KNOWLl^DGK. 



109 



part from the recent flagellates and in part from the red algae. 

 The early forms were marine, but after taking possession of 

 the upper waters of the sea and invading estuaries, they 

 became adapted to life inland in fresh water. 



Whether or not this ingenious theory of Brunnthaler's gains 

 much acceptance among botanists, it must be admitted that it 

 has the merit, conspicuously absent in various other specula- 

 tions on the evolution of plants, of taking into account the 

 probable conditions of life in ancient times. 



EVOLUTION OF THE FLOWER.— In the New 

 Phytologist (1911. Nos. 5 to 8) there are two further 

 instalments of the interesting paper on " Floral Evolution, 

 with particular reference to the Sympetalous Dicotyledons," 

 by H. F. Wernham. The first two articles in this series have 

 alreadv been summarised in " Knowi.kdgi: " (1911, pages 

 231, 277). 



Engler's group Pentacyclidae includes the cohorts or orders 

 Ericales, Primulales, and Ebenales. .After discussing the 

 general characters of the Ericales (including the large family, 

 or natural order, Ericaceae, and five other smaller families), 

 we find the somewhat startling suggestion that the Bilberry 

 tribe iVaccinioideae) should be removed from the Heath 

 family altogether. It is true that the Vacciniiiiii tribe 

 differs from the rest of the Ericaceae in having an inferior 

 ovary, but in modern systems of classification (he position of 

 the ovary is regarded as being of relatively small importance 

 in deciding affinities, and it appears very unlikely that 

 botanists will be inclined to follow Wernham in making even 

 a separate family of Ericales for the Vacciniitm tribe, 

 much less in regarding them as having had an entirely 

 different origin. It is suggested that while the remaining 

 Ericales probably arose from the ancestral Kanalian 

 (Buttercnp-likel stock along the line which led to the 

 Geraniales, the \'accinioideae had their origin from the line 

 which led from the same ancestry to the Rosales. In a 

 course of lectures on Systematic Botany which has recently 

 been given in the University of London, and of which it is 

 hoped to include a summary in " Knowledge " shortly, Dr. 

 C. E. .Moss dissented from Wernham's theory regarding the 

 position of the \'accinioideae, while agreeing warmly with his 

 views on the evolution of the Gamopetalae in general, and 

 pointed out that if the Vaccinioideae are to be severed from 

 the Ericaceae, the Arbutus tribe (Arbutoideae) must 

 necessarily go along with them. Since the two sections 

 (Vaccinioideae and Arbutoideae) agree in all essential 

 characters, excepting that the ovary is superior in Arbutoideae. 

 and the .'Vrbutoideae are admittedly related very closely to the 

 remaining sections (Ericoideae and Khododendroideac) of the 

 family Ericaceae, the result of the adoption of Wernham's 

 revolutionary proposal would be an unnatural and, in fact. 

 chaotic arrangement. 



.■\part from this criticism, one can have nothing but praise 

 for the careful and brilliant working out of Wernham's views 

 regarding the origin of the various groups of Gamopetalae 

 (Sympetalae) from different groups of lower Dicotyledons 

 (.■Archichlamydeae or " Polypetalae "I. He. of course, adopts 

 the theory that the Primulales have descended from the large 

 group Centrospermae. which includes such well-known 

 families as the Chenopodiaceae and Caryophyllaceae, and 

 gives a '■ family tree '' illustrating the details of the descent 

 according to his views. Here we feel on much safer ground, 

 though further work on the development of the flower in the 

 various families is. naturally, required before the detailed 

 course of evolution can be traced. 



The Ebenales are less familiar to the British botanist than 

 the two preceding cohorts, since this group consists almost 

 entirely of tropical trees. The view is put forward, and 

 worked out in some detail, that the Ebenales have arisen from 

 the Parietales. hence the three cohorts dealt with are regarded 

 as having arisen independently " by the grafting of sympetaly 

 upon at least three archichlkmydeous stocks — the geranial, 

 the centrospermal, and the parietalian." 



