

MISSOURI BOTANICAL GARDEN, 



of the stem or procumbent branel 



appears to be 



characteristic of this species only, for the other two species 

 possess the tubercles on ordinary roots, like Ceanothus. 



masses 



Elaeagnus, yet many 



m 



on Ceanothus which 



have attained nearly the same size. The dense collection of 

 tubercles of Myrica is due more to the intricate interlacing 

 of the branches than to the close proximity of the individual 



tubercles. 



between 



tubercles of Elaeagnus and Ceanothus is their color. The 

 former are very dark, due to the black cortical layer which 

 peels off and is changed to a darker color, probably by the 

 oxidizing agents of the air and soil. In Ceanothus the young 

 growing tubercles are almost colorless, but later they assume 

 a flesh color which turns still darker as they become older. 

 In Myrica the color varies from glistening white growths to 



brown or almost black mature 



As to form, 



those of Ceanothus and Elaeagnus are quite similar. The 

 former attain by far the greater length, but do not form so 



i 



thick or irregular a growth as in Elaeagnus. However, there 

 seems to be more symmetry in the growth and branching of 

 those of Ceanothus than in any other type. A great devia- 

 tion from those described is found in the Myrica forms. 

 Their branching is far more irregular and abundant. It is 

 difficult to compare them as to length and thickness. 







no 



the 



long slender thread in which may be found the central cyl- 

 inder. No indication of such a growth is found among the 

 other tubercles, however great a length they may attain. 



All the foregoing facts may be but minor differences, due, 

 more or less, to environmental conditions in which the plant 

 finds itself. The more important variations 

 in the fun mis of each host with its vnrions \ 



may 



jus of each host with its various tissues. In the 

 tubercles of Elaeagnus and Ceanothus the fungus shows 

 some similarity in behavior. In both a definite region of the 

 cortex is infected, the mode of infection does not differ very 

 materially. One variation exists in the fact that, in 



W 



Elaeagnus, the fungus is unable to dissolve, either directly 

 or indirectly, portions of cell walls after it has entered and 



