Journal of Neurology, Neurosurgery, and Psychiatry, 1974, 37, 1022-1030

Relationship between visual field defect and arterialocclusion in the posterior cerebral circulation

S. N. KAUL, G. H. DU BOULAY, B. E. KENDALL,AND R. W. ROSS RUSSELL

From the National Hospitals for Nervous Diseases, Londonz

SYNOPSIS (1) The extent and severity of visual field loss has been compared in a series of 14patients with occlusions of the posterior cerebral artery or its branches, all verified angiographically.Atheroma, embolism, and migraine were the commonest types of underlying vascular disease.(2) Occlusion of the main trunk of the artery was associated with severe and permanent field lossusually with some sparing of the central area and, in one case, of some field adjacent to the verticalmeridian. It is suggested that this is due to collateral blood flow reaching the margins of the posteriorcerebral territory from the adjacent middle cerebral territory via pial anastomoses. (3) Single or

multiple occlusions of the main branches of the posterior cerebral artery gave variable amounts offield loss with considerable recovery in some cases. Collateral blood flow from the middle cerebralterritory and from other branches of the posterior cerebral artery was demonstrated and the variationmay be due to the size and number of pial anastomoses and to systemic factors such as bloodpressure and vascular reactivity. (4) Occlusion of smaller branches of the calcarine artery producedlocalized zones of capillary underperfusion near the posterior cerebral pole. These corresponded toscotomatous paracentral defects in the visual field which were often permanent and showed no

central sparing. The potential capacity of the collateral system may be limited by occlusion of intra-cerebral arteries or by involvement of segments of the pial arteries in the disease process.

The detailed clinicopathological studies ofHolmes (1931) and Henschen (1926) in patientswith defect of the visual fields after trauma allowus to delineate lesions of the occipital lobes withconsiderable accuracy, although it may not bepossible to distinguish a lesion of the calcarinecortex from one involving the optic radiation.More recently, improved methods of cerebral

angiography have been applied to patients withvascular lesions in the circulation of the posteriorcerebral artery and the introduction of selectivecatheter techniques combined with subtraction,colour subtraction, and radiographic magnifica-tion have enabled occlusion of smaller branchesand regions of delayed or diminished capillaryperfusion to be identified (Hoyt and Newton,1970).The relevant anatomy of the posterior cerebral

artery and its branches may be summarized asfollows: The posterior cerebral artery originatesfrom the terminal bifurcation of the basilar

1022

artery; its circum-mesencephalic portion passesaround the midbrain immediately below thetentorium cerebelli and becomes the internaloccipital segment as it crosses the free edge of thetentorial notch to lie on the medial aspect of theoccipital lobe. It divides into terminal parieto-occipital and calcarine branches at a variablepoint along its course. The named branches arethe thalamo-perforate, posterior communicating,medial and lateral posterior choroidal, splenialand anterior, middle, and posterior temporalarteries.

Passing from its origin in the interpeduncularcistern around the cerebral peduncle and throughthe crural cistern, the circum-mesencephalic seg-ment curves backward with a downward con-vexity around the midbrain. In the ambientcistern it lies medial to the hippocampal gyrus,above and slightly lateral to the superior cere-bellar artery. As it reaches the quadrigeminalcistern it courses medially and upwards through

guest. Protected by copyright.

on October 15, 2020 by

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.37.9.1022 on 1 Septem

ber 1974. Dow

nloaded from

Visual field defect and arterial occlusion in the posterior cerebral circulation

the tentorial notch towards the calcarine fissurebelow the splenium of the corpus callosum andmedial to the anterosuperior portion of thelingual gyrus. It may bifurcate into its terminalbranches at any point beyond the crural cistern,often near its passage supratentorially.

