Anat. Embryol. 152, 73-87 (1977)
Anatomy and Embryology 9 by Springer-Verlag 1977
The Cerebellar Projection from the Raphe Nuclei in the Cat as Studied with the Method of Retrograde Transport of Horseradish Peroxidase Elizabeth Taber Pierce, Grethe H. Hoddevik, and Fred Walberg Department of Anatomy, Harvard Medical School, Boston, Massachusetts, USA and Anatomical Institute, Universityof Oslo, Oslo, Norway
Summary. Following injections of horseradish peroxidase (HRP) in the cerebellar cortex and nuclei of the cat, the distribution of labeled ceils in the raphe nuclei was mapped. The findings confirm those made previously in studies of retrograde cell degeneration following cerebellar ablations (Brodal et al., 1960a), and in addition reveal new details in the projection of the raphe nuclei onto the cerebellar cortex and nuclei. All the raphe nuclei except nucleus linearis intermedius and nucleus linearis rostralis project onto the cerebellar cortex. The nuclei raphe obscurus and pontis contribute the greatest number of afferents to the cerebellum. With the exception of lobule VI which probably is the recipient of a weak projection, all parts of the cerebellar cortex receive afferents from the raphe nuclei. The heaviest projection is to the vermis of lobules VIIA and X, and to crus II. The afferents to the cerebellar nuclei are few in number (Tables 2-6). The observations indicate that each raphe neuron probably projects to more than one terminal site in the cerebellum. The findings are discussed with reference to other efferent and afferent studies of the raphe nuclei. All these studies indicate that the raphe nuclei have widespread efferent and afferent connections, making them capable to participate in a variety of regulatory functions. Key words: Raphe nuclei - Cerebellar cortex and n u c l e i - Experimental study Horseradish peroxidase.
List of abbreviations, fiapm. Ansoparamedian fissure;fiicul. Intraculminate fissure;fiin.cr. Intercrural fissure;ft. Flocculus;fpe. Preculminate fissure;fpft. Parafloccular fissure;fppd. Prepyramidalfissure; fpr. Fissura prima; fpre. Precentral fissure;fprc.a Precentral fissure a; fp.l. Posterolateral fissure; f.p.s. Posterior superior fissure;fsee. Fissura secunda; HII-HX Hemispherallobules II-X; HVIIA cr.L er. H Crus I and II of lobule HVIIA; HVIIIA,B Sublobules A and B of lobule HVIII; Li Nucleus linearis intermedius; Lr Nucleus linearis rostralis; l.ans. Ansiform lobule; N.f. Nucleus fastigii; N.i.a. Nucleus interpositus anterior; N.i.p. Nucleus interpositus posterior; N.l. Nucleus lateralis;pft.d. Dorsal paraflocculus;pfl.v. Ventral paraflocculus;Rd Nucleus raphe dorsalis; Rm Nucleus raphe magnus; Rob Nucleus raphe obscurus; Rpa Nucleus raphe pallidus; Rpo Nucleus raphe pontis; Sc Nucleus raphe centralis superior; s.int.er.1 Intracrural sulcus 1; s.int.er.2 Intracrural sulcus 2; I - V I Vermian lobules I-VI; VIIA,B Anterior and posterior sublobule of lobule VII; VIIIA,B Anterior and posterior
sublobule of lobule VIII.
74
E. Taber Pierce et al.
Introduction
Some years ago it was reported that neurons within specific raphe nuclei project to the deep cerebellar nuclei o f the cat, but not to the cerebellar cortex (Brodal et al., 1960a). The report was based on observations from kitten material with extensive cerebellar lesions and the presence o f retrograde degenerative changes in the cell b o d y (modified G u d d e n method, Brodal, 1940). M o r e recently H r k f e l t and Fuxe (1969), using the histofluorescent technique, demonstrated serotonergic terminals within the cerebellar cortex o f the rat. Since previously it had been shown ( D a h l s t r r m and Fuxe, 1964) that m o s t o f the serotonergic containing cell bodies are located in the r a p h e nuclei, the inference was m a d e that the cerebellar cortex did indeed receive a direct input from the raphe nuclei. Recently, Shinnar e t al. (1975) after injections o f horseradish peroxidase into the vermis o f lobules VI and VII demonstrated labeled neurons in all the raphe nuclei except nucleus linearis intermedius and linearis rostralis. The positive neurons were few in number and labeled faintly. In the present study the projections from the raphe nuclei to the cerebellum have been determined following injections o f small amounts o f horseradish peroxidase into most areas of the cat's cerebellar cortex and nuclei. The study documents that the nuclei raphe obscurus and raphe pontis are the main sources o f inflow from the raphe complex. The d a t a show, also, that lobule VII, in particular lobule VII A, and crus II receive the largest input.
