486 



NATURE 



[August 21, 1919 



On my classification we shall be able to study 

 the peculiarities and differences of the two 

 systems; and a valuable test is in this way 

 provided. 



These considerations are certainly fundamental 

 enough, and there are others. 



The similarity of the third curve to the first and 

 second justifies the return to some considerations 

 which I referred to in the fourth Bulletin regarding 

 the kinks in the curves. The descending arm 

 of the curve is much more continuous than the 

 ascending one ; the greatest change from the more 

 vertical to the flatter shape of the ascending arm 

 occurs at the Aldebarian and Crucian stages — 

 that is to say, the greatest number of stars at 

 nearly the same temperature occurs in those two 

 regions. It is suggested that this is due to the fact 

 that the stars involved reach their highest tem- 

 perature in these regions, so that we may assume 

 that not all stars first visible as Antarian reach 

 the highest temperature, but one set may reach 

 it near the Aldebarian stage, and another at the 

 Crucian stage, or rather between the Crucian and 

 Alnitamian stages, only a very small number of 

 stars reaching the Argonian stage. It is very 

 remarkable what a small percentage of stars 

 reach the Argonian stage. It is fair to assume 

 that the power of reaching these various stages 

 of temperature must depend on the initial equip- 

 ment of the swarm, and from this point of view 

 a close inquiry into the mass and density con- 

 ditions may be expected to help matters. 



In all that has gone before I have dealt with 

 a rise followed by a fall of temperature. I am 

 bound to say that for years after I put this view- 

 forward as the only one acceptable on the meteor- 

 itic hypothesis it was generally scouted. This 

 would not have mattered so much had the 

 Harvard classification, with its thousands upon 

 thousands of stars, not taken the other view of a 

 continued fall of temperature, as demanded by the 

 views formulated by Kant and Laplace. 



There have been many signs lately that the 

 opposition to my views is weakening; but the 

 more they are accepted, the more is it necessary 

 that a large number of stars should be added to 

 those I have classified. We want tens of thou- 

 sands of stars in homogeneous groups in order 

 that inquiries may be prosecuted with advantage. 



I showed in Fig. i of the fourth Bulletin that 

 the letters A, B, F; K of the Harvard classifica- 

 tion occurred in the spectra of stars located on 

 both sides of my temperature curve, and although 

 differences were indicated by sub-numbers, it is 

 a common practice to use the descriptive letters 

 alone, and it is difficult, therefore, to ensure 

 homogeneity. 



One of the great desiderata of the moment, 

 therefore, is to inquire whether something cannot 

 be done to render the stupendous and long- 

 continued work of classification carried out at 

 Harvard available under conditions which would 

 ensure the complete homogeneity of the stars 

 classed together. In order to study this question 

 1 have prepared tables which show the Harvard 

 NO. 2599, VOL. 103] 



classification of the stars included in the Hill 

 Observatory catalogues of 354 and 287 stars 

 (Plate i). I chose these catalogues because the 

 classification was carried on by the same three 

 observers and with the same instrument, and the 

 classification by each observer was carefully 

 checked by the others. The dispersion employed 

 between K and H^, 927 Angstrom units, is equal 

 to 28 mm. 



My hope was that the same sub-numbers of the 

 Harvard classification would not be found on both 

 sides of the temperature curve. 



In the comparisons I have previously made of 

 the Harvard classification and my own I have 

 indicated the Harvard classification of the stars 

 chosen as the type star in each of my groups, 

 but it will be seen from the present comparison 

 that the Harvard classification, in consequence of 

 the much greater detail which it attempts to 

 secure, does not justify us, as I hoped it would, 

 in giving a distinction between the letters and 

 their accompanying numerals used on both sides 

 of the curve. 



But this difficulty is not common to all parts of 

 the curve. Near the top, at the Crucian and 

 Achernian stages, the greatest number of stars in 

 which, on both sides, are classified B3, it is not 

 of the highest importance to draw the distinction. 

 In the case of the Sirian and Cygnian stars, where 

 it is imperative that a complete separation should 

 be chosen, the majority of stars in both are classi- 

 fied in A, with the exception of two classified as 

 F, which probably may be due to misprints. But 

 when we come to the difference between the 

 Polarian and Procyonian and the Aldebarian and 

 Arcturian, it will be seen that the attempt is 

 hopeless. Twenty-two Aldebarian stars are classi- 

 fied as K, and forty-two Arcturian stars also 

 classified as K. Norman Lockyer. 



Hill Observatory, 



August ,21, 1919. 



THE SUPPLY OF IXRUGS DURING THE 



WAR. 

 '1117' HEN war broke out, the National Health 

 » * Insurance Commission was charged by the 

 Government with the duty of safeguarding the 

 position of this country with regard to the supply 

 of drugs, and the Commissioners have just issued 

 a memorandum ^ describing the work done in this 

 connection. The work fell mainly into two 

 categories, viz. (i) conservation of existing 

 suppHes by such means as restriction of exports 

 and the most economical use of the materials 

 available, and (2) encouragement of home pro- 

 duction of fine chemicals used in medicine. The 

 second is, of course, much the more interesting, 

 and some of the results of this work were illus- 

 trated in the exhibits shown by various fine 

 chemical manufactureres at the recent British 

 Scientific Products Exhibition. Certain manu- 

 facturers took up on their own initiative the pro- 



1 Memorandum on the Special Measures Taken by the National Health 

 Insurance Commission (England) in Relation to the Supply of Drugs and 

 her Medical Stores duringthe War. Cd. 183. (1919 ) 



