280 
feet had five or six well-marked stout spines, also two or three scat- 
tered ones in the middle, the tubercle being rounded, convex, not 
flattened at the end. 
‘‘On the sixth segment, following the fourth pair of normal ab- 
dominal legs, is a pair of tubercles like those on the second seg- 
ment and exactly corresponding in situation with the normal legs; 
situated externally are two long straight spines, but none homol- 
ogous with those forming the crown. At the base in front of each 
tubercle is a tuft of sparse hairs, and on the outside is a chitinous 
spot bearing a dense tuft of hairs; these two tufts precisely agree 
in situation and appearance with those at the base of normal abdom- 
inal legs. 
‘‘TIn the fully fed caterpillar the tubercles are exactly the same. 
It thus appears that in the Lagoa larva the first abdominal segment 
is foatless; the second bears rudimentary feet ; segments 3-6 bear 
normal proplegs ; the seventh bears a pair of rudimentary legs; seg- 
ments 8 and 9 are footless, while the tenth bears the fully 
developed anal or fifth pair of genuine proplegs. 
‘¢ While these two pairs of tubercles differ from the normal legs 
in being much smaller and without a crown of curved spines, they 
are protruded and actively engaged in locomotion, and in situation, 
as well as the presence of the basal tufts, are truly homologous with 
the normal abdominal legs. 
<« When we turn to the work of Kowalevsky on the embryology 
of Sphinx, we find that it has ten pairs of abdominal legs which 
arise in the same manner as the thoracic or chitinous, jointed legs. 
Of these ten pairs one-half disappear before hatching, leaving the 
five pairs usually present. Itseems to us that the two pairs of rudi- 
mentary legs in Lagoa are survivals of these embryonic temporary 
feet. Although the proplegs are not popularly regarded as true 
legs, they are undoubtedly so, as embryology proves. In the 
lower Noctuide, <uch as Catocala, Aletia, etc., the larvz are at first 
geometriform, having but three pairs of proplegs ; in the geometrids 
there are but two pairs, while in the Cochlidiz there are not even 
any rudimentary feet, thoracic or abdominal. As we have else- 
where observed, the primitive lepidopterous larva must have had a 
pair of feet on each abdominal segment, and may have descended 
from Neuroptera-hke forms allied to the Panorpidze as well as 
Trichoptera.’’ 
As this and the case of Chrysopyga are unique, no other lepidop- 
