i6o 



NATURE 



[August 4, 19 10 



devialion from the normal, whidi, taken over a 3-incli 

 range in the lirst-order sptrtium, only amouni!; to 

 01 A ipsirom unit. 



The important part played in solar and stellar spectro- 

 fropy by the H and K lines of caleium renders it essential 

 that the absolute wave-lengths of these lines should be 

 known with the greatest possible accuracy. For this 

 reason Mr. C. St. John, working at the Mount Wilson 

 Soiar Observatory, has recently made a series of wave- 

 length determinations for these lines in the arc, spark, and 

 electric furnace, and in No. 2, vol. xxxi., of the Astro- 

 physical journal he gives his results in terms of the 

 secondary standards of Fabry and Buisson adopted at the 

 Meudon meeting of the International Solar Union. The 

 mean results are 3968-476 and 3933-667 for H and K 

 respectively, and are estimated to be certain within 

 o-ooi Angstrom. Mr. St. John also discusses the behaviour 

 of these lines under the various conditions employed, and, 

 from his results, concludes that the wave-lengths are 

 identical for the absorption and the fine emission lines, and 

 are the same in arc, spark, and furnace. The mean ratio 

 of the width of K to H is 1-28, and the mean ratio of 

 the respective intensities is 1-47. The experiments 

 described were preliminary to an exhaustive comparative 

 study of the corresponding solar lines. 



PLANT DISTRIBUTION. 

 "T^WO recent papers furnish a supplement to the 

 ■*■ magnificent phytogeographical memoir on South Africa 

 by Dr. R. Marlotii. The one is an article, by Dr. L. 

 Diels, on formations and flora-elements in the north-west 

 of Cape Colony, published in Engler's Botanische Jahr- 

 biicher (vol. xliv., part i.). This is a detailed and localised 

 account of botanical observations made in the country 

 lying between the mouth of the Olifant River, Clanwilliam, 

 and Calvinia. Near Clanwilliam lie sandy stretches where 

 Com'positae and Scrophulariacea; provide the bulk of the 

 conspicuous vegetation. As the land rises, succulents, not- 

 ably species of Euphorbia and Crassulacea;, become pre- 

 dominant. At a height of 500 metres the vegetation begins 

 to show elements natural to the true Cape flora, culminating 

 in a " proteaceous-macchi " association on the Bokkeveld 

 ridge. A special object of the trip was the exploration of 

 the Hantam-bcrg flora, which is classed by the author with 

 the botanical formations associated with Namaqualand. 



The second paper is a contribution by Dr. H. H. W. 

 Pearson to the Royal Geographical Society, published in 

 the Geographical Journal (May), giving a general sketch 

 of a botanical expedition through the dry western districts 

 of- Cape Colony and the adjoining German territory to 

 Luderitzbuch, and thence from Mossamedes in Angola to 

 Ft. Rosadas on the Kunene River. The regions of vegeta- 

 tion through which Dr. Pearson travelled are very clearly 

 set out in the accompanying map. The succulent Karroo 

 vegetation was first travel sed until this gave place to a 

 composite flora near Calvinia. F'urther north, floras known 

 as the Namaqualand montane and Bushmanland were met 

 with. The former is characterised by the presence of Aloe 

 dichotoina, Vogclia ajricuna, and other plants, while species 

 of .-\ristida and I'arkinsonia africana are typical of the 

 latter type. The district lying immediately south of 

 Mossamedes is the historic locality in which Welwitschia 

 v;as discovered, and here the author found it more at home 

 than in Damaraland, which suggests that it is a tropical 

 species, and therefore more closely related to the genus 

 Gnctum than to Ephedra. 



Attention is frequently directed to weeds produced in 

 new countries by exotic plants. There is, however, more 

 interest attaching to the spread of indigenous plants caused 

 by a disturbance of natural conditions, of which a striking 

 instance in the case of Celniisia spcclahilis is described 

 by Dr. L. Cockayne in the Canterbury ,'\griculturists and 

 Planters Association's Journal (.April). This plant is a 

 composite and endemic, growing naturally w-ith other 

 species of the genus at elevations above 3000 feet. It has 

 a woody, creeping stem furnished with numerous cord-like 

 roots. The end of the stem bears a rosette of ttiick 

 tom.entose leaves with long sheathing bases; the rosettes 

 are crowded together, forming a circular mat or cushion. 

