3 0 0 D O C U M E N T 1 7 5 A P R I L 1 9 2 8 There was, then, the question of illumination. This needed to be exceedingly strong, and at the same time of uniform intensity throughout an experiment which might last for 2 or 3 hours. I am not sure whether any ordinary illumination would suffice to mark a fairly wide range of light intensities at the P.E.[19] cell, corre- sponding to the different opacities of the various points of the section. An ordinary carbon arc might just suffice. But it seemed to me that the true line to follow was to pursue the method employed, I understand, by Lummer,[20] of enclosing the arc in a chamber under pressure of many atmospheres. As the P.E. cell is much most sensitive to blue light, and as the exceedingly high temperature attainable when one uses such high pressure, shifts the chief energy region of the spectrum down to the blue, an arc enclosed in this way would probably be more than a hundred times as effective for the purposes of this method as an ordinary carbon arc. It would of course be necessary in any case to use an exceedingly sensitive P.E. cell such as is employed for astrophysical research. With light as effective as sunlight one could I think get results with a particle of c.m in diameter which would probably mean a microscopic resolution of about c.m. As to securing a perfect uniformity in the illumination over the period of exper- iment, there seemed to be no great difficulty in designing a simple automatic device which would effect this. The simplest way seemed to be to interpose a vessel con- taining some imperfectly transparent liquid between the arc and the microscope condenser. Portion of the beam after passing through the liquid is reflected on to a P.E. cell or selenium element, which affects a galvanometer needle, and if the nee- dle shifts its position, a little cock is opened which dilutes automatically the liquid in the vessel, until with the increased transparency, the reflected beam becomes stronger and brings the needle back to its normal position. —This could be adapted to make sunlight usable. I am not sure how near the biological section could be brought to the surface of the quartz plate, without impairing the totality of the reflection. Lord Rayleigh[21] states that total reflection takes place in an intermediate region between the two mediums of about c.m. in depth. But I do not know whether two surfaces can be brought within, say c.m or c.m of one another without at all im- pairing the totality of the reflection. This seems to be the only theoretical limitation imposed on the method. Everything else is one of perfection in technique. A word seems necessary regarding surface films of contamination. These are, of course, all important when one is dealing with magnitudes of the order of c.m. 10–6 210 –6 10–6 10–6 510 –6 10–6