Variables anfr jfatnt Stars 153 



that Capella is no near stellar neighbour that sends us 

 greeting from out the northern glow. 



VARIABLES AND FAINT STARS 



" The general conclusion with regard to all the 

 regions I have under observation is that there are 

 practically no variables among the very faint stars," 

 announced Mr. C. B. D'Esterre, of Tatsfield, Surrey, 

 in the spring of 1914. And he added the following 

 suggestive remarks : "It appears as though when we 

 move out towards the confines of our universe we 

 leave the region and perhaps the cause of variable 

 stars. I have now a very considerable amount of 

 evidence on this point. It is dangerous to announce 

 variability on the evidence of a few plates only. A 

 speck of dust, a pimple on the sensitised surface, or an 

 insect crawling across the plate, can so easily produce 

 the effect of irregular variability. I am not sure that 

 observers, especially in tropical regions, have always 

 taken this possibility sufficiently into account. I 

 have one splendid case where a star is almost com- 

 pletely obliterated on one plate, though showing quite 

 unchanged on the companion plate taken simultane- 

 ously. I feel almost sure that the result must be 

 traceable to insect interference." 



Mr. D'Esterre, in the course of a letter to me on the 

 subject, said that he conceived our stellar universe to 

 be built up on somewhat similar lines to the Andromeda 

 Nebula. Condensation had gone on much further 

 towards the central parts, and the stars forming the 

 more or less globular portion (including our sun) had 

 atmospheres much denser than those of the average 



