August io, 191 1] 



NATURE 



i95 



set oscillating the block executes movements the ampli- 

 tudes and periods of which diminish in such a way that 

 resonance effects are prevented. By experimenting on a 

 movable platform, Galitzin found that the acceleration of 

 the enforced movement was the determining factor in the 

 overthrow of each block. 



Prof. Reid, of Baltimore, described a new method of 

 estimating the intensity of an earthquake, the fundamental 

 proposition being that the energy associated with an earth- 

 quake was proportional to the square of the area within 

 a given isoseismal line. 



Mr. Napier Denison, of Victoria, Vancouver, gave an 

 account of his observations of secular movements of the 

 horizontal pendulum, and made a strong claim for Victoria 

 as a place well fitted for seismological work. 



Prof. Omori, of Tokyo, described some of the recent 

 volcanic and seismic phenomena of Japan, the most 

 interesting being the rise of a new hill during the eruption 

 of Usu-san in Hokkaido, and the increasing activity of 

 Asama Yama, a volcano in the centre of Japan whicli 

 rises to a height of Sooo feet. The dull red contents of 

 the crater have been steadily rising in level for some time, 

 and fairly large blocks of stone have been projected from 

 it. A station has been built on the flank of the moun- 

 tain, and seismographs installed in it. The character of 

 the tremors is markedly different according as they do or 

 do not accompany a volcanic eruption. 



In their less strenuous moments the delegates and their 

 friends enjoyed to the full the hospitality of Manchester, 

 more especially the Lord Mayor's reception in the Town 

 Hall; Prof. Schuster's garden-party at Kent House, when 

 a remarkably good photograph was taken of all attending 

 the congress ; and the closing dinner given by the Uni- 

 versity Council, when Prince Galitzin, in a humorous 

 impromptu speech, proposed the health of Prof. John 

 Milne, whose characteristic reply was a fitting close to a 

 great conference. 



THE MIGRATION OF A RACE. 1 



'FHE theory of a relationship between the numerous 

 languages spoken in the islands of the Indo-Pacific 

 Ocean from Madagascar and Sumatra to the Philippines, 

 and thence far eastward to Melanesia and Polynesia, is 

 almost universally accepted. But the connection of these 

 languages with the Asiatic Continent, their origin, and 

 the means by which they reached their present settlements, 

 are still uncertain. 



In the volume before us Mr. Churchill essays to trace 

 (he migration of the Polynesian people from their first 

 home on the borders of Indonesia, through Melanesia, to 

 Nuclear Polynesia, that is, to the region round about 

 Samoa, Tonga, and Niue. Dealing with the languages 

 only, he recognises two streams of voyagers who have left 

 (races of their passage in the loan words adopted from 

 their speech by the Melanesians with whom they came in 

 contact ; these words being most numerous in the languages 

 of the islands along the coasts of which the Polynesians 

 passed, and less frequent in the languages more remote 

 from their route. 



One stream of these primitive Polynesians, or Proto- 

 Samoans, passed north of New Guinea, bv wav of the 

 Admiralty, Bismarck, Solomon, and Santa Cruz Archi- 

 pelagoes to Samoa. Another stream came southward 

 through Torres Straits, by the south-eastern shores of 

 Papua, through the New Hebrides to Fiji. In Nuclear 

 Polynesia the two streams resumed their ancient fellowship, 

 and thence despatched colonies to Hawaii, New Zealand, 

 and the Far Eastern Pacific. 



Later, there came upon these Proto-Samoans a swarm of 

 kindred people whose origin and migrations Mr. Churchill 

 regards as indefinite and obscure, and to whom he gives 

 the name — Tongafiti— by which they are known in Samoan 

 history. The Tongafiti are considered to have left no 

 definite trace of a passage through Melanesia, though 

 their presence in Nuclear Polynesia is clearly evident. The 



'The Polynesian Wanderings." Tracks of the Migration deduced from 

 .in Examination of the Proto-Samoan Content of Efate'and other Languages 

 •of Melanesia. By W. Churchill. Pp. ix + 516. (Washington : Carnegie 



NO. 2l8o, VOL. 87] 



origin and migrations of the Tongafiti are not discussed in 

 the present volume. 



