368 PEOCEEDINGS OP THE AMERICAN ACADEMY 



Probably the definition could not be sufficiently good for the measure- 

 ment of the closer double stars, but if clockwork was attached, faint 

 companions could be measured or approximate positions of the coarser 

 pairs determined very rapidly. The positions of nebulaj could also be 

 observed with a view to detecting their proper motion. Stars having 

 a large proper motion might be observed, and the observations so 

 arranged that any very large parallax would be detected. A similar 

 search for a large parallax of variable stars, short-period binaries, 

 minute planetary nebulte, or stars having singular spectra, might lead 

 to interesting results. The argument that no ordinary star is very 

 near does not apply to such objects. 



III. Spectroscopy. The increased dimensions which could be given 

 to the spectroscope, and its steadiness, would compensate in a great 

 measure for a defect in definition. By Zollner's reversion spectroscope 

 the slit might be dispensed with, and also the necessity of clockwork. 

 So many stars could be observed in a single evening that systematic 

 errors could be in a great measure eliminated, and as the spectroscope 

 would not be moved, we should have a great assurance that the deviations 

 were real. Of the six thousand nebulae hitherto discovered we know 

 nothing of the spectrum of more than three or four hundred, while the 

 observation of all the others with a large horizontal telescope would 

 not be a very formidable undertaking. It would also be interesting to 

 observe the spectra of all the clusters. It is possible that some may 

 consist of stars having singular spectra, or even of disconnected nebu- 

 lous masses, in fact forming clusters of planetary nebulje. The inter- 

 esting discovery by Dr. Copeland that Burnham's double nebula in 

 Cygnus is gaseous, shows the same tendency to aggregation in these 

 bodies as in stars. Observations of the spectra of all the red stars 

 and variables would also probably lead to interesting results. 



IV. Photometry. Should the instrument be devoted to photometry» 

 numerous problems suggest themselves. Variable stars could be ob- 

 served near their minimum when too faint to be identified witli an equa- 

 torial without great loss of time. Faint stars in zones or faint com- 

 panions to bright stars could be measured very rapidly. The relative 

 light of all the asteroids would be an interesting problem. Many 

 coarse clusters appear to consist of stars of nearly equal brightness. 

 Their light compared with their distances apart might aid our study 

 of their formation. Another useful investigation would be to measure 

 the brightness of all the nebulae. 



In the application of physics to astronomy doubtless many other 

 problems will suggest themselves. Thus no satisfactory results have 



