294 Report of tiu; Botanical Departmp:xt of the 



example, gummosis may be produced in various ways, but that does 

 not show that in nature it is actually due to any of the agents that 

 may have been used to induce it artificially. The first requisite in 

 the investigation of a disease of plants is a thorough study of the 

 environment in relation to the life and seasonal history of the host 

 and the selection for experimentation of the most hkely environ- 

 mental factor or factors and conditions of the host that when com- 

 bined may result in the disease. That the selection of the chief 

 causal factors is often very difficult is attested by the numerous 

 failures reported in endeavors to make natural inoculations on plants 

 with Avhat was thought to be the real cause of disease. 



Fungi not the first cause of crown-rot. — In a former paper it was 

 shown that crown-rot had been attributed to various causes by dif- 

 ferent authors; many suggesting winter-injury as the cause and others 

 fungi, etc. The observations recorded in the present paper show 

 that winter-injury is the first cause, or more accurately that the pri- 

 mary injuries occur during winter. The fact that fungi nearly always 

 appear on affected areas in the summer following the time of injury, 

 while some bark is still partially alive and sometimes found exuding 

 discolored " sap," has doubtless given rise to the idea that fungi are 

 the cause. But since fungi seem to be confined to dead areas or to 

 dead spots in severely winter-injured areas of bark it seems more 

 logical to hold, at least until the matter can be more definitely deter- 

 mined, that they are only the agents of decay. In the case of the 

 wood-rotting fungi found in connection with this disease a similar 

 conclusion is reached, because the wood they invade is usually only 

 that which had been stained by the after effects of winter-injury and 

 that killed by exposure. 



Can alkali he the cause?— The somewhat plausible assumption that 

 crown-rot is due to an excess of alkali^ in the soil in some sections of 

 the west where vegetation is apparently often killed by alkali, seems 

 rather unlikely in view of the fact that a very similar disease is equally 

 common in regions where alkali is not present; but more especially 

 in view of observations made by Headden^ in the same alkali sections, 

 which show that the roots of typically crown-rotted trees at some 

 distance away from the " corroded trunk " are usually normal. It 

 would seem that the more delicate peripheral roots and root-hairs of 

 such a tree would be killed by the alkaline soil solution before the 

 tree trunk could be " corroded " at the surface of the ground. 



Arsenic from spray mixtures probably has no relation to the disease. — 

 A httle more might be said here about arsenic as the causal agent in 

 relation to this disease although much of the pertinent matter was 

 discussed in a former paper. 



1 E. D. Ball. Is arsenical spraying killing our fruit trees? 



Jour. Econ. Ent. 2:142-48. 1909. 



2 W. P. Headden. Arsenical poisoning of fruit trees. 



Colo. Agrl. Expt. Sta. Bui. 157:1-56. 1910. 



