112 



NATURE 



[September 30, 19 15 



by the other. The calculation of deflections by 

 "Mohr's method" is good. But the whole sub- 

 ject would have been much clearer if the well- 

 known principle had been explained, that the load, 

 shearing force, bending moment, slope and deflec- 

 tion curves form a series each derived from the 

 next by graphic integration. 



The phrasing is often loose. " For most metals 

 . , . the yield point is reached when there is a 

 sudden large increase in strain." Only in a few 

 metals is there a yield point in this sense. The 

 whole account of fatigue is unsatisfactory. The 

 author is in confusion between the old view of 

 Claxton Fidler, who must always be spoken of 

 with respect, who believed that the Wohler effect 

 was purely the result of dynamic action, and what 

 is now clear that there is a real " fatigue " due to 

 slip fissures developing at first slowly, but at last 

 rapidly and locally. It .is true that American en- 

 gineers have mixed up an allowance for fatigue 

 and for impact in the case of railway bridges. 

 But that is wrong in principle. Strictly, the 

 fatigue allowance is not one of the factors which 

 make up the factor of safety, but like the yield 

 point, the fatigue limit is a stress not to be ex- 

 ceeded when the total stress due to load, impact, 

 and contingencies has been calculated. To the 

 author " the subject is still full of difficulties " 

 (p. 89). At this late date it is odd to read "we 

 will make the following assumptions In developing 

 our theory of torsion" (p. 317). The use of align- 

 ment charts in designing springs seems new 

 (P- 342). 



PLANTS OF FORMOSA AND THE 



RIVIERA. 



(i) Icones Plantarum Formosanarum nee non et 



Contributiones ad Floram Formosanam. Vol. 



iv. By B. Hayata. Pp. vi + 264 + xxv plates. 



(Taihoku : Government of Formosa, 1914.) n.p. 

 {2) Flowering Plants of the Riviera. By H. 



Stuart Thompson. With an Introduction on 



Riviera Vegetation. By A. G. Tansley. Pp. 



xxix + 249. (London: Longmans, Green, and 



Co., 1914.) Price I05. 6d. net. 

 (i) ^T^HE fourth volume of Dr. Hayata's great 

 X work on the flora of Formosa, modestly 

 described as "materials for a flora of the island," 

 based on a study of the collections of the Botani- 

 cal Survey of the Government of Formosa, con- 

 tains descriptions of 285 species, of which no 

 fewer than 167 are new to science, including a 

 new orchid genus Arisanorchis of considerable 

 interest as adding another member to the small 

 list of leafless saprophytes in this family. More- 

 over, thirteen genera are mentioned as new to 

 the Formosan flora, these including such well- 

 NO. 2396, VOL. 96] 



known genera as Aconitum, Cornus, and Hydro- 

 charis. This volume brings the total number of 

 species of the flora up to more than 3000, and we 

 are promised several more volumes, indicating the 

 remarkable floristic richness of the island. 



In his interesting introduction the author sug- 

 gests two resolutions regarding the description 

 of new species of plants for consideration at the 

 meeting of the International Botanical Congress 

 fixed for the summer of 191 5 in London, "if the 

 present overwhelming war is then over." This 

 introduction is dated August 31, 1914. As usual 

 in Dr. Hayata's works, the illustrations are 

 numerous and beautifully executed, and the whole 

 reflects the greatest credit upon the author and 

 upon the enlightened Government of Japan. It is 

 worthy of note that the author has preferred to 

 follow the classification of Bentham and Hooker 

 rather than that of Engler and Prantl, and this 

 appears a wise choice, for although the former 

 system may be in various respects inferior as a 

 method of presenting the genetic affinities of the 

 families of plants, the latter is far from being a 

 natural system, and is constantly undergoing 

 modification. This being the case, adherence to 

 the older system, for purposes of ready comparison 

 between different floras, is perhaps on the whole 

 the best course to adopt. 



{2) The author's previous volumes on alpine and 

 subalpine plants fully demonstrated his facility for 

 popular yet accurate plant description, and he has 

 had long personal acquaintance with the region 

 here dealt with. The flora is arranged on the 

 Bentham and Hooker system, thus facilitating 

 comparison with French works on the flora of the 

 Riviera, in which the system used is practically 

 identical with the English one. The author has 

 spared no pains to make it easy for those with 

 even a slight knowledge of botanical terminology 

 to identify the plants described, for the actual 

 descriptive part is preceded by synopses of family 

 characters and keys to the tribes and genera, and 

 there is also a full and clear glossary of terms used. 

 As the subtitle ("A Descriptive Account of 1800 

 of the More Interesting Species ") indicates, 

 the book does not profess to be an exhaustive 

 "flora," but though the botanist may regret the 

 incompleteness which is particularly noticeable in 

 the case of some of the larger families and genera, 

 these cases nearly all refer to groups which the 

 non-professional botanist would find unattractive. 

 The author wisely refrains from the attempt, which 

 mars many books of this scope, to manufacture 

 so-called English names for species. 



The plates, both coloured and plain, are exceed- 

 ingly good, and a valuable feature of the book is 

 the interesting but disappointingly brief introduc- 

 tion by Mr. Tansley, which deals with the ecology 



