[ 053 ] XV .
Polacanthus Foxii , a large undescribed Dinosaur from the Formation in the Isle of Wight .
ByJ .
W. Hulke , F.R.S. Received January 3 , \#151 ; Read January 27 , 1881 .
[ Plates 70-7G .
] For the opportunity of studying the remains described in this note I am indebted to the courtesy of the Rev. W. Fox , of Brixton , Isle of Wight , who last autumn gave me free access to his rich collection of fossils obtained in that locality .
Much shattered by being very hastily dug out , and since much damaged by the accidental breakages and the dissociations scarcely avoidable in the absence of a suitable place for their safe-keeping , there is risk of these remains becoming before long lost to the palaeontologist .
In view of this not improbable eventuality I venture to offer to the Royal Society these notes , in writing which I have been reminded that it was to this Society the late Dr. G. A. Mantell , now more than fifty years since , communicated his first discoveries of Iguanodont and Hylaeosaurian remains .
The remains of Polacanthus were found by Mr. Fox in 1865 in a bed of blue shaly clay , which occurs near the middle of the cliff , a short distance east of Barne 's Chine .
The bed is easily recognisable by the large quantities of lignite which it contains .
Professor R. Owen , to whom Mr. Fox showed some of these fossils soon after their discovery , suggested for the animal indicated by them the name Polacanthus\#151 ; many-spined\#151 ; P. Foxii , and this name Mr. Fox adopted in an account of his disco very read by him at the next meeting of the British Association .
A brief notice of the discovery with a rude woodcut also appeared about the same time in the " Illustrated London News .
" Both these communications have only the value of preliminary notices by persons without anatomical training , and no description of the fossils sufficient for the use of palaeontologists has yet appeared .
Mr. Fox 's MS .
, read at the meeting of the British Association , cannot now be found , and his paper does not appear in the " Reports .
" An abstract which I made of it in 1869 gives the following list of the parts he believed he had secured .
" Sacrum and pelvis ; 7 lumbar , 7 anterior dorsal vertebrae with their ribs ; 20 caudal vertebrae ; 2 femora ; 1 tibia with fibula ; 3 metatarsals , phalanges , and 3 unguals ; 20 to 30 large dermal spines , and as many scutes .
" The scattered remains which last autumn I succeeded in bringing together again do not quite agree with these numbers ; the vertebrae , foot-bones , and dermal spines are fewer , and I was not able to identify any portions of Ilium , Ischium , or Pubis .
Vertebral column.\#151 ; This is now represented by 11 praesacral , five sacral , and 15 postsacral or caudal vertebrae .
The 11 praesacral vertebrae comprise six disconnected and five anchylosed in a continuous series .
Of the six disconnected vertebrae three are fairly complete ( Plate 70 , figs. 1 , 2 ) .
The form of the centrum is cylindroid ; it is long relatively to its breadth , slightly constricted at its middle and expanded at its articular ends , which are plane or very slightly concave , the concavity of the posterior surface being most evident .
The antero-posterior extent of the neurapophyses at their attachment to the centrum nearly equals the length of this latter .
Their anterior margin rises nearly vertically from the centrum , whilst their posterior margin has a strong forward slant .
The spinous processes of all the praesacral vertebrae are broken off and missing .
The transverse processes in this series show a double costal articulation .
In the level of the crown of the arch is a large , conspicuous , capitular , costal facet borne jointly by the arch and root of transverse process .
It is directed outwards , and against it in two instances the rib-head , of an expanded discoid form , still abuts .
Above this , the process , slender and trihedral in cross-section , is prolonged outwards and upwards above the rib-neck .
It bore at its free end , as is shown by detached pieces , an articular surface for the tubercle of the rib .
The length of the vertebral centrum ( represented in Plate 70 , figs. 1 , 2 ) is 75 millims. ; the horizontal diameter at the articular ends is 52 millims. , and the vertical diameter here 51 millims. ; and the horizontal diameter at the middle of the centrum is 35 millims. The double costal articulation places these six disconnected vertebrae in the front of the trunk .
Between them and the five anchylosed vertebrae several are doubtless missing , since these last are demonstrated to belong to the loins by the anchylosis of the hindmost of the series to the first sacral centrum .
