominal segments are still 
free; and the dorsal spines of Papilio, duplicated on 5, 6 and 9, are 
present; the egg is quite Papilionid. 
3. Luehdorfia .—In this, we still find susj^ension in Papilionid-fashion, 
and the pro-tlioracic spiracles have the same pit-like aspect as in Thais 
and Sericinus, but the general form is now squat (like that of Parnassius ) 
the surface rough as in Doritis, and there is movement only at the 4th 
abdominal incision ; the nose-horns are still quite distinct. 
4. Doritis. —This much resembles Luehdorfia in form and appearance, 
but has lost the cremastral hooks, as well as all power of movement; 
the pit of the pro-thoracic spiracle still appears, but is much less pro¬ 
nounced, whilst the nose-horns are reduced to obsolescence. The egg 
is dome-shaped as in Papilio. 
5. Parnassius. —This is nearly identical with Doritis, except that 
the rough surface is replaced by a very smooth one, which is covered by 
a resinous bloom ; this gives it an extremely different aspect, and is 
associated, no doubt, with the habit of making a strong, though loose 
cocoon, often in moist situations. The ventral position of the head 
indicates the thoracic pitting otherwise smoothed out. 
All these forms are evidently related, but they do not admit of being 
arranged in a linear series. The earlier forms, at least, possess the 
intersegmental sub-segment, whilst the head gradually assumes an 
inferior position, almost as pronounced as in the LycEenids. 
The Nymphalids agree with the Pierids in two very important 
points; the egg is ribbed, and the pupa possesses only lateral motion 
—all ant.ero-posterior movement having been lost, although, from its 
manner of suspension, free movement in all directions would have been 
of advantage to it. Since the loss of this movement must have resulted 
from suspension by a girth, it seems certain that the Nymplialid coul 
not have acquired it independently. It forms a remarkable illustration 
of the law, that movement once lost is never regained. 
If we imagine a Pierid pupa to have got rid of the girth and to 
hang by the tail, we have substantially a Nymphalid pupa, the 
duplication of the head-spines and some other changes being of a 
minor character. 
