A MONOGRAPH OF THE RECENT CEPHALOPODA A MONOGRAPH OF THE RECENT CEPHALOPODA BASED ON THE COLLECTIONS IN THE BRITISH MUSEUM (NATURAL HISTORY) PART I OCTOPODINAE By G. C. ROBSON, M.A. ASSISTANT KEEPER IN THE DEPARTMENT OP ZOOLOGY LONDON : PRINTED BY ORDER OF THE TRUSTEES OF THE BRITISH MUSEUM SOLD AT THE BRITISH MUSEUM (NATURAL HISTORY), CROMWELL ROAD, S.W. 7 AND BY B. QUARITCH, LTD.; DULAU & CO., LTD.; THE OXFORD UNIVERSITY PRESS; AND WHELDON & WESLEY, LTD., LONDON; ALSO BY OLIVER & BOYD, EDINBURGH 1929 Issued 21th Jidy, 1929] (All rights reserved) Printed in Great Britain by Richard Clay & Sons. Limited, BUNGAY, SUFFOLK. PREFACE This volume is the first instalment of a systematic review of recent Cephalopoda, and deals with the large subfamily Octopodinse, containing forms to which the popular name “ Octopus ” is applied. The collection of these animals in the Zoological Department is a rich one, containing the types of species described by Dr. J. E. Gray and Dr. W. E. Hoyle (“ Challenger ” collection) and other specimens of historical importance. In addition, the author has visited several of the larger continental museums, and other museums have generously allowed him to borrow valuable material or have supplied photographs of types which could not be lent. For these and similar courtesies the thanks of this Department are due to the following institutions and individuals : — Zoologisches Museum (Berlin), Indian Museum (Calcutta), Royal Scottish Museum (Edinburgh), Senckenbergisches Institut (Frankfurt a.M.), University Museums (Gottingen and Jena), ’s Rijk’s Museum (Leiden), University Museum (Leipzig), Musee d’Histoire Naturelle (Paris), Royal Museum (Stockholm), Musee Zoologique de l’Universite and de la Ville (Strasbourg), United States National Museum (Washington) ; Dr. P. Bartsch, Dr. L. Blote, Dr. S. S. Berry, Prof. G. Grimpe, Dr. F. Haas, Dr. H. Hoffmann, Prof. L. Joubin, Miss A. L. Massy, Prof. N. Odhner, Dr. B. Prashad, Dr. B. Rensch, Mr. J. R. le B. Tomlin, and Mr. R. Winckworth. It has long been recognised that the species of Octopodinse are par¬ ticularly difficult to define. Their external characters are often subject to alteration, as, for example, by the action of preservatives, making it hard to obtain series of specimens in comparable condition. Moreover, as in many other groups of animals, the description of “ new species ” has gone on without any critical revision of the species supposed to be already “ known.” Much of the obscurity that has thus arisen still defies, and will probably always defy, elucidation. Meanwhile it is hoped that the revision here attempted will facilitate future study of the group. The author has indicated on pp. 1, 4 and 31 some problems relating to evolution and adaptation in this group of animals, as they present them¬ selves to the systematist. W. T. Calm an, Keeper of Zoology. British Museum ( Natural History), April 22, 1929. CONTENTS PAGE I, Introduction . 1 (a) Scope of the work ....... 1 (b) Systematic position of the subfamily Octopodinae . 2 (c) Classification ........ 3 (d) Historical ........ 5 II. Structure . 6 III. Habits, etc . 19 IV. Development . 23 V. Variation in Octopus vulgaris . 24 VI. Phylogeny and Classification . 31 VII. Geographical Distribution . 36 VIII. Method of Measurement : Bibliographical Method : Abbreviations . 38 IX. Synopsis of Classification . 40 X. Table of Measurements . 42 XI. Systematic Treatment . . 56 Bibliography . 222 Index . . 231 vii LIST OF PLATES PLATE I. II. III. IV. V. VI. VII. “ Sepia octopodia ” (presumed syntype; Uppsala Museum). Octopus berenice (holotype). Octopus areolatus , var. ovulum. Octopus rugosus , eggs. Octopus teuthoides (holotype). Octopus ocellatus (holotype). Octopus hardwickei (syntvpe). Hapdlochlaena lunulata. Octopus medoria (holotype). Octopus aegina (syntype). Octopus cephaea (holotype). Octopus joubini (holotype). Octopus taprobanensis (holotype) . Enteroctopus sp. Octopus areolatus (syntype; Leiden Museum). Octopus ocellatus (syntype of Octopus areolatus , Orb. ; Leiden Museum). Fig. 1. 2. 3. 4. 1. 2. 1. 2. 1. 2. 1. „ 2. ,, 3. „ 1. 2. viii LIST OF TEXT-FIGURES PIG. PAGE 1. Types of web in the Octopodinae ...... 8 2. Outline drawings of various types of funnel-organ ... 10 3. Ocelli of various species of Octopus . . . . . .11 4. Diagrams of (a) simple, and (b) asymmetrical seriation in the rhachidian tooth of Octopus . . . . . . .13 5. Diagrams illustrating the relation of ink sac to liver . . .15 5 a. Octopus vulgaris. Mandible ....... 28 6. Octopus vulgaris. Variation in hectocotylus .... 60 7. Octopus vulgaris. Penis . . . . . . . .61 8. Octopus rugosus. Radula ....... 66 9. Octopus rugosus. Hectocotylus of various examples ... 69 10. Octopus verrucosus. Hectocotylus . . . . . .75 11. Octopus tonganus. Radula . . . . . . .78 12. Octopus tonganus. Hectocotylus . . . . . .78 13. Octopus bimaculatus. Hectocotylus . . . . . .79 14. Octopus tenebricus. Hectocotylus ...... 81 15. Octopus duplex (type). Ra’dula ...... 83 16. Octopus berenice. Funnel-organ ...... 85 17. Octopus micropintlnalmus. Funnel-organ ..... 89 18. Octopus horridus. Hectocotylus ...... 92 19. Octopus horridus. Penis ........ 92 20. Octopus globosus. Hectocotylus ...... 94 21. Octopus cyanea. Radula ....... 96 22. Octopus cyanea. Penis ........ 97 23. Octopus cyanea. Hectocotylus ...... 97 24. Octopus tetricus. Radula ....... 99 25. Octopus tetricus. Penis ........ 99 26. Octopus tetricus. Hectocotylus ...... 99 27. Octopus macropus. Outline of head and mantle in two Japanese specimens .......... 103 28. Octopus macropus . Penis . . . . . . .104 IX X LIST OF TEXT-FIGURES MG. 29. Octopus macropus (U.M., Leipzig). Hectocotylus 30. Octopus macropus. Female reproductive organs illustrating varia¬ tion of oviducal gland ........ 31. Octopus aegina. Hectocotylus . . . . . . • 32. Octopus aegina. Penis and diverticulum ..... 33. Octopus ocellatus (Holotype). Hectocotylus .... 34. Octopus ocellatus. Penis ........ 35. Octopus ocellatus. Oviduct ....... 36. (a) Octopus ocellatus (type), (b) 0. areolatus. (“ Challenger.”) Funnel-organs ......... 37. Octopus areolatus. (“ Challenger.”) Radula .... 38. Octopus pallida. Radula ....... 39. Octopus calif ornicus. Hectocotylus ...... 40. Octopus dojleini. Hectocotylus ...... 41. Octopus gilbertianus ........ 42. Octopus teuthoides. Funnel-organ ...... 43. Octopus amboinensis (type) ....... 44. Octopus amboinensis (type). Funnel-organ .... 45. Octopus defilippi. (a) Var. dama. (b) Typical form . 46. Octopus defilippi. Radula. (a) Var. dama. (b) Typical form . 47. Octopus defilippi. Reproductive organs ..... 48. Octopus defilippi. Male reproductive organs .... 49. Octopus defilippi. Penis ........ 50. Octopus niveus. (a) Funnel-organ, (b) Hectocotylus 51. Octopus australis. Hectocotylus . . . . . 52. Octopus schultzei. Funnel-organ ...... 53. Octopus schultzei. Hectocotylus ...... 54. Octopus pusillus, Hectocotylus ...... 55. Octopus validus. Hectocotylus ...... 56. Octopus joubini ......... 57. Octopus joubini. Funnel-organ ...... 58. Octopus joubini. Radula. ....... 59. Octopus verrilli, var. palliata ....... 60. Macrotritopus Jcempi ........ 61. Macrotritopus elegans (type). Mantle . 62. Macrotritopus elegans. Funnel-organ . 63. Enteroctopus megalocyathus. Radula . 64. Enteroctopus megalocyathus. Hectocotylus . . . . 65. Enteroctopus megalocyathus. Penis ...... 66. Enteroctopus eureka. Male reproductive organs . . . . 67. Enteroctopus eureka. Hectocotylus ...... PAGE 104 105 114 114 121 121 121 122 122 127 129 130 131 133 134 134 136 137 138 139 140 142 145 148 148 150 154 161 162 162 163 170 171 171 176 177 177 179 180 LIST OF TEXT-FIGURES xi PIG. PAGE 68. Enteroctopus eureka. Penis . . . . . . .180 69. Enteroctopus sp. Radula . . . . . . .181 70. Cistopus indicus. Hectocotylus ...... 183 71. Joubinia fontaniana. Penis . . . . . . .188 72. Joubinia fontaniana. Hectocotylus ...... 188 73. Joubinia campbelli. Hectocotylus ...... 190 74. Joubinia campbelli. Penis . . . . . . .190 75. Macrochlaena winckworthi. Male reproductive organs . .194 76. Macrochlaena winckworthi. Penis . . . . . .195 77. Paroctopus digueti. Funnel-organ . . . . . .198 78. Paroctopus digueti. Pallial sculpture . . . . .198 79. Paroctopus digueti (type). Hectocotylus . . . . .198 80. Paroctopus hongkongensis (type). Radula ..... 200 81. Paroctopus hongkongensis. Hectocotylus ..... 200 82. Paroctopus apollyon. Penis ....... 203 82a. Paroctopus yendoi. Outline of head and mantle .... 206 83. Hapalochlaena lunulata. Radula ...... 209 84. Hapalochlaena lunulata. Mandibles ...... 209 85. Hapalochlaena lunulata. Hectocotylus ..... 210 86. Hapalochlaena lunulata. Penis ...... 210 87. Hapalochlaena maculosa. Hectocotylus . . . . .213 88. Hapalochlaena maculosa. Penis ...... 213 89. Octopus hoylei, var. annae. Hectocotylus ..... 220 BRITISH MUSEUM 7 AUG 29 NATURAL HISTORY. A MONOGRAPH OF THE RECENT CEPHALOPODA I. INTRODUCTION. (a) Scope of the Work. The Octopodinae, which form the subject of the first part of this monograph, are commonly known as “ devilfish ” or “ octopods.” They are a subfamily of the eight-armed Cephalopoda and include those forms which have been usually placed in the genus Octojpus, such as the Common Octopus of the East Atlantic and Mediterranean. These animals are exclusively marine, and, like the Amphineura and Scaphopoda among other Molluscs, are not found in fresh or even brackish waters. The majority of the species live in depths of less than 100 fathoms, and are found in all the tropical and temperate seas of the world. They are not found in high latitudes, or in water of low temperature. In size they range from giant forms like Octopus apollyon, which spans nearly 28 feet with its arms, to small species which span barely two inches. As far as it is possible to speak of their habits and disposition they may be described as carnivorous, solitary and aggressive. The Octopodinae form the largest division of the Order Octopoda, and the species placed in this subfamily easily outnumber those in the other groups. As far as the divergence of species is concerned they are in a far more active evolutionary state than (e.g.) the Cirrata and Eledonellidae. Morphologically considered, however, the group is rather featureless. Most of the species are of an unspecialized type and differ from one another in trivial and apparently insignificant details. The work of which this volume is the first instalment has for its object a complete study of the morphology and evolution of the Class Cephalopoda, and the second volume will deal with the more interesting and specialized Bathypolypodinae, Argonautidae, etc. As a prelude to this study, however, it is necessary to deal with the classification of the less specialized Octopodinae, primarily with the object of seeing if in that large and confused assemblage of species there are any well-marked morphological tendencies which might cast some light on the evolution of the more specialized forms. The present work deals with inevitable questions of specific identity and variation, synonymy, etc., but an attempt has been made in Section VI and at various points in the systematic part to direct attention to wider and more interesting topics. The web, for example, is highly differentiated (see p. 7) in size and shape. It is possible that it may be of use either in brooding or in locomotion, but we do not know if the various characteristic shapes which it assumes are related to differ ences of function or due to non- adaptive tendencies. Nevertheless its modification is a highly distinctive B. M. CEPH. B 2 A MONOGRAPH OF THE CEPHALOPODA feature in Octopod evolution. The first signs of modification are to be found in the Octopodinae. Insignificant in forms like Octopus defilippi the web begins to increase in size in Octopus vulgaris. It is very deep in Hapalochlaena and Bathypolypus and finally in the Cirrata it forms a large bell-like structure conferring on the animal a medusiform appearance. There is no doubt that in the structure of the Octopodinae are to be seen many tendencies which are more fully realised in the deep-water Bathypolypodinae and Cirrata. The shortening of the arms, the deepen¬ ing of the web, the reduction of the ink sac and gills, the narrowing of the pallial aperture are features which occur in the Octopodinae and find their maximum expression in the Cirrata.* This is perhaps the fact of most general interest which emerges from the study of these animals, and its interest is heightened, if we recollect that these tendencies are seen in animals which are principally of littoral habitat. This subject will be more fully discussed in the second volume. For the present it is enough to say that the Cirrata and the true “ Octopods ” probably diverged from one another at a comparatively early stage in Octopod evolution. The special features in which certain of the Octopodinae tend to resemble their deep-water relatives are more likely to be due to adaptation to certain factors common to the abyssal environment and certain kinds of shallow- water habitats (“ epharmonic convergence ”) or to some ortho - genetic process than to close relationship. (See especially Robson 1926a, p. 1356, and Kemp, 1917.) This work is primarily a catalogue of the specimens of Octopodinae in the Zoological Department of the British Museum. It contains in addition a survey of the group as a whole and an account of the described species, together with a sketch of their morphology and a survey of variation in the Common Octopus. Information as to the specimens in certain European museums is included. There is no attempt to deal with the physiology of these animals or to give a complete account of their anatomy and development. Certain anatomical features of systematic and bionomic importance are, however, described in detail and a summary of the habits is given. The characteristic features of the young are described; but, as it was not possible to undertake a special study of the postembryonic stages, such young specimens as have been recorded by other workers (e.g. Hoyle) are not discussed except in certain special cases. Previous records in which no specific name is used (“ Octopus sp.”) are similarly ignored. (6) Systematic Position of the Subfamily. The Octopodinae together with the Eledoninae and Bathypolypodinae form the family Octopodidae. The systematic position of this family is as follows : — Order OCTOPOD A, Leach. Suborder 1. CIRRATA, Grimpe ( Cirroteuthis , Opisthoteuthis , etc.). „ 2. PALAEOCTOPODA, Naef. „ 3. INCIRRATA, Grimpe. * The question as to which of the special characters of the Cirrata are archaic and v hich are the result of specialization w ill be considered in the next volume. INTRODUCTION 3 Family 1. Eledonellidae, Sasaki. ,, 2. Amphitretidae, Hoyle. „ 3. Octopodidae, Orbigny. Subfamily 1. Eledoninae, Grimpe. Ink sac present : eggs large : suckers uniserial. ,, 2. Octopodinae, Grimpe. Ink sac present : „ eggs usually small : suckers biserial. „ 3. Bathypolypodinae, Kobson. Ink sac ab¬ sent : eggs ( ?) : suckers biserial. Family 4. Argonautidae, Naef. On p. 40 is given a table of the genera at present recognized in the Octopodinae. (c) Classification. Our knowledge of the broad outline of Octopod classification is due to Naef and Grimpe and is eminently satisfactory. It is far otherwise, however, with the classification of the subordinate group treated in this volume. During the last fifty years no comprehensive study of this subfamily has been attempted, although it contains some of the most remarkable and characteristic of littoral animals. Students have con¬ fined themselves to the description of species and, as a result, the Octo¬ podinae have become an unwieldy mass of species with few subdivisions. Through this mass the systematist gropes his way in hopeless confusion, and may be easily excused for shirking the task of reducing it to order ; for it presents special difficulties of more than one kind. The species of Decapod Cephalopods are not as a rule difficult to distinguish. The Octopodinae, on the other hand, confront the systematist with diffi¬ culties not unlike those encountered, for example, by the student of Corals and Hy droids. These difficulties are of four kinds. Three of them are largely adven¬ titious and the fourth seems to be rooted in the constitution of the group. I. The characteristic sculpture of the skin seems to be easily effaced, altered by preservation and subject to marked post-mortem changes. The cutaneous warts and tubercles, which are a marked feature in many forms, evidently contract after death and still more under the influence of preservatives. The strong muscular system devoid of any rigid internal support is the seat of distortion at and after death, which often obscures the natural shape, especially of the visceral sac and the head. The liability of the delicate extremities of the arms to breakage and consumption by the animal itself (and possibly by its partners) often renders their relative length difficult to ascertain. Colour is changeable even in life and liable to rapid alteration after death. II. These animals, owing to their activity, solitary habits and secre¬ tive mode of life are rarely caught in large numbers. The amount of material available for the study of a single species is generally limited, so that the amount of variation of a given species is often entirely unknown. The difficulties enumerated in I.— II. might of course be largely obviated, if freshly-caught and specially-preserved specimens were 4 A MONOGRAPH OF THE CEPHALOPODA used. A large part of the work of systematic revision is concerned, however, with the indifferently-preserved type-specimens and other historical material. A large supply of well-preserved specimens and observations on the living animal would materially assist us in this work; but the accumulation of such material is very difficult and can only be accomplished very slowly. III. These animals require very exact and exhaustive description, more particularly of the parts which are not affected by lesions and post-mortem changes. The earlier students, however, confined them¬ selves to the description of more or less valueless characters; measure¬ ments were often given without a precise definition of the points between which they were taken; no attempt was made to observe a uniform system of diagnosis and useful systematic characters, such as the pro¬ portion of the web, size of the suckers, etc., were often summarily expressed in general terms. It is desirable therefore to exercise very great caution in ascribing to a species the characters of all the specimens referred to it by previous workers. IV. A more fundamental difficulty is indicated by Appellof (1898, p. 570), who states that “ in dieser Gattung eine Artbildung noch vor sich geht, aus welcher bis jetzt nur wenige verhaltnissmassig charak- teristische und konstante Arten hervorgegangen sind.” Wiilker (1910, p. 8) expresses a similar opinion after a study of the Japanese forms. These authors seem to have been led to this view by the fact that in this group there is a very great amount of individual variation, so that each individual constitutes a separate diagnostic problem. As a rule the limits between species are extremely indistinct, and intermediate forms are very common. One does not encounter even in those forms (e.g. Octopus rugosus , vulgaris, macropus) that are numerically well-represented in museum collections the relatively homogeneous groups which are some¬ times found, e.g. in Gastropods. In the Octopods this difficulty is no doubt augmented by the numerical poverty of the material. A few individuals collected at points remote from one another cannot be expected to be completely identical. For this reason we should not exaggerate the genetic diversity of Octopod populations. But among the populations found in more circumscribed areas species are often very ill-defined, and even the individuals taken in a single haul may be conspicuously diverse. Some allowance must be made for the fact that preservation and the circumstances of death alter the colour, sculpture and shape of these animals. Very probably in life the specific characters are more easily distinguished, and colour, shape and sculpture are more constant. The difficulty, however, remains even after those characters which are modified by accident are disregarded. The cause of this diversity is hard to discover. One explanation may be that the group is, as Appellof suggested, in a very active evolutionary state with numerous new muta¬ tions arising and spreading through