aqueous flow into the perivascular space of the rabbit ciliary body

Aqueous flow into the perivascular spaceof the rabbit ciliary body

W. L. Fowlks and Virginia R. Havener

I

Through the use of the histochemical reagent, nitroblue tetrazolium chloride, to label anteriorchamber aqueous in the eyes of living albino rabbits, we have been able to trace the pathwayof aqueous leaving the anterior chamber angle. Although toe could find no evidence that anyof the marker had entered the trabecular veins (sometimes called Schlemm's canal) or thetrabecular meshwork directly in front of the veins, there was evidence that the aqueous hadmoved through channels in the ciliary body posterior to the ciliary cleft. The pigmented epi-thelium of the ciliary body, but not the ciliary processes, was intensely stained with marker,particularly in the areas adjacent to the perivascular aqueous channels. From a detailed studyof the distribution of the nitroblue formazan in serial sections of the ciliary body region ofeyes injected in vivo with marker 10 minutes before excision and fixation, we have concludedthat aqueous leaving the anterior chamber first enters the ciliary cleft from which it flows intothe perivascular spaces and perivascular aqueous channels of the ciliary body. The aqueouschannels lie adjacent to the pigmented epithelium in the posterior part of the ciliary body andsome of them connect with the perivascular space of the suprachoroid. The uptake of stainby the pigmented epithelium of the ciliary body has been interpreted to mean that fluidfrom this channel system may be resecreted into the posterior chamber.

n 1943 Kiss1 reported some experimentswhich confirmed earlier reports2'3 describ-ing the infiltration of colloidal pigmentsinto the ciliary body of the rabbit, othermammals, and man following injection invivo of India ink or other pigments into theanterior chamber. Kiss stated that if theanterior chamber was labeled with col-loidal pigment without an increase of theintraocular pressure during the process,some hours later particles of pigment wereto be found infiltrating the ciliary body

From the Department of Ophthalmology, Collegeof Medical Sciences, University of Minnesota,Minneapolis, Minn.

This work was supported in part by Grant No.B1979 from the Neurological Diseases andBlindness Section of the National Institutes ofHealth, United States Public Health Service.

but were not found in the structure werefer to as trabecular veins, nor were theparticles concentrated especially in orabout the trabecular meshwork in frontof the trabecular veins. Presumably theparticles which infiltrate the ciliary bodywere transported there by some sort offluid flow from the anterior chamber al-though other possible modes of transport,i.e., diffusion, were not convincingly ruledout. Seidel,4 for example, in 1921 was verycritical of earlier workers, who had in-jected pigments and soluble crystalloidsinto the anterior chambers of various liv-ing animals, on the grounds that when herepeated their experiments and examinedthe eyes microscopically a day or two af-ter the injections, the colloidal pigmentwhich was observed infiltrating the ciliarybody was found mostly intracellularly in

374

Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/932895/ on 03/23/2018

Vohttne 3Number 4

Aqueous flow into perivascular space 375

phagocytes. He thought the marker pig-ment which initially entered the outsidelayers of the tissue by diffusion was car-ried further into the ciliary body by thephagocytes. In addition, his experimentswith injection of colloidal dyes under ex-cess pressure into excised eyes was inter-preted by him to mean that aqueous leftthe rabbit eye only via an episcleral venoussystem.

In a comprehensive description of theanatomy of the rabbit eye, Davis5 sug-gested (1929) that the aqueous may leavethe anterior chamber of these animals byfiltering into the perivascular lymph spacesin the root of the iris and also into similarspaces found "behind the lace-like uvealmeshwork of the iris angle" as well as bySchlemm's canal. The structure which Da-vis" labeled "Schlemm's canal" was called"trabecular veins" by Troncoso.G In therabbit this structure is located posterior tothe iris-corneal junction about at the apexof the cilioscleral sinus, a structure calledthe ciliary cleft by Duke-Elder.7 In thisreport, we will call these structurestrabecular veins and ciliary cleft, respec-tively, not out of preference or convictionbut simply because a name for an identi-fiable structure is convenient. When weuse these names we mean simply the spe-cific structures as they are labeled in thefigures of the references cited.

