iS6 



NA TURE 



[June 20, 1895 



rotation — a rotation the other way about. What we really 

 perceive then is acceltralton of rotation, using the word ac- 

 celeration in its technical sense. If the turn-table is stopped, 

 this is a negative acceleration, and what we feel is that we are 

 l>eing turned round in a negative sense, and at the same time «e 

 feel our eyeballs jerk. The sense of rotation and the jerking die 

 away in this as in the former case. 



If, while we are Iwing turned round with uniform angular 

 velixity, but after all sense of rotation and all jerking of the 

 eyeballs have ceased, we ojK'n our eyes, we still feel ourselves 

 quite at rest, but we .see all external objects turning rounil us ; as 

 has been well .said by Prof. .Mach, the external «orId seems 

 to turn round inside an outer unseen fixed worhl. It is in 

 reference to this imaginary fixe<l world that our glance-lines are 

 now fixed. If the rale of rotation is changed while the eyes 

 are open, the sensation of rotation is exactly the same as if 

 they were shut, we feel the acceleration — [xjsitive or negative — 

 as a rotation in the one or in the other sense, and the jerks of 

 the eyeballs take place as if the real external world were not 

 there, and we were looking l)eyond it at the unseen fixed world 

 outside of it, that imaginary world in reference to which our 

 glance-lines are now fixed. 



If while the experiment I have described is going on, we 

 move so as to change the direction, in our head, of the axis of 

 rotation — for instance, if, after uniform rotation about a vertical 

 a.xis has gone on, with the head in its usual ui>right jiosition, 

 until the sense of rotation has cea.sed, we 1k)w our liead forwards 

 so that the axis of rotation is now parallel to a line from the 

 <Kciput to the chin, a very striking, and somewhat alarming, 

 but most instructive sen.salion is ex|x;rienced. What we feel is 

 that we are being turned round with a rotation which is the 

 resultant of two rotations of equal angidar vclixity — one the 

 real rotation about what is now the vertical, the other the 

 imaginar)- (but equally perceived) rotation in the op|X)site sense 

 alxiut the line in the head which was vertical. If the angular 

 movement of the head is small, so that the angle between what 

 is the vertical and what was the vertical is small, then the two 

 component rotations nearly neutrali.se one another, and the 

 .strange and alarming resultant is slight ; but if the head is l)ent 

 so that the old and new verticals are at right angles to one 

 another, the real and the imaginary comijonenLs are bf>lh felt in 

 full, and the effect is ver)' startling. If the rate of rotation is 

 changed simultaneou.sly with the change of position of the head, 

 we have a resultant of two rotations of different angidar 

 veliKity. The most easily observed c,i.se of this kind is when 

 the rotation is stopped altogether at the moment of change of 

 p)sition of the head. Mere the rc-al component is zero, and we 

 have only the imaginary one. This is the case of the well- 

 known practical joke : a man is asked to plant the poker l)efore 

 him on the flixir. pl.-ice his forehead on the end of it, walk 

 round it three times, and then rise and walk to the diHir. The 

 preliminary |iart of this ex|)erimenl presents no ditliciilty ; the 

 victim pl.ints the poker, puts his forehead cm it, walks round it 

 with the greatest ca.se and with no sense of anything unusual. 

 But when he rises, the line in his heail which w.is vertical is now 

 horizontal, and he feels himself turned round alwut that 

 horizontal line. The external world he also .sees turning round 

 this line, objects on the one side rising up and objects on the 

 other side sinking down. In this visibly sw.aying world he has 

 to guide his .sensibly rotating lifxly, and if his friends do not 

 catch hold of him he is pretty sure to fall. All these ex|XTi- 

 mcnLs are most conveniently made on a smrxjthly working turn- 

 t.ablc of .such a size that one can comfortably lie down U|xjn it. 

 liy the kindness of Messrs Dove, lighthouse engineers, I had 

 the Use of a large turn-table made for the revolving lantern of a 

 lighthouse. It could lie turned round smiKithly and uniformly, 

 at the mrxicralc si)eed that is most .suit.ible for eX|K-rimenls of 

 the kinil in question. A few exixriments with such an ap- 

 jiaraiiis will convince any one that we have here to do with a 

 |>crfectlv definite .scn.sc, and not with any vague sensations 

 caused dy the inertia of the soft jxirts of the Ixwly. 



This is one of the ways in which the ])henomena have Ijccn 



''-■ ' ' ' ■ '■ '• who hesitate to Ix-lieve that there can Ix; a 



'■ -e 1 inly discovered »ilhin the last few years. 



