42 
Psyche 
[March 
that are attracted to tomato. Even granting that the inabil- 
ity to matuye on tomato is not absolute it is obvious that 
natural selection will completely prevent the development of 
any strain attracted to this plant for oviposition. 
The relation of the potato-beetle to tomatoes is an extreme 
case of mistaken instinct, but the broods in our other cages 
showed that a similar less pronounced relationship prevails 
with at least some other species of Solarium. As mentioned 
in a preceding paragraph nine species on which the beetles 
fed readily were set out in separate breeding cages in the 
spring of 1923. Each cage was supplied with twenty beetles 
(ten of each sex). No further attention was paid to them 
except to note the progress of the experiment and to replace 
from a reserve stock any plants that suffered excessive de- 
foliation. Replenishment was necessary especially in the 
case of S. rostratum as these plants were smaller and ap- 
peared to grow less vigorously. 
Numerous larvse developed in all of the nine cages and 
beetles of a second generation appeared in all of them. Plants 
were kept growing in the cages until the end of the season 
and outdoor conditions prevailed during the winter. 
The following season (1924) the cages were again sup- 
plied with plants of the same species, but, of course, no fur- 
ther beetles were added. Beetles appeared in all but four of 
the nine cages. These four included S. marginatum, S. har- 
bisetum and the two unidentified Panamanian species, indi- 
cating that the beetles had not survived the first year on 
these plants. 
In the spring of the following season (1925) only the five 
remaining cages were planted as before and three of these 
failed to develop any adult beetles. These were “wonder- 
berry”, S. torvum and S. dulcamara, showing that under the 
conditions of the experiment the beetles had been unable to 
maintain themselves on these three species beyond the second 
season. 
At this time (1925) only the two remaining cages pro- 
duced beetles of the spring brood and the experiment was 
carried no further as it appeared clear that the beetles were 
able to continue on these plants. They were S. rostratum, the 
plant believed to be the original food-plant and S. melongena, 
the egg-plant. We know of course from innumerable obser- 
