March 30, 1905] NATURE 



523 



is 0015 calorie per square centimetre of leaf-S!ir/ace per 

 minute, for a temperature gradient of 1° C, so that the 

 temperature difference X-t' will be represented by 

 o 04938/2 X 0015 = i''-64 C. 



For the simultaneous determination of the temperature 

 difference *-(' and the amount of water transpired, we 

 employed two differential platinum-resistance thermometers 

 each consisting- of about 24 metres of fine wire arranged 

 in a mica and ebonite plate so as to form a flat grid, 

 against the two sides of which two similar leaves were 

 lightlv pressed and held in position by ebonite frames 

 furnished with cross-threads of silk. The two leaf-laminae 

 were thus in close apposition to the resistance-coils, which 

 were favourably placed for rapidly acquiring the mean 

 temperature of the leaves, which were supplied with water 

 from two small fubes attached to the frames. \ definite 

 area of leaf-surface was exposed, amounting in each case 

 to 139-4 square centimetres. The loss of the water of 

 transpiration was determined by weighing the apparatus 

 at suitable intervals. 



The difference in temperature between the two coils was 

 determined by means of a Callendar's recorder. Instead 

 of determining the difference of temperature between the 

 leaf and the surrounding air, it was found more con- 

 venient to clothe both coils with leaves, but to arrange 

 them in such a manner as to produce differential transpira- 

 tion between the two pairs, a result which can in most 

 cases be brought about by arranging one pair of leaves 

 with their dorsal sides turned to the platinum roils, and 

 the other pair with their dorsal sides facing outwards. 

 Owing to the comparatively rapid thermal adjustment 

 which takes place, the results are not affected by the 

 gradual closing of the leaf stomata during an experiment, 

 provided the record is correctly integrated so as to give 

 the mean difference of temperature. From this mean 

 difference of temperature between the two pairs of leaves, 

 and the differential transpiration corresponding to this, 

 the thermal emissivlty of the leaves is readily calculable. 



As an example, we may take an experiment with the 

 leaves of Liriodcndron tulipifera, in which the experiment 

 lasted 129 minutes. The difference in the amount of water 

 transpired by the tw'o pairs of leaves was o-^io gram, and 

 the mean temperature difference was i°-4i C. Taking the 

 latent heat of water at 593-6 calories, it follows that 

 0.510x5936 = 3027 represents in calories the excess of 

 energy which must have entered the cooler pair of leaves 

 from their surroundings, an excess which is conditioned 

 solely by the temperature gradient of i''.4i representing 

 the difference of temperature between the two sets of 

 leaves. The surface area of the leaves exposed was 1394 

 square centimetres, so that the thermal emissivity of a 

 square centimetre of leaf-surface per minute for a 1° C. 

 temperature gradient will be 



302-7/129 X 139.4 X '-41 =001194 calorie. 



As examples of the extent to which the thermal emis- 

 sivities of leaves of various plants differ, the following 

 may be given. They represent the emissivity under con- 

 ditions of still air : — 



Thermal Emissivity of Leaves of Various Species of 

 Plants, under Still-air Conditions. 



Therma emissivity in calories 



per sq. cm. of leaf-surface for a 



Species of Plant I'C. excess of temperature 



Per minute Per second 



I.iriodendron tulipifera (a) ... ooiiq ... 0000199 



,, ., (b) ... 00127 •■• 0000212 



Heliaiithus viultiflorus 0.0150 ... 0000249 



Tropoeolum niajus 0-0142 ... 0-000237 



Tilia europoea 00159 ... 0000266 



Under ordinary outdoor conditions we never have to deal 

 with perfectly still air, and the inquiry had therefore to 

 be extended to the influence of moving air currents on the 

 thermal emissivity of leaves. 



This was investigated by observing the differential 

 temperat^jre and differential transpiration when the two 

 pairs of leaves were placed in a shaft through which a 

 current of air was passed having a definite and steady 



NO. 1848, VOL. 71] 



velocity. The results of two such experiments with leaves 

 of Liriodcndron tulipifera and Helianthus multifiorus are 

 given in the figure. It will be seen that the effect of the 

 cooling or heating due to the air is a linear function of 

 the velocity, the coefficient of thermal emissivity of the 

 leaf-surface increasing at the rate of 0017 calorie per 

 square centimetre per minute for an increased velocity of 

 the air current of 100 metres per minute. This effect of 

 moving air in dissipating the excess of radiant energy 

 falling on a leaf is a very iiTiportant fact in. the economy 

 of some plants in which transpiration is reduced to a 

 minimum, and it is one of nature's means for preventing 

 the rise of temperature in strongly insolated plants from 

 reaching a dangerous point. 



We must now turn our attention to the thermal relations 

 of a leaf to its surroundings when it is receiving direct 

 solar radiation, and here again, for the purpose of simplify- 

 ing my argument, I inust ask yoLi to imagine an ideal set 

 of conditions under which a healthy leaf, well supplied 

 with water, is exposed to sunlight of constant intensity, 

 and that there is no variation in the temperature, humidity, 

 or degree of movement of the surrounding air, or in the 

 dimensions of the leaf stomata. 



As in the previous case, a state of thermal equilibrium 

 will be speedily established between the leaf and its 

 environment, when the simultaneous loss and gain of 

 energy will just balance. 



When this condition is attained, let R represent the 

 total radiation falling on i square centimetre of the leaf 



-Influence of moving i 



tyofle. 



in one minute, and, further, let the " coefficient of absorp- 

 tion " of the leaf for this radiation be represented by a ; 

 then Ra will represent the radiant energy absorbed per 

 square centimetre of leaf-lamina per minute. 



At this stage it is of some interest to give absolute 

 values to R and a in order to see what would be the 

 thermometric effect produced on a leaf by ordinary sun- 

 shine in default of there being some ready means of dis- 

 sipating the absorbed energy. 



If we denote the mass of a square centimetre of the 

 leaf-lamina by m, and its specific heat by s, then on the 

 above assumption the rise of temperature of the lamina 

 per minute will be represented by Ro nis. 



Let R = o-8 calorie per square centimetre per minute, 

 which represents the intensity of ordinary summer sun- 

 shine in these latitudes. 



Let a, the coefficient of absorption, be 0-78. a value 

 which is determinable by a method presently to be de- 

 scribed : further, let the mass, m, of a square centimetre 

 of leaf be 0-020 gram, and its specific heat 5 = 0-879, then 

 the rise of temperature of the leaf under the conditions 

 postulated will be at the rate of 



0-8 xo-78/oo2X 0879 =35°-4 C. per minute, 



a result which would be speedily fatal to the leaf. 



The dissipation of the absorbed energy necessary to keep 

 the temperature of the leaf within working limits is pro- 

 vided for, on the one hand, by the internal work of the 

 leaf, consisting mainly of the vaporisation of water, and 



