^0^ 
 
 -^ •* A> , 
 
 
 ; ^ ^'y' *^^- '^ ^^^ '-^ 
 
 
 ..^^ 
 
 »** 
 
.^ .^^ "o . 
 
 '^^ "-: 
 
 
 
 
 \^* ^^ %, ' 
 
 V *•' 
 
 
 0^ o-JL"". '^o. ^^ -• 
 
 
 
 
 
 
 
 **'\ 
 
 •^^ . 
 
Digitized by the Internet Archive 
 in 2010 with funding from 
 The Library of Congress 
 
 http://www.archive.org/details/strumelladiseaseOOheal 
 

 . BULLETIN No. ,0 J™^> ""• 
 
 / 
 
 Commonwealth of Pennsylvania 
 
 DEPARTMENT OF FORESTRY 
 
 (Issued by Direction of the Commissioner of Forestry) 
 
 The Stmmella Disease of Oak 
 and Chestnut Trees 
 
 BY 
 F. D. HEALD 
 
 AND 
 
 R. A. STUDHALTER 
 
 Agents, Investigations in Forest Pathology 
 Bureau of Plant Industry 
 
 HARRISBURG. PA.: 
 
 WM STANLEY RAY. STATE PRINTER 
 
 1914 
 

 
 D. of D. 
 JAN 16 :I15 
 
 (2) 
 
THE STRUMELLA DISEASE OF OAK AND CHESTNUT 
 
 TREES.^ 
 
 BY 
 
 F. D. HEALD and R. A. STUDHALTER, Agents, 
 Investigations in Forest Pathology, Bureau of Plant Industry. 
 
 INTRODUCTION. 
 
 The study of chestnut tree blight in Pennsylvania has brought 
 to light a serious disease of the chestnut due to an entirely dif- 
 ferent organism. The trouble is not confined to the chestnut but 
 attacks various species of oaks. The disease is obscure in its 
 early stages and in some of its more advanced phases, which 
 probably accounts for the fact that it has not been generally 
 observed by foresters. Our recent studies reported by the senior 
 author^ lead to the belief that it is a rather widespread disease 
 and responsible for the death of many trees, both young and old. 
 
 What was probably the same trouble was briefly described in 
 1899 by Dr. W. A. Buckhout^vas ca,using the death of various 
 species of oak trees, but ijo meijtion.was made of the occurrence 
 of the disease on chestnut trees. While he regarded the disease 
 as of fungous origin, no definite determination of the causal 
 organism was made. The correctness of our inference that he 
 was dealing with the same trouble is based largely on the similar 
 symptomology as may be noted by comparison of our descrip- 
 tion with the following quotation: "The black oak is particu- 
 larly subject to a disease apparently of the same nature if not 
 identical with Nectria ditissima, the canker of the apple tree. 
 The early stages of the disease are obscure and inconspicuous, 
 and one's attention is attracted only after serious injury has been 
 done. Apparently the starting point is at the base of a small 
 branch, causing the death and slow decay of surrounding bark 
 
 1. Published by permission of the Secretary of Agriculture. 
 
 2. Heald, F. D. A little-known disease of chestnut and oak trees. Phytopathology 
 
 4: 49. 1914. 
 
 3. Buckhout, W. A. The undesirability of red and black oak because of fungus disease. 
 
 Report Penna. Station, 1899: 250-252. 
 
 (3) 
 
and wood. Both continuation of decay and the effort of the 
 trunk to repair the injury go on together, but it is generally a 
 losing battle. The dead region spreads in spite of the efforts 
 made to produce a callus, covering an enlarging elliptical area 
 on which the little branch shoot appears in the middle as a sort 
 of centre about which circle the lines representing the different 
 stages in the process. As time proceeds, the rest of the trunk 
 receives the normal additions of wood and bark, while the dead 
 parLs become more and more an element of weakness." 
 
