- 256

THE SHANKS of hooks of Type 3 were slender and straight or approximately straight, and the points were probably very uniform in shape. It is remarkable that not a single complete shank or even a fragment of a wooden shank of Type 3 has yet been recovered in Murihiku. Some fragments of bone shanks of the type are all that is at present known from that area. It is probable that the two slender hooks in Fig 34 are from Murihiku; even if from another district, they still represent Type 3 satisfactorily. Fig. 35 is after Beasley's Plate 62, there localized as Tahitian. Beasley states that he guessed this locality but that the correctness of the guess was confirmed by a similar hook in the Peabody Museum, Salem, which was recorded as from Tahiti. The specimens in the British Museum are stated1 to number a score, of which some were brought home by Vancouver. “The shanks of all these hooks are round lengths of hard red wood, of which the longest measures 11¼ inches, while the shortest is only 4¾ inches. The barbs are of bone, neatly attached to the base of the shanks with fine lashings, which have been covered with black gum.” I have not seen these hooks, and when reviewing Beasley's book in the American Anthropologist2 I rejected Tahiti and suggested the North West Coast, basing this suggestion on the presence of gum or wax covering the junction of shank and point. Such gum or wax is unrecorded in New Zealand or elsewhere in Polynesia. I have since examined the hooks in Fig. 34, all of which are proved by their Phormium tenax cord to be Maori. It may be suggested that the wax on the British Museum specimens is cobbler's wax added by some former collector-owner to stop the unravelling of the cord. Fig. 36 is a stone shank from the Nelson district with the very rare feature of a perforation at the snooding end of the shank. Fig. 37 is a slightly curved specimen in bone from Coromandel. Fig. 38 represents a wooden shank and bone point from Monck cave, Sumner, stated by the vendor to have been found in association. This specimen illustrates what is probably the normal method of junction of wooden shank and point, namely grooving of the former and bevelling of the latter. Fig 39 represents four specimens in the British Museum collections, no localizations being recorded. While the bone shank of the top right specimen is what is here regarded as normal the wooden shanks are all strongly curved. The specimen on bottom left has a point with barb on same side as the notches to take the binding. Fig. 40 represents twenty-five points all of which probably belong to Type 3. Those with bevelled edge were presumably designed to fit grooved wooden shanks, while the - 257 rest were presumably attached to bone (or rarely to stone) shanks. Fig. 41 is an unusual and interesting human-bone point, from Shag river mouth, in the Hornsey collection.

Age of Type 3. In Murihiku the type of point represented in Figs. 35, 38, 39, 40 is found, though rarely, in moa-hunter deposits. It is exceedingly common in surface deposits. In Otago Museum about 600 perfect localized specimens are on exhibition and several times that number, mostly imperfect, are in store. In the North island it seems to be rather uncommon. No similar points from other parts of Polynesia are on record. However, points with very similar barb-ends, though with quite dissimilar bases, appear on bonito hooks in a good many Polynesian groups, and many one-piece hooks in hone and shell from widely separated points in Polynesia have very similar barbs. This is true also of Micronesia. No related forms are at present known from the Solomons. The British Museum Handbook3 shows a composite hook from the American N.W. Coast which closely resembles Type 3 both in shank and point. It seems fair to conclude that Type 3 is as old as any type in Polynesia; that it was not common in antiquity; and that in Murihiku it boomed in recent times.

