184 



SCIENCE 



[N. S. Vol. XLYII. No. 1208 



witliout it experiments illustrate rather the 

 fallibility of human nature than the laws of 

 nature. Much time and effort must be given 

 to it. It was formerly a common notion that 

 in the laboratory the student should by him- 

 self rediscover the laws of nature, doing in a 

 few youthful work periods what during two 

 thousand years the greatest minds only gradu- 

 ally accomplished. Nobody thinks of that 

 now; the laws of machines, of motion, of heat 

 and electricity can only be illustrated, not dis- 

 covered, in the student laboratory. The be- 

 ginner has not only to make the illustrative ob- 

 servations, he must learn how to make them. 

 The technique of observation is as real as that 

 of the shop; just as in wood shop or forge or 

 machine room the ways of handling materials 

 and metals which experience has proved good 

 are not only taught but insisted on, so the 

 technique of the laboratory must be enforced; 

 the setting up of apparatus, the making of 

 measurements, the interpretation of the meas- 

 urements, all these must be done according to 

 good ways developed by experience. Hence 

 there must be' rather explicit directions for 

 work placed in the hands of the student, deal- 

 ing not merely with apparatus in general, but 

 with the apparatus of the laboratory where the 

 work is done. It is justly to be questioned 

 whether general laboratory manuals are a suc- 

 cess; were all laboratories equipped with the 

 same patterns of apparatus it might be pos- 

 sible to write a manual which would serve 

 them all; as the case is, different laboratories 

 have such different equipments, inherited and 

 constructed, that the detailed directions which 

 the beginner must have are unique in each in- 

 stitution. This does not mean, of course, that 

 manuals can not be written in book form 

 (though I hold the loose-leaf system better, as 

 more flexible), but that the book which fi.ts the 

 conditions of one place can hardly fit those of 

 another. 



'Nat only is the technique of observation the 

 result of a long historical development ; equally 

 so is the technique of the interpretation of ob- 

 servations when made. For the elementary 

 laboratory the experiment directions should be 

 laid out with this in mind, and good methods 



of reduction and a reasonable standard of ac- 

 curacy insisted on. Of course the degree of ac- 

 curacy can not be very high — usually a tenth 

 of a per cent, is very good for student work, 

 and sufficient, in well-planned experiments, for 

 the illustrations desired. Better than this is 

 hardly to be expected, as the apparatus for be- 

 ginners can not have many refinements, and 

 must also be fairly fool-proof, if the staff is not 

 to spend most of its time making repairs. For 

 this degree of accuracy and for the purposes of 

 illustration careful sliderule calculations and 

 careful use of plotting paper are excellent in 

 most cases; I have not found that the average 

 student is any more able to learn the accurate 

 use of these materials than he is of apparatus, 

 unless he is obliged to take the pains, and 

 taught how. 



Since the object of the laboratory work is 

 the illustration of the laws of nature, the in- 

 terpretation of the results should not be made 

 difficult by clouds of calculation, myriads of 

 curves or apparatus planned badly for the pur- 

 pose in hand. I have sometimes been obliged 

 to force students through laboratory directions 

 subject to these drawbacks, and the reading of 

 their reports has driven me to the belief that 

 simple experimental methods, direct and well- 

 planned computations and the use of curves 

 when curves are the shortest way to bring out 

 the facts, are what they need to be taught with 

 great pains. In the elementary laboratory busy 

 work has no place. Particularly, I am sure 

 that in some cases where students are forced to 

 compute results by the method of least squares 

 the labor is worse than wasted, as certain basic 

 assumptions of the application are not justified 

 in the conditions of the observations. The ex- 

 istence of such cases was pointed out in Nature 

 a good many years ago; Merriman has dealt 

 with their theory and computation; a glance 

 at his paper will convince any one that such 

 treatment has no place in the elementary lab- 

 oratory. 



When the student has got along in his 

 knowledge of laboratory technique he should be 

 set at problems which involve reading and ad- 

 vanced observation, but this should in general 

 be called research only if the undergraduate is 