Material and Methods
Altogether 55 cats were used in this study (Table 1). The material was selected from cases prepared to systematically examine afferent cerebellar connections from precerebellar relay nuclei. A number of papers based on these connections, including one which describes problems related to the uptake of peroxidase in the cerebellar cortex, have been published. Since most of the cases used for this study have been described in detail in previous publications, the reader should refer to them for details concerning the injection sites (Brodal, P., 1975; Brodal et aL, 1975; Hoddevik, 1975; Rinvik and Walberg, 1975; Brodal, A., 1976; Hoddevik et al., 1976; Walberg et al., 1976; Brodal and Walberg, 1977). Figures 6 and 7 of the present publication give the outlines of the injections in the cerebellar cortex of the animals. Under Nembutal or Mebumal anesthesia, small amounts of horseradish peroxidase (HRP) suspension (Table 1) were injected with a Hamilton syringe, either stereotaxically or under direct observation. After one to three days the animals were killed under deep Nembutal or Mebumal anesthesia by intracardiac perfusion of a mixture of 0.4% paraformaldehyde and 1.25% gluyaraldehyde with phosphate buffer. After removal of the brain from the skull, the cerebellum was separated from the brain stem. Both were divided into blocks and further fixed. Serial sections of 50-60 gt were cut on the freezing microtome (the brain stem transversaUy, the cerebellum sagitally or horizontally). Two sections Table 1
Cat B.St.L.
Weight kg
Peroxidase concentration, in % (w/v)
625 630 634
0.50 2.40 3.00
50 Sigma 50 Sigma 50 Sigma
amount,/~1
Survival time in days
0.25 0.50 0.50
2 2 3
75
CerebeUar Projection from Raphe Nuclei Demonstrated with HRP Table 1 (cont.) C at B.St.L,
Weight kg.
Peroxidase concentration
amount,/~1
Survival time in days
in % (w/v) 635 637 640 641 642 646 647 648 649 650 65 l 652 653 654 655 656 657 658 665 666 667 668 679 681 682 684 685 686 687 689 690 692 695 698 700 701 702 704 706 707 708 709 710 714 722 723 748 754 755 757 762 763 772
3.20 2.70 0.35 2.40 2.70 2.60 1.80 1.80 2.70 0.75 0.78 2.70 2.30 1.00 1.00 1.60 1.20 1.90 3.20 3.40 3.40 2.00 4.00 2.70 0.75 0.40 0.38 L20 1.20 1.80 1.80 1.80 2.00 2.00 2.10 1.90 4.20 3.00 3.20 3.20 3.20 3.20 2.00 2.50 0.09 2.20 3.20 2.40 2.60 2.50 2.50 2.50 2.50
50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 100 Sigma 100 Sigma 100 Sigma 75 Sigma 75 Sigma 75 Sigma 75 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Serva 50 Serva 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Serva
0.50 0.50 0.15 0.50 O.5O 0.25 0.25 0.25 0.30 0,05 0,05 0.25 0,40 0.08 0.07 0.40 0.08 0.20 0.50 0.50 0.40 0.40 0.50 0.40 0.30 0.20 0.20 0.30 0.30 0.40 0.30 0.40 0.25 0.20 0,25 0.25 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.30 0.20 0.30 0.20 0.20 0.20 0.30 0.30 0.10
2 2 1 1 2 2 2 2 3 1 1 2 3 1 1
2
1 1 2 2 1 1 2 2 2 1 1 3 3 2 2 1
3 2 2 1 3 2 2 2 2 2 3 2 1 3 1
2 2 2 2 2 1
76
E. Taber Pierce et al.
from every group of five were treated with 3,3'-diaminobenzidine tetrahydrochloride as described by Graham and Karnovsky (1966) and mounted on slides. One section Was weakly stained with cresyl violet, the other was left unstained. In approximately serial sections through the cerebellum the area of HRP staining was identified and transferred to a diagram of the cerebellar surface imagined unfolded as depicted by Larsell ( 1970).