In 80% of cases the posterior temporal arteryarises as a single trunk, usually from the ambientsegment of the posterior cerebral artery, occas-

ionally from the quadrigeminal segment. Inabout 20%, however, there are two or more

trunks of origin (Margolis et al., 1971). Theposterior temporal artery or arteries then course

FIG. 1. J.P., a patient aged 64 years who wakened onemorning with a headache and generally impairedvision. After four days the visual defect had becomeconfined to the left field and proved to be a homony-mous hemianopia sparing the macula. He sufferedfrom severe hypertension and generalized arterialdisease. The hemianopia improved and this vertebralangiogram, performed three months after the onset ofthe disturbance, shows almost complete occlusion ofthe right posterior cerebral artery about J10 cm fromits origin.

posteriorly, laterally, and slightly downwardalong the parahippocampal gyrus to the lingualgyrus, anastomosing with the posterior temporalbranch of the middle cerebral artery laterally andinferior branches of the calcarine artery pos-teriorly.The parieto-occipital artery, the larger of the

two terminal branches of the posterior cerebralartery, arises from its internal occipital segmentin 4000 of cases (Margolis et al., 1971). In theremainder the bifurcation moves proximally andthere is then no internal occipital segment. Whenit occurs the internal occipital artery passesposteriorly and upward over the cingulate gyrus,terminating in branches to the precuneus,parieto-occipital fissure, and cuneus. It anasto-moses with the pericallosal branch of the anteriorcerebral artery and with the calcarine artery.The calcarine artery arises as one or two

trunks, usually, but not invariably, on thelateral side of the parieto-occipital branch. Thecalcarine artery or arteries then cross the parieto-occipital on a posteromedial course to enter thecalcarine fissure. It is no longer possible to sup-port the contention of Shellshear (1927) that thecalcarine artery exclusively supplies that area ofbrain which functions as the primary visualreceptive field. In fact, that striate area itselfoften receives a contribution from the posteriortemporal or parieto-occipital branches (Abbie,1938; Smith and Richardson, 1966).

In patients with vascular occlusion, it is ofinterest to relate the site of blockage in thearterial tree to the type of visual field defectbecause it is possible to deduce the extent of theinfarcted region of brain and the adequacy ofcollateral blood supply as well as other circula-tory factors influencing survival of focal areas ofcerebral ischaemia.

These cases necessarily form a highly selectedgroup since not all cases of vascular hemianopiaare subject to angiography. For instance, manypatients who sustain only temporary visual fielddefects may not be included since they may not bereferred to hospital. Patients of advanced ageand those with widespread vascular or cardiacdisease or with severe hypertension are alsounder-represented, since angiography carries a

higher morbidity in this group. Similarly, incases of embolism where there is a cardiac sourcefor embolus, angiography may not be justified.

1023

guest. Protected by copyright.

on October 15, 2020 by

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.37.9.1022 on 1 Septem

ber 1974. Dow

nloaded from

S. N. Kaul, G. H. Du Boulay, B. E. Kendall, and R. W. Ross Russell

FIG. 2. P. W., a woman aged42 years with a history suggesting episodic vertebrobasilar ischaemia who developeda transient right hemianopia which later returned and became permanent. There was central sparing. (a) Thevertebral angiogram, A.P. view, shows occlusion of the calcarine artery and narrowing of the parieto-occipital,perhaps because it has recanalized. (b) The carotid angiogram, lateral view, shows collateral flow from theposterior temporal branches of the middle cerebral artery into the calcarine region.

A number of additional patients were exam-ined; they had suffered an ischaemic episode andhad a residual field defect but had no vascularocclusion on angiography; these patients havenot been included in this report.

RESULTS

OCCLUSION OF MAIN TRUNK OF POSTERIORCEREBRAL ARTERY (Table 1) There were fourpatients who had occlusion of the main trunk ofthe vessel, usually at a point one to two centi-metres from the origin (Fig. 1). Three of thesehad evidence of severe atheroma with diabetes or

hypertension. The fourth patient had a history ofrecurrent migraine.

In two patients the hemianopia was dense andincluded the whole of the half-field with about10° of central sparing. In the other two patientshemianopia was incomplete, with sparing of thecentral area and a portion of field adjacent to the

vertical meridian (Fig. 1). All these patientscomplained of headache at the onset. In twopatients there was evidence of mild motor orsensory impairment on the affected side; onehad spontaneous pain. Two patients with occlu-sion of the left posterior cerebral artery showedtransient amnesia and mild dysphasia. The routeof collateral blood supply was seen in one case tobe from the posterior temporal branch of themiddle cerebral artery.

There was a single case of bilateral posteriorcerebral occlusion; he had evidence of additionalupper brain-stem damage and died rapidly.

