40 MOUNTAIN OBSERVATORIES IN AMERICA AND EUROPE. 



Californian mountains — on the liigh, dry plains of Arizona — in Mexico 

 — in the Andes — in Japan — or in which of these regions ? 



Wherever it is found, there is the place to install a large equatorial to 

 be devoted to observations of the most difficult and delicate nature — 

 to the settling of mooted questions — to discovery, in short. It is onh' 

 in such situations that a great telescope will do full justice to its con- 

 structor and will afford the fullest scope to an accomplished and diligent 

 observer. 



Professor W. H. Pickering has been kind enough (in a letter of April 

 8, 1895), to give me a direct comparison between the astronomical con- 

 ditions at Arequipa, Mt. Wilson, and Flagstaff, which he is better able 

 to make than anyone else : 



Regarding a comparison of the three observatories, Arequipa, Mt. "Wilson, 

 and Flagstatf , I think the sky was somewhat more transparent at the former, 

 since fainter stars could be seen in the horizon. On the other hand I doubt if the 

 difference was a practical one at altitudes over 30°. In fact, even at Cambridge. 

 I do not beUeve there is very much light lost on a really clear night. The great 

 advantage of the former stations, to my mind, was that the transparency was the 

 same night after night, for perhaps a month at a time, so that similar observations, 

 on comets for instance, could be conducted under identical conditions for con- 

 siderable periods. I think there is no doubt that the moon appears whiter in 

 Arequipa than in Cambridge, and the same remark applies, possibly in a less de- 

 gree, to the two other stations. My only knowledge of the steadiness at Mt. Wil- 

 son is derived from photographs, and the statement of Mr. Lowell that on the 

 night he looked through Professor Swift's telescope, which was said to be an 

 average night, the seeing seemed to him to be about as good as at Flagstaff. Upon 

 one or two occasions it seemed to me that the seeing at Flagstaff was as good as 

 anything I had seen at Arequipa, but the difference was that while at Arequipa 

 such seeing was common, at Flagstaff it was very rare.* 



Still the average at Flagstaff was probably better than the best we have at 

 Cambridge — it was certainly as good. My observations at Flagstaff lasted practi- 

 cally from June 1 to December 1. After the latter date there were many clouds 

 and the seeing was very bad. Regarding the number of nights on which photo- 

 graphs could be taken, I am inclined to think that there was little to choose be- 

 tween the three places ; perhaps 70 per cent of the nights were suitable. In Are- 

 quipa, however, the clouds were often very thin and of such a character that 

 while showing structure they were not accompanied by bad seeing. Therefore 

 u}X)n these nights visual observations could be conducted upon bright objects, like 

 the moon and the planets, and perhaps 80 to 85 per cent of the nights throughout 

 the year could be so utilized. 



One characteristic of Arequipa was that the seeing was excellent in the after- 

 noons beginning at perhaps 4 o'clock. In the early morning, i. e. after midnight, 

 the seeing at Arequipa was frequently bad. 



This was apparently due to a local cause — the formation of a cold stream of 

 air which followed down the river-bed from the interior. It is probable that at 

 a site located a few miles back from the river this current would be avoided. 



* In making this statement I allow for the difference in the apertures of the 

 Arequipa and Flagstaff telescopes. With the latter instrument one would neces- 

 sarily be more critical. 