Observations
A. The Cytoarchitecture of the Raphe Nuclei A detailed account of the normal topography and cytoarchitecture of the raphe nuclei in the cat has been given previously (Taber et al., 1960). For details the reader is referred to this extensive publication. Figure 5 of the present publication shows the various raphe nuclei as these appear in a parasagittal section through the brain stem of the cat.
B. Projections of Raphe Nuclei to Cerebellar Cortex as Determined Using HRP as a Tracer The distribution of labeled neurons (Figs. 1-4) within the raphe nuclei varied, depending on the site of injection, but no obvious topographical patterns could be discerned. Positive neurons in the raphe nuclei are labeled less intensely than neurons labeled in the same sections in nuclei known to project to the cerebellar cortex (especially the pontine nuclei, nucleus reticularis tegmenti pontis of Bechterew, the inferior olive and the later reticular nucleus). The number of labeled neurons is scarce, except in the projection to lobules VII (particularly lobule VIIA) and X and the posterior part of crus II. The cases to be described are selected into the following groups: 1) Those with injections of the vermis of the anterior lobe, 2) those with injections of the intermediate-lateral part of the anterior lobe, 3) those with injections of the vermis of the posterior lobe, 4) those with injections of the intermediate-lateral part of the posterior lobe, and 5) cases with injections of the cerebellar nuclei. To facilitate the reading, the number of labeled cells found in the stained series in the various cases will be presented in tables. In most of the nuclei positive cells are of all categories, small, medium-sized and large. They are ususally found bilaterally, a unilateral distribution is mostly present when only a few positive cells occur in a nucleus. 1. Injections of Vermis of Anterior Lobe (Fig. 6) The table presented on p. 78 gives the finding of labeled cells in the raphe nuclei following injections in this part of the cerebellar cortex (Table 2).
Comments
Vermis of Anterior Lobe. Cases B.St.L 665 and 700 show that lobules I and II receive a projection from nuclei raphe obscurus, pallidus, magnus, pontis and
Cerebellar Projectionfrom Raphe Nuclei Demonstratedwith HRP
77
Fig. 1. Photomicrographof HRP positive cell in nucleus raphe obscurus in cat B.St.L 700. Dark field. x280
Fig. 2. Same cell as shownin Fig. 1. Interferencemicroscopy. • Fig. 3. HRP positive cell in nucleus parvicellularis of cat B.St.L 714. Interferencemicroscopy.• Fig. 4. HRP positive cellin nucleus raphe pontis in cat B.St.L 704. Interferencemicroscopy.•
dorsalis. Cats B.St.L. 701 and 704 indicate that lobule III is the target of fibers from nuclei raphe obscurus, pallidus, magnus and pontis. Lpbule IV (cases B.St.L. 709, 723 and 706) apparently has a very weak connection with the raphe nuclei, it receives afferents from nuclei raphe obscurus and pallidus. Cat B.St.L. 648 shows that lobule V receives afferents from nuclei raphe obscurus, pallidus and magnus. Likewise, cat B.St.L. 654 gives evidence that lobule V receives fibers from these nuclei, but in addition has a limited input from nucleus raphe pontis. The three last cases in this group (cats 689, 714 and 685) corroborate this conclusion.
78
"-=.
E. Taber Pierce et al.
2"
" " 2- ".- - "'.'~
" -.
"-~,
,,
.. -....:..-'a..,,,r:-
",--- >
.'.."
'
" .c ":..
,.
