March 20, 1890] 



NA TURE 



46, 



length of each which responds to radiation is a kind of measure 

 of specific intensity. They form (speaking sensationally) an 

 electric eye with a definite range of colour sensation. It would 

 be easy to supply it with a pitch or paraffin lens. 



There is no need to suppose the retinal bodies to be conduct- 

 ing : a body of high refractive index should be subject to 

 electric vibrations, and its surface to spurious electrifications, 

 when radiation falls upon it ; and the optical density of the rods 

 and cones is known to be high. They may, however, be 

 electrolytic conductors ; and I find that a liquid sphere — e.g. a 

 flask of inky water — responds to radiation, giving a glow to 

 a point touching its glass. 



The diameters of the rods, as measured by various physio- 

 logists, are not very different from dimensions adapted to 

 respond to actual light-vibration frequency ; and if this idea 

 substantiates itself, these bodies can be supposed to constitute 

 a sort of Corti's organ responding to etherial instead of to 

 aerial vibrations, and stimulating in some still unknown, but 

 possibly mechanical, manner, the nerve-fibre and ganglion with 

 ■which each appears to be associated. Oliver J. Lodge. 



University College, Liverpool, March ii.J 



" Peculiar Ice-forms." 



May I add another to the long series of communications 

 which from time to time have been addressed to you under the 

 above heading? Most of them have described and discussed 

 the occurrence of ice in the form of filaments. One signed by J. 

 D. Paul (Nature, vol. xxxi. p. 264) seems (the description is 

 somewhat vague) to refer to a mode of ice formation which is of 

 somewhat frequent occurrence here, and is the only reference to 

 this mode which I can find in that portion of the literature of 

 physics which is accessible to me. 



It happens now and again in our variable climate that a loose 

 porous soil which has been thoroughly soaked with rain is made 

 by a sudden and a sharp frost to produce a crop of little columns 

 of ice. I observed a striking instance lately on a piece of hard 

 compact ground, which, not being quite smooth, had been 

 ■covered with an inch or so of loose pebbly soil for levelling pur- 

 poses. Before the loose soil had been rolled or trampled upon, 

 it becai*e saturated with water through two days of continuous 

 rain ; and while it was still saturated, a sharp frost set in at 

 night. In the morning the ground, to the extent of 60 square 

 yards, was found to be covered with little columns of ice, some 

 of them about two inches in length. They were roughly circular 

 in section ; and each column had approximately the same section 

 throughout. Their diameters ranged from one-tenth to one-third 

 of an inch. They were not transparent, but were whitish inap-. 

 pearance, and carried on their summits pebbles or frozen earth. 

 They were thus obviously not ice crystals, such as Brewster de- 

 scribes in the Edinburgh Journal of Science, vol. ix. p. 122, as 

 occurring in similar circumstances. The columns started from 

 the ground at various inclinations to the vertical, and in the 

 great majority of cases they curved upwards to a greater or less 

 ■extent. I had never noticed this upward curving of the ice 

 columns before, but other persons familiar with the phenomenon 

 iissure me they have observed it. 



Tlie explanation of this mode of ice formation seems pretty 

 obvious. The sudden frost solidifies the crust of the soil ; and 

 it may therefore sometimes happen (in the above case it clearly 

 must happen) that water becomes imprisoned between the frozen 

 crust and the impervious sub-soil. Further freezing enables 

 nature to perform Major Williams's experiment for us. If the 

 crust does not give way as a whole, it must at its weak points ; 

 and the internal pressure is relieved by the protrusion of ice 

 ■columns through apertures formed at these points. Theje 

 columns would naturally carry portions of the crust on their 

 summits, and during their protrusion might be expected to have 

 innumerable minute fissures or cracks produced in them so as to 

 ■exhibit a whitish snowy appearance. At the base of any 

 column, at points where the freezing-point has been lowered by 

 ihe pressure to the actual temperature, melting is continually 

 occurring, and the water thus formed will flow into the fissures 

 referred to. If the axis of the column is inclined to the vertical, 

 and if we assume that the fissures and the points at which melt- 

 ing occurs are pretty uniformly distributed, more water will flow 

 into the fissures of the lower side of the column than into those 

 of the upper side. When the water re-freezes therefore, the 

 jower side must elongate more than the upper, and the column 



must consequently in general curve upwards. That in 

 exceptional cases the upward curving may not occur is obvious. 



