2 42 



NA TURE 



[January i6, 1908 



var. ro]i, Watt. It can safely be said that the last 

 varietal name is superfluous, being synonymous with 

 G. obtusifolium, Roxb., as interpreted in this book. 



As examples of cases in which a plant is wrongly 

 assigned to an old-established species, the following 

 cases may be cited : — 



(i) G. penivianum, Cav., is stated (p. 217) to have 

 /uzz-coated seeds, though Cavanilles states that the 

 seeds are black, i.e. devoid of fuzz, and figures them 

 so. This discrepancy would have been immaterial 

 had not the presence or absence of fuzz on the seed 

 been made the basis of classification (see later). This 

 plant is stated to produce the Imbabura cotton of 

 Peru, while the Piura cotton is stated to be produced 

 by a somewhat similar but hairy plant, which is 

 identified as G. vitifoUiim, Lamk. It suffices to say 

 first that Piura cotton is produced by the plant 

 Cavanilles called penivianum, which has, as that 

 author states, naked seeds ; and secondly, that 

 Lamarck states that his species vitifolium has the 

 under sides of its leaves glabrous. 



(2) Of G. microcarpitm, Tod., its author mentions 

 specially that the two lobes on either side of the 

 central one were unequal, and gives a good figure of 

 this. Our present author not only gives (plate 36) 

 a plant with much broader lobes, but one which does 

 not display the peculiarity of lobes mentioned, is 

 called " red Peruvian cotton," and differs from 

 ordinary Peruvian only in bearing brown cotton in- 

 stead of white. The reviewer has grown the real 

 G. tnicrocarpum, which is an exceedingly character- 

 istic variety, and can be distinguished with certainty 

 at a glance. It is, indeed, the plant for which our 

 author has created a new species, viz., G. Schottii, 

 two specimens cited as types in the British Museum 

 being exactly the plant as figured and described by 

 Tcdaro, and as grown bv the present writer. 

 Examples of this kind might be multiplied almost 

 indefinitely. 



Turning to the system on which the varieties are 

 classified in the volume under review, we find it is as 

 follows : — 



Section i.. Species with fuzz but no floss. 



Section ii., Fuzzy seeded cottons with united brac- 

 teoles. 



Section iii,, Fuzzy seeded cottons with free brac- 

 teoles. 



Section iv.. Naked seeded cottons with bracteoles 

 free, or ne.irly so, and glands conspicuous. 



Section v.. Naked seeded cottons with bracteoles 

 quite free and floral glands absent. 



It will be noted that the presence or otherwise of 

 a fuzzy covering to the seed below the cotton is made 

 the primary basis of classification. Now it is hardly 

 too much to say that every cultivated species of cotton 

 comprises varieties some of which bear a fuzzy and 

 others do not. The present writer has found among 

 others completely naked seeded varieties in the species 

 (to adopt the nomenclature of our author) G. 

 iiniikiiii; (Chinese and Japanese cottons), G. nanking, 

 Var. roji., G. obtusifolium, var. Wightiana, G. 

 herbaccum, G. punctatum, G. hirsutum. The naked- 

 seeded varieties show not the slightest trace of 

 hybridisation with a member of sections iv. or v., 

 NO. 1994, VOL. 77] 



and, indeed, a hybrid between any of the first four 

 varieties named and a member of sections iv. and v. 

 is by no means readily produced even by artificial 

 means. Yet our author seeks to explain the occur- 

 rence of naked seeds in the " jowari hathi " 

 ( = country cotton) of Madras by the supposition of 

 a naturally produced cross between G. obtusifolium. 

 var. Wightiana (section ii.), and Bourbon cotton (G. 

 purpurascens, section iv.). Similarly, those of Ameri- 

 can upland varieties that have naked seeds are said to 

 be crosses with a naked-seeded variety for this reason 

 alone. 



If any further proof of the fallacy of this method is 

 required, it is found in the fact that fuzzy-seeded 

 American has in India been converted into a naked- 

 seeded variety in a few generations by the present 

 writer through the simple process of growing it under 

 irrigation in well-manured soil. That such a change 

 occurs is well known to cultivators in the West 

 Indies and other parts of the world. 



If we take the second point on which the classifi- 

 cation is based, viz. whether the bracteoles are free 

 or united, we find the same impossibility of applying 

 the characteristic in the field, in some varieties there 

 being found, on tlic same plant, flowers with the 

 bracteoles all free, others with them all united, and 

 still others with two united and one free. 



The last chapter (thirty-one pages) is devoted to a 

 discussion of " The Improvement of the Cotton 

 Plant." This consists merely of a general description 

 of the process of selection equally applicable to al! 

 crops, an attempt to trace the history of some varieties 

 now grown, and a description of the pollen grains of 

 some species of cotton. 



Throughout the book no attempt is made to give 

 the character of the cotton produced under given con- 

 ditions of soil and climate by the several varieties de- 

 scribed, though the author hopes in his preface that 

 the book will be useful to " planters and seed pro- 

 ducers throughout the world." 



F. Fletcher. 



A CONCISE WORK ON EVOLUTION. 



Evolution and Animal Life. An Elementary Discus- 

 sion of Facts, Processes, Laws and Theories 

 relating to the Life and Evolution of Animals. By 

 David Starr Jordan and Vernon Lyman Kellogg. 

 Pp. xi + 489; illustrated. (New York: D. Apple- 

 ton and Company, 1907.) Price 2.50 dollars net. 

 THERE is growing up a generation of biological 

 students that does not read its Darwin, its 

 Weismann, or Galton ; instead, it cons manuals and 

 text-books on the works of these masters. It is so 

 very much less trouble, if the student's object is to 

 satisfy an examiner, to " get up " a text-book on 

 evolutionary problems than to extract from original 

 sources a clear conception of the authors' theories; 

 and yet what a world of difference is there between 

 thT ipsissima verba of a master and the cut-and-dried 

 phrases of the manual-maker ! The one is the advo- 

 cate pleading his cause with all the eloquence in his 

 power, the other the reporter compressing the living 

 Words and phrases into the limits of a column. The 



