. \ 
III that Singapore is convenient for the passing scientists; 
XV that the herbarium should he near the university; and 
V that it would he better for the federated Malay States to XKM 
to have their own systematic botanist and herbarium* 
As regards( I) I believe the Colony has never been asked* 
]*!► * ■ * * > •*■ ’ ■ 
(II) aserne unimportant as the assistant Director In Singapore 
f 
I * r , 
can look after the Singapore Gardens under the general spper- 
% 
▼ision of the Director in the earns way as the Penang hardens 
are now managed. Moreover there is room for the development of 
the puhlio Gardens in Kuala Lumpor and far sure land near 
Kuala Lumpor than near Singapore for the extension of the 
experimental economic Gardena. The passing scientist/ is not 
r, 1 * A- 
worth eonsideartion; the real scientist, who a&ae matters, will 
not object te the short additional journey to Kuala Lumpor. I am 
not competent to express an opinion on (TV) but I have heard it 
suggested that Singapore ie not an Ideal site for a unverslty 
and I think It is perhaps open to argpment sdiether a university 
of Malaya in Kuala Lumpor would not be preferable to a Raffles 
College in Singapore. IT) seems to involve a totally unnoessary 
duplication of wottc. The forest and Agricultural Departments 
must always have small referenoe herbaria, but there is no need 
for two botanies! departments. 
IS. In oonoluslon X would make the following 
suggestions of a general nature:- 
taj dll research work should as far as possible be aooerding to 
programs, both In order to prevent overlap ana the stash 
sacrifice of the general good to research officers' individual 
fancies and in order to secure co-operative between the different 
departments. 
* a 
(b) In India there ie a beard of Scientific Advice* I do not 
know precisely what its functions are* but it might be worth 
while to make enquiries and to secure copies of recent records 
of their proceedings* The establishment of an advisory board might 
prove useful here* 
(o) the central research institute ummat should include a 
