26 BULLETIN 1450, U.-S. DEPARTMENT OF AGRICULTURE 
dependent plant. Only once during the course of these investigations 
has a sterile timothy shoot on a plant growing in a meadow been 
observed to become naturally bent over so that one or more of the in- 
novations at its tip had come into contact with the soil and produced 
roots. In some semiprocumbent European forms of Phleum pratense 
or allied species, however, the stems frequently become rooted at 
nodes in the culm (4-i, p. 155). It may be concluded, therefore, 
that while timothy plants do sometimes reproduce themselves by 
means of innovations growing at the tips of sterile shoots, yet this 
method of reproduction is not a common one in timothy plants of 
the types which commonly grow in meadows in America. 
The habit of vegetative reproduction at the tips of sterile shoots 
is not peculiar to timothy, though it has not been found in all 
perennial grasses which have been observed. In a plat of redtop 
(Agrostis alba L.) on the timothy-breeding field station in the fall 
of 1919, buds and innovations were growing in the axils of leaves 
near the tip on a considerably larger proportion of the sterile shoots 
than in the near-by plats of timothy. On the other hand, in plats of 
tall oat grass (Arrhenatheru?n elatius L.) and brome grass (Bromus 
inermis Leyss.) observed in November, 1919, although sterile shoots 
were numerous, on none of them was a bud or an innovation found 
in the axil of a leaf near the tip. 
RELATION OF THE TWO TYPES OF SHOOTS TO EACH OTHER 
No indication has been found that the buds from which fertile 
lateral shoots develop have a position on the primary shoots different 
from that of the buds which develop into sterile shoots. Whether 
any shoot may produce an inflorescence or not depends upon the cul- 
tural conditions under which the plant is growing, rather than upon 
the position of the bud from which the shoot has grown. 
One question which suggests itself in any study of fertile and 
sterile timothy shoots is whether either kind of shoot may originate 
from a bud at the base of a shoot of the other type. In order to get 
data which would answer this question, an experiment was conducted 
in 1916 in an old meadow which had not been mowed in 1915. 
On different dates from May 19 to 24, 1916, 21 partially elongated 
shoots which had developed from buds at the base of shoots which 
produced heads in 1915 were marked and their later development ob- 
served. The plants were growing in a comparatively dense stand, 
under conditions which were fairly typical of any ordinary timothy 
meadow. Thirteen of the twenty-one shoots which were marked 
continued their growth until July ; on each one of five, or 38.5 per cent, 
of these shoots a head was produced, and eight, or 61.5 per cent of 
the shoots were of the short, leafy type with no head. 
On 10 other plants in the same meadow small lateral shoots grow- 
ing from the base of shoots of the short, leafy type, which did not 
produce heads in 1915, were also marked at the same time for observa- 
tion. After these plants had been selected the surrounding ones were 
removed, loose soil was placed about the plants, and a small quantity 
of 2-12-0 grade of commercial fertilizer was applied about them on 
May 25, 1916. Conditions during the remainder of the season were 
therefore favorable for a vigorous growth. Eleven of the twelve 
shoots marked on these plants continued growth until July. On 10, 