 ^^ ithin the shelter of the leaves lips tlie bud, -.vhich throws 



out dai.sy-like flowers above the leaves, and subsequently 

 develops downy fruits. As a result of burning and over- 

 gia.'.ing, the tussock formations at a lower level, which 

 consist of useful grasses, are being replaced by the 

 Celmisia. The remedy suggested is to reinstate natural 

 conditions, when the grasses should win back the lost 

 gi ound. 



Captain A. A. Dorrien-Smith contributes to a recent 

 number of the Keiv Bulletin (No. 4) an account of his 

 botanical excursions in Chatham Island with ■ the primary 

 object of collecting specimens of Olcaria semidentata, 

 Aciphylla Dieffenbachii, and other local plants for intro- 

 duction into the Scilly Isles. In the south and boggiest 

 pai't of the island Olearia semidentata covers acres of 

 ground, and here the author discovered a pure white form 

 of this normally purple daisy-like flower, and his com.- 

 panion found a pink variety. The article provides an 

 interesting sketch of the vegetation, and is illustrated with 

 several photographs, two of which represent bushes of the 

 normal type and white variety of the Olearia respectively. 



THE MAINTENANCE AND ADMINISTRATION 



OF ROADS. 

 A L THORITIES having control of highways have now 

 to consider the problem of road construction and 

 maintenance from a new point of view in consequence of 

 the conditions of modern traffic. The problem is an acute 

 one everywhere, and various solutions of it have been put 

 forward by highway engineers. Evidence of experts as to 

 the causes of the increased wear and tear of roads and 

 their opinions as to remedies will be found in the report 

 of the important conference on roads, held last year at the 

 Institution of Civil Engineers, and also in various publica- 

 tions of the Roads Improvement .'\ssociation and the Royal 

 Automobile Club. The subjoined extracts from these publi- 

 cations, and summaries of papers, provide the essence of 

 a large amount of evidence given by road engineers befot-e 

 several conferences and associations upon important ques- 

 tions relating to roads. 



A very large number of our roads, except those of recent 

 construction, may be said to have grown, or developed, 

 rather than to have been made. Many of them were 

 originally mere tracks, and have arrived at their present 

 state through the accretion of coats of ground-up stone, 

 often of poor character, possibly faced with a thin crust 

 of granite or some inferior material. The fact that many 

 roads have been built up by the use of metalling, without 

 foundations, other than the subsoil upon which the metal- 

 ling is placed, accounts for the difficulties, troubles, and 

 e.xpensive maintenance now experienced in connection with 

 most existing rural main roads, for where the foundation 

 of a road is weak, the surface is always difficult and costly 

 to maintain. 



The greatest practicable improvement in the construction 

 of macadamised roads is to be found in the use of the 

 very hardest and toughest coating materials well con- 

 solidated by rolling, with the addition of just suflicient fine 

 chippings during the consolidating process to fill completely 

 the spaces between the stones. The common method (con- 

 demned by every road engineer) of binding together the 

 aggregate of an ordinary macadam road by the use of 

 road scrapings is productive of the greater part of the mud 

 and dust found so objectionable. One of the resolutions 

 referring to macadamised roads adopted by the International 

 Road Congress held at Paris in 190S was : — " To use as 

 far as possible only hard and homogeneous road materials, 

 regularly broken ; to make choice of a binder suitable tn 

 the structure of the road material used, reducing, moreov- 1 

 the binder to a minimum." 



As to the wearing characters of various rocks used ;i^ 

 road metals, some definite information is available. The 

 Town Council of Hornsey possesses a machine by which 

 the effect of w^ear and tear on road stones can be tested. 

 The stones to be tested are all broken to a 2-inch gauge 

 and placed in cast-iron cylinders, which are made to re- 

 volve 8000 times at a speed of twenty revolutions a minute. 

 They are tested both wet and dry, and as the result of the 

 shaking they receive a certain amount of chips and dust 

 is produced. The percentage loss of weight experienced 

 bv the stones is then determined. As the treatment is the 



NO. 2127, VOL. 84] 



