Reply to Swedland

Dear Sir,—Mr. Swedland's considerations as to the essence of science are interesting but too restrictive. His demand—"controlled experiment"—excludes even a large part of chemistry besides sciences such as meteorology, geology.

Lord Kelvin considered the essence to be measurement. All sciences as they become more complex become mathematical and metrical—this would seem due to innate limitations of the human mind. A third school considers scientists intuitively know what science is.

If science cannot be defined, certain aspects can be shown. Essential to a science is a theory to correlate many facts. Conclusions arrived at statistically without a theory or with unsubstantial theories are of little interest, even though often of utility. Much meteorology is still statistical with no unifying theory.

Discrimination of facts is a second desideratum. Coincare calls it selection and I understand "controlled" means discriminating. There is an infinity of facts but most facts are too special to be of interest. Implicit in the above is observation of the external environment.

The argument is the same as between formalist and intuitivist schools in other branches of knowledge.