Caliber Spectra of Fibers in the Fasciculus Gracilis of the Cat Cervical Spinal Cord: A Quantitative Electron Microscopic Study 1,2 YANN-CHING HWANG,%4 EDWARD J . HINSMAN 4 AND 0 . F. ROESEL Depart ment s of vet eri n ary A n a t o m y , and 5 Ve te r inar y Physiology and Pharmacology, School of Veterinary Medicine, Purdue U niv e r s ity , West L afayet t e, Indiana 47907
5
ABSTRACT In order to obtain a better understanding of the microscopic structure of the cat fasciculus gracilis, entire cross-sections of the fasciculus were examined with the electron microscope. Twenty-five thousand, two hundred and eighty-four fibers were encountered in one fasciculus. The fiber caliber spectra obtained from the study show that the fasciculus gracilis at cervical level has a unique fiber distribution pattern. The fiber diameters range from less than 1 p to 15 p , however, 97% of fibers have diameters less than 8 p ; and the majority of the fibers are in the 2-5 p range.
It is generally acknowledged that the ascending fibers of the fasciculus gracilis maintain an orderly arrangement of lamination representing dermatomal layering of the body. However, overlap of the lamination has been noticed (Kodama, '42; Walker and Weaver, '42; Carpenter et al., '68). Recently, Whitsel and his coworkers ('70, '72) reported that fibers of the fasciculus gracilis re-sort while they extend from the lumbosacral to the cervical portions, and the map of the hind limb in the cervical gracile fasciculus more closely resembles the map of the hind limb in sensory cortical areas than does the representation at lumbar levels of the fasciculus. On the other hand, fibers of the fasciculus gracilis are believed to cluster in the beta (group 11) size range (Mountcastle and Darian-Smith, '68), which is in conflict with the results of several fiber spectral studies involving this fasciculus in both human and the cat (Haggqvist, '36; Szentagothai-Schimert, '4 1; Hildebrand and Skoglund, '71). The present study, by employing techniques of quantitative electron microscopy and with the aid of the computer, was to reinvestigate fiber distributional pattern and fiber size range within the fasciculus gracilis of the cat at cervical level. MATERIALS AND METHODS
Adult cats raised in the colony of the Department of Veterinary Anatomy, PurJ.
COMP.
NEUR., 162: 1 9 5 2 0 4
due University were used. Cats were perfused by Karnovsky's ('65) immersion fixative under deep sodium pentobarbital anesthesia following a modified retrograde aortic perfusion technique of GonzalezAguilar and DeRobertis ('63). The spinal cord at C3 level was removed and selected parts of the dorsal column were dissected free and trimmed. Pieces of tissue were then postfixed in 2 % osmium tetroxide, and after alcoholic dehydration embedded in Epon 812. For consistently producing precise cross sections of the tissue, a special embedding technique (Hwang, '70) was followed. The thin epon section was cut perpendicular to the long axis of the fasciculus and picked up on a loop of thin film (Padgett, '63). Each section with film was then precisely mounted on a slot grid (Gay and Anderson, '54). Sections were double stained with uranyl acetate (Watson, '58) and lead citrate (Venable and Coggeshall, '65) before examination, and photographed with a Philips 200 electron microscope. In the primary case (cat no. 526), 297 electron micrographs, covering most of the fasciculus gracilis and vicinity, were taken from two consecutive serial sections at X 1,900. The negatives were then project1 Published as Paper Number 5744, Purdue Agriculture Experiment Station, West Lafayette, Indiana. 2 Research support in part from Purdue Agricultural Experiment Station -project 1418-36-1174. 3 Present address: Department of Anatomy, School of Medicine, Howard University, Washington, D. C. 20001.
195
196
YANN-CHING HWANG, EDWARD J . HINSMAN A N D 0. F. ROESEL
ed at X 10,000 onto a tracing board, an image of each axon circumference was traced, and myelin sheath thickness was marked on tracing paper. With the aid of printed photographs ( x 4,000), each fiber on the tracing paper was identified, and its circumference was measured by a OTT planimeter, and myelin sheath thickness measured by a millimeter ruled scale. Data of each fiber were then processed with computer assistance. A digital computer program written by one of the authors (0. F. Roesel) was used to aid in computation of the fiber diameter and in formation of the caliber histogram of each photographic negative. By studying primarily the peak and the mode of each histogram, a caliber spectral chart (fig. 1) was made. Based upon this chart, the total area was partitioned into four groups (subareas). The group’s grand histogram and fiber diameter range were then produced by the aid of the computer. In another case (cat no. 162), only four electron photographs were taken from each group, however, all similar data processing was followed. The thick epon sections for cat no. 526 and cat no. 162, together with five other animals’ were examined under the light microscope for comparison of patterns of fiber distribution within fasciculus gracilis, and boundaries of the two fasiculi of the dorsal column. RESULTS
In the primary experiment (cat no. 526), a total area of about 541,000 pL2of tissue was studied, which covered most of the fasciculus gracilis and some of its vicinity. The total nerve fiber count in this area is 29,206. As we already mentioned, the caliber histogram chart (fig. 1) was produced from the primary computer manipulation which considered each electron photomicrographic negative as a sector unit and a histogram was produced for each sector unit. Because of regional differences observed in the caliber histogram chart, the area was partitioned into four subareas (Sa): SA-111, SA-IV, SA-V, and SA-P. A demarcation of the four subareas is shown in figure 2. Based upon these four subareas, the caliber histograms and the fiber
diameter range of both the primary and the additional studies were then considered. Caliber his togram s The histograms of all four subareas of both cat no. 526 and cat no. 162 are shown in figures 3 and 4. In the primary study (cat no. 526) histograms of SA-111, SA-IV, and SA-V have prominent unimodal peaks which exceed 25% of the total number of fibers involved; on the other hand, the histogram of SA-P is bimodal with a more gradual slope. The peak of each subarea was used to designate that particular subarea; i.e., SA-I11 has its peak at 3 p , SA-IV at 4 p , and SA-V at 5 p ranges, however, P of the SA-P symbolizes a mosaic grand polymodal histogram. Although the peak of SA-I11 histogram locates at 3 p , there are also fairly high percentages of fibers at 2 p and 4 p diameter ranges. This is primarily because the subarea includes not only the individual sector unit histograms with peaks at 3 p but also the sector unit histograms with peaks at at 2 p and some at 4 p . The histograms of SA-IV and SA-V both show very sharp unimodal peaks. In contrast, the histogram of SA-P is unique. The area of SA-P as seen in figure 1 is not homogeneous, but an area with different pattern of histograms and fiber diameter peaks. When all these different histograms were put together, the mosaic grand histogram for the area of SA-P was then formed, which was characterized by wide spreading fiber diameter range with its major peak at 4 p. In cat no. 162, four areas (approximately 10,000 p2 each) identifiable as SA-111, SA-IV, SA-V and SA-P were sampled. Histograms (fig. 4) from each area, in general, do agree with the results of cat no. 526 (fig. 3). However, the following differences were noticed: (1) The peak of SA-111 histogram shifted to 2 p . This result is mostly due to difficulty encountered in sampling such a 10,000 p2 area from a slender band-like SA-I11 (fig. 2). To ensure the electron photomicrographic sample was inside the subarea, we had no choice but to take the photographs from the most wide open portion of the SA-111, which is the more dorsolateral portion of this subarea where more fine fibers were seen (fig.
