100 TCAXSACTIONS AND PROCEEDINGS OF THE [Sess. lxii. 



The other curves may be taken as accurate, because 

 they all are derived from the records of standard-tested 

 instruments. One may err in trying to account for the 

 rise and fall of the increment in separate years, but there 

 is a marked periodicity of years which synchronises so 

 closely with the meteorological conditions as to place 

 beyond dispute the question of its external cause. 



During the 38 years there were never more than four 

 out of the six growing months without frost in any one 

 year. In 1881, 'S5, and '92, there was only one month 

 each year without frost, but it is almost impossible to 

 trace the effect of those summer frosts in the growth of 

 the trees ; it is intei'esting to note, however, that each of 

 these three years corresponds with a minimum or slight 

 depression in all the five trees under observation. 



Finally, it seems clear that, temperature being uniform, 

 increase of rainfall and decrease of sunshine depress the 

 increment ; but it is also evident that, rainfall and 

 sunshine being uniform, the increment decreases when 

 the mean temperature of the growing six months falls to 

 or near 48° F., while the increment increases when the 

 temperature rises to or near 50° F. 



F is the second larch, but as its age was probably nearly 

 twice that of A, its curve is necessarily a very low one. 



It shows maxima at 1859, '62, '70, '75, '82, '89, and 

 '93 agreeing with those of A. 



It has minima at 1861, '85, and '88 agreeing with 

 those of ji, but the minimum of 1864 is wanting, and, 

 instead, it has a minimum at 1863, and also another in 

 1880, i.e. a year after the 1879 minimum of A. The 

 section was above the level of the lowest and largest 

 branches, and at that level the growth seems to have been 

 more directly influenced by the rise and fall of temperature 

 each year, although there are also traces of one year affect- 

 ing the growth of the next, as in 1860, '61, '79, '80, '84, 

 and '86. 



The curve of larch G is much more like tliat of A, both 

 in size and in rise and fall. The tree grew on the level, 

 with a N.K. exposure. It lias maxima in 1859, '75, '82, 

 '87, '89, and '93 agreeing with those of A. 



The only parts of the curve that are diflicult to under- 



