Talk:Paper - Correlated changes in nervous tissues in malformations of the central nervous system (1946)
[ 88 ]
CORRELATED CHANGES IN NERVOUS TISSUES IN MALFORMATIONS OF THE CENTRAL NERVOUS SYSTEM
By A. BRODAL, Anatomical Institute, University of Oslo, Norway
In the present communication some observations made on a rare type of malformation of the cerebellum will be briefly reported, and the inferences which can be drawn from them in regard to the understanding of the factors which determine the occurrence of correlated changes in the nuclei connected with the malformed part will be discussed.
A perusal of only a limited part of the vast literature dealing with malformations of the central nervous system at once reveals extensive differences and variations within the cases belonging to one type or the other, as e.g. microgyria, anencephalia, cerebellar malformation and others. This is true not only in regard to the intensity of the ‘primary’ malformation and the accompanying abnormalities in other organs so frequently met with, but also as concerns the accompanying changes within the central nervous system itself. These usually are termed correlative changes, indicating thus a causal connexion of some sort or other between them and the principal, major malformation.
The study of cases of this type has, especially in earlier days, been undertaken partly to obtain information concerning the fibre connexions between different regions of the central nervous system, partly to ascertain the factors influencing the orderly development of the central nervous system, the malformations representing, as has often been stated, experiments by nature itself to a certain degree. However, it must be admitted that although certain major features concerning the fibre connexions have been recognized from cases of this
. type, the conclusions drawn will always be more or less hypothetical, on account of our still very
fragmentary knowledge of the factors at work in .
normal development.* The value of information concerning normal development is likewise restricted, since secondary changes of different kinds may produce extensive disturbances in the picture. Thus very diverging hypotheses have been set forth on this point by different authors.
Experimental and comparative anatomical research has brought forward a considerable mass of evidence, elucidating the factors which are at work in the morphogenesis of the nervous system. Necessarily, however, the experimental studies have for the most part been performed on lower verte
- This has been more fully elaborated in regard to the
cerebellum in a previous paper (Brodal, 1940, pp. 127 ff.).
brates, and on systems connecting the central nervous structures with the peripheral organs. In regard to the connexions within the central nervous system, we are thus mainly forced to content ourselves with conclusions of analogy. In single instances only the situation is more favourable, special circumstances allowing more exact conclusions to be drawn directly from findings in higher vertebrates. An opportunity of this sort was afforded by the material forming the basis of the present communication, a strain of mice affected with hereditary hydrocephalus with a partial maldevelopment of the cerebellar vermis.
In his extensive study of cerebellar malformations Brun (1917-18) distinguishes two types of correlative changes in cerebellar malformations, especially of the neo-cerebellum. One type he calls primary correlative checking of development (‘ primarkorrelative Entwicklungshemmung’). In this instance it is assumed that the same noxious factor, which arrests the development of, e.g. the neocerebellum, will produce a simultaneous checking of development of those ‘subcerebellar’ structures which develop previous to or at the same time as the neo-cerebellum. These structures then will be present in such cases, but they will remain at an early foetal stage, eventually, however, presenting further histo-tectonic differentiation.
The second type of correlative changes is named by Brun secondary correlative checking of development (‘sekundar-korrelative Entwicklungshemmung’). Here the changes in the ‘subcerebellar’ structures cannot be explained as a consequence of the same noxious factor which has affected the cerebellum, but they must be conceived of as being sécondary to it, produced through a repression of the ‘morphogenic stimulation’ normally exercised by the neo-cerebellum on the structures which are dependent on it (v. Monakow’s ‘Kleinhirnanteile’). In this case the effect will be limited to structures developing later than the neo-cerebellum, and the histo-tectonic differentiation will fail to take place. As such structures Brun mentions especially the pontine nuclei. /
It will appear from his nomenclature that Brun is inclined to explain all malformations as being due to a checking of development. Furthermore, it appears that insufficient emphasis is placed on the possibility of hereditary factors as responsible for the ultimate manifestation of the maldevelopment, Correlated changes in nervous tissues
and instead presumable intoxications of the germ or of the foetus are over-emphasized as causal factors. On these points modern conceptions are not in full accord with Brun’s rather schematic views, but as this problem falls beyond the scope of the present paper, it will not be treated more fully.
