302 SECTIONAL TRANSACTIONS.— A. 



Dr. R. A. HousTOUN. 



The speaker was certain that sensation could be measured. The results, 

 though not very accurate, were much better than nothing, and would 

 eventually throw light on the mechanism of sensation. In classifying the 

 visible stars into six magnitudes the measurement of the sensation of intensity 

 had been used for centuries, and in different parts of the world, with con- 

 sistent and satisfactory results . He thought that those who took the contrary 

 view defined measurement too narrowly and in such a way as would exclude 

 many physical quantities. But it was undesirable to strain too much after 

 definitions. Simple experiments made on a photometer bench with a wedge 

 photometer and two lamps, on building up a scale of the sensation of 

 intensity and halving intervals on such a scale, would bring more conviction, 

 and it was desirable that such measurements should be made as widely as 

 possible. 



In conclusion he alluded to his own investigations, which showed that the 

 sensation of intensity was much better expressed as a function of the loga- 

 rithm of the stimulus by a probability integral than by the conventional 

 logarithmic function. The integral approximated to the latter at the middle 

 point of the range. 



Mr. R. J. Bartlett. 



As originally stated Weber's law records the fact that, within limits, the 

 change that must be made in the objective physical basis of a stimulus in 

 order that the change may be just noticeable is proportional to the amount 

 present initially. All measurements made are physical measurements. 

 Sensation is not measured, but the amount of change in the objective basis 

 is governed by subjective judgments based in sense experience. The judg- 

 ments rest on awareness of sameness or difference. The change necessary 

 for detection of difference fluctuates in amount, but the various readings 

 obtained tend to concentrate about a mean value. The necessary change 

 can be expressed as a ' constant error ' and a scatter value about that 

 ' error.' 



In work reported more fully to Section J it is shown that : 



(i) The ' constant error ' is a regression towards a central, accus- 

 tomed value of stimulus. 



(2) Weber's law holds for a limited central zone to which we are 

 adapted. 



(3) Beyond this zone, in either direction, deviation from the law 

 increases rapidly, and when a geometric series of units is used, 

 apparently, the deviation is proportional to the cube of the 

 ' distance ' of the objective basis of stimulus from the central 

 datum value of (i) above. 



Dr. S. G. Barker and Mr. C. G. Winson. — The psychological basis of 

 wool-sorting. 



When wool arrives in the raw state for processing purposes, the first 

 stage is to place it in the hands of the wool-sorter, who disintegrates the 

 fleece into its constituent qualities. Both the tactile and visual senses are 

 employed, and with almost uncanny precision the experienced sorter classi- 

 fies the wool according to its subsequent manufacturing performance, 

 or in particular to the degree of fineness to which the ensuing yarn may be 

 spun. An examination of the methods employed reveals that whilst fibre 