If aqueous leaves the anterior chamberby flowing into the ciliary cleft, as sug-gested by Davis,5 then if aqueous entersthe intrascleral vascular plexus which in-cludes the trabecular veins, as was sug-gested by Troncoso,6 this pathway shouldbe marked by any label which had beeninjected into the anterior chamber a suf-ficient time before enucleation. Thus, mi-croscopic sections from eyes which hadbeen injected with a suitable label foraqueous, i.e., colloidal pigment or histo-chemical reagent, should show evidence ofthe label in the ciliary cleft and all othercanals, ducts, or tissue through which theaqueous moves as it exits from the eye.

As part of a study of fluid movement in

the eye,8 we have repeated some of theprevious experiments and have extendedthe work to include experimental methodsand controls which we believe take intoaccount the question of passive diffusioninto the ciliary body. To eliminate anypossible role of phagocytes in the trans-port scheme and to minimize the effectsof diffusion, we have used a very shorttime instead of the days or hours used byprevious investigators. As a label andmarker for fluid movement, we have againemployed the water-soluble histochemicalreagent, nitroblue tetrazolium chloride,which is reduced by all cellular tissues ofthe eye to insoluble nitroblue formazan.

Materials and methodsThe materials employed and the methods used

were essentially the same as those reported inprevious researchs where applicable. In the pres-ent experiments injections of 2 to 15 f\ of nitro-blue tetrazolium chloride (nitro BT) (10 mg. permilliliter) solution were made after insertion ofa needle through the cornea of albino rabbits insuch a way that the point of the needle was posi-tioned in the filtration angle without the tip touch-ing the iris or cornea.

At die expiration of the time of an experiment(which for all experiments reported was 10 min-utes), the rabbit was beheaded and the experi-mental eye was enucleated, frozen in a dry ice-petroleum ether mixture, and transferred toBouin's fixative which had been previously frozento a mush. The bottles with the eyes and fixativewere immediately placed in the freezing compart-ment of the refrigerator at about -10° C. for 24hours, then were allowed to fix at room tempera-ture another 24 hours or more before processingfor sectioning.

After enucleation of the experimental eye, thehead with the remaining eye was placed in apan kept at 37° ± 1° C , and the second eye wasinjected in a manner as identical as possible tothat used for the experimental eye. The secondeye, which will be referred to as the control eye,was also enucleated after 10 minutes and wasfrozen, fixed, and prepared in the same way as theexperimental eye.

After fixation and washing were completed, acoronal slice of each eye was prepared by cuttingoff the cornea and sectioning the eye just posteriorto the ora serrata. After the zonules were cut thelens was removed. Microscopic sections of thesepreparations 8 /* thick were cut parallel to, or ata very acute angle to, the plane of the corona


376 Fowlks and Havener Investigative OphthalmologyAugust 1964

Fig. 1. Semischematic drawing of the filtrationangle and ciliary body of the rabbit. A, the cornea;B, the iris; C, sclera; D, the pectinate ligamentat tine scleral end; E, anterior open spaces of theciliary cleft; F, iris pillar; G, perivascular spacearound a large vein of the ciliary body just pos-terior to the apex of the ciliary cleft; H, the tra-becular veins; /, perivascular aqueous channel;K, large vein posterior to the perivascular region,which lies above this level, showing absence ofperivascular spaces; L, the distorted S-shaped twistof the large veins at the root of the iris.

ciliaris. Such sections, cut through the region ofthe nitroblue formazan stained portion of the eyes,were mounted serially on slides without additionalstaining. Occasional sections were stained withHarris's hematoxylin and eosin stain for identifica-tion of doubtful structures.