The lumbar centra ( Plate 71 , fig. 1 ) have a more attenuate form than that of those referred to the front of the chest .
Their lateral surfaces slope inwards and meet somewhat angularly below .
This is very apparent in the second in the chain .
The spinous and transverse processes of all are broken off and missing .
The average length of the centrum in this series is 75 millims. , the horizontal diameter at the middle is 33 millims. , 30 millims. , 30 millims. , 34 millims. ; and that of the articular ends is 40 millims. , 37 millims. , 36 millims. , 50 millims. Sacrum ( Plate 71 , fig. 1).\#151 ; This has the usual dinosaurian structure ; it consists of a chain of five anchylosed centra much larger than the slender lumbar centra .
Accidental cross-sections made by fractures show the form of the centrum to be remarkably depressed ( fig. 2 ) ; it is , however , not improbable that this great excess of the horizontal over the vertical diameter may have been increased by pressure .
Ihese diameters are in the second centrum 70 millims. and 25 millims. respectively .
The MR. J. W. HULKE OR THE POLACANTHUS FOXII .
under surface of the centrum is cylindroid ; in the three foremost centra a shallow median groove indents it longitudinally .
The piers of the neural arches rest each on two centra , and the sacral nerves escaped from the vertebral canal across the middle of a centrum except the last nerve , which passed out intervertebrally between the last sacral and the first caudal vertebra .
A stout lower transverse process stands out from the side of the sacrum along the line of junction of each two centra , and doubtless the outer ends of these processes coalesced in the usual looplike manner : all have been broken off , and are now missing .
The upper transverse processes which project from the crown of the arch are small and inconspicuous ; they form , with the expanded summit of a dwarfed spinous process , the support of a stout dermal armour to be presently described .
Comparison of the following measurements of the sacral centra with those of the prsesacral centra already given will make the greater bulk of the former very apparent .
Sacral centra .
No. 1 .
No. 2 .
No. 3 .
No. 4 .
No. 5 .
Horizontal diameter at the articular ends .
* 87 87 87 87 , , , , middle of centrum . .
70 70 70 70f 82 Length of centrum 67 67 67 62 58 Post-sacral vertebrae ( Plate 72 , figs. 1,2 ; Plate 73 , figs. 1 , 2 ; Plate 75 , figs. 3 , 4).\#151 ; The 13 now remaining are certainly not a continuous series\#151 ; many intermediate ones are missing ; most are from the root of the tail .
The first caudal , known to be such by its exactly fitting the last sacral , differs from this most obviously by the centrum bearing entirely its own arch .
It and those centra , which by their great size are referable to the root of the tail close to the sacrum , bear no chevron mark .
These centra are short and wide ; their articular ends are both gently concave , and they have a depressed heart-shaped contour .
In one of the best preserved the horizontal diameter of the articular end is 100 millims. , that of the anterior end being about 90 millims. , whilst the vertical diameters of the same ends are 64 millims. and 61 millims. , and the length of the centrum measured along its neural surface is 52 millims. At their middle these centra are much contracted .
The transverse process stands off from the side of the centrum horizontally just below the line of the neurocentral suture .
The arch is low ; the spinous process has a strong backward slant ; the post-zygapophyses are large and prominent .
Other caudal vertebra smaller than those just described differ from them mainly in the presence of chevron marks and in the greater concavity of their articular ends .
At least 13 , probably about 17 or 18 , of the vertebra3 in the fore part of the tail had transverse processes ; two smaller vertebrae , which by their very small size belonged to near the end of the tail , have a simple cylindroid figure ; the processes have disappeared , and the arch is reduced to a tubular hoop ( Plate 7 5 , figs. 3 , 4 ) .
# Rather less tlian others .
+ The numbers in this line , except the last , are approximate .
The osseous tissue of all the vertebrae is coarse , and the outer surface wants the closeness of texture and the smoothness so observable in Iguanodont bones .
This textural difference serves to distinguish very mutilated pieces .
It is not peculiar to this particular skeleton , for it is equally apparent in a thoracic vertebra which I dug a quarter of a mile to east of the place where this skeleton was discovered\#151 ; a distance which makes it almost impossible that it should have ever formed part of it .