the general population. Alterna¬ tively there may be little isolation and such new mutations as arise may be rapidly and effectively diffused. These animals seem to deposit their eggs inshore, where they are fixed down to rocks, leaves, debris, etc., but they probably pass a considerable time offshore in a pelagic condition when young. The special powers of adhesion may facilitate their dispersal by coastal traffic and floating debris. INTRODUCTION 5 It is clear that intensive studies on the variation of single species are still much needed, and that such investigations should be carried out on the living animal. (d) Histokical. The special study of the Octopoda may be said to date from Aristotle, who devoted much attention to these animals and described with a wealth of detail and great accuracy the habits and appearance of the Mediterranean forms. In the long interval between Aristotle’s work and the first modern comprehensive study of the group there are a certain number of good descriptions of single forms such as that of Jonston (1657) and Kolreuter (1761). The writings of delle Chiaje, Lamarck’s 44 Memoire sur les genres de la Seche, du Calmar et du Poulpe ” (1799) and Cuvier’s 44 Memoire sur les Cephalopodes” (18176) are notable contributions on a limited scale. H. de Blainville (1826) made the first attempt to treat the group comprehensively. But Orbigny’s masterpiece, the “Histoire Naturelle des Cephalopodes acetabuliferes,” of which the first section of the text was published in 1835, is actually the first critical and exhaustive systematic treatment of the group. In it Orbigny attempted to formulate subdivisions of Octopus founded on the length of the arms. In his 44 Catalogue of the [Cephalopoda Antepedia] in the Collection of the British Museum ” J. E. Gray (1849) made another attempt to subdivide the group by using the size and arrangement of the suckers and the character of the epidermis. He recognized a family (Octopidae) containing Octopus , Cistopus, Pinnotopus [sic], Eledone and Cirroteuthis. Try on (1879) recognized seven genera and utilized a combina¬ tion of the methods of Gray and Orbigny for subdividing the genus Octopus. Hoyle, Appellof, Brock, Verrill and Joubin made notable contributions to the study of the group, but their work was mainly of a detailed kind and they made no attempt to introduce order into the growing mass of species. Appellof ’s discovery of the absence of the ink sac in a deep- sea form * ( Octopus piscatorum) (1892, p. 5) was of considerable value and formed the basis of a subdivision of the group made by Grimpe (1921). In the same year the latter produced a more fundamental scheme involving the recognition of ten genera. In 1928 I proposed the elevation of the two abyssal genera Bathypolypus and Benthoctopus with the new genus Grimpella to the status of a subfamily. Since the beginning of the present century valuable faunistic work has been pub¬ lished by Joubin, Massy, Wiilker, Berry and Sasaki. Naef’s monograph on the Cephalopoda of the Gulf of Naples (1921-8) contains much valuable matter on the structure and morphology of the group. * It had previously been shown to he absent in 0. arcticus, but this was regarded as an exceptional case. II. STRUCTURE. The external and internal structure of these animals is fairly well known. There is, however, no exhaustive monograph on any member of the subfamily, and for a detailed modern account of an Octopod recourse must be had to Isgrove’s monograph on Eledone. Cuvier and Orbigny both published rather summary descriptions of Octopus, and there have been numerous papers on individual organ-systems. Meyer’s monograph (1913) on Sepia and Octopus is extremely useful; but it does not treat Octopus with the fullness of detail to be attained in a work devoted exclusively to one genus. In recent years Grimpe (1913, circu¬ latory system), Marchand (1907, 1913 male genitalia, spermatophores), Robson (1925, radula) and others have described individual systems in some detail, and Naef (1923) has discussed the general morphology of the group. The following description is not an exhaustive account of the anatomy; it contains a special description of those parts which are important in classification together with fresh information on several structures and organs. In the second part of this work an account of the morphology of the family Octopididae will be given, in which the general architecture of the body and the modification of its parts will be discussed. The following account of the anatomy of the Octopodinae contains only such matter as is of importance in the classification of this subfamily. Attention is drawn to the remarks on terminology. External Anatomy. The body of an Octopus is formed of two main parts, the body proper containing the viscera, and the head (“ capito- pedal mass”) the edges of which are drawn out into the characteristic eight arms. In this work the word “ body ” is used for the visceral mass and the mantle which invests it, and “ head ” for the head-foot. In the past no standard orientation of the Cephalopod body has been adopted; authors have variously adopted an orientation based on morphological principles or one dependent on the posture assumed by the animal when swimming (cf. Hoyle, 1886, p. 53). Lankester (1884, p. 664) gave a clear explanation of the orientation of the Cephalopod body according to morphological principles. The plantar surface of the head-foot was defined as ventral and the pallial cavity as posterior. It would be preferable to retain this scheme of orientation and to use the terms “ ventral ” and “ dorsal,” “ anterior ” and “ posterior ” in describ¬ ing the Octopod body exactly as they are applied to an Amphineuran or Gastropod. Nevertheless the orientation based on the posture adopted when swimming horizontally, i.e. having the mantle cavity below, has become so general that I am disinclined to suggest that the morpho¬ logical orientation should be generally adopted. I have accordingly treated the morphologically posterior surface as “ ventral ” and the morphologically anterior surface as “ dorsal.” The head is intimately fused with the mantle in the anterior dorsal (nuchal) region, and the ventral flap of the mantle is joined to the visceral 6 STRUCTURE 7 mass by the median pallial adductor muscle. The visceral mass, however, is not connected with the dorsal portion of the mantle behind the nuchal fusion. The connection between visceral mass and mantle is limited to (a) the posterior region, (b) the nuchal area and (c) the median ventral line, and is in strong contrast to the architecture of the Decapod body. In Macroctopus (Robson, 1928c?) the ventral union of the mantle and visceral sac attains its maximum. It should be pointed out that Sepiola (nuchal union of head and mantle), Idiosepius, Lepidoteuthis and the Sepiolidae (median pallial adductor) among the Decapoda resemble the Octopoda in isolated features of the attachment of the mantle to the viscera. The significance of these various modifications will be discussed in Yol. II. The body of the Octopodinae is usually saccular and broadly oval, but it exhibits on the one hand a tendency to become globular (e.g. 0. tonganus) and on the other to attain a narrow squid-like form ( Octopus teuthoides, fusiformis, Macr octopus, Macrochlaena) . The sexes tend to differ in the shape of the body (cf. p. 15). The circumference of the body sometimes exhibits a continuous tegumentary ridge. The genus Pinnoctopus, according to the only available account, has this ridge developed as lateral fins like those of a Sepia. The status of this form is, however, questionable (see p. 184), and, as the ridge is found sporadically in species usually devoid of it ( Octopus rugosus, Scaeurgus unicirrus, etc.), I am inclined to regard it as devoid of direct morphological interest.* The head is usually well defined by praeocular and postocular con¬ strictions ; but it is sometimes merged very closely with the body. The integument of the head is continued forwards as a membranous expan¬ sion, the web, which is stretched between the arms. In the Octopodinae this structure, which in the Cirrata is often extensive and confers on the animal a medusiform appearance, usually attains a depth of about 25% of the longest arm; but in Hapalochlaena, Pteroctopus, etc., it may be as much as 40-50% of the arms. In others, again, it is very shallow (Octopus niveus, defilippi). I have already j)ointed out (1928a, p. 642) that this organ, though variable, is of considerable systematic value. (a) The depth, i.e. the length of the interbrachial sectors measured from the mouth to the circumference of the web (fig. 1) is rather constant within the various species, (h) The form of the web is very varied. To appreciate this, reference should be made to fig. 1, where the four most typical forms of web are illustrated. The various interbrachial sectors are collectively arranged according to one of the following plans : — A. All sectors equal (radial). B. Sectors bilaterally symmetrical (bilateral). (i) Depth of the sectors diminishing dorso-ventrally. (ii) Depth diminishing ventro-dorsally. (iii) Lateral sectors larger than sagittal. I have suggested that a formula analogous to the arm-formula may be used for denoting the various types of web. Each sector of one side is lettered as in fig. 1, and the form of web is denoted by a particular series of the letters placed in the order of size of the sectors which they represent. Thus the web shown in fig. 1 (A) would be represented as * See, however, Naef (1923, p. 675). 8 A MONOGRAPH OF THE CEPHALOPODA “A=B=C = D = E”; that shown in fig. 1 (B) would be A.B.C.D.E. ; and fig. 1 (D) would represent the formula C.D.B.E.A. This type of formula is largely used in this work, (c) The web also differs a good deal in the extent to which it is prolonged up the sides of the arms as lateral or brachial membranes. In the genus Cistopus and, as far as I know, in no other genus, the oral surface of the interbrachial sectors of the web is perforated by A B C D Fig. 1. — Types of Web in the Octopodinae. (Diagrammatic.) a pore, one such pore occurring in each sector. The pores lead into oblong subcutaneous pouches or crypts, which are disposed with their long axis parallel to the arms and with the apertures at their inner extremity. The function of these pouches, which are of simple construction, is unknown * ; an examination by means of sections would be of interest. The eight arms are usually stout at the base and taper towards the extremities. In certain forms (e.g. Octopus filosus) they are very attenuated and filiform. This condition may, however, be due to * They are not homologous with the “buccal pits” of Loligo ; but, though they are present in both sexes, they may possibly be used in the female for the reception of the spermatophores . STRUCTURE 9 accident. The arms are muscular and undergo characteristic post-mortem contraction, which in some species is of diagnostic value. They are paired and bilaterally disposed on each side of the (morphologically) antero-posterior axis. It is customary to number each pair (the members of which, except those of the third pair in males, are equal), beginning from those on each side of the median dorsal line, and each arm can be designated by the number of its pair and the letter L or R for “ left ” or “ right.” Thus 2L means the second arm on the left. The pairs are not always equal in size, and it has long been customary to use the relative size of the pairs for systematic purposes. The order of size is expressed by a formula in which the numbers of the pairs are placed in their order of length. Thus 1234 means that the first pair is the longest, the second pair next in order and so on. This feature is of dubious systematic value, as the arms are often damaged and subject to regeneration. It is only used in this work with very considerable reserve. The total length of the arms varies considerably from over 90% of the total length of the animal to under 50%. The arms are relatively shorter and less differentiated in young animals. The modification of one or more of the arms to form a copulatory organ is discussed on p. 16. The ventral surface of the arms is flat and furnished with two rows of suckers. The latter are sessile, i.e. not placed on pedicels as in the Decapoda, and are devoid of a chitinous armature. They are usually arranged in pairs, except for the first three or four suckers which are commonly uniserial, and they are counted from the mouth outwards. The first (adoral) suckers of the arms are occasionally found somewhat separated from the second suckers and forming a close peribuccal ring (cf. Robson, 1926, p. 166). The significance and taxonomic value of this arrangement is obscure. The suckers are not as a rule equal in size. The diameter of the 12th pair usually is largest and the suckers decrease in diameter distally and proximally. In a good many species the lateral arms (more rarely all the arms) of the male bear specially and abruptly enlarged suckers at this point. These enlarged suckers are sometimes found in old females. It is usually assumed that these enlarged suckers are of service in copulation and are thus comparable to the “ apparatus copula tor ” (fixator) of certain Decapoda (Naef, 1923, pp. 580, 604, etc.; Grimpe, 1925, p. 79). But it is by no means certain that the suckers are thus used, as the male appar¬ ently does not grasp the female in such a way as to make the adoral suckers specially important. It is a curious fact that some Octopoda which have a very large copulatory organ have no specially enlarged suckers (Bathypolypus , etc.). The diameter is the only character of the suckers which is used for systematic purposes in this work. I suspect that the fine structure of the cups may be of considerable value, but this matter is in need of special investigation. On the posterior surface of the mantle is seen the pallial aperture, a transverse slit from which the funnel projects. This slit is usually wide and extends on each side almost as far as the eyes. In the subfamily there is a tendency for it to become narrow and confined to the middle line. In Hapalochlaena and Pteroctopus it is scarcely wider than the base of the funnel, and thus foreshadows the complete closure of the 10 A MONOGRAPH OF THE CEPHALOPODA aperture seen in certain Cirrata. I have suggested a rough method of indicating the degree of closure (1926a, p. 1334) by designating the widely open stage as C, the condition in which the aperture extends on each side halfway to the eyes from the median line as B and the stage at which the aperture is confined to the base of the funnel as A. The funnel is a muscular organ formed from the tissues of the head- foot. It can be divided into three regions : (a) The most anterior and exposed part is in the form of a tube with a more or less circular aperture. There is no siphuncular groove formed to receive it in the head nor any internal valve, such as occur in the Decapoda. It is free from the tissues of the head for about one-half to one-third of the total length of the whole organ; but in some instances it is entirely fused to the head, (b) The basal portion is expanded just inside the mantle aperture, and its ven¬ tral edge is reflected to form the cephalic component of the “ locking apparatus ” (infra). This flanged or reflected portion is either con¬ tinuous from side to side or resolved into right and left lateral elements by a deep median incision, (c) The posterior extremities of the funnel Fig. 2. — Outline drawings of various types of funnel-organ. A, W-type ; B, 'VV-type ; C, Octopus tenuipulvinus ; D, Octopus macropus , var. minor. (C and D after Sasaki). are continued into the siphonal depressor muscles which form a pair of conspicuous ridges on each side of the anus. On each side of the funnel in some species are found conspicuous pouches. Inside the cavity of the funnel on its dorsal wall is the characteristic “ funnel organ ” (“ Muller’s organ ”), a superficial patch of glandular tissue which secretes a lubricant, no doubt for promoting the expulsion of debris of various kinds which might otherwise obstruct the bore of the funnel. This organ is composed of four short oblong sections arranged as a W, A or V V-shaped patch. It is sometimes absent ( e.g . in Macroctopus) ; but I think this is due to temporary physiological causes rather than to permanent loss. The discovery by Sasaki (1920, p. 181) that a varietal form of Octopus macropus has a radically different form of funnel-organ (cf. fig. 2D) from those hitherto described in the subfamily makes the systematic value of this organ uncertain. For the time being I am inclined to suspend judgment on this matter pending a confirmation of Sasaki’s identification. The “ locking ” apparatus. The ridges already noticed at the base of the funnel fit into shallow grooves on the inner edge of the mantle, and no doubt, when they are so engaged, the pallial cavity is sealed, as it is by the stud-and-socket articulation found in the Decapoda, and the contained water is concentrated in the locomotor jet expelled from STRUCTURE 11 the funnel. In most forms the flanged ridges are weak, the receptor grooves shallow and the apparatus seems much weaker than the firm cartilaginous articulation of the Decapoda. On the other hand, it is continuous from side to side in many forms and affords a more extensive articulatory surface. The external surface of the body is either smooth or covered by epidermal projections usually known as “ sculpture.” These are in the form of granules, warts or ridges. Around the eyes and in certain other areas these are much larger and sometimes branched and are known as i $ V* 2 2 0. herdmani. 1. After Hoyle (description). 2. After Winckworth (description). 1. After Verrill. 2. After Berry. 0. ocellatus. U'-? y- ’.hr A frf' O. areolatus (U.M., Leipzig). 0. membranaceus (after Orbigny, text). Fig. 3. — Ocelli of various species of Octopus. cirrhi. In one case already investigated (Octopus arborescens) these cirrhi have a very peculiar structure and are highly contractile. The so-called “Kolliker’s bristles” of the larva are discussed on p. 23. Special colour-patterns are uncommon. The colour, when present, is usually diffused in irregular maculae. Characteristic patterns, how¬ ever, are occasionally seen ( e.g . rings in Hapalochlaena lunulata , Octopus horridus, and stripes in 0. chierchiae, ornatus). A special feature is the occurrence of ocelli or circular patches of colour (often consisting of concentric rings) placed between the eyes and the edge of the web. Skeleton. Skeletal structures in the Octopodinae are represented by the cephalic cartilages and “ dorsal stylets ” alone. The latter are, as Appellof (1898a, p. 1) showed, homologous with the shell-rudiment of 12 A MONOGRAPH OF THE CEPHALOPODA Decapoda. They are situated one on each side of the middle line on the dorsal surface of the muscular mass immediately underlying the cuticle. I suspect that they may be of considerable systematic value ; but for reasons given elsewhere (p. 39) I have not examined them in a large number of cases. In Macrochlaena they are absent, and in Joubinia fontaniana they are shorter and smaller than in other Octopodines (Dali (1909, p. 181) says “no internal shell or endostyle” (sic)). Alimentary System. The horny jaws each consists of a beak (or cutting portion) and insertion plate ('palatine lamella in the upper jaw, gnlar lamella in the lower jaw), on which the beak is fixed. The beak of the lower jaw is produced laterally as wing-like rostral lamellae and may be irregularly toothed. The form of the jaws differs very little in the genera under review. In some forms the lower jaw is weak, the rostral lamellae being rather rudimentary and the beak poorly developed. Some variation is seen in the size of the beak relative to that of the insertion plate ; but I have not explored this matter very fully. The radula is essentially of carnivorous type. It is distinguished at once from the Decapod radula by (a) the frequent occurrence of two or more cusps on each side of the mesocone of the rhachidian tooth, and (6) the greater amount of differentiation in the individual teeth. Each row of teeth always consists of a median tooth (rhachidian) with three teeth and a marginal plate on each side. The identity of these teeth with the “ laterals ” and “ marginals ” of other Molluscs is obscure. I name the first three side teeth “ laterals 99 and the marginal plates “ marginals.” The radula of the Octopodinae is unique among the Mollusca on account of certain growth-phenomena seen in the form of the median tooth. These phenomena, which were first noticed by Dali and briefly commented upon by Hoyle, were described fully by me ( 1 925) . The rhachi¬ dian teeth of any radula, when examined serially from the latest formed back to the oldest and most worn, are arranged in metamerically repeated series. In a simple case (e.g. 0. gardineri) we find, if we compare the first formed tooth with those which succeed it, that the cusps of the second are not arranged like those of the first, while those of the third are again arranged differently. The fourth tooth, however, is like the first, the fifth like the second and the sixth like the third, the seventh again is like the first and so on, the same type of tooth recurring at every third row.