July 5, 1888] 



NATURE 



225 



effect, and that it was only when the current was dimin- 

 ished by increasing the resistance of the battery that the 

 elongation of the iron became well developed. This view 

 clearly involved the assumption that the common notions 

 as to magnetic saturation must be at least in part erro- 

 neous, and I therefore endeavoured to find some other 

 explanation of the apparent anomaly. In particular, I 

 suspected that it might be due to electro-magnetic action 

 of the kind known as " solenoidal suction " between the 

 iron rod and the coil ; but a few careful experiments con- 

 vinced me that, although this might well have been the 

 case, yet in fact it was not so. Nor did any other hypo- 

 thesis present itself which would bear examination, and I 

 accordingly fell back upon the first and natural interpreta- 

 tion of the facts, which implies that magnetizing force may 

 exert an important molecular influence upon iron even 

 when its magnetism is saturated. 



A fuller investigation of the phenomenon was then made 

 with very delicate apparatus and greater battery power, 

 and the results were communicated during the next year 

 to the Royal Society, the principal conclusion arrived at, 

 so far as regards iron, being the following : When an iron 

 rod is subjected to a continually increasing magnetizing 

 force, its length at first increases to a maximum and then 

 diminishes, ultimately becoming actually less than when 

 the rod is unmagnetized. 



I have since published accounts of further experiments, 

 and amongst others of a series in which iron rings 



surrounded by magnetizing coils were used instead of 

 straight rods. The changes produced by magnetization 

 in the diameter of the rings were of exactly the same 

 nature, showing conclusively that the effects before ob- 

 served could not have been due to any unexplained action 

 of the ends of the rods. 



By the kindness of Mr. W. H. Preece, F.R.S., who 

 placed at my disposal the large secondary battery used in 

 lighting his house at Wimbledon, I have recently been 

 able to repeat some of my experiments with magnetic 

 fields of exceedingly high intensity. Rods of iron, nickel, 

 and cobalt were thus tested, and the results are clearly 

 shown in the accompanying curves, where the abscissae 

 represent the magnetic fields due to the coil in C.G.S. 

 units, and the ordinates the elongations and retractions of 

 the rods in ten-millionths of their lengths. 



The retraction of iron, it will be seen, becomes ulti- 

 mately greater in amount than its maximum elongation, 

 and reaches a limit in a field of 1000 or 1100 units, after 

 which its curve becomes sensibly parallel to the horizontal 

 axis. Nickel, unlike iron, retracts 1 from the very com- 

 mencement, rapidly at first and afterwards more slowly, 

 until in fields of 800 units and upwards its length becomes 

 apparently constant. Cobalt behaves in a very remark- 

 able manner. While the field is comparatively weak, no 

 sensible change in either direction can be detected. After 



1 The retraction of nickel under magnetization was first observed by Prof. 

 Barrett (Nature, xxvi. 585). 



about 50 units of magnetizing force, the rod begins to 

 contract, attaining its minimum length with 300 or 400 

 units. But instead of remaining unchanged in fields 

 stronger than this, it again becomes longer. At 750 it 

 regains its original length, and thence up to 1400, the 

 highest field reached in the experiment, it continues to 

 elongate steadily. 



It should be understood that so far as mere details are 

 concerned the curves in the diagram relate only to par- 

 ticular specimens of the metals in question. With different 

 rods there will be certain small variations, dependent upon 

 the purity of the metals and their physical condition. 

 But I have always found that under increasing magnetizing 

 force iron is at first extended and then contracted, nickel 

 is contracted from the beginning, while cobalt is first 

 contracted and afterwards extended. 



My best thanks are due to Mr. Preece, not only for 

 having given me the opportunity of carrying out the 

 experiments described above, but also for the exceedingly 

 kind and cordial manner in which he did so. 



Shelford Bidwell. 



A METEOROLOGIST AT THE ROYAL 

 ACADEMY. 



ARTISTS and poets are supposed to draw their in- 

 spirations from communing with Nature ; but it is 

 well known that painters in oil are rarely successful with 

 the cloud portion of their pictures. 



For some reason or other, skies and clouds are always 

 far more satisfactory in the water-colour exhibitions than 

 in galleries devoted to oils. The transparency of the 

 former medium enables a painter to put an amount of 

 detail into his clouds which would make the sky far too 

 heavy if attempted in oil ; so there is no doubt that oil 

 as a medium is peculiarly unsuitable for the reproduction 

 of cloud-forms, and that the utmost skill is required to 

 give even passable results. 



Painters generally do pretty well when they only try 

 to represent shading of the sky, or what Mr. Lockyer has 

 called the zoning of colour in the heavens. They can 

 paint the blue sky overhead gradually getting whiter and 

 grayer as you approach the horizon, or the red round the 

 horizon at sunset surmounted by a zone of orange shading 

 through green into the blue above, as only shade and 

 colour have to be rendered. But when artists try to 

 delineate the form, and, still more, the texture, of clouds, 

 the difficulties are so great that few painters attain 

 excellence in this branch of their work. 



Few have yet learnt that, putting the difficulties of the 

 medium aside, the structure of a cloud has an anatomy 

 as definite as that of a man ; and that the perspective of 

 cloud-forms obeys the same laws as that of bodies on the 

 earth's surface. Everybody paints ordinary objects so as 

 to show a characteristic texture or structure : the silk 

 dress, the woollen carpet, the wooden floor, are all care- 

 fully distinguished ; but how many realize the essentially 

 different structure of cloud-forms — the hairy cirrus, the 

 lovely fleecy sky, or the rocky masses of cumulus ? Nobody 

 would dare to draw a building, a road, or a tree out of its 

 due perspective ; but many seem to think that the forms 

 and distances of the sky can be rendered by daubing 

 white and blue and gray promiscuously over the canvas. 



The chapters on clouds in Mr. Ruskin's " Modern 

 Painters " sin against every canon of literature. They 

 are disjointed, discursive, irrelevant, and wander into 

 many by-paths ; for one note brings in the causes of the 

 failure of the Reformation in Germany, and the whole ends 

 with a commentary on the nineteenth Psalm. But, in 

 spite of all this, they preach in brilliant and poetic lan- 

 guage the two great truths that clouds have distinctive 

 characteristic structures, and that their perspective must 

 be as carefully drawn as that of a building. In one of the 