CALIBER SPECTRUM OF F. GRACILIS I N CAT
Fig. 1
Caliber Specbal Chart of the Cat Fasciculus Gracilis.
197
198
YANN-CHING HWANG, EDWARD J. HINSMAN AND 0. F. ROESEL
Fig. 2 Partition of the Fasciculus Gracilis and Vicinity. The area photographed is partitioned into four subareas according to the caliber spectral chart (fig. 1). The fasciculus gracilis is bounded between the dorsal median sulcus (DMS) and septum and dorsal intermediate sulcus (DIS), however, the boundary between the fasciculus and the fasciculus cuneatus is primarily the lateral edge of SA-111. (III), Band-like area of SA-111 with more 2 p or 3 p fibers; (IV) Band-like area of SA-IV predominated by 4 p fibers; (V), Band-like area of SA-V with more 5 p fibers; (P). Band-like area of SA-P which has unique fiber distributional pattern and is not considered as a part of the fasciculus; (FC), part of the fasciculus cuneatus.
CALIBER SPECTRUM OF F. GRACILIS I N CAT
FIBER DIAMETER ( p i
Fig. 3 Histograms of Fasciculus Gracilis of cat No. 526 at Cs level.
1). (2) The most bimodal peaks of SA-P are located at 4 p and 7 CL instead of 4 p and 2 p . However, in both animals, the major peaks are at 4 p . Fiber diameters Diameter percentages of the caliber spectra of cats no. S26 and no. 162 are shown in tables 1 and 2. On the primary study (table 1) the majority of the fibers within SA-111, SA-IV and SA-V belong to the delta fiber size range (2p-5 F ) , while only 12.5% of 8,969 fibers in SA-111, 14.1% of 12,217 fibers in SA-IV, and 21.1% of 4,098fibers in SA-V belong to the beta fiber s u e range. In SA-P, however, 40.3% of its 3,922 fibers fall into the beta size range. In the additional study (table 2), SA-111 had 5.2% of its 761 fibers, SA-IV 13.5% of its 507 fibers, and SA-V 28.1% of the 331 fibers belonging to the beta size range, while SA-P had 35.6% of its 295 fibers falling into the beta size range. DISCUSSION
One of the more prominent results obtained from the present study is the uniqueness of SA-P. This subarea is characterized by being located at the ventral lateral portion of the identifiable fasciculus gracilis,
199
and showed a bimodal histogramwith a wide range of fiber sizes. Such characteristics are clearly in contrast with all three other subareas (figs. 3, 4; tables 1, 2), and, therefore, the question arose whether to include SA-P in the fasciculus gracilis. Inclusion of SA-P into the present study was due to the lack of a clearly definable boundary between the two fasciculi of the dorsal column at cervical level of the cat spinal cord. For this reason, the area was taken into account in this study since there was an insignificant boundary existing between SA-P and the remaining fasciculus cuneatus (fig. 2). The region of the fasciculus gracilis is frequently defined as an area demarcated medially by the dorsal median septum and laterally by the dorsal intermediate septum of the spinal cord (e.g., Ranson and Clark, '59). The dorsal median septum is a well defined boundary, however, the dorsal intermediate septum at the cervical level of cat spinal cord projects regularly into the white matter for only a short distance and fades away without making contact with other definable structures (fig. 2). Therefore the ventral portion of the fasciculus is difficult to distinguish from its neighboring fasciculi. 30 25
20
~II
8
10
5
0
1 2 3 4
10 5
0
30
331
25
i7891011
1
2 3 4 5 6 7 8 91011
35
SA-Y
bL
!5
15
20 15
: l
0
1 2 3 4 5 6 7 8 9 1 0 1 1
FIBER DIAMETER (pl
Fig. 4 Histograms of Fasciculus Gracilis of cat No. 162 at C3 level.
