hy iransmissiofi through Fluids. 43 1 



in nil probability*, become apparent and tinted, if not deci- 

 dedly so, raise the liquid by means of the syringe until it dues, 

 and observe the depth ; then move the vernier in a right- 

 handed direction until the image X attains its maximum of 

 obscuration, and read off the amount of rotation; continue 

 the movement of the vernier until O next becomes obscured, 

 then again observe X, and then again O. Thus you will ob- 

 tain four readings corresponding to the depth observed, each 

 of which should differ from the preceding 90^ if your observa- 

 tions have all been correct, but if not, a mean must be taken, 

 and it will in all cases be better to take a mean of several ob- 

 servations, and again to compare such mean with that ob- 

 tained from an observation of the height necessary to rotate 

 the ray 180°, which is procured as follows: — Flaving placed 

 the vernier so that X is at its greatest point of obscuration, 

 and observed the height at which the liquid stands, pump in 

 air to raise the liquid until O becomes most obscured, and 

 observe the additional height necessary to effect this, corre- 

 sponding to a rotation of 90° ; continue the raising of the liquid 

 until X again becomes obscured, and again observe the addi- 

 tional height, which, if the experiment be correctly conducted, 

 should be equal to the former, ami of course the total height 

 will correspond to a rotation of 180°. It is unnecessary to 

 say anything more about the mode in which the direction of 

 the rotation is to be ascertained, this having been already fully 

 explained. I shall conclude, therefore, with a few further ob- 

 servations on sources of error which may tend to mislead or 

 to interfere with the accuracy of observations. When exa- 

 mining different heights care must be taken that the fluid which 

 adheres to the sides of the tubes has time to subside, otherwise 

 the apparent amount of rotation will be thereby increased. 



Great care must be taken that the position of the mirror 

 is not altered after the adjustments are made, as with some 

 Nicol's prisms I have found a very slight change of the incli- 

 nation of the mirror alter the plane of polarization altogether. 



After the adjustments are made, any double refracting me- 

 dium placed upon the shelf xvill cause the image X to become 

 depolarized and therefore luminous; this imist not be con- 

 founded xmth circidar polarizatiuu, which I suspect has been 

 the case. Many oils, and some solutions of grape and potato 

 sugar, exhibit this phaBUomenon. The difference from circular 

 polarization will be readily understood by attending to our 

 former instructions, and observing that in these cases there is 

 no alteration of the position of the plane of polarization exhi- 

 bited on rotating the analyser, more especially on increasing 



* Alluding to the accidental circumstance of the depth corresponding to 

 ISr rotation. 



