August 25, 1923] 



NATURE 



281 



Some of these organisms are described as possessing 

 a single flagellum and an undulating membrane, 

 others as biflagellate ; their general resemblance to 

 trypanosomes is also claimed. 



When Nelson's paper was received in this country 

 some few months ago, I was engaged in a study of 

 the mosaic of hops, a disease probably to be classed 

 as a virus disease. A search for protozoa similar to 

 those described by Nelson was accordingly made in 

 the phloem of hops thus affected. No such organisms 

 were to be observed, but elongated deeply-staining 

 structures having a marked resemblance to those 

 figured by Nelson and described by him as protozoa 

 were found, as shown in Figs. la, ib. In the case of 

 the hop-mosaic these structures were undoubtedly de- 

 generate nuclei, for all transitions could be observed 

 between them and the normal nuclei of the phloem. 

 These degenerate nuclei were not observed in the 

 phloem of healthy hop plants, but they 

 were to be seen in the phloem of an 

 unhealthy bean plant that had been 

 1 \| ' \\ kept some time in the poor light of 



■ \ ^*^\ ^ laboratory and the leaves of which 



■ \ ^!3 ^ ^e^re attacked by Botrytis (Fig. i , c) . 

 ^ ' 11 •, - ■ These results do not, of course, dis- 



1 !l ' I ^ prove the observations of Nelson as to 

 ' \* the association of pro- 



tozoa with virus dis- 

 eases, for the diseases 

 which he investigated 

 have not been studied . 

 Considering, however, 

 how important the 

 discovery of a causal 

 organism in virus dis- 

 eases would be, it 

 seemed advisable to 

 put on record the re- 

 sults obtained with 

 diseased hops and 

 beans. 



Such results indi- 

 cate clearly that the 

 theor}^ of the associa- 

 tion of protozoa with 

 virus diseases requires 

 fuller evidence than 

 has yet been supplied. 

 It is to be noted 

 that Nelson describes 

 the protozoa in the 

 plants he examined as 

 usually existing singly 

 in the cells, and as always elongated in the direction of 

 the axis of the stem, i.e., the organisms stand perman- 

 ently on end in the plant. These somewhat remarkable 

 results would find an easy explanation if the structures 

 in question were no more than the degenerating nuclei 

 of the elongated cells of the phloem. 



M. S. Lacey. 

 Department of Plant Physiology and Pathology, 

 Imperial College of Science and Technology, 

 South Kensington, S.W.7, August 8. 



Fig. I. — Longitudinal sections of the phloem. 



a and b, a mosaic hop stem. X looo ; 

 c, an unhealthy bean plant. X 600. 



The Scattering of Light by Liquid and Solid 

 Surfaces. 



It is a well-known fact of observation that most 

 reflecting surfaces usually also scatter a little light 

 and are thus rendered visible. The effect is usually 

 dismissed, however, as due to dust or imperfect 

 polish of the surface, and little attention has been 

 given to the problem of determining whether, when 

 these disturbing factors are eliminated, any scattering 



NO. 2808, VOL. 112] 



by the surface persists. Experiments carried out by 

 the writer in collaboration with Mr. L. A. Ramdas to 

 test this matter have led to some interesting results. 



Freshly split cleavage faces of crystals show extra- 

 ordinarily little scattering. In fact, it is found that 

 a clean good piece of mica has surfaces which are 

 invisible even when placed at the focus of a lens 

 illuminated by sunlight against a dark background. 

 This is what one would expect theoretically. Cleavage 

 surfaces of rock-salt and Iceland spar are also good, 

 though not so perfect. The conchoidal fracture- 

 surfaces of quartz are relatively very imperfect 

 optically. Blocks of thick plate glass when freshly 

 broken open exhibit surfaces which apparently are 

 quite smooth, but when illuminated by sunlight they 

 show a blue superficial opalescence. Freshly-blown 

 bulbs of glass when held in a strong light also show 

 this surface opalescence very well. 



Coming to liquids, the most interesting case is 

 that of metallic mercury. After carrying out a 

 series of chemical purifications, washing and drying 

 the mercury and then distilling it in vacuum from 

 one bulb to another and transferring it back again 

 repeatedly, Mr. Ramdas succeeded in obtaining 

 surfaces which were dust-free and chemically clean. 

 When sunlight is concentrated on such a mercury 

 surface in a vacuum, the focal spot shows a bluish- 

 white opalescence, the scattered light when observed 

 in a direction nearly parallel to the surface being very 

 strongly polarised with the electric vector perpendi- 

 cular to the surface and of nearly similar intensity 

 in all azimuths. The opalescent spot when examined 

 under a microscope appears perfectly structureless, 

 showing that it is a true molecular phenomenon. 



To test whether the surface-opalescence exhibited 

 by mercury is due to the mobility of the dispersion- 

 electrons usually assumed to exist in metals, or 

 whether it is due to the rugosities of the surface 

 caused by molecular bombardment, observations 

 were also made with transparent liquids in enclosed 

 bulbs made dust-free by repeated distillation. 

 Various liquids tried, e.g. ether, alcohol, benzene, 

 carbon tetrachloride, liquid carbon dioxide, all 

 showed the surface-opalescence conspicuously under 

 strong illumination. The character of the effect in 

 these cases was, however, quite different from that 

 shown by a clean mercury surface. 



The surface-scattering by transparent liquids is un- 

 doubtedly due to the effect of molecular bombardment 

 of the surface. It is much more intense when observed 

 in directions adjacent to that of regular reflection and 

 refraction than in other directions. It is less blue 

 than the internally-scattered light, and shows re- 

 markable changes in its state of polarisation with 

 varying angles of incidence and observation. There 

 were notable differences in this respect between the 

 cases in which the light is incident respectively 

 within and outside the liquid on the interface. There 

 is a rapid falling off in the intensity of the surface 

 opalescence when the angle of incidence is increased 

 much beyond the critical angle. These facts clearly 

 indicate that the effect shown by transparent liquids 

 is essentially due to the imperfect planeness of the 

 surface. The scattering by a metallic liquid surface, 

 on the other hand, has probably a different origin, as 

 suggested above. 



The interface between two non-miscible dust-free 

 liquids also shows strong opalescence under illumina- 

 tion. For the particular case in which the interfacial 

 tension is very small or negligible, the opalescence 

 becomes greatly exaggerated. Some observations by 

 Mandelstamm [Ann. d. Phys. vol. 41, 1913) on the 

 critical state of liquid mixtures are of interest in 

 this connexion. 



H 2 