However, Brun has stressed the importance of the time factor in developmental disturbances, especially in the cerebellum. For an understanding of the correlative changes in such instances this is obviously of great value. Although such correlative changes are the rule in cerebellar malformations, other cases are met with where the ‘subcerebellar’ structures, i.e. the nuclei sending fibres to or receiving fibres from the cerebellum, are quite intact, although parts of the cerebellum are severely malformed or even lacking. For example, Vogt and Astwazaturow’s Case I (1912) presented extensive changes in the hemispheres of the cerebellum and anomalies in the vermis, but in spite of this there were no changes in the inferior olive, the brachium conjunctivum or the nucleus ruber, but the pontine grey was devoid of cells. In Verhaart’s case (1933) there was a cyst and changed cortical structure in the left hemisphere, but no changes in the nuclei connected with the cerebellum. Obersteiner (1914) describes a brain where most of the vermis as well as the nuclei fastigii were lacking, the two cerebellar hemispheres fusing directly in the midline. Obersteiner states specifically that there was no reduction of the afferent or efferent cerebellar tracts and, although it is not specially mentioned, it is not likely that there had been changes in the nuclei in question.
This lack of correspondence between cerebellar and subcerebellar lesions is usually explained by assuming a certain degree of ‘self-differentiation’ of the nerve cells. It is imagined that the outgrowing neurites which are prevented from reaching their normal ending place will nevertheless have the capacity of further development, and eventually will take an aberrant course, to end in other regions, preferably in other parts of the same nuclear complex. This assumption appears rather probable, but obviously a process of this type can occur only on the supposition that the nucleus of origin is developed at a time when the major changes are already established in the nucleus or grey matter which is normally the destination of the fibres. If the outgrowing fibres have already reached their ending place or if they have no other way to go, the result will presumably be a secondary affection of the nucleus of origin. (Some observations, however, tend to show that smaller structural aberrations, e.g. in the cerebellar cortex, do not interfere with the ingrowth of neurites, e.g. a case described by Miskolezy (1932).)
As far as can be seen from the literature, observations which clearly support the above-mentioned
89
assumption have not been recorded. The study of the foetal stages of the mouse strain with hereditary hydrocephalus, mentioned above, has, however, made possible some conclusions concerning this topic. Below, therefore, a brief survey of the findings in adult and foetal hydrocephalic mice will be given, followed by a discussion of the findings in regard to the problems in question.
The mouse material consists of a strain of the albino house mouse, affected with hereditary hydrocephalus, which has been studied by Professor Kristine Bonnevie, Zoological Laboratory, University of Oslo. The anomaly has been shown to be inherited as a monohybrid recessive, manifesting itself in a hydrocephalus of varying degree (Bonnevie, 1943).
In studies of foetal stages of these mice, Bonnevie (1944) has shown that the first signs of developmental disturbances have occurred already in the blastocyst at or before its implantation in the uterus, manifested first as anomalies in the foetal coverings. Later on the disturbances are transmitted also to the embryo itself, ie. as an abnormal imbibition of extra-embryonic fluid. The hydrocephalus makes its appearance first on the 12th day of foetal life (Bonnevie, 1943, 1944). In the embryos, as in the adult mice, the degree of hydrocephalus varies from the most extreme dilatation of all the cerebral ventricles to con . ditions just beyond the normal. The hydrocephalus is
constantly accompanied by a splitting of the cerebellum into two more or less complete halves (Fig. 4 (1944)).*
The cerebellar anomalies were made the object of a special study (Brodal, Bonnevie & Harkmark, 1944), and likewise the development of the cerebellar changes during foetal life were studied in embryos from different stages (Bonnevie & Brodal, in the Press).
It was found that, in the least affected adult specimens, the splitting was confined to the caudalmost part of the vermis, e.g. in one cerebellum, shown in Fig. 5 (1944), the lobulus 5, the uvula, was present as two halves entirely separated from each other; the lobulus a, the nodulus, was nearly completely subdivided in two halves, which were only connected rostrally; the most superficial folia of the anterior lobe (lobuli 3 and 4) did not meet in the midline, otherwise the two cerebellar hemispheres were interconnected almost as usual; lobulus c, the tuber and pyramis, however, was lacking except for its most caudal folium which was present as two bilateral halves.
