PHYSIOLOGICAL ANIMAL GEOGRAPHY 571 
been found in soils with a moisture content similar to that near 
Chicago. 
b. Migration of the larvae. The larvae of this species rarely 
migrate. I have watched the larvae that appeared in the exper- 
imental cages after the soil had been permitted to become very 
dry, but none of them moved during several weeks. 
c. Variation in local distribution. The distribution of larvae 
in Gary in 1906 corresponded to that of the white blossoms of 
Parnassia caroliniana which are shown in fig. 12. Their position 
varies from year to year, according to the rainfall as in the case 
of C. limbalis. 
d. Relation of the larval distribution to the distribution of 
the adults. On the pond margins where herbaceous plants have 
been displaced by shrubs, C. tranquebarica is not present, al- 
though the higher parts of the ridges are bare and much like they 
are where C. tranquebarica is present, indicating that the adults 
disappear with the larvae. 
4. Experimental studies of habitat selection a. Method. 
Adults of C. tranquebarica were placed in cages which were much 
like those which were used in the work on C. limbalis, but the 
soil was all essentially level. 
b. Results. The results were as follows: 
TABLE 2 
The distribution of ovipositor holes and larvae of C. tranquebarica under exper imental 
I 
conditions 
SAND, 9 PTS.) 
CLAY HUMUS HUMUS, lpr. CLEAN SAND) ees 4 
elo) (see eee eet 7 0 13+ 19 wanting 
1907 | TAY Oe ier ces ak 4 0 25 1 wanting 
jeelolests =: ? 3 25+ 18 wanting 
bimanvae. 09S & 11 3 31 1 wanting 
1908 { Holes. A202 oe, a: 16 wanting 29+ 11 46 
GATE AOC Abr cccer 5 | wanting 41 a 24 
c. Factors controlling egg-laying. One striking difference is 
that the females did not lay with the same precision as did the 
