200 



NA TURE 



[December 30, 1897 



LARGE REFRACTING AND REFLECTING 

 TELESCOPES. 



''PHE Yerkes Observatory, which has recently been 

 -^ completed and inaugurated, contains in its instru- 

 mental equipment the largest refractor in the world, the 

 diameter of the object-glass spanning 40 inches. 



The late Mr. Alvan Clark, the constructor of the lens 

 in question, expressed the hope that still larger apertures 

 might be successfully made, but he pointed out that the 

 effect of flexure in larger discs was to be mostly feared, 

 although he felt that it might be perhaps possible to still 

 further increase the aperture without endangering the 

 performance of the objective. Being therefore ap- 

 parently near the limit to which such large object- 

 glasses can be successfully constructed and mounted, it 

 is only quite natural that attention should be turned 

 to the other form of telescope, namely the reflector, and 

 inquire whether this type of instrument is restricted in 

 the same way as the refractor, or whether it can step 

 beyond these bounds and open up fields which would 

 otherwise be lost to us. 



Discussions as to the capabilities of these two types 

 of telescopes have been very rife, and while some 

 observers hold that the reflector is the instrument of the 

 future, others again take the other side and advocate 

 refractors. It is now generally conceded that for 

 definition the refractor is the instrument par excellence., 

 but for purposes where light-grasping power is the main 

 requirement the reflector takes the first place. In the 

 cases of very large apertures reflectors can be made of 

 diameters far exceeding anything that can be attempted 

 for refractors. A point of initial importance in large 

 instruments is the question of the focal length of the 

 object-glass or mirror, as the case may be, for on this 

 factor depends the length of the telescope tube. Now if 

 this be of considerable length, the telescope mounting 

 and dome have to be of considerable proportions, render- 

 ing the instrument both expensive and subject to many 

 possible errors. To retain the size of the aperture of the 

 instrument and reduce the focal length is a natural 

 means of overcoming this difficulty, and this has been 

 attempted in many instruments. Such a reduction is, 

 however, accompanied by several optical drawbacks 

 which detract from the efficiency of the instruments. 



In the case of reflectors of large aperture and very 

 short focal length a most striking deficiency becomes 

 apparent, and, curiously enough, this has practically been 

 passed unnoticed until Prof. Schaeberle {As/r. Journal., 

 vol. xviii. No. 413) quite recently brought attention to it. 

 So large is this source of bad definition, that he refers to 

 it as a "fundamental optical defect." How he came to 

 alight on this source of blurring factor will be best 

 gathered from the following brief extract in his own 

 words. 



" On a very favourable night, I recently had the 

 opportunity of testing the great Crossley reflector of the 

 Lick Observatory, and found the surface of the same to 

 be a practically perfect paraboloid of revolution ; but on 

 examining certain celestial objects — Saturn among others 

 — ^^I was very much surprised to find that the instrument 

 failed most signally to come up to expectations. While 

 puzzling and pondering over the probable cause of the 

 poor results given by what I knew to be a finely figured 

 surface, it occurred to me to ascertain the exact amount 

 of the error introduced in the form of the image, result- 

 ing from the well-known fact that the focal point is not 

 at the centre of curvature of the parabolic mirror." 



To understand the origin of this bad definition, one 

 must imagine a small circular disc situated in the focal 

 plane of the paraboloid of revolution, and concentric 

 with the optical axis. Viewing this disc from different 

 points on the surface of the mirror, it is obvious that it 

 will appear circular only when the eye is in the optical 



NO. 1470, VOL. 57] 



axis, but in all other positions it will appear elliptical, 

 the eccentricities of the ellipses becoming greater the 

 further the eye is moved away from the optical axis. 

 Further, the angular diameter of both axes of the disc 

 will decrease as the eye moves away from the optical 

 axis, in consequence of the increase in distance from the 

 focal point. The result of such a source of error as this 

 would be that if the rays from the components of a 

 double star be reflected by the mirror, the linear distance 

 between their focal images as formed from different areas 

 along any radius of the mirror will vary from a minimum 

 for the area on the optical axis to a maximum for that 

 area furthest away. In the case of a planetary disc, 

 there will be produced a blurring effect caused by the 

 numerous images of different sizes overlapping one 

 another. 



Having investigated this source of error. Prof. 

 Schaeberle made a comparison of the efficiency of the 

 more prominent reflectors now in use. The result is of 

 such interest, that we must refer to it at some length. 



In the following table the blurring effect for each 

 mirror is tabulated in the fourth column, the fifth and 

 sixth columns representing the computed difference of 

 angular diameters of the outer ring of Saturn and the 

 solar or lunar disc based on the given ratio of diameters. 



The last column gives the values for the limiting 

 radius of the field of view of best definition for each 

 telescope, on the assumption that the radius of the field 

 of measured distances with the optical axis as centre, 

 does not exceed a value which would introduce an error 

 of more than o"'o5. The most striking feature of this 

 column is the smallness of the fields in the several cases 

 mentioned which are not influenced by this error. 



A glance down the fifth column brings out clearly the 

 fact that the smaller the ratio of focal length to diameter 

 of mirror, the larger the difference of angular diameter of 

 the objects observed. Thus in the case of the Draper, 

 Schaeberle, and the two Lassell instruments, where this 

 ratio is comparatively large, about 9:1, the differences 

 are small, while for the two Common mirrors and 

 Schaeberle's 12-inch, where the mean ratio is approxi- 

 mately 4:1, the differences increase very rapidly. 



Not only will this blurring effect, caused lay these dif- 

 ferences of angular diameters of the images, be notably 

 increased as the focal length is decreased, but the greater 

 diameter of the image, and therefore distance from the 

 optical axis, will bring this defect more in evidence. 



The main result of the investigation, summed up in a 

 few words, is that large parabolic mirrors having a ratio 

 of focal length to aperture less than fourteen to one are, 

 as regards definition, " theoretically unfit for making 

 observations of extreme delicacy." It may be mentioned 

 that this defect does not mar the efficiency of such 

 instruments for certain kinds of work, such as spectro- 

 scopic, bolometric, &c. 



