October 29, 1S91] 



NA. URE 



613 



OUR BOOK SHELF. 



The Melanesians : Studies in their Anthropology and Folk 

 Lore. By R. H. Codrington, D.D. (Oxford : Clarendon 

 Press, 1 891.) 



In this book Dr. Codrington gives us the results of ob- 

 servations and inquiries made in the Melanesian Islands 

 from 1863, when he first visited them, to 1887, when he 

 left the Melanesian Mission. He does not profess to 

 offer a complete account of the Melanesian people ; 

 nevertheless, the work is one of great value, for it is in 

 the main a record, not of what Europeans say about the 

 natives, but of what the natives say about themselves. 

 The most careful of European inquirers may, of course, 

 mistake the real significance of what natives tell them ; 

 but Dr. Codrington seems to have been at all times fully 

 conscious of this danger, and to have done his best to 

 guard against it. 



He begins with a chapter on the discovery of the 

 Melanesian Islands, and on their geology and zoology. 

 The ethnology of Melanesia he dees not attempt to deal 

 with ; but he discusses thoroughly the facts relating 

 to kinship and marriage connection among the Mela- 

 nesians, starting wiih the proposition that the division of 

 the people into two or more classes, which are exogam.ous. 

 and in which descent is traced through the mother, is the 

 foundation of native society. He also gives a good 

 account of the position of the chiefs. A chapter is 

 devoted to property and inheritance, and this is followed 

 by a description of secret societies and clubs, a knowledge 

 of both of which is essential to a proper comprehension 

 of Melanesian life. 



The religion of the Melanesians, like that of all savage 

 and barbarous peoples, is a subject of great difficulty ; 

 but Dr. Codrington is able to present clearly what seem to 

 be at least its main outlines. Students of the evolution of 

 religious conceptions will read with especial interest what 

 he has to say about "mana," a supernatural power or 

 influence which is supposed to act in all kinds of ways 

 for good and evil, and which everyone tries to possess 

 or control. The objects of worship are spirits, some of 

 which were formerly men, while others belong to an inde- 

 pendent and higher class. All these beings are full of 

 "mana," and many suggestive facts about the popular 

 belief in them will be found in the chapters on sacrifices, 

 prayers, spirits, sacred places and things, magic, posses- 

 sion, and intercourse with ghosts. There are also good 

 chapters on birth, childhood, and marriage ; death, burial, 

 and "after death." 



The chapters on the arts of life, and on dances, music, 

 and games, contain an immense number of interesting 

 facts, well arranged ; and in a chapter entitled " Miscel- 

 laneous," the author treats of several disconnected sub- 

 jects, such as cannibalism, head-taking, and castaways. 

 The concluding chapter is in some respects the best of 

 all. It consists of stories, divided into three groups — 

 animal stories, myths and tales of origins, and wonder 

 tales. These stories are not only pleasant to read, but 

 provide excellent materials for those who devote them- 

 selves to the comparative study of folk-tales. 



We may note that there are some very good illustra- 

 tions, especially in the chapter on the arts of life. 



Guide to Examinations in Physiography., and Answers /^ 

 Questions. By W. Jerome Harrison, F.G.S. (London • 

 Blackie and Son, 1891.) 



The author of this little work of forty-eight pages is 

 well known as a successful teacher, of wide experience in 

 connection with classes recognized by the Science and 

 Art Department. It is avowedly a guide to the art of 

 passing an examination, the author giving it as his 

 opinion that " knowledge of any subject is not the only 

 requisite to successfully passing an examination in it." 



NO. II 48, vol. 44] 



Unfortunately, this is, to a certain extent, true. Some 

 candidates are apt to make an injudicious choice of 

 questions, while others, again, spend too little time in 

 studying them, and consequently wander from the point. 

 Few who read Mr. Harrison's notes will fail to profit by 

 the sound advice which he gives. 



The first part gives general information about the 

 Science and Art Department and its objects, and 

 applies equally to all the subjects in which its exami- 

 nations are held. The questions which have been given 

 in the elemen y stage since 1882 are answered in 

 Part III. The appear to be sufficiently good to satisfy 

 the examiners. 



LETTERS TO THE EDITOR. 



[The Editor does not hold himself responsible for opinions ex- 

 pressed by his correspondents. Ntither can he undertake 

 to return, or to correspond with the writers of, rejected 

 manuscripts intended for this or any other part of ^ ArrVKt.. 

 No notice is taken of anonymous communications. '\ 



A Difficulty in Weismannism. 



Wkismann's theories of heredity and sexual reproduction 

 have been criticized from many a priori points of view. The 

 following remarks are an attempt to apply to his theory of repro- 

 duction a test familiar to llie mathematician ; and assuming its 

 truth, to follow out the deductions from this assumption. The 

 result is a startling one I believe the following theses will be 

 accepted as an impartial statement of the main points of the 

 theory : — 



I. Each primitive germ-cell, of either sex, contains a number 

 of ancestral germ-units, the Ahnenplasmas ; and this number is 

 constant, for the s[ ecies at least. 



II. These ancestral germ-units are far more constant and un- 

 changeable in character than the species itself. 



III. They lie associated together in the germ-cell without loss 

 or alteration of their individual personalities. 



IV. The number contained in the mature ovum and spermato- 

 zoon is reduced by one-half; and in the fertilized ovum or 

 oosperm the number is restored to the normal by the summation 

 of the Ahnenplasmas of the two fusing ceils. This process is 

 comparable to the shuffling of two packs of cards by taking half 

 from each and joining the talons or remainders to form a new 

 pack. 



V. The possible combinations under this process are so nume- 

 rous as to explain the variations among the oiTspring of sexual 

 union. 



Accepting these statements, we next inquire. How are we to 

 conceive of these ancestral units, the Ahnenplasmas ? Two 

 hypotheses may be given in answer to this question :— 



A. Each Ahnenplasma unit corresponds to an individual of 

 the species itself; and if put under proper trophic conditions 

 would, singly, reproduce such an individual. 



B. The Ahnenplasmas correspond to the primitive Proto- 

 zoan ancestors, which, according to theory, 'could alone reproduce 

 modifications due to external causes (acquired modifications). 



According to hypothesis A, the Ahnenplasmas of living 

 man are Anthropic ; those of our Simian forebears were Simian ; 

 and so we get Protochordate, and finally Protometazoan 

 Ahnenplasmas in the germ-cells of our more and more remote 

 ancestors. In other words, the Ahnenplasms have varied in- 

 definitely, and at the same rate with the race. This inference 

 not only renders the shuffling process unnecessary to explain 

 variation ; but it is inconsistent with thesis II., the very founda- 

 tion of Weismann's theory of heredity. 



According to hypothesis B, the Ahnenplasmas of all Metazoa 

 being similar and Protozoan, if the numbers arc equal and the 

 shuffling fair any two parents may beget any offspring what- 

 ever; on the plane of thesis V,, a lioness might be expected to 

 bring forth a lobster or a starfish or any other animal, which, 

 as we know, does not take place in Na'ure. The only escape 

 from this result is to assume the postulates— (i) that the 

 ' "Hereditary variability . . . can only arise in the lowest unicellular 



organisms ; and . . . necessarily passed over into the higher organisms 



■ V first appeared" (Weismanr, "On Heredity," English edition, 



p. 279). This passage would seem to render hyp-thesis B necess\ry for the 



when thev first appeared ' 



p. 279). ""■ " 



theory. 