Pertinent anatomy of therabbit ciliary body

The ciliary body of the rabbit is tra-versed by closely spaced, meridionally ori-ented, relatively large blood vessels. Theblood vessels are veins which drain a capil-lary plexus of the iris.5 These veins alsotend to have a distorted S-shaped coursethrough the root of the iris.1 In meridionalsections the ciliary cleft7 of the rabbit eye

is shaped similar to an acute triangle withthe base in the plane of the anterior sur-face of the iris. The outside of the ciliarycleft is defined by Descemet's membrane,an extension of the corneal endothelium,and the sclerocorneal trabeculum. Theapex lies at the posterior terminus of thesclerocorneal trabeculum. The inside isbordered by stromal cells of the iris rootand ciliary body except where this side isbordered by the walls of the large veinsreferred to previously as they begin theirmeridional course posteriorward. At itsapex the ciliary cleft is crossed by nu-merous iris pillars" which decrease mark-edly in number toward the base so thatthe anterior half of the ciliary cleft iscrossed by very few iris pillars. The an-terior surface of the ciliary cleft is closedwith a membranous extension of the irisroot, the pectinate ligament,7 which is at-tached to Descemet's membrane in such away that a series of small openings isformed which connect the ciliary cleftwith the anterior chamber.

The structures called the trabecularveins are found embedded in the scleralstroma immediately adjacent to the ciliarybody. They are located just posterior tothe terminus of Descemet's membrane andare covered on the ciliary body side withtrabecular tissue. This tissue which Davisr>

labeled sclerocorneal trabeculum extendsanteriorly to where it inserts itself betweenDescemet's membrane and the sclerocor-neal stroma and posteriorly to the apex ofthe ciliary cleft. These anatomical featurescan be recognized in Fig. 1.

Results

Regardless of how slowly the nitro BTsolution was injected into the anteriorchamber angle of the control eyes in theseexperiments, the nitro BT was carriedthroughout the anterior chamber by con-vective currents so that most of the an-terior chamber contained some nitroblueformazan at the end of the experiments.In contrast, when nitro BT solution was in-jected into experimental eyes very slowly


Volume 3Number 4


Fig. 2. Photomicrograph of an unstained oblique section cut from a coronal ring, which in-cludes the iris and corona ciliaris of an experimental eye of an albino rabbit. This photomicro-graph shows part of the iris, the angle, and about one half of the length of the ciliary bodyregion measured from the iris angle to pars plana. This eye was injected in vivo at the filtrationangle with 10 MI of nitro BT solution 10 minutes before the eye was excised, frozen, andfixed. Dense deposits of nitroblue fomiazan are to be seen in the open spaces of the ciliarycleft, in the perivascular spaces around the large veins, and in the pigmented epithelium of theciliary body. Fomiazan also stains certain structures in the iris and the pigmented epitheliumof the iris. The five large veins seen in this view, each cut at a different distance posteriorto the angle, illustrate the gradual conversion of the ciliary cleft into the perivascular spaces.(x280.)

(maximum rate 1 [A per 30 seconds) mostof the nitro BT solution entered the uvealtissue immediately posterior to the site ofinjection and little or none escaped intothe surrounding aqueous. Thus, only asmall area of the filtration angle wasstained with nitroblue formazan in theseeyes.

On gross examination of the coronalslices, which includes the corona ciliarisand iris and were prepared as describedunder materials and methods, it was foundthat the most obvious difference betweenthe experimental and control eyes from thesame albino rabbit was the amount of bluestain visible in the region of the coronaciliaris as viewed from the posterior sideof the preparation. A blue-stained area was

always found in the segment opposite theinjection site but only in those prepara-tions from experimental eyes. The size ofthe area correlated well with the amountof nitro BT solution injected—the moreinjected the larger the area. The stainedarea of the experimental eyes injected withthe larger amounts of nitro BT solutionoften extended onto the iris, and the con-trol eye from the same animal sometimesshowed a little stain on the posterior ofthe iris. The stained area on the experi-mental eyes was confined to the tissue ofthe iris and ciliary body between theprocesses. The ciliary processes themselvesshowed no sign of formazan stain.

