188 Mr. A. Mallock. Growth of Trees, with a Note on 



For the much more rapid growth of trees, however, it is convenient to use 

 an arrangement which demands a greater variation of angle to cause the 

 same amount of shift in the bands, and this can be secured by merely forming 

 the bands by light having a grazing instead of a normal incidence. For this 

 purpose the plate A is replaced by a right-angled prism. The bands now 

 formed are not equidistant and have several peculiarities which need not be 

 here particularised. The theory is given in the note at the end of this paper. 

 What is of importance for the present purpose is, that the alteration of angle 

 between A and B necessary to shift one band to the position formerly occupied 



Bat ler>- aud Lamp 

 for 



■ r 



Fig. 1. 



by its neighbour is more than ten times as great as when the incidence of the 

 light is normal. 



The arrangement for using these bands in the measurement'of 'the growth 

 of trees is shown in figs. 1 and 2. At the place of measurement (usually 

 about 5 feet above the ground) a tape of " invar " is passed round the trunk, 

 the roughnesses of the bark having been previously smoothed with a rasp. 