 Unfortunately no specimens of oak showing this disease are 
 preserved in the State College collections, but the occurrence of 
 the Strumella disease in the vicinity of State College at the 
 present time lends added support to the descriptive evidence that 
 Buckhout was certainly dealing with the trouble described in 
 the following pages. It seems strange however, that a disease 
 recognized fifteen years ago should continue its ravages and es- 
 cape further study until the present time. 
 
 Although our knowledge concerning this disease is still very 
 imperfect, sufficient data have been accumulated to justify us in 
 presenting a short account. The main object at the present time 
 is to call attention to the importance of the disease and describe 
 its characters in sufficient detail to enable foresters and path- 
 ologists to recognize it. 
 
 SYMPTOMS AND EFFECTS. 
 
 There is a rather remarkable parallelism between this new dis- 
 ease and the well-known chestnut tree blight as far as general 
 symptoms and effects are concerned. 
 
 Young infections of the Strumella disease on smooth-barked 
 trunks can be easily recognized by the presence of yellowish 
 or yellowish-brown patches, slightly raised, and standing out in 
 marked contrast to the normal bark which is darker in color. 
 The area invaded by the fungus may be fairly regular in out- 
 line, but it is very frequently irregular in outline, the latter 
 showing what may be designated as the amoeboid type of in- 
 fection (fig. 1). The most noticeable external difference between 
 young lesions of the chestnut blight (Endothia), and those of 
 
the Strumella disease, is the presence in the latter of small black 
 nodules, scattered over the surface of the yellowish area which 
 marks the extent of the lesion. In young infections of the type 
 described the dark nodules are either aborted or rudimentary 
 fruiting- pustules No fertile pustules have ever been found in 
 these young infections, but spore production occurs in the later 
 stages of the disease. The yellowish-brown color is caused by 
 the whitish, or very pale buff-colored vegetative body, or my- 
 celium, of the causal fungus, which is covered only by the thin 
 external layer of cork cells. If some of this cork layer is re- 
 moved from the advancing edge of the lesion, the mycelium 
 exposed will show its pure white color, but in more central and 
 older parts the mycelium has changed to a pale buff color. In 
 general it is lighter in color than the mycelium of the chestnut 
 blight fungus, and definite "fans" characteristic of the latter are 
 not noticeable. Lesions of this type have been found on shoots 
 up to 4 or 5 inches in diameter, and varying from minute inci- 
 pient infections to those completely encircling the trunk. The 
 young infections are much more obscure on older or rough- 
 barked trunks, but there is generally sufficient diff'erence in color 
 to indicate the extent of infection. 
 
 Two forms of the disease may be recognized with further 
 development of the trouble : first, the canker type, in which the 
 progress of the fungus is slow with a more or less pronounced 
 formation of callus at the advancing edge of the lesion; second,, 
 the diffuse type, in which the fungus spreads more rapidly, kill- 
 ing the tissues so quickly that the formation of callus is not 
 possible. Apparently it was only the conspicuous canker type 
 that was observed and described by Buckhout. 
 
 In the well-defined canker there may be more or less zonal 
 growth (fig. 7), as the result of the repeated formation of cal- 
 lous tissue at the advancing edge, but it is the exception for the 
 callus to check the advance of the fungus. The lesion is gen- 
 erally extended more rapidly lengthwise of the trunk than trans- 
 versely, thus giving rise to cankers of somewhat elliptical out- 
 line (figs. 2 and 3). Cankers five feet or more in length have 
 been observed on trunks of chestnut trees 6-8 inches in diameter 
 
that had not yet been girdled. In old cankers there is likely to 
 be more or less destruction of tissue at the center, so as to 
 leave a somewhat open wound exposing the decaying wood 
 (figs. 3 and 5). Old cankers are likely also to show the' work 
 of woodpeckers and insect larvae. 
 
