;74 



NATURE 



[January io, 191 8 



r 



THE LEARNING PROCESS IN A SNAIL.^ 



N his well-known experiments (1904), the Russian 

 physiologist Pavlov showed that salivary secre- 

 tion in a dog, prirnarily induced by the odour or sight 

 of food, could eventually be induced by a sound or 

 colour which had been for a time synchronised with 

 the primary stimulus. The dog, according to the 

 experiments, was soon able to establish an organic 

 association between the primary and the secondary 

 stimulus. When Pavlov slightly changed the secondary 

 stimulus there was a change in the dog's salivary 

 reaction, and this was taken as evidence of the 

 animal's power to discriminate between stimuli. 



With noteworthy clear-headedness. Miss Elizabeth 

 Lockwood Thompson has seen how to apply Pavlov's 

 method to a water-snail, Fhysa gyrina, which glides 

 about in ponds, with foot and mouth upwards, sus- 

 pended from the surface-film. When a part of the 

 body within a millimetre or two of the mouth is 

 touched with a bit of food, a chewing motion of the 

 mouth-parts is started. With the application of food 

 to near the mouth there was synchronously associated 

 a pressure with a clean glass rod at a fixed distance 

 from the mouth. The next step in the ingenious ex- 

 periment was to apply tue associated or auxiliary 

 stimulus alone in the absence of food, in order to 

 determine from the presence or absence of reactions 

 whether or not an association had been formed be- 

 tween the two sets of stimuli. Miss Thompson de- 

 serves to be congratulated, we think, on this extension 

 of Pavlov's method, which he himself did not regard 

 as applicable except to a limited number of mammals. 

 It is now possible, along this line of investigation, to 

 test a snail's power of "learning." 



When food was applied to the mouth and at once 

 withdrawn, response followed in 61 per cent, of the 

 tests, the mouth being opened and closed on an 

 average 3-93 times. I3y means of an apparatus a 

 simultaneous application of pressure to the foot and 

 food to the mouth was secured. In the first 60-110 

 trials of simultaneous stimuli no response followed; 

 in the remaining trials, out of 250 in all, a response 

 was always given. The snails were thus " trained." 

 After forty-eight hours a response followed the pressure 

 by itself, i.e. in the absence of any food-stimulus, 

 but only for a limited period. Cessation of re- 

 sponse to pressure after training is sudden and final. 

 The limit of the effect of training (which simulates 

 memory) is about ninety-six hours. An interesting 

 waning of response (marked by a reduction in the 

 number of mouth movements) was observed in, some 

 series of trials; it showed that the snails became 

 adapted to a stimulus which was not followed by its 

 wonted reward. The relation between length of train- 

 ing and training effect (as measured by response to 

 pressure only) requires further investigation. 



Miss Thompson also devoted many experiments to 

 inquiring whether the snail could learn to solve a 

 simple U-shaped or Y-shaped labyrinth with a picket 

 fence of wires, one arm leading from near the foot 

 of the tank to the air (the reward), the other not (the 

 punishment). In some cases error was punished bv 

 an electric shock, and roughness of the path was used 

 as a warning stimulus. The result was Interesting. 

 The snails showed no abilitv to learn that the one path 

 was to be preferred to the other. But in if. per cent, 

 of a total of Q30 trials in one series, the snails changed 

 their course from the wrong to the right path after 

 contact with a warning stimulus (in this case, slight 

 irritation of the tentacles and the back of the head 

 with a hair) before the shock (punishment) was re- 



1 " An An'ilysis of the Learning Proces'; in the Snail, P'lV-ia g-vrina, '^ny." 

 (Behav'oiir Monoaraphs, vol. iii., No. 3, 1917, pp. 1-89 + 8 plates + 12 tables.) 

 (Cambridge, Mas';.) 



NO. 



1515, VOL. 100] 



ceived. There was formed a weak association between 

 two stimuli, the hair and the shock, the former serv- 

 ing as a warning of the punishment to follow if the 

 course be not changed. But the capacity to form 

 associations, already proved by the method of using 

 simultaneous stimuli, does not suffice for the solution 

 of the simplest labyrinth. There was no evidence of 

 I "selective " ability. 



Miss Thompson has made a very interesting con- 

 tribution to the study of animal behaviour; the details 

 of the experiments show the punctilious carefulness of 

 her work. 



SCREW GAUGES. 



HP HE production of a satisfactory screw gauge 

 ■*■ is a matter of considerable difficulty as regards 

 both manufacture and testing, and the pamphlet 

 on this subject just issued by the National Physical 

 Laboratory ' will be found to contain much useful 

 information. 



In the case of a plug screw gauge, it is essential 

 that it should enter a standard check ring gauge, but 

 tnis test is insufficient, since it may be complied with 

 by a plug gauge having such a combination of errors 

 as to enable it to enter the check and yet be useless 

 for the purpose of gauging screws. "Not go" tests 

 are also essential, and certain .errors can be detected 

 only by carrying out measurements on the gauge of 

 either a mechanical or an optical character. The full 

 (or major) diameter is measured 'by use of a micro- 

 meter in conjunction with a set of Hoffmann roller 

 gauges. The core (or minor) diameter and the effec- 

 tive diameter are also measured by means of a micro- 

 meter, together with a pair of Vee-pieces and a pair 

 of small cylinders respectively. The lathe in which 

 the gauge is machined should be furnished with an 

 attachment for holding the micrometer so that its 

 axis intersects the axis of the gauge at right angles, 

 and arranged so that the instrument can be readily re- 

 moved. This permits of the gauge being measured as 

 the work proceeds, without the necessity for removing it 

 j from the machine In instruments used for measuring 

 these diameters in the inspection room, the micro- 

 meter should be held mechanically so as to comply 

 with the same condition. 



The machine described in the pamphlet for measur- 

 ing the pitch of the screw appears to be both simple 

 and effective. The actual measurement is made by 

 means of a micrometer having a large dial reading to 

 o-oooi in. An ingenious arrangement, partly mechan- 

 ical and partly optical, ensures that the axial move- 

 ment of the micrometer point shall be exactly equal 

 to the pitch of the screw under test. Both periodic 

 and progressive errors in the pitch can be detected 

 from the readings obtained in this machine. 



As a general rule, optical measurements of screw 

 gauges cannot be made to the same accuracy as 

 m'echanical measurements, but optical methods are of 

 great service from the consideration that the whole of 

 the screwed surface of a gauge can be examined in 

 detail. Errors in angle, want of straightness of the 

 threads, eccentricity between different diameters, and 

 local bumps and hollows can be detected readily by 

 optical means. Until recently, microscopes having- 

 cross wires in the eyepiece were alone emploved for 

 measurements, and a machine embodving this prin- 

 ciple is made by the Cambridge Scientific Instrument 

 Co., and is described in the pamphlet. Such methods 

 have now been displaced to a great extent by a pro- 



1 " Notes on Pcrew Ga>i<?es." By the Staff of the Gauee-testine Berart- 

 m^nt. National Physical Laboratory. Knlarged issue II. (Teddington : 

 W. E. Parrott, The Causeway, 1917.) Price 2^. 6ci. 



