FOREST TYPES IN CENTRAL ROCKY MOUNTAINS. 91 



air temperatures are not to be expected until four decades later, an 

 unusually late time, as has been explained in the discussion of air 

 temperatures. The highest average soil temperatures occur one 

 decade earlier than the highest mean air temperatures, the former 

 being slightly in excess of the latter, a fact which is made the more 

 certain by considering the low value of the soil temperatures recorded 

 at 8 a. m. The extremes at 4 feet occur two decades later than 

 those at 1 foot. 



(7) At this station and at practically all others the soil temperature 

 rises in the first or second decade of September. This results from 

 the usual cessation of rains and cloudiness near the end of August. 

 Six years out of eight show this characteristic. 



Soil temperatures evidently vary by years in about the same man- 

 ner and degree as air temperatures, and corresponding periods of 

 different years may exhibit quite different characteristics. Not- 

 withstanding these facts, it has been decided to accept each soil- 

 temperature record from the other stations at its face value, without 

 attempting, as was done with air temperatures, to compare it with 

 the control station for the same period. The reasons for this decision 

 may be stated as follows : 



(1) Most of the records here presented cover periods of three years 

 or more, with the consequent opportunity for partial compensation 

 of variations. 



(2) The failure to determine mean 1-foot temperatures more 

 directly makes great precision in this phase of the study impossible. 



Absolute soil temperatures of the types. — The records for each of 

 20 stations, by months, are accordingly presented in Tables 27 and 

 28, the period of observation for each station being shown. Figures 

 5 and 6 show some typical soil-temperature relations. 



Table 27 discloses that considerable differences in soil temperatures 

 may exist between sites which are similar as to forest cover. These 

 are equal to or greater than the air-temperature differences between 

 the same sites. Since it is apparent that soil temperatures at a 

 depth of a foot or more can have little direct bearing on growth, as 

 temperature factors, there will at first be an inclination to say that 

 soil temperatures are not so good criteria of the possibilities of the 

 site as air temperatures are, provided the latter are taken close 

 enough to the ground to represent those conditions most directly 

 affecting germination and seedling growth. As has been said, how- 

 ever, the tables show that mean soil temperatures, even for whole 

 years, do not bear a constant relation to the air temperatures of the 

 respective sites. When it is also considered that the latter have not 

 always been measured where they would best show the conditions 

 surrounding seedlings, and as the soil-temperature measurements do 

 bear a certain fairly definite relation to surface conditions, it is seen 

 that soil temperatures may have a special significance, at least in 

 indicating the degree of insolation of the site, and the maximum 

 temperatures which the seedlings will encounter. It is now believed 

 that the maximum temperatures near the surface of the soil arc 

 of the greatest importance in the distribution of the species, and 

 it is for this reason that the 1-foot temperatures during the growing 

 season will be considered in the greatest detail. 