Ribs.\#151 ; Many of these show a double vertebral articulation by head and tubercle .
In those referable to the front of the chest , as has been already mentioned , the head of the rib is very swollen and the neck is relatively slender .
These parts are wanting in all the disconnected ribs from this part of the chest , but they are fortunately preserved in situ upon three vertebrae ( Plate 70 , figs. 1 and 2 ) .
In this region the neck of the rib is short , and it joins the body of the rib in an almost uniform curve .
In ribs referable to the middle of the trunk a very decided angle marks the junction of the tubercle and body ( Plate 76 , fig. 2 ) .
In all ribs with double vertebral articulation the neck is slender , and it is so compressed that the long diameter of its cross-section is vertical .
Beyond the tubercle the upper border ( here become outer ) expands so widely as to deserve the term " surface/ ' whilst the lower ( here inner ) border continues thin .
This gives a triquetrous form and great strength to this part of the rib .
Limbs.\#151 ; The femur ( Plate 74 ) is remarkable for the largeness of its articular extremities and the slenderness of its shaft .
Both ends are much damaged by pressure .
My description is taken from the right , which is better preserved than the left .
The proximal end bears at its inner angle a sessile sub-hemispherical head { cap. ) , external to which and nearly in the same level with it is a well-developed trochanter The inner or posterior trochanter ( fig. 2 , tr.i .
) , characteristic of dinosauria , is also strongly developed ; it is rather nearer to the proximal than the distal end of the bone : * The knee condyles are very large ; the inner is somewhat the larger of the two ; a deep narrow groove separates them in front , and a wider , relatively shallow , depression divides them behind .
The length of the right femur is 5 5 centims. ; the diameter of the distal end across the condyles is 15*5 centims. , and that of the proximal end is 18 centims. The diameters of the caput femoris are 9 centims. and 8'5 centims. , and that of the middle of the shaft is 8 centims. The Tibia ( Plate 75 , figs. 1 , 2 ) , the left one , is much shorter than the femur , being only 35 centims. long .
It resembles the femur in the largeness of its joint ends and the slenderness of its shaft .
The proximal end , distorted by pressure , shows obscurely a division of the articular surface into two parts answering to the femoral condyles ( cd .
) , and a remarkably large prsecnemial process ( pro .
) .
The distal end , flattened , is of the usual dinosaurian pattern , having a longer and narrower outer and a shorter and stouter inner division .
When the bone is held vertically so that a line drawn between the proximal condyles is directed forwards , the longer axis of the distal end intersects this , making with it an angle of about 60 ' .
This different direction of the ends makes the shaft appear twisted .
Relatively to the large size of the articular ends the shaft MR. J. W. HULKE ON THE POLACANTHUS FOXII .
looks singularly short and slender .
Its cross-section at the middle is roughly trigonal , and the diameters here are 4'5 centims. and 5'5 centims. Owing to the twist of the shaft , that which below is the inner border becomes as it ascends the broad anterior surface above ; and the narrow surface at the upper end lying between the outer condyle and the prsecnemial crest as it descends becomes towards the distal end the broad antero-external surface .
Of the fibula only a small fragment attached to the outer border of the antero-external surface near the lower end now remains .
Foot.\#151 ; The only parts of this which can with certainty be identified are two metatarsals ( Plate 72 , fig. 3 ) .
In the great size of their joints and shortness and slenderness of the shaft they repeat the most striking features of the femur and tibia .
Their distal end has the usual pulley form .
The condyles project strongly towards the sole , and the pits for the attachment of the lateral ligaments are large and deep .
The proximal end is most expanded vertically .
The length of the best preserved metatarsal bone is about 8'5 centims. , the breadth of its distal end is 4*5 centims. , the longer diameter of its proximal end about 7'5 centims. , and the diameter of the middle of the shaft nearly 2 centims. Respecting the bones considered unguals by Mr. Fox I cannot speak confidently .
Their form is broad , depressed , and blunt .
Dermal armour.\#151 ; Together with the endoskeletal bones a highly developed dermal armour was found .
Mr. Fox told me that when he first laid this bare it formed so continuous a mail that his first impression was that he had exposed the carapace of a huge turtle .
It measured 3 feet by 3 feet 3 inches , covered the loins , and it was thinner at its middle near the vertebral column than towards its borders .
