August 12, 1920] 



NATURE 



741 



oil lamps are more likely to have had an indefinite 

 position, as they must Ix; handled to fill, but they 

 had probably been much longer close to the situations 

 in which we saw llicni. The less vegetative growth 

 round them and tin shadow effect behind the electric 

 lamps would seem to show that it was the shorter 

 light-waves which were requisite for this plant-life 

 rather than contact-warmth or longer heat and reddish 

 light. 

 I With regard to the transport of the spores to the 



^ depths of the caves, some experiments by Profs. 



Zeleny and McKeehan arc of interest. At the Winni- 

 peg meeting of the British Association in 1909 they 

 read a paper, followed by a discussion, on experi- 

 mental verifications of Stokis\ law for the fall of 

 spherical bodies in a viscous lluid. They deduced a 

 discrepancy between theor\ and experiment which 

 would seriously affect the cloud estimate of gaseous 

 ionisation. A fuller paper and further experiments 

 were published in Physikalische Zeitschrift for 

 F"ebruarv, 1910, in which they showed that while a 

 cloud of minute smooth paraffin spheres or mercury 

 droplets obeyed Stokes's law, \*i similai- <'.\periments 

 with the spores of Lycopcrdon, i'olx trii iuni, and Lyco- 

 podium (all nearly spherical) gave only about half 

 the terminal velocities required by mathematical 

 theory. In Nature of January 6, 19 10, I offered an 

 explanation of the apparent discrepancy shown by 

 their results. By using a large-aperture microscopic 

 objective with oblique illumination and spectrum- 

 sifted blue solar light, the sjjori's can be seen, just 

 within the limits of visibility, to be coated with a 

 mass of very fine hairs more than a radius in length. 

 Substituting in Stokes's formula for the terminal 

 velocity 



9 M 



where a is the radius, /x the air viscosity, and d the 

 density of the spores, the effective diameter comes 

 out to be just double that of the measured diameter 

 as seen in an ordinary microscope. This increase of 

 effective diameter is what should be expected if a 

 mass of air be entangled with the spore, or a tail of 

 eddies formed. Hence the physical measurement of 

 the terminal velocity of fall confirms the microscopic 

 observation of the hirsute coating in all the three sets 

 of cases where spores were vised. The spores are 

 enabled to be wafted great distances, therefore, much 

 as are the seeds of a dandelion. No Brownian motion 

 or rotation was observed, and this also suggests the 

 coating of hairs. Since the spore-walls are not abso- 

 lutely spheres or smooth in the sense that surface 

 tension makes the droplets, some Brownian motion 

 would have been expected if the external air molecules 

 could strike directly on the spore-wall. The air en- 

 tangled in the chevaux de frise of hairs will, however, 

 soften down the averag-e result of individual impacts 

 of external air molecules by making the effect slower, 

 and therefore the resultant average smoother. 



Yet another indication of this coating of long hairs 

 is the difficulty of wetting Lycopodium dust until it 

 has lain on the water long enough to get water- 

 logged, viz. long enough, probably, for the entangled 

 air to be dissolved out. While the air is so entangled 

 the effective density is more nearly one-eighth than a 

 little above unity as measured by Profs. Zeleny 

 and McKeehan. 



That this hair\- coating provides these spores with n 

 sp)ecial mechanism which enables them to be carritd 

 great distances, is onlv to make them resemble man\ 

 other wind-borne fruits, and the fact is therefore 

 likely from general considerations. 



NO. 2650, VOL. 105] 



The method of verifying a difficult, almost ultra- 



microscopic, obstrvaliii 

 terminal velocit\, as 

 viscous fluid, is ixThai 



b(iian\ 1)\ measuring the 

 small failini.; body in a 

 I common. 



Edith A. Sm )m;v. 

 King's ('oll<g( for \\\)men, London. 



Curious Formation of ice. 



In Nature of December 12, 1912, was published a 

 letter wherein 1 described a curious formation of ice 

 in the hope that some of your readers would be able 

 to explain the cause, but there was no reply. After 

 five years the formation occurred again in similar 

 circumstances, and 1 submit a partial explanation 

 which occurred to me on seeing this second example 

 of the phenomenon. The ic<' was again forni«d on 

 water in a rough hole or pond (about 2 ft. b\ i ft.) 

 in the garden in clay soil. it was ot)><r\(d at 

 3.30 p.m. on January 13, 1918. Th« •dark, sinuous 

 lines " in this case were about ^ m. w idt , ami again 

 ran "about parallel to the major axis " of the small 

 pond. These dark lines were again due to the ridges 

 of ice on the under-side of the ice covering the water, 

 but were closer together than before, being about 

 i^ in. apart. The cross-section of the ridges was 

 again of "dovetail" shape, the attachment being at 

 the smaller end of the ' dox ctail. " 



The partial explanation apjx'ars to be as toilow s : 

 A uniform layer of ice about I in. thici^ forms o\(m- 

 the w^hole surface of the water. The water slowl\ 

 leaks out of the pond. The ice sags in the middle, 

 keeping in contact with the water over its central 

 area, but, owing to the support of the sides of the 

 pond, the edges do not sag, and an air-spac<- forms 

 under the ice round its margin. Tlie m itiial < ross- 

 section of this air-space is a long, narrow triangle 

 lying on one long side (the free surface of the water) ; 

 the under-side of the ice forms the other long side, 

 and the mud-bank of the pond the short side. At 

 night, or at any other tinn- when the tem]xiatun' again 

 falls below freezing point, die water at tlv margin 

 (where the ice and surfai-c of tlie water meet at an 

 acute angle) fre<'zes to the slab of ice and forms a 

 ridge on the under-side of the ice. Tlie water 

 leaking slowly from the ])ond all the wliib would lielp 

 the formation of the ridj^t . Tlie ui xt tla\ , or when 

 the temperature is again slightly above freezing, the 

 water, continuing to leak away, allows a further slight 

 sagging of the ice and the enlargement of the air- 

 space, thus g^iving the space between the ridges of 

 ice. The next freezing forms the second ridge, and 

 so on. 



This explanation appears to account for the ridges, 

 their spacing, and their being roughly parallel to the 

 major axis of the pond, but it does not ai count for 

 the beautifully sharp, regular, and s\mm(trical 

 formation of the cross-section of the ri<lg<s. One 

 expects an asymmetrical cross-section instead of the 

 symmetrical "dovetail." It has been suggested to 

 me that the "dovetail" shape is duo to the ridge 

 being partly melted (where it is joined to tJTi top slab 

 of ice) during the period when the temperature is 

 above freezing bv the comparatively warm top surface 

 of the water. This seems to be a possible explana- 

 tion if the cross-section of the ridg(> when first formed 

 is rectangular. 



T hope that with this as a basis som( on.' will be 

 able to complete or modify the e\]ilanation of the 

 riirious formation of ice observed. 



-Alfred S. K. .\( kkkmaw. 



2^ Victoria .Street, Westminster, London, 

 S.W.I, August 3. 



