FUNDAMENTAL CONCEPTIONS 361 



extremely unlike the earth in structure and temperature, present 

 a most serious obstacle to the application of correct principles of 

 reasoning. 



Notwithstanding the strangeness of the phenomena with which 

 they have had to deal, it is worthy of remark that practical investi- 

 gators in astronomy have indulged sparingly in speculations about 

 things which cannot be firmly apprehended through our perceptions, 

 or which have no adequate analogies in human experience. They 

 have shown little inclination to dwell on such questions as that of the 

 probable limits of the universe; whether the planets are inhabited; 

 what gravitation really is; whether there are burned-out suns in space 

 and few or many of them; what is implied in the motions of the stars, 

 as to the past and future of the visible universe; and many other 

 points of a similar nature. These are matters of high philosophic 

 interest and legitimate subjects for speculation; but they are not yet 

 fully ripe for treatment according to the logical methods of scientific 

 research; and therefore the astronomer usually refers to them in his 

 private, rather than in his professional capacity. 



The charm of astronomical work seems partly to reside in the fact 

 that the senses, aided or unaided by mechanical appliances, are not 

 directly sufficient to provide needful material of investigation without 

 important aid in the art of interpretation. That the hypotheses and 

 theories to be inferred from observations are not immediately sug- 

 gested by customary modes of thought is another attractive feature 

 of the work. 



In the astronomy of motion the question at issue is not always 

 whether a certain thing can be seen, or not seen; it more often turns 

 on probability of evidence. A small planet appears exactly like the 

 stars and is primarily distinguished from them only by its motion. 

 The object suspected to be a planet either distinctly appears to move 

 relatively to surrounding stars, or it does not so appear. Decision 

 relative to that fact determines the question, which is not strictly 

 quantitative. On the other hand, does the annual motion of the earth 

 in its orbit cause an apparent shifting from side to side of nearer stars 

 relatively to those more distant? It should do this if the earth really 

 moves and the stars are not so distant as to render this apparent shift- 

 ing, or annual parallax, inappreciable. The largest annual parallax thus 

 far detected for any star is less than one second; and in only a few in- 

 stances does it amount to one tenth of this quantity. To measure the 

 tenth of a second on the sky, with absolute certainty, in a single set 

 of operations on any one night, is undoubtedly beyond the present 

 possibilities of any means now employed. The question then arises 

 whether, by multiplying observations and varying the circumstances 

 of measurement, the casual and systematic errors of observation can 

 be so far eliminated through compensation of positive and negative 



