MEMOIRS.

NEW OBSERVATIONS upon the MINUTE ANATOMY of thePAPILLAE of the FEOG'S TONGUE. By LIONEL S. BEALE,M.B., F.R.S., Fellow of the Royal College of Physicians,Professor of Physiology and of .General and Morbid Ana-tomy in King's College, London; Physician to King'sCollege Hospital, &C.1 With Plates I—IV.

IN this paper I propose to give the results of some recentinvestigations upon the minute anatomy of the beautifulfungiform papillae of the tongue of the little green tree-frog{Hyla arborea). The specimens have been prepared accord-ing to the principles laid down in former communications.The success I have met with in this and other minute anato-mical inquiries is, I believe, almost entirely due to the pro-/ 0cess of investigation which I have adopted for some yearspast, and which enables me to render specimens very trans-parent, and to demonstrate all the tissues in one specimen,a result which, as far as I am aware, can be obtained in noother way. By this plan sections are obtained so exceedinglythin, without the destruction even of the most delicate tissues,that they may be examined under the highest powers which itis possible to obtain (VV magnifying 1700 linear, and -3^ magni-fying about 3000 linear).

The following are among the most recent contributions tothe anatomy of the papillae of the frog's tongue:

WALLER: "Minute Structure of the Papillse and Nervesof the Tongue of the Frog and Toad," ' Philosophical Trans-actions,' 1847.

^ BILLROTH : " Ueber die Epithelzellen der frosch-zunge,«IQ sowie uber den Bau der cylinder-und flimmerejiithelien und•"—•.jhr Verhiiltniss zum bindegewebe," ' Archiv fur Anat. Phys./

^ 1858, S. 163.^v^. HOYBHI: " Mikroskopiche Untersuchungen uber die zunget^des Frosches," 'Archiv fur Anat. Phys.,' 1859, S. 488.

1 This paper lias been reprinted from the 'Phil. Trans.,' but lias beencarefully revised by the author for this Journal. Several additions have been

ade to it, and a new figure added.VOL. IX. NEW SER. A

AXEL KEY : " Ueber d. Endigungen d. Geschm'acksnerveniu der zunge Frosches," ' Miiller's Archiv,' 1861, S. SH9.

HAKTMANN: " Ueber die Endigungsweise der nerven inden PapillEe-fungiformes der Froschzunge,"' Archiv fiir Anat.Phys.,' 1863, S. 634.

ENGELMANN : " On the Terminations of the GustatoryNerve in the Frog's Tongue," 'Siebold und Kolliker'sZeitschrift,' Bd. xviii, heft i.

In Part I of ' Max Schultze's Archiv' for 1868 will befound " A Memoir on the Taste Papillae of the Tongue/' byDr. Christian Loven, translated from the Swedish, to whichattention may be directed. The author records the demon-stration of some peculiar cells which are probably connectedwith the sense of taste in the higher vertebrata. The cells inquestion, which are interspersed amongst the ordinaryepithelial cells, perhaps correspond with the mass of cells ofthe papillse of the frog's tongue described in this paper, whichare, without doubt, very intimately connected with, the nerve-fibres of the plexus, which I have demonstrated, at thesummit of the papilla.

Although the views of Axel Key are supported byschematic figures which do not accurately represent the realarrangement of the tissues, they approach much nearer to thetruth than those of other observers. He describes two kindsof cells at the summit of the papilla, epithelial cells and specialcells concerned in taste. I have not been able to verify hisstatements in this particular. He has not demonstrated thepeculiar network at the summit of the papilla wliich is seenso distinctly in my specimens, and his delineations of the pro-longation of the axis-cylinder alone, and its division intofibres far too fine to be visible by the magnifying powersemployed, and the abrupt cessation of the white substancedelineated by him, are evidently schematic,—indeed, he doesnot pretend that the figures referred to are copies from nature.Still his inferences regarding the division of the nerve-fibresinto very fine fibres which pass into the epithelium-like tissueat the summit of the papilla, approach much nearer to theactual arrangement than those of any other observers withwhich I am acquainted.

Among the later researches upon the mode of termination ofthe nerves are those by Dr. Hartmann. These are concluded inthe number of Reichert and Du Bois-Reymond's * Archiv' for1863. The drawings of the papillse accompanying this memoir,especially fig. 65, PI. 18, which I have copied (see fig. 24,PI. IV, of this memoir), form an excellent illustration of the

fact that the most beautiful and well-defined structures maybe rendered quite invisible by being soaked in aqueous solu-tion of bichromate of potash for three days, one day in car-mine solution, and then in caustic soda !

In order that I may not express myself against the mode ofpreparation followed by this and many other observers inGermany in the present day more strongly than is justifiedby the results obtained as shown by their own drawings, Iwould refer to Hartmann's figure. Of this drawing it isnot too much to say that it represents nothing sufficientlydefinite to enable any one to form an idea of the structure ofthe part. The drawing, and I conclude the preparation fromwhich it was taken, are far behind the day; and it seems tome most remarkable that after all the anatomical research ofthe last twenty years an observer should publish such a figureas this as a representation of natural structure. The nerve-fibres are completely altered by the mode of investigationfollowed, and the finer fibres are of course destroyed orrendered invisible. Nor can I admit that the epitheliumupon the summit of the papillae represented in his fig. 64gives a correct idea of this structure.