This is the New Zealand version of the composite hook with pearl-shell called pa in most parts of Polynesia and generally used in catching bonito. As pearl-shell does not exist in New Zealand the incoming Polynesians must substitute some other material. This was found at first in brightly coloured stones which are present though not common in the “Polynesian continent”. Greenstone was rarely used, due probably to its precious character and to its weight. Slate was commonly employed in the North island, but in Murihiku the most prized material appears to have been argillite. The site which, up to the present time, has produced the most numerous and the most handsome shanks of Type 4 is the moa-hunter camp at Shag river-mouth excavated by David Teviotdale.4 Here Teviotdale found Fig. 42 with its perforated, unbarbed, bone point lying in position beside the shank. Dr. J. E. L. Simcox tells me he found a similar stone shank associated with a similar unbarbed point on an East Coast, North island site. Mr. J. M. Gray has in his collection a smaller, but beautiful, stone shank and a similar point found on the surface of blown sand at Curio bay, south Otago, about four feet separating the two pieces. The present writer found a similar point associated with moa bones on surface sand near Bottle point, Durville island. A number of stone shanks lacking points have been found in moa-hunter deposits.5 The antiquity of the type in New Zealand is thus demonstrated. No example with flax binding or snood is at present known, so the fashion of using stone as shank-material must have been of brief duration. Double perforation of the point must likewise have been short-lived, for this feature is of extreme rarity in New Zealand. Points of this shape have what may be described as a marginal distribution in Polynesia—Tikopia, Ellice, Samoa, Manihiki, Rakahanga, Pukapuka, and New Zealand. In the Marquesas the two perforations in the point remain, but otherwise the shape of the point has been changed. In the Society group the upper perforation remains, but a notch and projection have been substituted for the lower perforation and a barb has been added to the point. Many of the Murihiku stone shanks - 258 retain the pearl-shell shape in that they are deep at the snood-end but become shallow distally. Figs 42, 43 and 44 have a sharklike mouth, which seems to be a Murihiku feature. In the perforation at distal end Fig. 45 has a feature unique among stone shanks, though the Otago Museum has a bone shank6 with two perforations at distal end. Fig. 46 has had a perforation begun at the widest part, but, a fault developing, the perforation has been completed near the proximal end. Fig. 44 has the archaic Polynesian feature of notched ornamentation.

Variety B, bone shanks. It is not known why stone shanks went out of fashion, but it may be conjectured that their weight made them difficult to handle. Bone was substituted, and in the North island this material was used until European times for shanks of hooks derived from this type—the interesting variety apparently limited to the East Cape area and often figured.7 Moa-hunter sites in Murihiku have provided a number of shanks made of moa-bone, exemplified by Fig. 47. Moa-bone is never thick enough to provide a triangular cross-section, and the perforation therefore becomes dorso-ventral. It would appear that when once this shape had been established in bone it was copied in stone. Fig. 48 is an example of this variety made of inferior nephrite, and the Otago Museum has a second, closely similar, example.

The shape of points used with these two varieties of shanks calls for discussion. The unbarbed point with two perforations has already been established as one form. Fig. 49 shows one of these bi-perforate points and five imperforate points which are thought to have been used with the same shanks. They are found in the same deposits and are much commoner than the perforate forms. Those who excavated them regarded these imperforate unbarbed points as belonging to shanks of the variety illustrated by Fig. 47. However, at the Shag river site stone shanks were numerous, but there was only one bi-perforate unbarbed point. The imperforate points were numerous. At Nauru island similar points were used with shanks made from stalactite. It seems fair to conclude that both bi-perforate and imperforate unbarbed bone points of the kinds figured were used in Murihiku with stone and with bone shanks. Figs. 50 and 51 represent imperforate points which vary in minor respects from those shown in Fig. 48.

Variety B, wooden shanks. The two wooden shanks in the Otago Museum collection of which Fig. 52 is an example, follow Variety B closely except that the bone point was attached through a hole and not by binding. Thus the shape of the point differed widely from those of Figs. 49, 50, 51. In many cases the proximal end was drilled through, and the point was held in the shank by means of a peg passed through the hole. Where no hole was drilled, the point was probably secured by wedging, a method which must have tended ultimately to split the shank. This change in shape of point meant the use of a point identical with that already being used for Type 5, the barracouta hook. So Fig. 52 is to be regarded as a hybrid, its shank belonging - 259 to Type 4 but its point being of the shape characteristic of Type 5. In classification, shank is more important than point.

Fig. 53 represents three points which might equally well be used for Type 4, Variety B, or for Type 5. The perforation of the middle point is begun but not finished. Upper and middle points are made from canine teeth of sea-leopards, and are the handsomest of all Murihiku fish-hook points. They are especially common on Long beach and Murdering beach where they are always found in surface-layers. They have never been recorded from moa-hunter sites and may therefore fairly be regarded as of recent origin.

Age of Type 4. From its definition Type 4 must go back to the beginning of culture in New Zealand, and this generalization is fully supported by stratigraphic evidence. It is possible that the stone shank of northern districts, which is circular in cross-section as contrasted with the triangular section of the south, and differs in some other features from the southern stone shank, has a different descent, and comes from an ancestor closely allied to the stalactite shanks of Tikopia and Ocean islands.8 Though in the north shanks continued until European times to be made from moa-bone there is at present no evidence that moa-bone shanks were made in Murihiku in any but the earliest period. Wooden shanks of the shape exemplified by Fig 52 were probably in use in early European times. There is at present no evidence of their antiquity.

