1881.] 
295 
[Penhallow. 
minimum and quite sensitive. It was necessary that they should 
enter the tree to such a de]ith as to indicate the true interior tem- 
perature, while at the same time it was essential that they could 
be easily and quickly drawn for reading and that some means be 
provided to completely arrest all communication between the 
closed chamber and the external air. It was furthermore very 
essential to the success of experiments upon the flow of sap, 
which were being made upon the same trees, that there should 
not be the slightest leakage. These objects were all successfully 
secured in the following manner : 
With a five-eighths inch augur a hole six inches deep was 
bored in the tree. A tube of thin zinc, half an inch in diameter 
at one end and closed, but regularly enlaro’ino* to six-eighths of an 
inch at the other end, which was left open and provided with a 
flange one-fourth of an inch in width, 1 was [then wound with 
wicking and white lead at the larger end and driven firmly into 
the hole. The obvious use of a tapering plug was to provide a 
small sjnace for the sap to accumulate while the plug was being 
driven. With such a hole in the tree, it was but a simple matter 
to insert the thermometer bulb, allowing the long stem to pass 
through a perforated cork which completely closed the tube and 
made a small chamber within the tree, the temperature of which 
would be that of the tree itself. This method gave very satisfac- 
tory results and was sufficiently accurate for all the requirements 
of the case. 
The diameter of the oak at 3 ft. 3 in. from the ground was 34.8 
inches; maple Xo. 1 at 3ft. 1 in. from the ground, 10.8 inches; 
and of maple Xo. 2 at 3 ft. 1 in. from the ground, 20.9 inches. It 
should also be noted that, while the bark of the oak was an inch 
and one half thick on the average, the bark of the maple was 
hardly half an inch thick. 
1 A possible source of error has been indicated in radiation from the metal flange, of 
heat conducted outward by the metal tube. The error, however, if any, is inapprecia- 
ble and could hardly be measured. During previous experiments upon temperature, 
Mr. Wm. Trelease found that if the thermometer were simply inserted into the tree 
and the hole stopped with cotton, a high temperature maintained in a closed chamber 
about the point where the thermometer was inserted, by means of an alchohol lamp, 
did not noticeably affect the thermometer readings even after several hours. 
