474 ME. A. E. TUTTOX OX THE THERMAL DEFORMATIOX OF THE CRYSTALLISED 
The results of the determinations of linear thermal expansion are presented in the 
next section in tabular form. Each table represents the results for one axial direction 
of a particular salt, and is divided into three jDortions. In the first portion is given 
the essential experimental data afforded by the observations and measurements. is 
the measured thickness of the crystal block, that of the compensator, I the length of 
the platinum-iridium screws, and d the thickness of the air-film, each measured in the 
manner described on p. 471. Next come the temperatures, and subsequently the corres¬ 
ponding barometric pressures. The next column contains fo, the number of inter¬ 
ference bands which effected their transit past the reference point during the mterval 
between L and to. In the succeeding column is given the small correction to be 
applied to the number of bands, rendered necessary by the alteration in the wave¬ 
length of the monochromatic light employed, which accompanies the change in the 
refraction of air consequent on the considerable rise of temperature and possible 
.alteration of pressure. The nature and a.mount of this correction were fully discussed 
ill the previous memoir {loc. cit. p. 350), and the formula for it there given was 
invariably followed. The barometric pressures and d are essential terms of that 
formula. The corrected number of bands, fi, is given in the next column, and the 
three remaining columns contain the number of bands for the temperature interval 
between b and its correction, and the corrected number for that interval. 
In the second portion are given, in the first two columns, the calculated values of 
the apparent expansion, obtained by multiplying the corrected number of bands by 
half the wave-length of the red C hydrogen light emjiloyed, 0•0003281 millim., 
according to the fundamental principle of the method; in the next six columns the 
calculated quantities involved in the correction to be applied to the apjDarent expan¬ 
sion for lack of compensation are recorded; and in the last two columns the actual 
expansion of the crystal obtained by use of the correction. For a fuller discussion of 
the principle of the method as touching the first tv/o columns, the memoir concerning 
the dilatometer may be referred to ; it need only be remarked here that the transit 
of each band past the reference spot corresponds to an alteration in the thickness of 
the air-film c/, between the compensator and the cover-wedge at the position of the 
reference spot, equal to half a wave-length of the monochromatic light employed. 
The determination of the correction for non-compensation involves the calculation of 
the actual expansion of the platinum-iridinm screws, which is given in the third and 
fourth columns for the two respective temperature-intervals, and of the aluminium 
compensator, which is given in the fifth and sixth columns. These values were 
calculated with the aid of the coetficients of linear expansion of the two metals, as 
previously determined with the greatest cai'e by the author and published in the 
memoir concerning the dilatometer (pp. 356 and 360). The following were the actual 
expressions used, / and being the values given in the fourth and third columns 
respectively of the first portion of the table 