* The series of rhachidian teeth therefore consists of a number of identical series metamerically repeated. The change of form in each series always involves the same phenomenon, viz. a change in the size and position of the small cusps (ectocones) on each side of the main cusp (mesocone). Thus an ectocone which is found close to the main cusp in the first tooth of a series will be found to be situated more externally in the next tooth, and in the third will have reached the margin of the tooth. In the fourth tooth this cusp disappears from the margin and appears in the position which it occupied in the first tooth and the centrifugal migration is repeated. This change of position on the part of the cusps is of course the reflec- * Inasmuch as the oldest (first formed) teeth are always very worn the seriation only becomes perceptible at about a third of the way along the radula from the worn to the “ nascent ” end. STRUCTURE 13 tion of a rhythm in development in the radular coecum. The secretory process forming the cusps gradually shifts laterally in successive growth- periods and then resumes its activity in a more axial position. I have found a hint of this phenomenon in certain Gastropod radulae (e.g. Peristerna). The process outlined above is complicated in many species by the fact Fig. 4. — Diagrams of (a) simple, and (6) asymmetrical seriation in the rhachidian tooth of Octopus. that the seriation is asymmetrical, i.e. the cusps on each side of the mesocone do not occupy similar positions. There is the same centrifugal displacement of the cusps, but it begins later on one side than on the other, and is completed (i.e. the ectocone attains a marginal position) later. The process of seriation and asymmetry is illustrated by the diagrams (Fig. 4). By the aid of a simple formula it is possible to indicate the type of seriation. Simple symmetrical seriation is indicated by the letter A, asymmetrical by B, and the numbers attached to the letter indicate how 14 A MONOGRAPH OF THE CEPHALOPODA many teeth are occupied by a series. Thus A5 means that the seriation is simple and is completed in five successive teeth. At present I do not attach very great weight to the seriation in the diagnosis of species. In a general way species distinguished by other characters tend to have different types of seriation, but the data on p. 29 show that the individual variation in the type of seriation is considerable, and I think that the form of the radula undergoes changes with increasing age, so that the rhachidian tooth of young forms may have a different type of seriation from that found in old ones. The whole question requires very careful and special study. The lateral teeth are of considerable diagnostic value. The first is a small imicuspidate tooth. The second is large; it has a long curved base, a large mesocone and often an entocone. When the latter is absent the mesocone may be placed at the inner margin, but more often it is nearer the centre of the base and a narrow inner “ heel ” is found. The third laterals have a long, usually slender and sword-like blade arising from a square or oblong base. The marginals are usually plain oblong plates. No very marked evolutionary tendencies are to be seen in the radula. The most interesting is the presence of simple (unicuspidate) rhachidians in 0. defilippi and Enteroctopus sp. and signs of degeneration in the first lateral and marginals of Joubinia. No Octopodine radula attains the degree of reduction seen in Eledone rotunda (Robson, 1926a, p. 1346). The Ink Sac. In the Octopodinae, as in other Octopoda, the ink sac is involved in the liver, and in this respect the Octopoda may be regarded as more specialized than the Decapoda. The sac itself or reservoir (Girod) lies in a depression on the ventral face of the liver, and at least the initial part of the duct is free. In Joubinia and Hapalochlaena the duct is contained in a groove on the surface of the liver for at least half its length (fig. 5) and about three-quarters of its length is similarly placed in Enteroctopus and 0. defilippi. In Hapalochlaena the reservoir is degenerate, and we see the initial stage of the process which ended in the complete disappearance of the organ in the Bathypolypodinae. There seems to be a good deal of variation in the degree to which the sac is covered in by the surrounding liver. In 0. vulgaris it is completely invested by the capsule of the latter; in others it occupies a more superficial position. In Cistopus it is very deeply imbedded and is scarcely visible from the exterior.* The Gills. These organs consist of a number of filaments suspended from opposite sides of a central axis in such a manner as to form two parallel and closely opposed series. Each of these series is known as a demibranch. The filaments are attached to the wall of the mantle- cavity at their tips (afferent border of the gill), and the whole gill forms a compact, roughly rectangular, mass. The filaments are very much folded. In the Octopoda generally the most striking feature in the evolution of the gills is the atrophy of these organs in relation to the abyssal mode of life (cf. Robson, l.c.). This atrophy is seen in the reduc¬ tion of (i) the size and (ii) the number of filaments. The latter pheno¬ menon alone is considered here. The average number of filaments in * In the single specimen of Paroctopus conispadiceus (p. 205) which I have seen the sac is entirely free of the liver. STRUCTURE 15 each demibranch in the Octopodinae is eleven, but there are species in which the number of filaments is considerably less, tending towards the condition found in the Bathypolypodinae and Cirrata. Thus it is as low as five or six in 0. horridus. The inner demibranch, i.e. that facing the dorsal wall of the mantle-cavity, is somewhat less developed than the outer, even in species with a large number of filaments. In Macrochlaena and a few others its filaments are markedly reduced in length, though not Fig. 5. — Diagrams illustrating the relation of ink sac ( i ) to liver (£) in the Octopodinae. in number. I believe that neither sex nor age plays any important part in determining the number of filaments in the adult (Robson, 1926a, p. 1338, and this work, p. 28), though very young forms have fewer filaments than adults. Reproductive Organs. The males and females are distinguished by certain differences of bodily proportion. I am not sure that the nature of these differences is yet understood. Naef says that the body of the male is narrower in the Octopoda generally (1923, p. 667, though cf. 0. macropus ( l.c . fig., p. 703)). Actually I have found (p. 25) that the male in 0. vulgaris is relatively broader than the female (as in Naef’s 16 A MONOGRAPH OF THE CEPHALOPODA figures, p. 703), tliougli I agree that it is of a less full and rounded shape. Very probably the proportions vary with the season. The matter is in need of exact investigation. It is generally held that in the Cephalopoda the females tend to be more numerous than the males. Pelseneer (1926, p. 42) states that the proportion in 0. vulgaris is $ 2 : 1. In 0. bimaculatus and 0. hong- kongensis Berry (1912a, pp. 280-3) found a slight excess of females. Isgrove, however (1909, p. 472), has already suggested that the numerical disparity may be more apparent than real, as the female breeds in littoral stations and is thus more often taken than the male. Figures in support of this suggestion have been published by me (1926a, p. 1325). Sometimes males and females are obtained in equal numbers and the males may even preponderate. The structure of the male and female reproductive system has been fully explored by Brock (1879, 1882) and Marchand (1907). Male System. The testis is apical (posterior) in position and may be sufficiently large to displace many of the adjacent organs. The vas deferens is long, slender and very much coiled. It ultimately passes into the long spermatic gland in which Marchand (1907) recognizes three main areas. This organ opens in common with Needham’s organ and the “ accessory gland ” into a common duct. Needham’s organ, in which the spermatozoa are stored, is usually long, cylindrical and pointed at its apex. The accessory gland is club-shaped. At the point where it enters the common duct a short coecum is sometimes found, the “ appen¬ dix ” of Marchand. The common duct terminates in the penis, which is partly external, partly internal. It is nearly always furnished with an appendage (“ diverticulum ” of Marchand). This diverticulum is usually globular or reniform, but may be long and slender (as in Joubinia and Enter octopus). A second diverticulum also is found in Joubinia. Although the presence of spermatophores in the penis may alter its shape, they do not seem to modify the diverticulum. For a long time I was under the impression that the long diverticulum of Joubinia was adventitious and merely due to its excessive enlarge¬ ment by the spermatophores. I have to conclude, however, that it is a fixed and permanent feature for the following reasons : — 1. The diverti¬ culum of other forms often contains spermatophores and is not enlarged. 2. The enlargement occurs regularly in all the males of Joubinia and Enteroctopus that I have seen and in no other forms. (In O. aegina an analogous enlargement is found, but in a different part of the system.) 3. I have found the diverticulum enlarged without any spermatophores being present (see also Robson, 1929a). The spermatophores have been well described in 0. vulgaris by Marchand (1913), and recently Sasaki has attempted to use those of other species for systematic purposes. I have myself noted considerable difference in the structural details, and some of these may be correlated with the form of the hectocotylus. The Hectocotylized Arms. The third arm of the right side is modified for copulation in all Octopodinae except Scaeurgus and Pteroctopus, in which the third arm of the left side is thus modified. The modi¬ fication expresses itself in three ways : — (1) In most, but not all, species, the arm is reduced in length. (2) The velar membrane of one side of STRUCTURE 17 the arm is converted into an open seminal channel by thickening and infolding of its rim. This channel terminates at the tip of the arm in (3) the hectocotylus proper, or end-organ (“ Loffel ” of German authors). Typically this has the following structure. A portion of the distal end of the arm is devoid of suckers and somewhat flattened dorso-ventrally. The seminal groove terminates in a conical papilla (the calamus) which is usually adjacent to the last sucker. Beyond this the surface (ligula) usually is longitudinally grooved. This groove may be very feebly developed or deep, well marked and provided with transverse ridges (laminae copulatoriae). In certain forms the apparatus is feebly developed and the essential parts ill defined. I do not know if there is any seasonal change in its form, but I suspect that this may occur. Winckworth (1928, p. 49) also makes this suggestion. Ignorance on this point makes it difficult to assess the importance of the very undifferentiated organ in Macro - chlaena , etc. On the other hand, there is a well-marked tendency in the group for the end-organ to become large and to occupy a progres¬ sively larger part of the arm. This process gives rise on the one hand to the long and pointed type of organ (Paroctopus) and on the other to the coarse, heavy Bathypolypoid type (0. australis, Scaeurgus). Further modifications of the arms of the male are seen in (a) specially enlarged suckers, and (b) conversion of the suckers of the distal part of the arms into papillae (0. chierchiae). For the use of the hectocotylized arm see p. 21. Within the same species there is very considerable variation in the form of the hectocotylus (cf. fig. 9). As examples of approximately the same size sometimes have well- or ill-defined end-organs the likelihood that this organ undergoes seasonal change is increased. Female System. The ovary is apical and is sometimes so large that it displaces the adjacent organs (cf. Robson, 1921, p. 438). The system is very simple as compared with that of the male ; but it has been very little studied. It consists, as in most of the Egopsida, of two oviducts which are divided into two parts, distal and proximal, by the development on their course of an “ oviducal ” gland. The proximal section (i.e. that which passes from the ovary to the oviducal gland) is usually the shorter. Little is known concerning the nature of the oviducal gland in the various genera. A matter of considerable morphological importance is involved herein. In the Argonautidae Brock (1882) found a recepta- culum seminis involved in the “ oviducal gland.” He stated that this structure does not occur in Octopus. Bergmann (1903), however, found evidence that part of the oviducal gland functions in 0. defilippi as a receptaculum seminis, and on sectioning the organ in an example of that species I discovered ample evidence that he was correct. Sections of the organ in 0. vulgaris also revealed the presence of spermatozoa in the gland; but the structure of the latter is not so complex as in 0. defilippi. At the present moment I have not completed this study; but it will be of great interest to discover if this receptacular function of the “ oviducal ” gland, the activity of which has hitherto been regarded as limited to secreting the egg-capsules (cf. Meyer, 1913, p. 73), is found in many other species.* * For a fuller discussion see p. 137. B. M. CEPH. C 18 A MONOGRAPH OF THE CEPHALOPODA Externally the oviducal glands are partly subdivided by a median constriction in some forms. Their size varies very considerably; but I am unable to say if this is due to differences in physiological activity. In Bathypolypus they always seem to be large. The distal part of the oviduct is in a few species expanded to form a vagina. The external female apertures are usually mere orifices in the wall of the pallia! cavity. Occasionally, however, they open on well-developed genital papillae. Structural Abnormalities , etc. Among the specimens which I have examined structural abnormalities are very uncommon, and the literature of this group contains few records of such phenomena. There are at least three instances of bifid arms in Octopus (s.s.). Parona (1900, p. 4) records a subdivided dorsal arm in Octopus vulgaris , and Hanko (1913) described a fourth arm thus subdivided. Smith (1907) gave an account of a far more remarkable abnormality in Octopus cephaea (see p. 90), in which all the arms were divided. Lonnberg (1907, p. 51) describes a specimen of Octopus patagonicus which had only seven arms ; “ of the eighth not even a trace can be seen ” (id., l.c.). I have described a very rare and striking abnormality in a specimen of Octopus rugosus from Cura£ao (Robson, 1929). In this specimen there was a double hectocotylus, the second arm on the left side bearing a fully- formed ligula and calamus. The seminal groove, however, is incomplete. Specimens showing signs of disease are singularly rare. I have opened the mantle-cavity of about 400 specimens in all and have noted only two cases of disease of the viscera and pallial complex. In one specimen there were abundant signs of kidney disease, the walls of one kidney being thickened and hardened. In another (a female) the oviduct was partly atrophied. I have discovered but one case of infection by parasitic worms.* The presence of calcareous concretions in the skin of Pter octopus tetracirrhus described by Troschel (q.v.) may have been pathological (see pp. 23, 193, 197). * The alimentary canal was not, however, opened in many specimens. III. HABITS, etc.* The Octopodinae are an exclusively marine group of animals. Like the Amphineura and Scaphopoda among Molluscs and the majority of Echinoderms and Brachiopods ( e.g .) among other groups, they seem quite incapable of living permanently in brackish water. No doubt they occasionally find their way into areas of low salinity, but I can find no record of their acclimatization to such habitats. They undoubtedly are to be found in estuaries. Winckworth (1926) records Octopus hongkongensis from Lake Tamblegam (a “ sea-loch ”) in Ceylon and Allen and Todd (1900, p. 151) state that Octopus vulgaris lives at a considerable distance up the Salcombe estuary. But when these records are closely examined it will be found that the animals live only in those estuaries of which the salinity is nearly equal to that of normal sea-water. Thus the salinity of the Salcombe estuary is not markedly different from that of the English Channel (Allen and Todd, l.c.) ; whereas in the Exe estuary, which is much fresher, Octopus vulgaris is absent (Allen and Todd, l.c., p. 295). There is a record of 0. vulgaris in the River Crouch (Essex) and of 0. lunulatus in the Swan River (Australia) (Brit. Mus.). Unfortunately no details are given in either case. Hoyle (1907a, p. 38) records 0. horri- dus from the mouth of a freshwater (?) canal in E. Africa. On the other hand, the recent study of the fauna of the Suez Canal (Robson, 1927a, p. 321) revealed Octopus horridus at Toussoum (density 1034) and Kabret (density 1033-36). In the intermediate stations of lower density no Octopods were found. The number of occurrences is low and may not be significant; but, if representative, they suggest that these animals are better able to live in water the density of which exceeds that of normal sea-water, than in a density below that of the latter. Most of the Octopodinae live in shallow water. Of 50 species of which the vertical distribution is given or may be inferred, no less than 33 are found in water of under 100 fathoms and 17 were taken in over 100 fathoms. Many species are recorded from rock-pools and reefs and can evidently tolerate exposure in the intertidal zone. In deeper water the family is principally represented by the Bathypolypodinae. Certain Octopodine forms are recorded from deep water, but it is not yet certain whether some may not be species of Benthoctopus and Bathypolypus. Our knowledge of the exact vertical range of these forms is subject to a limita¬ tion familiar to students of the fauna of deep water, viz. that, unless a specimen is taken in a closing-net, the record of the depth to which the net has been lowered cannot be accepted as that at which the animal was living. This matter has been discussed by me with special reference to the Octopoda (1926a, p. 1326). Actually we believe that the adults of this group are benthic or at least keep near to the bottom. Such animals are not likely to get into nets near the surface, when the haul is made at any considerable distance from the land ; so that the statement that an Octopus has been taken in 500 fathoms may be reasonably taken at * See Bartsch (Rept. Smithson. Instn., 1917, p. 347) fcr an interesting account of octopus-fisheries, attacks on Man, etc. 19 20 A MONOGRAPH OF THE CEPHALOPODA its face- value, even if the net used was not of the “closing55 pattern. However, we do not know how strictly benthic these animals are. Cer¬ tain well-known forms (0. vulgaris) seem to keep to the bottom and only swim off it when attacking their prey or avoiding an enemy (Hempelmann, 1926, p. 196). But we do not know if this is a universally developed characteristic, though I suspect that it is so. The young postembryonic stages of some Octopods are pelagic (Lo Bianco, 1903, p. 170; this work, pp. 23, 170) ; but no certain records of pelagic adults are known to me. Joubin (1900, p. 33) mentions that O. vulgaris was taken at an offshore station over 748-1262 metres (cf. p. 60). As to the exact habitats of Octopus and its allies we can only speak very generally. They seem on the whole to prefer rocky bottoms in adult life, but some are found in muddy and sandy places, and