MAIN BRANCH OCCLUSION (Table 2) Patientshaving occlusion of one or more of the threeprincipal branches of the posterior cerebralartery included two with occlusion of the cal-carine artery. In both the aetiology was atheromaand one had previous attacks of vertebrobasilarischaemia. In both cases there was a dense and

1024

guest. Protected by copyright.

on October 15, 2020 by

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.37.9.1022 on 1 Septem

ber 1974. Dow

nloaded from

Visual field defect and arterial occlusion in the posterior cerebral circulation

TABLE 1OCCLUSION OF MAIN TRUNK POSTERIOR CEREBRAL ARTERY

Sex, age Aetiology Associated Field defect Outcoine(yr) features

Unilateral1 M 38 Atheroma

DiabetesEmbolism

2 M 64 AtheromaHypertension

3 F 42 Migraine Migraine continuesHemianaesthesiaSpontaneous painTransient amnesiaDysphasia

4 M 62 Atheroma HemiparesisHypertension Dyspraxia

Transient amnesiaDysphasia

Bilateral5 M 55 Atheroma

EmbolismQuadriplegia

permanent loss of a complete half field but withcentral sparing of about 10° and normal visualacuity. In one of these patients carotid arterio-graphy showed collateral flow into the posteriorcerebral territory from the posterior temporalbranch of the middle cerebral artery (Fig. 2a, b).There were three examples of multiple branch

occlusion with diverse aetiology. One was a

young woman six to seven months pregnant whodeveloped a fluctuating visual defect of quadran-tic type, most marked in the ipsilateral eye. Theextent and severity of the field loss could be in-creased by hyperventilation and sometimes byeating. There was occlusion of the parieto-

occipital, calcarine, and posterior temporalbranches (Fig. 3). The second case was a man

aged 40 years with a history of vertebrobasilarinsufficiency; both calcarine and parieto-occipitalbranches were occluded. The field defect was a

transient incomplete hemianopia which re-

covered completely in a few weeks. The thirdcase was a man aged 64 years with giant-cellarteritis of the vertebral arteries. Angiographyshowed multiple occlusions involving the leftposterior temporal, left calcarine, and rightcalcarine arteries. Field defects again showedconsiderable fluctuation but a central 10° ofvision was constantly retained.

Headache

Headache

No change

ImprovedVBI

Unchanged

ComaDeath

1025

guest. Protected by copyright.

on October 15, 2020 by

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.37.9.1022 on 1 Septem

ber 1974. Dow

nloaded from

S. N. Kaul, G. H. Du Boulay, B. E. Kendall, and R. W. Ross Russell

FIG. 3. W.M., a woman aged 30 years, who, whensix to seven months pregnant, suddenly developed a

left-sided visual disturbance which later proved to be a

quadrantic type of hemianopia. She also complainedof transient left-sided numbness. This Towne's view ofa vertebral angiogram carried out eight months latershows occlusion of parts of the parieto-occipital andcalcarine arteries and also non-filling of the posteriortemporal branch.

SMALL VESSEL OCCLUSION (Table 3) In thisgroup there were three men and one woman,

ages ranging from 29 to 52 years. Underlyingfactors thought to predispose to occlusion weremigraine in three and the puerperium in one.Angiography showed normal main vessels butabsence of branches of the calcarine artery andareas of unperfused capillary circulation, bestshown on the subtraction film (Fig. 4). Fielddefects in all cases were similar and consisted ofa scotomatous hemianopia, quadrantic in shapeand extending to within 20 of the fixation point.There was no evidence of central sparing. Evi-dence of collateral blood flow was seen in onlyone case from the parieto-occipital branch of theposterior cerebral artery.

DISCUSSION

Beevor (1909), in a classical study of the distribu-tion of cerebral arteries by the method of post-

FIG. 4 G D a man aged 52 years who sufferedfrommigraine and developed a right quadrantic visual fielddefect with no central sparing. On this subtractedTowne's view of the vertebral angiogram, the smallbranches of the left calcarine artery are difficult to seeand there is substantially less capillary blush on thisside. Colour subtraction highlighted the slow and poorcapillary filling of the left calcarine cortex.

