66 
W. B. McDOUGALL 
fungus, offers an explanation of the coral branching clusters of mycor- 
hizas. It is well known that when the growing tip of a shoot is cut 
off, or otherwise inhibited from further development, the shoot is 
stimulated to excessive branching. If the growing tip of a root, 
therefore, is inhibited from further development by a mycorhizal 
fungus, it is reasonable to expect that branching will result, and, if the 
branches are in turn infected by the fungus, the ultimate result will 
be a coral cluster of mycorhizas. 
In regard to the development of endotrophic mycorhizas, little 
can be said. We have shown that infection is necessary to the pro- 
duction of the beads of Acer mycorhizas, and that each bead is pro- 
duced by separate infection from the mycelium in the soil. Future 
work must show whether several beads on the same root may be pro- 
duced simultaneously, or whether they must be produced consecu- 
tively. Shibata (22) found well developed nodules on the roots of 
Podocarpus which were uninfected by the endophyte, and con- 
cluded that infection is not necessary to the production of nodules. 
It has since been shown by Spratt (23), however, that the nodules of 
Podocarpus are due to a bacterium which inhabits them, and that the 
endophytic fungus is of secondary importance. 
4. The Seasonal Relations of Mycorhizas 
Moller (14) is the only previous worker who has made any serious 
attempt to study the seasonal variations of mycorhizas. He worked 
with one and two year old seedlings of Picea. His results agree with 
mine so far as the time of mycorhiza formation and development is 
concerned. He seems, however, to have missed the fact that the 
mycorhizas are annual. To sum up the results of my observations, it 
may be said that new mycorhizas are formed throughout the summer 
months. They reach their fullest development in late autumn, and 
persist in this condition throughout the winter. In late spring they 
die. They are, therefore, annual. No doubt the character of the 
season may hasten or retard the death of old mycorhizas and the forma- 
tion of new ones, so that these processes may take place earlier some 
years than others. Since the formation of new mycorhizas usually 
begins before all of the old ones are dead, there is no time of year when 
no specimens at all can be found. The best time to collect for demon- 
stration purposes, or for class use, is in late autumn just before the soil 
freezes up, since they can be found most easily and abundantly at 
this time. 