Unfortunately, Engelmann, who is the last writer on thesubject, does not appear to have seen my paper, which was pub-lished three years before his own memoir. He, however, con-firms some of my observations, but has evidently failed to seethe numerous nuclei connected with the subdivisions of tbenerve-fibres, and represented in my figures 1, 2, and 3.That remarkable and intricate plexus of the very finestnerve-fibres which I have described and figured has alsoescaped his observation. But it would have been impossiblefor him to demonstrate these and many other points by theprocess of investigation which he followed, while the papillaeof the Rana temporaria are not at all well adapted for study-ing the finest ramifications of the nerve-fibres.

It may be proved conclusively by experiments that soakingdelicate animal tissues in dilute aqueous solution of bichro-mate of potash renders invisible and destrO)S structureswhich can be demonstrated by other means. Iuquiries con-ducted by the aid of such plans of preparation retard ratherthan advance anatomical inquiry, for some of the most im-portant anatomical characters are rendered completely invi-sible. The very conflicting opinions now entertained byobservers in Germany upon the structure of these papillfe,render it important that they should be studied again withthe advantage of the highest powers and the most advan-tageous methods of preparation which we now possess.

In this communication I sliall only attempt to describebriefly those poiuts which I believe to be new, and which are,I conceive, demonstrated in my specimens for the first time.Most of the points described in this paper were demonstratedin 1862, and the specimens have been repeatedly studied andshown to other observers. The points described could still bedemonstrated in the same specimens in June, 1864.

The structures entering into the formation of the papillaare the following:

1. The connective tissue which forms the body of thepapilla.

2. The " epithelium."3. The nerve-fibres in the body of the papilla, and the

fibres prolonged from them which form a plexus uponits summit.

4. Nerve-fibres ramifying in the connective tissue, uponthe capillary vessels and amongst the muscular fibves.

5. The muscular fibres.6. The vessels.

The Connective Tissue.The nerves, vessels, and muscular fibres are embedded in a

very transparent basis-substance which exhibits a slightlystriated or fibrous appearance when stretched, but this struc-ture in all the papillae of the Hyla is exceedingly delicate andtransparent.

The great majority of the masses of germinal matter (nuclei)seen in this basis or connective substance are undoubtedly con-nected with the nerves, vessels, and muscular fibres, but thereare a few which seem to belong to the connective substancealone, and may therefore be called " connective-tissue cor-puscles." It is possible that these at an earlier period mayhave been connected with nerves or muscles; at any rate,they are the descendants of the same nuclei or masses of ger-minal matter as the nuclei taking part in • the production ofthese tissues.

I consider that indefinite connective tissue of this kindresults principally from the accumulation of the remains ofhigher structures, especially nerve-fibres, which were in astate of functional activity at an earlier period of life. At anearly period of development nuclei (masses of germinal mat-ter) can alone be detected. As development proceeds, tissueis formed by these nuclei, and increases as age advances. Thelarge and fully-formed fungiforai papillee have twice as manynerve-fibres as smaller and younger ones. During thedevelopment of such an organ as one of these papillse many

changes occur, and much texture is probably preduced andremoved before the papilla attains its fully developed state. Thatpassive substance called connective tissue which remains andoccupies the intervals between the higher tissues, whichpossess active and special endowments, slowly accumulates,but undergoes condensation as the organ advances in age.Amongst this are a few nuclei which can no longer produceanything but indefinite " connective tissue" of the same cha-racter. In PL II, fig. 9, it would have been impossible, hadthe specimen been prepared in the usual manner, to havedetermined if the nuclei marked a, b -were nuclei of themuscle concerned in producing muscle, or connective-tissuecorpuscles concerned in the formation of connective tissueonly. This question requires restudy from a new point ofview. It is quite certain that many of the masses of germi-nal matter (nuclei) figured in all my drawings in connexionwith nerves, vessels, muscles, and other tissues, would, if~thespecimens had been prepared in a different manner, so thattheir connexions were not so very distinctly seen, have beenCalled " CONNECTIVE-TISSUE CORPUSCLES."

The drawings accompanying my paper explain the rela-tion which I believe the essential structures entering into theformation of the papilla bear to the indefinite connectivetissue in which they lie embedded.

Epithelium.

The so-called epithelium upon the summit of the papilla ofthe frog's tongue (PL I, fig. 1, a) differs from the epitheliumattached to its sides (b), that covering the simple papillae (c),and that on the surface of the tongue generally, in many im-portant characters. As is well known, it. is not ciliated. Thecells differ from the ciliated cells in several points. They aresmaller than these. The nucleus is very large in proportionto the entire cell. The cells are not easily separated from oneanother, as is the case with the ciliated epithelium. Thesecells form a compact mass, the upper surface of which is con-vex. This is adherent by its lower surface to the summit ofthe papilla, and it is not detached without employing force.The cells do not separate one by one, as occurs with the ordi-nary epithelium, but the whole collection is usually detachedentire, and then it is, I believe, torn away.

Although some observers would assert that the two or threelayers of cells represented in my drawings do not exist, butthat the appearance is produced by the cells of a single layerbeing pushed over one another by pressure, I am convinced

that in this mass upon the summit of the papilla of the Hylathere is more than the single layer of cells represented byHartmann, who is the latest observer on this point.

Hartmann's representation (I. c.) of this very same struc-ture from the summit of the papilla of the Hyla is verydifferent from my drawings. Not only do we represent thesesame cells of very different shapes, but the nucleus in myspecimens is 'three or four times as large in proportion to thecell as represented by him.

The general outline of the free surface is convex (a, a, a,fig. 1, PI. I), and the tissue which intervenes between thenuclei appears very transparent and projects a little, so as togive the convex summit a honeycombed appearance (PL II,fig. 7).

