OF TREE BRANCHES AT FREEZING TEMPERATURES 30 
position of the medullary canal in some species had a marked 
influence on the phenomenon, but there were some anomalies 
which could not be explained. Experiments made in 1865-1867 
showed a slight relation of the water content of the branches to the 
thermometric movement of the branches, but the connection was 
not at all definite. The work was undertaken largely to find an 
explanation of Caspary’s work at K6nigsberg, already referred to. 
has M ee ee eS ie 
~e~e—4 HEIGHT OF BRANCH _| w 
TUT Branch A _| : 
IN Tree a1] ef 
es LL | AAGTAL [\ ee bi a 
im CO 8 FA PR ee LO 
| Xt | PLOT a terror et be 
: eS: \ YT act \ =. : ed 
é an acl iN AA\ [ | sRelative Humidity % ais, : i 
60 kK or. YN [TN U.S. Weather Bureau | : i 
Fd \ 4 adcuk 40 
at no 38 
§ = nacre hy aha 30 
55 § Daily Record 25 
: ; i 
| 10° 
[ 5° 
50 58410 IG 16 11 16 19-29 2425 21 20-12 18 - 
Fic. 5. Daily record of Branch ‘A, Tree No. 1 (European linden) during A pril 
and May, 1912. Little or no movement appears to accompany the large changes 
of both the air temperatures (all above 32° F.) and the relative humidity. The 
depression in the branch during May is due to the weight of the leaves. 
A letter in Nature, entitled, ‘‘ Position of boughs in summer and 
winter,’ by Agnes Fry,* records the height of branches of both a 
mulberry and a walnut tree in summer and in winter. The 
measurements show the reduced height in summer due to the 
weight of the leaves. The changes amounted to one foot or 
slightly more. This note is not pertinent to the subject under 
discussion since the movements observed were not thermometric. 
In 1898, there appeared a paper of similar character to the 
last, entitled, ‘‘ Preliminary observations of the seasonal variation 
of a branch of a horse chestnut tree,” by Miller Christy.T 
* Nature 54: 198. 1806. 
+ Jour. Linn. Soc. Bot. 33: 501. 1898. 