mortem injection, was the first to show thepotential value of pial anastomosis. He notedthat when the posterior cerebral artery wasoccluded its territory could easily be filled fromthe adjacent middle cerebral field by the border-zone anastomoses which in most subjects werefound in the U-shaped region on the infero-lateral aspect of the cerebral hemisphere, thebase of the U being near the occipital pole. Thatthe anastomoses have a function during life isknown from the finding that in posterior cerebralartery occlusion the area of infarction is normallymuch smaller than the anatomical territory ofthe vessel. Ischaemic damage is most severe inthe middle of the posterior cerebral arterialfield (the anterior part of the striate area) and theborder zones are spared (Zuilch and Behrend,1961).The results of the present study, which

attempts to correlate the site of posterior cerebralocclusion with the type and extent of visual fielddefect, provide further confirmation of Beevor'sviews. Our four cases of occlusion of the main

1026

guest. Protected by copyright.

on October 15, 2020 by

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.37.9.1022 on 1 Septem

ber 1974. Dow

nloaded from

Visual field defect and arterial occlusion in the posterior cerebral circulation

TABLE 2OCCLUSION OF CALCARINE, POSTERIOR TEMPORAL, PARIETO-OCCIPITAL ARTERIES

Sex, age Aetiology Associated Field defect Outcome(yr) features

CalcarineI F 42 Atheroma

Embolism

2 M 52 Atheroma

VBI

Nil

ciC

ciC

Unchanged

Unchanged

Multiple, unilateral3 F 30 Hemianaesthesia 3 main branches

Fluctuating defectFull recovery

4 M 40 Atheroma

Multiple, bilateral5 M 64 Giant cell

arteritis

VBI

NilS ci

Calcarine and parieto-occipital

Full recovery

L posterior temporalL calcarine completeR calcarine partialPartial recovery

trunk of the artery had severe permanent visualfield defects but showed at least 10° centralsparing and, in one case, sparing of a furthervertical strip. In anatomical terms this impliesintact macular projections to the posterior poleand, in addition, survival of an area in the peri-phery of the striate cortex. This area is furthestfrom the middle of the posterior cerebral fieldand nearest to the surrounding vascular terri-tories. Conversely, that part of the visual fieldmost consistently lost was the area adjacent tothe horizontal meridian which corresponds tothe depths ofthe calcarine fissure in the middle ofthe posterior cerebral territory.

There was also radiological confirmation thatthe route of collateral blood supply was via theposterior temporal and angular branches of themiddle cerebral artery. In fact, posterior cerebralartery occlusion may sometimes be diagnosed bycarotid angiography when in the late arterialphase retrograde filling of posterior cerebralbranches is seen.

Radiological/clinical correlation is less easy inthe case of branch occlusions since there is con-siderable variation in the number, size, anddistribution of the three main branches-posterior temporal, parieto-occipital, and cal-carine. In general, the calcarine artery is the

PregnancyEmbolism c.I"

1027

guest. Protected by copyright.

on October 15, 2020 by

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.37.9.1022 on 1 Septem

ber 1974. Dow

nloaded from

S. N. Kaul, G. H. Du Boulay, B. E. Kendall, and R. W. Ross Russell

TABLE 3OCCLUSION OF SMALL BRANCHES, DELAYED CAPILLARY PERFUSION

Sex, age Aetiology Associated Field defect Outcome(yr) features

1 M 45 Migraine Nil Unchanged

2 F 29 Puerperium Nil Unchanged

3 M 52 Migraine Nil Unchanged

4 MI 36 Migraine Nil Unchanged?Polyarteritis Nil

W ~~~~lmain blood supply to the striate cortex but thereare also contributions from the posterior tem-poral and/or parieto-occipital branches in over

500% of brains. Our cases of branch occlusionshowed no constant relationship between severityand extent of visual field loss and the type ofarterial occlusion. On the one hand there were twopatients with unilateral calcarine artery occlu-sion who had permanent loss of all but thecentral 5-10° of visual field; on the otherwerethree cases with multiple branch occlusionspossibly embolic in origin, all of whom showedan incomplete hemianopia, fluctuating in extentand eventually recovering to a substantial degree.The area of unperfused capillary bed shown on

the vertebral angiogram of such patients is con-

siderably greater than the anatomical extent ofbrain damage deduced from visual field studies(Hoyt and Newton, 1970). This is due to col-lateral supply from the carotid territory and can

be verified by performing a carotid angiogramwhen contrast medium can be seen reaching thearea via the posterior temporal (and perhaps theangular) branches of the middle cerebral artery.In our cases of isolated calcarine artery occlusionthere was also radiological evidence of collateralflow from the parieto-occipital branch of theposterior cerebral artery, although Beevor andlater Van der Eecken and Adams (1953) showedby injection studies that the anastomoses be-tween branches of the same artery were less welldeveloped than those between adjacent maincerebral arteries.The third group showing small branch occlu-

sion or areas of capillary avascularity is of par-ticular interest since this abnormality wouldprobably escape detection by older techniques.The visual field defect in these cases was of auniform type and comprised a scotomatouscongruous hemianopic defect, triangular in