The under concave surface of this hemispheroidal masswhich adheres to the summit of the papilla of the Hyla'stongue, corresponds to the exact area over'which the nerve-fibres of the papilla are distributed, as will presently beshown. The shape of these cell-like bodies, of which themass is composed, and their connexion with fibres, is shown inPI. II , fig. 3, and in the very highly magnified specimen repre-sented in fig. 2, PL I. These figures, I believe, represent theactual arrangement, but the point is most difficult of investiga-tion. In the intervals between what would be called, if theywere capable of complete separation from one another, the in-dividual cells, fibres are seen. These fibres do not, I think,arise simply from the pressure to which the masses have beensubjected. I have represented trie arrangement as I believeit to be in PL II , fig. 6, from the central part of one of thehemispheroidal masses. I regard the entire hemispheroidalmass as resembling in its essential structure the network Ihave described at the summit of the papilla, but the massesof germinal matter are so very close together and the fibresso much interlaced with one another, that it is most difficultto unravel the mass without destroying it. The arrange-ment at the surface is seen in PL II , fig, 7.• The epithelium of the tongue generally is easily removed,but many of these hemispheroidal masses remain connectedwith the summits of the' papillce to which they belong. Fromwhat I have stated, it will, I think, be admitted that the con-stituent parts of the mass at the summit of the papilla couldnot be properly called epithelial cells, so that, with referenceto the termination of the nerves in the papilla, I think it ismore correct to say that nerves may be traced to specialbodies or cells which form a hemispheroidal mass attached tothe summit of the papilla, than to assert that the separate

bodies, which, compose the mass in which nerves terminate,are actual epithelial cells.

In the simple papilla (PI. I, fig. 1, d) of the frog'stongue, a " nucleus " of a nerve sometimes projects beyondthe outline of the papilla and lies amongst the epithelium.This nucleus, however, adheres to the papilla when all theepithelial cells have been detached. It might from its posi-tion be easily mistaken for an epithelial cell, but it is nomore really related to this structure than is a ganglion-cell,or a caudate nerve-cell of the spinal cord. The cells of theciliated epithelium of the frog's tongue are not in any in-stance, as far as I am able to observe, connected -with thenerve-fibres. It is probable that the opposite inference,which is still held by many observers, has resulted from theobservation of such a nucleus as is represented in PI. I,fig. 1, d, projecting beyond and adherent to the surface ofthe papilla. It is really continuous with the delicate nerve-fibres (e) ramifying in the substance of the papilla, but it isnot an epithelial cell, and remains adherent after every par-ticle of epithelium has been removed.

More recently, the view that the finest ramifications ofthe nerve-fibres run amongst the epithelial cells of varioustissues has been gaining ground, but it is probable thiswill turn out incorrect.

The nervous tissue is in all cases structurally distinct fromevery other tissue, in every part of its distribution. It neverblends with epithelium any more than it blends with fibroustissue, cartilage, bone, or mus.cle. If nerves exert any directinfluence upon the nutrition of any of these tissues, the in-fluence must be exerted through some distance. The resultsof anatomical research render any physiological doctrineAvhich maintains that nerves act through their structuralcontinuity with other tissues untenable. My own observa-tions lead me to conclude that nerves do not directly influencethe processes of nutrition, growth, or development at all.They act only indirectly, and affect the supply of nutrientmatter distributed, by modifying the calibre of the vessels,and hence regulate the supply of blood which passes to thecapillaries. The nerves, I believe, really exert their in-fluence upon the contractile muscular coat of the smallarteries and veins alone, and do not act directly upon anyother tissues.

The Nerves.

With regard to the trunks of the nerves, I remark the fol-lowing facts of importance:

L That the bundle of nerve-fibves distributed to a papillaalways divides into two bundles which pursue opposite direc-tions. The division of the bundle may take place just at thebase of the papilla, or at some distance from it, but it alwaysoccurs (PI. I, fig. 1).

£. Fine pale nerve-fibres pass from the same trunk of dark-bordered fibres as that which gives off the bundle of nervesto the papilla. The fine fibres ramify—

a. Amongst the muscular fibres of the tongue (figs. 1, 9).b. Upon the vessels (fig. 1, i, i, i).c. In the connective tissue of the tongue generally, and

also in the simple papillse (fig. 1, d, e).The division of the bundle at the base of a papilla is

shown in PI. I, fig. 1, and in PL III , fig. 10, is a diagramto indicate the manner in which the nerve-plexuses at thesummits of the papilla are connected together by commissuralfibres. Thus, in action the papillae may be associated to-gether. The bearing of this arrangement upon the existenceof complete nervous circuits is discussed in my 'Archives,'vol. iv. The bundle in the central part of the papilla con-sists of dark-bordered fibres, which frequently cross andinterlace with one another in this part of their course. Theyvary much in diameter, some being so fine as scarcely to bevisible.

As the bundle passes towards the summit of the papilla,the individual fibres divide and subdivide into finer branches.Now, as I have before remarked, nerves so near their distri-bution as these do not usually possess an axis-cylinder as astructure distinct from the white substance. The white sub-stance does not abruptly cease, while the axis-cylinder isalone prolonged onwards by itself as is often described, butthe entire fibre divides and subdivides. In fact dark-borderednerve-fibres, near their ultimate ramifications, always consistof fatty albuminous material embedded in a transparent matrixof connective tissue. The " tubular membrane," " white sub-stance," and " axis-cylinder " can never be demonstrated asdistinct structures near the peripheral distribution of nerves.The " tubular membrane " is nothing more than the trans-parent matrix in which one or more nerve-fibres are em-bedded.

