December 29, 192 1] 



NATURE 



567 



light intensity. Many workers resort to the sub-stage 

 iris diaphragm for this purpose, with the result that 

 definition and resolution suffer. Where colourless or 

 lightly stained preparations are under examination 

 the advantage of light regulation is considerable, 

 while for dark-ground illumination it is absolutely 

 essential for accurate observation. It is f>ossible that 

 there would be a limited demand for objectives cor- 

 rected for some small region of the visible spectrum, 

 but whether the advantages to be derived are sufficient 

 to induce any manufacturer in this country to take the 

 matter up I do not know. J. E. Barnard. 



National Institute for Medical Research, 

 Ham[>stead, London, N.W.3, December 20. 



Prismatic Structure in Optical Glass. 

 The specimens of columnar structure illustrated in 

 the accompanying reproductions of photografrfis 

 Figs. I and 2) were obtained during the manufacture 

 f optical crown glass. 



The dimensions of the pot were 24-in. diameter 

 and 24-in. height. After the final stirring, during 

 which the temperature fell to about 1000° C. the 



Fig. I. 



pot of glass was withdrawn from the furnace and 

 quenched overall externally with cold water. The 

 pot was then regularly cooled during about eighty 

 hours. Fracturing occurred radiallv from the 

 centre to the sides and base of the ix)t. Without 

 further consideration it might be assumed that the 

 columnar fracture commenced at the pot surface, 

 which had been severely chilled, and spread radially 



Towards a central nodule the diameter of which was 

 about 6 in., but such a conclusion is scarcelv justified. 

 It is the central nodule that is of most interest, as 

 it gives a clue to the history of the specimen. The 

 prisms of glass (Fig. i). which are about 

 6 in. long, ended abruptly at the nodule surface, 

 which was uniform. To break the nodule was 



NO. 2722. VOL. 



108] 



difficult ; it behaved like toughened glass. There can 

 be little doubt that its surface-layers were held in 

 compression by the tension of the interior, and, there- 

 fore, any surface cracks were held closed, thus pre- 

 venting fracture. .\s the prismatic structure ended 

 abruptly at the surface, there being no sign of its 

 continuation within the nodule, it is evident that the 

 surface of the nodule constituting the terminal joints 

 of the prisms was not of later date than the surfaces 

 of the prisms. 



.\t the moment of separation of the nodule its 

 material would tend to move towards its centre. Its 

 surface would move to a place of smaller area, and, 

 being thereby compressed, the formation of normal 

 fractures would be prevented. But the material 

 external to the nodule would withdraw towards the 

 outside. The surface adjacent to the nodule would 

 move to a position of greater area and would crack" 

 |X)lygonally. These cracks would instantly spread 

 outwards. That the action started from the centre 

 and not the outside is indicated by the close adhesion 

 of the prisms at their base, as compared with their 

 inner ends. Indeed, it is difficult to trace any passage 

 of the cracks as far as the pot surface. 



This sequence of events may throw some light upon 

 the time-relationship of joints and prism surfaces in 

 natural columnar basalt dykes. 



To explain the sf)ecimen which is reproduced in 

 Fig. 2, about one-sixth full size, is not so easy. In 

 this case the diameter and height of the fireclay pot 

 were each 12 in. It contained a small experimental 

 melt of optical crown glass. At a temperature of 

 about 1300° C. the whole pot of glass was plunged 

 into cold water for about five minutes. It was then 

 cooled regularly during twelve hours. In the glass 

 itself there was no sign of prismatic structure, which 

 was confined to the fireclay base of the pot as illus- 

 trated. Most probably the whole action took place 

 during the drastic cooling of the fireclay, the bottom 

 thickness of which was only about i in. 



The glass itself within a short distance of the 

 surface must have been still quite plastic after five 

 minutes' quenching. Any trace of structure that 

 might exist in the intermediate chilled layer would 

 be lost in the inevitable shattering of this layer that 

 usually occurs. James Weir French. 



Anniesland, Glasgow. 



The Calendar of Scientific Pioneers. 



With this issue of Nature the series of short bio- 

 graphical notes on the leading men of science of the 

 past, which it has been my privilege to contribute 

 week by week, comes to an end. It has, however, 

 been considered that it may be of interest to supple- 

 ment these notes by another series referring to the 

 great pioneers of industry, and the Editor has again 

 entrusted the preparation of this new Calendar to me. 



In bringing the Calendar of Scientific Pioneers to 

 a close, I should like to express my thanks to Prof. 

 J. Arthur Thomson, Prof. G. B. Matthews, and 

 others who kindly scrutinised the list of names to be 

 included, and to various correspondents who have 

 supplied me with information. Mv thanks are 

 esf)ecially due to Mr. E. T. Warner, of the Roval 

 Naval College, Dartmouth, who has carefully checked 

 all the notes and has assisted me in other wavs. 



Great care has been taken to give the correct dates, 

 and where authorities differ, as they often do, in- 

 quiries have been made in order to trA- to trace the 

 source of error. In the case of Joseph Black, for 

 instance, who died on December 6, 1799, but whose 

 death is often stated to have taken place on Novem- 



