NA TURE 



[Flir.KUAKV 2, 1899 



Harvarh Astroi'uvsicai CoxrERENCE. — Several items 

 in Ihf report iif this conference have already lieen noticed, but 

 in (he pani|ihlet now distributed by I'rof. M. B. Snyder, of the 

 I'hiladelphia Observatory, there arc some points of interest. 



Mrs. Fleming presented a paper on stars of the \'th type 

 in the Magellanic clouds. These stars have spectra consisting 

 chiefly of bright lines, and are usually known as Wolf-Kayet 

 stars. Up to 1897 the number of these objects known was 67, 

 and all lie closely along the central line of ihe Milky Way. 

 Photographs taken at Arequipa with the Bruce telescope 

 revealed 21 of these objects in the large Magellanic cloud, 3 

 others in the Milky Way, and I in the small Magellanic cloud, 

 bringing the total number up to 92. Of these fifth-type stars 22 

 are thus in the Magellanic clouds, and this large proportion 

 renders probable the connection of these objects with the Milky 

 Way. 



Mr. E. S. King described an ingenious method of converting 

 prismatic spectra into normal spectra. To do this the original 

 plate is inclined to the plate on svhich the copy is to be made, 

 by an amount calculated to make the scale exact for three 

 points in its length, while at the same time maintaining good 

 focus. As illustrations he showed several stellar spectra 

 compared with Rowland's map of the solar spectrum. Prof. 

 Pickering said that the method was perfectly general, and all 

 scales were thus reproducible. 



Mrs. Fleming, in a paper on classification of spectra of vari- 

 ables of long period, describes the characteristics of these objects. 

 About 100 stars of this class are known, and all have the hydogen 

 lines bright. Examination of the relative intensities of these 

 lines has led to the class being divided into eleven groups, of 

 which f.vo are mentioned in detail. One group, represented by 

 R Lyncis, has a spectrum resembling a Tauri, but having H3 

 and 1 17 strongly bright, while HS is barely visible. Another 

 group, typified by R Leonis, shows a continuous spectrum 

 with 113 invisible, H7 barely visible, and 115 strongly marked. 



Prof. G E. Hale presented a review of work on the spectra 

 of stars of Secchi's fourth type. These are the blood-red 

 stars, and the spectra of twenty-two of them have been obtained 

 with the spectrograph of the 40inch Verkes telescope, mostly 

 with a dispersion of one prism. For the two brightest stars of 

 this class, 132 and 152 .Schjellerup, three'prisms were employed. 

 The photographs show a large number of lines hitherto unre- 

 corded, most of which are dark ; but it is stated that there seem 

 to be a number of bright lines. 



The presence of bright lines in 152 Schjellerup was, it is stated, 

 confirmed by Profs. Keeler and Campbell at the Lick Observatory. 

 The wave-lengths of two of the brightest of these lines agree very 

 closely with those of the two brightest lines in the WolfRayet 

 stars, as measured by Campbell ; but no connection between 

 these two classes of bodies is yet indicated. A reproduction of 

 the photographed spectrum of 152 Schj. is published in the 

 A si. Phys. Journ., November 189S. 



THE PLAN OF THE EARTH AND ITS 

 CAUSES} 



T N a passage in the "Novum Organum'' Bacon pointed out 

 resemblances between the continents of the Old and New 

 Worlds, which he thought showed that their shapes were not 

 due to chance, but to the action of a common cause. Similar 

 coincidences have been repeatedly noticed by geographers, who 

 have accordingly been led to the belief that the distribution of 

 land and water on the globe is based on a definite plan. Any 

 such plan can only be recognised in broad outline, since geo- 

 graphical shapes depend on an intricate series of local acci- 

 dents. ToiK)graphical form depends on such inconstant, in- 

 calculable factors that the stages of its growth are often un- 

 itraceable. The missing links of geographical evolution arc as 

 numerous as those of organic evolution. Nevertheless, belief 

 in the existence of a fundamental geographical plan is as old as 

 geography. It was expressed in some of the earliest classical 

 maps ; it possessed the minds of medi.rval cartographers, and 

 led to their fantastic wheel maps ; and it was popularised in 

 the first half of the present century by the teaching of Humboldt 

 and Elie de Beaumont. But with ihe growth of Lyellism and 

 its doctrine of the interchange of lan<l and sea under the in- 

 fluence of local variations in level, the idea fell into discredit. 



t Ab>lracl of lecture lo the Royal Gc0grapluc.1l Sociely on January 23, 

 l>v Dr I. W. tlrcRor). 



