II. In normal development, specific analytic observations are originally connected almost exclusively with the imperative need for noting means and ends in carrying on activities. When one is _doing_ something, one is compelled, if the work is to succeed (unless it is purely routine), to use eyes, ears, and sense of touch as guides to action. Without a constant and alert exercise of the senses, not even plays and games can go on; in any form of work, materials, obstacles, appliances, failures, and successes, must be intently watched. Sense-perception does not occur for its own sake or for purposes of training, but because it is an indispensable factor of success in doing what one is interested in doing. Although not designed for sense-training, this method effects sense-training in the most economical and thoroughgoing way. Various schemes have been designed by teachers for cultivating sharp and prompt observation of forms, as by writing words,--even in an unknown language,--making arrangements of figures and geometrical forms, and having pupils reproduce them after a momentary glance. Children often attain great skill in quick seeing and full reproducing of even complicated meaningless combinations. But such methods of training--however valuable as occasional games and diversions--compare very unfavorably with the training of eye and hand that comes as an incident of work with tools in wood or metals, or of gardening, cooking, or the care of animals. Training by isolated exercises leaves no deposit, leads nowhere; and even the technical skill acquired has little radiating power, or transferable value. Criticisms made upon the training of observation on the ground that many persons cannot correctly reproduce the forms and arrangement of the figures on the face of their watches misses the point because persons do not look at a watch to find out whether four o'clock is indicated by IIII or by IV, but to find out what time it is, and, if observation decides this matter, noting other details is irrelevant and a waste of time. In the training of observation the question of end and motive is all-important. [Sidenote: Scientific observations are linked to problems] [Sidenote: "Object-lessons" rarely supply problems]