268 PART V. WORKING STAGE. 



Quantitative accuracy should be naturally proportionate to the 

 needs of the investigation. Calculable relations, of the kind 

 enumerated in our table of Primary Categories, need special 

 attention. 



130. (K) Instruments. Observation should be, wherever 

 possible, instrumental. Dynamometers, ergographs, telescopes, 

 spectroscopes, transparent and graded glass vessels, scales, 

 diagrams, etc., and mathematical methods and formula? should 

 be employed. Instruments should be adapted, or new instru- 

 ments invented, to suit novel requirements. The unassisted 

 senses have wrested little from nature: they are altogether 

 too gross and clumsy for this purpose. The acquired capacity 

 of devising fresh and effective instruments constitutes, in some 

 sciences, an integral portion of the outfit of the man of science, 

 though enterprising firms of instrument makers materially 

 second his efforts. While the naked eye can detect only about 

 3000 stars, instruments acquaint us with a 100,000,000, and 

 while the reason wonders what the ocean depths harbour, the 

 deep-sea dredge lays their marvels at our feet. 



131. (/) Experiment. Observation should, of course, as- 

 sume the form of experiment when circumstances are propitious. 

 A scrupulously arranged and conducted experiment, for instance 

 as to the solubility of food-stuffs with and without the ad- 

 mixture of certain glandular juices, singles out constituents and 

 factors with the greatest assurance, but only when Conclusions 5 

 and 20 are complied with. Consider the problem of the pro- 

 tective value of colouring: "An Italian naturalist, Cesnola, 

 tethered twenty green mantis among green herbage and twenty 

 brown mantis among withered herbage; they were all alive 

 seventeen days afterwards. He then tethered brown mantis in 

 a green environment, and green in brown grass, and found 

 that thirty-five out of forty-five were devoured within seventeen 

 days. Professor Poulten . . . fastened 600 pupa? on leaves, 

 fences, etc., and found that the mortality of the more con- 

 spicuous was ninety-two per cent. Professor Davenport found 

 that, of 300 chickens in a field, twenty-four were quickly killed 

 by crows, and that only one of the twenty-four was of the 

 less conspicuous spotted variety." (Joseph McCabe, The Prin- 

 ciples of Evolution, 1913, p. 117.) Or examine a very simple 

 problem. Walking in a certain direction at the rate of 3 J /2 miles 

 per hour, I experience no wind; returning at the same rate 

 in the opposite direction, I calculate the velocity of the wind 

 to be apparently 7 miles per hour. Standing still, however, I 

 simplify the conditions to the utmost, and am enabled to decide 

 whether there is a wind blowing and, if so, what is its direc- 

 tion and velocity. 



Experiments are sadly needed to solve some of the problems 

 of heredity and instinct. Thus, as already adverted to, various 

 specimens of each of the domestic and of some other animals 