In general, the same features were characteristic also of the more severely affected specimens. In some (exemplified by the drawings in Fig. 8 (1944)), the vermis was nearly completely split in two halves, which were only connected at the middle levels by a narrow bridge of cerebellar substance. Also in cerebella affected to this degree the only
- For illustrations the reader is referred to the original
papers. In the present paper each figure to which reference is made, is indicated by its number in the original paper and its year of publication (1944 referring to the article by Brodal, Bonnevie & Harkmark, 1945 to that by Brodal).. 7-2 90
lobule lacking was the lobulus c, with the exception of its caudalmost folium. The hemispheres, the paraflocculi and the flocculi were normal in all these specimens.
At those places where the two cerebellar halves were separated from each other, traces were found of a tiny membrane connecting them. In many of the adult cerebella this membrane had been broken, but from the study of the foetal cerebella it appeared clear that the two halves had originally been connected by a membrane, covered with flattened ependymal cells on the ventricular surface, by the pia mater on the outer surface, and containing as a rule some nerve fibres.
Apart from minor aberrations in the structure of the cerebellar cortex in the vicinity of the median split, some small heterotopias in a few of the cerebella and a probable reduction of the nucleus medialis of the cerebellum, no other anomalies were ascertained. It is to be noted that no abnormalities could be-detected in the inferior olive, the pontine nuclei, the nucleus cuneatus externus, the nucleus reticularis lateralis and the vestibular. nuclei, in spite of the lack of the largest part of the lobulus c, a constant feature in all the investigated hydrocephalic cerebella.
In only one of the microscopically investigated cerebella, which was very severely affected, were there any changes present in the nuclei sending their fibres to the cerebellum. In this specimen (from which the drawings in Fig. 10 (1944) were made), the splitting was complete and the two halves drawn widely apart. Furthermore, the caudal part of the left hemisphere was greatly reduced in size, and part of the nervous substance was replaced by connective tissue. At some places groups of macrophages, partly laden with blood pigment, were present in the connective tissue, and in the cerebellar substance itself other groups of non-pigmented, presumably glial, macrophages were met with. In this specimen there was a considerable reduction of the cell mass of the external cuneate nucleus and the nucleus reticularis on the left side (homolateral to the most affected half of the cerebellum) and in the pontine nuclei, especially the peduncular and lateral grey, and in the inferior olive on the right side. In the inferior olive the medial accessory olive was most reduced, but also the principal olive, especially its caudal parts, and the dorsal accessory olive was poorer in cells than normal. .
From the study of the adult cerebella it was concluded that the cerebellar anomalies were a consequence of the hydrocephalus, and this was confirmed by the findings in the foetal material. The hydrocephalus, which sets in on the 12th day of foetal life, will produce a stretching of the cerebellar plate. As this is normally thinnest in its median region, it is to be expected that the attenuation will mainly affect this region, especially at the caudal levels. The attenuation may in extreme cases proceed so far that only a thin membrane is left of the cerebellar plate in the midline, whereas in the lesser affected brains the rostral part of the cerebellar plate will develop nearly normally, thus resulting in a practically normal anterior lobe,
A. BropaL
combined with a more or less complete splitting of the more posterior parts of the vermis. The quantitative reduction of the cerebellar substance, especially the partial lack of the lobulus c, is explained according to the hypothesis of Dow (1940). Dow assumes that a process disturbing the regular development of the cerebellar commissures (commissura lateralis in the flocculo-nodular lobe and commissura cerebelli in the corpus cerebelli) . will interfere with the proper migration of neuroblasts from the ‘anlagen’, and then will affect preponderantly the material destined for the formation of the vermis, whereas the hemispheres will be less affected or entirely normal. As the anterior part of the vermis is the first to develop, an explanation is also obtained of the fact that in all cases of partial lack of the vermis so far reported in the literature the vermis of the anterior lobe is always best preserved.*
Astonishingly similar findings as in these hydrocephalic mice were made in a human case (Brodal, 1945). The patient was a girl, 10 years old. From her first year of life she had presented an enlarged head, which later on continued to grow out of proportion. Clinically and by ventriculography the diagnosis of hydrocephalus was made. In addition to the enlarged head the child presented other symptoms, i.e. reduced vision, an unsteady gait and reduced mental powers. Death occurred some days after a surgical intervention for loosening adhesions in the posterior cranial fossa. At the operation it was disclosed that the caudal part of the cerebellar vermis was lacking.
