180 EXPERIMENT STATION. [Jan. 



often produce some injury, which, however, may not prove 

 fatal. We believe that it is essential, in applying any sub- 

 stance to a tree as a means of protecting it against insects, 

 that one should be well within the limit of safety. 



In this connection it should be noted that all of the man- 

 ufacturers of these substances are, so far as we have learned, 

 honorable business men, as is shown by their perfect frank- 

 ness, and desire not to place any substance on the market 

 which is not reliable. However, one should be cautious in 

 purchasing newly advertised banding materials, since we 

 have reason to believe that some of these new mixtures will 

 not prove reliable. We have had opportunities to test a large 

 number of new mixtures during the past two years, none of 

 which, in our estimation, answers the requirements of a good, 

 reliable banding substance. 



Effects of Escaping Illuminating Gas on Trees. 



Undoubtedly a larger number of trees suffer from the 

 effects of escaping illuminating gas at the present time than 

 in previous years. The increased death rate from this cause 

 may be accounted for by the fact that gas is now more ex- 

 tensively used than formerly, and the larger pipes now in 

 use, together with modifications in the methods of laying 

 these pipes and calking, may be in part responsible for the 

 increased leakage. At any rate, it would seem that where 

 small pipes have been in the ground for many years with 

 thread joint connections there is much less leakage than where 

 larger pipes are used, and where the calking is done with 

 Portland or Roman cement and oakum or lead. 



There is a large amount of gas manufactured by companies 

 which is unaccounted for. According to the twenty-first 

 annual report of the Gas and Electric Light Commissioners of 

 Massachusetts, the production of gas for the year 1905 in 

 this State was 6,418,024,954 cubic feet. The amount un- 

 accounted for during that year was 622,304,044 cubic feet; 

 in other words, there was a loss of about 10 per cent. Prob- 

 ably this loss represents more than mere leakage, since part of 

 it may be accounted for by differences in temperature which 

 the gases are subjected to when measured. Nevertheless, there 