On examination of the serial sectionsmade from experimental eyes, we found



Fig. 3. The trabecular vein (TV) region from a microscopic section of the same eye of Fig. 2.The portion of the trabecular meshwork (TM) directly in front of the trabecular vein did notcontain formazan, although the ciliary cleft (CC) both anterior and posterior contained for-mazan in large amounts, and the sclera (S) both anterior and posterior to the trabecular veincontained formazan. The trabecular meshwork both anterior and posterior to the trabecularvein has appreciable stain. The layers of trabecular meshwork in front of the trabecular veinhas some fonnazan stain but only those layers adjacent to the ciliary cleft. (x800.)

that the open spaces of the ciliary cleftcontain large amounts of nitroblue forma-zan (Fig. 2). The same region in the con-trol eyes has only barely detectableamounts of nitroblue formazan in the cili-ary cleft, at the iris surface, inside the irisstroina (if more nitro BT than was usuallyused had been injected), and nowhereelse except the anterior chamber. The ex-perimental eyes which were injectedslowly with 5 to 10 /A of nitro BT solutionhave dense deposits of granular formazanwhich extend posteriorly in the ciliary cleftall the way to the apex. In these eyes thetrabecular meshwork directly in front ofthe trabecular veins was noted to be freeof formazan even though the same struc-ture anterior and posterior to the trabecu-lar veins was stained. In some sections,particularly from eyes which had been in-jected with the larger amounts of nitro

BT solution, the trabecular meshwork lay-ers adjacent to the ciliary cleft werestained lightly in the region in front ofthe trabecular veins (Fig. 3). The trabecu-lar veins and their lining were invariablyfree of the stain, except in the eye thathad been injected with 15 /A of markersolution in less than 5 minutes. The dis-tribution of formazan in the vicinity of thetrabecular veins is particularly significantbecause in the same eyes appreciablequantities of formazan were invariablyfound in certain parts of the ciliary bodyposterior to the trabecular veins and in thesclera both anterior to and posterior to theveins.

When the pathway defined by the de-posits of nitro BT formazan was tracedfrom section to section it was found thatthe open spaces near the apex of the ciliarycleft are continuous with perivascular


Volume 3Number i


1

Fig. 4. Photomicrograph of a large vein (V) in the ciliary body posterior to the ciliary cleftand trabecular veins showing well-defined perivascular spaces (P), some of which containclumps of nitroblue formazan. This section from an albino rabbit eye was unstained exceptfor the residual yellow of Bouin's fixative and the blue deposits of formazan. The adventitiaof the vein is stained with nitroblue formazan on one side only, the side nearest the sclera.This photomicrograph was taken with the condenser adjusted to simulate phase contrast.(x800.)

spaces of the large, meridionally ori-ented veins. The well-defined perivascularspaces, such as those shown in Fig. 4, fol-low the vein posteriorly for a short dis-tance, the perivascular region. At the an-terior end of this perivascular region thevein has large open spaces on the scleralside only. The spaces gradually surroundthe vein until they have a typical peri-vascular appearance. Open spaces with noevidence of endothelial lining form con-tinuous, tortouous channels which branchoff from the apex area of the ciliary cleftand also from the perivascular region.These channels wind their way inside theciliary body until they are adjacent tothe pigmented epithelial layer. The chan-nels can be identified on microscopic sec-tions by the clumps of formazan seen inthem and by the formazan stained cellsadjacent to them. As seen in the usual

microscopic section stained with hema-toxylin and eosin or other stains, thesespaces could easily be mistaken for arti-facts since the continuity of the spaces ismasked by their normal tortuous path-ways which, when cut in a plane, resultsin a number of isolated spaces on anysingle section. In this respect they re-semble the perivascular spaces which ap-pear as continuous passageways or con-nected chambers only in reconstructionsfrom serial sections.