 When a weU-defined canker is not produced the extent of the 
 lesion may be indicated by an elliptical discolored area which is 
 generally somewhat depressed. Even when the bark of this 
 area is not destroyed it is rougher than surrounding portions 
 due to more or less shrinking and cracking. In all cankers the 
 characteristic black nodules may be noted on those portions on 
 which the bark has remained intact. 
 
 In some infections there is a pronounced enlargement or 
 hypertrophy of the entire invaded area (fig. 6), instead of 
 atrophy or lessening of growth. In the open cankers there may 
 be more or less enlargement or distortion of the trunk (figs. 3, 
 5 and 7), due to the stimulating effect of the parasite. Such 
 effects appear to be more frequent on the oaks than on the 
 chestnut. 
 
 The disease is least evident in its diffuse form. In this type 
 there is an absence of well-defined cankers and the fungus ap- 
 pears to spread so rapidly that little or no callus is developed. 
 In such cases the young tree is killed before time has been given 
 for the development of a canker. This points to the fact that 
 the larger trees showing well-defined cankers represent the few 
 individuals that have offered the most resistance to the attacks 
 of the parasite. The diffuse type of the disease has not yet been 
 found on anythmg but young trees up to 3-4 inches in diameter. 
 It should not be understood that there are two entirely distinct 
 types of the disease, for there are all gradations between the 
 pronounced canker types and the diffuse types ; in the inter- 
 mediate types the lesion may be marked by a more or less con- 
 centric zonation (figs. 2, 4 and 7), while in the extreme diffuse 
 type no zonatioii is evident (fig. 8). 
 
 The disease appears to start in a branch axil. In all young 
 infections or old cankers this relation is an invariable rule, and 
 in lesions of any age the dead stub of a branch marks the center 
 
of the inflection (fig. 7). After the branch has been killed the 
 fungus grows more rapidly through its tissues, and soon pro- 
 duces the characteristic fruiting pustules over its surface in the 
 form of dark brown, powdery, and erumpent nodules 1-3 mm. 
 (i-^ inch) in diameter (fig. 8). 
 
 The production of sprouts or "suckers" just below a lesion 
 is one of the marked characteristics in either the canker or semi- 
 diffuse forms of the disease (figs. 11 and 12). This production 
 of vigorous sprouts occurs as soon as the trunk has been girdled, 
 and they may be few in number or sufficiently numerous to make 
 a conspicuous clump. 
 
 The final result of the disease is the death of the tree, but in 
 the extreme canker type there is a rather prolonged struggle. 
 No cases have been observed where the advance of the parasite 
 has been permanently checked. It is apparently the diffuse type 
 of the disease that is exacting the larger toll, if we may judge 
 from observations in certain localities where the disease is es- 
 pecially prevalent. The disease is especially serious since there 
 is a marked tendency to the production of body or trunk cankers 
 only. They may occur at any point from the ground up but 
 they appear to be more common in the lower portions of the 
 trunk. As soon as girdling has been completed the parts beyond 
 the canker must succumb, but even before such a fatal termina- 
 tion, the trunk may be so weakened by the inroads of the fungus 
 as to be unable to withstand the force of strong winds. Many 
 fallen oaks which have been broken at a lesion are silent wit- 
 nesses of the destructiveness of this trouble (fig. 12). Unlike 
 the chestnut bark fungus, the organism in question grows equally 
 well in both bark and wood. It is this disintegration of the 
 wood by this parasite and not by secondary decays that causes 
 the pronounced weakening of the trunk in the region of a canker. 
 The extent to which the fungus invades the wood even in the 
 canker type is shown in Fig. 14, the light area being the por- 
 tion infected. In Fig. 13 the only normal wood remaining shows 
 a light color also. 
 
THE CAUSAL ORGANISM. 
 