YANN-CHING H W A N G , EDWARD J. HINSMAN A N D 0 . F. ROESEL
200
TABLE 1
Diameter percentage of t h e caliber spectra of animal no. 526. T h e u p p e r line representing percent (55) i n e a c h diameter range; the lower line representing its accumulative percent ( % ) Fiber diameter ranges ( p ) Areas
1
2
3
4
5
6
7
8
9
10
SA-I11
5.8 5.8
20.6 26.4
25.8 52.2
21.8 74.0
13.5 87.5
7.2 94.7
3.3 98.0
1.4 99.4
0.4 99.8
0.1 99.9
0.0 99.9
SA-IV
4.3 4.3
14.3 18.6
20.3 38.9
28.9 67.8
18.1 85.9
8.8 94.7
3.1 97.8
1.4 99.2
0.5 99.7
0.2 99.9
0.1 100.0
SA-V
4.4 4.4
10.2 14.6
13.4 28.0
21.0 49.0
28.9 77.9
15.1 93.0
4.0 97.0
2.4 99.4
0.6 100.0
SA-P
6.5 6.5
13.2 19.7
11.7 31.4
14.3 45.7
13.8 59.5
10.7 70.2
10.0 80.2
7.9 88.1
4.7 92.8
3.4 96.2
2.4 98.6
11
12
13
0.1 100.0
1.2 99.8
0.2 100.0
TABLE 2
Diameter percentage of t h e caliber spectra of animal no. 162. The u p p e r line representing percent ( r4 of e a c h diameter range; the lower line representing its accumulative percent (‘r,)
)
Fiber diameter ranges (/A) Areas
1
2
3
4
5
6
7
8
9
SA-111
4.1 4.1
27.9 32.0
23.9 55.9
24.3 80.2
14.6 94.8
3.1 98.9
1.6 99.5
0.0 99.5
0.3 99.8
0.1 99.9
SA-IV
0.0 0.0
8.9 8.9
19.7 28.6
32.1 60.7
25.8 86.5
10.3 96.8
2.4 99.2
0.2 99.4
0.4 99.8
0.2 100.0
SA-V
0.3 0.3
3.0 3.3
7.6 10.9
27.2 38.1
33.8 71.9
11.2 83.1
6.0 89.1
6.4 95.5
3.3 98.8
1.2 100.0
SA-P
9.5 9.5
12.9 22.4
13.6 36.0
16.9 52.9
11.5 64.4
9.1 73.5
10.8 84.3
7.5 91.8
5.1 96.9
2.4 99.3
Such a problem, although often neglected, was noticed by Sherrington (1893) who reported that the dorsal intermediate septum in some places is not an absolute guide to limit the fasciculus gracilis. Therefore SA-P of the present study is probably not a part of the fasciculus gracilis, nevertheless, it was included into our primary consideration. As we have observed (fig. 2), SA-I11 is a band-like structure extending from surroundings of the short dorsal intermediate septum obliquely toward the deeper portion of the dorsal median septum. Such an observation was also made by Sherrington in not only the cervical cord sections of the cat but also in the dog, monkey and of man. He named the band-like structure the “band of condensation” in which more densely packed uniform nerve fibers were present and suggested that this band is a guide to the boundary between gracile and cuneate fas-
10
11
0.1 100.0
0.7 100.0
ciculi. We, therefore, concluded that in the cat cervical spinal cord the lateral edge of SA-I11 serves as a more reliable boundary between the two major fasciculi of the dorsal column, and SA-P is not a part of the fasciculi gracilis. Fiber distribution pattern of the fasciculus gracilis The fasciculus gracilis is generally considered a tract in which the fibers are uniform in size (e.g., Glees, ’61). Because of such an understanding, it is logical to imply results obtained from partial sampling studies as conclusive anatomical observations of the fasciculus. Nevertheless, results from such studies are vague. Haggqvist (’36) reported a bimodal caliber spectrum with its major peak at 1-3 p from T3 of a 13 year-old girl. Szentagothai-Schimert (’41), on the other hand, reported a unimodal histogram in six human gracile
CALIBER SPECTRUM OF F. GRACILIS IN CAT
fasciculi at Cs level with its peak at 3-5 p range. In the cat fasciculus gracilis Hildebrand and Skoglund ('71) observed a bimodal caliber spectrum with its peak at 1-2 p and 3 4 p. To avoid such misinterpretation, our study covered the whole fasciculus: and to obtain more accurate results the electron microscope, planimeter, and the computer were introduced. Our results thus reveal that the fasciculus gracilis of the cat at the cervical level does not have a homogeneous fiber distribution, rather it is characterized by three different regional spectra. Although all of them show unimodal histograms, their peaks vary from one to another (figs. 3, 4), and a unique fiber distribution pattern is therefore seen. Our caliber spectra are obviously in contrast with Hildebrand and Skoglund's which show bimodal histograms and about 95% of the total fiber count with diameter smaller than 5 p . Since methods of fixation and processing for epon preparations in both studies are very similar, most of the conflicting results undoubtedly come from the different methods of measurement. In the present study the area of each axon was measured and converted into a diameter under an assumption that the axon is round, then twice the thickness of the myelin sheath was added to form the final diameter of the nerve fiber. Greater accuracy of this method of measurement was proved (Hwang, '73). On the other hand, Hildebrand and Skoglund measured only the shortest diameters of the fibers with oval or pear shapes, and disregarded fibers with higher degree of asymmetry. Since data obtained from our study showed only 15% of fibers symmetrically shaped, the bias of their histogram evidently would incline toward the left of the fiber size distribution, and possibly produced the bimodal appearance of their histogram. Our fiber distribution pattern although in conflict with the trditional view of the fasciculus, does show a positive correlation between it and the recent neuroanatomical and electrophysiological studies (Whitsel et al., '70, '72). They reported a highly specific fiber re-sorting process taking place between lumbar and cervical levels of the fasciculus gracilis. Such an observation undoubtedly establishes the possibility that the caliber spectrum of the fasciculus at
20 1
the cervical level may undergo some kind of modification. The unique fiber distribution pattern, therefore, can be interpreted as a result of the fiber re-sorting which reflects the transformation of the segmental fiber arrangement into the more functionally oriented fiber arrangement at the cervical fasciculus gracilis. In our primary study, there were 25,284 fibers counted in the fasciculus gracilis (not including fibers from SA-P). Since the great majority of them are primary sensory fibers and according to Glees and Soler ('51) only 22-25% of the primary fibers reach their relay nucleus in the cat; we therefore, imply that in the cat about 100,000 fibers originate from the unilateral spinal ganglia and extend into the ipsilateral fasciculus gracilis. It is generally suggested that the beta fibers are the only group remaining as the primary sensory component in the dorsal column (Bishop, '64; Mountcastle and Darian-Smith, '68; Goldberg and Lavine, '68). However several caliber spectral studies showed the majority of fibers fall into the delta size range (Haggqvist, '36; Sentagothai, '41; Hildebrand and Skoglund, '71). In addition, several neurophysiological studies demonstrated reduced conduction velocities in the dorsal columns, and suggested that the size of fibers in the columns is smaller than in peripheral nerves (Lloyd and McIntyre, '50; Holmgren, '54; Brown, '68; Petit and Burgess, '68). In Petit and Burgess' study, for example, peripheral conduction velocities for the cat's mlsec., sural fibers ranged from 4-5 while in dorsal column (C, to TI,,) the conduction velocities for the same sural fibers ranged from 1 4 5 m/sec. The conduction velocity (mlsec.) for cat peripheral myelinated fibers is six times their fiber diameter in microns (Hursh, '39), and Bishop and Clare ('55), and Chang ('56) further indicated that a factor of six also holds reasonably well for cat central system. Therefore, the corresponding fiber diameters in the periphery of Petit and Burgess' results would be 7-14 N , whereas in the dorsal column 2-7 p. In terms of fiber classification, the majority of the dorsal column fibers will be in the delta fiber size range (2-5 p ) instead of the beta P).
In the area of SA-111, SA-IV, and SA-V
202
YANN-CHING HWANG, EDWARD J. HINSMAN AND 0. F. ROESEL
DORSAL MEDIAN SULCUS
..