Macroscopical and microscopical examination postmortem showed that the cerebellar hemispheres and the vermis of the anterior lobe were nearly normal, only the right tonsil was abnormal in its position and folial pattern. The declive also was almost normal, but the folia caudal to the declive appeared to be lacking. However, on a closer analysis, rudimentary lobules were found which could be identified as the tuber, the pyramis, the uvula and the nodulus. But most of these folia were present as bilateral structures (cf. Figs. 4-12 (1945)). The rudimentary folia representing the tuber, pyramis and uvula were attached to a membrane (v.m.d. in Fig. 10 (1945)), which on the ventricular surface was covered with ependyma, on the outer with pia mater and which contained some myelinated fibres. The membrane was attached to the central cerebellar mass on each side, and extended rather far dorsally as a sack-like bulge, but had apparently been damaged at its extreme rostral point during the operation. A velum
- Three cases described by Brun (19186, Cases VI, VII
and IX), however, are different in some respects, in so far as in these instances a cleft was found only in the vermis of the anterior lobe. But in all cases there were also rather prominent other anomalies as well in the cerebellum as in other parts of the nervous system (e.g. spina bifida in two instances), which makes it probable that Brun’s cases belong to another type of malformation than those treated in this paper. Correlated changes in nervous tissues
medullare posterius was also present, continuous both with the flocculi, which presented smaller aberrations in their finer structure, and with the rudimentary folia representing the nodulus, likewise present as small bilateral structures. The nucleus fastigii on the right side was smaller than normal and the nuclei dentati showed slighter structural anomalies.
In this brain also no aberrations from the normal picture in the inferior olive, the pontine grey or other nuclei sending their fibres to the cerebellum could be detected.
An analysis of the findings in this human case thus disclosed striking similarities with the anomalies in the hydrocephalic mouse strain, and it was concluded that the mechanism which was responsible for its origin was of the same nature in both instances, viz. an early developing hydrocephalus, which had impeded the migration of neuroblasts to the future vermis region through attenuation of the cerebellar plate.
The finding in these cases, which is of special interest for the problem to be discussed in the present communication, is the intactness of the nuclei sending their fibres to the cerebellum. Apart from one very severely affected mouse brain, the nuclei in question were intact, in spite of the fact that parts of the cerebellum were always lacking, or clearly underdeveloped in the case of the human brain. Now it must be admitted that a minor reduction of, e.g. the pontine nuclei, the external cuneate nucleus or the nucleus reticularis lateralis might have escaped attention, since it is clearly very difficult to verify a slight reduction in these systems where there exists no clear-cut localization within the projection (cf. Brodal & Jansen, 1943; Brodal, 1941, 1943). In regard to the inferior olive, however, the situation is different. It has long been known from observations made on human material, that the olivo-cerebellar projection system is clearly organized on a localization principle (Henschen, 1907; Holmes & Stewart, 1908; Brouwer & Coenen, 1919; Koster, 1926). By experimental investigations on rabbits and cats by means of a modified Gudden method, it has been possible to ascertain the details of this localization (Brodal, 1940). Each lobule of the cerebellum has its own projection area in the complex of the inferior olive, and also the intra-cerebellar nuclei have their distinct projection areas. The lobulus c, which especially concerns us here, receives its fibres in the rabbit from a wellcircumscribed area (about one-quarter) of the medial accessory olive, but in the cat from a considerably larger area in correspondence with the more developed lobulus c in this animal. It can scarcely be imagined that a reduction of such a large portion of the medial accessory olive will be overlooked, especially when particular attention is paid to it.
From the observations made in the adult mice it was concluded (Brodal et al. 1944) that the intactness of the ‘subcerebellar’ structures, especially of
91
the inferior olive, was probably due to the early development of the cerebellar anomalies, these setting in before the neurites from the inferior olive had reached the cerebellum. The neurites destined for the lacking parts were probably directed to intact regions of the cerebellum. The analysis of the brains of the foetal mice (Bonnevie & Brodal, in the Press) lent active support to this view. A brief survey of the principal findings will elucidate this.
The hydrocephalus, as previously mentioned, makes its appearance on the 12th day of foetal life. On'the 11-12th day the cerebellar plate is still undifferentiated into folia. Ts is considerably thinner in its median part, especially caudally, than in the major lateral parts and there are no signs of fissures or of any cortex. A thin superficial marginal veil, practically free of cells, covers the intermediate mantle layer which, without clear limits, passes into the matrix on the ventricular side of the plate. In the brain stem most of the nuclei of the cranial nerves can be recognized, but there is no inferior olive and no pontine nuclei, and the corpus ponto-bulbare, which provides the material for the pontine nuclei and the inferior olive (Essick, 1912) has just begun its mitotic activity.
