Feb., 1923] SEIFRIZ — CAUSES OF GREGARIOUS FLOWERING IO5 
Dendrocalamus giganteus to give up flowering and continue its purely 
vegetative growth. Similar observations must have been made by Brandis, 
since he writes that “offsets taken from a clump some time before it flowers 
come into flower at the same time as the parent clump” (1, p. 662). 
The hypothesis of depletion of nourishment as the cause of flowering in 
bamboos could never be applied to those bamboos which flower gregariously. 
It is quite untenable that each individual of the multitude of plants in a 
forest of Dendrocalamus strictus, one thousand square miles in area in India, 
or of Chusquea abietifolia extending over a region ten miles in length in the 
mountains of Jamaica, should simultaneously exhaust the supply of food in 
the soil where they are growing. 
Injury as a Cause of the Flowering of Bamboos 
Several interesting cases have been reported which support the theory 
that injury may cause anthesis in bamboos. While injury has no bearing 
on our problem of the natural cause of gregarious flowering in plants, yet it 
is worthy of consideration, since it is a probable stimulus which apparently 
arouses some bamboos to sexual activity. 
Gamble states that single clumps of Bambusa Tulda, “if badly treated 
by over cutting or partly uprooted, will afterwards produce flowers without 
any general flowering” (5, p. 31). 
Another instance of the flowering of bamboos being caused by injury is 
reported by Merrill from the Philippines. In an extensive bamboo forest 
of Schizostychium one single culm was seen in flower. This culm had been 
cut by a bolo (machete). The culm was about two thirds severed and in 
full flower. 
Knowledge of these two instances reported from India and the Philip¬ 
pines caused me immediately to suspect that two injured clumps of Bambusa 
arundinacea which I noticed in flower in the Buitenzorg Gardens had also 
flowered because of the injury received. In each clump several culms were 
in profuse flower, and these culms were broken off about midway of their 
length, while all those culms which were not in flower were healthy, un¬ 
injured shoots. It seemed possible, therefore, that the broken culms had 
flowered as a consequence of injury. On second thought it was evident 
that there was no way of determining without previous data whether the 
culms had flowered as a result of injury or whether they had broken as a 
result of flowering. The culms of Bambusa arundinacea die after flowering. 
A dead culm is much less resistant than is a live one to strain from wind, 
which may be very great on a culm forty to sixty feet in height. In order 
to ascertain which event, the flowering or the breaking, had preceded the 
other, I had several culms cut in a large and healthy clump of Bambusa 
arundinacea. These culms when observed one year after cutting had not 
flowered. The injured culms above mentioned had in all probability broken 
as a result of flowering and dying and consequent weakening of the culms. 