1028

guest. Protected by copyright.

on October 15, 2020 by

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.37.9.1022 on 1 Septem

ber 1974. Dow

nloaded from

Visual field defect and arterial occlusion in the posterior cerebral circulation1

shape with the apex extending to the fixationpoint, and with no evidence of central sparing.This would correspond to a wedge-shaped areaof striate cortex extending to the occipital poleand including either the upper or lower lips ofthe calcarine fissure (Fig. 4). An exactly similarcase with pathological confirmation is recordedby Polyak (1957). The lack of sparing of thecentral visual field in these cases of branchocclusion has been discussed elsewhere (Russell,1973). It may be explained by involvement of alength of a single vessel or by involvement of anumber of vessels in the pathological process,thus limiting the potential collateral supply. Suchconditions may obtain in multiple small emboli,propagating thrombosis, or vasospasm involvinga segment of vessel.

AETIOLOGY OF VASCULAR OCCLUSION Our find-

ings illustrate that posterior cerebral artery occlu-sion is not due to a single pathological process.The various types of pathology found should notbe taken as a true representation since olderpatients or those with hypertensive cardiacdisease or with a known source of embolizationhave been excluded. The number of cases withsimultaneous bilateral or multiple occlusionssuggests that embolism with fragmentation ofemboli in the circulation is more common than isgenerally thought. Pathological studies haveshown that the source of emboli in these casesmay be mural thrombus of the vertebral orbasilar arteries; there is usually extensiveatheroma and sometimes aneurysmal dilatationof the basilar artery. Giant-cell arteritis of thevertebral arteries is another recognized cause(Wilkinson and Russell, 1972).

In younger patients the most importantaetiological factor was a history of severe recur-rent migraine. The preceding migraines in allcases were accompanied by prominent visualsymptoms which were sometimes bilateral andthe permanent hemianopia appeared in thecourse of a severe migraine attack. The site ofocclusion was the main posterior cerebral trunkin one case and small branch vessels in three. Itmust be emphasized that vascular occlusion andprominent neurological deficit is exceptionallyrare in migraine but in reported cases theposterior cerebral artery territory is more oftenaffected than other areas (Pearce, 1968). The

fact that vascular occlusion can be demonstrateddays or weeks after the onset of symptoms sug-gests that the condition is not simply due to anexceptionally severe episode of vascular spasmor oedema but probably represents a superaddedthrombosis. Other aetiological factors in youngpatients included contraceptive medication andthe puerperium, both recognized as predisposingto thrombosis (Bickerstaff and Holmes, 1967).Delayed endarteritis after radiotherapy may alsoaffect arteries of this size (Kagan et al., 1971).The angiographic examinations were done at

various intervals, after the onset of visual lossand the appearance on x-ray examination cannotbe assumed to be constant throughout the ill-ness. Pathological studies in cases of cerebralinfarction have shown that the earlier theexamination the higher the proportion of occlu-sions. Lysis and fragmentation of arterialemboli has been demonstrated by pathologicaland angiographic studies and by direct observa-tions on the retina (Fisher and Adams, 1951;Russell, 1961; Dalal et al., 1965). The impactionof an embolus at the bifurcation of the basilarartery and its subsequent displacement into theposterior cerebral artery on one side probablyaccounts for those patients who suffer from totalloss of sight which clears in the space of minutesor hours to leave a unilateral defect (Symondsand Mackenzie, 1957). Simultaneous smallvessel blockage, smooth regular vascular out-lines up to the point of obstruction, and thevisualization of the sharply defined profile of theproximal end of the thrombus lodged in thevascular lumen are further angiographic pointsin favour of embolism.The time factor must also be taken into

account in the comparison of visual fields. It hasbeen found that complete and dense hemianopiato all visual stimuli which is present 48 hoursafter a stroke frequently persists indefinitely, butif a patient retains even vague perception of handmovements in the affected field then considerablerecovery can be expected. This takes the form ofgradual enlargement of the central spared areaand extension of vision across the vertical meri-dian. Incongruity between the two eyes is amarked feature of these cases, the eye showing thetemporal field loss usually being the moreseverely affected. In the final stage the defect maybe demonstrated in this eye only.