It is now I fear irreparably damaged and beyond reconstruction .
Broken up into countless pieces through hasty and incautious removal from the cliffs , these have in 15 years cracked and fallen into numberless smaller fragments ; the attempt to rejoin which would be a hopeless undertaking .
Scutes of three forms are readily distinguished : a , simple flat scutes ; f3 , keeled scutes ; and y , spined scutes .
Those of the first kind are most numerous .
Since none are now entire their shape and dimensions can only be approximately ascertained .
They varied greatly in different situations , since some were certainly more than 26 centims. across , whilst the breadth of others did not exceed 1 centim. Their thickness ranges between 3 centims. and 5 centims. Their deep or inner surface is smooth and their outer surface is studded with scattered tubercles .
( Plate 71 , fig. 3 .
) A smaller number of scutes are keeled .
The keel , in some , itself unsculptured , rises within a circle of one or more rows of tubercles which are separated from the margin of the scute by a sunken groove , the inner lip of which is thin and projects .
( Plate 70 , figs. 3 , 4 , and Plate 72 , fig. 4 .
) It is probable that into this groove fitted the thin edge of the adjoining scute , a mode of articulation giving flexibility with security against dislocation .
The deep surface of these scutes is smooth and sinuous .
A few of the keeled scutes have the deep surface angularly excavated .
( Plate fig. 7 .
Plate 73 , figs. 1-4 .
) The spined scutes are fewer than either of the other two forms .
They are all asymmetrical .
Their form is rudely triangular , the shorter are obtuse and the longer acute .
( Plate 71 , figs. 4-6 .
Plate 76 , fig. 1 .
) Their base is very stout , its outline is a rhomboid .
The blade projects in one of the most perfect , which , however , wants the tip , to 30 centims. beyond the base .
The long diameter of the base of this scute is 21 centims. and the shorter 11 centims. One edge of the blade is relatively straight and the other is incurved .
A similar difference in the direction of the borders is seen in the smaller scutes of this kind .
When a spined scute is placed upon its base on a flat surface the slant of the blade is seen to be considerable ; one surface , which in this position is upper , is nearly plane or sinuous transversely , whilst the other surface is transversely convex .
The plane or sinuous surface has its distal moiety deeply furrowed by vascular grooves .
With regard to the distribution of the different forms of scutes we have to guide us : cl , Mr. Fox 's impressions of the armour as he saw it first before it was disturbed and broken up\#151 ; he says that from its relations to the other bones he thought it covered the loins as a continuous shield ; / 3 , inferences drawn from the scutes themselves ; y , the preservation of a few scutes in situ in two regions .
The upper surface of the sacrum is still overlaid by a continuous flat scutal covering ornamented with tubercles , which dot it irregularly without definite grouping ( Plate 71 , fig. 3 ) .
I did not detect in it any marks of joints , and am therefore disposed to regard this as forming part of one large plate , which is certainly in its natural position .
The fortunate recovery of the piece sketched in Plate 73 , figs. 1-3 , shows that an upper row of carinate angularly excavated scutes covered the neural , and a lower row of similar scutes embraced the haemal spines of the tail .
One of these keeled hollowed scutes , which from its large size was probably situated at the root of the tail , is 21 centims. long , 12*5 centims. high , and the angular excavation of its base is 4 centims. deep ( Plate 73 , fig. 4 ) .
Another from near the end of the tail is only 2'5 centims. long by 5 centims. across , and its keel is quite dwarfed ( Plate 75 , fig. 5 ) .
The upper and the lower row of these keeled scutes did not quite meet , but they left a lateral interval filled by a series of smaller fiat scutes ( Plate 73 , fig. 1 ) .
Both forms with the diminution of the bulk of the vertebrae underwent a corresponding reduction , and they became towards the end of the tail small button or buckler-like studs , one of which is shown by Plate 75 , fig. 6 .
Thus the whole tail was sheathed in armour .
It has been already mentioned that the spined scutes are asymmetrical .
This alone would make it most unlikely that they formed a median dorsal crest .
That they were not so placed in the lumbo-sacral region is demonstrated by the preservation of flat scutes there in situ .
