429 



were measured, and the thirteenth was remeasured this morning 

 by sunlight. A heliostat was used to give a steady beam of light, 

 and blue glass to make it monochromatic. The index error of the 

 cobweb micrometer was redetermined for the occasion ; three ob- 

 servations gave F 9 T > A' A» revolutions respectively. This cor- 

 rection was applied in all the calculations. All measurements 

 were repeated ; all the striae measured were counted at least twice. 

 In the first five diatoms, as is well known, the stria? are not of 

 uniform fineness all over the surface of the frustule ; in these, one 



noticed on the frustule ; and another was made on the same strice, 

 but at the point where they converge nearest to each other. Care 

 was taken not to be deceived by spectral lines. An immersion 

 sixteenth, by Tolles, having a maximum angle of 178 degrees, was 