 A single species of fungus is invariably associated with this 
 disease. It has repeatedly been isolated from all of the different 
 types of the disease on a variety of hosts/ The fungus grows 
 well in ordinary culture media, but 3% dextrose agar has been 
 used in making most of the isolations. Tissue transfers made 
 by planting small fragments of diseased bark or affected wood 
 in agar plates give pure cultures in the majority of cases. 
 Typical isolation plates are shown in Figs. 15 and 16. The 
 fungus produces a cottony aerial mycelium and this turns to a 
 dirty gray after one to three weeks. The under side of the 
 colony shows a dark brown or black coloration which first be- 
 comes noticeable at the center and gradually spreads towards 
 the periphery of the colony. Different strains show great varia- 
 tion in the amount and intensity of darkening (fig. 17). No 
 spore formation has been observed in any cultures. 
 
 The fungus in question, Strumella coryneoidea Sacc. & Wint., 
 is not a new one, but is an old species which has not pre- 
 viously been accused of being a parasite. It was first collected 
 at Perryville, Mo., in 1883, by C. H. Demetrio and described by 
 Saccardo and Winter.^ The fungus has also been collected at 
 London, Ontario, on oak bark by J. Dearness (No. 94, June, 
 1889). The two following collections referred to this species 
 were apparently incorrectly determined : 
 
 1.— Flora of Washington. No. 307. Collected Jan. 28, 1894, 
 on dead branches of Rhus diversiloha J. and G., by W. N. 
 Suksdorf. Columbia River, W. Klickitat Co. 
 
 2.— Flora Tudoviciana. No. 1799. Collected Mar. 30, 1889, on 
 young dead twigs of Carya olwacforniis by A. B. Langlois. 
 St. Martinsville P. O., La. 
 
 A careful comparison of our specimens with those of Deme- 
 trio (Ellis and Everhart. North American Fungi, No. 1653), 
 and Dearness (No. 94), has failed to reveal any morphological 
 
 1. Many of the isolations were made by Mr. R. C. Walton, who was associated with 
 the writers during the earlier work on this disease. 
 
 2. Saccardo and Winter. Rabenherstii Fungi europaei extraeuropaei. No. 2984. 
 Hedwigia, 1S83: 175-176. Ellis and Everhart. North Am. Fungi, No. 1653, June 1885 
 
differences, so we feel justified in referring the parasite to the 
 species originally described by Saccardo and Winter. The 
 causal relation is based not only on the constant association of 
 the fungus with the disease but also upon successful inocula- 
 tions in the field. That ^trumella coryn eoidea wa.s not connected 
 with the disease by Dr. Buckhout^ is evident from his statement 
 that the disease is "apparently of the same nature i'f not identical 
 with yectria ditissima, the canker of the apple tree." 
 
 The fruiting stage. — During the time previous to the girdling 
 and consequent death of a branch or trunk, the fungus produces 
 only mycelium and aborted or rudimentary fruiting pustules. 
 As soon as the girdling has been completed, however, an abund- 
 ance of fertile pustules appear in the form of brownish 
 erumpent powdery cushions, the sporodochia, scattered either 
 sparsely or abundantly over the surface of the invaded por- 
 tions and also beyond the extent of the original lesion (figs. 8 
 and 18). 
 
 The sporodochia vary in size from 1-3 mm. in diameter (fig. 
 19), and each consists of a dense aggregate of fungous tissue 
 which bursts through the external cork layer in the form of a 
 rounded nodule which produces over its entire free surface a 
 multitude of erect, branched, spore-bearing hyphse or conidio- 
 phores. Spores are developed in profusion from the terminal 
 portions of the brownish conidiophores (fig. 20), and either 
 loose or attached, give the granular or powdery appearance to 
 the fruiting pustules (figs. 8 and 18). The spores (fig. 21) are 
 light or ver}^ dark brown, single-celled, vary from nearly glob- 
 ular to pear-shaped or irregular in form, and vary in size from 
 .5.1 — 13x'1 — 7.3 mikrons. No ascus stage has been found in 
 the life cycle of this fungus, but the profusion of conidiospores 
 would appear to offer an effective means of dissemination. As 
 the accumulated spore mass is powdery when dry it seems prob- 
 able that wind dissemination is one of the important means 
 of spread of the fungus. The loose spores are undoubtedly 
 washed down by rains. The latter statements are based however 
 on observational evidence rather than on direct experimental 
 evidence. 
 