There are a few similar channels in thestroma between the veins in the perivas-cular space region of the ciliary body. Thechannels are not numerous nor were theyfound to contain formazan in amountsequal to the formazan deposits in the chan-nels in the perivascular region. The peri-vascular spaces of the perivascular regionleave the large veins rather abruptly about


380 Fowlks and Havener Investigative OphthalmologyAugust 196<t

Fig. 5. Higher magnification of the root of a ciliary process at the midpoint of the ciliarybody of Fig. 2 showing the intensely stained pigment epithelial layer of this albino rabbitwith little or no stain in the epithelial layer adjacent to the posterior chamber. Peri vascularaqueous channels are seen adjacent to the pigmented epithelium near the two sides of theciliary process. (x4,500.)

two thirds of the distance from the pec-tinate ligaments to the border between theciliary body and pars plana. The largeveins continue posteriorly from that levelwithout perivascular spaces, but the spacescontinue as channels both posteriorly andtoward the inside with multiple branchingsuntil the channels also are adjacent to thepigment epithelium of the ciliary body. Inall respects they resemble the channelswhich branch off from the apex of the cili-ary cleft and the perivascular spaces. Someof the channels proceed posteriorly in thetissue adjacent to the pigmented epithe-lium of the ciliary body until at the level ofpars plana they become continuous withthe perivascular spaces of the suprachor-oid. We will refer to these channels in theremainder of this paper as perivascularaqueous channels.

Microscopic examination of sectionsfrom experimental eyes also reveals that

the pigmented epithelium nearest a peri-vascular aqueous channel is usually moreintensely stained than other portions ofthe pigmented epithelium in that area(Fig. 5). Cells were found in the ciliaryepithelial layer adjacent to the' most in-tensely stained pigmented epithelial cellswhich also contained recognizable amountsof nitroblue formazan.

While no quantitative measurementshave yet been made, it appears from ex-amination of our serial sections that thetotal cross-section area of the open spacesof the ciliary cleft, the perivascular spaces,and the perivascular aqueous channels de-creases markedly from any given level toa more posterior level. It also appeal's thatthe total number of perivascular aqueouschannels which reach the level of parsplana are only a fraction of the total num-ber that branch off from the ciliary cleftand perivascular spaces.


Volume 3Number 4

Aqueous floio into perivascular space 381

Another observation which should benoted is that the walls of the veins in theperivascular region are usually unstainedeven though the adjacent perivascularspace or ciliary cleft contains dense de-posits of nitroblue formazan. Perhaps thestain that is occasionally found in the ad-ventitia of these veins diffuses there dur-ing the time between beheading the ani-mal and freezing the excised eye. Sincethe walls of these veins are not usuallystained, it would appear that aqueous ab-sorption into the veins in this region doesnot occur with great facility.

Fig. 1 is a fairly accurate schematic rep-resentation of the open spaces of the filtra-tion angle and ciliary body of the rabbit.This illustration was drawn to show theopen spaces as we were able to reconstructthe area from our serial sections. Thephotomicrograph (Fig. 2) of an obliquecut through the ciliary body with five ofthe large veins of the ciliary body visible,each cut at a different distance posteriorto the iris, illustrates the typical changesin the ciliary cleft and perivascular spaceregion of the large veins. It also shows thedistribution we observe of nitroblue for-mazan in the open spaces of the ciliarycleft, perivascular spaces, perivascularaqueous channels, and in the ciliary bodyand pigmented epithelium of that struc-ture. Fig. 5 shows the shadings in the in-tensity of formazan stain in the pigmentedepithelium of the ciliary body at differentdistances from perivascular aqueous chan-nels.

From examination of Fig. 2 it can beseen that the pigmented epithelium of theiris and the most anterior portion of theciliary body are also deeply stained withformazan deposits. Certain areas of the irisare also visibly stained although withsomewhat less intensity than the pig-mented epithelium. Since our purpose wasto report only the findings in the ciliarybody, which involve the ciliary cleft andperivascular channels, and also becauseour researches on the uptake of nitro BTvia the iris are not completed, discussion

of these findings will be deferred to asubsequent paper. Suffice it to say at thistime that the route from anterior chamberto the pigmented epithelium of the irisand the anterior part of the ciliary bodydoes not appear to include movement ofreagent from the ciliary cleft.

