202 CARNEGIE INSTITUTION OF WASHINGTON. 



zium stars this gas is rendered luminous; brilliantly so, with excitation of the 

 nebular lines, in the near neighborhood of the stars; faintly, without bright 

 lines in the outlying regions. In some places this secondary radiation is 

 partly or completely obscured by intervening dust clouds, producing the dark 

 structures seen visually and in photographs (p. 223). 



The measurements of Mr. van Maanen upon three additional spiral nebulae 

 confirm results already obtained from four others and point to motion out- 

 ward along the spiral arms rather than rotational motion. From these values 

 combined with radial velocities, and from a discussion of the proper motions 

 of a large number of spirals and the theoretical considerations of Jeans, van 

 Maanen arrives at parallaxes for the spirals of from O'fOOOl to ''0010. This 

 would indicate distances considerably less than would be required by the 

 island universe theory (p. 224). 



Van Maanen, assisted by Gingrich, has carried on measurements of paral- 

 lax and proper motion and has now determined the trigonometric parallaxes 

 of 160 stars. A new discussion of the systematic errors of these parallaxes by 

 several methods, some of which depend upon objects whose distances are 

 known with high accuracy from independent sources, yields a result of 

 — 0'' 0024 ±0". 0003. Several objects of especial interest have been investi- 

 gated and the probable distance of the nebula near o Persei is found to be 

 about 350 light-years (p. 225). 



The continuation by Mr. Pease of his measurements of stellar diameters 

 with the 20-foot Michelson interferometer has given results for a Orionis and 

 a Scorpii in fair agreement with those obtained previously. A slightly larger 

 value for the angular diameter of a Orionis perhaps is to be ascribed to the 

 effects of seeing. A study of the visibility curves of a Tauri, a Bootis, and (3 

 Pegasi indicates that the fringes of these stars would disappear at a separation 

 of the interferometer mirrors of about 25 feet, which would correspond to an 

 angular diameter of 0T019. A more extensive extrapolation gives probable 

 values of 30 to 40 feet for the separation in the case of y Andromedse, and 40 

 to 50 feet in that of a Arietis (about 0!013 and 0''010, respectively). Numer- 

 ous stars have been observed for which the visibility of the interference 

 fringes is from 50 to 80 per cent (p. 236). 



The use of an auxiliary interferometer with one variable aperture, as sug- 

 gested by Michelson, for the purpose of calibrating the principal set of fringes, 

 has been investigated extensively by Pease. Under good conditions of seeing 

 this has proved most valuable for determinations both of visibility and of 

 seeing. With poor seeing, however, the two sets of fringes show marked rela- 

 tive variations, and simple visual estimates based upon experience appear to 

 be preferable (p. 237). 



Immediately after the first successful measurements of the diameter of 

 a Orionis, Messrs. Hale and Pease made provisional designs of a 50-foot inter- 

 ferometer, on independent equatorial mountings. These were too expensive, 

 however, for serious consideration, and the problem was dropped for the time. 

 As it is now evident that a large instrument of this kind must be provided, Hale 

 has devised a 50-foot interferometer telescope, the detailed design of which has 

 been worked out by Pease and Nichols. The essential feature of this instru- 

 ment is its extreme simplicity of form, and the precautions which have been 

 taken to avoid the necessity of large and expensive machine work. Theoretical 



