DEVELOPMENT AND LIFE OF THE CORM 29 
Growth during the third and fourth seasons is not marked by 
any peculiarity. The corms increase in size because of additional 
storage of starch, and new lateral buds are produced each season 
(PLATE 3, FIG. 49). Attheend of the fourth season of growth the 
largest corms showspikesof staminate flowers. Such cormsmeasure 
15 mm. or more in thickness. Of the plants grown from seed by 
the writer only about 10 per cent. produced flowers the fourth 
year, the remainder failing to produce flowers before the fifth or 
sixth year. 
After the first appearance of the flowers the growth and 
activity of the corm is quite regular, there being new lateral buds 
formed (PLATE 3, FIG. 49) as before and new food material stored up 
each growing season. The increase in size is not uniform from 
year to year, for, while constantly increasing amounts of food are 
removed from the corm each year for use in producing leaves and 
flowers, the amount of new storage is always dependent upon the 
length of the season of growth. So after several consecutive short 
growing seasons some old corms may be greatly reduced in size 
as a result of the drain to produce early growth and the failure to 
replace the food so used. The mass of starch is not divided into 
sections as it is stored up, but the dividing layer of phelloderm is 
formed each year, cutting off the portion to be used at that time. 
In the case of several consecutive poor growing seasons the avail- 
able amount of food becomes so reduced that growth of leaves 
and flowers is curtailed, and the production of flowers may even 
be entirely suspended. In the majority of cases examined in the 
field, the appearance of buds and root scars seems to indicate that 
the oldest portion of the corm is four years old or thereabout. 
Another point worthy of note is that a small number of the 
corms, 3-5 per cent., lie dormant during whole growing seasons. 
This is true with corms of all ages from one year up. As yet no 
reason for this phenomenon has been suggested. Neither is there 
apparent any regularity to indicate a cyclic occurrence of resting 
periods. It has been noted, however, that in a very few cases the 
resting period covers only a part of the season, and, consequently, 
the plants appear in late summer. This is probably related to the 
phenomenon of double seasonal growth mentioned above. 
The formation of lateral buds has already been mentioned. 