in all probability some of the genera (e.g. Pteroctopus) are distinguished by special adaptations to burrowing in mud. From all accounts the Octopodinae are wholly predatory and carni¬ vorous. The prey consists mainly of Crustacea, supplemented by Molluscs. Lo Bianco (1909, p. 652) states that O. vulgaris feeds regularly on Maia, Carcinus and Portunus. Haliotis (Stephenson, 1925, p. 492) and various Lamellibranchs (Dautzenberg and Durouchoux (1913, p. 7), Joubin (1907a, p. 48) and Jeffreys (1869, pp. 144-5)) are mentioned as Molluscan prey. Kollmann (1875, p. 8) described a battle between a Lobster and Octopus (sc. 0. vulgaris) ending in a victory for the latter. Hempelmann’s account ( l.c .) seems to suggest that 0. vulgaris eats fish in captivity, but Tanner (1916, p. 22) noted that a captive specimen always refused that diet. Lo Bianco (l.c.) has shown that O. vulgaris paralyses its prey by means of a toxin secreted by the second pair of salivary glands. The nature of this toxin has been investigated by several workers and it seems to be a crystallizable alkaloid. Little is known as to the enemies of Octopods. They are probably eaten by most of the larger carnivores of coastal waters. Lee (l.c., p. 52) says that the Conger is amongst the worst enemies of the Octopus on British coasts. The means of defence against enemies which these animals possess seem to be numerous. Their strength and agility is considerable, and their habit of lurking in crevices of rocks must be of advantage. In addition the ink sac provides them with the means of baffling their enemies and the highly developed cutaneous system of chromatophores is said (Lo Bianco, l.c.) to enable them to assume a protective resemblance to their background. There seems to be little doubt that Octopus vulgaris has a well- developed nesting or lair-making habit. Kollmann (l.c., p. 14) has described the transport by this species of suitable stones for the making of the lair. Whether the nest is made for the protection of the eggs or for some other purpose is uncertain. Cyclical Occurrence. Garstang (1900, p. 260) states that after being rare in the waters adjacent to Plymouth Octopus vulgaris became extra¬ ordinarily plentiful in 1900. Mr. R. Winckworth informs me that a similar “ Octopus year 55 was noted at Brighton in 1913 and probably in 1922. Garstang (l.c.) is inclined to attribute the sudden increase in 1900 to the prevalence of optimum conditions in previous years. Such fluctuations in the numbers of a particular species are, of course, familiar HABITS, ETC. 21 to most naturalists. It is uncertain whether the above-mentioned increases in population are due to local environmental causes or whether the local population was augmented by migratory shoals coming into the Channel. Autotomy and Regeneration of the Arms. It is well known that the Octopoda in common with other Cephalopods have the power of regenerat¬ ing parts of lost arms, and I have seen many traces of this process in the specimens which I have examined. The subject is reviewed, and an account of the histological phenomena found during regeneration is given by Lange (1920). Portious of the arms are often missing, and I have even found examples with all the arms reduced to mere stumps. No doubt this loss is often due to the attacks of enemies. It is, however, certain that it may also be due to autotomy (Octopus defilippi, Jatta (1896), Riggenbach (1901)), or to the animal eating its own arms ( Octopus vulgaris, Eisig (1901), Lo Bianco (1899)). Breeding Habits and Oviposition. The courtship and coitus of Octopus vulgaris have been studied by Racovitza (1894). The process consists essentially of the introduction of the extremity of the hectocotylized arm into the mantle-cavity of the female, during which operation Raco- vitza found that the animals remained a short distance apart. The male did not grasp the female during insemination, though it was seen on one occasion to hold the female with another arm for a short time previously to coitus. The exact use of the hectocotylus in manipulating the spermatophores is, however, very obscure and the significance of the various types of end-organ (see p. 17) is unknown. When we com¬ pare, e.g. the small and unspecialized hectocotylus of Octopus vulgaris with the large organ of Scaeurgus or Paroctopus hongkongensis, we are compelled to suspect some marked differences in the mode of insemination or in the spermatophores. The function of the enlarged suckers in the male (cf. Racovitza, l.c., and this work, p. 9) and of the modified suckers of Octopus chierchiae is likewise unknown. Hempel- mann (l.c., p. 200) states that the males of “ Octopus ” have been observed fighting, presumably for the possession of certain females. The eggs of very few species are known. Those of Octopus vulgaris and Paroctopus digueti are described very fully by Naef (1928, p. 70, 262 foil.) and Rochebrune (1896, p. 75). They are laid singly (Paroctopus digueti) or in clusters (Octopus vulgaris) and are encased in a capsule. One end of the latter is drawn out into a stalk by which it is attached. The following table gives the available records of the size of the eggs. Species. Authority. Size (in mm.). Octopus vulgaris Naef (l.c.) 1 X 1-8-2 ,, salutii > > 9 9 ,, macropus ,, ,, , , defilippi ,, 0-9 X 1-6 ,, rugosus Robson (MS.) 1-5 X 2-4-2-S ,, ochotensis Sasaki (1920) 9 (length of ovarian eggs) Paroctopus digueti Rochebrune (l.c.); Robson (MS.) 3 0-3-5 X 9-10 ,, yendoi Sasaki (l.c.) 7 X 17 (ovarian eggs) ,, conispadiceus „ (1917, p. 367) 30 (length of ripe ovarial eggs) 22 A MONOGRAPH OF THE CEPHALOPODA Lo Bianco (l.c., p. 650) and Naef ( l.c ., p. 263) state that the eggs of Octopus vulgaris are deposited during the summer (May-August). The size given by Sasaki for the eggs of P. conispadiceus is inordinately large and may possibly be an error. The eggs are usually attached to a solid object, usually a stone or rock (Jatta, Fisher). There is, however, a well-developed tendency to place them inside shells of other Molluscs. This is noted by Perrier and Rochebrune (1894; Octopus digueti), Robson (19286, p. 646; Octopus rugosus), Hornell (1922, p. 202; ? Octopus rugosus). Lonnberg (1907, p. 49; Octopus fontanianus) and Winckworth (1926, p. 325; Octopus fontanianus ( ?)) note the occurrence of adults inside the shells of Voluta and Chank. There is a specimen of Octopus digueti in the Musee d’His- toire Naturelle in Paris, which was found with its eggs in a broken bottle. These animals seem to have a very definite habit of entering any hollow object that will accommodate them. According to Orbigny (1840, pp. 69-70) this habit is exploited by Japanese fishermen in order to capture certain species. Lo Bianco (l.c.) relates that a specimen of 0. macropus was taken inside a human skull fished off Posilippo. Aristotle was the first to observe that the Octopus broods over its eggs. This observation has been confirmed by Lee (1875, p. 58), Schmidt- lein (1879, p. 135), Lo Bianco (l.c., p. 650), Monticelli (1921, p. 138) for Mediterranean forms and by Fisher (1923, p. 148) for 0. (?) apollyon in California. The female cradles the eggs in her web and performs some kind of incubatory process by syringing the eggs with her funnel. Parasites. Practically nothing is known of the parasites of Octopods. Lameere (1916, p. 1) described Dicyemids from 0. vulgaris and several species of Aggregata are recorded from the same form by Dobell (1925, pp. 8, 20, 31). Behaviour, etc. The behaviour of certain species has been studied by various workers and their results are summarized by Hempelmann (1926). Apart from the simpler reflexes involved in food-getting and mating, more complex manifestations have been noted, such as those involving memory and manipulative performance. Our knowledge of the behaviour of these animals is, however, very deficient and consequently the use of many peculiar structural features is unknown. In common with the Decapoda these animals may justly claim to represent the climax of invertebrate evolution, in so far as strength, size and agility are concerned, and probably also in physiological efficiency. We do not, however, know how to rate them from the point of view of intelligence. Unlike the living Tetrabranchia and some of the Decapoda which are probably gregarious, the Octopodinae seem to be solitary,* and, if it is not reading too much into our evidence, irritable and ferocious. From their very obvious preoccupation with parental responsibility seen in nesting and brooding and their wary aggressiveness and furtiveness we receive the impression of a distinctive and peculiar disposition. * The gregarious tendency noted by Garstang (l.c. p. 266) seems to have been exceptional. IV. DEVELOPMENT. f The larval and postlarval stages of the Octopoda have been studied by Naef (1923, pp. 668, 686; 1928, p. 255 and foil.). His observations, though they relate to the Mediterranean forms, are no doubt relevant in general to the majority of Octopodine species. The particular life- histories of the majority of described species are practically unknown, however, and the determination of the various young forms which have been recorded must in consequence await a fuller exploration of the subject. We may note the following characteristics of young forms. (1) The arms are very short relatively to the total length and are usually subequal in length. Marked differentiation in length is an adult feature. Thus in the planktonic stage the young 0. macropus has arms equal in length (Naef, 1921, pi. 9, f. 1). (2) The skin is usually covered with minute bristles. Naef (1923, p. 687) discusses the adaptive significance of these structures and considers that they may have a special value during the early (pelagic) stages of life.* (3) Owing to the marked increase in size of the arms in the adult, the web which seems to grow less rapidly, is sometimes deeper in the larval stage (cf. p. 27). I think it is also more evenly developed in the young. (4) The eyes are usually larger and more prominent. (5) There appears to be a special type of larval coloration. Numbers of large, well-defined (usually light brown) chromatophores often occur on the dorsum and there are sometimes double rows of such chromatophores up the arms. Naef states ( l.c ., p. 686) that “ altere Iugendstadien ” (sc. about 4 mm. long, to judge by his text-figure 401) are already benthic and do not occur in the plankton. Berry (1912a, p. 287) comes to the interesting conclusion that it is easier to make a specific diagnosis from young specimens than from the adult. * These bristles have been specially studied by F. R. von Querner ( Zeits . f. Zellforsch. mikr. Anatomie, 4Bd. Hft. 2, p. 237; 1926). 23 V. VARIATION IN OCTOPUS VULGARIS. I give below the results of an examination of a number of characters in 21 specimens of Octopus vulgaris. Most of the measurements are expressed as percentages of the length of the mantle measured from the apex of the latter to the level of the eyes. The mean, maximum and minimum of the measurements obtained are supplied. The numerical data for individual specimens are deposited in the Zoological Department. The recognition of species in this group, as already explained, is very difficult. Octopus vulgaris itself is, as a rule, tolerably distinct from its nearer Mediterranean relatives, Octopus rugosus and 0. macropus, and many specimens can be referred without much difficulty to one or another of them. I have, however, seen individuals of intermediate appearance which cannot be readily referred to any of the three species. Each of these “ species ” exhibits a combination of characters which is not found in the others. Individuals are, however, frequently found which display recombinations of these characters. This fact must be carefully borne in mind in considering the variation of these animals. The range of variation found in any sample of a species is, of course, dependent on the way in which the sample is taken. In closely allied species having many characters in common the result is also influenced by the basis on which we select our “ array.” It should be understood that individuals which in this work have been included in “ Octopus vulgaris ,” because they differ in several characters from (e.g.) Octopus macropus, may actually resemble the extreme or even average types of macropus in one particular character. Thus the length of the arms of Octopus vulgaris is 63-86% of the total length, with a mean of 78%. The individuals of arm-length 83-86% resemble the average macropus in respect of that character rather than the average vulgaris. They are included in Octopus vulgaris, because they are vulgaris- like in the majority of their characters. Data such as those supplied in this section are of interest, as they permit us to compare the variability of the various organ-systems. Thus of structures that can be actually measured the length of the arms is more variable (has a wider range) than that of the web. The interocular index is more variable still than the length of the latter. Of structures less susceptible to exact measurement the mandibles seem to be less variable than the radula. Finally we may gain some knowledge as to the effect of growth, sexual dimorphism and liability to lesion and distortion on the variation of certain structures. The number of individuals available for this study is rather low, and consequently the biometrical significance of the data is not very great. It is hoped, however, that as a contribution to a neglected aspect of Octopod systematics they may be of some use. 24 VARIATION IN OCTOPUS VULGARIS 25 1. The Width of the Mantle. The measurement from the apex (or posterior end) of the mantle to the line joining the centres of the two eyes is taken as “ length of mantle,” and the maximum width of the mantle is expressed as a percentage of this length ; the figure obtained is the “ width-index.” The range of this index is extensive viz. 37 — 62 — 91. The size of the animal seems to have some influence on the shape of the mantle. A series of nine large * specimens (average mantle-length 119 mm.) had an average index of 56 — 64 — 91 ; a series of smaller animals (61 mm.) had an average range of 37 — 58 — 68. The influence of sex in this character is, if anything, more marked, a series of eight males and eight females of the same average size having 68 and 57 respectively as their average width-indices. It is a little curious that the male should be broader, as one would expect that on an average the ovary would be larger and confer on the female a fuller mantle. Naef actually states that this is the case in 0. macropus ( l.c ., p. 703). It must be borne in mind, as pointed out elsewhere (p. 15), that the males may be broader, but the apical portion of the female body is usually rounder and fuller. I have, however, been repeatedly struck by the size attained by the ripe testis in male octopods. It is evident that the shape of the mantle can only be used for the discrimination of species with very great caution. 2. The Inter ocular Width. The distance between the outermost point of the eyes is expressed as a percentage of the mantle-length (“ interocular index ”). This dimension varies rather considerably, viz. 31 — 45 — 62; though its range is less extensive than that of the mantle- width. The difference between old and young specimens is not very marked (31 — 42 — 51 for old examples, 38 — 47 — 58 for young), but the head is somewhat narrower in old specimens. The difference between males and females is less marked ($,40 — 47 — 58; $, 32 — 44 — 58). The head of the male thus tends to be a little wider. 3. The Arms. (1) Order of size. The order of size of the arms is shown by the usual formula (p. 9). The formula 3. 2. 4.1. is most frequent and occurs eleven times in forty series; 3. 4. 2.1. occurs seven times; 2. 3. 4.1. four times. The remaining combinations (3. 2. 1.4., 2. 4.3.1., etc.) were found only once or twice. The formulae of thirteen series were rendered un¬ certain by lesions, etc. Age and sex do not seem to have any influence whatever on the relative size of the arms in adults. The character is of some systematic value; but on account of the liability of the arms to damage it is not always available. Perhaps the most constant feature is the small size of the first arms. (2) Length of the arms relative to the total length. The length of the longest arm is expressed as a percentage of the total length (arms + body). The range is fairly considerable, viz. 63 — 78 — 86. The difference between old and young specimens in this respect is trifling and quite negligible, as is that between males and females. In younger specimens than those * Large animals are not necessarily older than smaller ones, but it may be reasonably assumed that, on the whole, size is a fairly accurate index of age. 26 A MONOGRAPH OF THE CEPHALOPODA dealt with here the arms are relatively shorter and in the postembryonic stage (p. 23) they are shorter still. 4. The Width of the Pallial Aperture. The width of the aperture is assessed according to the system described on p. 10, as “ wide ” (C), “ half-open ” (B) and “ narrow ” (A). Type B occurs four times in seventeen specimens, B-C seven times and C four times. Neither age nor sex have any effect on the variation of this character. 5. The Funnel. The shape is so much subject to distortion that it is neglected here. The length of the free portion is rather constant. Expressed as a per¬ centage of the total length of the organ it has a range of 41 — 52 — 65 in sixteen examples. The length is uninfluenced by age and sex. 6. The Locking Ridge. No expression of the variability of this organ can be found. In all the examples I have examined it is continuous from side to side and well developed. 7. The Funnel Organ. The general form is fairly constant, viz. W. There is some variation in (a) the thickness of the limbs, and ( b ) the length of the outer limbs, which are not always as long as the inner limbs. The organ is not always found in a good state of preservation, and is sometimes entirely absent. Whether this is due to periodical physiological change or the mode of preservation, I do not know. I suggest that the latter is responsible, as the organ is usually absent in specimens which by their flaccid condition seem to have been picked up dead rather than killed and properly fixed before preservation. Of sixteen specimens the organ was absent in seven ; of the remaining nine the inner arms were longer than the outer in six, equal to them in three. Neither age nor sex has any appreciable influence on the occurrence or form of the gland. 8. The Skin. In sixteen examples the sculpture of the skin varies as follows : — (1) In ten examples the surface of the skin is raised into low, rather broad warts of irregular shape which are closely set. Single ocular cirrhi are present. The sculpture of these examples is not appreciably different from that seen in living animals. (2) In one example the sculpture is as in (1), but the cirrhi are absent or almost imperceptible. (3) In one specimen the warts are multifid and widely spaced. (4) In one specimen the cirrhi are present, but the character of the sculpture is uncertain. (5) In one specimen both the cirrhi and sculpture are very obscure. (6) In one specimen the warts are replaced by contiguous scales and the cirrhi are absent. (7) In one specimen the sculpture is like that of (1), but the cirrhi are numerous. VARIATION IN OCTOPUS VULGARIS 27 I agree with the opinion which Naef (1923, p. 697) seems to express, viz. that as a result of softening (? the tissues having been fixed some time after death) the appearance of rather wide, low warts is produced from papillae which during life were more prominent and perhaps more widely spaced. 9. The Web. The structure and method of designating the parts of the web have been already described (p. 7). (a) General form. It is plainly difficult to find a convenient method of expressing the variation of a complex structure like the web. A web-formula, in which each section is lettered and arranged in order of size, is useful, but it is better to find an expression of the variation in respect of size exhibited by each sector. In the accompanying table each sector is shown with the number of times it occupies the first, second, third, fourth and last position in point of depth. Order of Size (number of times each position is occupied). hector. 1st. 2nd. 3rd. 4th. 5th. 1-1 (A) . 0 0 2 8 10 1-2 (B) . 1 3 7 8 1 2-3 (C) . 16 3 1 0 0 3-4 (D) . 8 8 4 0 0 4-4 (E) 0 0 7 6 7 It follows that the most frequent order is 1-2 \ 4-4! 