202 REPORTS ON THE STATE OF SCIENCE.—1917. 
as each experiment proceeds: on uo account must this be done at any other 
time. A reasoned account of the work should then be written out at leisure, 
in flowing language, with due regard to style, never in the inexcusable form of 
a statement in advance of the conclusion to be arrived at ultimately, nor in 
the graceless hackneyed form of Experiment, Observation, Inference. It should 
never be forgotten that the prime object in view is to develop ‘habits of 
logical thought and logical statement, together with the habit of inquiry. 
The clearest possible distinction must be drawn, therefore, between an experi- 
mental, reasoned inquiry into an undetermined issue and the practical demonstra- 
tion or verification of a stated fact. It must be made clear that an experiment 
is an act performed with the definite object either of finding out something 
novel in the experience of the worker or of testing an assumption—that the mere 
demonstration or verification of the truth of a statement is not an experiment. 
The accounts should be fully illustrated by drawings and photographs. 
In order to teach the use of books and develop the habit of purposed, serious 
reading, as wide a course as possible of reading should be associated with the 
experimental work. The books used should be mainly of general interest, and 
informative—books of reference, books of travel, &c.—though technical books 
may be consulted occasionally with advantage. 
— 
