48 TROWBRIDGE: THE THERMOMETRIC MOVEMENTS 
concave upward, and observing the change of the tension on the j 
reduction of temperature. These cords which were taut at 32° F. 
(or 0° C.), sagged at 15° F., showing an upward movement of 
branches which were concave upward. This isin agreement with 
observations made by Geleznow, who cut off branches, inverted 
them, and observed that their movement was opposite to those 
in the normal position. 
RELATION OF FROST CRACKS TO THE THERMOMETRIC MOVEMENT 
It was noticed both during the winters of 1911-2 and 1912-3 — 
that a linden showed a decided so-called frost crack or longi- 
The width of this crack varied with the temperature and closed 
up completely as soon as the temperature rose to 32° F. This was 
particularly noticeable during the coldest period of about February 
10, 1912. Frost cracks in the linden are evidently due to the same 
origin as, or connected with, the thermometric movement of the 
branches, and therefore their study is intimately associated with 
the main theme of this paper. 
The fact that frost cracks open in cold, and close in warm 
weather is well known.* In this connection an interesting and 
reliable observation reported to the writer illustrating the well- 
known noises caused by frost cracks is as follows. In very 
were found to be the sudden formation of cracks in the lindens 
which were observed at the time. 
In Fic. 17 a frost crack in a linden is shown, which was about 
three-quarters of an inch broad. Measurements of this crack 
with curves of the tree temperatures are given in Fics. 18 and 19 
some sort, not necessarily ordinary thermal expansion, acting 08 
the cells of the tree, and probably those just below the bark. 
Bye hot OO a ee 
*See J. G. Giinliniieiichiie. Science II. 38: 204. 1913. 
