Jax. 1898.] BOTANICAL SOCIETY OF EDIXBUEGH 99 



The second great maximum is thus found in 1870, 

 when all the most favourable conditions of growth for 

 larch occurred. The curve between 1870 and 1875 first 

 falls and then rises with the sunshine and temperature, 

 while the lowest point of this period coincides with the 

 very heavy summer rainfall of 1873, almost equalled by 

 the summer rainfall of 1872. The rise of 1875 marks 

 the third of the great maxima, when the conditions were 

 similar to those of the first maximum in 1859, 



A decrease of increment took place from 1875 to 1879, 

 during a cold, wet, and cloudy period. There was a rapid 

 rise in 1880, due to a warm, dry, sunny season; a fall in 

 1881 coinciding with a cold, cloudy summer; and another 

 rise — the fourth maximum — in 1882, when sunshine and 

 temperature were greater than in 1881, and, although the 

 rainfall was greater, the relative humidity was very little 

 above the mean. From 1882 to 1889 there was a fall, 

 and then a rise with the sunshine and temperature. A 

 decided rise occurred in 1887, when sunshine and temper- 

 ature were above the mean, and when rainfall and humidity 

 were low. The fifth great maximum of 1889 coincided 

 with high temperature, an average rainfall, low humidity, 

 but low sunshine and somewhat cloudy skies. 



The fall from 1889 to 1892 corresponded with 

 decreasing sunshine and temperature; 1892 was the 

 third great minimum, and was probably due to low 

 temperature and sunshine. 



The rapid rise of 1893 appeared with the highest mean 

 summer temperature after 1865, while the sunshine, 

 though below the mean, was higher than during the 

 preceding five years. 



With regard to the curve for sunshine, the first half of 

 the period of 38 years seems to have been sunnier than 

 the second half ; but one must eliminate the personal 

 equation, because during the earlier years the estimation 

 of sunshine was very largely guess-work. 



The curve would probably be nearer the truth were one 

 to depress the first half, and raise the second half. This 

 would alter the mean, but the rise and fall of successive 

 years would remain. In this way the year 1891 would 

 be below, and 1893 perhaps above the mean. 