Type 5. Composite fish-hooks with straight and rather massive wooden shanks, used for catching barracouta. Fig. 54 illustrates two examples of this type.9 The wood of the shank is rimu (red pine). The hook is attached by a six-foot line to a stout rod of about the same length. European fishermen copy the Maori hook exactly, except that instead of a bone point they use a nail which is driven through the shank from the back at an angle of about 45°. The nail continues for about half an inch beyond the front of the shank, and the point is then bent round through about 90°. It is, I believe, on the basis of these nail points that the very numerous bone points exemplified by Fig. 55 are classified by collectors as points of barracouta hooks. I have never seen points of this kind set in shanks, but the evidence of the nails formerly seemed conclusive. So fish-hook points of shape represented by Fig. 55 have always been called “barracouta points”. This identification is shaken by the fact that a number of points of otherwise identical shape are grooved at the base presumably to hold a flax tie. Such grooving seems to forbid the method of attachment used in Figs. 52 and 54. Consideration of this point is deferred to the next section. Fig 56 is a greenstone point of the same shape as the bone ones of Fig. 54, found by Mr. S. V. Johnson at Kaikai's beach.

Fig. 57 represents a group of points, known to collectors from their shape as ‘doll's legs’. They suggest attachment to a hook of ruvettus type, but there is actually no evidence as to their place in classification. Fig 58 represents a group which in the past has always been associated with Fig. 55. But if these latter points were driven through a hole in the base of a wooden shank then the grooved examples - 260 of Fig. 58 cannot have been attached to a shank in the same way as the points of Fig. 55. It is safest to leave the matter in suspense until additional evidence is available. The same must be done with the characteristic shape of Fig. 59, a rarity belonging to the “human-foot” group. Fig 60 is also an uncommon shape. The splendid left-hand specimen is in Mr. Murray Thomson's collection from Murdering beach, and suggests a fish-spear point rather than the point of a fish-hook. Figs. 61 and 62 represent the numerous large unbarbed points found in the earliest layers in Murihiku, together with stone and moa-bone shanks but never found in recent deposits. Nearly all are too large for the stone shanks, and a majority are too large for the largest shanks made of moa-bone. It is possible that they were attached to very large wooden shanks of shape resembling Fig. 47, but some of them suggest rather strongly that there existed at that time in Murihiku a hook of the halibut hook type. No part but the points has survived.

Only two one-piece utility hooks in stone are at present known from Murihiku, both being fragmentary. One, in greenstone, is represented by Fig. 13. The other is an unfinished example in altered limestone from Pounawea. The Otago Museum has a greenstone point and a large one of schist of Type 3, three greenstone points of Type 5, and nine unfinished or fragmentary basalt examples also of Type 5.

Many points belonging to Type 5 are ground out of mandible and ascending ramus of dog or seal, and it is probable that the same region of the human jaw was also used. Occasionally a mandible with the canine tooth still embedded (Fig. 63) was utilized as a point. Beasley figures a beautiful example (Plate 24) in which the tooth, a molar, has been ground so as to form the heel of the “foot”. Canine teeth were also frequently used as points for composite hooks.

It is not profitable to discuss age or origin of Murihuku types unless consideration is given to the distribution of these types outside New Zealand. From several points of view it would be interesting to discuss fish-hooks area by area, for example in the northern Cooks or in the Carolines. But in a typological study each type and variety is best run through area by area in succession. Most of the hooks here-figured are from the Otago Museum collections, and none of these have been figured before. Unless some special point is involved, as in the case of the two Hawaiian hooks, it does not seem necessary to figure hooks from areas like the Society group, the Tuamotus, the Marquesas, or Hawaii, all of which have already been fairly adequately published.