Discussion

The earlier workers, including Kiss,1 whowere concerned with the pathways ofaqueous excretion apparently failed to ap-preciate the need to demonstrate a differ-ence between the living and dead eye.Since we made injections into the deadeyes (control eyes) in situ about 5 min-utes after the head was severed from therabbit's body, it actually would be moreaccurate to describe them as eyes withouta blood supply.

The question of the role played by thedifference in the intraocular pressure ofthe experimental and control eyes, as afactor in the distribution patterns observedfor the fluid movement tracer we em-ployed, is difficult to answer with absolutecertainty. It does not seem reasonable thatthe small reduction of intraocular pressureexperienced by the control eyes could bethe sole cause of the very marked differ-ence in the distribution pattern of nitro-blue formazan which was observed. It is abasic principle of fluid mechanics that afluid in a closed vessel does not move inthe absence of a specific driving force. Thephysical driving forces which have beenso far discovered are mechanical, i.e., by apump of some sort; electrical, of the end-osmotic variety; or magnetic if the fluidhas magnetic properties. Osmosis orprocesses associated with active transportcan result in a net movement of water inone direction in certain biologic systems.If it is true that the intraocular pressureis supported by the sclera, cornea, andblood vessels, then it would be a gross vi-olation of the basic principles of fluidmechanics to find fluid movement insidethe eye which was primarily responsive toa change in intraocular pressure. At or



near a hole through the sclera, cornea, orblood vessel, flow, which was responsive tochanges in the intraocular pressure, wouldoccur, since flow within a fluid will alwaysoccur along a pressure gradient from anygiven region to a region of lower pressure,except along the gravitational-inducedpressure gradient. At steady state the ab-solute pressure must be the same at allpoints inside the sclera and cornea if theinfinitesimal corrections due to rigidity ofthe uveal tissue and also due to gravita-tional differences are ignored. The appli-cation of these principles of fluid mechan-ics to the present problem leads to theprediction that more flow should havebeen observed in the control eyes if themajor exit of fluid from the eye was intothe equivalent of a hole through the scleraor into a vein.

The 10 minute time period between theinjection of tracer into the anterior cham-ber angle and freezing of the eye was soshort, that the role of any possible trans-port mechanism except convective flow isquite minimal. If diffusion alone could ac-count for the distribution patterns ob-served the control eyes should have shownthe same distribution of formazan as wasfound in the experimental eyes. Whenpieces of tissue were incubated in nitroBT solution 10 minutes, then frozen, fixed,and cut perpendicular to the surface, for-mazan was found only in a 0.5 mm. thicklayer adjacent to the surface. Within thislayer the intensity of stain is greatest atthe surface and decreases to none at theindistinct border. Fixation stops all reduc-tion of nitro BT to formazan and sincesome of the formazan is chemically com-bined with protein as it is reduced it can-not move about appreciably during prep-aration of the tissues for sectioning al-though some of the unbound formazanis probably lost through solubilization.There was little evidence on any of thesections we examined that diffusion dur-ing fixation had occurred to any appreci-able extent. Since large quantities of for-mazan were found in the eye tissue more

than half a millimeter posterior to the an-terior chamber, nitro BT must have beencarried there by flow of a fluid.