 3. Loc. cit. p. 250. 
 
10 
 
 HOST RELATIONS. 
 
 This disease must be viewed with some additional concern 
 since it is not confining its attacks to a single species. As a 
 disease of the chestnut it is much less virulent than the End- 
 othia blight but it occurs on the chestnut beyond the present 
 range of the chestnut bark disease. Various species of oaks are 
 affected and in some localities the disease has made such head- 
 way as to cause some alarm. In the short experience with this 
 disease up to date it has been found attacking the following 
 species : 
 
 American chestnut, Gastanea dentata (Marsh.) Borkh. 
 
 Chestnut oak, Quercus Prinus L. 
 
 Black oak, " velutina Lam. 
 
 Red oak, '' ruha L. 
 
 Scarlet oak, " coccinea Muench. 
 
 White oak, '' alha L. 
 
 The disease appears to be more severe on the red and black 
 oaks, and also produces more pronounced cankers on these 
 species. The most typical development of the disease upon oaks 
 has been observed in fairly dense stands in which a single 
 species predominates. The disease is more obscure on the white 
 oak than on any other species of oaks. This is due to the very 
 rough character of the normal bark, which may render incon- 
 spicuous lesions of some extent. The Strumella cankers on the 
 chestnut (fig. 9) are rarely as conspicuous as those caused by 
 Endothia. In some Strumella infections in this species the lesion 
 may reach a considerable size without showing much discolora- 
 tion or depression of the bark. There is but rarely any indica- 
 tion of concentric zonation in lesions on this species, while it 
 ma}^ be recalled that this character is especially marked in the 
 oak species. Ii. old cankers in all species there is generally 
 more or less corrosion at the center, so as to leave a rough, open 
 wound. It seems probable that all species of oaks will be found 
 susceptible to this disease. 
 
11 
 
 GEOGRAPHICAL DISTRIBUTIOJi. 
 
 The Strumella di=fc</t^ h^r .^tn r .; 1:=:1 from spedmens col- 
 lected at widely separated points in Pennsylvania. The first 
 specimens studied were collected in Somerset cotmty in the 
 southwestern part of the State. In this region the disease was 
 prevalent on both chestnut and chestnut oak. A little later the 
 trouble was found on chestnut oak in ^^liiflin cotmty in the cai- 
 tral part of the State. It is common on the red and black oaks, 
 and occasional on the white oaks in the State forest in Hunt- 
 ingdon county. It has been observed in Elk county in the north- 
 w-estem part of the State, where it is especially severe on the 
 chestnut in a number of localities. Studies made in Carbon 
 county in the eastern part of the State show that the trouble is 
 fairly common. It has also been reported from Pike county. 
 The known occurrence of the disease at localities in the southern, 
 middle, northern and eastern portions of Pennsylvania suggest 
 that a more detailed scouting of our forests will show a rather 
 general distribution throughout the State. 
 
 The causal fungus, Strumella corpneoideayhas been collected 
 in the fruiting stage on both chestnut and oak species in several 
 of the localities studied- The spedmens received from some of 
 the localities represent only the canker tjpe, probably due to 
 the fact that only conspicuous examples of the trouble were 
 observed by the collectors. In the locahties worked in detail 
 by the writers, the fruiting stage has always lesr. fiur.d. In 
 t' : - : r.nection reference may be made to the f: r - : . : . . t . i r - : 
 z:.z f^ingus in Missouri, and its later collectiir. frin Ir^iirii. 
 Canada, as indicating a wide range. All of the evidence at 
 present points to the parasite as a native of the northeastern 
 portion of Xorth America. If this supposition is correct, the 
 ci^riEt :; "ess likely to prove as virulent as though it were in- 
 :r.l::tl :r,' some foreign countr}-. The behavior of the 
 chesmui bi^rxc cisease is prophetic of what might happen if the 
 Strumella disease should be introduced into new localities. 
 