In his earlier chapters the author devotes some attention 

 to the two most prominent theories yet put forth as to the 

 settlement of the Oceanic peoples, namely, those of Dr. 

 Macdonald and Dr. Thilenius. The former refers the 

 island races to an immigration from Arabia, and affirms 

 their languages to be modern representatives of a Semitic 

 tongue (cf. Nature, March 19, 1908, p. 460). The latter 

 regards the Polynesians as entrants into the Pacific by 

 way of the Micronesian Islands, and sees in the Melanesian 

 Islands, and the Polynesian settlements bordering them, the 

 meshes of a net which has caught the drift of castaways 

 blown westward from their homes in Eastern Polynesia. 

 In his second chapter and elsewhere in his book Mr. 

 Churchill utterly demolishes the Semitic theory of 

 Macdonald, mainly on the grounds of illogical and forced 

 etymologies, and perverted definitions of words. Mr. 

 Churchill finds difficulties and fallacies in the argument 

 of Thilenius, chiefly with regard to the ascription of feeble 

 navigating powers to the Polynesians, and to the unlikely 

 survival of castaways among an anthropophagous people. 



Mr. Churchill regards the Proto-Samoans as a seafaring 

 race, who, driven by some expulsive force, set out from 

 Indonesia in double canoes. The principal difficulty in 

 their navigation was the victualling of their vessels, and 

 this led to coasting voyages wherever there were coasts to 

 follow. When supplies ran short, a food colony was 

 established on a suitable island until a sufficient crop was 

 raised to carry the voyagers farther on. For these food 

 colonies there were three requisites — a sufficient water 

 supply, an encouraging area of soil for tilth, and an 

 autochthonous population insufficient in number, or too 

 weak, to prevent the settlement of strangers. Mr. 

 Churchill considers the eastward impulse to have ceased in 

 Bismarck Archipelago and Torres Straits, so that the crop 

 settlements tended to become, in suitable places, permanent 

 colonies. Also, as the fleet kept to windward in the seas 

 along which they passed, the lands which became fixed 

 settlements (as, e.g. Nuguria, Liueniua, Sikayana, Aniwa, 

 and Futuna) would be found on the windward side of the 

 archipelago with which they are associated. The apparent 

 exceptions, Rennel and Bellona. though leeward of the 

 Solomons, are, however, on the weather side of the 

 voyagers in the southern stream. 



Mr. Churchill bases his evidence of this migration upon 

 an examination of the Polynesian words contained in the 

 languages of the Solomon and New Hebrides islands. 

 From comparisons of the material available to him, which 

 are given in an appendix on data and notes, and the 

 Polynesian content therein, set forth in a series of elaborate 

 tables, the author deduces the amount of likeness shown 

 by the individual languages to Polynesia. His method in 

 the comparisons is sound and accurate, and a very welcome 

 contrast to the wild guesses of some writers on the subject. 

 But the deficiencies and imperfections of his material are a 

 source of serious error. For a proper estimate of Polynesian 

 likeness the vocabularies compared should be of equal size 

 and range of signification, else the presence or absence of 

 certain words would unduly exaggerate or diminish the 

 likeness. Mr. Churchill's tables show this. He gives the 

 coefficient of likeness to Polynesian of Belaga as 100. 

 whilst Nggela, of which Belaga is a dialect, is given only 

 So. So also the likeness of Nguna to Polynesian is 

 expressed by 93, whilst that of Sesake (the same language, 

 Sesake being a colony from Nguna) is only 76. 



Mr. Churchill makes no comparisons of grammar. All 

 the Polynesian words present in Melanesia are regarded as 

 loans, but if this is so, it is remarkable that many of the 

 languages have borrowed 80 or 90 per cent, of Polynesian 

 vocables without borrowing a single grammatical form. 

 In some cases, not pointed out by Mr. Churchill, gram- 

 matical forms which survive in Polynesian have been 

 preserved more fully in Melanesian languages which are 

 not on the supposed migration route. 



The backward track deduced by Mr. Churchill breaks off 

 short at Moanus (Admiralty Island) in the north, and at 

 Motu (New Guinea) in the south, and he leaves the 

 starting point of the Proto-Samoans in a waste of empty 

 sea, where non-Polynesian and non-Melanesian languages 

 occupy the whole seaboard. He states that " only a -few of 



