rvoL. 2 



630 ANNALS OF THE MISSOURI BOTANICAL GARDEN 



shoot infected but not deformed. These leaves may be almost 

 normally green on the upper side or they may be more or 

 less reddened, sometimes to carmine red; on the under side 

 they become clothed with the felty fructification of the fungus 

 but the leaves are not deformed. This is merely a more gen- 

 eral infection than the leaf type a, described above, and is 

 often associated with it on the same plant as well as with the 

 leaf spot and leaf concavity forms. 



(c) Bag gall of Andromeda lignstrina. This is the ex- 

 treme in gall production. This gall finally becomes a hol- 

 low bag which attains a maximum size of 10-15 cm. in length 

 by 5-10 cm. in diameter. These bag galls are either terminal 

 or lateral on leafy shoots of the current season's growth. 

 When lateral, such a gall has the morphological position of 

 a leaf. 



(d) Bud gall of Symplocos tinctoria. The expanding leaf 

 buds are deformed into a subglobose mass which may be 3-3| 

 cm. in diameter. In this gall, the undeveloped stem of the 

 bud is greatly enlarged and the individual leaves of the bud 

 are greatly thickened and deformed. 



In North America, we have a large number of species of 

 Ericaceae which produce galls when infected by Exohasidium. 

 The specimens w^hich have accumulated under Exobasidimn 

 in herbaria show that none of the gall forms which I have 

 designated under distinctive names in the preceding para- 

 graph are isolated forms. Favorable hosts show a connection 

 and gradation between the various gall forms as intimate as 

 that presented by Woronin for Vaccinium vitis-idaea. How- 

 ever, the terms which I employ are useful for contrasting and 

 comparing the data presented by the specimens which I have 

 studied. These data are later given in tabular form. 



The microscopic examination of an Exobasidium gall shows 

 that it is composed principally of the tissues of the host plant. 

 Hyphae of the fungus ramify about between the cells of the 

 host and, in the galls in which deformation has taken place, 

 the presence of the fungous hyphae has caused the host both 

 to multiply and enlarge its cells in the infected region. The 

 gall is, therefore, a direct product of the host plant, which 



