NATURE 



[May 4, 1905 



study of the tropisins, and we find him declaring that 

 it is almost impossible to describe the behaviour of 

 the unicellulars intellig-ibly without using terms like 

 ■"perception," "discrimination," and "intelligence." 

 Of course these are used in an " objective sense," and 

 " when their objective significance is kept in mind 

 there is no theoretical objection to them, and they 

 have the advantage that they bring out the identity 

 of the objective factors in the behaviour of animals 

 with the objective factors in the behaviour of man." 



From our point of view, Prof. Jennings does not 

 strengthen his position by using these pre-occupied 

 psychological terms; "'perception' of a stimulus," 

 he says, " means merely that the organism reacts to it 

 in some way; 'discrimination ' of two stimuli means 

 that the organism reacts differently to them; ' intelli- 

 gence ' is defined by the objective manifestations 

 mentioned in the te.xt. " But this does not seem to 

 us the sound line of progress ; it leads back to saying 

 that the lucifer match perceives the sandpaper on the 

 box. It seems safer, in the meantime, to say that 

 infusorians alter their behaviour, and alter it 

 €ffectively, in respect to their experience. 



" Stentor does not continue reacting strongly to a 

 stimulus that is not injurious, but after a time, when 

 such stimulus is repeated, it ceases to react, or' reacts 

 in some less pronounced wav than at first. To an 

 injurious stimulus, on the other hand, it does continue 

 to react, but not throughout in the same manner. 

 When such stimulus is repeated, Stentor tries various 

 difl'erent ways of reacting to it. If the result of re- 

 acting by bending to one side is not success, it tries 

 reversing the ciliary current, then contracting into 

 Its tube, then leaving its tube, &c. This is clearly 

 the method of trial and error passing into the method 

 of intelligence, but the intelligence lasts onlv very 

 short periods." 



With such difficult subjects any evidence of the 

 registration of experience was not to be expected, and 

 the author is to be congratulated on having discovered 

 considerable evidence in support of the thesis that 

 the behaviour of unicellulars is largely a method of 

 trial and error, one reaction by trial and error be- 

 coming the basis for a succeeding reaction. This is 

 surely a pathway leading to the high-road of intelli- 

 gence. 



It is easy to make an inanimate system— a little 

 potassium pill on a basin of water, or a tiny wound- 

 up engine on a smooth table— which, once set 

 a-going, will charge against an obstacle, will fail to 

 ■overcome this, will recoil passively and charge again, 

 and some observers have thought that, mutatis 

 mutandis, the animalcule did little more. But Prof. 

 Jennings has shown that the infusorian, in relation 

 to its experience of " error," changes its little tactics, 

 and changes them again, until it succeeds. In a 

 word, it profits by experience. The very essence of 

 vitality, as Spencer pointed out, is in effective re- 

 sponse to environment; but when we find an in- 

 fusorian " trying " one response after another, 

 abandoning those that spell "error," we cannot but 

 feel that vitality has been raised to a second power; 

 it is just beginning to be intelligent. The infusorian 

 is more than a tropic automaton, it is playing a 

 little game of tactics; perhaps if we could educate 

 NO. 1853, VOL. 72] 



one it would develop the rudiments of strategy. It 

 is, of course, extremely difficult to keep to a scrupu- 

 lous objective record of what occurs, but we incline 

 to think that Prof. Jennings has supplied what com- 

 parative psychologists have been waiting for, namely, 

 quite trustworthy accounts of the beginnings of 

 selective or controlled behaviour. 



" The method oT trial and error involves some way 

 of distinguishing error, and also, in some cases at 

 least, some method of distinguishing success. The 

 problem as to how this is done is the same for man 

 and for the infusorian. We are compelled to postu- 

 late throughout the series certain physiological states 

 to account for the negative reactions 'under error, and 

 the positive reactions under success. In man these 

 physiological states are those conditioning pain and 

 pleasure. The ' method of trial and error ' is 

 evidently the same as reaction by ' selection of over- 

 produced movements,' which plays so large a part in 

 the theories of Spencer and Bain, and especially in 

 the recent discussions of behaviour by J. Mark 

 Baldwin. The method of trial and error, which forms 

 the most essential feature of the behaviour of these 

 lower organisms, is in complete contrast with the 

 tropism schema, which has long been supposed to 

 express the essential characteristics of their be- 

 haviour." 



Instead of referring in detail to the author's 

 studies— (i) reactions to heat and cold in the ciliate 

 infusorians; (2) reactions to light in ciliates and 

 flagellates ; (3) reactions to stimuli in certain rotifers ; 

 (4) the theory of tropisms ; (5) physiological states as 

 determining factors in the behaviour of lower 

 organisms; and (6) the movements and reactions of 

 amoeba— we have sought to explain the chief result 

 of his studies in the infant school of life, and to 

 emphasise its importance in relation to the general 

 theory of animal behaviour. Prof. Jennings has 

 rescued the animalculae from the bonds of automatism 

 too hurriedly thrust on them, and has afforded a 

 secure basis for the study of the evolution of intelli- 

 g^^ce- J. A. T. 



MECHANISM. 

 Mechanism. By Pr'of. S. Dunkerley. Pp. vi + 4oS. 

 (London : Longmans, Green and Co., iqo:;.) Price 

 95. net. 



■YiyRITERS of text-books on mechanism have, of 

 • » late years, been much influenced by the views 

 of Realeaux on the classification of mechanisms, and 

 the present work shows clearly the impress of these 

 views ; but the author has not hesitated to depart from 

 the order in which Realeaux presented his theory of 

 machines in order to suit the needs of beginners, who 

 are apt to find the elaboration of the systematic theory 

 somewhat dry if not accompanied by a wealth of 

 illustration drawn from actual machines, even if these 

 contain elements the properties of which have not 

 been fully explained. 



The author, as appears from his preface, is fully 

 alive to the difficulties which the logical treatment of 

 the subject presents, and he expressly states that his 

 work is not intended to be a philosophical treatise on 

 the subject. 



From this standpoint the arrangement of the sub- 



