1913.] PUBLIC DOCUMENT — No. 31. 47 



off. The iippor limbs iisuallj lose tlicir bark first, and even- 

 tually the larger limbs present the same appearance. Disinte- 

 gration may become so rapid that in one and a half to three 

 years most of the larger branches will break off, and soon noth- 

 ing bnt a portion of the trunk and a few stubs remain. 



All these conditions refer merely to the way in which a tree 

 succumbs to gas poisoning, and do not necessarily constitute 

 reliable symptoms of this type of injury, as these symptoms 

 may be found in trees dying from other causes. The tissue 

 furnishes the most reliable symptoms for diagnosis. The writer 

 has for years been examining thousands of trees killed by gas, 

 and has taken exhaustive notes on every symptom shown by 

 trees dying from various causes, and from the first it was found 

 necessary to make a thorough examination of the tissue to war- 

 rant any degree of accuracy in the diagnosis. From these notes 

 may be had much interesting data on the relative importance of 

 various symptoms. Since some of the poisonous constituents of 

 illuminating gas are absorbed by the roots and are circulated to 

 a certain extent through the tissue of the wood, it becomes nec- 

 essary to note these svibstances and learn to detect their pres- 

 ence, either by chemical means or from direct observations and 

 experience. 



As already stated, no two species of trees present precisely 

 the same symptoms, much variation occurring from location, 

 seasons, etc. A certain species, for example, examined in the 

 fall, will show slightly different symptoms from those given by a 

 spring examination, and this is true of trees poisoned by gas 

 from different manufacturing plants, due undoubtedly to the 

 fact that there is considerable difference in the gas manufac- 

 tured by different concerns. The different chemical constitu- 

 ents of the soil in remote localities might to a certain extent 

 be held responsible for the variations in the reaction of gas on 

 the tissues, but this is probably not very important, since these 

 variations are likely to occur even in a single to^\'n supplied 

 from one source, and as a rule the symptoms in trees injured 

 by gas from a single manufacturing plant are alike. From 

 tables giving the results of gas analysis from various corpora- 

 tions we find that there is considerable difference in the composi- 

 tion of gas, and furthermore, that gas from a single corporation 



