142 Professor Gordon on the Temperature of the Earth. 



Rudberg at Upsala, of Arago at Paris, of Muncke at Heidelberg, of 

 Bischoff at Bonn, and others. 



Dove's researches on the non-periodic changes of the distribution of 

 heat on the earth's surface, published in 1838, 1839, 1842, show with 

 great clearness and certainty, that years of failure of crops, in general, 

 are distinguished by a sinking of the mean temperature at each place of 

 observation, for a considerable length of time. Yet, when a large portion 

 of the earth's surface is taken into view, the apparent irregularities of 

 particular seasons counteract one another, so as to give no countenance 

 to the idea, that more heat falls upon the earth generally one year than 

 another. 



As, however, the mould or plant soil is exposed to direct isolation and 

 nightly radiation, and therefore under different circumstances from those 

 of a thermometer in shade, it becomes a question, whether the temperature 

 of the upper soil surface varies uniformly with that of the air, in its 

 periodic and non-periodic changes, as in it the roots sink to greater or less 

 depth — and so, whether the soil is affected by the anomalies which 

 frequently distinguish the temperature of the air of one given year so 

 considerably from that of another ? 



It is clear, that without the solution of this problem, the temperature 

 which any plant requires for its complete development, cannot be even 

 approximately determined. 



From Dove's discussion of the Brussels, Upsala, and Heidelberg 

 observations, it is manifest that the invariable stratum referred to periodic 

 changes alone, has a determinate distance from surface, discoverable as 

 above. Considering won-periodic change likewise, however, this invari- 

 able layer oscillates. 



In years of " sea climate^ it gets nearer the surface ; in years of greater 

 difference of summer heat and winter cold, it falls deeper under the sur- 

 face. What has been said of the invariable stratum, is true, in like 

 manner, of those above it. They have a constant mean position, and 

 oscillate in particular years up and down. This oscillation determines in 

 each particular depth, the non-periodic change of the stratum. 



Diagrams of the Brussels experiments were exhibited, projected on a 

 different plan from those of the Edinburgh experiments ; the curves of 

 temperatures in the deeper strata cutting the curves of the temperatures 

 of the upper strata. The points of intersection are likewise the times at 

 which the air has its yearly mean value. It is, perhaps, for the develop- 

 ment of life in plants, a matter by no means indifferent, that, in winter, 

 when vegetation is interrupted, the higher temperature is found at the 

 roo ts — in summer, at those parts of the plant in immediate contact with 

 the atmosphere, that the times of awakening from the winter sleep and of 

 falling into it again, agree with the transition of one division into the 

 other. 



When the plant seeks heat, nature leads it to go upwards for it in 

 summer; in winter it finds it the more certainly the deeper it goes. In 