The first clear-cut indication of vigorous cellular proliferation in the ponto-bulbar body is encountered at the 12th day stage. Approximately on the 13th day, the olivary migration along the lateral border of the medulla oblongata makes its appearance, and on the 13-14th day the cells from this migration are found to accumulate at the site of the future inferior olive. Not until approximately the 17th day are the separate subdivisions of the olive recognized, but at this time many of the cells still appear immature. The olivary migration subsides about the 17-18th day.
The pontine migration, and the formation of the pontine nuclei, which begins about the 14th day lags a little behind the olivary migration and the differentiation of the inferior olive, although the delay is not nearly so pronounced as in man.*
The first indication of lobulation in the cerebellum of the normal mouse is encountered on the 17th day. Then the fissura postero-lateralis of Larsell (Larsell & Dow, 1935) makes its appearance, closely followed by the fissura prima. The cortical differentiation begins on the 14th day as a clear-cut embryonic granular layer separated from an inner granular layer by a fibre-rich narrow zone, and still at the 18th day stage the Purkinje cells are not fully developed. The separate lobules of the vermis can be identified from the 18th day stage, but at birth the lobulation of the cerebellum and likewise the cortical differentiation are not yet completed.
In the abnormal foetal mice the hydrocephalic dilatation with the ensuing attenuation of the cerebellar plate, beginning on the 12th day, leads to a more or less severe
- This can be explained partly as a consequence of the
much more rapid development of the brain in the mouse, partly owing to the greater development and complexity of this nuclear mass in man as compared with the mouse. 92
reduction of the median part of the plate. Finally, the two halves are connected partially or in their entire rostrocaudal extension by a thin membrane only. The ultimate degree of the cerebellar changes, however, as mentioned previously, is highly variable. The changes taking place in the cerebellar plate, however, clearly do not interfere with the processes leading to the formation of the pontine nuclei and the inferior olive. At all stages these, like the ponto-bulbar body and the migrations arising from it, develop exactly as in the normal specimens, and finally they present a completely normal picture in the adult hydrocephalic mice. The process affecting the cerebellar plate appears to be entirely mechanical.
It thus appears that in the abnormal mice the hydrocephalus, and the attenuation of the cerebellar plate which is the consequence of. it, sets in earlier than the full activity of the corpus pontobulbare. The subdivisions of the inferior olive are differentiated first at the 17th day, approximately 5 days after the beginning of the damage to the cerebellar plate, and at the time when the first signs of lobulation appear, only one day before the subdivision of the vermis into its different lobules is clearly recognized. Unfortunately, the sections at my disposal were not fibre stained.* However, judging from the findings in the sections employed (van Gieson stain, fixed in Bouin’s fluid) it appears safe to conclude that the fibres from the inferior olive cannot have reached the cerebellum during the lapse of one day, the interval between the recognition of the separate subdivisions of the olive and the separation of the vermis into its lobules. Of course, it cannot be strictly denied that some of the cells of the olive at this time might have rather well-developed neurites, but certainly a half of them have not, as many cells in the olive are still distinctly undifferentiated. The same may be said of the accessory olives and the pontine nuclei.
Assuming now that at least the major portion of the fibres from the olive have not reached the future vermis region at the 18th day, when the cerebellar changes are clearly established, it is likely that the fibres destined for this part of the cerebellum, especially for the constantly lacking major part of the lobulus c, will be directed to other parts of the cerebellar cortex. They will find an aberrant ending place in the cerebellar cortex, and at this point establish qualitative normal connexions. In this manner no changes will appear in the inferior olive. Similar considerations can obviously be applied to the pons and the ponto-cerebellar fibres.
The assumption was made above that the proliferating cerebellar cortex will attract the incoming cerebellopetal fibres. Without embarking on a detailed discussion of the factors concerned in the morphogenesis of the brain some data of special
- The mouse strain died out before the microscopical
study was begun.
A. BrRoDAL
interest will be briefly mentioned.* Thus it is learned from the experimental investigations made especially on amphibia by Coghill (1924), Detwiler (1928) and Burr (1932) that proliferating centres have the capacity of attracting ingrowing fibres. That the reverse also holds good, ingrowing fibres having the capacity of stimulating the activity of a proliferating centre may also be mentioned (Burr, 1920; May & Detwiler, 1925).
