250 BOTANICAL GAZETTE [APRIL 
harvested so completely by rock squirrels that, out of the hundreds 
on the slopes near the laboratory, only two or three were allowed 
to stand in place until the seeds were escaping from the capsules. 
The three surviving treated stalks of this plant that were saved 
were preserved by wrapping the stems in cloth. The losses in 
Eschscholizia were very great, by reason of the quick response of 
this plant to injuries. Echinocereus failed to set a single fruit 
that had been treated during the first year, although quantities 
of seeds from treated ovaries have since been secured. Sever 
dozens of ripening capsules of Argemone were destroyed by birds; 
Mentzelia matures but few seeds in each capsule; Opuntia fails to 
set fruits in many cases, owing to the injurious effects of the opera- 
tion and failure to pollinate. Carnegiea gigantea drops many 
flowers that have been operated upon and fails to pollinate many, 
so that not more than 12 or 15 per cent of the operations were suc- 
cessful with this plant. 
These data concerning the fatalities encountered are of interest, 
since they deal with plants wholly in a state of nature, and repre- 
sent with fair measure the chance of survival in similar cases that 
might occur without human intervention. 
Some attention has already been given to the conditions found 
in Pentstemon Wrightii, a small beard-tongue native to southern 
Arizona. This plant displays a wide range of variability in its 
various habitats. The plants treated on Tumamoc Hill, however, 
are restricted in the range of somatic characters, their progeny 
from treated ovaries displaying many supposedly new qualities. 
A most diligent observation has been made of the wild plants to 
distinguish naturally recurring characters from those induced. ¥ 
now seems fairly certain that some new characters are found in the 
treated progenies. The actual examination of these forms 1S 
fraught with great difficulty, since it has not been found possible 
to secure fertilization in guarded inflorescences, due probably to 
the fact that the action of visiting insects has not been correculy 
imitated. 
_ It is evident that the operations described above ent 
introduction of reagents into the complex ovular mechanism, ' 
no rational interpretation would be possible without an analysl 
ail the 
and 
5 of 
