102 TRAXSACTIOXS AND PROCEEDINGS OF THE [Sess. lv. 



2. The progressive rate of girth-increase in the growing 



season. 



3. The correlation between it and development of the new 



shoots. 



To ascertain these points I aimed at fortnightly observa- 

 tions in 1888 and 1889, and at weekly ones in 1890. Bad 

 weather, my distance from the Garden, and other causes 

 interfered with the regularity of the measurements, however, 

 so that in 1888 and 1889 the periods varied between twelve 

 and eighteen days, and in 1890 the weekly intervals were 

 strictly observed only in July, August, and September, when 

 the measurements were kindly taken for me by my brother 

 Mr John Christison, and by Messrs Lindsay and Eichardsdn 

 of the Botanic Garden. Fortunately, however, I had taken 

 observations much more frequently than weekly in March, 

 April, May, and June of that year, so it was easy to reduce 

 them to weekly values, and thus to equalise the intervals for 

 the whole season. 



In experiments at such frequent intervals, more delicate 

 measurements were necessary than I had used in my previous 

 annual and monthly observations, and I recorded to the 

 fortieth instead of to the twentieth of an inch. This was 

 done by using a Chesterman steel-tape, graduated to tenths 

 of an inch, the intermediate degrees being estimated by eye, 

 which, after a little practice, can be done with wonderful 

 accuracy. But in experiments this season I am using a 

 Chesterman tape graduated to millimetres. I may mention, 

 for the benefit of any one who may wish to make similar 

 delicate observations, that to ensure accuracy, I have had the 

 " Piing" S(piared and reckoned in the graduation. The squared 

 " King " is also made somewhat wider than the tape, so that 

 it can be held in position by the nail of the forefinger a 

 little below the level of the tape. Thus the tape can be 

 brought fairly over the " lling " and read off, where it crosses 

 its outer straight margin. 



Great care was also necessary in selecting suitable trees 

 for such delicate records. They were all young, perhaps 

 between ten and twenty years of age, and known to be 

 (piick growers. As far as possible, trees with smooth 

 cylindrical stems were chosen, as it is only in such that 

 measurements to the fortieth of an inch are at all reliable. 



