Investigating Diatom Structure. By H. Hartridge. 371 



dependent on D, that the ratio of sin d to sin d' underwent but 

 little variation ; so that where there was evidence of the period of 

 the markings being approximately regular, the method would give 

 results of sutficient accuracy to fall within the limits of the 

 experimental error. 



The measurements were carried out as follows. The apparatus 

 having been set up as for tiie first series of measurements, a small 

 axial beam of parallel light was caused to fall on to the diatom. 

 Above the movable screen was placed the micrometer eyepiece, 

 and by this the distance between the first order spectra and the 

 direct beam could be estimated ; in the present case the observa- 

 tions were limited to those spectra lying at right angles to the long 

 axis of the diatom. The readings were, as before, obtained in terms 

 of X. A., and the ratio of sin d to sin d' at once obtained by division. 

 Eeference to the graph then gave the angle at which the area of 

 diatom examined lay. 



To determine the exact character of the markings of the diatom 

 valve is a matter of some complexity requiring the accumulated 

 evidence of several modes of investigation : certain of these will 

 now receive consideration. At the olTset the field of enquiry can 

 be materially reduced by at once eliminating the possibility of the 

 markings being either opacities, or points of different refractive 

 index, in an otherwise uniform envelope. This can be done by 

 immersing the valve in media of different E.I., and observing the 

 difference that takes place in the spectra of different orders. For 

 spectra are set up by diffraction not only by the light which 

 traverses the transparent framework between the markings, but 

 also by that through the markings themselves, whether they be 

 holes, pits, spines, or otherwise.* These spectra being of the same 

 period will overlie one another, and interference will take place, 

 the positions of the minima and maxima depending on the path- 

 difference of any given wave-length. Xow this path-difference 

 will be altered in the case of the perforation, etc., by a change in 

 the E.I. of the medium in which the diatom is placed ; whereas it 

 will remain unchanged if the marking be an opacity in an other- 

 wise uniform structure. But further, if a change of intensity is 

 found to follow the change of E.I. of the mounting medium, then 

 by measuring the intensity for a light of given wave-length with 

 media of different E.I., it is possible to calculate the path-difference 

 and thus to estimate the depth of the depression in the case of a 

 pit, or the height in the case of an elevation. Lastly, the evidence 

 for or against the markings being complete perforations may be 

 completed by comparing the value obtained above with a direct 

 measurement of the thickness of the valve, obtained by an inter- 

 ferometer method. In a case where the perforations go through a 



* Quekett Journal (1901), p. 13. 