DORSAL INTERMEDIATE SULCUS SAIP
,,,‘ \
SA-P ,,,’
\
---..---,,
fibers; SA-IV in the dorsal portion of the fasciculus is dominated by 4 p diameter fibers; SA-V medially located has more larger caliber (5 p ) fibers. However, all of the three subareas of the fasciculus show unimodal histograms. The majority of fibers in the fasciculus are in the delta rather than the beta size range.
,,----_
SA-IU,,,,,,’
FASC IC U LUS CU N EATUS
Fig. 5 Schematic Demarcation of the Cat Fasciculus Gracilis at CBlevel.
of the present study fiber diameters ranged from less than 1 p to 15 p , 97% of fibers having their diameters less than 8 p , and 80 % or more of fiber diameters being smaller than 6 p . Our results hold quite a satisfactory correlation with the facts obtained from electrophysiological study of the dorsal columns. In the area of SA-P, however, only about 60% of the fiber diameters fall into the range of 1-5 p , and the remaining 40% are in the beta size range. This result further indicates that SA-P differs from the three other subareas not only by its topographical location but also by its unique fiber spectrum. There was a slightly higher percentage of unmyelinated fibers encountered in our study. However, it is difficult to evaluate such a result, since in cross sections of the fibers many “unmyelinated fibers might simply result from being cut through the nodes of Ranvier. This problem, therefore, requires further investigation. CONCLUSION
The present study clearly demonstrated that the fasciculus gracilis of the cat at C, level has a unique fiber distribution pattern: SA-I11 located dorsolaterally as a band-like zone has more smaller caliber
ACKNOWLEDGMENTS
The authors acknowledge Dr. M. W. Stromberg for his critical reading of the manuscript. We are indebted to Mr. C. A. Lynn for photographic processing and Mr. N. Harris for his art work. LITERATURE CITED Bishop, G. H. 1964 Fiber size and myelinization in afferent systems. In: Henry Ford Hospital International Symposium of Pain. R. S. Knighton, ed. Little, Brown & Co., Boston, pp. 83439. Bishop, G. H., and M. H. Clare 1955 Organization and distribution of fibers i n the optic tract of the cat. J. Comp. Neur., 103: 264-304. Brown, A. G. 1968 Cutaneous afferent fiber collaterals in the dorsal columns of the cat. Expl. Brain Res., 5 : 293-305. Carpenter, M. B., B. M. Stein and J. E. Shriver 1968 Central projections of spinal dorsal roots i n the monkey. 11. Lower thoracic, lumbosacral and coccygeal dorsal roots. Am. J. Anat., 123:
75-1 18.
Chang, H. T. 1956 Fiber groups in primary optic pathway of cat. J. Neurophysiol., 19: 224-
231.
Gay, H., and T. F. Anderson 1954 Serial sections for electron microscopy. Science, 120:
1071-1 073.
Glees, P. 1961 Spinal cord. Experimental Neurology. Clarendon Press, Oxford, p. 200. Glees, P., and J. Soler 1951 Fibre content of the posterior column and synaptic connections of nucleus gracilis. Z. Zellfors., 36: 381-400. Goldberg, J. M.,and R. A. Lavine 1968 Nervous system: Afferent mechanisms. Ann. Rev. Physiol., 30: 31S358. Gonzalez-Aguilar, F., and E. DeRobertis 1963 A formalin-perfusion fixation method for histophysiological study of the central nervous system with the electron microscope. Neurology,
3: 758-771.
Haggqvist, G. 1936 Analyse der Faserverteilung i n einem Ruckenmarkquerschnitt (Th 3). Z. Mikrosk.-anat. Forsch., 39: 1-34. Hildebrand, C., and S. Skoglund 1971 Calibre spectra of some fiber tracts in the feline central nervous system during postnatal development. Acta Physiol. Scand., Suppl. 364: 5-42. Holmgren, B. 1954 Conduction along the dorsal tracts of the spinal cord. J. Physiol. (London),
123:324-337.
Hursh, J . B. 1939 Conduction velocity and diameter of nerve fibers. Am. J. Physiol., 127:
131-1 39.
CALIBER SPECTRUM OF F. GRACILE IN CAT Hwang, Y. C. 1970 A modification for orientation by the use of silicone rubber molds for embedding tissue in epoxy resins. J. Electron Microscopy, 19: 18S190. 1973 A quantitative electron microscope study of the fasciculus gracilis of the cat cervical spinal cord. Ph.D. thesis, West Lafayette, Ind. 1973,Purdue University. Karnovsky, M. J. 1965 A formaldehyde-glutaraldehyde fixative of high-osmolarity for use in electron microscopy. J. Cell Biol., 27: 137A. Kodama, Y. 1942 Uber das zentrale Verhalten der Hinterwurzelfasern bei der Katze. Okajimas Folia Anatomic Japonica, 21: 291-334. Lloyd, D. P. C., and A. K. McIntyre 1950 Dorsal column conduction of group 1 muscle afferent impulses and their relay through Clarke’s column. J. Neurophysiol., 13: 3S54. Mountcastle, V. B., and I. Darian-Smith 1968 Neural mechanisms in somesthesia. In: Medical Physiology. V. B. Mountcastle, ed. C. V. Mosby Co., St. Louis, p. 1378. Padgett, F. 1963 Some observations on new fixation and staining procedures of biological material and on a new supporting film. Scienti. Instru., 8:8-11. Petit, D., and P. R. Burgess 1968 Dorsal column projection of receptors in cat hairy skin supplied by myelinated fibers. J. Neurophysiol., 31: 84-55,
203
Ranson, S. W., and S. L. Clark 1959 The Anatomy of the Nervous System. W. B. Saunders Co., Philadelphia, p. 169. Sherrington, C . S. 1893 Note on the spinal portion of some ascending degenerations. J. Physiol. (London), 14: 255-302. Szentagothai-Schimert, J . 1941 Die Bedeutung des Faserkalibers und Markscheidendicke in Zentralnervensystem. Z. Anat. Entwicklungsch, 1 1 1 : 201-223. Venable, J. H.,and R. Coggeshall 1965 A simplified lead citrate stain for use in electron microscopy. J. Cell Biol., 25: 407408. Walker, A . E., and T. A. Weaver, Jr. 1942 The topical organization and termination of the fibers of the posterior columns in Macaca mulatta. J. Comp. Neur., 76: 145-158. Watson, M. L. 1958 Staining of tissue sections for electron microscopy with heavy metals. J. Biophys. Biochem. Cytol., 4: 475-478. Whitsel, B. L., L. M. Petrucelli, G. Sapiro and H. Ha 1970 Fiber sorting in the fasciculus gracilis of squirrel monkeys. Expl. Neurol., 29: 227-242. Whitsel, B. L., L. M. Petrucelli, H. Ha and D. A. Dreyer 1972 The resorting of spinal afferents as antecedent to the body representation in the postcentral gyrus. Brain Behav. Evol., 5: 303341.