J. G. MacGregor. 

 Dalhousie College, Halifax, N. S., March i. 



On a Certain Theory of Elastic After-Strain. 



In a recent paper (Proc. Lond. Math. Soc, April 11, 1889), 

 Prof. Karl Pearson has discussed at some length the possible 

 forms of the additional terms which may be introduced into the 

 general equations of elasticity by a consideration of the mutual 

 action of the molecules and the ether, and has examined what 

 physical phenomena may admit of explanation in this way. In 

 particular, certain terms which thus appear admissible are made 

 to yield a theory of the phenomenon known as '^ elastische 

 Nachwir/cung," or "after-strain." The attempt to explain such 

 a comparatively slow process by the intervention of the ether 

 certainly invites scrutiny, and in fact a very slight examination 

 serves, I think, to show that the theory in question rests on a 

 mistake. The author, after writing down the equations which 

 (on his view) represent the steady application of stress to a 

 portion of matter, proceeds to integrate them in the usual way 

 by assuming a time-factor e"'\ and arrives at a quadratic in m" 

 whose roots are /i//*' and (3A -f 2/*)/(3\' -f 2;u'), where A, jit are 

 the ordinary elastic constants of Lame, and A.', /*' are the co- 

 efficients of the additional terms referred to. He continues : — 

 " Now m cannot be positive, so long at least as we are dealing 

 with elastic-strain. For A.' and jx' are small as compared with 

 A and n, the effects we are considering being only of the second 

 order. Hence niP' is large, and if m were positive the strain 

 would rapidly grow immensely large, which is contrary to ex- 

 perience. Thus, we must give m the negative values - /^{ft/fi) 

 and - VKSA + 2/i)/(3A' + 2/)'." The positive values of m 

 are certainly inconvenient, but they are on the same footing with 

 the negative ones ; all are solutions of the author's equations, 

 and all are required for the purpose of satisfying arbitrary initial 

 conditions. The proper inference is surely that the substance 

 is unstable, so long as the constants fi/ij.' and 3A' + 2^' are (as 

 the author has tacitly assumed them to be) positive. If, to avoid 

 this disaster, we change the signs of these constants, we get 

 circular instead of exponential functions, and all analogy to 

 elastic after-strain of course disappears. In its place we have 

 vibrations (not molecular, but " molar") whose period is intrin- 

 sic to the substance and independent of the dimensions of the 

 portion considered. To what physical reality these may corre- 

 spond I do not undertake to say. Horace Lamb. 



The Owens College, March 4. 



Foreign Substances attached to Crabs, 



If, as Mr. Garstang seems to suppose, the presence of tuni- 

 cates on a crab is to be regarded as a danger-signal to its ene- 

 mies, then Hyas must belong equally to both his groups a and 3. 

 I have found simple tunicates {A. sordidd) on two small speci- 

 mens of .^. coarctatus. In one example they almost completely 

 hid, and several were larger than, .the crab. I do not know if 

 anyone has observed Hyas "dressing" itself with tunicates. I 

 should think it was an operation of some difficulty, at least in 

 the case of A. sordida, which adheres pretty tighily to stones 

 and shells. It cannot be said to be brilliantly coloured, so that 

 its assumption by Hyas might be regarded as only an adaptation 

 for concealment, as in the case of Algcc — belonging, therefore, 

 to group o. It seems to me, however, very doubtful whether a 

 small Hyas would, even if it could, willingly burden itself with 

 such a serious incubus as half a dozen tunicates. Probably their 

 presence is in no way due to any act of the crab's. 



The shore-crab, as pointed out long ago by Prof. Mcintosh, 

 frequently suffers loss of sight by the usurpation of its orbit by 

 a growing mussel, and the Norway lobster has been found with 

 one eye grown over by a Polyzoan. Such foreign bodies are 

 surely rather hurtful than protective, and the same may perhaps 

 be said of the tunicates on Hyas. It is also a question whether 

 the crab likes the smell of tunicates any better than its neigh- 

 bours. 



I think Mr. Garstang is wrong in assuming the inedibility of 

 tunicates. Prof. Mcintosh, in " The Marine Invertebrata and 

 Fishes of St. Andrews," speaks of Molgula arenosa as being 

 found abundantly, and of Pelonaia corrugata as occasionally in 

 the stomach of the cod and haddock ; and Mr. W. L. Calder- 



