73

O.62

might be omitted); and in the next place, hardly a discovery is made that does not show us that natural conditions are far more complex and intricate than we had imagined them to be. But what astounded me most was to find that in notable cases what I have just called "superfluous embroideries" have turned out the most valuable parts of theories. Dalton's atomic theory may be said to be wholly made up of gratuitous material; since, although he aimed to account for simple multiple proportions, his theory neither necessitates nor renders more probable that