92 



NA TURE 



[May 26, 1892 



rapidly, owing to the differentiation which takes place in the 

 water drops. 



The spectroscope shows that when the light is blue there is a 

 general darkening of the whole spectrum, but the absorption is 

 greatest in the red end, and the red end is also much shortened. 

 When the transmuted light was yellow, the blue end was cut 

 out, and the yellow part was much the brightest. 



The Cause of the Colour. 

 In the steam jet when the condensation is dense some of the 

 yellow colour in the transmitted light is due to some of the particles 

 being so small that they reflect and scatter the blue rays. This 

 blue reflected light is polarized. The colours, however, seem in 

 most cases to be produced in the same manner as the colours in 

 thin plates ; only a few of the colours of the first order or 

 spectrum are visible, whilst those of the second and third orders 

 are very distinct. 



A " Green " or "Blue" Sun. 

 It is thought that these phenomena give the explanation of the 

 "green " or " blue " sun seen in India and elsewhere in Septem- 

 ber, 1883, and also on other occasions. The eruption of 

 Krakata"b had taken place a few days before the green sun was 

 observed in India. The volcano threw into our atmosphere a 

 great quantity of water vapour, and a vast amount of dust, the 

 very materials necessary for producing a green sun by small drops 

 of water. Prof. C. Michie Smith's observations made in India 

 show that there was a great amount of vapour present in our 

 atmosphere at the time, and most observers frequently refer to 

 a fine form of haze which covered the sky on the days the green 

 sun was seen. It is therefore in the highest degree probable 

 that, under the conditions existing at the time, this haze was 

 greatly composed of water. 



A Nezv I nstrument for Detecting Dust-polluted Air. 



The colour phenomena produced when air is suddenly ex- 

 panded has led to the construction of a new instrument for in- 

 dicating roughly the amount of dusty pollution in the air. This 

 instrument has been called a " koniscope," and it is hoped it will 

 be found useful for studying sanitary questions. The instrument 

 consists simply of an air pump and a tube provided with glass 

 ends. The air to be tested is drawn into the tube, where it is 

 moistened and expanded. The depth of colour seen on looking 

 through the tube indicates the amount of impurity in the air. 

 With about 80,000 particles of dust per cubic centimetre the 

 colour is very faint; 1,500,000 gives a fine blue; while 

 4,000,000 gives an extremely dark blue. These colours are 

 for an instrument having a tube half a metre long. By means 

 of this instrument it is easy to trace the pollution taking place in 

 our rooms by open flames and other causes. We can trace by 

 means of it the pure and impure currents in the room, and note 

 the rate at which the impurity varies. 



May 5.— "The Potential of an Anchor Ring." By F. W. 

 Dyson, Fellow of Trinity College, Cambridge. Communicated 

 by Prof. J. J. Thomson, F. R. S. 



In this paper the author develops a method of dealing with 

 physical questions connected with anchor rings. He applies it 



(i) To find the potential of a solid anchor ring at all external 

 points. The result is obtained in a very convergent series of 

 integrals, each of which may be reduced to elliptic functions. 

 The equipotential surfaces are drawn, when the ratio of the 

 radius of the generating circle to the mean circle of the ring is 

 i, I, I, I. I. 



(2) The density, at any point, of a ring charged with elec- 

 tricity is found ; and the charge is calculated. 



(3) The velocity potential of a ring moving in an infinite 

 fluid is found, the kinetic energy calculated, and a few cases of 

 motion discussed. 



(4) The annular form of rotating fluid is considered, and the 

 form of the cross-section determined. The cross-seciion even 

 for large rings is, roughly, of an elliptic shape ; the minor axis 

 being parallel to the axis of revolution of the annulus. 



May 12. — " On the Embryology of ^«j^/£'//'^m evecta, Hoff'm." 

 By J. Bretland Farmer, M.A., Fellow of Magdalen College, 

 Oxford. Communicated by S, H. Vines, M.A., F.R.S. 



The germination of the spore and the development of the 

 prothallium have been described by Jonkman,^ who also 

 observed the formation of the sexual organs. The antheridium 

 ^ "De geslachtsgeneratie der Marattiaceeen," door H. F. Jonkman, 



NO. II 78, VOL. 46] 



is formed from a superficial cell of the prothallium, which 

 divides by a wall, parallel to the surface, into an outer shallow 

 cell and an inner cubical cell. The former, by walls at right 

 angles to the free surface, gives rise to the cover-cells ; while 

 the inner one, by successive bipartitions, originates the anthero- 

 zoid mother-cells. 