In his fourth paper, the author begins upon the large series 

 of cohorts grouped under the name '" Tetracyclidae." in which 



the floral organs (sepals, petals, stamens, carpels) are each con- 

 fined to a single whorl, as compared with the Pentacyclidae in 

 which there are two whorls of stamens and therefore five 

 whorls in all of floral p.arts. The Contortae comprise the 

 families Asclepiadaceae, Apocynaceae, Oleaceae, Gentian- 

 aceae. and so on, few of which occur in Britain. The ancestry 

 and affinities of the Contortae present great difliculties — the 

 floral structure is remarkably constant throughout the cohort 

 and is of a relatively advanced character — and the Contortae 

 " have left no traces of their progress from polypetaly to 

 sympetaly in the shape of pentacyclic forms ; neither a second 

 staminal whorl nor any hint of it ever occurs." Still, it is 

 suggested that the group may have arisen from the Geranial 

 stock, like the Ericales. " While the latter have employed 

 their evolutionary powers, so to express it, in the direction of 

 specialization of habit and details of floral structure, the 

 Contortae have reserved their efforts for the realization of the 

 economy tendency." In tracing the detailed affinities of the 

 Contortae, Wernham points out the isolated character of the 

 two orders Oleaceae and SaKadoraceae, which agree with 

 each other and differ from the remaining famiHes of Contortae 

 in so many striking respects that the author is led to regard 

 them as deserving the rank of a special cohort, for which he 

 proposes the name Jasniinales. 



cm-.Mi.srRv, 



By C. Ai.NSWORTii Mitchell. B.A. (Oxon.), F.I.C. 



CANADIUM: A NEW ELEMENT.— .\n account of a 

 supposed new element, to which the name of " canadium " has 

 been assigned, is given by Mr. .\. G. French in a recent issue 

 of the Chemical News (1911. CIV, 283). It was discovered 

 in association with platinum and other metals of the platinum 

 group, in the trap-dyke of the Nelson District of British 

 Columbia, its quantity ranging from a few grains to about 

 three ounces per ton of its companions. 



The new metal, which is found in the form of grains or 

 scales, is white and possesses a brilliant lustre. It melts at 

 about the same temperature (964° C.) as silver, and is not so 

 hard as platinum, ruthenium, palladium or osmium. Like 

 platinum, it is not altered by being heated in the air, or by 

 exposure to a moist atmosphere. It is not affected by iodine, 

 hydrogen sulphide or solutions of sulphides, but is dissolved 

 by nitric or hydrochloric acid. Unlike silver, it is not 

 precipitated from its solutions by alkali chlorides or iodides. 



PHOTOGRAPHIC EFFECT OF CHEMICAL RE- 

 ACTIONS. — .A discovery, which appears likely to have far- 

 reaching effects, both in theory and practice, is described by 

 Messrs. Matuschek and Neiming in the Clieiniker Zeituiig 

 (1912, X.XXVI. 21). It has been found that chemical 

 reactions of the most different kinds produce light waves, or 

 that part of the heat energy developed during the re-action is 

 transformed into light energy. 



In the striking series of experiments described, a beaker, 

 to the bottom of which was fixed a star of tin-foil, was placed 

 upon a sensitive plate in the dark, while the interacting sub- 

 stances were placed within the beaker. .After several hours 

 the plate was developed, and in each case a sharp image of 

 the star was obtained. F"or instance, strips of various 

 metals, such as zinc, copper, tin and lead, were suspended in 

 sulphuric, hydrochloric or nitric acids, and notes were taken 

 of the periods of exposure necessary to obtain sharp outlines 

 of the star. 



The distance of the reacting bodies from the sensitive plate 

 and the nature of the surface of the metal had an influence 

 upon the results : and the smaller the chemical affinity of the 

 acid for the particular metal, the less the intensity of the 

 photographic action, and the longer the time required. Thus, 

 in the case of the metals and acids mentioned above, four 

 days were necessary to obtain a sharp image in the reaction 

 between lead and nitric acid. The treatment of copper oxide 

 or hydroxide with acids, or of caustic alkali with water, gave 

 analogous results. A thin layer of quicklime or of calcium 

 carbide when spread on a glass plate with a device of tin-foil 



