HOW TO OBSERVE THE HEAVENS. 



are also magnified, which latter circumstance is a serious inconvenience, but 

 one which has been partially surmounted by using threads of extreme tenuity 

 If no special means were provided in the telescope, these threads would r jc 

 be visible at night, for the light of a star would be insufficient to illuminate 

 them. To remedy this, there is an orifice near the middle of .he tube, 

 close to which a lamp is placed, the light of which is reflected on the wires 

 and produces a general illumination of the field of view. This orifice can be 

 expanded or contracted to any desired extent, and may even be altogether 

 closed, so that the illumination of the field may be var ; ed at the discretion of 

 the observer. Now when two stars are of such a decree of brightness that an 

 opening may be given to the orifice which will produce such an illumination 

 of the field as will extinguish them, we may compare their brightness by com- 

 paring those degrees of light, exposed to which they become invisible. This 

 is still, however, but an approximation. It is not only difficult to get a lamp 

 which will always yield light of the same intensity, and to know whether any 

 given lamp be such or not, but as various stars are of various colors and tints 

 of color, the same lamp will extinguish a star of its own color with an opening 

 of the orifice by which it will not extinguish an equally bright star of a differ- 

 ent color. Thus a red light would extinguish a star of the same red tint, while 

 a bluish star, even of inferior lustre, would continue to be visible when ex- 

 posed to it. It is However by no means impossible that a diligent and judicious 

 employment of limits of different colors might be made to add to our knowledge 

 of this part of astronomy; and it is more especially in such fields that the pri- 

 vate observer may become a useful assistant to the public one. Let us con- 

 sider more fully, then, one or two more of the various ways in which a person 

 fond of looking at the heavens, provided only with a moderately-good telescope 

 arid micrometer may make himself useful even without mathematical knowl- 

 edge. 



First, then, with regard to the variation of the fixed stars in magnitude and 

 color. It is evident that the question whether a fixed star revolve on an axis 

 or not, as our sun does, can never be settled except by some variations of ap- 

 pearance presented by its different parts as they come one after another under 

 the eye of the observer, and also that a regular succession of repeated appear- 

 ances in a star, is a very strong presumption of a rotation round an axis. For 

 instance, the star ft Persei will at eight o'clock on Monday evening appear as 

 a star of the second magnitude. On Tuesday, at midnight, it will be decidedly 

 smaller, and on Wednesday night it will resume its original magnitude. If it 

 be watched again on the nights of Thursday, Friday, and Saturday, the same 

 succession of changes will be observed. By repeated observations of this 

 kind properly compared together, it has been calculated that the exact period 

 of this succession of changes in ft Persei is two days, twenty hours, and forty- 

 eight minutes. 



It will probably be asked how such accuracy can be attained when the 

 changes observed are so gradual that it is evidently impossible to determine 

 even the five minutes when the star resumes the same degree of apparent 

 lustre 1 As the answer to this very pertinent and natural question involves a 

 point of universal importance in almost every class of astronomical observa- 

 tions, we shall explain it pretty fully. 



In all cases of natural phenomena submitted to experimental inquiry, or to 

 observation, rough approximations are first made, and these imperfect estimates 

 afterward become the means of obtaining others of greater accuracy, and so on 

 until the highest degree of precision has been attained. As an example of 

 the application and use of this principle, let us suppose that the length of the 

 year is to be determined, that is, the exact interval of time which elapses be- 