Type 1, Variety A. Sub-circular one-piece hooks. Fig 64 is an example in black-lipped pearl-shell, collected by Captain James Fowler at Manihiki. The bait-line is attached to the snooding-point. Fig 65 is an example in gold-lipped shell from Pukapuka collected by Drury Low. Fig. 66 is a fine Easter island hook in black basalt, originally in the Brander collection, presented by Charles Nordhoff. Fig. 67, a hook in dark turtle-shell from the Carolines, has transferred the snooding-knob from its normal position to the inner bend. Fig. 68, - 261 gold-lipped pearl-shell from the Carolines, has the snooding-knob elongated in a way unparalleled in Polynesia. Fig 69, abalone shell from Santa Barbara, California, has the snooding-knob elongated in an identical way. Fig. 70, abalone shell, from Santa Barbara, California, has its snooding-knob of a shape which is found throughout Polynesia and the Carolines.

Type 1, Variety B. One-piece hooks with point strongly bent. Fig. 71 is a black-lipped pearl-shell hook collected at Pukapuka by F. W. Platts. Fig. 72 is a finely cut gold-lipped pearl-shell hook found by L. E. Trenn in the course of excavation at Manihiki. Fig. 73, a well-shaped hook in basalt, was secured on Pitcairn by James Norman Hall from an islander who had recently dug it up in his garden.

Type 1, Variety C. One-piece hooks with legs approximately parallel. Fig 74 is a pearl-shell example dug up on Manihiki by L. E. Trenn. Fig. 75, an Easter island hook in human bone, was collected there by Henri Lavacheri. Fig. 76, in black-lipped pearl-shell, was collected in Ulawa, Solomons, by W. G. Ivens.

Barb on shank leg. Fig. 77, collected on Easter island by Henri Lavacheri, is the shank-leg of a broken hook in human bone. Fig. 78 follows J. G. McAlister, and is from Kohoolawe, Hawaiian islands.

Barb and shank simulating human foot. Fig. 79 is a specimen in pearl-shell from Pukapuka in the collection of Mr. J. D. Campbell, Rarotonga. Fig. 80 was collected by F. W. Platts at Pukapuka, and is of dark-lipped pearl-shell. Fig. 81 is of the same material and was collected at Manihiki by Capt. James Fowler. Fig. 82, after McAliser, is from Kahoolawe, and appears to be made from human bone. Fig. 83, from Beasley, after Kubary, indicates that the “foot” goes back at least to the Carolines.

Type 2. Composite hooks conforming in shape to Type 1. It is not necessary to document the varieties of this type so fully as was done with Type 1. Fig. 84 is a fine example of the shark-hook of Tongarewa. Fig. 85 is a small example of the kind of hook usually described as “ruvettus hook”. This specimen is stated to be from Manihiki. Fig. 86, collected by Dr. Ritchie at Fakaofo, Union Group, is of interest from its similarity to Fig. 24.

No instances of barb on shank-leg are known in Type 2, nor, outside New Zealand, can satisfactory examples of “human foot” be quoted in this type.

Type 3. Composite hooks with straight slender shanks. It has been shown that doubt attaches to Beasley's attribution of this type to Tahiti, and it is not at present certainly known in Polynesia except New Zealand. Fig. 87 illustrates the type in the Carolines.

Type 4. Composite hooks resembling the bonito-hook of tropical Polynesia. Since the earlier part of this paper was written Duff has published an account of his excavation of a site of moa-hunter date, including burials, at the Wairau bar.10 Among many important finds was a typical hook of this type, with stone shank triangular in cross section and bi-perforate unbarbed bone point This was found with skeleton 5 which was also accompanied by a perforated moa-egg. The hook now becomes type specimen so far as antiquity of type is concerned. The burials at the Wairau site represent the oldest phase of culture in New Zealand. Earlier in this paper I suggested that - 262 Type 4 evolved in New Zealand. Duff's specimen shows that it was already fully evolved in the earliest New Zealand cultural phase. Fig. 88,11 after a Bishop Museum photograph of material found in a grave on Fanning island, indicates that Type 4 was already in existence before the Polynesian colonization of New Zealand, and that the theory of its New Zealand origin must be discarded. Such being the case, it would be quite possible to argue that the North island hook with a stone shank circular in cross-section also originated outside New Zealand, and that it is related to the stalactite shaft, circular in cross-section, of Nauru, and even, though less probably, to the stalactite shaft of Tikopia. Until further evidence is available such suggestions remain merely tentative.

