812 
Journal of Agricultural Research 
Vol. XXI, No. II 
superficial observations lead to the necessary conclusion that culm for¬ 
mation in the interior of a grass plant, within the boundary of a more 
or less narrow peripheral ring, is sooner or later slowed down to a nearly 
zero rate. A few culms in the interior may be formed, but certainly 
anything like continued doubling is quite impossible. Culm increase is 
mainly always centrifugal in character. 
USE OF CONVENTIONAL PLANTS 
Relative merely to the rate of culm formation, four conventional plants 
were conceived with one, two, three, and four culms per plant, respectively. 
It was recognized that the rate of culm increase would presumably 
lessen as the total number of culms per plant increased. Even the mini¬ 
mum rate of increase of a i-culmed plant could not be maintained for 
any length of time, as was pointed out above. In dealing with the four 
conventional plants just mentioned, notice was taken only of the number 
of culms added at each culm accretion period. The time element, the 
time elapsing between successive accretion periods, could be modified 
at will. With the groups of plants actually in hand, the time element 
was a factor to be reckoned with. As the time between counting periods 
was lessened the number of culms added each period was lessened, and 
contrariwise. 
With the foregoing ideas in mind, the four conventional plants were 
assumed to double their entire culmage at definite intervals so long as 
the entire culmage was peripheral in character. In the more advanced 
states, when a differentiation became existent between the interior and 
peripheral areas, only the peripheral culms were conceived to produce 
sister culms. It is obvious that as a plant increases in size, the peri¬ 
pheral ring, assuming it to remain of constant thickness, increases in 
area relatively more slowly than the interior area of the plant. In 
terms of rate of culm formation, this would mean that with the increase in 
size of plant, coincident with its advancing age, the ratio of culm increase 
would constantly decrease as the total number of culms per plant 
increased. 
The earlier stages of the conventional plant initialing with four culms 
is presented. This one suffices, since the plants initialing at one and 
two culms pass through the 4-culm stage. The plant initialing with 
three culms has its stages somewhat modified. Indicating culms by 
dots we have the following scheme: 
