May 1891.] BOTANICAL SOCIETY OF EDINBUHGH. 103 



I. — The Pkecise Limits of Gieth-Inceease at the beginning 



AND END OF THE GROWING SEASON. 



The general limits of seasonal girth -increase in many 

 species of trees in the Botanic Garden, and at Craigiehall, near 

 Edinburgh, as far as monthly observations go, were de- 

 termined by my father and myself, and the results com- 

 municated to this Society and the Eoyal Society of Edinburgh 

 in various papers since 1880. The present observations, 

 therefore, are a natural sequel to the earlier ones, with the 

 object of attaining to greater precision. As the weekly 

 observations of 1890 give much more precise results than the 

 fortnightly ones of the two previous years, I give only the 

 former. 



(a) Period of the Commencement of Girth-Increase in 



Spring. 



The results, as given in Table I., being derived from weekly 

 observations, are, of course, only approximative, but they 

 cannot be more than a few days in excess of the truth. 

 They show a considerable range, the various examples of the 

 Pinacese starting between the 6th April and 3rd May, and 

 the Deciduous trees between 20th April and I7th May. 

 ^N'os. 19, 76, and 88 may have been in an abnormal condition, 

 as, without appearing unhealthy, they grew at a much 

 slower rate than their average for three previous years. 

 Probably the start was unusually soon, as the early spring 

 months were mild, and vegetation made rapid progress in 

 March and April. 



(b) Period of the Cessation of Girth-Increase in Autumn. 



The range lay between September 20th and October 11th 

 in the Pinacese, and between August 23rd and September 

 27th in the Deciduous trees. 



(c) Duration of the Season of Girth-Increase. 



The range varied between twenty and twenty-seven weeks 

 in the Pinacete, and between nineteen and twenty-two weeks 

 in the Deciduous trees. 



The periods given in this review are those marked by un- 

 doubtedly reliable results, but it is necessary to observe that 