2-3, 3-4, i-n 4-4/’ or, using the formula suggested on p. 8, C, D, B1 A1 A] EJ 'BJ E/- The variation in the position which any sector may occupy is fairly large ; but 2-3 (C) and 3-4 (D) are regularly first and second in order ; 1-2 (B) tends to be larger than A- 4 (E), and the latter is usually larger than 1-1 (A). (6) The depth of the web relative to that of the longest arm. The depth of the deepest sector is expressed as a percentage of the longest arm. This percentage is very uniform and varies but little, the range being 15 — 20 — 25 for 22 specimens. The larger specimens have a slightly larger average, viz. 21, as compared with 19 for the smaller ones. There is no difference at all between males and females in this respect. (c) Disparity in level between the various sectors of the web. The difference between the deepest and shallowest sections of the web is expressed as a percentage of the mantle-length. The range is very extensive, viz. 11 — 31 — 71 in 21 individuals. Age seems to be an influential factor in this case, larger specimens tending to have a higher “ disparity index ” than smaller specimens (average 32 in old specimens as against 24 in young ones). Sex has no appreciable effect. 28 A MONOGRAPH OF THE CEPHALOPODA 10. The Suckers. (1) Maximum size. The outside diameter of the largest sucker is expressed as a percentage of the mantle-length. The range of this index is 7 — 13 — 23. The males differ markedly from the females in respect to this character, the largest suckers having a range of 12 — 16 — 23 in the male, and of 7 — 11 — 15 in the female. It will be noted that this is not a sex- limited character, as certain females tend to have markedly enlarged suckers. The relative size of the suckers is not noticeably influenced by age.* (2) Position of the largest suckers. Either a single sucker or more usually a pair of suckers is markedly enlarged (see p. 9). Of thirteen specimens the large sucker or suckers were on the second arm in seven, on the third arm in five and in one animal they were on both the second and third arms. I believe that the third arm more frequently bears the largest suckers in the male, the second in the female. Internal Structures. 11. The Gills. The filaments of the inner and outer lamellae of each gill were counted and the four figures obtained in each individual were averaged. The range in the average number of filaments of each demibranch is very limited, viz. 8 — 9 — 10. There is a slight amount of uncertainty as to which of the small terminal branchial tufts should count as a filament. Comparison of the age-groups shows that there is no difference between them as regards the number of filaments. The number of females in which the gills were well preserved was too small to justify a comparison with the males. 12. The Mandibles. The character selected for measurement is the depth of the rostrum of the upper mandible expressed as a percentage of the total length of the mandible. Fig. 5 a gives a key to this measurement, the length ab being made a percentage of ac. The average depth of the rostrum is 29% with a range of 22-37% (12 specimens), by no means a wide range, when the variability of other parts is considered. I feel, however, that the mandibles require special study, as the differences of shape are obscure and diffi¬ cult to express. The variation within the species is probably neither more nor less than that of other organs. Fio. 5 a. — Octopus vulgaris. Mandible 13. The Radula. Eleven radulae were available for study. Of the twenty-six specimens in the collection some were in too poor a condition to justify examination, and it was considered desirable to keep a certain number intact. (a) The number of teeth in each transverse row is constant. The average number of rows in the complete radula is 111, the maximum obtained being 140. In a radula having the average number of rows it is necessary to examine about the last 60 rows, the teeth formed earlier * It is usually the old females which have enlarged suckers (cf. Robson, 1929a). VARIATION IN OCTOPUS VULGARIS 29 than this being usually worn and disfigured. Except in the rhachidian tooth no essential difference in form is to be noted between teeth at the growing end and those formed earlier. (b) Rhachidian tooth (see p. 12). The seriation is asymmetrical (type B (Robson, 1925)) in six cases, symmetrical (type A) in four cases. One of the latter showed a “ lag,” the right-hand cusps being more eccentric than the left-hand ones. One specimen was anomalous in that some of the older teeth were symmetrical, while the later ones were asymmetrical. The number of teeth in a series varies very much. The six radulae of type B are as follows : — B3_6, B3_5, B2 (irregular), B3 (irregular), B2_4, ? . The radulae with symmetrical seriation are again very variable in the cusp-succession and vary from A2 to A4. The following diagrams show examples of the types of cusp-succession found : — ■ f b - a 1 a - b 1. -[ a - 1 - a ( a - 1 - a -a 1 a - a a 1 a - a — -a 1 a - a - 1 - a {b - a 1 a - b a - 1 — - - a b - a 1 a - - b It will thus be seen that the form of the rhachidian tooth is very variable both as to type of seriation and the number of teeth in a single series. (c) The first lateral tooth. The base is straight in three out of eleven cases, arched and sinuous in the remainder. Naef (1921, pi. 16) and Meyer (1912, p. 31) figure straight bases. The inner end of the tooth is usually devoid of a cusp, but in one case the upper angle of the inner side practically attains that status. The ectocone is low to moderate in nine cases, prominent and high in two cases. (d) The second lateral. There is a marked endocone in eight radulae. In two it is but vaguely indicated, and it is absent in one. The indenta¬ tion of the base line varies from moderate to deep ; but it is impossible to classify our examples satisfactorily. ( e ) The third lateral. This tooth has always a moderate curvature. Naef (1921, pi. 16, fig. 5) shows a straight shaft, but I am satisfied that this appearance can be produced accidentally. The shaft varies some¬ what in thickness, four radulae having rather solid teeth with chisel¬ shaped tips. (/) The marginal plates. These are always present. In one specimen they were more or less degenerate. There is some variation in their antero-posterior depth, five being ranked as shallow, two as very shallow and two as deep. The plates are always about as long as the second lateral. 14. The Pallial Septum. The length of the insertion of the adductor pallii medius is measured and expressed as a percentage of the mantle-length (Robson, 1928d). The range of variation is slight, 12 — 18 — 23. There is no appreciable differ- 30 A MONOGRAPH OF THE CEPHALOPODA ence between males and females or old and young specimens in tbe relative length of tbe septum. 15. Reproductive Organs. A. Male. (a) The length of the penis (including tbe diverticle). Tbe range in tbe length expressed as a percentage of tbe length of the mantle is 11 — 15 — 24. There is no noticeable alteration with age. (h) The shape of the penis. This is difficult to assess and no attempt is made here to give it an expression. Apart from accidental distortion the size and shape seem rather constant and the proportions of the diverticulum and penis proper are little variable. In two out of seven examples the terminal part was much longer than the diverticulum and in one the latter was circular, not conical. The shape of the diverticulum is not, as far as I can see, influenced by the presence of the spermatophores. (c) The hectocotylus. (1) Length. The average length of the ligula (measured from the distal end to the last sucker) is 4-4% of the third arm with a range of 3*3 — 5-5%. (2) Position of the calamus. This is always below the middle point of the ligula. (3) Glandular “ cheeks .” In five examples these are present in three, absent in two. The number available for studying (l)-(3) is very low. No age- classes could be formed for testing the difference (if any) between young and old specimens. (See also p. 69.) B. Female Organs. (a) Position of the oviducal aperture. This is on a level with the anterior end of the septum in two examples, 6 mm. behind it in one and 8 mm. behind it in a fourth. (h) The length of the distal part of the oviduct. This is 38% of the mantle-length on an average with a range of 37% to 45%. The organ is measured from the oviducal aperture to the point of entry into the oviducal gland. (c) The size of the “ oviducal gland.” The average length is 5-3% of the length of the mantle with average of 3-5% to 7-0%. In none of the above features does there seem to be evidence of difference between young and old specimens, but again the number of individuals is not enough to justify comparison. I believe that the size of the oviducal gland may be influenced by its state of activity. VI. PHYLOGENY AND CLASSIFICATION. Our knowledge of the evolution of these animals is very defective. It is entirely dependent on the structure of the living forms, as there are no fossil representatives known at present. Moreover, there are but very scanty means of interpreting the various modifications of the arms, web, hectocotylus, etc., in the light of the animals’ habits and mode of life. So little is known of the latter that the diversity of structure among the numerous genera and species appears meaningless at present. The only lines of bionomic divergence which are apparent in the subfamily are (a) the tendency seen in Macrochlaena , Pteroctopus, etc., to approach the deep-water Bathypolypodinae possibly in relation to a life spent in mud and sand, and ( b ) certain specialized modes of coitus probably indicated by the enlarged hectocotylus of Paroctopus on the one hand and that of J oubinia on the other. The preparation of a scheme of classification based on the phylogeny of these animals is rendered additionally difficult by two circumstances. 1 . Our knowledge of their structure is very defective and in many species no information is as yet available as to such important features as the funnel-organ, web, ink sac, eggs, radula, etc. As long as these structures remain undescribed, a very large number of the species known cannot be dealt with in a classificatory scheme based on these organs. 2. Even when these structures are known the recognition of genera, subgenera and groups of species is by no means easy. I have already shown that the species are difficult to distinguish owing to the very marked lack of correlation of characters. This tendency is to be seen in the larger groups. It is not easy to discover any broad lines of evolutionary divergence upon which a satisfactory classification could be founded. Marked structural divergence is indeed common ; but it is usually seen in single characters, unaccompanied by other modifications. The subfamily seems to be broken up into a large number of disconnected and often monotypic groups rather than along well-defined lines of evolutionary divergence. The matter is complicated by the fact that the phylogeny of the suborder Incirrata is very obscure. If we knew the most primitive characteristics within the suborder, it might be possible to distinguish at least the most primitive of the Octopodinae from the more specialized. But if we review the chief organs and parts it will be found difficult to come to a conclusion as to which phase of these is to be regarded as more specialized and which as more primitive. As some such conception is, however, indispensable it will be necessary to consider this in detail. I. The Arms, {a) In all probability the arms were primitively equal in length and the condition in which either the terminals are larger than the laterals or vice versa, is more specialized. 31 32 A MONOGRAPH OF THE CEPHALOPODA (6) I am uncertain if there is any significance to be attached to either excessive length or excessive shortness. I am inclined to think that lengthening of the arm beyond 80% is a specializa¬ tion. This conclusion and also that stated under (a) are largely founded on the fact that in young forms the arms are usually equal and of moderate length (p. 23). II. The Web. In the Decapoda the web is usually equally and moder¬ ately developed between the arms; and in the Octopoda a deep and well-developed web is found in the otherwise specialized Cirroteuthidae. I therefore believe that in the Octopodinae the web was primitively low and equally developed and that increase in its depth and asymmetry is a mark of specialization. III. The Funnel Organ. It is difficult to decide whether the fre¬ quently occurring W or W shape is more primitive than the rarer VV shape. The shape most commonly encountered in the Decapoda is quite different, viz. \a/, and it is possible to derive either of the Octopodine forms from it. IV. The Gills. A reduction of the number of filaments in each demi- branch is a marked feature of abyssal forms (Robson, 1926, p. 1338). It is, however, impossible to say if a number above the average (11) is a sign of specialization or whether it is a primitive feature. The Decapoda have many more filaments, so that the higher number may be an index of primitive status among the Octopodinae. V. The Hectocotylus. The spoon-like ligula is a structure confined to the Octopoda, and no arguments from the other members of the class can be employed. In all probability the large ligula with deeply-folded sides and well-developed laminae copula- toriae, such as is seen in Bathypolypus, is a specialization, and the small undifferentiated ligula of ( e.g .) Octopus vulgaris is more archaic. VI. The Radula. The Octopodinae as a whole show very little variation in the lateral teeth and it is not easy to refer such modification as may occur to a more specialized or a more archaic status. The character of the rhachidian tooth is, however, more varied, and I think it is reasonable to suppose that the simpler type of tooth without elaborate seriation and multiple cusps is more primitive. VII. The Ink Sac. In Nautilus there is no ink sac. In the Decapoda it is well developed and is not intimately associated with the liver. In those Octopoda which possess this organ it is im¬ bedded in the ventral surface of the liver and covered over by the integument of the latter. It is uncertain whether the absence of the sac in the Cirromorpha is primitive or an adaptation to abyssal conditions. The absence of the sac in the Bathypolypodinae is probably a specialization. As far as I am able to see any difference in the degree to which the ink sac is involved in the liver in the Octopodinae I think it is legitimate to infer that the more superficial position is more primitive. PHYTOGENY AND CLASSIFICATION 33 VIII. The Mantle-aperture. I believe that the widely-open mantle- aperture is a primitive feature. In Nautilus and the Decapoda the mantle is widely open, and it is only in the Cirromorpha and some of the abyssal Bathypolypodinae that it is partly or wholly closed.* As regards the other characters of morphological value mentioned on p. 39, I do not think we can regularly employ them in treating of the evolutionary tendencies of the group. Some of them (eggs, suckers, dorsal cartilages, adhesive apparatus) are but little known, others are of doubtful value (sculpture and outline of head and mantle) or of special and isolated occurrence (interbrachial pouches), and they can scarcely be used in a general scheme. If the conclusions arrived at in I.-VIII. are correct, then we would imagine our primitive Octopodine as having (1) rather short, equal arms, (2) a low equal web, (3) probably a W-shaped funnel-organ, (4) numerous gill-filaments, (5) a small, undifferentiated hectocotylus, (6) simple rhachidian teeth, (7) a superficially placed ink sac and (8) a widely-open mantle aperture. Up to the present some fifteen f genera of Octopodines including those proposed in this work have been described. Of these two ( Schiz - octopus and Amphioctopus) are not recognized here. The first was mentioned casually (though with definition) by Hoyle J (1886, p. 31). The only characters by which it is defined (a deep incision of the web between the dorsal arms and the occurrence of sculpture on the oral surface of the web are of trivial systematic value. Amphioctopus was proposed by Fischer for Octopus membranaceus, Quoy and Gaimard, the peculiar feature of which (lateral membranes of the body) is almost certainly of abnormal occurrence (see p. 7). Of the rest Tritaxeopus (p. 172) and Pinnoctopus (p. 184) are enigmatic forms, and since their description no student of the group has had an opportunity of studying them. Macrotritopus is likewise of problematical status (p. 167). The remaining genera have well-marked and individual characteristics, and the following view is taken of their relationship. The features considered to be archaic (see above) in the group are realized most frequently in Octopus (s.s.). The complete array of characters assumed to be primitive is not, as far as I know, found in any one species (0. defilippi presenting the nearest approximation to this condition). Nevertheless the subgenus, as defined on p. 57, contains the greatest number of species having these characters most often associated. Macr octopus is probably a closely-related form. Among the others there are, I think, two definite lines of specialization, though the genera placed in these are not necessarily related and may resemble each other through convergence. The one manifests an approximation to the Bathy poly pod type and is represented by Pteroctopus, Macrochlaena and Hapalochlaena. These genera exhibit some primitive traits. The web and arms are subequal. In Macrochlaena the hectocotylus is undifferentiated and * I exclude from this generalization the special sipbono-pallial fusions found in the “ Oegopsida consuta” and Amphitretidae. f I do not include Pseudoctopus (Grimpe, 1925, p. 93), as Grimpe himself noted that Naef had previously given the name Paroctopus to the group in question. I Hoyle ascribes Schizoctopus to Steenstrup ; but, as far as I (and others) can ascertain the name was never published by that author. B. M. CEPH. D 34 A MONOGRAPH OF THE CEPHALOPODA there is no appendix in the male genital tract. They all, however, show specialized features in (a) the soft gelatinous skin, (6) deep web, (c) narrow mantle aperture, (d) more or less reduced gills. As in Hapalochlaena the ink sac is manifestly reduced, it should perhaps be regarded as the most evolved. The other main line of specialization is represented by Joubinia, Enter octopus, Scaeurgus and Par octopus. Of these the first two are obviously allied and exhibit the same tendency. It consists of the development of a long penial diverticle in the male genitalia. Some marked specialization of the hectocotylus is seen in this group. In Paroctopus the hectocotylus tends to be conspicuously enlarged and the eggs are much larger than in other forms. This genus was proposed by Naef for 0. digueti, the ovipository habits of which are well known. I include 0. hongkongensis, 0. apollyon, 0. conispadiceus and 0. yendoi in this genus, though I think the evidence suggests that 0. apollyon (to which 0. hongkongensis is closely related) has small eggs. All these forms have a large hectocotylus. I believe that 0. californicus and 0. dofleini may belong to this group ; and it is possible that more of Sasaki’s species (1920) should be included here. Cistopus, which has the remarkable feature of well-defined inter- brachial pouches of unknown function, is worthy of a separate generic position. Its low web and plain tricusped rhachidian teeth ally it with the more primitive Octopod radicle. There remain for consideration Macroctopus and Macrotritopus. These forms do not, in my opinion, show the same degree of differentiation as the forms hitherto treated. The former is very closely allied to one of the subdivisions of Octopus. The number of its gill-filaments is remarkably high and the locking apparatus may be primitive. On the other hand, if all the accounts are to be believed, the arms of the female are peculiarly specialized in the same way as they are in Eledone and in the male of 0. chierchiae. Macrotritopus is of very doubtful value as a separate group. It is only distinguished from Octopus by its large third arms and is at present principally known from young specimens.