However, it is obvious that these factors alone cannot explain the orderly morphogenesis of the nervous system. The investigations of Coghill, Detwiler and others clearly demonstrate that there are other forces at work. Obviously the proliferation of the cellular masses is to a great extent determined by ‘intrinsic dynamic forces’, of whatever sort they may be, especially early in development. Thus as concerns the medulla spinalis of amphibia Coghill (1933, p. 345) states: ‘There is, therefore, an intrinsic dynamic substratum of progressive development in the region irrespective of peripheral, afferent nerves; and there is evidence... that the intrinsic dynamic substratum persists after the establishment of the afferent nerve roots.’ That the development of the inferior olive takes its usual course in the hydrocephalic mice in spite of the considerable changes taking place simultaneously in the cerebellum is a fact confirming this conception. In this instance conditions are also especially favourable for such a judgement, as the factor responsible for the cerebellar malformation is clearly purely mechanical, produced by the hydrocephalus, and can scarcely be thought to affect the ponto-bulbar body to any considerable extent.
It will be recalled that in one case only (Fig. 10, 1944) were clear-cut changes found in the pons as well as in the inferior olive, preponderantly in the parts corresponding to the lacking parts of the cerebellum. This, however, probably does not make untenable the hypothesis set forth above. It cannot be denied that the severe change in one cerebellar hemisphere in this case had taken place later and was associated with haemorrhage and secondary ingrowth of mesenchyme (cf. Brodal ef al. 1944). On the other hand, the affection has from the beginning been severe, and the reduction of the one hemisphere can be dated back to an early foetal stage. But when in this brain the fibres from the normally developing inferior olive have reached the cerebellum, it can be assumed that the cortical areas left have not been extensive enough to receive them all. On this account some of them have fallen the victim of atrophy, which in the foetal brain is followed by a total disintegration with a rapid clearing up by the neuroglia. That the damage to the olive is found preponderantly in the regions corresponding to the lacking areas is in accordance with the considerations made previously.
- Excellent reviews of these problems have been published by Detwiler (1926) and Kappers (1934).
Correlated changes in nervous tissues
A mechanism similar to that in the hydrocephalic mice will also reasonably explain the lack of changes in the inferior olive and the pons in the human case briefly reported above. In the analysis of the findings in that case (Brodal, 1945) it was regarded as probable from a comparison with the normal development of the human cerebellum as described by Hochstetter (1929) that the hydrocephalus had set in when the embryo had a length of approximately 48mm. According to Essick (1912) the first cells at the site of the future inferior olive in man can be recognized before the 30 mm. stage. But not until approximately the 143 mm. stage is the pattern of the olive established with its typical subdivisions. The characteristic foldings, however, are not yet visible. Thus also in this instance the hydrocephalus must .be assumed to have arisen considerably earlier than the full development of the inferior olive and still longer prior to the pontine nuclei. When in some of the previously published cases of the same type (defective development of the vermis) changes were observed in the inferior olive, e.g. in the cases of Lyssenkow (1931) and Castrillén (1932), and Dow (1940), they might be due to a comparatively later setting in of the hydrocephalus. Besides, in Lyssenkow’s and Dow’s cases the hemispheres also were partially affected.
The chief interest in the cerebellar malformations discussed in this paper is the lack of correlative changes. It appears safe to state that in the case of the hydrocephalic mice it has been possible to prove that this lack can be explained by the very early development of the primary cerebellar malformation, some time in advance of the full development of the nuclei in connexion with it. It appears plausible that similar conditions may be responsible for the lack of the usual correlative changes now and then observed also in other types of cerebral malformations.
93
SUMMARY AND CONCLUSIONS
The findings in the brains of a strain of mice affected with a hereditary hydrocephalus are reported. The most prominent change apart from the hydrocephalus was a partial or complete splitting of the cerebellar vermis with lack of the lobulus ec, except its caudalmost folium. The nuclei sending fibres to the cerebellum were intact. A closely similar human case is also briefly described. In both instances there is reason to assume that the cerebellar anomaly has been produced by an attenuation of the cerebellar plate caused by the hydrocephalus.