I „ ' the sensiition is not in the soft parts of the 



Ixxly generHJIy, Init in the head, is made ix;rfeclly plain by the 

 fact that the (xisilion of the head and the changes of that 

 |«>siii<.n alone determine Ihe sensations. We must therefore 

 |i-ik in the head for Ihe organ of this .sen.se. 



In clone proximity to the cochli-a. which is universally re- 



NO. 1338, VOL. 52] 



garded as the oi^n of hearing, there is an organ of very 

 striking, and we might say mysterious, form. It occurs in all 

 vertebrates, and occurs in them fully developed, except in tlie 

 lowest forms of fish. It is contained in a bony or cartilaginous 

 cavity, which communicates with the ctxrhlea t»r lagena. This 

 ca\-ity may lie divided into the vestibule and the three seniiciicular 

 canals. The canals i^jk-u at both ends into the vestibule, and eacli 

 has at one end an enlargement called the ampulla.' Within this 

 bony case is contained a membranous structure, consisting of the 

 utricle, situated in the vestibule, and three membranous canals, 

 each in one of the bony canals, each with an ampulla in the 

 Irany ampulla, and each opening at both ends into the utricle. 

 The vestibule contains, besides the utricle, the s.-icculc. a mem- 

 branous Ixig continuous with the cochlear tiuct, and has in the 

 side next the tympanic cavity a hole in the bony wall fdled in 

 by a membrane, and known .as theye«<',r/ra m'alis. The s;iccule 

 and the utricle have each a spot on the lower wall supplied with 

 nerves ending in hair-cells, and known as the matiiln aaistua. 



The maciilic aciisticit are probably, as suggested by Mach and 

 Breuer, organs fitted to ]ierceive acceleration of translatory 

 motion, and are iu)( connected directly with the function of the 

 semicircular canals. The fiiirslra 07Hi/is belongs to the organ of 

 hearing, which may thus be said to have a right of way througli 

 the vestibule. We need not therefore here consider any further 

 these organs, but confine ourselves to the semicircular canals and 

 the utricle in its relation to them. As already stated, each Ixmy 

 canal contains a membninous canal. The membranous canal is, 

 except at the ampulla, nuich smaller in bore than the bony 

 canal, so that the sjiace outside the mend)ran»>us canal filled « ilh 

 |x;ril)Tnph, is much greater than the sjwce inside filled with 

 endolymph. The membranous ampulla much more nearl)- fills 

 the bony ampulla, so that here the perilymph sjxice is compara- 

 tively small. The membranous canal is pretty firmly attached 

 (in .some animals, at all events) to the periosteum of Ihe bony 

 canal, so that in man a section has somewhat this form : 



ETX 



'iT^TUpK 



endolynipK 



Each canal is, in all animals I have examined, approxim.ately 

 in a plane, and it is iniporlanl to consider the relations of these 

 planes to one aimlher and to Ihe mesial plane of the head. 



As I have limughl part of the api«ratus with me, I may 

 shortly describe the melluxl I used to measure the angles which 

 these planes make with one another, and also an improved 

 methixl, of uhich I have not yet had time to make any \ery h\\\ 

 trial. 



[The method illustrated by tlie luinian skull shown is fully 

 described, with woodcuts from nholographs, in I'rof. McKend- 

 rick's " Text-book of I'hysiolDg^-, ' vol. ii. pp. 697-699, and there- 

 fore need not be reprinted here. The other metluxl will, I ho|K-, 

 give more accurate measurements.] 



It consists in alt.aching the preimration -either a cast of the 

 canals, or, in the case of a bird, the dissected and cleaned Ixjny 

 canals — to one ann of a branched rod, and a lump of wax to the 

 other. The ro<l is then fixed to the large apparatus already 

 referred to. The canals are successively made liorizontal, and a 

 small plate of glass fixed horizontally in each case- parallel 

 therefore to e.ich canal lo tlic lump of wax. We can alsii 

 att.ach a glass plate parallel lu the mesial plane. We can thus 

 have, on a com|«ratively small piece of wax, glass plates parallel 

 to all Ihe planes, the relations of which to one another are to be 

 me.-usured. The lump of wax is then remove<l from the rod, and 

 the angles between the iilanes of the glass jilatcs measured by 

 means of an ordinary reflexion goniometer. 



The general results art : 



(1) The canals do not lie rigorously in ])lanes, but sulliciently 

 nearly so to give closely accordant results. 



(2) The external canals are very nearly at right angles to the 

 mesial plane, and therefore, from the bilateral symmetry, Ihe two 

 external canals are very nearly in one plane. 



{3) The superior and |x).sterior canals of the .same side make 



* In .-ill anim.iU (he non-.impuU.irx- ends of the superior and the iiosterior 

 canal li.ivc .1 common opening into tlie vcstiliiilc. 