Fig. 89 (Beasley, Plate 147) represents a hook and point resembling Type 4 in a general way, excavated by Christian at Ponape in the Carolines. The point is similar. The principal difference in the shank is the presence of a knob in place of a perforation for attachment of snood. The same feature is present in Fig. 90, from S.E. Solomons. This hook appears to present a number of homologies with Type 4 in its Murihiku homeland, but the time seems not yet ripe for their discussion. Fig. 91, the familiar trolling-hook of the eastern and central Solomons also presents homologies.

Type 5. The barracouta-hook. Literature and collections alike show few hooks, or none, from other parts of Polynesia that are at all closely related to Type 5.

Indeterminate types. In view of the absence of shanks it is not at present profitable to attempt comparisons of these forms.

A consideration of the Murihiku hooks here figured will show that they differ a good deal from the fish-hooks of the North island, and that the difference increases as we move northward in that island. The comparative forms figured from overseas indicate clearly that Murihiku relationships are closest with Marginal Polynesia, in which must be included the northern Cooks. Back of Marginal Polynesia stands the Caroline group, closely related and showing many ancestral forms. All the hooks of Melanesia, from the eastern Solomons northward, are derived from the Carolines. The failure of diggers in the Philippines to publish their finds for more than twenty years prevents comparative studies from pushing westward beyond the Carolines. We have seen that the abalone hooks of Santa Barbara, California, seems to reflect recent Caroline influence. Fig. 92, from the Handbook to the Ethnographic Collections of the British Museum (1910 edn., Fig. 237) illustrates allied forms from the American North-West Coast. The relationship here is ancient, and the allied forms must have passed from coastal Asia into America by the Aleutian islands or by Behring strait.

- i

Fig. 39.

Illustration

- ii Page is blank

- 263

Figs. 35–38.

Illustration

- 264

Fig. 40.

Illustration

- 265

Figs. 41, 42.

Illustration

- 266

Figs. 43, 44.

Illustration

- 267

Figs. 45, 46.

Illustration

- 268

Fig. 47.

Illustration

- 269

Fig. 49.

Illustration

- 270

Figs. 48, 50, 51.

Illustration

- 271

Fig. 52.

Illustration

- 272

Figs. 53, 55–57.

Illustration

- 273

Fig. 54.

Illustration

- 274

Figs. 58–61.

Illustration

- 275

Fig. 62.

Illustration

- 276

Figs. 63–66. (66, loc. Easter id.)

Illustration

- 277

Figs. 67–69. (69, loc. Santa Barbara.)

Illustration

- 278

Figs. 70–73. (70, loc. Santa Barbara; 73, Pitcairn id.)

Illustration

- 279

Figs. 74–77. (75, 77, loc. Easter id.; 76, Ulawa.)

Illustration

- 280

Figs. 78–81. (78, B. Mus.,Hawaii; 79-81, loc. Pukapuka.)

Illustration

- 281

Figs, 82–85. (83, loc. Carolines.)

Illustration

- 282

Figs 86–88. (87, loc. Carolines; 88, loc. Fanning id.—Bishop Museum.)

Illustration

- 283

Figs 89–91. (89, loc. Ponape, Carolines; 90, Eastern Solomons.)

Illustration

- 284

Fig. 92 (loc. American N.W. coast.)

Illustration

- 285

(All natural size except where otherwise specified.)