5
ABSTRACT In order to obtain a better understanding of the microscopic structure of the cat fasciculus gracilis, entire cross-sections of the fasciculus were examined with the electron microscope. Twenty-five thousand, two hundred and eighty-four fibers were encountered in one fasciculus. The fiber caliber spectra obtained from the study show that the fasciculus gracilis at cervical level has a unique fiber distribution pattern. The fiber diameters range from less than 1 p to 15 p , however, 97% of fibers have diameters less than 8 p ; and the majority of the fibers are in the 2-5 p range.
It is generally acknowledged that the ascending fibers of the fasciculus gracilis maintain an orderly arrangement of lamination representing dermatomal layering of the body. However, overlap of the lamination has been noticed (Kodama, '42; Walker and Weaver, '42; Carpenter et al., '68). Recently, Whitsel and his coworkers ('70, '72) reported that fibers of the fasciculus gracilis re-sort while they extend from the lumbosacral to the cervical portions, and the map of the hind limb in the cervical gracile fasciculus more closely resembles the map of the hind limb in sensory cortical areas than does the representation at lumbar levels of the fasciculus. On the other hand, fibers of the fasciculus gracilis are believed to cluster in the beta (group 11) size range (Mountcastle and Darian-Smith, '68), which is in conflict with the results of several fiber spectral studies involving this fasciculus in both human and the cat (Haggqvist, '36; Szentagothai-Schimert, '4 1; Hildebrand and Skoglund, '71). The present study, by employing techniques of quantitative electron microscopy and with the aid of the computer, was to reinvestigate fiber distributional pattern and fiber size range within the fasciculus gracilis of the cat at cervical level. MATERIALS AND METHODS
Adult cats raised in the colony of the Department of Veterinary Anatomy, PurJ.
COMP.
NEUR., 162: 1 9 5 2 0 4
due University were used. Cats were perfused by Karnovsky's ('65) immersion fixative under deep sodium pentobarbital anesthesia following a modified retrograde aortic perfusion technique of GonzalezAguilar and DeRobertis ('63). The spinal cord at C3 level was removed and selected parts of the dorsal column were dissected free and trimmed. Pieces of tissue were then postfixed in 2 % osmium tetroxide, and after alcoholic dehydration embedded in Epon 812. For consistently producing precise cross sections of the tissue, a special embedding technique (Hwang, '70) was followed. The thin epon section was cut perpendicular to the long axis of the fasciculus and picked up on a loop of thin film (Padgett, '63). Each section with film was then precisely mounted on a slot grid (Gay and Anderson, '54). Sections were double stained with uranyl acetate (Watson, '58) and lead citrate (Venable and Coggeshall, '65) before examination, and photographed with a Philips 200 electron microscope. In the primary case (cat no. 526), 297 electron micrographs, covering most of the fasciculus gracilis and vicinity, were taken from two consecutive serial sections at X 1,900. The negatives were then project1 Published as Paper Number 5744, Purdue Agriculture Experiment Station, West Lafayette, Indiana. 2 Research support in part from Purdue Agricultural Experiment Station -project 1418-36-1174. 3 Present address: Department of Anatomy, School of Medicine, Howard University, Washington, D. C. 20001.
195
196
YANN-CHING HWANG, EDWARD J . HINSMAN A N D 0. F. ROESEL
ed at X 10,000 onto a tracing board, an image of each axon circumference was traced, and myelin sheath thickness was marked on tracing paper. With the aid of printed photographs ( x 4,000), each fiber on the tracing paper was identified, and its circumference was measured by a OTT planimeter, and myelin sheath thickness measured by a millimeter ruled scale. Data of each fiber were then processed with computer assistance. A digital computer program written by one of the authors (0. F. Roesel) was used to aid in computation of the fiber diameter and in formation of the caliber histogram of each photographic negative. By studying primarily the peak and the mode of each histogram, a caliber spectral chart (fig. 1) was made. Based upon this chart, the total area was partitioned into four groups (subareas). The group’s grand histogram and fiber diameter range were then produced by the aid of the computer. In another case (cat no. 162), only four electron photographs were taken from each group, however, all similar data processing was followed. The thick epon sections for cat no. 526 and cat no. 162, together with five other animals’ were examined under the light microscope for comparison of patterns of fiber distribution within fasciculus gracilis, and boundaries of the two fasiculi of the dorsal column. RESULTS
In the primary experiment (cat no. 526), a total area of about 541,000 pL2of tissue was studied, which covered most of the fasciculus gracilis and some of its vicinity. The total nerve fiber count in this area is 29,206. As we already mentioned, the caliber histogram chart (fig. 1) was produced from the primary computer manipulation which considered each electron photomicrographic negative as a sector unit and a histogram was produced for each sector unit. Because of regional differences observed in the caliber histogram chart, the area was partitioned into four subareas (Sa): SA-111, SA-IV, SA-V, and SA-P. A demarcation of the four subareas is shown in figure 2. Based upon these four subareas, the caliber histograms and the fiber
diameter range of both the primary and the additional studies were then considered. Caliber his togram s The histograms of all four subareas of both cat no. 526 and cat no. 162 are shown in figures 3 and 4. In the primary study (cat no. 526) histograms of SA-111, SA-IV, and SA-V have prominent unimodal peaks which exceed 25% of the total number of fibers involved; on the other hand, the histogram of SA-P is bimodal with a more gradual slope. The peak of each subarea was used to designate that particular subarea; i.e., SA-I11 has its peak at 3 p , SA-IV at 4 p , and SA-V at 5 p ranges, however, P of the SA-P symbolizes a mosaic grand polymodal histogram. Although the peak of SA-I11 histogram locates at 3 p , there are also fairly high percentages of fibers at 2 p and 4 p diameter ranges. This is primarily because the subarea includes not only the individual sector unit histograms with peaks at 3 p but also the sector unit histograms with peaks at at 2 p and some at 4 p . The histograms of SA-IV and SA-V both show very sharp unimodal peaks. In contrast, the histogram of SA-P is unique. The area of SA-P as seen in figure 1 is not homogeneous, but an area with different pattern of histograms and fiber diameter peaks. When all these different histograms were put together, the mosaic grand histogram for the area of SA-P was then formed, which was characterized by wide spreading fiber diameter range with its major peak at 4 p. In cat no. 162, four areas (approximately 10,000 p2 each) identifiable as SA-111, SA-IV, SA-V and SA-P were sampled. Histograms (fig. 4) from each area, in general, do agree with the results of cat no. 526 (fig. 3). However, the following differences were noticed: (1) The peak of SA-111 histogram shifted to 2 p . This result is mostly due to difficulty encountered in sampling such a 10,000 p2 area from a slender band-like SA-I11 (fig. 2). To ensure the electron photomicrographic sample was inside the subarea, we had no choice but to take the photographs from the most wide open portion of the SA-111, which is the more dorsolateral portion of this subarea where more fine fibers were seen (fig.