It is not improbable that the spined scutes and the unexcavated carinate scutes encircled with tubercles were grouped in lateral rows .
MR , J. W. HULKE ON THE POLACANTHUS FOXIT .
As no part of the endoskeleton referable to the scapular region and neck was discovered , it cannot be ascertained how far forwards the dorsal shield reached , neither from the material in its present state can any inference be drawn of the presence of ventral armour .
The presence of a haemal series of tail scutes suggests that the belly as well as the back may have been mailed .
The tissue of the scutes is distinctly bony .
The vascular canals in the cortex are large and very numerous ( Plate 70 , fig. 4 ) , suggestive of a stout epidermal covering .
These remains indicate an animal of low stature whose height at the rump probably did not exceed 3 feet .
Its strongly marked bones and their large joints speak of its immense muscular power , whilst the shortness of its limbs and the anchylosis of the lumbar vertebrae welding the loins and the sacrum into a long inflexible rod suggest an absence of the lithesome and agile movements of a terrestrial carnivore , and give probability to its having been a slowly moving vegetable feeder .
As regards its zoological position , its dinosaurian marks\#151 ; the inner femoral trochanter , the lower end of the tibia , and the forked ribs\#151 ; are so plain that its reference to this Order cannot be doubted .
Its place within the Order is also not uncertain .
From the Iguanodont family , as represented by its two best known genera and Iguanodon , Polacanthus differs widely in the form and proportions of its limb-bones and vertebrae , and by its very highly developed dermal armour , in comparison with which the scutes of Hypsilophodon and Iguanodon may without inaccuracy be called flimsy .
In its stoutly sheathed , crested tail , and its strong body mail Polacanthus repeats two striking features of the Liassic Scelidosaurus .
The scuted caudal vertebra of Polacanthus sketched in Plate 73 , fig. 1 , presents a resemblance to the figure of a corresponding vertebra accompanying Professor P. Owen 's " Monograph on Scelidosaurus , " which must strike the most superficial observer .
The trunk armour of Scelidosaurus is , however , much less developed than that of canthus , the spined scutes of the former , so far as these are known from the types preserved in the British Museum , are smaller than those of Polacanthus .
The resemblance of Polacanthus and Scelidosaurus is not restricted to their armour , for massive joints are a feature common to both ; the differences of shape and proportion of the limb-bones and vertebrae more than suffice , however , to prove their generic distinctness .
Omosaurus armatus , Owen , of oolitic times , had as large dermal spines as Polacanthus , but their shape is very different .
The dermal spines of the Jurassic Stegosaurus figured by Professor O. C. Marsh , in * American Journal of Science/ vol. xix .
, March , 1880 , plate x. , are apparently not smaller than those of Polacanthus , but their form differs from these ; the limb-bones of Stegosaurus are also more slender and the femur wants the inner trochanter .
From the lower chalk ( Huxley ) , which has also a somewhat similar mail , Polacanthus differs in the greater development of this protective covering and in its very dissimilar vertebrae .
It is to the Wealden Hylceosaurus that Polacanthus appears most closely related .
4 q 2 The resemblance of their dermal spines is very close , and their tibiae are remarkably alike .
Indeed , I do not know in any public or private collection any bone which the tibia of Polacanthusso nearly resembles as the type tibia of preserved in the national collection and a tibia which a few years since I brought before the Geological Society and provisionally referred to this dinosaur.* It was obtained from Brixton Bay , the locality which yielded these remains of Polacanthus .
We may not safely compare the metatarsals of this skeleton with those accredited to saurus , since the type specimen ( No. 2556 Brit. Mus. Cat. figured in the Brit. Foss .
Kept .
, Monog .
Hylceoscmrus , plate xi .
) is only conjecturally assigned to it , and may have belonged to a very different animal , for the length and slenderness of these metatarsals are not in harmony with the Hylaeosaurian tibia so short and with such expanded articular ends .
Neither are the type specimens of sacrum and the disconnected vertebrae in the British Museum available , because their reference to Hylceosaurus is also conjectural and still requires confirmation .
Such comparison , whatever its value may finally prove , however , shows that the sacrum of Polacanthus is much more massive and the thoracic vertebrae are longer and less stout than those reputed Hylaeosaurian fossils .
Explanation of Plates .