 The supposition concerning the wide distribution of the dis- 
 ease has been further substantiated bv the collection of authentic 
 
121 
 
 specimens on both the black and white oak in Massachusetts 
 by Mr. R. G. Pierce. It is hoped that this brief description will 
 aid and stimulate further observations and study on this inter- 
 esting tree disease, so that we may have more complete informa- 
 tion, as to its range and prevalence. 
 
 EXPLANATION OF PLATES. 
 
 All of the photographs used for illustrating this bulletin were 
 made under the writers' supervision from specimens collected 
 in Pennsylvania. 
 
 PLATE L 
 
 Fig. 1. — A young Strumella infection of the "amoeboid" type 
 on chestnut oak 3 inches in diameter. The minute black 
 nodules scattered over the surface distinguish this super- 
 ficially from a young chestnut blight lesion. 
 
 Fig. 2. — A semi-diffuse type of infection showing three distinct 
 zones of growth, and the characteristic dark nodules scat- 
 tered over the invaded area. The cut stub of a branch shows 
 at the center, and its lighter peripheral portion indicates 
 the depth to which the fungus had penetrated the wood. 
 
 PLATE IL 
 
 Fig. 3. — A well-defined canker on red oak 2.5 inches in diameter. 
 The advance of the fungus was checked on the right by a 
 marked formation of callus. In the central part the bark 
 was completely destroyed so as to leave an open wound 
 exposing the wood. (See also Fig. 13 for transverse sec- 
 tion through this canker). 
 
 Fig. 4. — The same specimen shown in Fig. 3 viewed from the 
 opposite surface. The zonal advance of the fungus may be 
 noted. Only a narrow bridge of uninfected tissue remains. 
 (See also Fig. 13). 
 
13 
 
 PLATE III. 
 
 Fig. 5. — Canker on a black oak 3 inches in diameter showing 
 considera:ble hypertrophy and also the destruction of tissue 
 to form an open wound. 
 
 Fig. 6. — A young lesion on chestnut oak 1 inch in diameter 
 showing hypertrophy, with characteristic longitudinal crack- 
 ing of the bark. The fungus has completely encircled the 
 trunk. 
 
 PLATE IV. 
 
 Figf. 7. — A canker on a scarlet oak 2.5 inches in diameter show- 
 ing pronounced zonal growth and the stub of a branch at 
 the center. (See Fig. 14 for transverse section through this 
 canker). 
 
 Fig. 8.— Portion of a small black oak tree 1.5 inches in diameter 
 killed by the diffuse type of the disease. The characteristic 
 fruiting pustules, or sporodochia, are very numerous over 
 the entire surface. 
 
 PLATE V. 
 
 Fig. 9. — Typical Strumella canker on a chestnut trunk 6 inches 
 in diameter. Note that on this species the extent of the 
 lesion is indistinct. 
 
 Fig. 10. — A body canker from a young white oak tree. The 
 elliptical outline of the canker was especially noticeable in 
 its earlier stages, and callus-zones mark its progress. 
 
 PLATE VL 
 
 Fig. 11. — Portion of a small black oak tree killed by the Stru- 
 mella disease. Vigorous sprouts or "suckers" were pro- 
 duced belov/ the lesion which was of the semi-diffuse type. 
 The center of the infection is marked by a short stub of a 
 branch. 
 
14 
 
 PLATE VII. 
 
 Fig. 12. — A black oak which has been broken by the wind at 
 the lesion. The copious sprouting below the lesion is evi- 
 dent and the canker also shows the work of woodpeckers. 
 
 PLATE Vm. 
 
 Fig. 13. — Transverse section through the middle of the canker 
 shown in Fig. 3. The light area on the left is the only 
 normal tissue that remains. 
 
