38o SECTIONAL TRANSACTIONS.— J. 



differences * is a necessary and sufficient condition for the existence of a 

 single general factor, three theoretical difficulties arise. The first is the 

 difficulty of the sampling problem, but this is not discussed in the present 

 paper. The second difficulty is that the condition that the ' tetrad differ- 

 ences ' should vanish is not sufficient to determine ' g ' uniquely. Some 

 experimental work is given to show the effect this would be likely to have in 

 practice. The third difficulty is that if new test scores are formed by linear 

 combinations of the old, even if the ' tetrad differences ' vanish for the new 

 scores, the latter will not necessarily lead us to the same value of ' g ' as the 

 old. Those linear transformations for which ' g ' is invariant should have 

 a particular psychological importance. 



Dr. S. J. F. Philpott. — An approximation to a theoretical curve of output. 



It has been suggested by the present writer (' Fluctuations in Human 

 Output,' B.y.P. Man. Supp., 1932) that the curve of output can be expressed 

 as the sum of a large series of waves. Necessary conditions seem to be that 

 the periods concerned shall be whole number multiples of a common unit, 

 and that the waves converge on a common trough somewhere near the 

 origin of the curve. 



In arriving at an approximate outline for such a curve, waves were assumed 

 of periods 2 to 360 times the conamon unit, inclusive. Amplitude 

 measures were, however, not calculated in the ordinary way. With so large 

 a number of waves, it seemed sufficient to assume that the mean amplitude 

 over any given short length of curve should be inversely proportional to 

 the number of waves therein coming to the centres of troughs. 



Comparisons between the curve so obtained and actual output curves 

 show that there are definite points of resemblance. 



Sunday, September 4. 



Visit to the Retreat Mental Hospital. 



Monday, September 5. 



Presidential Address by Prof. B. Edgell on Current constructive theories 

 in Psychology. (See p. 169.) 



Prof. W. McDouGALL, F.R.S. — A third report on a Lamarckian experiment. 



Since my second report, made in 1929, the stock of rats trained to a 

 specific task has shown further increase of facility, the thirtieth generation 

 making in the curve of training twenty errors per rat, where the ancestral 

 stock made 148 errors. This report is mainly concerned with experiments 

 designed to test two questions : {a) Can this increased facility be due to 

 ' social ' transmission ? {b) Can it be due to favourable selection ? The 

 experiments of the former group are of three kinds: (i) omitting training 

 of one or more generations ; (2) rearing under foster-mothers ; (3) cross- 

 breeding females of untrained stock with males of trained stock. The 

 latter experiment consists in practising strongly adverse selection on eleven 

 generations during training. Facility continues to increase under training 

 in spite of adverse selection. 



Dr. T. G. Maitland, — Disorientation and vertigo. 