CALIBER SPECTRUM OF F. GRACILIS I N CAT
Fig. 1
Caliber Specbal Chart of the Cat Fasciculus Gracilis.
197
198
YANN-CHING HWANG, EDWARD J. HINSMAN AND 0. F. ROESEL
Fig. 2 Partition of the Fasciculus Gracilis and Vicinity. The area photographed is partitioned into four subareas according to the caliber spectral chart (fig. 1). The fasciculus gracilis is bounded between the dorsal median sulcus (DMS) and septum and dorsal intermediate sulcus (DIS), however, the boundary between the fasciculus and the fasciculus cuneatus is primarily the lateral edge of SA-111. (III), Band-like area of SA-111 with more 2 p or 3 p fibers; (IV) Band-like area of SA-IV predominated by 4 p fibers; (V), Band-like area of SA-V with more 5 p fibers; (P). Band-like area of SA-P which has unique fiber distributional pattern and is not considered as a part of the fasciculus; (FC), part of the fasciculus cuneatus.
CALIBER SPECTRUM OF F. GRACILIS I N CAT
FIBER DIAMETER ( p i
Fig. 3 Histograms of Fasciculus Gracilis of cat No. 526 at Cs level.
1). (2) The most bimodal peaks of SA-P are located at 4 p and 7 CL instead of 4 p and 2 p . However, in both animals, the major peaks are at 4 p . Fiber diameters Diameter percentages of the caliber spectra of cats no. S26 and no. 162 are shown in tables 1 and 2. On the primary study (table 1) the majority of the fibers within SA-111, SA-IV and SA-V belong to the delta fiber size range (2p-5 F ) , while only 12.5% of 8,969 fibers in SA-111, 14.1% of 12,217 fibers in SA-IV, and 21.1% of 4,098fibers in SA-V belong to the beta fiber s u e range. In SA-P, however, 40.3% of its 3,922 fibers fall into the beta size range. In the additional study (table 2), SA-111 had 5.2% of its 761 fibers, SA-IV 13.5% of its 507 fibers, and SA-V 28.1% of the 331 fibers belonging to the beta size range, while SA-P had 35.6% of its 295 fibers falling into the beta size range. DISCUSSION
One of the more prominent results obtained from the present study is the uniqueness of SA-P. This subarea is characterized by being located at the ventral lateral portion of the identifiable fasciculus gracilis,
199
and showed a bimodal histogramwith a wide range of fiber sizes. Such characteristics are clearly in contrast with all three other subareas (figs. 3, 4; tables 1, 2), and, therefore, the question arose whether to include SA-P in the fasciculus gracilis. Inclusion of SA-P into the present study was due to the lack of a clearly definable boundary between the two fasciculi of the dorsal column at cervical level of the cat spinal cord. For this reason, the area was taken into account in this study since there was an insignificant boundary existing between SA-P and the remaining fasciculus cuneatus (fig. 2). The region of the fasciculus gracilis is frequently defined as an area demarcated medially by the dorsal median septum and laterally by the dorsal intermediate septum of the spinal cord (e.g., Ranson and Clark, '59). The dorsal median septum is a well defined boundary, however, the dorsal intermediate septum at the cervical level of cat spinal cord projects regularly into the white matter for only a short distance and fades away without making contact with other definable structures (fig. 2). Therefore the ventral portion of the fasciculus is difficult to distinguish from its neighboring fasciculi. 30 25
20
~II
8
10
5
0
1 2 3 4
10 5
0
30
331
25
i7891011
1
2 3 4 5 6 7 8 91011
35
SA-Y
bL
!5
15
20 15
: l
0
1 2 3 4 5 6 7 8 9 1 0 1 1
FIBER DIAMETER (pl
Fig. 4 Histograms of Fasciculus Gracilis of cat No. 162 at C3 level.
YANN-CHING H W A N G , EDWARD J. HINSMAN A N D 0 . F. ROESEL
200
TABLE 1
Diameter percentage of t h e caliber spectra of animal no. 526. T h e u p p e r line representing percent (55) i n e a c h diameter range; the lower line representing its accumulative percent ( % ) Fiber diameter ranges ( p ) Areas
1
2
3
4
5
6
7
8
9
10
SA-I11
5.8 5.8
20.6 26.4
25.8 52.2
21.8 74.0
13.5 87.5
7.2 94.7
3.3 98.0
1.4 99.4
0.4 99.8
0.1 99.9
0.0 99.9
SA-IV
4.3 4.3
14.3 18.6
20.3 38.9
28.9 67.8
18.1 85.9
8.8 94.7
3.1 97.8
1.4 99.2
0.5 99.7
0.2 99.9
0.1 100.0
SA-V
4.4 4.4
10.2 14.6
13.4 28.0
21.0 49.0
28.9 77.9
15.1 93.0
4.0 97.0
2.4 99.4
0.6 100.0
SA-P
6.5 6.5
13.2 19.7
11.7 31.4
14.3 45.7
13.8 59.5
10.7 70.2
10.0 80.2
7.9 88.1
4.7 92.8
3.4 96.2
2.4 98.6
11
12
13
0.1 100.0
1.2 99.8
0.2 100.0
TABLE 2
Diameter percentage of t h e caliber spectra of animal no. 162. The u p p e r line representing percent ( r4 of e a c h diameter range; the lower line representing its accumulative percent (‘r,)
)
Fiber diameter ranges (/A) Areas
1
2
3
4
5
6
7
8
9
SA-111
4.1 4.1
27.9 32.0
23.9 55.9
24.3 80.2
14.6 94.8
3.1 98.9
1.6 99.5
0.0 99.5
0.3 99.8
0.1 99.9
SA-IV
0.0 0.0
8.9 8.9
19.7 28.6
32.1 60.7
25.8 86.5
10.3 96.8
2.4 99.2
0.2 99.4
0.4 99.8
0.2 100.0
SA-V
0.3 0.3
3.0 3.3
7.6 10.9
27.2 38.1
33.8 71.9
11.2 83.1
6.0 89.1
6.4 95.5
3.3 98.8
1.2 100.0
SA-P
9.5 9.5
12.9 22.4
13.6 36.0
16.9 52.9
11.5 64.4
9.1 73.5
10.8 84.3
7.5 91.8
5.1 96.9
2.4 99.3
Such a problem, although often neglected, was noticed by Sherrington (1893) who reported that the dorsal intermediate septum in some places is not an absolute guide to limit the fasciculus gracilis. Therefore SA-P of the present study is probably not a part of the fasciculus gracilis, nevertheless, it was included into our primary consideration. As we have observed (fig. 2), SA-I11 is a band-like structure extending from surroundings of the short dorsal intermediate septum obliquely toward the deeper portion of the dorsal median septum. Such an observation was also made by Sherrington in not only the cervical cord sections of the cat but also in the dog, monkey and of man. He named the band-like structure the “band of condensation” in which more densely packed uniform nerve fibers were present and suggested that this band is a guide to the boundary between gracile and cuneate fas-
10
11
0.1 100.0
