208 BOTANICAL GAZETTE [SEPTEMBER 
ries, herbaria and gardens, is a subject deserving of the most 
careful study, whether viewed from the standpoint of the 
endower or administrator of an institution of education or 
research, or from that of the botanist whose reputation is built 
up in the performance of the duties assigned to him in such an 
institution. 
While there is every reason to expect large returns from the 
endowment of such independent departments of research, free- 
dom from the duties of the class room, while leaving more time 
available for investigation, will not prove an unmixed blessing. 
I believe it to be the experience of the best investigators in this 
country that research is promoted by the necessity of imparting 
some or all of its results in the class room. In no other way, 
after specializing in the small field to which it seems necessary 
for most of us to confine ourselves, can one make sure of pre- 
serving the breadth of view needed for the investigation of sie 
a limited specialty in the most successful manner. It must be 
admitted further that the power of application and concentra- 
tion varies with different men, so that up to a certain point the 
interruptions introduced by limited teaching or looking ae 
collections in many cases may give fresh zest to the pursuit of 
knowledge in the time remaining for research. And it may be 
that at this very point lies the greatest difficulty to be — | 
surmounted in the development and management of reseae’ — 
institutions. 
While there is no doubt that some supervision and p ie 
are conducive to the performance of the greatest aie 
amount of investigation, as of other work, since they 1m 
sistent planning and close application, it cannot be overlooxe' : 
that this is the extent to which scientific work can profitably PF 
crowded. To require more of an investigator than that he si 
be reasonably busy with thoughtfully planned study 
always been antagonistic to the performance of his 
emanate from each department at stated intervals, 
7, expe 
insures quantity in publication, generally does so at the — 
ressute 
sure con : 
is and “s . 
best worki 
etin shall 
and the requirement of some institutions that a bull Co 
while " 
i a 