The dark-bordered fibres divide into finer fibres about thelevel of the ring or half-ring of capillaries at the summit ofthe papilla. As the fibres are exceedingly transparent, theyare usually lost from view about this point. For example,Hartmann's figures convey the idea that distinct dark-bordered fibres can be followed as high as this point, but

that they cannot be traced further. Above this spot thepapilla is a little thickened and the tissue more granular,and hence it is not to be wondered at that great difficultyshould have been experienced in demonstrating the furthercourse of the nerves, or that many different views should beentertained upon the oft-debated question of the mode ofending of nerves in this situation; but it is most certain thatthe fibres do divide and subdivide into finer and much moretransparent fibres at this point, and that these again divideand subdivide and form an elaborate plexus in the summit ofthe papilla, which has not been before described.

By reference to my figures, the arrangement, which, is noteasily described with accuracy, will be at once understood, sothat a minute description of it would be superfluous.

Above the plexus c (PI. II, fig. 3), and below the epithe-lium-like organ at the summit of the papilla (a), is a layer (b)which appears to be composed of granular matter. In mymost perfect specimens, however, this " granular layer,"when examined by very high powers under the influence ofa good light, is seen to consist of a plexus of extremely finefibres which interlace with one another in every direction,but which pass from the plexus above to the epithelium-likenervous (?) organ upon the summit of the papilla (PI. I,fig. 2). I believe this granular appearance to result fromthe extreme delicacy and fineness of the nerve-fibres at thispart of their course. In like manner the " granular matter "seen in the grey matter of the cerebral convolutions and thatof the retina, results mainly from the breaking down of veryfine and delicate nerve-fibres, which undergo disintegrationvery soon after death, unless they are subjected to specialmethods of preparation.

Of the existence of the elaborate network of nerve-fibreswith the large nuclei, represented in PL II, fig. 3, c, therecan be no question whatever; but there may be some differ-ence of opinion regarding the exact relation of the very finenerve-fibres at the summit of the papilla, to the peculiar cellswhich surmount it, and the ultimate ramification of thefibres of the elaborate plexus just described. However, thereare but two possible arrangements:

1. That the nerves form a network of exceedingly finefibres upon the summit of the papilla, upon which the basesof the epithelium-like cells impinge.

2. That the very fine nerve-fibres are really continuouswith the peculiar and epithelium-like cells; in which casethese bodies must be regarded as part of the nervousapparatus.

10

There seems to me to be so much strong evidence infavour of the last view, that I venture to express a decidedopinion that this is the truth. In many specimens I haveseen, and most distinctly, the delicate network of fibres re-presented in PL II, fig. 3, continuous with the fine nerve-fibres in the summit of the papilla, and I have demonstratedthe continuity of these fine fibres with the matter of whichthe outer part of these peculiar cells consists (Pis. I & II, figs. 2,3, 6). I have also seen what I consider to be nerve-fibres inthe intervals between some of these cells (fig. 7). Upon .the whole I am justified in the inference that there is astructural continuity between the matter which intervenesbetween the masses of germinal matter at the summit of thepapilla and the nerve-fibres in its axis, and I consider thatan impression produced upon the surface of these peculiarcells may be condurted by continuity of tissue to the bundleof nerve-fibres in the body of the papilla. These peculiar

. cells in the summit of the papilla cannot therefore be re-garded as epithelium, and the mass constitutes a peculiarorgan which belongs, not to epithelial structures, but to thenervous system.

Although there can be no doubt whatever as to the exist-ence of an intricate and 'exceedingly delicate nervous net-work or plexus at the summit of every papilla, such a plexusmight he connected with the nerves according to one of twovery difFerent arrangements:

1. The plexus might be formed at the extremity of a nerveor nerves, as represented in diagram (PI. IV, fig. 17).

2. The plexus might form a part of the course of a nerveor nerves, as represented in diagram (PI. IV, fig. 18).

If the first be true, the network must be terminal, and im-pressions must be conveyed along the fibre, of which theplexus is but the terminal expansion, direct from peripheryto centre. If the second arrangement is correct, the networkforms a part of a continuous circuit or of continuous circuits.I believe the division of the nerves at the base of the papilla,already adverted to, is alone sufficient to justify us in accept-ing the second conclusion as the more probable; but whenthis fact is considered with reference to those which I haveadduced in my paper published in the ' Transactions' for1863, and that in the ' Proceedings' for June, 1864, and theobservations published in several papers in vols. ii, iii, aridiv of my 'Archives/ and in the Croonian Lecture for 1865,I think the general view in favour of complete circuits isthe only one which the anatomical facts render tenable. Themode of branching and division of trunks and individualfibres is represented in PL IV, figs. 20, 21, 22, 23.

11

From the number and size of the nerve-fibres constitutingthe bundle in. the centre of the papilla, we should infer thatthe finest ramifications resulting from the subdivision of thesebranches would be very numerous, since it has been shownthat the fine fibres resulting from the subdivision of a singledark-bordered fibre in the frog's bladder, palate, skin, andmuscle, constitute plexuses or networks which pass over avery extended area. The mode of formation of a nerve-plexus is represented in PI. I l l , figs. 11 and 14. In thesebeautiful Little organs the numerous fibres resulting from thesubdivision of the dark-bordered fibres are distributed over acomparatively small extent of tissue, forming the summit ofthe papilla. Still, we have the same formation of plexuses,the constant change in the direction taken by fibres, and thesame crossing and intercrossing which have been noticed inother situations. In fact, the nerve distribution in theseorgans presents the same typical arrangement as is met within other tissues, but it is compressed into a very small space.