The pigmented epithelial layer of theposterior portion of the ciliary body ismore intensely stained than the adjacentstructure on the anterior side. The epi-thelial layer adjacent to the posteriorchamber was sometimes found to bestained very lightly but only adjacent tothe most intensely stained pigmented epi-thelial cells. The reverse of this is ob-served if nitro BT solution is placed in theposterior chamber. This observation is astrong argument for the hypothesis thatthe reagent had moved from the anteriorchamber by convective flow and was takenup by the pigmented epithelial cells be-cause these cells were in the process ofsecreting at least some substances fromthe aqueous into the posterior chamber.The fact that a blood supply to the eyeappears to be necessary before an appre-ciable flow will occur may indicate somesort of active role of the blood in thisprocess. Two possibilities suggest them-selves. The first, and most obvious, is therole of the blood as a supplier of oxygento an active transport system such as thetissue layer made up of the pigmentedepithelium and epithelial cells whichwould probably not operate efficientlywithout oxygen. A second possible role ofthe blood supply is suggested by the peri-vascular space region of the large veinswhich could serve to pump fluid throughthe perivascular spaces by utilizing rhyth-mic pulsations in the veins, the same me-chanical principle which applies to cer-tain laboratory pumps. Obviously bothmechanisms may be operative. But, re-gardless of the role played by the bloodsupply in the transport of nitro BT to thepigmented epithelial layer of the posteriorparts of the ciliary body, the pathway ofaqueous from the anterior chamberthrough the ciliary cleft and into the peri-vascular aqueous channels is clearly evi-dent from the distribution of formazan inthe experimental eyes.


Volume 3Number 4


The continuity of aqueous-filled spacesfrom the anterior chamber through thefull length of the ciliary body and into thesuprachoroid has apparently not been de-scribed before, although Davis5 referredto perivascular lymph spaces around thelarge veins in the iris root "near the basesof the iris pillars." He does not refer tothe perivascular spaces around these largeveins posterior to the level of the trabecu-lar veins, nor does he define very preciselythe structure to which he refers when heuses the phrase "lace-like uveal meshworkof the iris angle" although he probablymeans the structure called the ciliary cleftby Sir Duke-Elder.7 Although Davis5

stated that the iris pillars are covered withan endothelial layer and the open spacesof the iris angle and ciliary body are linedwith endothelium, this does not appear tobe true in our preparations. It appears tous that the iris pillars are partially coveredwith endothelial cells but most of the othertissue surrounding the open spaces doesnot appear to have an endothelial layer.For this reason we would object to usingthe term "lymph spaces" because this termis usually applied to spaces which are linedwith endothelium and do contain lymph.We prefer the term "perivascular aqueouschannels," since they are perivascular for

some portion of their course and containfluid throughout their length which is con-tinuous with anterior chamber aqueous.

The help and encouragement of Professor JohnE. Harris and Dr. Robert Hugh Monahan and theassistance of John E. Parker of the Medical Artand Photography Department, who made thedrawing (Fig. 1), were invaluable aids to use incompleting this project. We acknowledge theircontributions with many thanks.

R E F E R E N C E S1. Kiss, F.: Der Blutkreislauf des Auges, Ophthal-

mologica 106: 225, 1943.2. Nuel, J. P., and Benoit, F.: Des voies d'61im-

ination des liquides intra-oculaires hors de lachambre anterieure et au fond de l'oeil, Arch.cl'Opth. 20:161, 1900.

3. Erdmann, Paul: uber experimentelles Glaukomnebst Untersuckungen am glaukomatosen Tie-rauge, Arch. f. Ophth. 66: 325, 1907.

4. Seidel, Erich: iiber den Abfluss des Kammer-wassers aus der vorderen Augenkammer, Arch,f. Ophth. 104: 357, 1921.

5. Davis, F. A.: The anatomy and histology of theeye and orbit of the rabbit, Tr. Am. Ophth.Soc. 27: 401, 1929.

6. Troncoso, M. U.: The intrascleral vascularplexus and its relations to the aqueous veins,Am. J. Ophth. 25: 1153, 1942.

7. Duke-Elder, Sir Stewart: System of ophthal-mology, vol. 1, The eye in evolution, St. Louis,1958, The C. V. Mosby Company, p. 460 ff.

8. Fowlks, W. L.: Meridional flow from thecorona ciliaxis through the pararetinal zone ofthe rabbit vitreous, Invest. Ophth. 2: 63, 1963.


aqueous flow into the perivascular space of the rabbit ciliary body

Documents