24 BULLETIN 1426, U. S. DEPARTMENT OF AGRICULTURE 
excavates a fine threadike mine, usually at right angles to the ege 
gallery, for from 3 to 5mm. At this point the larva molts for the 
first time and starts off again at right angles to the first part oi the 
burrow, but usually parallel to the longitudinal axis of the root. The 
first several millimeters of the young larva’s burrow are eaten out 
very cleanly, leaving as a residue only a blackish excrementitious 
material. As the larva increases in size its burrow often becomes 
sinuous, but conforms in a general way to the longitudinal axis of the 
root. Frequently its later extensions are nearly straight, and at this 
stage the tunnel often runs down the center of the root. This latter 
part of the burrow is closely packed with a bright brown frass in which 
the broken remains of the molted head capsules are found at intervals. 
An extensive study of such burrows indicated that these head capsules 
occurred approximately as follows: The first at 3 to 5 mm. from the 
egg niche; the second at 5 to 8 mm. from the first; the third at 7 to 
10 mm. from the second; the fourth at 10 mm. from the third, or, in 
case of apparently only four molts, at 15 to 16 mm. from the third; 
the fifth at 6 to 8 mm. from the fourth. 
Larval mines vary much in length, the shortest being nearly 20 mm., 
the longest 35 to 40 mm. long. The clean, smooth-sided pupal cham- 
ber occurs at the end of the mine, usually well down in the root, but 
sometimes on or near the crown. The pupa usually hes with its head 
toward the surface of the soil; this was believed to be invariable and 
is so stated by Schmitt (39, p. 595). However, afew pupe lying head 
downward were found September 24, 1921, on clover roots. It is 
possible that the position of the pupa is reversed soon after pupation. 
The actual number of larval molts is very difficult to determine, 
because of the difficulties met in rearing the larve under observation 
and because of the fragile nature of the lightly chitinized molted head 
capsules leftin thelarval burrows. Another difficulty encountered was 
that of tracing individual burrows in heavily infested clover roots, as 
burrows often crossed and frequently ran together, so that two larvee 
were sometimes found at different parts of one burrow. [Evidence se- 
cured from the examination of many larval burrows indicates that 
there may occur a variable number of molts, four or five, or perhaps 
occasionally only three. There was also so great a variation in the 
size of the larve at all stages, except perhaps the first, that it was not 
possible from larval measurements to assign a given larva to any 
instar. The measurements of the larval head capsules, as the most 
chitinized parts, were used for comparison. In some cases measure- 
ments immediately following the molt showed very little increase in 
the size of the head capsule. This fact was occasionally noted by 
H. L. Parker, at Hagerstown. Head capsules varied in the relative 
proportions of their length to their width, and it seems possible that 
this variation may be partly sexual, as the male adult head is rela- 
tively narrower than the head of the adult female. 
The larval head capsules cast at the first molt measured as follows: 
Length (without mandibles), 0.30 to 0.35 mm.; pleurostoma, 0.07 
mm.; width of front, 0.13 mm.; epistoma, 0.13mm. The last molted 
heads taken from the pupal chambers measured as follows: In length 
‘without mandibles), 0.55 to 0.70 mm.; pleurostoma, 0.13 to 0.14 mm.; 
width of front, 0.32 to 0.836 mm.; epistoma, 0.23 to 0.26 mm. ; 
, 
ee ee ee ee eee see re Wee ee a a 
BR rc Re lt Sal a 
neal: taf ast oo ae at ee st a a 
nt a Me 
4 
i 
i 
4 