-V&r.. -
". -..-~!-
~./2~'. :",.'/:.-:.:~.~....., ~ "
Anatomy and Embryology 9 by Springer-Verlag 1977
The Cerebellar Projection from the Raphe Nuclei in the Cat as Studied with the Method of Retrograde Transport of Horseradish Peroxidase Elizabeth Taber Pierce, Grethe H. Hoddevik, and Fred Walberg Department of Anatomy, Harvard Medical School, Boston, Massachusetts, USA and Anatomical Institute, Universityof Oslo, Oslo, Norway
Summary. Following injections of horseradish peroxidase (HRP) in the cerebellar cortex and nuclei of the cat, the distribution of labeled ceils in the raphe nuclei was mapped. The findings confirm those made previously in studies of retrograde cell degeneration following cerebellar ablations (Brodal et al., 1960a), and in addition reveal new details in the projection of the raphe nuclei onto the cerebellar cortex and nuclei. All the raphe nuclei except nucleus linearis intermedius and nucleus linearis rostralis project onto the cerebellar cortex. The nuclei raphe obscurus and pontis contribute the greatest number of afferents to the cerebellum. With the exception of lobule VI which probably is the recipient of a weak projection, all parts of the cerebellar cortex receive afferents from the raphe nuclei. The heaviest projection is to the vermis of lobules VIIA and X, and to crus II. The afferents to the cerebellar nuclei are few in number (Tables 2-6). The observations indicate that each raphe neuron probably projects to more than one terminal site in the cerebellum. The findings are discussed with reference to other efferent and afferent studies of the raphe nuclei. All these studies indicate that the raphe nuclei have widespread efferent and afferent connections, making them capable to participate in a variety of regulatory functions. Key words: Raphe nuclei - Cerebellar cortex and n u c l e i - Experimental study Horseradish peroxidase.
List of abbreviations, fiapm. Ansoparamedian fissure;fiicul. Intraculminate fissure;fiin.cr. Intercrural fissure;ft. Flocculus;fpe. Preculminate fissure;fpft. Parafloccular fissure;fppd. Prepyramidalfissure; fpr. Fissura prima; fpre. Precentral fissure;fprc.a Precentral fissure a; fp.l. Posterolateral fissure; f.p.s. Posterior superior fissure;fsee. Fissura secunda; HII-HX Hemispherallobules II-X; HVIIA cr.L er. H Crus I and II of lobule HVIIA; HVIIIA,B Sublobules A and B of lobule HVIII; Li Nucleus linearis intermedius; Lr Nucleus linearis rostralis; l.ans. Ansiform lobule; N.f. Nucleus fastigii; N.i.a. Nucleus interpositus anterior; N.i.p. Nucleus interpositus posterior; N.l. Nucleus lateralis;pft.d. Dorsal paraflocculus;pfl.v. Ventral paraflocculus;Rd Nucleus raphe dorsalis; Rm Nucleus raphe magnus; Rob Nucleus raphe obscurus; Rpa Nucleus raphe pallidus; Rpo Nucleus raphe pontis; Sc Nucleus raphe centralis superior; s.int.er.1 Intracrural sulcus 1; s.int.er.2 Intracrural sulcus 2; I - V I Vermian lobules I-VI; VIIA,B Anterior and posterior sublobule of lobule VII; VIIIA,B Anterior and posterior
sublobule of lobule VIII.
74
E. Taber Pierce et al.
Introduction
Some years ago it was reported that neurons within specific raphe nuclei project to the deep cerebellar nuclei o f the cat, but not to the cerebellar cortex (Brodal et al., 1960a). The report was based on observations from kitten material with extensive cerebellar lesions and the presence o f retrograde degenerative changes in the cell b o d y (modified G u d d e n method, Brodal, 1940). M o r e recently H r k f e l t and Fuxe (1969), using the histofluorescent technique, demonstrated serotonergic terminals within the cerebellar cortex o f the rat. Since previously it had been shown ( D a h l s t r r m and Fuxe, 1964) that m o s t o f the serotonergic containing cell bodies are located in the r a p h e nuclei, the inference was m a d e that the cerebellar cortex did indeed receive a direct input from the raphe nuclei. Recently, Shinnar e t al. (1975) after injections o f horseradish peroxidase into the vermis o f lobules VI and VII demonstrated labeled neurons in all the raphe nuclei except nucleus linearis intermedius and linearis rostralis. The positive neurons were few in number and labeled faintly. In the present study the projections from the raphe nuclei to the cerebellum have been determined following injections o f small amounts o f horseradish peroxidase into most areas of the cat's cerebellar cortex and nuclei. The study documents that the nuclei raphe obscurus and raphe pontis are the main sources o f inflow from the raphe complex. The d a t a show, also, that lobule VII, in particular lobule VII A, and crus II receive the largest input.
