325 
in the aifected region; or an indirect result of their destruction, 
by reason of the clogging of the tracheae in the contiguous zone 
immediately above, by the morbid products of their degeneration ; 
no evidence in support of either of these views could finally be 
discovered. 
lt was evident therefore that some other factor existed in the 
were exuberantly healthy and turgescent; above the limit o 
although the er conducting tissues of the diseased region 
of the stem were free from fungus, they must yet be mechani- 
cally occluded in some other way. To determine the nature of 
ance of tylose formation in the several regions of the stem. In 
the duramen xylem of all portions of the woody cylinder tyloses 
were abundant, and in the peripheral actively-conducting 
tissues of the healthy wood they were not infrequently present. 
They are thin walled and in almost all cases extend in a single 
series along the narrow vessels. The tyloses were difficult of 
been to stimulate the wood-parenchyma cells to the active forma- 
tion of tyloses. : 
Marshall Ward* demonstrated that in the invasion of tissues 
by Botrytis the host cells are killed in advance of the hyphae by 
an enzymic body secreted by the young fungus cells. Brownt 
has recently investigated this active principle and has shown that 
it is possibly a protoplasmic toxin of an enzymic nature and 
that it possesses a relatively high coefficient of diffusion. 
Although it has not yet been proved of Botrytis extract it is a 
* Marshall Ward, H. A Lily Disease, Ann. Bot. vol. ii, 1888. 
+ Brown, loc. cit. 