• Fig. 35. After Beasley, Plate 62.
• Fig. 36. Tonga bay, Nelson. Fels Collection. Otago Museum. D.24. 1208.
• Fig. 37. Slipper island, Coromandel. Bollons. Dominion Museum.
• Fig. 38. Moncks cave, Sumner. Fels Fund. Otago Museum. D.29. 690.
• Fig. 39. British Museum.
• Fig. 40. Otago Museum.
• Fig. 41. Hornsey Collection, Timaru.
• Fig. 42. Shag river-mouth. Teviotdale. Otago Museum. D.28.133.
• Fig. 43. Shag river-mouth. Teviotdale. Otago Museum. D.27.1054.
• Fig. 44. Shag river-mouth. G. Griffiths Colln. Red argellite.
• Fig. 45. Southland. Red argellite. Otago Museum. D.29.59.
• Fig. 46. Pounawea, Otago. Red argellite. Otago Museum. D.29.58.
• Fig. 47. Little Papanui, Otago. Moa-bone. Otago Museum.
• Fig. 48. North Otago. Nephrite. Otago Museum. D.30.969.
• Fig. 49. Otago. Bone. Otago Museum.
• Fig. 50. Shag river-mouth. Human bone. Hornsey Colln.
• Fig. 51. Kings rocks. Bone. Otago Museum. D.39.3250.
• Fig. 52. Otago. Rimu and bone. Otago Museum. D.18.149.
• Fig. 53. Otago. Bottom example bone, others sea-leopard canines. Otago Museum.
• Fig. 54. Otago. Otago Museum.
• Fig. 55. Otago. Bone. Otago Museum.
• Fig. 56. Kaikai beach. Greenstone Johnson Collections.
• Fig. 57. Otago. Bone. Otago Museum.
• Fig. 58. Otago. Bottom example human skull, others limb-bone. Otago Museum.
• Fig. 59. Otago. Bone. Otago Museum.
• Fig. 60. Otago. Bone. Left example, Thomson Colln., others Otago Museum.
• Fig. 61. Otago. Bone. Otago Museum.
• Fig. 62. Otago. Bone. Otago Museum.
• Fig. 63. Otago. Bone. Otago Museum.
• Fig. 64. Black-lipped pearl. Manihiki. Colln. Capt. Fowler. Otago Museum. D.36.765.
• Fig. 65. Gold-lipped pearl. Pukapuka. Colln. Drury Low. Otago Museum. D.34.395.
• Fig. 66. Basalt. Easter island. Colln. Charles Nordhoff. Otago Museum. D.33.1592.
• Fig. 67. Turtle shell. Carolines. Otago Museum. D.24.1905.
• Fig. 68. Gold-lipped pearl. Carolines. Otago Museum. D.28.341.
• Fig. 69. Abalone. Santa Barbara, California. Otago Museum. D.31.1330.
• Fig. 70. Abalone. Santa Barbara, California. Otago Museum. D.33.13.

- 286

• Fig. 71. Pearl. Pukapuka. Colln. F. W. Platts. Otago Museum. D.21.140.
• Fig. 72. Gold-lipped pearl. Manihiki. Excav. L. E. Trenn. Otago Museum. D.37.351.
• Fig. 73. Basalt. Pitcairn. Colln. J. N. Hall. Otago Museum. D.34.1006.
• Fig. 74. Pearl. Manihiki. Excav. L. E. Trenn. Otago Museum. D.41.167.
• Fig. 75. Human bone. Easter. Colln. H. Lavacheri. Otago Museum. D.36.990.
• Fig. 76. Black-lipped pearl. Ulawa. Colln. Walter Ivens. Otago Museum. D.26.302.
• Fig. 77. Human bone. Easter. Colln. H. Lavacheri. Otago Museum. D.36.992.
• Fig. 78. Human bone. Kahoolawe, Hawaii. McAllister, Bishop Mus. Bull., 115, Fig. 5, 9.
• Fig. 79. Pearl. Pukapuka. Colln. J. D. Campbell, Rarotonga.
• Fig. 80. Black-lipped pearl. Pukapuka. Colln. F. W. Platts. Otago Museum. D.21.139.
• Fig. 81. Black-lipped pearl. Manihiki. Colln. Capt. Fowler. Otago Museum. D.36.764.
• Fig. 82. Human bone. Kahoolawe, Hawaii. McAllister, Bishop Mus. Bull., 115, Fig. 5.
• Fig. 83. Carolines. Beasley, Plate 153, Fig. 4. After Kubary.
• Fig. 84. Wood. Tongarewa. Otago Museum. D.39.1702.
• Fig. 85. Wood. Manihiki. Otago Museum. D.24.1909.
• Fig. 86. Wood. Fakaofo, Union Group. Colln. Dr. Ritchie. Otago Museum. D.26.3.
• Fig. 87. Turtle bone and turtle shell. Carolines. Otago Museum. D.23,865.
• Fig. 88. From Bishop Museum photo. Objects dug from tomb of Tongan type on Fanning island.
• Fig. 89. From Beasley, Plate 147. After Christian “Caroline islands”.
• Fig. 90. Pearl shell and turtle shell. S.E. Solomons. Otago Museum. D.23.264.
• Fig. 91. Pearl shell and turtle shell. Ulawa. Otago Museum. D.26.341.
• Fig. 92. N.W. Coast, N. America. From B.M. Handbook, 1910 edn., Fig. 237.