Now, with regard to the epithelium-like structure uponthe summit, it has been shown that the nerve-fibres are pro-bably continuous with the material lying between the largenuclei. In fact, if the interpretation of the appearanceswhich I have given be correct, the' arrangement may be ex-pressed thus:

The material marked a (PI. I, fig. 2) is a continuationof the nervous structure or tissue, while the matter marked bbears the same relation to this as the so-called nucleus of anerve bears to its fibre, or that of an epithelial cell to its wall.If this be so, the matter which is freely exposed at the verysummit of the papilla is at least structurally continuous withnerve-tissue, if it is not to be regarded as nerve itself. Myown opinion is that it is just as much nerve-tissue as a finenerve-fibre is nerve-tissue, or the caudate process of a nerve-cell is nerve-tissue. The formed matter is produced by thelarge masses of germinal matter which are so very numerous,just as the formed matter of a central nerve-cell results fromchanges occurring in its germinal matter.

It may not be out of place here to consider how theelaborate organ connected with the bundle of nerve-fibres ofthe papilla may act during life. As already stated, the freesurface is uneven, and the arrangement is such that there aremany elevations projecting, like fibres, by slightly varyingdistances, from the general surfaces. Now, from the intricateinterlacement of the nerve-fibres in the summit of a papilla,as well as at the point between this and the peculiar organ0*1. II, fig. 3, b), it is obvious that a fibre given off from one

coming from the extreme left of the papilla, for example, maybe situated a very short distance from a fibre coming fromthe opposite side. Any object, therefore, which connects theexposed projections would produce a temporary disturbancein the nerve-currents which are traversing these fibres, andthis alteration in the current would, of course, produce achange in the cell or cells which form part of the same circuitin the nerve-centre. Any strong pressure would influence allthe fibres distributed to this delicate nervous organ.

The supposed mode of action is explained by the plan(PI. II, fig. 4).

Nerve-fibres ramifying upon the capillary vessels, in the con-nective tissue, and upon the muscular fibres.

Many of the so-called connective-tissue corpuscles, withtheir anastomosing processes or " tubes," are really nerve-nuclei and very fine pale nerve-fibres, as has already beenshown in observations upon the frog's bladder. In the tongueI have followed these fine fibres in very many specimens.They can only be seen and traced in specimens prepared insyrup, glycerine, or other viscid medium miscible in all pro-portions with water.

In PI. I, fig. 1,/ , and in fig. 8, one of these fine branches,coming off from a bundle of dark-bordered fibres, is repre-sented. Now, if examined by a low power, this might bemistaken for a fibre of connective tissue ; but it really con-sists of several very fine fibres, which in their arrangementexhibit the same peculiarities observed in nerves ramifyingin larger trunks (PI. IV, figs. 20, 23). The fine branchesdivide and subdivide, and the delicate fibres resulting fromtheir division can be followed for a very long distance. Thefinest are composed of several finer fibres, and they form net-works or plexuses, the meshes of which vary much in size.

The branches which are distributed around the capillaries,in the connective tissue, and to the muscular fibres, seem toresult from the division and subdivision of the same fibres(PL I, fig. 1).

Nerves which are constantly distributed external to thecapillary vessels and in the connective tissue have beendemonstrated by me (PI. I l l , fig. 15) (see ' Archives,' vol. iv,p. 19). I consider these fibres as the afferent fibres throughwhich an impression conveyed from the surface or from thetissues around capillaries influences the motor nerves distri-buted to the small arteries from which the capillaries arederived. It is probable that these nerve-fibres pass to the

13

very same set of central cells as that from which the vaso-motor fibres take their rise. It is through these fibres thatchanges in the nutrition of the tissues may affect the circula-tion in the neighbouring vessels.

In these fungiform papillae, then, there are—1. The bundle of nerve-fibres which is distributed to the

sensitive nervous organ at the summit.2. Delicate fibres which may be traced to fibres running in

the same bundles as purely sensitive fibres. These delicatefibres are distributed—

o. Around the capillaries of the papilla (PI. I, fig. 1, i).See also PL III , fig. 15.

b. Some fibres ramify in the connective tissue of thesimple papillae (PL I, fig. 1).

c. Some are distributed to the muscular fibres (figs. 1&9).

Now, the first and second fibres are probably sensitive,excitor, or afferent, whilst the last must be motor. From thisobservation it follows that certain afferent and motor fibresare intimately related at their distribution, a conclusionalready arrived at in my investigations upon the distributionof the nerves to the frog's bladder, the palate, and pharynx.Moreover, I think that fibres passing from the plexus of sensi-tive fibres at the summit of the papilla, establish here andthere a structural continuity between these and the fibresramifying in the connective tissue and around the capillaryvessels. It is very difficult to obtain a specimen whichrenders this perfectly certain, but I have been led to a similarconclusion in investigations upon the corpuscula tactus of thehuman subject. The physiological interest and importance ofthis branch of anatomical inquiry are so great, and it pro-mises to lead to such important results, that it cannot be toominutely or too patiently worked out.

0 / the Muscles.

The muscular fibres of the papilla (PI. I, fig. 1, /*) are thecontinuations of muscular fibres in the substance of thetongue. They are excellent examples of branching stripedmuscle. The finest branches are less than xowoth of an inchin diameter, but these exhibit the most distinct transversemarkings. The markings, however, gradually cease, and thefibre becomes a mere line, which is lost in the connectivetissue at the summit of the papilla. The arrangement will beunderstood if figs. 1 & 9 be referred to.