Material and Methods
Altogether 55 cats were used in this study (Table 1). The material was selected from cases prepared to systematically examine afferent cerebellar connections from precerebellar relay nuclei. A number of papers based on these connections, including one which describes problems related to the uptake of peroxidase in the cerebellar cortex, have been published. Since most of the cases used for this study have been described in detail in previous publications, the reader should refer to them for details concerning the injection sites (Brodal, P., 1975; Brodal et aL, 1975; Hoddevik, 1975; Rinvik and Walberg, 1975; Brodal, A., 1976; Hoddevik et al., 1976; Walberg et al., 1976; Brodal and Walberg, 1977). Figures 6 and 7 of the present publication give the outlines of the injections in the cerebellar cortex of the animals. Under Nembutal or Mebumal anesthesia, small amounts of horseradish peroxidase (HRP) suspension (Table 1) were injected with a Hamilton syringe, either stereotaxically or under direct observation. After one to three days the animals were killed under deep Nembutal or Mebumal anesthesia by intracardiac perfusion of a mixture of 0.4% paraformaldehyde and 1.25% gluyaraldehyde with phosphate buffer. After removal of the brain from the skull, the cerebellum was separated from the brain stem. Both were divided into blocks and further fixed. Serial sections of 50-60 gt were cut on the freezing microtome (the brain stem transversaUy, the cerebellum sagitally or horizontally). Two sections Table 1
Cat B.St.L.
Weight kg
Peroxidase concentration, in % (w/v)
625 630 634
0.50 2.40 3.00
50 Sigma 50 Sigma 50 Sigma
amount,/~1
Survival time in days
0.25 0.50 0.50
2 2 3
75
CerebeUar Projection from Raphe Nuclei Demonstrated with HRP Table 1 (cont.) C at B.St.L,
Weight kg.
Peroxidase concentration
amount,/~1
Survival time in days
in % (w/v) 635 637 640 641 642 646 647 648 649 650 65 l 652 653 654 655 656 657 658 665 666 667 668 679 681 682 684 685 686 687 689 690 692 695 698 700 701 702 704 706 707 708 709 710 714 722 723 748 754 755 757 762 763 772
3.20 2.70 0.35 2.40 2.70 2.60 1.80 1.80 2.70 0.75 0.78 2.70 2.30 1.00 1.00 1.60 1.20 1.90 3.20 3.40 3.40 2.00 4.00 2.70 0.75 0.40 0.38 L20 1.20 1.80 1.80 1.80 2.00 2.00 2.10 1.90 4.20 3.00 3.20 3.20 3.20 3.20 2.00 2.50 0.09 2.20 3.20 2.40 2.60 2.50 2.50 2.50 2.50
50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 100 Sigma 100 Sigma 100 Sigma 75 Sigma 75 Sigma 75 Sigma 75 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Serva 50 Serva 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Sigma 50 Serva
0.50 0.50 0.15 0.50 O.5O 0.25 0.25 0.25 0.30 0,05 0,05 0.25 0,40 0.08 0.07 0.40 0.08 0.20 0.50 0.50 0.40 0.40 0.50 0.40 0.30 0.20 0.20 0.30 0.30 0.40 0.30 0.40 0.25 0.20 0,25 0.25 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.30 0.20 0.30 0.20 0.20 0.20 0.30 0.30 0.10
2 2 1 1 2 2 2 2 3 1 1 2 3 1 1
2
1 1 2 2 1 1 2 2 2 1 1 3 3 2 2 1
3 2 2 1 3 2 2 2 2 2 3 2 1 3 1
2 2 2 2 2 1
76
E. Taber Pierce et al.
from every group of five were treated with 3,3'-diaminobenzidine tetrahydrochloride as described by Graham and Karnovsky (1966) and mounted on slides. One section Was weakly stained with cresyl violet, the other was left unstained. In approximately serial sections through the cerebellum the area of HRP staining was identified and transferred to a diagram of the cerebellar surface imagined unfolded as depicted by Larsell ( 1970).