0.7 100.0
ciculi. We, therefore, concluded that in the cat cervical spinal cord the lateral edge of SA-I11 serves as a more reliable boundary between the two major fasciculi of the dorsal column, and SA-P is not a part of the fasciculi gracilis. Fiber distribution pattern of the fasciculus gracilis The fasciculus gracilis is generally considered a tract in which the fibers are uniform in size (e.g., Glees, ’61). Because of such an understanding, it is logical to imply results obtained from partial sampling studies as conclusive anatomical observations of the fasciculus. Nevertheless, results from such studies are vague. Haggqvist (’36) reported a bimodal caliber spectrum with its major peak at 1-3 p from T3 of a 13 year-old girl. Szentagothai-Schimert (’41), on the other hand, reported a unimodal histogram in six human gracile
CALIBER SPECTRUM OF F. GRACILIS IN CAT
fasciculi at Cs level with its peak at 3-5 p range. In the cat fasciculus gracilis Hildebrand and Skoglund ('71) observed a bimodal caliber spectrum with its peak at 1-2 p and 3 4 p. To avoid such misinterpretation, our study covered the whole fasciculus: and to obtain more accurate results the electron microscope, planimeter, and the computer were introduced. Our results thus reveal that the fasciculus gracilis of the cat at the cervical level does not have a homogeneous fiber distribution, rather it is characterized by three different regional spectra. Although all of them show unimodal histograms, their peaks vary from one to another (figs. 3, 4), and a unique fiber distribution pattern is therefore seen. Our caliber spectra are obviously in contrast with Hildebrand and Skoglund's which show bimodal histograms and about 95% of the total fiber count with diameter smaller than 5 p . Since methods of fixation and processing for epon preparations in both studies are very similar, most of the conflicting results undoubtedly come from the different methods of measurement. In the present study the area of each axon was measured and converted into a diameter under an assumption that the axon is round, then twice the thickness of the myelin sheath was added to form the final diameter of the nerve fiber. Greater accuracy of this method of measurement was proved (Hwang, '73). On the other hand, Hildebrand and Skoglund measured only the shortest diameters of the fibers with oval or pear shapes, and disregarded fibers with higher degree of asymmetry. Since data obtained from our study showed only 15% of fibers symmetrically shaped, the bias of their histogram evidently would incline toward the left of the fiber size distribution, and possibly produced the bimodal appearance of their histogram. Our fiber distribution pattern although in conflict with the trditional view of the fasciculus, does show a positive correlation between it and the recent neuroanatomical and electrophysiological studies (Whitsel et al., '70, '72). They reported a highly specific fiber re-sorting process taking place between lumbar and cervical levels of the fasciculus gracilis. Such an observation undoubtedly establishes the possibility that the caliber spectrum of the fasciculus at
20 1
the cervical level may undergo some kind of modification. The unique fiber distribution pattern, therefore, can be interpreted as a result of the fiber re-sorting which reflects the transformation of the segmental fiber arrangement into the more functionally oriented fiber arrangement at the cervical fasciculus gracilis. In our primary study, there were 25,284 fibers counted in the fasciculus gracilis (not including fibers from SA-P). Since the great majority of them are primary sensory fibers and according to Glees and Soler ('51) only 22-25% of the primary fibers reach their relay nucleus in the cat; we therefore, imply that in the cat about 100,000 fibers originate from the unilateral spinal ganglia and extend into the ipsilateral fasciculus gracilis. It is generally suggested that the beta fibers are the only group remaining as the primary sensory component in the dorsal column (Bishop, '64; Mountcastle and Darian-Smith, '68; Goldberg and Lavine, '68). However several caliber spectral studies showed the majority of fibers fall into the delta size range (Haggqvist, '36; Sentagothai, '41; Hildebrand and Skoglund, '71). In addition, several neurophysiological studies demonstrated reduced conduction velocities in the dorsal columns, and suggested that the size of fibers in the columns is smaller than in peripheral nerves (Lloyd and McIntyre, '50; Holmgren, '54; Brown, '68; Petit and Burgess, '68). In Petit and Burgess' study, for example, peripheral conduction velocities for the cat's mlsec., sural fibers ranged from 4-5 while in dorsal column (C, to TI,,) the conduction velocities for the same sural fibers ranged from 1 4 5 m/sec. The conduction velocity (mlsec.) for cat peripheral myelinated fibers is six times their fiber diameter in microns (Hursh, '39), and Bishop and Clare ('55), and Chang ('56) further indicated that a factor of six also holds reasonably well for cat central system. Therefore, the corresponding fiber diameters in the periphery of Petit and Burgess' results would be 7-14 N , whereas in the dorsal column 2-7 p. In terms of fiber classification, the majority of the dorsal column fibers will be in the delta fiber size range (2-5 p ) instead of the beta P).
In the area of SA-111, SA-IV, and SA-V
202
YANN-CHING HWANG, EDWARD J. HINSMAN AND 0. F. ROESEL
DORSAL MEDIAN SULCUS
..
DORSAL INTERMEDIATE SULCUS SAIP
,,,‘ \
SA-P ,,,’
\
---..---,,
fibers; SA-IV in the dorsal portion of the fasciculus is dominated by 4 p diameter fibers; SA-V medially located has more larger caliber (5 p ) fibers. However, all of the three subareas of the fasciculus show unimodal histograms. The majority of fibers in the fasciculus are in the delta rather than the beta size range.