The antheridia are distributed both on the upper and lower 

 surfaces of the prothallium, and apparently without any 

 approach to regularity, though they are somewhat more frequent 

 on the lower surface. I may observe, however, that an antheri- 

 dium may often occur on the upper surface immediately above 

 an archegonium which has been fertilized. 



The archegonia occur exclusively on the lower surface. Their 

 development has been described by Jonkman, who also noticed 

 the division of the neck canal cell, by a transverse wall, into 

 two cells. The division is not, however, invariable, and in one 

 preparation in which the protoplasm had shrunk slightly from 

 the wall, I observed that the cell plate had not extended so as to 

 completely partition the neck passage into two cells. 



The oospore, after fertilization, speedily forms an ovoid 

 cellular body, and although I was not so fortunate, owing to 

 scarcity of material, as to see the formation of the earliest cell 

 walls, their succession could be determined with tolerable 

 certainty in the youngest embryo that I met with, consisting as 

 it did of about ten cells. 



The basal wall is formed, as in Isoctes, at right angles to the 

 axis of the archegonium. The next one in order of occurrence I 

 believe to be the median wall, which can easily be distinguished, 

 even in advanced embryos, as a well-defined vertical line. 



The tranverse wall ts much more indefinite, and early loses its 

 individuality owing to the unequal growth of the various parts 

 of the young embryo. The further cell-division is irregular, 

 and to a far greater extent than is the case with the leptospor- 

 angiate ferns as described by Hofmeister and Leitgeb. 



The anterior epibasal octants together give rise to the 

 cotyledon ; the stem-apex is formed, not as in the leptospor- 

 angiate ferns, from one octant only, but from both of the 

 posterior epibasal octants, though one of them contributes the 

 greater portion. The truth of this statement is seen on 

 examining vertical sections through the embryo cut at right 

 angles to the median wall, when a few cells on each side are 

 seen to be clearly marked out, by their dense protoplasmic 

 contents and large nuclei, as meristem cells. There is no single 

 apical cell in Angiopteris from which all the later stem tissue is 

 derived, and this fact is, without doubt, to be connected with the 

 character of the apical meristem just described. The root is 

 formed from one of the octants beneath the cotyledon, i.e. from 

 an anterior hypobasal one, and is at first indicated by a 

 triangular apical cell, which, in one fortunate preparation, 

 showed the first cap cell. The other octant, together with 

 the two posterior hypobasal octants (which together form 

 the rudimentary foot), round off the base of the embryo. 

 The root presents considerable difficulty in tracing the 

 course of its development, as the apical cell, at no time 

 very clear, is early replaced by two cells. Moreover, the 

 root grows in a somewhat sinuous manner in the embryo, and 

 the cells of its apex may easily be confounded with other tri- 

 angular cells which occur irregularly scattered in the lower por- 

 tion of the embryo. It finally emerges, not immediately 

 beneath, nor yet exactly opposite, the cotyledon, but at a distance 

 from it of between one-third and one-half of the circumference 

 of the embryo. The difficulties attending the exact following of 

 its growth, added to the scarcity of the material, have prevented 

 my elucidating completely the details of development, but the 

 important point, that, even before its emergence from the 

 embryo, its apex contains a group of initial cells, occupying the 

 place of the single one characteristic of other orders of ferns, 

 can be regarded as established with certainty. 



When the embryo has reached a certain size, it bursts through 

 the prothallium ; the root boring through below, whilst the 

 cotyledon and stem grow through the upper surface. This 

 manner of issuing from the prothallium at once serves to dis- 

 tinguish Angiopteris from those other ferns whose embryogeny 

 is known, and probably the peculiarity of its growth may be 

 reasonably connected with the direction and position of the basal 

 wall which separates the root and short portions of the embryo. 



Fresh leaves and roots speedily arise on the young plantlet 

 the second leaf appearing just above the place of exit of the first 

 root — that is, not quite opposite the first leaf. The third leaf 

 rises between the first and second ones, and nearer the first than 

 the second. Their roots observe the same rule of divergence as 



