Part I.] 
Troup : Teak Forests of Burma . 
15 
All the other forests hitherto enumerated have less than 15 per cent, 
of teak in their composition. The lowest proportion of teak in true teak 
reserves hitherto enumerated is 6 per cent., in the Gwethe and Saing 
working circles of the Toungoo Division. There are, it is true, certain 
plains forests with a lower percentage, namely, Thindawyo (5 per cent.), 
Tonkan (2 per cent.), and the Rangoon Plains forests (1 per cent.) : these 
forests, however, were reserved for the supply not of teak, but of miscel¬ 
laneous species, and cannot be classed as teak reserves. 
(4) Relative proportion of teak of different age-classes. 
Taking the average of all reserves in which all five girth classes of teak 
have been fully enumerated, we find, as a general average for the whole 
of Burma, that the percentage of teak in each girth class is as follows :— 
I. 7 feet girth and over ..... 7 per cent. 
II. 6 feet to 7 feet in girth ..... 5 „ 
III. 41 to 6 feet in girth . . . . 10 
IV. 3 to 41 feet in girth . . . . . 17 
V. Under 3 feet in girth . . . . .61,, 
100 
A reference to Appendix II will show that the proportions in the 
various reserves differ very widely from this average, and it is no doubt 
an open question whether the general average truly represents the state 
of affairs which should exist in a normal teak forest. It forms, how¬ 
ever, some basis of comparison between the several reserves dealt with : 
for example, the deficiency of large trees and the excess of small trees in 
the Pegu Yoma forests of the Pyinmana Division are strikingly illus¬ 
trated in this way. 
(5) Teak growing-stock in specially rich sample-plots* 
The query, “What constitutes rich natural teak forest?” is best 
met by the reply that it depends entirely on the area in question. We 
have dealt hitherto with large forest tracts, running in some cases to hun¬ 
dreds of square miles, and have considered the information furnished 
* In this section the term sample-plot is taken to include linear valuation 
surveys. 
[15] c 
