April r, 1886] 



NA TURE 



513 



observation which at first sight seemed to show the existence of 

 residual diamagnetic polarity in a diamagnetic substance after 

 exposure to a strong field, and remarks that this seemed an 

 incomprehensible result. It appears to me that this result, 

 should it be confirmed, is not incomprehensible on Weber's 

 theory of diamagnetism, if we supplement that by a modification 

 of the Ampere-Weber theory of ordinary magnetisation. 



Suppose that the induced currents in the molecules of a dia- 

 magnetic substance are confined to definite channels, that there 

 is little or no primitive current, and that the molecules are 

 capable of being deflected. Then as the field is increased, 

 each molecule is turned so that the plane of its channel becomes 

 more and more nearly parallel to the lines of force. We may 

 assume that this turning of the molecules is resisted, like the 

 turning of the molecules of iron, and that when the field is 

 withdrawn they return more or less completely towards their 

 initial positions. 



Experiments with iron and steel show that in the turning of 

 the molecules the resistance while the field is being applied is 

 on the whole greater than the restoring force, while the field 

 is being removed : in fact something very like static friction 

 acts on each molecule. There is what I have elsewhere called 

 "hysteresis," or lagging behind, in the relation of the mole- 

 cule's movement to the magnetising force. If this molecular 

 quasi-friction also exists in diamagnetic substances, and if the 

 molecular channels are turned at all, they will, during the 

 removal of the field, be in less favourable positions for the in- 

 duction of currents than they were in during the application of 

 the field. There will consequently be a residue of current in 

 each when the field is wholly withdrawn ; and these residues 

 will make the substance a permanent "diamagnet." 



But the fact that this result would be comprehensible is no 

 evidence of its truth, and apparently Dr. Lodge inclines to 

 interpret the experiment referred to in an entirely difterent — 

 indeed opposite — sense. Moreover, his other results show not 

 residual diamagiit-tisni, but residual paramapielism in dia- 

 magnetic substances which have been immersed in a very strong 

 field. 



Now I think this result may also be interpreted in terms of 

 the magnetic theory of magnetisation ; and the purpose of this 

 communication is to suggest an explanation which seems to me 

 so probable that it may perhaps serve, until Dr. Lodge confirms 

 these results, as a set-off against the suspicion he has cast on 

 them by suggesting the presence of iron in his diamagnetic 

 bodies. 



When we begin to magnetise iron by a field which increases 

 from zero, we find at first scarcely a trace of magnetisation. A 

 curve showing the relation of intensity of magnetism to mag- 

 netising force starts off (as nearly as can be judged) tangent to 

 the line along which the magnetising force is plotted, but soon, 

 of course, takes ra rapid bend as the permeability increases. 

 1 his is very consistent with the idea that the molecular electro- 

 magnets are held back from turning by a sort of static friction 

 which requires the field to reach a finite value (different perhaps 

 for different molecules) before the process of turning begins. 

 But what has happened before this process begins ? Diamag- 

 netic induction has been going on in each molecule that has not 

 begun to turn ; and hence, if the molecular configuration is 

 rigid for a magnetising force of any finite magnitude, the sub- 

 stance is diamagnetic in that and all weaker fields. 



If this be the case in iron (and the experimental evidence 

 certainly points to the existence of a finite frictional resistance 

 to the turning of the molecules) that metal is really diamagnetic 

 in excessively weak fields, because the molecules are fixed by 

 friction ; then very paramagnetic in stronger fields, because the 

 molecules are turning ; and, finally, diauiagnetic in a field 

 strong enough to turn the molecules as far as they will go, and 

 to induce currents in them which swamp the primitive Ampcrian 

 currents. 



Next, imagine a substance whose molecules are held by fric- 

 tion in a very tight grip, sd that no moderate magnetising force 

 is able to alter their configuration. The substance is then dia- 

 magnet'c, and when the field is withdrawn there is no residual 

 polarity. But let a field be applied strong enough to begin 

 turning the molecules. This will cause a decrease of diamag- 

 netic susceptibility. And when the field is withdrawn the 

 molecule; remain deflected, and the substance is a permanent 

 paraniagiiet. 



