368 



APPENDIX 



quantity of water. A similar fact had already been observed in 1878 and 1890 in the 

 reports published by Mathieu and M. Bartet ^ on the results obtained in those years by 

 the "Station de recherches" of Nancy. 



This is the more instructive, inasmuch as it is the winter rains which are the most 

 important from the point of view of the feeding of springs. 



The fact is very natm-ally explained by the condensation of aqueous vapor on the 

 considerable surface presented by the crests of forest trees.^^ 



At all s<;asons of the year and all hours of the day the trees are, in fact, colder than 

 the surrounding atmosphere. Nevertheless the difference is greater in summer than 

 in winter, and during the day rather than at night. It is also greater at the base of the 

 trunk than among the branches. 



Here are some figures on the subject. The Swiss observations give results recorded 

 on an average of 12 years; those in Bavaria were xmdertaken during two years only: 



DIFFERENCES (IN CENTIGRADE DEGREES) BETWEEN THE TEM- 

 PERATURE OF THE TREES * AND SURROUNDING AIR 



Larch (near Interlaken) 



Spruce (near Berne) 



Beech (near Porrentruy) 



Miscellaneous species (Bavaria) 



0.4 

 0.9 

 0.3 

 1.3 

 1.4 



* In the Swiss experiments the temperature of the tree was taken at breast-height, 

 while in Bavaria it was in the branches of the crown. 



These reports suffice to give us the key to the important phenomenon noted above. 

 Trees with considerably spreading crowns form very excellent condensers of vapor 

 from atmospheric water which they conduct to the soil in a hquid condition; this process 

 is naturally more marked at the beginning and end of winter, periods when the atmos- 

 phere of our latitudes, especially in forests, is very near the point of saturation. I'he 

 slightest lowering of the temperature is then sufficient to produce condensation. 



In conclusion, the results obtained by the observations conducted at Cinq-Tranchees 

 can be summed up in the following manner: Of a hundred miUimeters (3.94 inches) of 

 atmospheric water, the instrument placed under shelter of the crest of a young horn- 

 beam only received in an average year 92 m.m. 4 (3.64 inches); therefore 7 m.m. 6 

 (0.299 inch) were retained by the crown a balance produced by the amount of condensa- 

 tion from branches and the top of the trunk. 



22 " Meteorologie Comparee, Agricole et Forestiere. . . . Observations faites a 

 la Station de Recherches de I'Ecole National Forestiere," published by M. E. Bartet, 

 Paris, 1890 (" Bulletin du Ministere de I'Agriculture"). 



23 It happens sometimes that a considerable quantity of water is supplied to the 

 forest soil in a verv short period of time. This occurred in .January, 1882. The east 

 wind accompanied'by fog had deposited such a quantity of hoar frost on the trees that 

 a great number of branches were broken by the weight of it. This first took the form of 

 icicles ten centimeters in length. A twig covered with them, cut off with great care, 

 weighed 550 grams; relieved of its burden, it only weighed 70 grams ("Bulletin de la 

 Commission Meteorologique de Meurthe-et-Moselle"). This frost in melting on January 

 17 was equal to a rainfall of 7. m.m. 4 according to the ram-gauge under the trees at 

 Cinq-Tranchees. In Januarv, 1879, a branch of birch from the forest of Fontainebleau, 

 weighed by M. Croizette-Desnoyers, covered with a thick crust of rime, turned the 

 scales at 700 grams. Weighed again after the frost had melted, the result given was 

 only 50 grams. 