,,----_
SA-IU,,,,,,’
FASC IC U LUS CU N EATUS
Fig. 5 Schematic Demarcation of the Cat Fasciculus Gracilis at CBlevel.
of the present study fiber diameters ranged from less than 1 p to 15 p , 97% of fibers having their diameters less than 8 p , and 80 % or more of fiber diameters being smaller than 6 p . Our results hold quite a satisfactory correlation with the facts obtained from electrophysiological study of the dorsal columns. In the area of SA-P, however, only about 60% of the fiber diameters fall into the range of 1-5 p , and the remaining 40% are in the beta size range. This result further indicates that SA-P differs from the three other subareas not only by its topographical location but also by its unique fiber spectrum. There was a slightly higher percentage of unmyelinated fibers encountered in our study. However, it is difficult to evaluate such a result, since in cross sections of the fibers many “unmyelinated fibers might simply result from being cut through the nodes of Ranvier. This problem, therefore, requires further investigation. CONCLUSION
The present study clearly demonstrated that the fasciculus gracilis of the cat at C, level has a unique fiber distribution pattern: SA-I11 located dorsolaterally as a band-like zone has more smaller caliber
ACKNOWLEDGMENTS
The authors acknowledge Dr. M. W. Stromberg for his critical reading of the manuscript. We are indebted to Mr. C. A. Lynn for photographic processing and Mr. N. Harris for his art work. LITERATURE CITED Bishop, G. H. 1964 Fiber size and myelinization in afferent systems. In: Henry Ford Hospital International Symposium of Pain. R. S. Knighton, ed. Little, Brown & Co., Boston, pp. 83439. Bishop, G. H., and M. H. Clare 1955 Organization and distribution of fibers i n the optic tract of the cat. J. Comp. Neur., 103: 264-304. Brown, A. G. 1968 Cutaneous afferent fiber collaterals in the dorsal columns of the cat. Expl. Brain Res., 5 : 293-305. Carpenter, M. B., B. M. Stein and J. E. Shriver 1968 Central projections of spinal dorsal roots i n the monkey. 11. Lower thoracic, lumbosacral and coccygeal dorsal roots. Am. J. Anat., 123:
75-1 18.
Chang, H. T. 1956 Fiber groups in primary optic pathway of cat. J. Neurophysiol., 19: 224-
231.
Gay, H., and T. F. Anderson 1954 Serial sections for electron microscopy. Science, 120:
1071-1 073.
Glees, P. 1961 Spinal cord. Experimental Neurology. Clarendon Press, Oxford, p. 200. Glees, P., and J. Soler 1951 Fibre content of the posterior column and synaptic connections of nucleus gracilis. Z. Zellfors., 36: 381-400. Goldberg, J. M.,and R. A. Lavine 1968 Nervous system: Afferent mechanisms. Ann. Rev. Physiol., 30: 31S358. Gonzalez-Aguilar, F., and E. DeRobertis 1963 A formalin-perfusion fixation method for histophysiological study of the central nervous system with the electron microscope. Neurology,
3: 758-771.
Haggqvist, G. 1936 Analyse der Faserverteilung i n einem Ruckenmarkquerschnitt (Th 3). Z. Mikrosk.-anat. Forsch., 39: 1-34. Hildebrand, C., and S. Skoglund 1971 Calibre spectra of some fiber tracts in the feline central nervous system during postnatal development. Acta Physiol. Scand., Suppl. 364: 5-42. Holmgren, B. 1954 Conduction along the dorsal tracts of the spinal cord. J. Physiol. (London),
123:324-337.
Hursh, J . B. 1939 Conduction velocity and diameter of nerve fibers. Am. J. Physiol., 127:
131-1 39.
CALIBER SPECTRUM OF F. GRACILE IN CAT Hwang, Y. C. 1970 A modification for orientation by the use of silicone rubber molds for embedding tissue in epoxy resins. J. Electron Microscopy, 19: 18S190. 1973 A quantitative electron microscope study of the fasciculus gracilis of the cat cervical spinal cord. Ph.D. thesis, West Lafayette, Ind. 1973,Purdue University. Karnovsky, M. J. 1965 A formaldehyde-glutaraldehyde fixative of high-osmolarity for use in electron microscopy. J. Cell Biol., 27: 137A. Kodama, Y. 1942 Uber das zentrale Verhalten der Hinterwurzelfasern bei der Katze. Okajimas Folia Anatomic Japonica, 21: 291-334. Lloyd, D. P. C., and A. K. McIntyre 1950 Dorsal column conduction of group 1 muscle afferent impulses and their relay through Clarke’s column. J. Neurophysiol., 13: 3S54. Mountcastle, V. B., and I. Darian-Smith 1968 Neural mechanisms in somesthesia. In: Medical Physiology. V. B. Mountcastle, ed. C. V. Mosby Co., St. Louis, p. 1378. Padgett, F. 1963 Some observations on new fixation and staining procedures of biological material and on a new supporting film. Scienti. Instru., 8:8-11. Petit, D., and P. R. Burgess 1968 Dorsal column projection of receptors in cat hairy skin supplied by myelinated fibers. J. Neurophysiol., 31: 84-55,
203
Ranson, S. W., and S. L. Clark 1959 The Anatomy of the Nervous System. W. B. Saunders Co., Philadelphia, p. 169. Sherrington, C . S. 1893 Note on the spinal portion of some ascending degenerations. J. Physiol. (London), 14: 255-302. Szentagothai-Schimert, J . 1941 Die Bedeutung des Faserkalibers und Markscheidendicke in Zentralnervensystem. Z. Anat. Entwicklungsch, 1 1 1 : 201-223. Venable, J. H.,and R. Coggeshall 1965 A simplified lead citrate stain for use in electron microscopy. J. Cell Biol., 25: 407408. Walker, A . E., and T. A. Weaver, Jr. 1942 The topical organization and termination of the fibers of the posterior columns in Macaca mulatta. J. Comp. Neur., 76: 145-158. Watson, M. L. 1958 Staining of tissue sections for electron microscopy with heavy metals. J. Biophys. Biochem. Cytol., 4: 475-478. Whitsel, B. L., L. M. Petrucelli, G. Sapiro and H. Ha 1970 Fiber sorting in the fasciculus gracilis of squirrel monkeys. Expl. Neurol., 29: 227-242. Whitsel, B. L., L. M. Petrucelli, H. Ha and D. A. Dreyer 1972 The resorting of spinal afferents as antecedent to the body representation in the postcentral gyrus. Brain Behav. Evol., 5: 303341.