The so-called nuclei or masses of germinal matter in con-

14

nection with these fine muscular fibres present several pointswhich will well repay attentive study. These masses ofgerminal matter are sometimes twice or three times the widthof the fibre with which they are connected. In a paper pub-lished, in Part XIV of my ' Archives' I have adduced factswhich render it probable that these nuclei or masses ofgerminal matter change their position in a very remarkablemanner during life.

The conclusions I have arrived at upon this point are asfollows:

The masses of germinal matter appear to move along thesurface of the already formed muscular tissue, and as theymove part of their substance becomes converted into muscle(PI. I l l , fig. 13). It is in this way that new muscle is formedand new muscular tissue is added to that already produced.The germinal matter itself does not diminish in size, becauseit absorbs as much pabulum as will compensate for what itloses of its OAvn substance by conversion into tissue. In theyoung muscle the nucleus increases in size.

From what I have observed, I think that these oval massesof germinal matter move in different directions, but always ina line with the fibrillated structure, so that in a muscle somewill be moving upwards, some downwards ; and when thenuclei are arranged in rows or straight lines, the nuclei lyingin adjacent lines will be moving in opposite directions.During the formation of a muscle these masses undergo divi-sion in two directions, longitudinally and transversely. Thetwo masses which result from the division of one will pass inopposite directions.

As is well known, the position of these nuclei with respectto the formed muscular tissue is very different in differentcases. Sometimes they are in the very centre of the elemen-tary fibre, as in the constantly growing fibres of the heart,sometimes upon its surface only, sometimes distributed atveiy equal distances throughout its substance. Whereverthese nuclei are situated, new muscular tissue may be pro-duced, and it is only in these situations that muscular tissueever is produced; so that by the position of the nuclei welearn the seat of formation of new muscle at different periodsof life.

The facts which I regard as favorable to the view aboveexpressed concerning the movements of the masses of germinalmatter of muscle, are derived from many sources, but I willrefer to some observed in the case of the muscles of thepapillae of the tongue. Here the muscular fibre is very thinand delicate, and very favorable for observation. The mass

15

of germinal matter is very much wider than the muscle.Often three or four of these masses are seen close together(PL II , fig- 9), while for some distance above and below themuscular fibre is destitute of nuclei. The narrowest extre-mity of the oval mass is directed in some cases towards theterminal extremity of the muscle, in others towards its base.There are often three or four fine fibres branching off fromone stem, and gradually tapering iu fine threads towards theirinsertion at the summit of the papilla (figs. 1 & 9). Thenuclei are three or four times as wide as these fibres.The greatest difference is observed in the distance betweencontiguous nuclei connected with the very same fibre. If themuscle had gone on growing uniformly iu all parts since theearliest period of its development, the nuclei would be sepa-rated from one another by equal distances, or by distancesgradually but regularly increasing or diminishing from oneextremity of the fibre towards the other.

I think the irregular arrangement of the nuclei in thesemuscular fibres of the tougue is to be accounted for by theirmovements. Perhaps, of a collection of these nuclei closetogether, two may be moving upwards towards the narrowextremity of the muscle which is inserted into the connectivetissue, while the third may be moving in the opposite direc-tion.

In some instances a " fault " is observed in the productionof the muscular tissue, as if the nucleus had bridged over aspace and formed a thin layer or band of muscular tissue,which, when fully formed, was separated by a narrow spaceor interval from the rest of the muscle. See PI. I l l , fig. 12.

In cases in which the nuclei are distributed at intervalsthroughoutthe muscular tissue,as in thelarge elementary fibresof the muscles of the frog, the formation of the contractile mate-rial gradually ceases as the elementary fibre attains its fullsize. When this point has been reached some of the nucleigradually diminish in size, and their original seat is markedby a collection of granules. These granules are sometimesabsorbed, and the seat of the original nucleus is marked by ashort line which gradually tapers at the two extremities untilit is lost.

It is almost needless to say that no alteration produced bythe different contractions of the muscle in different parts,would account for the position of the nuclei observed in thefine fibres of the papillae of the frog's tongue.

These views, it need scarcely be said, differ entirely fromthose generally entertained upon the development and forma-tion of muscular tissue. They are supported by detailed

1(5

observations made in all classes of animals, and in the samespecies at different periods of age. There are some facts inconnection with the changes occurring in disease which affordsupport to this view, which involves three positions. Thatin the nutrition of muscle the pabulum invariably becomes con-verted into germinal matter; that the latter undergoes change,and gradually becomes contractile tissue; and that all the con-tractile material of muscle was once in the state of the materialof which the nuclei or masses of germinal matter are composed.It is not deposited from the blood, nor produced by the actionof the nuclei at a distance, but it results from a change in thevery matter of the nucleus itself. The manner in which thisoccurs has been already discussed in the paper above referredto ('Archives,' No. XIV). It was shown that the ovalnucleus could be followed into a very fine band of contractiletissue or fibrilla (PI. I l l , fig. 18). We pass from the matterof the nucleus into very transparent imperfectly formedtissue in which no transverse lines are perceptible, and fromthis into fully developed contractile material in which thecharacteristic transverse marking3 are fully developed.

Of the Capillaries.

The capillaries of the papilla of the frog's tongue are re-markable for their large size. In the common frog there is acomplete vascular ring at the summit of the papilla, throughwhich the bundle of nerve-fibres distributed to this part pass.In the Hyla the same is observed in some of the papillse, butthe more common arrangement may be described as a halfring or a simple loop, bent upon one side at its upper part(PI. I, fig. 1).

