SECTIONAL TRANSACTIONS.— J. 399 



Dr. R. H. Thouless. — Dr. Houstoun's Substitute for Weber's Law. 



It lias long been known that Weber's Law does not adequately represent the facts 

 of variation of differential threshold (AI) with stimulus (I). If AI/I is plotted against 

 I, an asymetrical curve is obtained with the two ends pointing steeply upwards. If 

 Weber's Law were true, this should be a straight line parallel with the base. Dr. 

 Houstoun has found, however, that for light sensation, if AI/I is plotted against log. 

 I, a curve is given which is approximate!}' the normal curve of error. The approxima- 

 tion is, in fact, remarkably close. If this proves to be true of all stimuli, Weber's Law 

 that AI/I is a constant must be replaced by the law 



_ (log.I-log.l 'r.i 

 I/AI=ifc.e '.va 



in which I' is the value of I at which the curve attains its maximum and a is the 

 standard deviation. 



Dr. Houstoun also proposes to replace Feehner's ' fundamental formula ' 

 (rfS = k.dljl) bv the formula 



,„ _ (log I -log 11)2 



'^^ he i^^- 



d(logl) 

 which by integration gives 



rx _ Oog I-log li)'^ 



J — QO 



in place of Fechner'a Law S = K (log . I/Io). 



Afternoon. 

 Mr. G. C. Grindley. — Psychological Factors in Peripheral Vision. 



Mr. C. A. Mace. — The Psycho-physics of Desire. 



Despite the prominence of the concept of 'purpose' in contemporary Psychology, 

 little has as yet been achieved in the way of controlling human activity by scien- 

 tific means even where this activity is most obviously purposive in nature. This 

 div^orce between general theory and applied science is due in the main to our ignorance 

 concerning the way in which general ' good intentions ' become specific to particular 

 operations, and concerning the precise influence of conditions which prima facie are 

 relevant to this process of specification. Here, then, is an enticing but neglected field 

 for experimental investigation. The chief problems are brought into view by investi- 

 gation of the incentive conditions operative in ordinary industrial tasks, and by the 

 experimental determination of the influence of similar conditions in the performance of 

 tasks commonly employed in laboratory tests of psycho-physical traits. 



An ' incentive condition ' may be defined as any factor which modifies the content 

 or direction, the intensity, the quality, the duration or the date of occurrence of a 

 specific intention controlling the performance of a task, and it is a reasonable postulate 

 that such characteristics of an intention will be reflected in, and therefore measurable 

 by reference to, corresponding characteristics of the performance expressive of the 

 intention. 



Preliminary experiments suggest that principles such as the following are involved : 



1. That under certain commonly realised conditions the prescription of a standard 

 operates as an incentive condition modifjdng each of the characteristic phases of an 

 intention, but that difierent types of standards influence the various phases in different 

 ways. Thus, the prescription of a fixed standard will intensify effort at one time and 

 for a certain period, whereas a 'sliding scale standard ' will intensify effort at another 

 time and for a different period. In the special investigations here referred to, the 

 fixed standard proved on the whole the more potent incentive. 



2. That the higher the prescribed standard the more potent it is as an incentive 

 condition so long as the task continues to be performed, but the higher standards 

 modify the ' quality ' of the intention in ways that weaken the tendency of the 

 intention to recur. 



3. That the increase of facility due to transitory incentive conditions is permanent, 

 and not limited to the period for which such conditions are directly operative. 



