404 DR. OTTO STAPF ON THE 
by Mr. A. E. Wild, late Conservator of Forests, Bengal, in the autumn of 1901 *, and it 
was on them that this paper was based in its original form as read before the Society 
on March 20, 1902. Before going to press, however, I received through Sir Dietrich 
Brandis another batch of Melocanna fruits from the same source, stated to have been 
collected in the Chittagong District, and later on a third set, communicated by Mr. N. K. 
Mukherji, Extra-Assistant Conservator of Assam, and consisting of unripe fruits and 
flowers gathered in the Garo Hills. As the fruits with which I had been working 
originally were almost without exception perfectly mature or even germinating, I was 
frequently obliged to form my ideas concerning the morphologie value of parts of the fruit 
and the general development by inference from the mature state. Hence more than one 
point necessarily remained doubtful. "The new material, particularly that from the Garo 
Hills, offered an opportunity for extending my researches to the earlier stages of develop- 
ment and revising my conclusions. I thought it only right to avail myself of it at the 
risk of considerably delaying the publication, as other work had meanwhile claimed my 
time. "These efforts of broadening and deepening the basis of my investigations resulted. 
in the confirmation—in the essential points— of the conclusions laid down in the short 
abstract of my paper contained in the Report of the Society's meeting of March 20, 
1902 t, in the explanation of certain particulars in the structure of the Melocanna fruits 
which were unknown or imperfectly understood before, and in a more complete account 
of their development based on the immediate observation of the successive stages. There 
also was some uncertainty as to the specific identity of the plants figured by Roxburgh 
in his * Plantee Coromandelian:e,' tab. 243, as Bambusa baccifera, and by Gamble in his 
“ Indian Bambuseze ” (in Ann. Bot. Gard. Cale. vii. tab. 105) as Melocanna bambusoides, 
and aecordingly also with respect to the correct determination of the fruits I had to deal 
with. I discussed therefore this question in full, so that it might be studied in the field. 
The new material, however, dispelled my doubts, which, as it now appears, had been 
caused by the very remarkable heteromorphy of the inflorescences of Melocanna bam- 
busoides, a condition hitherto unknown. Hence I cancelled the paragraph dealing with 
the suspected identity of the plant depicted by Roxburgh and Gamble respectively, 
reserving a full description of the inflorescences, as being beyond the limits of this paper, 
for another occasion. 
My original intention was to include the fruits of JMelocalamus and Ochlandra in the 
scope of my investigations; but considering the scantiness of the material pertaining 
to those genera to which I then had access, I preferred to confine myself to Melocanna. 
I hope, however, before long to have an opportunity of publishing an account of the 
structure of the fruits of Melocalamus and Ochlandra, of which I have since received 
good material through the courtesy of Sir Dietrich Brandis and Mr. T. F. Bourdillon, 
Conservator of Forests, Travancore. 
lam also under great obligation to Mr. J. S. Gamble for the loan of his herbarium 
* Kew had received a few equally good specimens from the same source in 1900, but as they were reserved for 
exhibition purposes in the Museum, I did not include them in my investigations beyond an examination of the 
external features. 
t See * Nature,’ vol. lxv. p. 548. 