When the large capillaries of the papilla are distendedwith transparent Prussian-blue injection, their walls are seento be of extreme tenuity and transparency. Connected withthe transparent tissue are numerous oval masses of germinalmatter (nuclei), which are separated from one another by veryshort intervals. Some of these masses project slightly fromthe inner surface of the vessel into its interior, but themajority seem to be upon its external surface. Of an ovalform, many of these latter gradually taper into thin fibreswhich are continuous with the tissue of which the vascularwall is constituted. The delicate membrane constituting thevascular wall exhibits longitudinal striaj, which are probablyproduced in its formation, and by its external surface is con-nected with the delicate connective tissue which forms, as itwere, the basis-substance of the papilla, and intervenes

17

between all the important tissues which are found in it.This is proved by the fact that the vessel is moved when thetransparent connective tissue at some distance from it is drawnin a direction from the vessel.

The most interesting point I have observed in connectionwith the anatomy of these vessels is the existence of very finenerve-fibres. These form a lax network around the capillary.I have traced these fine fibres continuous with undoubtednerve-trunks in many instances, and have followed the latterinto the trunks of dark-bordered fibres, from which thebundle in the papilla is derived. A similar arrangement offine nerve-fibres has been demonstrated in connection withother capillary vessels of the frog. These fine nerve-fibresare very distinct in several of my specimens.

I have indeed observed, in my paper published in the'Phil. Trans. ' for 1863, contrary to the statements of mostanatomists, that capillary vessels generally are freely suppliedwith nerves, but the latter and their nuclei have been re-garded as connective-tissue fibres and connective-tissue cor-puscles; I have shown in certain specimens that, of the twofibres resulting from the subdivision of a dark-bordered fibre,one was distributed to the fibres of voluntary muscle, whilethe other ramified over the vessels supplying the muscle(PI. I l l , fig. 15). These facts, it need scarcely be said, are ofgreat importance with reference to determining the structureof the mechanism concerned in nervous action.

I have not succeeded in demonstrating lymphatic vessels inthe papillae of the frog's tongue.

Besides the various nuclei described, there are severalround, oval, and variously shaped bodies, about the size of afrog's blood-corpuscle, which are composed principally ofminute oil-globules and granules. These are not coloured bycarmine. Many contain a small mass of germinal matter(nucleus) in the centre, which is, of course, coloured. Insome of the smaller ones this mass of germinal matter ismuch larger in proportion to the entire " cell." These bodiesresemble many of the fat-cells of the frog, and I think it pro-bable they are of this nature. It is, however, possible thatthese masses may be altered lymph-corpuscles. The Hylffiwhich I examined had been for some time in confinement,and contained very little adipose tissue.

Conclusions.

1. That fine nerve-fibres ramify in the connective tissue ofwhich the simple papillae are composed, and that connected

VOL. IX. NEW SEK. B

with these fine nerve-fibres are oval masses of germinalmatter or nuclei, which, with the fine nerve-fibres themselves,have been usually regarded as " connective-tissue corpuscles."

2. That neither the epithelial cells of the frog's tonguegenerally, nor those covering the simple papillae, are connectedwith nerve-fibres.

3. That the mass consisting of epithelium-like cells uponthe summit of the fungiform papilla is-connected with thenerve-fibres, but it is not an epithelial structure.

4. That the dark-bordered sensitive fibres constituting thebundle of nerves in the axis of the papilla divide near itssummit into numerous very fine branches, with which nucleiare connected. Thus is formed a plexus or network ofexceedingly fine fibres upon the summit of each papilla;from this network numerous fine fibres may be traced intothe special nervous organ, composed of epithelium-like cellsupon the summit, with every one of which nerve-fibres appearto be connected.

5. That the bundle of nerve-fibres distributed to a papillaalways divides into two bundles which pursue opposite direc-tions. The division of the bundle may take place just at thebase of the papilla, or at some distance from it, but it alwaysoccurs.

6. That fine pale nerve-fibres pass from the same trunk ofdark-bordered fibres as that which gives off the bundle ofnerves to the papilla. The fine fibres ramify—

a. Amongst the muscular fibres of the tongue.b. Upon the vessels.c. In the connective-tissue of the tongue generally, and

also in the simple papillae.7. That the fine 'nucleated' nerve-fibres ramify freely

amongst the delicate branching muscular fibres of the papillae,and form plexuses or networks which exhibit no nerve-endsor terminations, nor in any case does a nerve-fibre penetratethrough the sarcolemma or investing tissue of the fibre, orconnect itself with the nuclei of the muscle. As many of themuscular fibres are so very fine and narrow, the distributionof the nerves, and their exact relation to the contractile tissue,can be demonstrated very distinctly in the case of the musclesof the papillae of the frog's tongue.

20

island of Java, in which, while the ball-and-socket appen-dage is present, the fringe is wanting. Some who have seenthis scale have compared it to a lobster-pot, for the top ofthe scale is hollow, the points not meeting and closing as inother scales; this is figured from a slide prepared by Norman,of the City Road. Fig. 6 is from Anthocharis Danai, atropical representative of the English Orange-tip. The ma-jority of the members of this group have a patch of bright-red or orange on the tips of the anterior wings, while theunder side of the posterior wings is beautifully marked withgreen and pearl-white. Fig. 7 is from A. Antevippe, andfig. 8 from A. Evippe, two African members of this group.

Many other examples might be given, but a sufficientnumber have been adduced to show that through the familyof the Pieridce, in the males alone, are found scales of a cer-tain type differing in the different species, but whether suffi-ciently distinct from each other as to make them availablefor determining species seems doubtful. Though, as hasbeen seen, the fringe or tassel-like termination is not presentin all the Pieridce, yet they all appear to have the peculiarball-and-socket-like bulb at the base of the scale.