* Octopus chierchiae is remarkable for the modification of the distal suckers of all the arms in the male. O.fusiformis and 0. teuthoides , which have very narrow squid-like bodies and arms exceedingly short even for the immature state of the single individuals by which they are represented, possibly deserve subgeneric recognition, though I see no gain in raising them, and perhaps one or two more slender squid-like forms, to that rank until they are better known. With regard to the occurrence of subdivisions within the large sub- genus Octopus (s.s.) the difficulty encountered in dealing with the genera makes itself felt in an acute form. Very many species are known by only a few characters, and it is difficult to assemble data on a sufficiently large body of species to make subdivisions of any value. So far as our information goes at present I believe, however, that the following groups can be recognized. 1 . A. A large radicle of forms which are more or less like 0. vulgaris. The web is bilaterally symmetrical and of moderate depth, the arms are uneven (the laterals being larger than the terminals) * There is some possibility that these may be young examples of Scaeurgus (see p. 168). PHYLOGENY AND CLASSIFICATION 35 and of moderate length (about 80%), the radula has an A3 rhachidian and the hectocotylus is from 1*5 to 5-5% of the arms. The skin is usually covered with coarse simple warts. 2. A smaller number of other groups illustrating divers tendencies of which we may recognize B. (a) Arms, 1 2 3 4; web, A B C D E ; hectocotylus, 6% or more; arms long (over 83%). (Group of 0. macropus.) (b) As ini, but arms short (average 76%). (Group of 0. leioderma.) C. Arms, ^j.21 ; web, ^ Jc B A (A very shallow). (Group of 0. egina.) D. Hectocotylus long ; arms short (under 80%) ; funnel-organ double. (Group of 0. pallida.) E. Funnel-organ with median limbs far larger than lateral. (Group of 0. tenuipulvinus.) F. Body long and narrow, arms usually short, web subequal. (Group of 0. fusiformis.) G. Arms very long (over 83%); web low and subequal; the radula with a unicuspidate rhachidian ; a receptaculum seminis known to occur. (Group of 0. defilippi.) H. Arms short, web deep (over 30%), hectocotylus Bathy polypoid. (Group of 0. australis.) I have arranged under these groups those species which have been adequately described. It will be apparent, however, that, as the descrip¬ tions are often defective in one or more material respects, many species are given a position on somewhat debatable grounds. I have tried, however, to assess the position of each species on as comprehensive a basis as possible, and have, in one or two places, been guided by general facies (as indicated in the descriptions) rather than by any particular association of characters. There is a large residuum of forms which cannot be placed in this scheme. These and a certain number of juvenile forms are placed in a separate category. Species of which no data are available or the descriptions of which are so defective that they cannot be properly distinguished are placed in a separate list on p. 214. At present the large group of N.E. and N.W. Pacific species do not lend themselves to satisfactory grouping. Those placed in Paroctopus seem to form a natural assemblage. They have some affinity with those which I have placed in Group D of Octopus (s.s.), and some of the species temporarily arranged in that group may eventually be transferred to Paroctopus when something is known of their reproduction. Paroctopus also has affinities with the leioderma-group of Octopus. On the other hand, the affinities of some of Sasaki’s species (1920) are by no means clear, and we await a more complete description of them.* 0. alatus and probably 0. tsugarensis seem to have long penial diverticula, which suggests relation¬ ship with Enter octopus . Finally the Mediterranean Octopus salutii may be ultimately placed with the Pacific forms, as it has, like them, a long narrow hectocotylus. At present, however, it cannot be accommodated in any of the recognizable groups. * For example, there is a tendency among them to exhibit the 0. macropus arm- and web-formula. This has led me to suspect affinity with macropus and to place certain forms that cannot be accommodated in Paroctopus close to the macropus -group. I admit, however, that the position of 0. tsugarensis is very ambiguous. VII. GEOGRAPHICAL DISTRIBUTION. The Octopodinae are widely distributed in all temperate and tropical seas. They do not occur in high latitudes and the limits of their dispersal toward the Arctic and Antarctic circles may be thus summarized. In the N. Atlantic they are not found north of a line drawn from the Firth of Forth to New York * (approximately along the isotherm of 10° C.). North of this line they are replaced by Bathypolypods and Eledone. In the South-east Atlantic, Octopus (s.s.) is not found south of Tristan d’Acunha (isotherm of 15° C.). On the south American mainland the limit is vaguer, as the generic position of several Patagonian forms is uncertain. Repre¬ sentatives of the genus reach about 52-53° S. in S. Patagonia (isotherm of 8° C.). The subfamily is, however, better represented by Joubinia and Enter octopus in the Magellanic region (isothermal limits 6-10° C.). Further south towards the Antarctic Circle Eledone seems to replace the Octopodinae. I am as yet a little uncertain as to the southern limits of the group, as the generic position of certain recorded forms (e.g. Octopus sp. Joubin, 1906) is not known. In the Indo-Pacific area the Northern limit seems to be the Aleutian-Kamschatkan region (about 3° C.), where forms probably referable to Par octopus are reported. Again, the generic position of many N. Pacific forms is not known. I suspect that many of the Japanese forms described by Sasaki are referable to Par oc¬ topus. In the Southern Ocean true Octopus is found off the island of St. Paul and in Tasmanian waters, 38-40° S. (isotherm 15° C.), and the nearly allied Macroctopus off the New Zealand coast and Campbell Id. (50° S., isotherm of 8° C.). Joubinia is also found off Campbell Id. The above-mentioned ambiguity as to the position of the Magellanic species of Octopus is applicable to those of the west coast of Patagonia. In short, we may say that the Octopodinae are found in water of a mean annual temperature usually not under 10° C. and that they are replaced in colder water mainly by the Bathypolypodinae in the north and the Eledoninae in the south. True Octopus is an inhabitant of warmer water than J oubinia, Enteroctopus and Paroctopus. Concerning the detailed distribution of species very little can be said until more work has been done on the variation and identity of the species themselves. The following points are to be noticed : — (1) There are many regions of the world of which the Octopodine fauna is very imperfectly known. Little information has been published concerning their occurrence on the east coast of N. and S. America, Western Central and South America, the coasts of China, and of West Africa from the Cape Verde Islands * I am indebted to Mr. C. Johnson of the Boston Society of Natural History for the information that there are no records other than those of Verrill for the New England coast. The only species found in that area is Bathypolypus arcticus. 36 GEOGRAPHICAL DISTRIBUTION 37 southwards. The result is that any attempt at dividing the various oceans into regions on the evidence of this group is more or less valueless at present. (2) Certain species are remarkably widely diffused. 0. vulgaris and macropus seem to range from the Mediterranean to Japan. 0. rugosus is found in nearly all the subtropical and tropical waters of the world (with the possible exception of the E. Pacific). Scaeurgus unicirrus is found in the Mediterranean, Indian Ocean, Japan and Hawaiian Islands. The identification of forms from regions so widely apart may be questioned ; but I am satisfied that the range of the species above mentioned is correctly stated. (3) The Magellanic region and the North Pacific seem to be foci of local differentiation, giving rise to fairly distinctive groups. The rest of the world is occupied by a population in which peculiar local faunas are difficult to distinguish and the in¬ dividual range of species is wide. VIII. MEASUREMENTS : BIBLIOGRAPHICAL METHOD : ABBREVIATIONS. (a) Measurements . The method of taking the measurements given here and in the table (pp. 42-55) is stated under the various organs in section V. In most cases the actual measurement is reduced to a percentage of the dorsal mantle-length measured from the apex to a point midway between the centres of the eyes. Usually this is expressed simply as an index number. Thus “ suckers 15 ” means 15% of the mantle-length. Certain special cases are treated differently. Thus the depth of the web is given as a percentage (“ web-length 30 ”) of the longest arm and the length of the ligula as a percentage of the total length of the hectocotylized arm. In preparing the tables on p. 42 I have been faced with the difficulty alluded to on p. 4 that in the past no standard method of measurement has been used. The measurements used in this work are those most commonly employed, but occasionally it is impossible to bring the measure¬ ments given by an author into line with these. Sometimes, therefore, I have had to make approximations. When this is done I have indicated the fact by a query (?), unless it is otherwise apparent that the figure given is of this nature. The treatment of Dr. Stillman Berry’s valuable data (1912a : 1914a) requires special mention. He sometimes gives the main body-length as “ tip of body to base of dorsal arms.” As I understand this measurement it is a little longer than my standard apex-eye length, and I have con¬ sequently obtained the latter by subtracting a small amount from Berry’s apex-base of arms length. The amount subtracted is proportionate to the size of the animal. Thus in Berry’s data for Octopus leioderma (1912a, p. 288) “ apex-base of arms ” is given as 45 mm. in the type specimen. I have subtracted 5 mm. from this and treated the apex-eye measurement as 40. Another method of obtaining the apex-eye length is to subtract the arm-length (maximum) + x mm. from the total length (if given), x being a small amount proportionate to the size of the animal and equivalent to the distance from the base of the arms to the eye. “ Base of the arms ” as used by sundry authors is actually vague ; the arm-length should always be most precisely defined (e.g. “ tip of arms to outer edge of lip ”). (b) Bibliographical Method. In drawing up the bibliography of each species an attempt has been made to give only the essential references. In the past it has been part of the tradition of careful monographers to include under each species every reference that has been made to it. This practice is often carried to quite 38 MEASUREMENTS, ETC. 39 unnecessary lengths even to the inclusion of MS. names and valueless quotations from popular works. In the present work, in order to avoid a cumbrous and useless bulk, I have usually excluded ( a ) all references which are verbatim quotations from other authors accompanied by copies of earlier figures, (b) extracts from purely nominal lists of species from areas covered by earlier references, (c) geographical references which are loosely or generally expressed (“ English Channel,” etc.) or relate to places immediately adjacent to previously noted localities. The bibliography of each species contains references to such works only as are of systematic or distributional importance. The former is interpreted fairly broadly. Thus the lists contain references to papers dealing with habitudinal peculiarities and internal structures as they may be of systematic value. They do not record papers on physiology, on experimental work or on miscellaneous subjects, such as psychology, regeneration, etc. (c) List of Abbreviations used for Names of Institutions. Brit. Mus. = British Museum (Natural History). R.S.M., Edinb. = Royal Scottish Museum, Edinburgh. M.H.N., Paris = Musee d’Histoire Naturelle, Paris. Z.M., Berlin. = Zoologisches Museum, Berlin. M.U., Leipzig. = Museum der Universitat, Leipzig. M.U., Jena. = ,, „ ,, , Jena. Senck. Inst. = Senckenbergisches Institut, Frankfurt a/M. R.M., Leiden. = ’s Rijks Museum, Leiden. U.S.N.M. = United States National Museum. Examination of Specimens and Characters used in Taxonomy. The number of characters which should be examined for taxonomic purposes, especially in the discrimination of genera, is very considerable. I doubt very much whether a really satisfactory phylogeny of this group will be produced until the anatomy of each group of species is completely known. Naef (1923, p. 691) recommends that the following characters should be used in conjunction with one another for the diagnosis of genera: — (1) size of the eggs; (2) outline of head and mantle; (3) relative arm length; (4) form of the web; (5) sculpture; (5) funnel- organ ; (7) presence of lateral folds or fin-rudiments ( ?) ; (8) the size of the mantle-aperture; (9) form of the radula; (10) ink sac; (11) consistency of the skin; (12) the presence of interbrachial pouches; (13) hecto- cotylus and (14) occurrence of abnormal extremities on the arms of the mature male. I have no doubt that the gills, suckers, “ dorsal stylets,” form of the “ adhesive apparatus,” salivary glands, crop, etc., would also yield important characters. In a work such as this, however, it is necessary to limit one’s examination, and, in addition, it must be pointed out that in dealing with forms like the Octopods, when few specimens are available, it is often undesirable to dissect the whole animal, as it would in most cases leave the latter in an almost unrecognizable condition. I have in all cases opened the pallial cavity and funnel, and in many instances I have examined the radula and ink sac. I have unwillingly abandoned the study of the “ dorsal stylets,” as their extraction usually damages the shape of the mantle too much. IX. SYNOPSIS OF CLASSIFICATION. The genera recognized in this work are enumerated below, together with the genera of the Bathypolypodinae. Although the latter are to be treated in Part II, I consider it desirable that at least a provisional arrange¬ ment should be published now, so that the position of the various species originally referred to Octopus and now placed in the Bathypolypodinae may be indicated. Family Octopodidae, Orbigny. Subfamily I. Eledoninae, Grimpe. „ II. Octopodinae, Grimpe. Genus 1. Octopus , Lamarck Subgenus i. Octopus, Lamarck ii. Macrotritopus, Grimpe iii. Tritaxeopus, Owen (?) iv. Macroctopus, Robson v. Enteroctopus, Rochebrune & Mabille (pp. 2. Cistopus, Gray 3. Pinnoctopus, Orbigny (?) 4. J oubinia, n. gen. 5. Scaeurgus, Troschel 6. Macrochlaena, n. gen. 7. Pteroctopus, Fischer 8. Paroctopus, Naef. 9. Hapalochlaena, n. gen. [10. Schizoctopus , Hoyle (Steenstrup MS.)] [11. Amphioctopus, Fischer] (pp- (pp- (pp- (pp- (pp- (pp- (pp- (pp- (pp- (pp- (pp- (pp- (pp- 56- 182). 57- 167). 167-172). 172- 173). 173- 175). 175-182). 182-184). 184-186). 187-191). 191-193). 193-195). 195-197). 197-207). 207-214). „ III. Bathypolypodinae, Robson. 12. Bathypolypus, Grimpe. B . arcticus (Prosch.). (0. bairdii, V’ll., 0. groen- landicus, (Dewhurst), synonymous). lentus (Verrill). obesus (Verrill). sponsalis (Fischer). faeroensis (Russel). valdiviae (Chun). grimpei, Robson. (?) abruptus, Sasaki.* 13. Benthoctopus, Grimpe. B. piscatorum (Verrill). (0. normani, Massy, synonymous). * There is no ink sac in specimens of this species sent to me on loan from the U.S. National Museum. 40 SYNOPSIS OF CLASSIFICATION 41 januarii (Hoyle). ergasticus (Fischer). (0. prof undicola, Massy, synonymous) . (?) salebrosus (Sasaki).* (?) Inokkaidensis (Berry) (= glaber, Sasaki, 1920). berryi, Robson. sasakii, Robson. (?) hyadesi, Rochebrune & Mabille. levis, Hoyle. 14. Grimpella, Robson. G. thaumastocheir, Robson. 15. (?) Atlantoctopus , Grimpe. A. lothei (Chun). pseudonymus, Grimpe (— “ 0. levis, Hoyle,” Joubin, 1906). 16. Haptochlaena, Grimpe. H. chuni, Grimpe. alberti (Joubin). A new genus is to be defined in Vol. II for “ Polypus pricei,” Massy (1916a, non Berry). f Incertae Sedis. Polypus hoylei, Berry. „ pricei, Berry ( non “ P. pricei, Berry,” Massy, 1916a, see above). * There is no ink sac in specimens of this species sent to me on loan from the U.S. National Museum. f Since preparing the discussion on this form (p. 219) I have received Massy’s Arabian Sea specimens on loan from the Indian Museum and am able to state that there is no trace of an ink sac in them. X.— TABLE OF For explanation of the measurements in columns 1-9 and 11, see “ Variation ” and p. 38. The rhachidian formula given Unless otherwise stated the figures given are index-numbers 1. Dorsal length of mantle (mm.). 2. Width index. 3. Inter¬ ocular index. 4. Arms, formula. 5. Arms, length. 6. Gill- filaments. 7. Diameter of suckers. I. Octopus. (i) Octopus (s.s.). A. Group of 0. vulgaris. 1. vulgaris Brit. Mus., etc. dor¬ sal and ventral. Over 25% — - Minute, prominent warts; no ocular or dorsal cirrhi. Dark purplish-brown above, orange and brown below. b=c=d.a.e* 22* — — Smooth. Deep violet, pale orange maculae, dark purple maculae. BDCEA CDBEA 11-18 r=a2 1L, very large. 2L, no entocone heel present. 1-1 type. Mainly smooth, a few tubercles. “ Buff densely covered with dark brown chromatophores ” (Massy). CBDAE 18 ? — 1-4-1-7 Prominent unequal warts. Dark brownish grey, sometimes clouded. DOBAE 15 - 3-7 Smooth; a few scat¬ tered cirrhi. Uniform dark purplish- brown. CDBAE 27 (see p. 81) i "■ Coarse irregular warts. Orange-ochre, blue marbling and dark transverse crescents on arms and web. C=D = BAE Sc. A=E (Berry). 23-26 p Simple granulations. Slate grey or faint purple above; warm creamy-brown later¬ ally. E>A 32 (?) — ■ — Granular ; no cirrhi. Dull violet shading into ochre below. Ocellus present. CDA=BE Sc. subequal. 18 — — Simple and multifid warts. See p. 84. — d=c.b=e.a 19 — — Originally warty. Type very much worn. ? Purple. ? C.D=E.A=B 26 — — A large number of prominent warts, cf. text, p. 86. Dull brownish-grey, paler below. Ocellus. C.B=E.D.A 33 Sparsely granulated, ocular cirrhi. - 43 44 A MONOGRAPH OF THE CEPHALOPODA 1. Dorsal length of mantle (mm.). 2. Width index. 3. Inter¬ ocular index. 4. Arms, formula. 5. Arms, length. 6. Gill- filaments. 7. Diameter of suckers. 16. microphthalmus Goodrich, 1896, $ Massy, 1916A, ? * 31-36 58-64 38-48 3.4=21 4=32=1 123=4 69-72 8 4-8~5-6 17. cephaea Brit. Mus. 85 58 35 3241 83 ? 7-8 16 18. horridus Brit. Mus. D>A>B>E 22 ? 20-16 E=A3_5 2L, entocone pre¬ sent. 1-5 Numerous close ro¬ sette-like tubercles. Centre of rosette high. ? ? C=D = B.A.E Under 25 Numerous low, rough, conical tubercles, the skin between smooth or finely papillose. No ocular cirrhi. Dark slate; inner sur¬ face of suckers light brown sometimes shading to cream. A.B.O.D.E. 11-16-25 E=A3 2L, entocone and heel absent. 4-8-9-5-14 Pine warts tending to become granular. Eeddish. — ? 17 — . — Granular. “ Milky, blotched or speckled with ochre, sprinkled with brown.” ? a=b=c.d.e 14 Smooth. Buff with rough quin- cuncial pattern of purple chromato¬ phores. BAODE 10-11 (?) - 1-1-3-8 “ Coarsely reticulate- papillose,” ovoid brachial “ bullae.” Papillae generally linear (Berry). Deep orange with lon¬ gitudinal buff stripes. a=b=c=d.e ? 16 e=a4 1L, base narrow, cusp high. 2L, no entocone, no heel. Fine sparse granules, a few larger and coloured brown. Pale ochreous brown with dark patches. " Finely granular. Dark ochre, covered very closely with light and dark chromato¬ phores. 46 A MONOGEAPH OF THE CEPHALOPODA 1. Dorsal length of mantle (mm.). 2. Width index. 3. Inter- ocular index. 4. Arms, formula. 5. Arms, length. 6. Gill- filaments. 7. Diameter of suckers. (6) Sub-group of 0. leioderma. 1. leioderma Berrv 9 ? 25-40 ?100 ? 77 1234 70-71 — 2. tsugarensis Sasaki, 1920 2>3>4 L 2>3=1>4 85 ? 8 * Calculated, MEASUREMENTS 47 8. Web formula. 9. Web, depth. 10. Badula. 11. Hectocotylus, length of ligula. 12. Sculpture. 13. Colour. Sc. A=B=C>D>E 25 + - - Smooth ; some short simple papillae. Ocular cirrhi. Very pale grey buff suffused with pur¬ plish-brown. See text. ? 25 + ' — 9 Smooth faint warts about the eyes. Reddish-brown, three stripes on head. “ Broad ” 20 _ 10 Roundish warts best Pale purplish-brown. A=BCDE 25 9 developed above eyes. DEOBA 22-30 r=a4 1L, very long, cusp low. 2L, no entocone heel. 8 Smooth ?. Very characteristic re¬ ticulation on arms and anastomosing lines on dorsum. D=ECBA 22 - 5-4-G-5 Coarse close polygonal warts. Dirty olivaceous or brownish-grey, reticu¬ late. CDEBA C=D.E = B.A 20-33 R=A1 1L, long and nar¬ row. 4-4-13 (?) Closely shagreened with fine or coarse warts. Ocellus, interocular spot. CDBEA, etc. 21-27 - 6-1-5-7 Warts coarser than in areolatus and often multifid and rosette¬ like. Ocellus, interocular spot. D=CEBA (subequal) 28 Smooth ; three coni¬ cal ocular cirrhi, four dorsal cirrhi arranged in a dia¬ mond pattern. Dark greyish-brown marbled with darker brown. An ocellus. C=D = E.B.A 25 R=A5.e 1L, cusp high, long base. 2L, no entocone. 