NO. 1527, VOL. 59] 



But the discoveries of oceanography and geology have shown 

 what allowances should be made for the obscuring action of 

 minor oscillations, and thus have revealed fresh geographical 

 homologies and explained apparent exceptions. The introduc- 

 tion of such local changes is, however; unnecessary, since the 

 existence of a geographical plan of the earth is shown by three 

 features : ( i ) the concentration of land in the northern, and of 

 water in the southern hemisphere; (2) the triangular form of 

 the geographical unit.s, and the southward tapering of the land 

 masses. From these two features it follows that there is a 

 "northern land bell" from which three continents project 

 southward, separated by three oceans, which expand until they 

 form a " southern ocean belt." The third feature is the anti- 

 podal position of oceans and continents. The main problem 

 of geomorphogeny is the explanation of these three facts. 



The question is simplified by remembering that the earth 

 consists of three parts : (l) the unknown internal centrosphere ; 

 (2) the rocky crust, or lithosphere ; I3) the oceanic layer, or 

 hydrosphere. Oceans and continents occupy, respectively, de- 

 pressions and elevations of the lithosphere ; and their distri- 

 bution, therefore, directly depends on the distribution of the 

 irregularities in its surface. If the existing irregularities have 

 remained unchanged throughout geological time, then the 

 problem is astronomical rather than geographical and geological. 



The attempts to solve the problem on the basis of the per- 

 manence of the main geographical features may be grouped 

 into four sets : (i) Prof. G. H. Darwin has attributed the main 

 geographical lines to tidal wrinkles in the viscous crust ; (2) 

 Prinz has assigned them to torsion due to the acceleration of 

 the equatorial and southern belts of the earth, and retardation 

 of the northern land belt; (3) Sir J. Lubbock and Prof. Lap- 

 worth have independently explained continental form as due 

 to the intercrossing of two sets of folds, one parallel lo the 

 equator and one at right angles to it ; (4) Lord Kelvin refers 

 back the main geographical divisions to an even earlier period 

 than the previous theories ; for he regards them as due to 

 shoaling in the last molten layer of the globe, over areas 

 determined by previous chemical segregations in the nebula. 

 These theories are not necessarily inconsistent with the asym- 

 metry between the northern and southern hemispheres ; the 

 primitive wrinkles, the double folds, and the nebulous segrega- 

 tions are each in harmony with a considerable amount of geo- 

 logical evidence; but the theories are geographically inadequate, 

 because they do not explain how the existing asymmetry has 

 been developed. To do this, some more continuously acting 

 cause is required, such as the secular contraction of the earth, 

 on which F.lie de Beaumont based his famous theory. His 

 system, however, regarded the world as symmetrical, and was 

 too rigidly gecmietrical to apply satisfactorily to a heterogeneous 

 globe. In his " jient.agonal reseau " antipodal areas were 

 similar ; for he regarded the world as a spheroid based on a 

 pentagonal dodecahedron, which is a holodedral form ; whereas, 

 owing to the dissimilarity of antipodes on the earth, the litho- 

 sphere may be better regarded>as hemihedral. The recognition 

 of this fact led to the great advance on Elie de Beaumont's 

 theory made by Lowlhian Green, which has been advocated by 

 de Lapparent and Michel Levy. 



The tetrahedral theory does not regard the world as a regular 

 tetrahedron ; but suggests that, owing to the collapse of the 

 earth's crust due to contraction, the lithosphere has undergone 

 a tetrahedroid deformation. That collapsing shells tend to be- 

 come tetrahedral, just as short tubes often collapse trigonally, 

 is shown by experiments. " Nothing," says E. D. Preston, 

 " is more in accordance with the action of physical laws than 

 that the earth is contracting in approximately a tetrahedral 

 form. . . . Experiments on iron tubes, on gas bubbles rising 

 in water, and on rubber balloons, all tend to bear out the as- 

 sumption that a homogeneous sphere tends to contract into a 

 tetrahedron." 



If such be the case, then the lithosphere would be depressed 

 on four faces, which, being lower, would be naturally occupied 

 by oceans ; while the four projecting coigns would stand up as 

 continents. In such a tetrahedrally deformed earth there would 

 be a central ocean (the Arctic) on one face surrounded by a 

 land belt, from which three meridional continents would run 

 southward ; they would each taper to a point, below which 

 would be a southern ocean belt, from which three oceans 

 would project northward. South of the ocean belt would be 

 an Antarctic continent. 



The arrangement of land and water on such an ideal, de- 



