5°4 



NA TURE 



[April 2, 1885 



retinal receiving area. Il is however, a noteworthy circum- 

 stance that these natives are able to pass from the bright 

 tropical glare outside their dwellings to the dark interiors, and 

 vicevei sd, without showing the temporary derangement of vision 

 which the white man experiences whilst the iris is adapting itself 

 to the new condition. II. B. GurrY 



17, Wood Lane, Falmouth, March 30 



Mr. Lowne on the Morphology of Insects' Eyes 



In reference to the discussion between Dr. Sydney Hickson 

 and Mr. Benjamin Lowne, I beg to state that I have been 

 favoured by both of those gentlemen with opportunities of care- 

 fully studying their preparations, and I feel it to be my duty to 

 state that in my judgment Mr. Lowne's preparations do not 

 justify the conclusions which he has based on them, and are, in 

 fact, not made with that skill and knowledge of modern histo- 

 logical method which is necessary in order that trustworthy con- 

 clusions maybe obtained. On the other hand, Dr. Hiclcson's 

 preparations are thoroughly satisfactory as examples of histo- 

 logical manipulation. Dr. Hickson supports the accepted view 

 as to the termination of the optic nerve-fibres in the nervc-cml 

 cells of the retinulae. Mr. Lowne denies this connection. I 

 have no doubt that such a connection cannot be readily observed 

 in Mr. Lowne's preparations. At the same time I have no 

 doubt whatever that this is because the preparations are badly 

 made. Mr. Lowne's preparations fail to show many other 

 simple features in the structure of the insect's eye, which are 

 readily seen in preparations made by the application of methods 

 now recognised and approved, but not made use of by Mr. 

 Lowne. 



I am sorry to see the resources of the Linnean Society 

 employed in publishing a memoir the conclusions of which, 

 although startling in their novelty, are undeniably based upon 

 the mistaken interpretation of defective preparations. 



I think it is important that the Fellows of the Linnean 

 Society should know whether the memoir now published is the 

 same which was read a year or two ago at the Royal Society, 

 and whether the Council of the Royal Society took any step- to 

 ascertain the value of Mr. Lowne's preparations, or came to any 

 decision as to the fitness of Mr. Lowne's paper for publication. 



March 14 E. Ray Lankester 



On the Terminology of the Mathematical Theory of 

 Elasticity 



Engineers quite as much as " elasticians " have reason to 

 want some such terminology as that sought by Prof. Pearson 

 (Nature, vol. xxxi. p. 456), and have equal reason to be indebted 

 to him for undertaking the work which he has at present in 

 hand, which seems already to have given results of practical 

 value as great as their scientific interest. 



As I have for some years made a study of the physical side of 

 the problems mentioned by him, I should be glad to make some 

 suggestions as to terminology as contributions to the discussion 

 of the subject in your columns. I will confine what I have to 

 say to what may be called ductile materials (such as wrought 

 iron, ordinary steel, copper, &c ), because in these only the 

 whole phenomena are visible. The behaviour of such material 

 in tension is illustrated by the accompanying figure, in which 

 stresses are measured along the horizontal, and strains along the 

 vertical axis. 



It is extremely rare to obtain a piece of raw material already 

 in a state of ease. Wire, of course, is highly strained by its pro- 

 cess of manufacture, but that even ordinary bar and plate is also 

 slightly strained, is shown in the manner mentioned by Prof. 

 Pearson. Such initial strains as become visible as set by the first 

 stretching up to any load (within limit of elasticity) disappear 

 after one or two applications of that load. The material is then 

 in a state of ease up to that load, but higher loads (still within 

 the limit), on their first application, generally produce more set — 

 the state of ease thus extending only to the stress employed to 

 produce it. The sets are, along with the elastic strain, propor- 

 tional to the stress, their effect being simply to lower the 

 modulus of elasticity. Probably the process of annealing will 

 bring the material into a state of ease for all loads at which such 

 te is possible. I propose to examine this matter further by 

 aid, if possible, of the apparatus described by Prof. D. E. 

 Hughes in the Inst. M. Eng. Proc, 1883, p. 73. In the figure, 



a represents this condition ofpeifect elasticity (maximum state o 

 ease being presupposed) and n, the superior limit of this con- 

 dition, is the mathematical limit of perfect elasticity. 



After ii comes a stage i, within which the set is not propor- 

 tional to the stress, although it still remains small ; the total ex- 

 tension, therefore, increases faster than the stress. Occasionally 

 this stage does not occur at all, and both its higher and lower 

 limits seem — more than any other points in the life of the 

 material — to be susceptible of change depending on manipula- 

 tion. Accidental shock will shorten the stage considerably ; very 

 gradual loading extends it somewhat. For these and other 

 reasons I therefore suggest that this stage be called the condition 

 of instability, or of unstable equilibrium. 



This condition terminates at c, in what I have called a 

 "breaking-down" in the paper referred to by Prof. Pea- son, in 

 which paper I believe the phenomenon was described for the 

 fir-t time. This point is the one called by engineers the limit 

 of elasticity, because it is the only one markedly visible without 

 special apparatus. (The extension at B, on a length of 10 

 inches, may be about o'Oi inch ; at ccro3inch and at Cj, same 

 stress, it increases to o"20, o'25, and even occasionally 04 inch.) 

 If "breaking-down point" be too crude a name, I would sug- 

 gesl limit of. tail iff. It should be noted that the stress at this 



Loads) 



point does not remain constant, but in reality appears to diminish 

 as the extension goes on, as shown at c' (this dotted curve not 

 drawn to scale), a matter on which I am at present experiment- 

 ing. I should add that, during the application of load at this 

 point, extension appears to be occurring at different parts of the 

 length successively, and not at all parts simultaneously, as during 

 conditions a and c. 



In the next stage, c to D, the whole strains consist of a very 

 small ela-tic portion (apparently closely following the modulus), 

 and a very large set, increasing much faster than the stress. 

 The test bar remains at each load practically constant in its 

 cross-section at all points of its length, and rises in temperature 

 in-tead of (as in condition a) cooling. I would suggest for this 

 stage the name condition of uniform-flow, the physical applica- 

 bility of which will be obvious to any one who has seen ductile 

 metal in this condition. 



At some point, D, a maximum load is reached, and at about 

 the same point (generally, I think, a little earlier, but the differ- 

 ence is small, and not very easy to get at with certainty) the 

 metal begins to flow locally, apart becoming much more reduced 

 in cross-section than the rest, and eventually frpcture occurs at 

 this place under a less load than D, but with a greater extension, 

 as at E. This final stage, d, might be called condition of local 

 flow. The loads D and E (as Prof. Pearson suggests) would be 

 nuix imuni ami terminal loads respectively). (Their difference was 

 first pointed out, I think, by Mr. Daniel Adamson's experiments, 

 Journal I. and S. Inst., 1878). The maximum intensity of stress 