 Pig. 14. — Transverse section through the middle of the canker 
 shown in Fig. 7. The lighter area above the organic center 
 has been invaded by the fungus and the wood disintegrated. 
 This affected wood is soft and punky. 
 
 PLATE IX. 
 
 Fig. 15. — An isolation culture oi Strumella coryneoidea on 3% 
 dextrose agar, made by tissue transfers. The small frag- 
 ment of diseased tissue is still noticeable at the center of 
 each colony. 
 
 Fig. 16. — An isolation culture of Strumella coryneoidea on 3% 
 dextrose agar made by transferring marked colonies from 
 poured plates. 
 
 PLATE X. 
 
 Fig. 17. — A series of cultures of Strumella on dextrose agar 
 taken from the under surface to show the variation in the 
 development of the black coloration characteristic of the 
 species. Each culture represents a different strain. 
 
 PLATE XL 
 
 Fig. 18. — A piece of bark slightly magnified to show the char- 
 acteristic fruiting pustules, or sporodochia, in the form of 
 dark brown, powdery, erumpent nodules. 
 
15 
 
 Fig. 19. — Vertical section through a typical sporodochium. The 
 dense aggregate of fungous tissue has broken through the 
 periderm, and produced a superficial zone of branched coni- 
 diophores, in the terminal parts of which spores have been 
 formed. Semi-diagramatic from a camera drawing. X-75. 
 
 PLATE XII. 
 
 Fig. 20. — A single branched conidiophore showing characteristic 
 form and septation. The spores are loosely attached and 
 easily separate from the conidiophores. X-800. 
 
 Fig. 21. — Terminal portion of a conidiophore showing the origin 
 of spores. X-800. 
 
 Fig. 22. — Camera drawing of spores showing the variation in 
 form. X-800. 
 
i^»^ 
 
 W 
 
 (X6) 
 
'[..\'j"j-: I. 
 
 Fig. 1. 
 
 Fig. 2. 
 
PLATE II. 
 
 Fig. 3. 
 
 Fig. 4. 
 
PLATE III. 
 
 Fig. 5. 
 
 Fig. 6. 
 
PLATE IV. 
 
 1 
 
 V 
 
 i 
 
 
 
 x ■ 
 
 
 
 Fig. 7. 
 
 Fig. 8. 
 
PLATE V. 
 
 Fig. 9. 
 
 Fig. 10. 
 
rr.ATE VI. 
 
 FU. 11. 
 
PLATE yil. 
 
 Fig. 12. 
 
PLATE VIII. 
 
 Fig. 13. 
 Fig. 14. 
 
PLATE IX. 
 
 Fig. 15. 
 Fig. 16. 
 
PLATE X . 
 
PLATE XI. 
 
 
 
 x.h-'-'m~-fv 
 
 E-s^US^ i 
 
 &^ 
 
 
 
 "IW^-tfe 
 
 Fig. 18. 
 Fig. 19. 
 
PLATE XII. 
 
 


 
 cv -» 
 
 
 >* "'^ ^y 
 
 > "^^ A^ *^ 
 
 
 c.~ .r» ° V// xkVVw vv •<» &?^ll^^ > ,S ^^ 
 
 
 
 
 'oV 
 
 '>o* 
 
 L^^ o_ 
 
 o. * 
 
 
 ,0* »1V°'. > v^ .^ 
 
 
 
 
 0" 
 
 ■» o 
 
 
 
 ' -^^^^ ^'^^S ^<^^^\ °^^S ^^^\ "'^^S ^^^^\. ^^ 
 

 
 
 
 
 
 . .......uITo U ^"^ *t 
 
 
 
 '^0^ 
 
 
 
 ^^' 
 
 #'\ V 
 
 
 
 0^ 
 
 
 A <^ . . . 
 
 .<^ r O » O - <S> . 
 
 
 /%. '"TO^-" *"^'% •.^'" /% '^^p-- ^^"% ''-n