LEAF TEMPERATURE. 51 
errors from loss of weight by evaporation while the experiment is in progress; 
(4) a substance which when mixed with water shows no temperature 
changes (the high water content of most leaves makes this important) ; (5) a 
substance which is miscible with water; (6) a substance with which sub- 
stances in the leaf produce no chemical changes accompanied by heat; (7) a 
substance obtainable in the market at a reasonable price; (8) a good con- 
ductor of heat, in order that the final temperature may be reached quickly. 
This ideal substance has not yet been found, but there are several mate- 
rials which satisfy most of the important conditions. Water, the substance 
most commonly used by physicists in work of this kind, is ruled out by the 
first condition, as it collects in drops on the surfaces of hairy leaves, leaving 
bubbles of air on the leaf which will act as non-conductors of heat. The 
second condition also helps to rule out water. The third condition might 
be overcome by using a large enough quantity to reduce to a minimum the 
percentage of error from evaporation were it not for the fact that the specific 
heat is so high that a large amount of water requires a larger amount of 
leaves than it is practicable to use in the case of small-leaved plants. Alco- 
hol answers all the conditions with the exception of the third. It is so 
volatile that troublesome precautions would have to be taken to avoid a 
very appreciable error. Common glycerine causes a rise of temperature 
when mixed with water and therefore can not be used. 
Turpentine, the substance used in the following experiments, was found: 
(1) to penetrate the leaves immediately, so that they appeared ‘“‘cleared”’ 
under the microscope; (2) to have a satisfactory combination of specific 
heat and density for the use of small amounts; (3) not to be more volatile 
than the accuracy of the other factors in the experiment as a whole required; 
(4) to show no temperature change when mixed with either large or small 
amounts of water; (6) to cause the release of no measurable amount of 
heat when the leaves of three species of plants were immersed in it; (7) to be 
easily obtainable in the market at a reasonable price; but samples from dif- 
ferent distillations show variable specific heats; (8) to have a high enough 
heat conductivity for equilibrium temperature to be reached in 2 or 3 min- 
utes. Turpentine is thus seen to answer fully all of the conditions except 
the fifth and seventh. In regard to-the fifth condition it was found that 
although turpentine does not mix with water, yet.the absence of small 
bubbles under the microscope, when the leaves used had been immersed 
in turpentine, led to the assumption that no air-spaces of sufficient size to 
influence the results existed. As for the seventh condition, the disadvantage 
arising from the irregularity in the specific heat of turpentine can be over- 
come by a little extra labor, for the specific heat can be determined accu- 
rately to the second decimal place with the apparatus used in the rest of the 
experiment or with an ordinary double calorimeter. Samples taken at 
various times from the same purchase show an agreement in specific heat, 