Among the Polyommati all the continental and tropicalspecies of the " blues " proper, which I have been able toobtain, exhibit some modification of the " battledore " scalesfigured by microscopists. I have figured those found onblues, which, though common on the continent, have butseldom been taken in Britain, and are, therefore, at presentnot classed as British insects. They are P. Dorylas (fig. 9),P. Argus (fig. 10), and P. (Egon (fig. 11). A male of thislast has been taken near Brighton this season by a boy, andtherefore has as great a claim as P. bmticus to be consideredBritish. I have used the term " blues " proper advisedly,for strange though it appear, the " battledore " is found onlyon those males of this group which have blue or bluish wings.It would seem that when, as regards colour, the malesresemble in personal appearance the females, they are withoutwhat I call another male characteristic. As was mentionedbefore, the males of the blues proper are of some shade ofblue, while the females are brown and only dashed or spottedwith that colour. Other species are of brown hue and withoutthe blue scales, and among these the males, as far as my obser-vations are concerned, do not possess any peculiar scale. Mr.J. Watson, of Manchester, who has been working on plumulesand battledores, remarks,"No species,however, the males ofwhich are brown, yields these scales;" and he has foundbattledores on 121 species belonging to this group.

I

21

The Fritillaries, or at least that division of them in whichthe underside of the wings is marked with metallic spots,present tasselled scales of a very decided character, differingessentially from those of the Pieridse, on the one hand, andfrom the Hipparchite, which they somewhat resemble, on theother. Some are of a very long, narrow, and ribbon-likeform, with the tassel at the apex, while others are shorterand broader. Their position on the wings differs from thatof the other groups, for, instead of being placed in rowsbeneath the ordinary scales, they are situated on the nervures,or black veins of the upper surface, and are mingled in somespecies with scales or hairs of a unique shape. Fig. 12 is fromArgynnis Aglaia (dark green fritillary); tig. 13, A Paphia(silver-washed fritillary), on which are found also the Indianclub-like scale or hair (fig. 14). A. Adippe (fig. 15, highbrown fritillary), and figs. 16, 17, and 19, are from foreignfritillaries, in the last of which the scale (fig. 18) is alsofound. It is strange, too, the difference of opacity in thesescales; for while with the exception of the apices, which inall are nearly transparent, figs. 12 and 19 are opaque in thebasal half, fig. 15 in the upper half, figs. 13 and 17 in thewhole of the ribbon, and fig. 16 for four fifths of its length.

While examining the wings and scales of butterflies, andwatching the escape of some hundreds of butterflies andmoths from the pupa cases, I have been struck by severalpoints, some of which are worthy the consideration of micro-scopists. First, the almost endless variety of shape of thescales among the Lepidoptera j some of. the most commonforms are figured in all books on the microscope, but thereis room for plenty of work in this direction. The colour ofscales is not only of almost every shade possible, but whenviewed in situ, if the stage be rotated, as great a changeof colour will be often seen as with polarized objects; somecolours (as, for example, the green in the orange-tip), proveon examination to be yellow and black scales intermingled.Again, when scales are viewed as transparent objects, thecolour in the majority either vanishes or becomes a dullyellowish brown. Hence arises the question, To what isthe colour due? Many circumstances lead to a belief thatthe scales can be inflated, and that the insects possess notonly the power of inflation at will, but also of raising therows of scales. I have often caught and killed butterflies inwhich the rows of scales, instead of lying flat, were at anangle to the wing membrane. From this I have inferredthat the insects possessed the power of rendering them-selves more buoyant. I do not think the " battledores " or

22

" plumules" have much to do with rendering their maleowners more vigorous on the wing, for I have generallyfound the rule to be greater activity and speed among thefemales. The markings, too, are very varied; and thoughmuch has been done in this direction, much more remainsto be worked out. The last point has reference to the wayin which the wiugs expand and the scales are drawn outwhen an insect escapes from the pupa case. Some havethought that each scale expanded in size along with the wing-membrane itself, the air breathed by the newly developedinsect entering between the two laminee, and causing theirenlargement; others, that they depended on the age andvigour of the insect; in fact, being small in the newly deve-loped, and of full size in one of mature age—assuming, whatis not the case, increase in size as the insect becomes older.If a portion of that part of the pupa case which covers thewing be removed a day or two before an insect escapes, it willbe found the scales are all of full size; or if a wing of a newlyescaped insect be taken before the wing has time to expand,it will be seen that the scales are not only of full size, butpacked closely together both laterally and longitudinally, asfig. 20. As the wing-membrane expands, the rows of scalesare drawn further and wider apart, until they present theappearance seen in a fully-developed wing (fig. 21). Thereis, therefore, literally no growth of scales after the insectemerges from the pupa case, but simply a lateral and longi-tudinal opening out, or double telescopic unfolding of thewings. Fig. 22 represents the actual size of the wing of thelarge white P. brassicae when emerging from the pupa case;fig. 23 shows the full-sized wing of the same butterfly. Ananalogous state of things is seen in the leaves of some plantshaving scales or hairs, notably in the case of the Deutziascabra, whose stellate scales are well known. If one leaf ofan unexpanded bud be removed, the stellate scales will befound of full size, but so crowded as to completely cover theleaf; if either a fully-formed leaf be taken, or the leaf oppo-site to that removed be allowed to grow, the scales will befound not to have increased in size, but only to be more widelyseparated by the lateral and longitudinal expansion of theleaf itself.

I