QUANTITATIVE ESTIMATES OF SENSORY EVENTS 329 



with a single kind of stimulus) using equality of just noticeable differences 

 as a convention of measurement parallel to the use of equality of tempera- 

 ture differences causing equal volume changes of thermometry. Anyway 

 it is obvious that the restrictions are such that measurement in this sense 

 would be of little practical scientific value. 



(4) I am not convinced by the argument of pp. 317 ff. Surely apparent 

 shapes change with different angles of vision as do apparent sizes at different 

 distances. 



(5) I think Mr. Guild would agree that most of the quantitative experi- 

 ments psychologists do when they attempt to ' measure sensation ' could 

 be done equally well and equally meaningfully if they gave up that assump- 

 tion. He does say this, but I should like to see it emphasised. It con- 

 siderably narrows the apparent difference between his view and the oppos- 

 ing. For example : ' Thus the prevalent idea . . .' (p. 322) seems to me to 

 exaggerate this difference. I should like to see added that if Mr. Guild's 

 view is accepted and the mean-gradation experiment is merely a report of 

 sameness of relations, it is nevertheless a real question whether if one arranged, 

 let us say, a series of electric lamps so that each successive pair had the same 

 brightness relation between them, whether this series would be a geometrical 

 or arithmetical series of physical intensities. The phrase ' is devoid of any 

 basis ' seems to suggest that there is no real problem, whereas I think 

 Mr. Guild would agree that there is a real problem but that it should be 

 stated in other terms. 23.6.38 



B. By Dr. L. F. Richardson. 



Mr. Guild has made a logical analysis of the relation between sensation 

 and stimulus. Those whose chief reliance is on logic, take risks of passing 

 over assumptions without noticing them. Mr. Guild's analysis may be 

 summarised in three steps : — 



Step I. Is sensation an ^-magnitude ? No ! 



Step II. Is sensation a B-magnitude ? No ! 



Step III. Therefore sensation is not a magnitude of any sort. 



The tacit assumption is that the A and B magnitudes are the only kinds 

 of magnitudes that can exist. But there is abundant experimental evidence 

 {vide Dr. Semeonoff's report) from several independent investigators in 

 England and America that intuited magnitudes exist. Of course ^-magni- 

 tudes are usually the most reliable ; and A and B magnitudes, taken together, 

 are the only sorts of magnitude which are respectable in practical physics, 

 except for the estimation of tenths of small divisions. It is also evident 

 that intuited magnitudes are subject to variations with the occasion and with 

 the observer ; variations so large that they would not be tolerated in practical 

 physics. But if, as Mr. Guild avers, A and B magnitudes are not available 

 for sensation, then intuited magnitudes are not to be despised. The 

 progress of psychology towards the status of a quantitative science is more 

 likely to be advanced by experimental exploration of the relations of intuited 

 magnitudes than by refusal to allow them to be considered. 



Some people wish to see the word ' measurement ' restricted to mean the 

 determination of A and B magnitudes only. If that were done it would 

 be necessary to point out that a mere terminological convention must not 

 be allowed to prejudice discussions about the existence of magnitudes other 

 than those called A and B. 



There is a remark, independent of the foregoing considerations, to be 

 made about Mr. Guild's Step I. He avers (on p. 310) that the method of 



M 2 