From an analysis of normal and hydrocephalic mouse embryos it appears that the hydrocephalus sets in on the 12th day of foetal life, at a stage when the activity of the ponto-bulbar body is just beginning. The separate subdivisions of the inferior olive cannot be recognized until 5 days later, and even at this stage many of its cells are not fully differentiated. The lobulation of the vermis can be recognized approximately one day later.
Based on these findings, the hypothesis is set forth that the lack of correlative changes in the inferior olive (and the pons) can be explained in the following manner: At the time when the neurites from the cells of the olive reach the cerebellum, the malformation is already fully developed (as abnormal embryos from this stage also show). The ingrowing neurites which do not find their normal ending place are directed instead to the remaining intact regions, thus. preventing the occurrence of atrophy and changes in the inferior olive, pons and other nuclei.
A mechanism of the type described in this material might give a valid explanation in other instances of malformation, where the usual correlative changes are lacking.
REFERENCES
Bonnevisg, K. (1943). Norske Vid.-Akad. Skr. I. Mat.-Nat. KI. no. 4.
Bonnevig, K. (1944). Norske Vid.-Akad. Skr. I. Mat.-Nat. K1. no. 10.
Bonneviz, K. & Bropat, A. (1946). Norske Vid.-Akad. Skr. I, Mat.-Nat. K1. (in the Press).
Bropat, A. (1940). Z. Neurol. 169, 1.
Bropat, A. (1941). Z. Neurol. 171, 167.
Bropat, A. (1943). Acta Psychiat. Neurol. Kbh. 18, 171.
Bropat, A. (1945). Norske Vid.-Akad. Skr. I, Mat.-Nat. K1. no. 3.
Bropat, A., Bonnrevig, K. & Harxmarxk, W. (1944). Norske Vid.-Akad. Skr. 1. Mat.-Nat. K1. no. 8.
Bropat, A. & JANSEN, J. (1943). Norske Vid.-Akad. Skr. I. Mat.-Nat. K1. no. 10.
Brouwer, B. & Cornsrn, L. (1919). J. Psychol. Neurol., LDpz., 25, 52.
Brun, R. (1917). Schweiz. Arch. Neurol. 1, 51.
Bron, R. (1918a). Schweiz. Arch. Neurol. 2, 48.
Bron, R. (19180). Schweiz. Arch. Neurol. 3, 13.
Bure, H. 8S. (1920). J. exp. Zool. 30, 159.
Burr, H. S. (1932). J. comp. Neurol. 56, 347.
CastRritioén, H. A. (1932). Z. Neurol. 144, 113.
Coa@ui1t, G. E, (1924). J. comp. Neurol. 87, 71.
Coauit, G. E. (1933). J. comp. Neurol. 57, 327. Detwier, S. R. (1924). J. comp. Neurol. 37, 1. Dertwiier, S. R. (1926). Quart. Rev. Biol. 1, 61. Dertwiep, 8. R. (1928). J. exp. Zool. 51, 1. Dow, R. 8. (1940). J. comp. Neurol. 72, 569. Esstox, C. R. (1912). Amer. J. Anat. 18, 25. Henscuen, F. Jr. (1907). Z. klin. Med. 63, 115. HocustettEr, F. (1929). Beitrdge zur Hniwicklungsgeschichte des menschlichen Gehirns. Leipzig u. Wien: Deuticke. . Hoss, G. & Stewart, T. G. (1908). Brain, 31, 125. Kappgrs, C. U. A. (1934). Irish J. med. Sci. Sept. Kostsp, 8S. (1926). Acta Psychiat. Neurol. Kbh. 1, 47. LaRSELL, O. (1934). Arch. Neurol. Psychiat., Chicago, 31, 373. LaRsELL, O. (1937). Arch. Neurol. Psychiat., Chicago, 38, 580. LarsELL, O. & Dow, R. S. (1935). J. comp. Neurol. 62, 443. Lyssznxow, N. K. (1931). Virchows Arch. 280, 611. _ May, R. M. & Detwier, S. R. (1925). J. exp. Zool. 43, 83. Misxotczy, D. (1932). Arch. Psychiat. Nervenkr. 93, 596. OBERSTEINER, H. (1914). Arb. Neurol. Inst. Wiener. Univ. 21, 124. VERHAART, W. J.C. (1933). Acta psychiat. Neurol. Kbh. 8,691. Voet, H. & AstwazaTuRow, M. (1912). Arch. Psychiat. Nervenkr. 51, 75.