1029

guest. Protected by copyright.

on October 15, 2020 by

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.37.9.1022 on 1 Septem

ber 1974. Dow

nloaded from

S. N. Kaul, G. H. Du Boulay, B. E. Rendall, and R. W. Ross Russell

REFERENCES

Abbie, A. A. (1938). The blood supply of the visual pathways.Medical Journal of Australia, 2, 199-202.

Beevor, C. E. (1909). On the distribution of the differentarteries supplying the human brain. Philosophical Trans-actions of the Royal Society, Series B, 200, 1-55.

Bickerstaff, E. R., and Holmes, J. MacD. (1967). Cerebralarterial insufficiency and oral contraceptives. BritishMedical Journal, 1, 726-729.

Dalal, P. M., Shah, P. M., and Aiyer, R. R. (1965). Angio-graphic study of cerebral embolism. Lancet, 2, 358-361.

Fisher, M., and Adams, R. D. (1951). Observations on brainembolism with special reference to the mechanism ofhemorrhagic infarction. Journal of Neuropathology andExperimental Neurology, 10, 92-94.

Henschen, S. E. (1926). Zur Anatomie der Sehbahn und desSehzentrums. Albrecht v. Graefes Archivfiir Ophthalmologie,117, 403-418.

Holmes, G. (1931). A contribution to the cortical representa-tion of vision. Brain, 54, 470-479.

Hoyt, W. F., and Newton, T. H. (1970). Angiographicchanges with occlusion of arteries that supply the visualcortex. New Zealand Medical Journal, 72, 310-317.

Kagan, A. R., Bruce, D. W., and Di Chiro, G. (1971). Fatalfoam cell arteritis of the brain after irradiation forHodgkin's disease: angiography and pathology. Stroke, 2,232-238.

Margolis, M. T., Newton, T. H., and Hoyt, W. F. (1971).Cortical branches of the posterior cerebral artery. Ana-tomic-radiologic correlation. Neuroradiology, 2, 127-135.

Pearce, J. (1968). The ophthalmological complications ofmigraine. Journal of the Neurological Sciences, 6, 73-81.

Polyak, S. L. (1957). The Vertebrate Visual System. Universityof Chicago Press: Chicago.

Russell, R. W. Ross (1961). Observations on the retinalblood-vessels in monocular blindness. Lancet, 2, 1422-1428.

Russell, R. W. Ross (1973). The posterior cerebral circula-tion. Journal of the Royal College ofPhysicians ofLondon,7, 331-346.

Shellshear, J. L. (1927). A contribution to our knowledge ofthe arterial supply of the cerebral cortex in man. Brain, 50,236-253.

Smith, C. G., and Richardson, W. F. (1966). The course anddistribution of the arteries supplying the visual (striate)cortex. American Journal of Ophthalmology, 61, 1391-1396.

Symonds, C., and Mackenzie, I. (1957). Bilateral loss ofvision from cerebral infarction. Brain, 80, 415-455.

Van der Eecken, H. M., and Adams, R. D. (1953). Theanatomy and functional significance of the meningealarterial anastomoses of the human brain. Journal ofNeuro-pathology and Experimental Neurology, 12, 132-157.

Wilkinson, I. M. S., and Russell, R. W. Ross (1972). Arteriesof the head and neck in giant cell arteritis. Archives ofNeurology (Chic.), 27, 378-391.

Zulch, K. J., and Behrend, R. C. H. (1961). The pathogenesisand topography of anoxia, hypoxia and ischemia of thebrain in man. In Cerebral Anoxia and the Electroencephalo-gram, pp. 144-163. Edited by H. Gastaut and J. S. Meyer.Thomas: Springfield, Ill.

1030

guest. Protected by copyright.

on October 15, 2020 by

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.37.9.1022 on 1 Septem

ber 1974. Dow

nloaded from