PLATE 70 .
Fig. 1 .
Posterior view of a vertebra from the front of the chest , t Fig. 2 .
Oblique lateral view of the same .
Fig. 3 .
Fragment of a large keeled scute with grooved margin and submarginal rings of tubercles .
k. Beginning keel .
Fig. 4 .
Sectional view at x in fig. 3 .
( This and fig. 3 are slightly reduced .
) PLATE 71 .
( All the figures in this plate are represented rather less than one-half their natural size .
) Fig. 1 .
Ventral view of sacrum and anchylosed lumbar vertebrae .
The numerals 1-5 , s. , mark the sacral , and the letters l.l.l. the lumbar vertebrae .
ng .
Nerve-groove .
* ' Quarterly Journal Geological Society , ' vol. xxx .
, 1874 , plate xxxi .
, figs. 1 , 2 .
t In all the representations of vertebra ' , c. Centrum , ns .
Neural spinous process , Prsezj gapophysis .
psz .
Postzygapophysis .
cl. Diapophysis , p. Parapophysis .
c.c. Capitulum cost\#174 ; , r. Rib-shaft .
MR. J. W. HULKE ON THE POLACANTHUS FOXII .
Fig. 2 .
Cross sectional outline at fracture through the third sacral vertebra .
Fig. 3 .
Dorsal view of fragment of the large scutal shield resting on the fourth and fifth sacral vertebrae .
Fig. 4 .
Lateral view of a large dermal spine .
Fig. 5 .
Basal view of the same .
Fig. 6 .
Edge view of the same .
Fig. 7 .
Keeled scute with angularly excavated base .
The border x in this and in fig. 4 correspond .
PLATE 72 .
Fig. 1 .
Oblique view of a caudal vertebra .
Fig. 2 .
Anterior view of the centrum of the same .
Fig. 3 .
Oblique lateral view of two metatarsal bones .
Fig. 4 .
Fragment of a large keeled scute with grooved border and submarginal ring of tubercles .
kKeel .
g. Groove .
PLATE 73 .
Fig. 1 .
Lateral view of a fragment of the tail , sheathed by haemal and neural keeled scutes , between which are the remains of a lateral series of small peltate scute tubercles .
The vertebral centra , c.c ' .
, are overlaid by bundles of ossified tendons , t. Fig. 2 .
Sectional view of the fractured surface at .
c. Crushed centrum .
kh Haemal and neural scutes .
Fie ' .
3 .
Foreshortened view of scute at x. b. marks the same border in this and in fig. 1 .
Fig. 4 .
Lateral view of a large keeled scute with angularly excavated base .
PLATE 74 .
Fig. 1 .
Front view of the right femur ( nearly f ) .
Fig. 2 .
Back view of same ( nearly f ) .
( This is represented upside down .
) cap. Caput femoris .
tr.m.Trochanter major .
tr.i .
Inner trochanter .
PLATE 75 .
Fig. 1 .
Front view of the left tibia ( nearly -f ) .
Fig. 2 .
Back view of the same ( nearly f ) .
cd .
Condyles .
pr.c .
Prsecnemial crest .
f.Fragment of fibula .
Fig. 3 .
Side view of a caudal vertebra , where the transverse process has disappeared , and the spinous process has become dwarfed .
Fig. 4 .
View of posterior surface of fig. 3 .
Fig. 5 .
Small keeled scute upon a vertebra from near the end of the tail .
Fig. 6 .
Small buckler-like scute from near end of tail .
PLATE 76 .
Fig. 1 .
Lateral view of a large dermal spine ( slightly reduced ) .
Fig. 2 .
Fragment of a rib .
Phil Trans. 1881 .
70 Beguvwig Ke\amp ; i'* Fac , 3 Fiq .
4 .
Fiq .
7 .
Halke .
-Phil .
Tram .
1881 .
Plate 72 .
^ t KAA , C A sj Wait i\gt ; ( ''V i\gt ; /t/ : ' .
oUL .
\#171 ; * % litft .
Hvdke .
Phil. Trans. 1881 .
Plate 76 J.W. HuJicc \lt ; \gt ; *.\lt ; V H Wt'ti I**/ i West N\lt ; tWWi\lt ; nn \amp ; C ' imp

