lo • president's address. 



another half-century to elapse before we are in possession of 

 these 3,000,000 proper motions. Again, the labour, and with it 

 the cost involved is enormous, and will probably be in the neigh- 

 bourhood of los. a star. 



The drawback to these two methods of obtaining proper mo- 

 tions is the necessity for defining the exact position of each star at 

 different epochs, whilst what we want is not its exact position, 

 which is difficult to define, l)ut its change of position — that is, its 

 proper motion. At the beginning of this century it had been 

 suggested that there was no necessity to measure the places of 

 all the stars on photographic plates, but that if pairs of plates 

 were examined in the stereoscope, those stars which had moved 

 relatively would stand out in relief ; alternatively, that if pairs of 

 plates were superimposed, those stars which had moved by 

 proper motion woulcl easily be picked out. These suggestions 

 were tried, and led to the discovery of a few proper motions, 

 but the method was not workable on a large scale, mainly because 

 of fatigue or strain upon the eyes. A third alternative was 

 discovered by Dr. Pulfrich, of Jena, and described by him as a 

 blink method. By this method the pair of plates to be examined 

 is placed side by side, like the ])ictures in a sterescope, but they 

 are examined with one eye through an optical and mechanical 

 arrangement which rapidly lets the eye rest first on one plate and 

 then t)n the other, so that in one second the eye has looked at 

 each plate separately three or four times. This blinking makes 

 the eye wonderfully sensitive to the slightest shift uj^on the 

 plates. If one star relatively to its neighbours has shifted a 

 hundredth of a millimetre U]:)on a Carte dn Cicl plate, the change 

 is not only unmistakeable. it is obtrusive.. This blink-method 

 revolutionises astronomy of position as regards the stars. Both 

 with the meridian observations and the Carte du Cicl measure- 

 ments, each star had to be dealt with separately. In the blink 

 method the stars are dealt with in groups Indeed, one can say 

 that it is easier to deal with 1,000 stars by the blink method than 

 with one by the other methods. All that the blink method re- 

 quires is pairs of plates separated by as long intervals as pos- 

 sil)le. A few weeks ago Mr. Hough ( H. AI. Astronomer at the 

 Cape) placed in Mr. Voute's and my hands a pair of plates with 

 a time interval of nearly twenty-three years. There were about 

 10,000 stars on the two plates. Ijut in a few hours we were able 

 to announce that only twenty of these showed proper motion — 

 the rest were fixed stars — and we were able to find the proper 

 motions of many stars which were so faint that even the great 

 Carte (In Cicl would not have included them. Since then further 

 pairs of Cape plates have been placed at my disposal with inter- 

 vals of sixteen to eighteen years ; the results confirm the earlier 

 experience. We can therefore clearly state that astronomers 

 have now a weapon of attack which will in the course of time 

 reveal to them, without arduous or expensive labour, the proper 

 motions of all classes of stars from the brightest to the faintest. 

 This will lead to a knowledge of the structure of the sidereal 



