282 



NATURE 



[June i, 1916 



Dr. Taylor believes in the connection advocated 

 by Filchner between the Ross and Weddell Seas. 

 From the account of the researches by Dr. 

 Simpson it appears not improbable that the most 

 important of the scientific results of the expedition 

 will be the additions to Antarctic meteorology. 



The book is illustrated by numerous excellent 

 photographs, including- some by the expert 

 Antarctic photographer, Mr. Ponting, and also 

 many instructive and ingenious diagrammatic 

 sketches by Dr. Taylor. He publishes a photo- 

 graph of the Discovery Hut in which he lived 

 for a month, and the title directs attention to one 

 feature which shows that the hut was not built 

 as designed ; for it is raised on supports which 

 were only to have been used if the hut had to be 

 erected on ice into which they could have been 

 easily sunk. 



One interesting psychic incident is recorded. 

 .'Vt about the time when Amundsen turned back 

 from the South Pole his compatriot Gran had a 

 dream to that effect, and promptly recorded it in 

 one of Dr. Taylor's books. The author regards 

 it as a coincidence, but his remarks suggest that 

 he is not very confident of this explanation. 



PTOLEMY'S CATALOGUE OF STARS.-^ 



JUST forty years ago the late Prof. Peters, 

 of Clinton, New York, and Mr. Knobel 

 began independently, and without either of them 

 knowing of the other's work, to investigate the 

 Catalogue of Stars in Ptolemy's Almagest. 

 They soon, however, got into correspondence, and 

 eventually met in Paris in 1887. By that time 

 Peters had collated most of the manuscripts in 

 Continental libraries, and Mr. Knobel then under- 

 took to examine those in England. Peters died 

 in 1890, and in November, 1891, most of his 

 papers and notes bearing on the subject were 

 forwarded to Mr. Knobel, who completed the 

 work, and has now at last suceeded in getting it 

 printed. 



Only three editions of the Greek text of the 

 Almagest have been published, those of 

 Grynaeus (1538), Halma (1813-16), and Heiberg 

 {1898— 1903). A valuable German translation by 

 Manitius came out three years ago. Of the 

 Star Catalogue there have been several separate 

 editions, the best of which is that of Baily (Mem. 

 R. Astr. Soc, vol. xiii.). But from an astro- 

 nomer's p>oint of view no previous edition 

 can compare with the one we are considering 

 here, as this is founded on an examination of a 

 great number of codices — Greek, Latin, and 

 Arabic — and contains, besides, many other things 

 for which astronomers looked in vain in the 

 earlier editions. 



The investigation of Peters differs from those 

 hitherto made, as he began by calculating from 

 Piazzi's star-places and Madler-Bradley's proper 

 motions the longitudes and latitudes of all 

 Ptolemy's stars for the epoch a.d. 100, for the 



1 " Ptolemy's Catalogue of Srars. A Revision of the Alinagest." By Dr. 

 C. H F. Peters and E. B. Knobel. Pp. 207. (Washington : Carnegie 

 Institution, 1915.) 



NO. 2431, VOL. 97] 



purpose of identifying the stars and getting an* 

 idea of the accuracy of the positions. This was- 

 done before Auwers had published his new re- 

 duction of Bradley's observations, and it would 

 have been worth while to examine what difference 

 the adoption of Auwers 's proper motions would 

 have made, though the main results of the in- 

 vestigation W'Ould doubtless not have been 

 affected. The work also differs from all others 

 in the number of codices consulted. In all, 

 twenty-one Greek and eight Latin codices of the 

 Almagest were examined, and also three Arabic 

 codices of the Almagest, ten of Al Sufi's 

 Uranometry (the catalogue in which is that of 

 Ptolemy, with a constant correction for preces- 

 sion), and one of Nasir-ed-din Al Tusi's Com- 

 pendium of the Almagest. Detailed notes on the 

 first thirty-three of these codices and three photo- 

 graphic plates are appended ; the latter help tO' 

 make the reader understand the principal sources- 

 of error in the catalogue. 



The original catalogue was doubtless written 

 in the uncial Greek characters of the seconds 

 century, and the most common error in all manu- 

 scripts is that of confounding the uncial alpha 

 (=1) and delta ( = 4). Thus the magnitude of 

 d Eridani is given in all Almagests as i instead 

 of 4, which hitherto has puzzled everybody, while 

 the Bodleian Greek Almagest gives the magni- 

 tude of Sirius as 4. Errors are also caused by 

 confusion between A and A ( = 30) or 6 = 5 and 

 6 = g, etc. The Arabic MSS. are especially im- 

 portant for comparison with the Greek, as the 

 errors are of a different kind. Unlike the Greeks, 

 who wrote the minutes of longitude and latitude 

 as fractions of a degree, the Arabs wrote the 

 minutes in figures, and thus these two different 

 methods form a valuable check one on the other. 



The star-places finally adopted by the authors 

 are given in three catalogues. The first con- 

 tains for each star : Baily 's number, the number 

 and Latin description of the star from the Latin 

 edition printed in 1528, the Flamsteed number 

 and Bayer's letter, the longitude, latitude, and 

 magnitude. The second catalogue repeats the 

 last three items, and gives the longitude and' 

 latitude computed from Piazzi for the epoch 

 A.D. 100, and the difference between these and' 

 Ptolemy's values; also the magnitude from the 

 revised Harvard Photometry. The third cata- 

 logue gives Ptolemy's longitudes reduced by 

 2° 40', being the difference which Ptolemy states 

 he found between the longitudes of Hipparchus 

 and those of his own time, and the latitudes 

 unaltered ; also the positions computed for 

 130 B.C. After a lengthy set of notes on various 

 stars follow tables collating a number of codices 

 as regards longitude, latitude, and magnitude. 



Most writers have been of the opinion that 

 Ptolemy's catalogue was nothing but that of 

 Hipparchus, the longitudes being altered by 

 adding 2° 40' for precession. Peters had already 

 published in 1877 ^ paper showing that modern 

 star-places^ reduced to a.d. 100 and compared 

 with those of Ptolemy, gave a mean correction, 

 to his longitudes = +34*9', making his epochi 



