1906] WIEGAND—BUDS AND TWIGS IN WINTER fe) 
403 
be for all practical purposes not over about thirty minutes, no mat- 
ter whether it was cooled very much or only a little, providing it 
was plunged directly into the cooler temperature. 
We may also say, I belicve, that smaller buds with thinner scales 
and smaller shoots will show a time period correspondingly less 
than thirty minutes, and a time difference which will approach 
more nearly zero. In the case of the willow buds with only one 
thin bud-scale, the time period and time difference must be very 
small indeed. 
A number of readings were taken in which the thermometers 
were warmed up instead of cooled, and it was found, as expected, 
that the above generalizations applied in this case also. Providing 
that atmospheric changes out of doors are abrupt, I fail to see how 
the temperature at the center of a bud of medium size can be retarded 
more than five or ten minutes over practically all of the range of 
fall. A small bud would probably be retarded only about one to 
five minutes. Of course the retarding would be greater than this 
through the last degree and fraction of a degree, but this slight 
change, it seems to me, would be of little moment to the present 
question. 
Buds in nature, however, are under slightly different conditions. 
Instead of being transported from one temperature to another, 
the temperature itself changes. We should therefore conduct 
some experiments in which the air itself is varied. This change 
is either very gradual, as when a thaw approaches, or more abrupt, 
as when the sun shines from behind a cloud upon the bulb, which 
is the only way in which abrupt changes are produced in nature. 
In either case they are much less violent than were our laboratory 
experiments. During warming by the sun, radiation from surrounding 
objects may play an important part and introduce still another 
factor. We should thcrefore conduct some experiments in which 
the air itself is warmed. The experiments with the horsechestnut 
bud already described in the discussion of the function of color 
are to the point here. They show in a surprising way that instead 
of retarding the rise in temperature within the bud, under these 
very natural conditions the bud-scales actually scem to hasten it. 
These experiments were with direct sunlight. It seemed possible 