3L, slender. Mainly smooth ; a tract of low warts in mid-dorsal line. Discoloured ; reticulate pattern on web and arms. Hoyle, “ subequal.” Berry, ? ,, E>A 24 II 9-12 Numerous close ro¬ sette-like tubercles, centre of rosettes lo^y. Pale purplish-grey. a=b=cde 22 — Densely covered with large stellate pa¬ pillae. Livid pinkish-brown. ? B=G = D>A>E AD>E B.C=A>D>E ? 21 + - 12 16 Numerous minute rough papillae ; ocu¬ lar cirrhi. Deep brownish-claret- mottled with a darker shade above. BCD>A>E 25 — 6-2 Wrinkled dorsally and with a few warts. — “ Poorly developed, equal.” — — - Finely tessellated with grooves and beset with tubercles. - 48 A MONOGRAPH OP THE CEPHALOPODA 1. 2. 3. 4. 5. 6. 7. Dorsal length of Width Inter¬ ocular index. Arms, Arms, Gill- Diameter of suckers. mantle (mm.). index. formula. length. filaments. F. Group of 0. fusiformis. 1. fusiformis Brock 45-G7 26-35 26-32 1. 2.4.3 72-79 — 8-5 Massy 2.1. 3.4 2. teuthoides Brit. Mus. 9 16 37 26 1234 57 — — 3. amboinensis Gottingen Brock 17-18 41-44 — 3. 2.4.1 76-75 — — Joubin G. Group of 0. defilippi. 1. defilippi rest. 7-13 12 R unicuspid. (?) 2L = No. 3. 3-2 Probably smooth in adult; warts round eyes. Pale, golden yellow ; sometimes darkly maculate. ? Subequal. 4-12 4-1-4 Papillae and cirrhi on web and dorsal mantle ; ocular cirrhi. Lesson, “ white.” Orbigny, “ yellow-rose below.” Massy, “ lilac.” Subequal. 6-12 Smooth (Orbigny). Wrinkles and warts (Brit. Mus.), etc. Of. Wiilker. Blackish above ; inside of web white. BDA=CE B=CDAE 33 r=a2 1L, long base, cusp low. 2L, no entocone, low mesocone. 3L, broad base. (See p. 145.) 10-8 Thickly covered with small granular warts ; * ocular cirrhi and enlarged warts on body. Ochreous mottled and maculated with brown . Arms barred . B>A Lateral sectors “ ex¬ tend much further than from above backwards.” Sc. 20 Smooth. Ochreous, ? maculate black to ash colour, suckers white. — ? 14-15 — — Smooth with “ dorsal beards.” Reddish. 1 ? 20-25 from fig. - - Smooth. Blackish-brown. ABODE BCADE 32 (?) R=B 2L, no entocone. 3L, wide base. 2-4 No groove, a transverse pit. Smooth. iX Dull purple. B>rest. E absent I 20 — — “ Densely wrinkled.” Dark violet. C=AB=D ? E Sc. 33-30 r=b4 1L, cusp very pro¬ minent. 2L, no entocone, no heel. [8] Smooth, no cirrhi. Slate -coloured (dried). C=D.B.A=E 33-50 j R=A2 1L, very deep, base triangular. 2L, no heel. 2-5 Branched papillae. See text p. 151. Dull ochreous grey, circular and irregular dark purple mark¬ ings. # Warts coalesce to form ridges (Berry). B.M. CEPII. E 50 A MONOGRAPH OF THE CEPHALOPODA 1. Dorsal length of mantle (mm.). 2. Width index. 3. •Inter- ocular index. 4. Arms, formula. 5. Arms, length. 6. Gill- filaments. 7. Diameter of suckers. 8. chierchiae Jatta, ’89 and ’99 2>3>4 80(?) 11 ? 7-8 11. validus Sasaki, ’20 44 100+ ? 95 l>2>3>4 66 + n — 12. spinosus Sasaki, ’20 ? 23 — 95? Subequal 2.3>1.4 64 10-11 “ Small,” ?7% U.S.N.M., $ 19 84 78 ,2341 73 9 7-9 13. ochotensis Sasaki, ’20 2> 3>=4” (Sasaki) 75 ? 10-11 17. hawaiiensis Eydoux and Souleyet 23 (from fig.) 78 69 2. 1=3=4 76 - — _ 18. membranaceus Quoy & Gaimard Orbigny 17 (Orb.) 82 (Orb.) 82 (Orb.) 23 = 41 73 , — ■ 19. bermudensis Brit. Mus., etc. 12 81 72 1.2. 3. 4 86 ■ — — 20. joubini 9 Brit. Mus., etc. 16 75 62 2 = 3.4.1 67-69 - 6 21. v errilli Verrill “ pictus ” 8 100 100 Subequal 66 - - 22. sp. = brevipes Hoyle Brit. Mus. - 84 78 Subequal 48 10 - 22a. sp. (p. 164) Brit. Mus. — 53 60 2. 3.4.1 70 — — 23. sp. (p. 164) Brit. Mus. — . 60 44 ? 82 — - 24. superciliosus Quoy & Gaimard Orbigny Type 16(?) 59 (fig. Q- & G.) 50 (fig. Q. & G.) 2.4.3. 1 (Orb.) 77 (Orb.) — — 26. capensis Eydoux and Souleyet 5(?) 66(?) 66(?) Subequal 41 — 26. wolfl Frankfurt 2,x2 2. scorpio Berry, 1920 (?)6 64 64 3241 79 ? 9 3>2,x2+ 3. bandensis Brit. Mus. 6-7 83-78 100-92 321=4 80-89 _ _ a type $ 3421 Massy, ’16A $ 3124 Appellof ’98 ? 30 3>4 0)2 Xl-2-1-6 4. kempi Brit. MuS. 10 80-95 57-60 3241 72-78 _ _ 9 type 3>l-5-0-8 5. elegans Gdttingen (type) 21 38 — 3241 3>2,l-5 83 — ? 9 2. Cislopus indicus Brit. Mus., £ Orbigny, type 46-70 52-87 34-64 1243 84-87 10-11 13-17 8. Pinnoctopus Sub-equal Sc. 81 1. cordiformis Quoy & Gaimard, 175 (?) 69 (fig.) — — — 1832 1.4>2.3 2. kermadecensis Berry, 1916 43 46 39 1234 72 (?) 9 123=4 MEASUREMENTS 53 8. Web formula. 9. Web, depth. 10. Eadula. 11. Hectocotylus, length of ligula. 12. Sculpture. 13. Colour. A=B=C=D.E (?) 25-35 (?) 3-2 Smooth. Dull yellowish-grey sprinkled with small chromatophores. ABODE 18 R=a3 1L, large cusp. 2L, no entocone, heel present. 3L, straight, tips reflected. 6-2-8 Smooth (?), faint trace of low broad warts. Pale ochre or grey marbled greyish- purple. — “ “ Long.” “ Smooth.” “ Dark steel grey blotched irregularly with pale grey.” ? A = B— C = D.E 23-15 H-A-3-5 1L, low cusp. 2L, no entocone, heel weak. 3L, heavy and curved. M. degenerate ? 4-8-8-6 Smooth. See text. Extensive. — — — Smooth. Deep violet ; arms marbled. CBD = EA 23 r=a3? 6-9 Smooth ? Deep violet ; arms and OBDA=E 28 1L, see text. 2L, short base, low heavy cusp. 6-3 Smooth ? body marbled (dis¬ coloured). a=b=o=d.e 21 R unicuspid 1L, low, coarse cusp. 2L, narrow base. M. not degenerate. Rough. — ? 10+ - - Scattered warts on dorsum. Dullish-pink. (?) Equal. (?) 14 - - Smooth. See text Unequal. (?) 10 - - Papillate. See text. Equal (Massy). 11 ' . — — Warts on back and sides ; ocular cirr hi (?). See text. CB = DA=E 16 - — Smooth. See text. Subequal. 6 - - Smooth. See text. ABODE (0=BAED) etc. 13-18 R=A, A tricuspid, noseriation. See text. See text. Pine low widely- spaced warts. ? Dull purple. ? Subequal. 16-14 Tuberculate. Brownish-red, pale below, head and arms with light blue lunules ; fins bor¬ dered with bluish- green. ? A>B>C>D>E 16 Smooth; a few weak dorsal and ocular papillae. Dull gre'y - brown streaked and mottled with slate colour. 54 A MONOGRAPH OF THE CEPHALOPODA 1. Dorsal length of mantle (mm.). 2. Width index. 3. Inter¬ ocular index. 4. Arms, formula. 5. Arms, length. 6. Gin- filaments. 7. Diameter of Suckers. 4. Joubinia 1. fontaniana Brit. Mus. 2>3=4 80 10-12 — 5. conispadiceus U.S.N.M., $ 70 92 61 2143 79 10-10 15 Sasaki ( l.c . ’17), a " Sc. 50 1=2=3>4 75 10-12 9. Hapalochlaena 3241 3421 68-75 1. maculosa Brit. Mus. E D.C=BAE or EA 23 — 11-2-8 13-9 Numerous papilliform tubercles with stel¬ late bases and many irregular wrinkles ; ocular cirrhi ; bi¬ laterally arranged dorsal cirrhi. Dull brownish or purple heavily macu¬ late. b=c=dae C=D=BAE 23-6 r=a4 1L, very narrow base. 2L, no entocone. 11-20 Small, simple granules and short irregular rugae ; ocular cirrhi. ? Brownish-purple. A=B.C.D=E “ Broadest between ventral arms.” 25 - 6-7 Beset with roundish warts. Pale brown. BCADE (subequal) 30 — 12 20-16 Mainly smooth ; a few warts round head ; ocular cirrhi. . — Subequal. 20-48 b=a2 1L, wide base. 2L, very low cusp. 3L, very slender. 4-6-10 Rugose or with low close warts. See text. Subequal. 40 r=a2 1L, very weak cusp. 2L, no entocone, heel. Marginals ? dege¬ nerate. Tuberculate or smooth; ? gelatinous. See text. — 25 ± — — — - XI. SYSTEMATIC. Genus I. Octopus, Lamarck, 1798. Octopodinae with normal ink sac and moderate web. The web is usually bilateral in the adult and C or D is usually deepest. The penial diverticle is single. The hectocotylus is dextral, usually short ; the eggs (so far as known) are small (under 5 mm. long). No velar pouches or “ fins.” The mantle is aperture wide. Type of the genus. — Octopus vulgaris , Lamarck, 1798. Linne (1758, p. 658) placed an octopus in his genus Sepia, naming it Sepia octopodia. The identity of this form as an Octopod is established by Linne’s citations (e.g. Jonstonus (1657, t. 1, f. 1). Linne did not designate the type of his “ Sepia” This was accomplished in 1910 by Hoyle, Sepia officinalis (the Common Cuttlefish) being thus designated. At the same time Hoyle designated Octopus vulgaris as the type of Octopus, Lamarck. Sepia (with Sepia officinalis as type) was placed on the “ Official List of Genera ” by the International Commission on Zoological Nomenclature (Opinion 94, pp. 12-13) Any ambiguity as to the right generic name of the true Octopods is thus disposed of. The definition of the Linnean genus Sepia [“ Brachia sex interius adspersa cotyledonibus (praeter 2 Tentacula longiora quibusdam). Os inter brachia, terminale. Oculi ? infra tentacula versus corpus. Corpus vagina excipiens pectus ”] is very defective, and the words “ Brachia sex . . . praeter 2 tentacula ” (“ six arms in addition to two tentacles ”) are applicable neither to our Sepia nor our Octopus. Moreover, Sepia octopodia (“ tentaculis nullis ”) by his definition must have six arms only ! Apart from this error (obviously accidental because the name octopodia is given) the Linnean Sepia is a composite group containing “ officinalis ,” “ media,” “ Loligo ” and “ Sepiola,” forms recognizable from his citations as our Sepiola, Loligo, etc. Gmelin (1790, p. 3149) altered Linne’s octopodia to octopus ; but he contributed nothing to a better arrangement of the diverse elements contained in Linne’s Sepia. The relationship and names of the Octopus, Cuttlefish, etc., were regularized by Lamarck (1798, p. 130; 1799 p. 1), who distinguished Sepia (with Sepia officinalis and tuberculata) from Octopus (with vulgaris, granulatus, cirrhosus and moschatus). None of the constituent species are designated by him as types of these genera ; but his remarks on p. 2 of his Memoire (1799) and the nature of his definitions leave no doubt in the mind as to the relation¬ ship and character of the groups in question, and we now have no hesita¬ tion in accepting Octopus as the right name for the group in which our common Octopus is to be placed. The choice between Lamarck’s Octopus and Polypus of Schneider, 1784 (the latter name being recommended by Hoyle (1901) and freely used by many writers since that date) is, I think, quite easy. Schneider (1784, p. 116) obviously attempted to formulate definitions of certain groups of Octopods and he plainly grasped the dis¬ tinction between e.g. Eledone and his Polypus. It may, therefore, be 56 OCTOPUS 57 felt that such groups were equivalent to genera, though he actually calls them species. But (1) Schneider does not use binominal nomenclature and (2) his categories are not those accepted in traditional systematic practice (cf. Grimpe, 1920, p. 206). If evidence is wanted as to the ambiguous status of his groups “ Polypus ,” etc., it is forthcoming, as Mr. J. R. le B. Tomlin has pointed out to me, on p. 113 of his work, where his “ Teuthis ” is identified with Linne’s media {Sepia) ! It is thus quite impossible to accept the names proposed in this work. Subgenus i. Octopus, Lamarck (Orbigny). The arms are not conspicuously different in length. There are two rows of suckers. The penial diverticle is short. Suckers unmodified in female. Gill filaments rarely exceed eleven. Type of the subgenus. — Octopus vulgaris, Lamarck, 1798. Lamarck’s Octopus was subdivided by Orbigny (1835, p. 8) into two subgenera, Philonexis and Octopus (s.s.). Orbigny’s subgenus is here revived after many years of neglect. A. Group of Octopus vulgaris. Octopus (Octopus) vulgaris, Lamarck. (Plate I, fig. 1 ; text-figs. 6, 7.) Not Sepia octopodia, Linne (1758, p. 658), Fabricius (1780, p. 360 [? = Bathypolypus arcticus]) [Auctt.]; ? Sepia octopus, Gmelin (1791, p. 3149) ; Octopus vulgaris, Lamarck (1798, p. 130) ; “ le poulpe common Montfort (1805, p. 113, figs.); Sepia octopus, Bose (1802, p. 68); Sepia octopodia, Cuvier (1817a, p. 363), id. (18176, p. 2); Octopus vulgaris, Carus (1824, p. 319, PI. XXXI), Blainville (1825, p. 365, PI. II, fig. 1), id. (1826, p. 188); 1 Octopus brevitentaculatus , Blainville {l.c., p. 187); Octopus vulgaris, Payraudeau (1826, p. 172), Savigny (1826, p. 9, fig. 1), Risso (1826, p. 3), Orbigny and Ferussac (1826, p. 142), Ehrenberg (1828, a'), Sangiovanni (1829, p. 321), Wagner (1829, p. 387), delle Chiaje (1830, pp. 40, 55, PI. LVI, fig. 1) ; Sepia octopodia, Oken (1835, p. 536) ; Octopus vulgaris, Philippi (1836, p. 240), Rang (1837, p. 62, fig.), Ferussac and Orbigny (1840, p. 26, PL 2, etc.); Octopus vulgaris, Orbigny (1841, p. 11, PL 1, fig. 1 (as var. americanus)) ; Octopus vulgaris, delle Chiaje (1841, p. 2 and passim in anatomical section (p. 13)), id. (1841a, p. 65), Cantraine (1841, p. 18), Krauss (1848, p. 132), Gray (1849, p. 6) ; Octopus cassiopeia, id. {l.c., p. 9) ; Octopus vulgaris, Verany (1851, p. 16, PL 8), Forbes and Hanley (1853, p. 209, Pl. NNN, fig. 2), Aucapitaine (1863, p. 290), Fischer (1867, p. 12), Jeffreys (1869, p. 144) ; ? Octopus troscheli, Tozzetti (1869, p. 588) ; Octopus vulgaris, id. {l.c., p. 587), Lee (1875 passim) ; ? Octopus americanus, Guppy (1877, p. 136) ; Octopus octopodia, Tryon (1879, p. 113, Pis. 23-24); Octopus vulgaris, Fischer (in Tryon, l.c., pp. 62, 64), Stossich (1880, p. 157), (?) Yerrill (1880a, p. 253), Ninni (1884, pp. 159, 161), Hoyle (1886, p. 6, etc.), Appellof (1886, p. 7), (?) Jatta (1889, p. 64), Ortmann (1888, p. 642), Kolombatovic (1890, p. 7), Norman (1890, p. 466), Carus (1890, p. 459), Goodrich (1896, p. 19), Jatta (1896, p. 212, Pl. 4, etc.), Lonnberg (1896, p. 706), Joubin (1900, p. 33) ; Polypus vulgaris, Hoyle (1901, p. 1) ; Octopus vulgaris, Marchand (1906, p. 753, 1907, p. 311); Polypus 58 A MONOGRAPH OF THE CEPHALOPODA vulgaris , Hoyle (1907a, p. 35), Pfeffer (1908, p. 20, figs. 11-13); Octopus vulgaris, Zimmerman (1907, p. 293), Joubin and Fischer (1907, p. 322), Joubin (1907a, p. 48), Leidenfrost (1908, p. 160), Lo Bianco (1909, p. 650) ; Polypus vulgaris, Massy (1909, p. 1), Wiilker (1910, p. 5), Berry (19126, p. 386), Grimpe (1913, p. 531) ; Octopus vulgaris, Coen (1914, p. 1); Polypus vulgaris, Thiele (1915, p. 487); Octopus vulgaris, Tanner (1916, p. 24), Grimpe (1918, p. 593) ; Polypus vulgaris , Wiilker (1920, p. 56) ; Octopus vulgaris, Joubin (1920, p. 32), Monticelli (1921, p. 187), Naef (1921, p. 538), id. (1922, p. 288); Polypus vulgaris, Belcher (1922, p. 312) ; Octopus vulgaris, Naef (1923, p. 695), Joubin (1924, p. 10), Robson (1925, p. 106), id. (1926, p. 185), Grimpe (1925, p. 13), Degner (1925, p. 78), Massy (1928, p. 26) ; Polypus vulgaris, Boone (1928, p. 16) ; Octopus octopodia, Winckworth (1928, p. 49). Type specimen. — Not traced. Specimens examined. (а) In Brit. Mus. One specimen (? sex) from Dover : 70.10.1.1. One (2) from Brighton : 1928.4.15.1. One (. Ch., Cams (1890, p. 460) ; Octopus rugosus, Ortmann (1891, p. 669) ; Octopus granulatus, Brazier (1892, p. 4) ; Octopus tuberculatus, Fra Piero (1894, p. 270) ; Octopus granulatus, Goodrich (1896, p. 19) ; Octopus tuberculatus, Lonnberg (1896, p. 706) ; Octopus rugosus, Joubin (1897a, p. 99) ; Octopus granulatus var. rugosa, Joubin (1898, p. 22) ; Polypus granulatus, Hoyle (19046, p. 195) ; ? Polypus occidentalis, id. (1904a, p. 14) ; Polypus granulatus, id. (1907 a, p. 36) ; Wiilker (1910, p. 5) ; Octopus granulatus, Weindl (1912, p. 270) ; Polypus granulatus, Thiele (1915, p. 487) ; Polypus rugosus, Massy (1916a, p. 189), ead. (19166, p. 147, figs. 5-6); Polypus occidentalis, ead. (l.c., p. 148, figs. 7-8) ; Polypus granulatus, Wiilker (1920, p. 49); Octopus tuberculatus, Pallary (1920, p. 17); Polypus granulatus, Robson (1921, p. 440); Polypus granulatus, Thiele (1921, p. 436) ; Octopus tuberculatus, Odhner (1922, p. 221) ; Polypus rugosus, Robson (1924, p. 669, fig. 42) ; Polypus granulatus, Massy (1925, p. 222) ; Octopus rugosus and granulatus, Robson (1925, pp. 104-5) ; Octopus rugosus, Robson (19266, p. 188, fig. 17), Winekworth (1926, p. 324), Peile (1926, p. 98); Polypus granulatus, Massy (1926, p. 165). Type specimen. — Not traced (? Holotype). Specimens examined. — A. West Atlantic. (a) In Brit. Mus. One (?) [C. 157] from Porto Caballo, Venezuela : 49.12.7.60. One ((J) [C. 199] from Bermuda : 72.9.2.5. Two (??) [C. 125-6] from Jamaica (“ eudora ”) : 46.8.31.3. One (?) [C. 96] from “ West Indies ” : 69.5.22.27. One (?) [C. 259] from Bahia : 1903.9.17.9. One (?) [C. 258] from E. Brazil : 1908.12.11.1. One (A>E, and is less than 25% of the maximum arm- length. The arms are provided with lateral membranes. The radula, funnel-organ and internal and external genitalia are unknown. The dorsal surface is ornamented by “ numerous low rough conical tubercles.” The skin between these is either smooth or finely papillose. The sculpture is found on the inner surface of the web. There are no noticeable supra¬ ocular cirrhi. The colour is “ dark slate,” somewhat lighter below. The inner surfaces of the suckers are light brown “ sometimes shading to cream.” Maximum size. — 190 mm. Remarks. — This species is known only by two female examples. According to Oliver ( l.c ., p. 564), it is found in the Sargassum Belt and may be regarded as littoral, probably intertidal. As Berry rightly remarks (19146, p. 137), it is not in any way related to the other native Kermadec Octopod, Pinnoctopus (?) kermadecensis . Its immediate relationship is obscure. Berry rightly compares it with Octopus vitiensis , and I have followed him in including it in the same group. The narrow pallial aperture and shallow web distinguish it from that species, and, combined with the short arms and the fact that sector A of the web is deeper than E, tend to make it rather unlike the majority of the vulgaris- like group. B. Group of Octopus macropus. (a) Subgroup of Octopus macropus. Octopus (Octopus) macropus, Risso. (Text-figs. 27-30.) Octopus Lechenaultii, cuvierii, Orbigny (1826, plates only) ; ? Octopus granosus, Blainville (1826 (September),* p. 186) ; Octopus macropus, Risso (1826 (November),* p. 3), Wagner (1829, p. 387); “ Polpo rossastro,” delle Chiaje (1830, p. 40) ; Octopus macropus, id. {l.c., p. 56) ; Octopus macropodus, Sangiovanni (1829, p. 319); Octopus macropus, Rang (1837, p. 61); Octopus cuvieri, Orbigny (1840, p. 18, Pis. 4, 27); Octopus longi- manus, id. (l.c., p. 18) ; Octopus macropus, id. (l.c., PI. 24), delle Chiaje (1841, p. 3; 1841a, p. 65); Octopus ruber, Cantraine (1841, p. 18); Octopus macropodus, Philippi (1836, p. 240) ; Octopus cuvieri, Gray (1849, p. 13) ; Octopus macropus, Verany (1851, p. 27, PI. 10); ? Octopus alderii, id. (l.c., p. 32, PI. 7 bis) ; Octopus macropus, Steenstrup (1856, p. 202), Troschel (1857, p. 59), Targioni Tozzetti (1869, p. 588) ; Octopus cuvieri, * I am indebted to Mr. C. D. Sherborn for information as to the dates on which Blainville and Risso published these works. 102 A MONOGRAPH OF THE CEPHALOPODA id. (1869A, pp. 23, 24), Tryon (1879, p. 122) ; Octopus macropus , Tiberi (1880, p. 10), Stossich (1880, p. 157), Ninni (1884, pp. 159, 161), Hoyle (1886, pp. 11, 95); Octopus Cuvieri, Appellof (1886, p. 6, PL 6, fig. 1); Octopus cuvieri , Brock (1887, p. 597) ; Octopus macropus , Ortmann (1888, p. 643), Kolombatovic (1890, p. 7), Cams (1890, p. 460), Joubin (1894a, p. 212); Octopus ruber, Fra Piero (1895, p. 269); Octopus macropus, Lonnberg (1896, p. 706), Jatta (1896, p. 217, Pis. 6, 23, 24), (?) Goodrich (1896, p. 20), Joubin (1897a, p. 99), id. (1898, p. 22) ; ? Polypus macropus, Hoyle (1904a, p. 18), Hoyle (19046, p. 195), Hoyle (1907a, p. 36 (?)) ; Octopus macropus, Marchand (1907, p. 362), Leidenfrost (1908, p. 163), Lo Bianco (1909, p. 650); Polypus macropus, Wiilker (1910, p. 8), Berry (19126, p. 389); Octopus macropus, Weindl (1912, p. 270) ; Polypus macropus, Wiilker (1913, p. 456), Massy (1916A, p. 192) ; Octopus cuvieri, Odhner (1917, p. 70); (?) Polypus macropus, Sasaki (1920, p. 181), Wiilker (1920, pp. 50, 56); Octopus cuvieri, Pallary (1920, p. 17) ; Octopus macropus, Naef (1923, p. 702), Robson (19266, p. 187, fig. 16), Winckworth (1926, p. 325). Type specimen. — (?) Municipal Museum, Nice. (? Syntypes.) Specimens seen. (а) In Brit. Mus. One specimen (5) from Naples : 98.5.21.344. One (2) from “ the Mediterranean”: 79.1.20.2. One (?) from Yokohama: 89.4.24.39. One (9) from Chekiang : 1928.3.21.2. Three (29c?) from Amoy : 1928.3.30.3-5. Two ((J9) from Santuao, China : 1928.3.30.7-8. One (D>E, and to be about 25% of the arm’s length. The arm-membranes are broad and well-developed. The funnel is long and rather slender. The funnel- organ is not described. The hectocotylus is unknown. “ The surface is smooth except for a number of short, rather obscure simple papillae ” on the dorsal surface of the head, neck and mantle. There is a single tubercle * There is some discrepancy between the dimensions as shown in the measurements and those seen in the figure (PI. XL, fig. 4). 