134 CARNEGIE INSTITUTION 



the annual parallactic displacements of all the nearer stars relative 

 to the bulk of the 600 stars under observation ? In course of time 

 it would then be necessary to take these parallaxes into account in 

 determining the apparent places of the stars affected by them. 



Class 6 evidently calls urgently for a large refractor of the most 

 perfect kind, used at a station where the definition is of the very 

 best. Possibly this work could be best carried on at the proposed 

 high level astrophysical observatory, although I may mention that 

 at Jamaica, in the end of 1882, I found the definition near the level 

 of the sea exceedingly good on many nights. In that island it would 

 be possible to find a perfectly salubrious station at an elevation of 

 4,000 to 5,000 feet, and quite probably equally favorable localities 

 could be found in some of the more mountainous South Sea islands. 



The general sidereal part of your scheme ought, as you say, to 

 be carried out south of latitude — 30 , but in the southern hemisphere 

 the climate increases in severity at relatively moderate distances 

 from the tropics much more rapidly than it does north of the equa- 

 tor ; hence every care should be taken to avoid a site too far to the 

 south. At the same time I most earnestly support your view that 

 no pains should be spared in choosing a thoroughly salubrious cli- 

 mate ; otherwise the most devoted members of the expedition will 

 be just the most likely to fall victims to any error of judgment in 

 this most important matter. 



In conclusion, I would suggest placing your most important 

 timekeeper in a partial vacuum — say, under nine tenths of the nor- 

 mal local atmospheric pressure. This is easily secured in a cast iron 

 box with a three-quarter-inch glass face resting on a slip of rubber. 

 A ' ' quicksilver sleeve ' ' permits of winding the clock twice in the 

 week. A small syringe removes any slight leakage of air. We 

 have found here that the air in the box must not be dried artifi- 

 cially, or the oil necessary to the clock work will decompose and 

 the clock will stop. 



A great improvement would be to add an outer case, the air in 

 which, by a simple electric contrivance, could be kept at a uniform 

 temperature slightly in excess of the highest temperature likely to 

 occur naturally. Under these conditions, any well made clock 

 ought to have a rate subject only to very minute changes. 



