I860.] Soeby on Spectrum- Analysis applied to the Microscope. 201 



light may be obtained, if tbat be requisite. With day-light the con- 

 denser is not necessary, but the slit must usually be raised, the micro- 

 scope inclined at a less angle, and a screen of blackened cardboard, 

 with a suitable opening, placed in front of the slit, extending from 

 the table to some distance above b, so as to shut out extraneous light. 

 Since the uniform blackness of the field is often impaired by seeing 

 the reflection of various parts of the microscope from the upper sur- 

 face of the condenser, it is in all cases desirable to have a blackened 

 tube fixed over the end of the object-glass, extending down nearly to 

 the focus. 



When all is properly arranged, the narrow beam of light passing 

 through the slit at a, gives a spectrum at the focus of the achromatic 

 condenser (/), which is so adjusted that the focus may coincide with 

 any object placed on the stage at g. Both are magnified by the object- 

 glass (/*), and seen on looking down the microscope ; the spectrum 

 extending horizontally across the field, having the red at the top and 

 the blue and violet at the bottom. In my description it will there- 

 fore often be convenient to speak of the red as the upper, and the blue 

 as the lower, end of the spectrum. The breadth depends on the focal 

 length of the condenser and object-glass, and also on the distance of 

 the slit from the prism, and the thickness and refracting power of any 

 object placed on the stage at g. By using suitable lenses, we may 

 readily obtain a spectrum r7 ' 0T) th of an inch in width, so that the 

 whole may be seen through a crystal of that minute size, or we may 

 make the spectrum spread over the whole field of the microscope. 

 However, it is usually much better not to have the condenser and 

 object-glass of short focal length, since there is then more light, the 

 spectrum is of better quality, and the different portions in better 

 focus. A condenser of -frds of an inch focal length combined with 

 an object-glass of 1£, or a 1<? combined with a 2£, answer very 

 well. 



By using a micrometer in the eyepiece, the position of any dark 

 or bright band in the spectrum can be readily measured with sufficient 

 accuracy. I find it very convenient to use such powers, and fix the slit 

 at such a distance that the whole spectrum measures 12 divisions of 

 a micrometer, having lines -s^th of an inch apart. In that case the 

 sodium line is as nearly as need be 3 from the top and 9 from the 

 bottom. The slit can be easily kept at the proper distance by means 

 of a strip of wood resting above the prism and on the slit. To get a 

 correct datum with lamp-light, I have a platinum wire with a loop 

 holding common salt, so arranged that it can be turned into the flame, 

 and the sodium line seen as a bright yellow band in the spectrum. 

 After making one of the chief lines of the micrometer coincide with 

 this, the wire can be turned out of the flame, and the position of any 

 part of the spectrum measured. If day-bight be employed, and a good 

 object-glass used for the condenser, Fraunhofer's principal lines can 

 be readily seen, and D may be made use of as a datum in a similar 

 manner. 



I trust I have now made the general arrangements sufficiently in- 

 telligible, and will proceed to describe some of the objects to which 