112 A MONOGRAPH OP THE CEPHALOPODA over each eye, and the rest are bilaterally arranged, “ the most notable being a nearly equidistant row between the eyes.” The colour is very pale- grey buff suffused with purplish-brown. Maximum length. — 210 mm. Habits. — This species is evidently an inhabitant of rather deep water, as all the specimens are from 106-239 fathoms. Remarks. — I am a little uncertain as to the status of this species. Berry says that probably none of the specimens available is quite mature. His reliance on the lateral keel as a striking difference between it and its congeners is not quite justified (cf. p. 7). The shape, the shortness of the arms (? due to age) and form of the web are perhaps distinctive in combination. It is hoped that Dr. Berry will publish an account of the funnel-organ, radula, etc. It is possible that this species may be referable to the Bathypoly- podinae. The narrow mantle-aperture and short arms, in conjunction with the rather deep-water habitat, suggest this relationship. Octopus (Octopus) longispadiceus (Sasaki). Polypus longispadiceus, Sasaki (1917, p. 366; 1920, p. 178). Holotype. — In Science College, Tokyo. Specimen seen. — One (> S3 * s C •43 43 © u d c$ ft fe © §>! d > m *-h O O !=3 ft ft m d a xn 43 43 © g ft M © 03 tn d © 42 :d d 42 43 m ^ j> d f4 o g c6 cS © d O § ssj ■§ t= g^'S O M 02 O TO ft ^ £=d 81 Hect., % arm. * I | co ^ ? 9 9 t1 U || 1 1 1—1 rtf 1> CD CD IO II Web, % longest arms. 31 Very long 33 ? 20 25-33 26 32 21 27 Web. 18 22 24 25 22 C = D., B = E.A. _ /Well developed ( ? C = D = E.B.A. BCDEA C = D, B = E.A. CD.EBA C = D = E.B.A. C.D.B.E.A. Longest arm, % total length. 68 sc. 66 70 75-76(?) 69 78 75-74 81 69 77 81 Order of arms. 3214 Subequal I 4 S 83 3 44 « a . * >C o c© © ft EH oo HH 05 3 a" 2 2 £> «44 JO ’"© 5 o d §». eS lO a • eg ^ d 74—' © 45 © © ”**! 3 H No pH K £ § > 05 >» O w pc § -a ft ft <1 o £2 d^ 5S • g* d © o £4 © PP H c4 co Hi io CD 1> o ** '?7 . ■ . $ i4| ■ . ■ > , . - . ■ " -- • ’ . . ■ INDEX The letter (m) after a page-number indicates that the various measurements and proportions of the species are given on that page. Numbers in black type (133) indicate the pages on which the various species are fully described. Abbreviations, 38, 39. abnormalities, structural, 18. abruptus, Bathypolypus, 40. abyssal mode of life, 2, 14. aculeatus. Octopus, 48 (m), 81, 141, 142, 144. ,, Polypus, 91, 141. adaptation, 23, 193. adductor pallii medius, 29. adhesive (locking) apparatus, 10, 39. aegina, Octopus, 16, 35, 46 (m), 113, 114, 116, 137. africana, Joubinia fontaniana var., 189. alatus, Octopus , 35, 50 (m), 153. ,, Polypus, 153. alberti, Haptochlaena, 41. albus. Octopus , 215. alderii, Octopus, 101, 106, 108. aldrovandi, Octopus, 217. alimentary system, 12. amboinensis , Octopus, 48 (m), 111, 134, 172 “ americanus. Octopus ,” 73. ,, ,, vulgaris var., 57, 63, 73. ,, Polypus, 73. Amphineura, 19. Amphioctopus, 33, 40. anatomy, 6 annae. Octopus hoy lei var., 220. “ antillarum, ? Orb.”, Octopus [Dali], 217. antiquorum , Octopus, 215. aperture, mantle, 2, 26, 33, 39. ,, oviducal, 30. apollyon, Octopus, 22, 34, 80, 130, 197, 198, 199. ,, Paroctopus, 54 (m), 200, 201, 202, 203, 204, 207. apparatus copulator, 9. appendiculatus, Octopus, 214. appendix, 16. dranea, Octopus, 143, 144, 161. ,, Polypus, 143. arborescens. Octopus, 48 (m), 81, 151. ,, Polypus, 151. archaic characters, 33. arcticus , Bathy polypus, 40, 186. areolatus, Octopus, 46 (m), 79, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126. ,, and ocellatus (group), Octopus, variation, 118. „ Polypus , 120, 122. Argonauta sp., 215. “ argonautae ,” Octopus, 217. Argonautidae, 17. argus, Octopus, 91. arm -formula, 9. arms, 2, 3, 5, 8, 9, 23, 25, 31, 33, 39. ,, bifurcate, 18, 90. atlanticus, Octopus, 217. Atlantoctopus, 41. australis, Octopus, 17, 35, 48 (m), 90, 144. ,, Polypus, 144. autotomy, 21. backerii, Octopus, 63, 73. bairdii, Octopus, 40. bandensis, Macrotritopus, 52 (m), 168, 170, 171. ,, Octopus, 170. ,, Polypus, 170. Barkerii, Octopus, 73. Bathypolypodinae, 1, 2, 14, 15, 32, 40, 195 Bathy poly pus, 9, 18, 19, 32, 40, 62, 145, 187, 193, 195, 197, 211. behaviour, 22. Benthoctopus, 19, 40, 182. berenice, Octopus, 42 (m), 84. bermudensis. Octopus, 50 (m), 107, 160. berryi, Benthoctopus, 41, 111. bifurcate arms, 18, 90. bimaculatus, Octopus, 16, 42 (m), 79, 98, 125. ,, Polypus, 79. bodily proportions, 15. body, 6, 7, 39. boscii, Octopus, 69, 98, 100, 126, 128. ,, Polypus, 98, 126. ,, Sepia, 126. brevipes, Octopus, 164, 215. brevitentaculatus, Octopus, 57. bristles, 23. brocki. Octopus, 119 (m), 120, 122. brooding, 1, 22. brucei. Octopus, 176, 181. ,, Polypus, 176, 181. bursarius, Cistopus, 182, 184. calamus, 17. calif ornicus, Octopus, 34, 46 (m), 129, 197, 219. ,, Polypus , 129, 218, 219. 231 INDEX 232 campbeUi, Jovbinia, 54 (m), 190. ,, Octopus , 190 ,, Polypus , 190. capensis. Octopus, 50 (m), 216. carena , Octopus, 216. catenae. Octopus, 216. ,, Philonexis, 111. carolinensis, Octopus, 42 (m), 76. cassiopeia. Octopus, 57, 58. catena, Parasira, 135. catenulata, Philonexis, 213. catenulatus. Octopus, 216, 217. cephaea, Octopus, 18, 44 (m), 90. ,, Polypus, 90. cephalic cartilages, 11. characters used in classification, 4, 39. chierchiae, Octopus, 17, 21, 34, 50 (m), 152, 175. chromatophores, 20. chromatus, Octopus, 44 (m), 103, 107, 161. chuni, Haptochlaena, 41. Cirrata, 1, 2, 7, 10, 15. cirrhi, 11. . cirrhosus, Octopus, 217. Cirromorpha, 32, 33. Cirroteuthis, 137. Cistopus, 8, 14, 15, 34, 40, 52 (m), 182, 187 classification (general), 3, 31, 40. cocco. Octopus, 191, 192, 193, 216, 217. coccoi, Octopus , 191, 192. ,, Scaeurgus, 192. coerulescens, Octopus, 214. coerulescentes , Octopus, 216. coitus, 21. colossus, Octopus, 215. colour, 3, 23. colour-pattern, 11. communis, Macr octopus, 52 (m), 175. ,, Octopus, 175. ,, Polypus, 175. concretions, calcareous, 18. conispadiceus, Octopus, 34. ,, Paroctopus, 21, 22, 54 (m), 205. , , Polypus, 205. convergence, 2. Corals, 3. cordiformis, Octopus, 184, 185. ,, Pinnoctopus, 52 (m), 185, 186. cornutus, Tritaxeopus, 52 (m), 172. courtship, 21. cranchii, Octopus, 215. crop, 39. Crustacea, 20. cuvieri, Octopus, 101, 102, 106. cyanea, Octopus, 44 (m), 87, 94, 98 (m). Cytherea, 198. damn, Octopus defilippi var., 80, 81, 136, 140, 164. Decapoda, 3, 7, 9, 10, 11, 12, 14, 32, 33. defence, means of, 20. defilippi, Octopus, 7, 14, 17, 21, 33, 35, 48 (in), 135, 136, 137, 140, 164. „ Polypus, 135. degeneracy, 14. demibranch, 14. development, 23. diagnosis, specific, method of, 4. didynamus. Octopus, 214. differentiation, local, 37. digueti, Octopus, 22, 34, 197. ,, Paroctopus, 21, 54 (m), 197, 203. dimorphism, sexual, 15, 24. disease, 18. dispersal, 4. distribution, geographical, 36. ,, vertical, 19. diverticulum, 16. dofleini, Octopus, 34, 46 (m), 130. ,, Polypus, 130. j dollfusi, Octopus, 115. dorsal stylets, 11, 39. dubius, Octopus, 217. duplex, Octopus, 42 (m), 63, 82. ,, Polypus, 82, 145. eggs, 4, 21 (m), 22, 34, 39. Egopsida, 17. Eledone, 6, 34, 62, 199, 214, 217. Eledonella, 216. Eledoninae, 40. elegans, Macrotritopus, 52 (m), 171. „ Octopus, 90, 135, 171. enemies, 20. Enteroctopus, 14, 16, 34, 35, 36, 40, 52 (m), 148, 175, 187. epidermis, 5. equivocus, Macrotritopus, 52 (m), 169, 171. ,, Octopus, 167. ergasticus, Benthoctopus , 41, 191. estuaries, occurrence in, 19. eudora, Octopus, 63, 73. eureka, Enteroctopus, 52 (m), 179, 182. evolution, 1, 4, 14, 31. eyes, 23. eylais, Octopus , 217. faeroensis, Bathy poly pus, 40. fang-siao, Octopus, 120, 122, 124. ,, Polypus, 120, 122, 124. fasciata, Octopus pictus var., 211. fasciatus, Octopus pictus var., 211. ,, Polypus pictus var., 211. Javonia, Octopus, 63, 73. female apertures, external, 18. female reproductive system, 17. ferussacii, Octopus, 217. filamentosus , Octopus, 48 (m), 141, 142, 143. filosa, Octopus, 146. filosus, Octopus, 8, 48 (m), 107, 146, 165. fimbriatus, Octopus, 91, 93. fins, 7, 185. fishery, 19, 22. foemina, Polypus, 73. folds (membranes), lateral, 7, 39, 185. fontaniana, Joubinia, 12, 54 (m), 187, 189, 190, 191. Jonlanianus, Octopus, 22, 157, 167, 187. ,, Polypus, 94, 96, 187. food, 20. formula (arm), 9. ,, (radula), 13. „ (web), 7. INDEX fray edits, Octopus, 214. funnel, 10, 26. funnel-organ, 10, 26, 32, 33, 39. fttrvus, Octopus, 48 (m), 146. fusiformis, Octopus, 7, 34, 35, 48 (m), 90, 132, 134. ,, Polypus, 132. gardineri, Octopus, 12, 52 (m), 166. ,, Polypus, 166. gastropod radula, 13. geryonea, Octopus, 63, 73. gigas , Octopus, 215. ,, Sepia, 215. gilbertianus. Octopus, 46 (m), 131, 158, 197. ,, Polypus, 131. gill-filaments, 33. gills, 2, 14, 28, 32. glaber. Octopus, 95. gland, oviducal, 30. globosus, Octopus, 44 (m), 63, 73, 93, 145. ,, Polypus, 93. gracilis, Octopus, 167, 168, 169, 217. ,, ,, cyanea var., 98. granosus, Octopus, 101, 106. granulata. Sepia, 73. granulatus, Octopus, 63, 71, 72. ,, Polypus, 63, 113. granulosa. Sepia, 73. grimpei, Baihy polypus, 40. Grimpella, 41. groenlandicus, Octopus, 40. gular lamella, 12. habits, 3, 19. ,, breeding, 21. Hapalochlaena, 7, 9, 14, 15, 33, 34, 40, 54 (m), 207. Haptochlaena, 41, 211. hardwiclcei, Octopus, 46 (m), 115. harmandi, Octopus, 141, 142. hawaiiensis, Octopus, 50 (m), 98, 159. ,, Polypus, 159. hawiiensis, Octopus, 159. head, 6, 7. hectocotylization, double, 18, 82. hectocotylized arms, 16, 21. hectocotylus, 17, 30, 32, 33, 34, 39. ,, variation, 60. herdmani, Octopus, 42 (m), 86. ,, Polypus, 86. heteropodus, Octopus, 215. heteropus, Octopus, 215. hians, Argonauta, 217. historical, 5. hoeki, Octopus, 42 (m), 85. hokJcaidensis, Benthoctopus, 41. hongkongensis, Octopus, 16, 19, 21, 34, 129, 130, 158, 197, 199, 204. ,, Paroctopus, 54 (m), 199, 202, 203, 204, 205, 207. ,, Polypus, 199, 202. horridus, Octopus, 11, 15, 19, 44 (m), 81, 85, 91, 93, 141. ,, Polypus, 91. horsti, Octopus, 94, 96, 97, 98. ,, Polypus, 94, 95. 233 hoylei, Octopus, 191, 195, 219, 220, 221. ,, Polyprus, 41,219,220. hyadesi, Benthoctopus, 41. hyalinus, Octopus, 217. Hy droids, 3. Idiosepius, 7. incertus, Octopus, 63. inconspicuus, Octopus, 50 (m), 156. index, interocular, 24. ,, width, 25. indicus, Cistopus, 52 (m), 133, 182, 184. ,, Octopus, 182. ink sac, 2, 5, 15, 20, 32, 33, 39. insufficiently diagnosed species, 214. intelligence, 22. interocular index, 24. januarii, Benthoctopus, 41. jaws, 12. joubini, Octopus, 50 (m), 148, 161. Joubinia, 14, 16, 31, 34, 36, 40, 54 (m), 77, 148, 154, 162, 187, 192, 193, 207. kagoshimensis, Octopus, 113, 114, 115. kempi, Macrotritopus, 52 (m), 168, 170. kermadecensis, Octopus, 185. ,, Pinnoctopus, 52 (m), 101, 186. , , Polypus, 1 86. kodlikeri, Octopus, 216. kraken. Octopus, 217. lair, 20. laminae copulatoriae, 17. lechenaultii, Octopus, 101, 102, 106, 117. leioderma, Octopus, 35, 38, 46 (m), 111. ,, Polypus, 111. lentus, Bathypolypus, 40. Lepidoteuthis, 7. leucoderma, Octopus, 217. levis, Benthoctopus, 41. ,, Octopus, 41. ligula, 17. liver, 15. locking apparatus, 10, 26. longimanus, Octopus, 101, 102. longipes, Octopus, 214. longispadiceus, Octopus, 46 (m), 112. ,, Polypus, 112. lothei, Atlantoctopus, 41. lunulata, Hapalochlaena, 54 (m), 209, 212, 214. lunulatus, Octopus, 11, 19, 207, 208, 209. machikii, Octopus, 44 (m), 110. Macrochlaena, 7, 15, 17, 31, 33, 40, 54 (m), 192, 193, 197, 211. Macroctopus, 7, 10, 33, 34, 36, 40, 52 (m), 173, 174. macropodus. Octopus, 101. macropus. Octopus, 10, 15, 21, 25, 35, 37, 44 (m), 58, 62, 70, 81, 82, 101, 102, 106, 107, 108, 109, 110, 111, 117, 145, 146, 147, 151, 158, 159, 161, 173, 174, 186, 208, 214, 215, 216. ,, Polypus, 102, 186. 234 INDEX Macrotritopus, 33, 34, 40, 52 (m), 90, 135, 167, 168, 193. maculatus, Octopus, 215. maculosa , Hapalochlaena, 54 (m), 214. maculosus , Hapalochlaena, 211. ,, Octopus, 207, 208, 211, 213. madokai, Octopus, 50 (m), 159. ,, Polypus, 159. male reproductive system, 16. mandibles, 24, 28. mantle, 7. ,, width, 25. maorum, Macr octopus, 52 (m), 106, 173, 174, 175, 187, 208. ,, Octopus, 152, 174. ,, Polypus , 174. marmoratus, Octopus, 94, 97, 98, 159, 160. ,, Polypus, 94. mas, Octopus, 72. measurements, 4, 38. ,, table of, 42. medoria, Octopus, 44 (m), 110, 156, 206. megalocyathus, Enteroctopus, 52 (m), 176, 177, 178, 179, 181. ,, ,, variation, 177. ,, Octopus, 147, 176. melanic forms, 92. immbranaceus, Amphioctopus, 125. ,, Enteroctopus, 52 (m), 175, 179, 181. ,, Octopus, 11, 33, 50 (m), 84, 120, 122, 125. ,, Polypus, 125. method, bibliographical, 38. micropthalmus, Octopus, 44 (m), 89. ,, Polypus, 89. mimus, Octopus, 42 (m), 81. minor. Octopus macropus var., 10, 107. ,, Polypus macropus var., 107. • minimus, Octopus, 217. mollis. Octopus, 150, 216. monterosatoi. Octopus, 216. montevideo, Octopus, 217. moschatus, Octopus, 215, 217. moschites, Octopus, 217. Muller’s organ, 10. muscular system, 3. mycrostoma , Octopus, 217. Nautilus, 32, 33, 205. Needham’s organ, 16. nesting, 20. niyer , Octopus, 215, 217. niveus. Octopus, 7, 48 (m), 141, 142, 143, 144. normani, Octopus, 40. obelus, Bathy poly pus, 40. occidental is. Octopus, 63, 64, 70, 73 ,, Polypus, 63, 65. occurrence, cyclical, 20. „ pelagic, 20. orellatus, Octopus, 11, 46 (m), 79, 116, 117, 118 (m), 119 (m), 120, 121, 122, 123, 124, 126. ,, Polypus, 120. ocelli, 11. ochotensis, Octopus, 21, 50 (m), 155, 206. ,, Polypus, 155. octopodia. Octopus, 57, 58, 62. ,, Sepia, 56, 57, 58, 62. Octopodidae, 40. Octopodinae, 40. Octopus, 6, 34, 35, 36, 40, 42 (m), 56, 57, 168, 187. octopus, Sepia, 57, 62. oculifer. Octopus, 11, 42 (m), 84. ,, Polypus, 84. Ocyihoe, 157, 204. officinalis, Sepia, 56. oliveri, Octopus, 44 (m), 100. ,, Polypus, 88, 100. Onychoteuthis , 215. Opisthoteuthis, 137. orientation, 6. ornatus. Octopus, 44 (m), 108, 109. ,, Polypus, 108. orthogenesis, 2. ovary, 17. oviducal gland, 17, 18, 30. oviduct, 17, 30. oviposition, 21. ovulum. Octopus areolatus var., 123, 124 (m). ,, Polypus, 123. palatine lamella, 12. pallial aperture, 2, 9. ,, cavity, 6. palliata. Octopus verrilli var., 163 (m). pallida, Octopus, 35, 46 (m), 69, 99, 100, 126, 155. ,, ,, boscii var., 100, 126, 128. parasites, 22. Par octopus, 17, 34, 35, 36, 40, 54 (m), 131, 156, 197. parvus. Octopus, 50 (m), 153. ,, Polypus, 153. patagiatus, Scaeurgus, 192, 193. patagonicus. Octopus, 18, 48 (ml, 148, 149, 176. Pecten, 198. penis, 16, 30, 61 (variation). pennanti, Octopus, 217. pentherinus. Octopus, 42 (m), 76, 77, 149. Peristernia, 13. peronii, Octopus, 216. Philonexis, 57. Philonexus, 217. phylogeny, 31. pictus, Octopus, 11, 162, 163, 207, 208, 211, 213, 217. ,, Polypus, 211. pilosus, Octopus, 215. Pinnoctopus, 7, 33, 40, 52 (m), 184. piscatorum, Benthoctopus, 40. pisiformis, Octopus, 132. Polypus, 56, 57. polypus, Octopodia, 62. polyzenia, Octopus, 63, 65, 73. pores, aquiferous, 187. post -embryonic stages, 20, 23. post-mortem changes, 3. pouches, velar, 8, 34, 39, 182, 184. preservation, 3, 27. INDEX 235 pricei, Octopus, 90, 218, 219. ,, Polypus, 41, 218. prof undicola. Octopus, 41. Pseudoctopus, 197. pseudonymus, Atlantoctopus, 41. Pteroctopus, 7, 9, 15, 16, 20, 31, 33, 40, 54 (m), 191, 195. pulcher, Octopus, 46 (m), 125. punctatus, Octopus, 130, 159, 160, 199, 200, 201, 202, 204, 215. ,, Polypus, 199. pusillus, Octopus, 48 (m), 150, 151, 161. ,, Polypus, 150. pustulosus, Octopus , 50 (m), 214, 216. ,, Polypus, 159. quoyanus, Octopus, 217. radula, 12, 13, 24, 28, 32, 33, 39. range, vertical, 19. raricyathus, Octopus, 215. regeneration, 21. reproductive organs, 15, 30. reticularis, Octopus, 217. ridge, tegumentary, 7. rivalry of males, 21. robustus, Octopus, 144, 208, 211, 212, 213, 214. rostral lamellae, 12. rotunda, Eledone, 14. ruber, Octopus, 101, 102, 215. rufus, Octopus, 217. rugosa, Octopus granulatus, var., 63. ,, Sepia, 63, 71. rugosus, Octopus, 7, 18, 21, 22, 37, 42 (m), 58, 62, 63, 75, 76, 78, 83, 85, 87, 88, 94, 107, 113, 115, 128, 153, 155, 160, 163, 166, 214, 215. ,, ,, variation, 67. ,, Polypus, 63, 126. salebrosus, Benthoctopus, 40. salinity, 19. salivary glands, 20, 39. salutii, Octopus, 21, 35, 50 (m), 104, 157. ,, Polypus, 157. saluzzi, Octopus, 157. sanctae helenae, Octopus rugosus var., 74. saphenia, Octopus, 160. sasakii, Benthoctopus, 41. Scaeurgus, 16, 17, 34, 40, 54 (m), 168, 187, 191, 193. Scaphopoda, 19. Schizoctopus, 33, 40. schultzei, Octopus, 48 (m), 148. ,, Polypus,' 148. scorpio, Macrotritopus, 52 (m), 168, 169. sculpture, 3, 11, 26, 39. seasonal change (in hectocotylus), 17. sectors (of web), 7. semipalmatus, Octopus, 217. Sepia, 6, 186. Sepiola, 7. septum, pallial, 29. seriation, 14, 29. sex-ratio, 16. sexual differences, 7, 9, 15 and Section V passim. sinensis, Octopus, 116, 120, 122, 124. size, 1. skeletal structures, 11. skin, 11,23, 26,39. specialization, 33, 34. species, 4, 23, 24. sp. Enleroctopus, 52 (m), 181. ,, Octopus, 50 (m), 108, 164. ,, Scaeurgus, 192. spermatic gland, 16. spermatophores, 16. spinosus, Octopus, 50 (m), 155. ,, Polypus, 155. sponsalis, Bathypolypus, 40, 220. stages, developmental, 23. suckers, 5, 9, 28, 34. ,, enlarged, 9, 21. Suez Canal, 19. superciliosus, Octopus, 50 (m), 165. systematic position (general), 2. taprobanensis. Octopus, 44 (m), 108, 186. taxonomy, characters used in, 39. tchang-iu , Octopus, 124. tehuelchus. Octopus, 48 (m), 147, 149, 161, 176, 178, 179 and foil. ,, Polypus, 147, 179. tenebricus, Octopus, 42 (m), 80, 152, 157. tenuipulvinus, Octopus, 10, 35, 46 (m), 131. ,, Polypus, 131. terminology, 6. testis, 16. tetracirrhus, Octopus, 191, 195, 196. ,, Pteroctopus, 18, 54 (m), 195, 211. ,, Scaeurgus, 192, 195, 196. tetradynamus. Octopus, 215. tetricus. Octopus, 44 (m), 90, 98, 100, 128. teutlioides, Octopus, 7, 34, 48 (m), 133. thaumastocheir, Grimpella, 41. titanotus, Scaeurgus, 191, 193, 195, 196. tonganus. Octopus, 42 (m), 77, 167. ,, Polypus, 77. toxin (secreted by salivary glands), 20. Tremoctopus, 216, 217. Tritaxeopus, 33, 40, 52 (m), 172. tritentaculatus. Octopus, 217. troscheli. Octopus, 57. tschawytscha, Onchorhynchus, 218. tsugarensis. Octopus , 35, 46 (m), 113. ,, Polypus , 113. tuberculata, Ocythoe, 111, 135, 213, 216, 217. tuberculatus, Octopus , 63, 71, 72, 88, 215. ,, Philonexis, 217. uncertain generic position, species of, 218. unguiculatus , Octopus, 215. unicirrhus, Octopus, 191, 192, 193. ,, Scaeurgus, 7, 37, 54 (m), 192, 195,' 196, 197, 216. valdiviae, Bathypolypus, 40. validus , Octopus, 50 (m), 154. ,, Polypus, 154. variation, 5, 24, 67 and passim. ,, individual, 4. 236 INDEX variolatus, Octopus , 100, 126, 128, 214. ,, Polypus, 98, 126. vas deferens, 16. vdatus , Octopus , 217. velifer, Octopus, 217. Velodona , 169. veutricosus. Octopus, 217. venustus , Octopus, 216. verany\i\ , Octopus, 216. verrilli , Octopus, 50 (m), 162. verrucosa. Octopus octopodia var., 74-5. verrucosus. Octopus, 42 (m), 74. ,, Polypus, 74. violaceus. Octopus, 217. visceral mass, 7. vitiensis , Octopus, 42 (m), 63, 87, 88, 101. ,, Polypus, 87. Voluta , 189. vulgaris. Octopus, 14, 15, 16, 17, 18, 19, 20, 21, 22, 24 and foil., 32, 34, 37, 42 (m), 56, 57, 58, 59, 60, 61, 62, 65, 69, 70, 73, 75, 78, 85, 90, 101, 104, 105, 106, 121, 132, 137, 138, 152, 157, 158, 159, 165, 183, 192, 194, 199, 215, 216. vulgaris, Polypus , 57, 58. web, 7, 8, 27, 32, 33, 39. ,, evolution, 1, 2. ,, formula, 7. westerniensis, Octopus , 165. width-index, 25. ,, interocular, 25. winckworthi, Macrochlaena , 54 (m), 194. ,, Octopus, 193, 194. wolfi. Octopus, 50 (m), 166. ,, Polypus, 166. “ year,” “ Octopus,” 20. yendoi. Octopus, 34. ,, Paroctopus, 21, 54 (m), 205. ,, Polypus, 205. BRITISH museum a f T AUG 29 • • NATURAL Ml STORY. PLATE I “Sepia ociopodia ” (? syntype). Uppsala Museum BRITISH MUSEUM 7 AUG 29 NATURAL HISTORY. PLATE II Fig. 1. — Octopus berenice (holofcype). x 1. Fig. 3. — Octopus rugosus, eggs, x 1. Fig. 4. — Octopus teuthoides Fig. 2, — Octopus areolatus, var. ovulum. x 1. j BRITISH j MUSEUM \ 7 AUG 29 jj NATURAL j HISTORY. K-: ■I ' T > p ■ •/ PLATE Ilf ocrllutitH . . type), y 1. Kid. 2— Octopus hardwithi (syntype). BRITISH MUSEUM 7 AUG 29 NATURAL HISTORY. PLATE IV ■Octopus mat on' a (holotype). BRITISH MUSEUM 7 AUG 29 NATURAL HISTORY. PLATE V ■Ochpu* aegina (syntypo). x -8. Fio. i.— Octopus ctphta (holotype). British ! museum ! 7 AUG 20 1 NATURAL history PLATE VI ■Octopus joubini (holotype). x 2. Fig. 2. — Octopus iaprobanensis { holotype). x 2. Fig. 3. — Enteroctopus sp. I BRITISH MUSEUM ! 7 AUG 29 NATURAL history. PLATE VII urwlatus ' vtit \ |>' ; p. 117). Leiden MiiHoum. Pm. 2. — Octopus ocellatus (svntypo of (Octopus at colatus. Orbigny). Leiden Museum. [BRITISH I MUSFNM museum 7 AUG 29 natural history.