Observations
A. The Cytoarchitecture of the Raphe Nuclei A detailed account of the normal topography and cytoarchitecture of the raphe nuclei in the cat has been given previously (Taber et al., 1960). For details the reader is referred to this extensive publication. Figure 5 of the present publication shows the various raphe nuclei as these appear in a parasagittal section through the brain stem of the cat.
B. Projections of Raphe Nuclei to Cerebellar Cortex as Determined Using HRP as a Tracer The distribution of labeled neurons (Figs. 1-4) within the raphe nuclei varied, depending on the site of injection, but no obvious topographical patterns could be discerned. Positive neurons in the raphe nuclei are labeled less intensely than neurons labeled in the same sections in nuclei known to project to the cerebellar cortex (especially the pontine nuclei, nucleus reticularis tegmenti pontis of Bechterew, the inferior olive and the later reticular nucleus). The number of labeled neurons is scarce, except in the projection to lobules VII (particularly lobule VIIA) and X and the posterior part of crus II. The cases to be described are selected into the following groups: 1) Those with injections of the vermis of the anterior lobe, 2) those with injections of the intermediate-lateral part of the anterior lobe, 3) those with injections of the vermis of the posterior lobe, 4) those with injections of the intermediate-lateral part of the posterior lobe, and 5) cases with injections of the cerebellar nuclei. To facilitate the reading, the number of labeled cells found in the stained series in the various cases will be presented in tables. In most of the nuclei positive cells are of all categories, small, medium-sized and large. They are ususally found bilaterally, a unilateral distribution is mostly present when only a few positive cells occur in a nucleus. 1. Injections of Vermis of Anterior Lobe (Fig. 6) The table presented on p. 78 gives the finding of labeled cells in the raphe nuclei following injections in this part of the cerebellar cortex (Table 2).
Comments
Vermis of Anterior Lobe. Cases B.St.L 665 and 700 show that lobules I and II receive a projection from nuclei raphe obscurus, pallidus, magnus, pontis and
Cerebellar Projectionfrom Raphe Nuclei Demonstratedwith HRP
77
Fig. 1. Photomicrographof HRP positive cell in nucleus raphe obscurus in cat B.St.L 700. Dark field. x280
Fig. 2. Same cell as shownin Fig. 1. Interferencemicroscopy. • Fig. 3. HRP positive cell in nucleus parvicellularis of cat B.St.L 714. Interferencemicroscopy.• Fig. 4. HRP positive cellin nucleus raphe pontis in cat B.St.L 704. Interferencemicroscopy.•
dorsalis. Cats B.St.L. 701 and 704 indicate that lobule III is the target of fibers from nuclei raphe obscurus, pallidus, magnus and pontis. Lpbule IV (cases B.St.L. 709, 723 and 706) apparently has a very weak connection with the raphe nuclei, it receives afferents from nuclei raphe obscurus and pallidus. Cat B.St.L. 648 shows that lobule V receives afferents from nuclei raphe obscurus, pallidus and magnus. Likewise, cat B.St.L. 654 gives evidence that lobule V receives fibers from these nuclei, but in addition has a limited input from nucleus raphe pontis. The three last cases in this group (cats 689, 714 and 685) corroborate this conclusion.
78
"-=.
E. Taber Pierce et al.
2"
" " 2- ".- - "'.'~
" -.
"-~,
,,
.. -....:..-'a..,,,r:-
",--- >
.'.."
'
" .c ":..
,.
-V&r.. -
". -..-~!-
~./2~'. :",.'/:.-:.:~.~....., ~ "
