9 6 



NA TURE 



[July 15, 1922 



and their distances, thus leading to our knowledge 

 of the distribution of stars in space, and finally to the 

 structure of the universe. 



At the present time, however, data are most 

 lacking regarding stellar radial velocities, and any 

 attempt to increase our knowledge in this respect 

 deserves particular attention. It is, therefore, very 

 satisfactory to know that, when planning the equip- 

 ment of the Dominion Astrophysical Observatory of 

 Victoria, B.C., the first consideration was given to 

 its suitability for this class of work, which it is 

 proposed to make the great feature of the new 

 institution. Quite recently the first volume of 

 measures has been published, and this comprises the 

 determination of the radial velocities of 594 stars 

 between the fifth and eighth magnitudes, the velocities 

 of which have not been measured before. 



The work was commenced in May 1918 by the 

 Director of the Observatory, Dr. J. S". Plaskett, and 

 Dr. R. K. Young. In the following year Mr. W. E. 

 Harper and Mr. H. H. Plaskett joined the staff, so 

 that this volume is the outcome of the endeavours 

 of these four observers, who together made the 

 observations, the measurements and the reductions 

 of the spectrograms, as no computing assistance was 

 available. 



The introduction to the volume describes briefly the 

 telescope, the spectrograph, the observing arrange- 

 ments, measuring machines, wave-lengths of spectral 

 lines employed, etc. The telescope was the large 

 72-inch reflector, and the spectroscope was used with 

 the Cassegrain combination of the telescope. The 

 regulation and control of the temperature in the 

 spectrograph was at first attained by the use of a 

 mercury contact thermometer actuating a special 

 relay, but this was replaced later by the installation 

 of a Callendar recorder. Considerable attention was 

 devoted to the wave-lengths of the lines employed 

 in the reduction of the spectrograms, and the system 

 eventually adopted was one developed by Dr. Young, 

 which is to some extent a compromise of different 

 methods and systems used by various observers. 

 Tables showing the standard wave-lengths employed 

 are given, and they indicate the wave-lengths used 

 for B-type stars and for A- to F-type stars, together 

 with a list of iron comparison lines. 



The work accomplished may be briefly summarised 

 by referring to the results of the measures which are 

 brought together in various tables. The first con- 

 tains the mean velocities, with their probable errors, 



of all stars, 537 in number, assigned to be of con- 

 stant velocity, including their positions, magnitudes, 

 spectral types, etc. Next follows a summary table 

 of the velocities of the systems of all spectroscopic 

 binaries, the orbits of which have been determined 

 at the Victoria Observatory. Table IV. gives com- 

 plete information as regards the details of all the 3287 

 radial velocity plates of the 537 constant velocity 

 stars, an average of 6-i plates per star. This informa- 

 tion is concise, compact, and in a convenient form, 

 and will be valuable for detail reference when required. 

 Similar detailed information is given in Table V. 

 with respect to 206 plates of 35 probably binary stars. 

 It is interesting to make a short reference to the 

 accuracy of the determinations of the radial velocities, 

 because this accuracy depends to a very great extent 

 on the definition of the lines in the star's spectrum 

 under examination. Thus, some spectra have clean- 

 cut lines which render their measurement easy, while 

 others exhibit fuzzy lines, making measurements 

 difficult. The authors have, therefore, divided the 

 probable errors into three classes. The first includes 

 stars of spectral types between Fo and M (excluding 

 some early F stars) which give the most trustworthy 

 values ; the probable errors for these range between 

 ±o-i and +i-o kms. for the mean velocity obtained 

 from all plates, and between +0-2 and ±2-5 kms. 

 for a single plate. The second main class includes 

 about one-fourth of the A-type, about two-thirds of 

 the B-type, and the early F-type mentioned above. 

 The probable errors for this group are given as 

 ranging from + 0-5 to ± 1 -5 kms. for the mean velocity, 

 and from +1-2 to ±3-5 kms. for a single plate. 



The last group embraces mostly A-type stars with 

 tin- addition of a few B- and O-type, the lines of 

 which are diffuse, broad, and frequently weak. For 

 these the probable errors range from +i-o to +3-8 

 kms. for the mean and from ±2-5 to +io-o kms. for 

 a single plate. 



The original programme as to the number of photo- 

 graphs of the spectrum of each star to be secured, 

 namely, six plates for each star with well-defined lines 

 J and eight or ten plates with poor lines, was almost 

 carried out ; the whole system of velocities here 

 deduced is therefore homogeneous and a high grade 

 of accuracy has been maintained. The radial velocity 

 values are the result of a great amount of painstaking 

 care, and their early publication is evidence of the 

 industry that has been displayed in all stages of the 

 research. 



Geology of Antarctic Lands. 



A 



USEFUL reference to recent summaries of the 

 geological features of Antarctica occurs in the 

 Proceedings of the first Pacific Conference, part iii. 

 p. 644 (1921). It is unfortunate that the various 

 researches based on the results of different British 

 expeditions have not been carried out in a common 

 clearing-house and published as an interlocking 

 series. At present three sets of quarto publications 

 are appearing in our libraries, two of them under the 

 auspices of the Royal Society of Edinburgh, and one 

 inn hi those of the Trustees of the British Museum. 

 Mr. 1 . M. Wordie's observations on the Weddell Sea 

 area (Shackleton expedition, 1914-17) have been 

 already noticed (Nature, vol. 109, p. 218). The 

 geological results of the expedition from the Falkland 

 Islands in 1913, financed by Messrs. Salvesen of 

 Leith, are now described by the leader, Mr. D. 

 Ferguson (" Geological Observations in the South 

 Shetlands, the Palmer Archipelago, and Graham 

 Land," Trans. Roy. Soc. Edin., vol. liii. p. 29, 1921). 



NO. 2750, VOL. I IO] 



The unrest in the earth's crust in Oligocene and 

 Miocene times is illustrated by great outpourings of 

 basalt in the South Shetland Islands. The later 

 lavas of the series are notably columnar, and are 

 correlated with similar rocks in Patagonia. Volcanic 

 activity continued almost down to recent times, and 

 there is a series of andesitic tuffs and lavas that go 

 back to Jurassic or early Cretaceous age. The photo- 

 graphic landscapes in this memoir are of unusual 

 excellence. The rocks collected abundantly by its 

 author are described by G. W. Tyrrell in a separate 

 memoir {ibid. p. 57). They include the varied intru- 

 sive masses of Graham Land and its group of islands, 

 and the red adamellite of Mount Theodore, "the most 

 imposing natural feature " of the district. Mr. Tyrrell 

 regards these older igneous rocks as distinctly Andean 

 in type. 



Dr. H. H. Thomas (ibid. p. 81) deals with the 

 rocks and minerals collected from islands of the same 

 region by Mr. Innes Wilson, of the Falkland Islands, in 



