428 



NA TURE 



[August 31, 1905 



notice. The classification of the ground-sloths is 

 much more complicated than the one adopted by older 

 writers, the Megalotheriidae being- now split up into 

 a number of family groups. Very noteworthy is the 

 inclusion among the Monotremata of an extinct South 

 American family, the Dideilotheriidae, with four 

 generic modifications, as if this be justified it has a 

 most important bearing on former land connections 

 between the southern continents. We confess, how- 

 ever, to a certain amount of hesitation in accepting 

 the determination of these South .'\merican fossils 

 until it has been confirmed by a pateontologist of un- 

 questioned authority. In retaining provisionally the 

 South African Tritylodon among the mammalia, Dr. 

 Trouessart is in accord with opinions latelv e.xpressed 

 by Dr. R. Broom. ' R. L. 



How to Know Wild Fruits: a Guide to Plants when 

 not in Flower by Means of Fruit and Leaf. By 

 Maude Gridley Peterson. Pp. xliii + 340; illus- 

 trated. (New- York : The Macmillan Co. ; London : 

 Macmillan and Co., Ltd.) Price 6s. 6d. net. 

 " Ye shall know them by their fruits " might well 

 have served as the fore-word to this little volume. It 

 deals only " with those plants which bear attractively 

 coloured fruits," and might, therefore, be classed by 

 the reviewer among that very large class of books 

 which are made to look at rather than for any more 

 serious purpose. The very first chapter, on " Adapta- 

 tions of Fruits and Seeds for Dispersal and Protec- 

 tion," serves to dispel that notion. It consists only 

 of some half-dozen pages, but those pages are instruc- 

 tive, and, better still, suggestive. Then comes a list 

 of " definitions," few in number, but adequate to a 

 book of these pretensions, especially as it is supple- 

 mented by a glossary at the end. " A Guide to the 

 Plant Families Represented " comes next in order, 

 and consists of an analytical table by means of which 

 the several families may be discriminated by the 

 observation of the variations in the character of their 

 fruits. This seems to be carefully compiled, and is, 

 so far as we have seen, accurate, but its value can 

 only be tested by actual use in the field. 



In the second table the families and species are 

 grouped according to the colour of their fruits. Thus 

 the monocotyledonous families are arranged accord- 

 ing as the colour of the fruits is red, reddish-purple, 

 green, black, or dark-purple, or blue. Of course, 

 this is a highly artificial mode of grouping and one 

 subject to exception, but if these circumstances be 

 borne in mind the table will be found useful. 



Coming now to the individual plants, which are all 

 North American, the author gives a pretty full de- 

 scription of each, beginning with the fruit and pass- 

 ing on to the foliage and flowers. These descrip- 

 tions might have been materially abridged and com- 

 parison rendered easier by the omission of unneces- 

 sary particles and verbs. In this matter the example 

 of the author's fellow-countryman, Asa Gray, might 

 have been followed. Moreover, thev are not alwavs 

 botanically accurate; the " fruit" of the yew, for in- 

 stance, is only remotely " drupe-like," and is cer- 

 tainly not a " drupe," as it is said to be in the same 

 paragraph. Conversely, the leaves of the yew are 

 really spirally arranged, but appear to be disposed in 

 two planes only. 



It would obviously be unfair to treat this book as 

 if it were intended as a botanical text-book, but as a 

 help to the beginner and a means of stimulating ob- 

 servation it may be commended. It is well got up, 

 remarkably free from misprints, appropriately illus- 

 trated, and provided with an index of vernacular 

 names and one of the Latin designations of the plants 

 described. 



NO. 1870, VOL. 72] 



LETTERS TO THE EQJTOR. 



yriie Editor does not hold himself responsible for opinions 

 expressed by his correspondents. Neither can he undertake 

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

 manuscripts intended for this or any other part of Nature. 

 No notice is tal^en of anonymous conimunicatiofis.] 



The Kangra Earthquake of April 4, 1905. 



There have been certain papers on Indian geological 

 questions recently published in the Neues Jalirbuch and 

 associated Centralblatt fiir Mineralogie, Geol. und Pal. 

 (Stuttgart), in which either the data or the deductions, 

 generally both, have been unsound. In most cases the 

 authorship alone has been sufficient to enable us to 

 separate those papers that are worth careful study from 

 those that are not even worth the time necessary to 

 read. But in the latest production one of the editors of 

 the journal appears as a joint-author, and one wonders 

 consequently whether the papers we have been discarding 

 in India as untrustworthy are, after all, normal or 

 accessory constituents of a periodical which all geologists- 

 have regarded hitherto as essential to a working library. 



The paper I now refer to appeared in the Centralblatt 

 No. 11 (June), entitled " Das Erdbeben im Kangra-Tal 

 (Himalaya) von 4 April 1905," by E. Koken and 

 F. Noetling. The authors take eight pages of text and 

 a map to demonstrate the unusual features of scientific 

 interest shown by the recent earthquake — the time of its- 

 occurrence, the remarkable variation in the rate of trans- 

 mission of the earthquake waves in different directions, 

 the peculiar distribution of the isoseismal lines, and the 

 exceptional shape of the meizoseismal area. The whole of 

 this " scientific " discussion is built on a few newspaper 

 cuttings, in the collection of which the authors have not 

 been sufficiently industrious to escape certain tell-tale mis- 

 prints which appeared only in the newspapers of the 

 Presidency farthest removed from the earthquake centre. 

 One example will be sufficient to illustrate the care 

 exercised in collecting and checking their data. 



The authors on p. 336 refer to a town named Tagarmalli 

 as only very slightly damaged, and they consequently 

 adopt this point, which they determine to be fifty miles 

 from the epicentre, as the maximum extension eastwards 

 of the meizoseismal area. As a matter of fact, no such 

 place as Tagarmalli exists, and no such name appears in 

 any of the gazetteers of India ; the most casual attempt 

 at verification would have shown the authors that they 

 were basing their elaborate deductions on a misprint 

 which appeared in one newspaper only. In one of the 

 Lahore papers the names of the two places Nagar and 

 Manali, twelve miles distant from one another in the KuUi 

 valley, became contracted by the printer's devil to 

 Nagarmalli, and in this form it was telegraphed to Bombay 

 (Times of India, April 14) and to Calcutta (Englishman, 

 .'Kpril 14) ; but by the accidental omission of a single 

 Morse's dot the word reached Madras as Tagarmalli 

 (Madras Mail, April 15, and telegraphic summary, April 14). 

 Having found the clue to the authors' source of data, 

 we find it easy to explain other equally remarkable state- 

 ments in the paper. In an earlier part of their paper 

 (P- 334) 'h^y refer to the complete destruction of the 

 place Nagar (Naggar), without suspecting that it was 

 one of the roots of their mythical Tagarmalli ; but on this 

 occasion they have removed the little capital town of 

 Kulu, and, for purposes of seismological reasoning, have 

 carried it over the snowy range into the ICangra valley. 

 To base a series of scientific deductions on a few news- 

 paper cuttings may satisfy the devotee of precision (alias 

 accuracy) in Germany, but to neglect the simple means 

 of verifying their facts provided by the splendid maps 

 of the Punjab, the complete gazetteers, or even the four- 

 penny postal guide obtainable in nearly every village, 

 shows a carelessness that deserves the contempt of every 

 scientific man. 



But, after all, it is not the basis of data so much 

 that is at fault, though even the purchase of a few more 

 newspapers would have saved the authors from most of 

 their pitfalls; it is the "scientific" superstructure that 

 is so discreditable. When the authors noticed that the 

 earthquake was recorded by the Bombay seismograph at 