Now this is exactly what Dr. Lodge has observed in his 

 copper, coke, wood, and so forth. They behaved as diaraag- 



netics while in the field, but showed paramagnetic polarity 

 when withdrawn from it. 



My suggestion, then, is that in diainagnetics, as in paramag- 

 netics, there are strong primitive Amperian currents circulating 

 in the molecular channels. That in a strongly paramagnetic 

 substance such as iron there is comparatively little molecular 

 rigidity, so that the molecules begin to turn even in very weak 

 fields ; the induction of currents in their channels then plays a 

 very insignificant part in the magnetisation. That in a dia- 

 m.agnetic substance, on the other hand, the molecules stick so 

 fast that in any moderate field they have scarcely begun to turn ; 

 the induction of currents goes on independently of the existence 

 of the primitive currents, and is then practically the whole affair. 

 But if the field be made strong enough the molecules begin to 

 turn, not in the way spoken of in the earlier part of this com- 

 munication (where it was assumed that the induced currents 

 swamped the primitive currents), but in the way in which the 

 molecules of iron turn. Then common magnetisation becomes 

 superposed on diamagnetic induction. And when the field is 

 withdrawn the molecules are left with a paramagnetic alignment, 

 and with their primitive Amperian currents strengthened, if 

 anything, since they have been facing more favourably during 

 the withdrawal of the field. 



There is nothing to show that the primitive Amperian currents 

 are not as strong and as numerous in copper or bismuth as in 

 iron. If they are, and if we could only apply a field strong 

 enough to force them into alignment, we might expect to find, 

 in substances so hard to magnetise, a permanence in the 

 residual magnetism which would put steel to the blush. 



University College, Dundee, March 27 J. A. EwiNG 



Ferocity of Rats 



I H.4.\E recently had occasion to chloroform a number of wild 

 rats for the purpose of procuring their blood. The rats are sent 

 to me by a ratcatcher, who places frum six to twelve in the same 

 trap or cage. It usually happens that, within a few hours after 

 their imprisonment, some of their number are killed and eaten 

 by the others ; while they all exhibit scars as the result of their 

 struggle for existence in confined quarters. 



A few days ago I placed two wild rats in a cage, and for a 

 long time endeavoured unsuccessfully to catch the larger one 

 under a bell-jar let in through a doorway in the top of the cage. 

 The rat perfectly well understood my object, and for about ten 

 ininutes succeeded by his agility in thwarting it. This animal, 

 therefore, must have been in as great a state of alarm as it is 

 possible that a rat could be. Nevertheless, after the ten minutes' 

 chase inside the cage — during which he had been many times 

 very nearly caught — he appeared to be suddenly seized with a 

 violent outburst of ferocity against his fellow-prisoner ; for 

 he fell upon the smaller rat, drove it into a corner of the 

 cage, and killed it by biting its throat. By means of 

 a glass rod I drove him away, drew the dead body 

 of his victim beneath the doorway in the roof of the 

 cage, and held the bottom of the bell-jar just above the 

 dead rat. I had not long to wait before the living one again 

 fell upon his victim and began to devour the carcass. It was 

 then an easy matter to lower the bell-jar over both the living and 

 the dead, when, by pouring chloroform in at the open top of the 

 bell-jar, I quickly reduced the murderer to a state of insensi- 

 bility. But up to the very last moment of consciousness this 

 animal continued to bury his fangs in the body of the little rat, 

 and even after his head had dropped away in stupor the jaws 

 still continued to move as if he «ere enjoying the feast in his 

 dreams. 



Now, I do not believe that any instance of ferocity at all 

 approaching this could be found in any other animal. But it 

 has been suggested to me that the fact may have been due to a 

 kind of emotional insanity produced by extreme terror. I there- 

 fore write to ask whether any of your readers can supply me with 

 additional facts bearing upon the subject. In particular, is it the 

 habit of wild rats when not confined, or when in a state of 

 nature, to devour one another ? Or do they only do so when 

 shut up together in a cage ? George J. Romanes 



The Recent Weather 



The enclosed extract from the log of one of the "excellent" 

 observers for the Meteorological Oflice may be interesting to 

 some of your readers, as bearing upon the large amount of 



