August ii, 192 i] 



NATURE 



747 



As Mr. Synge says, the angles of reflection and 

 incidence as measured by an observer moving with 

 the mirror must be equal. When the motion of the 

 mirror relative to the earth is in its own plane, the 

 effect of the FitzGerald contraction is the same on 

 each angle, since it affects all distances parallel to 

 the motion in the same ratio, while leaving those per- 

 pendicular to the motion unaltered. Thus the angles 

 will appear equal to an observer fixed with regard to 

 the earth. They would also appear equal if the 

 motion was normal to the plane of the mirror, but 

 not if if were in any other direction than these two. 

 Even in the last case, however, the difference would 

 depend, not on the motion of the earth, but on the 

 motion of the mirror relative to the earth. 



Again, it is true that the apparent distance between 

 two stars must varj' during the year on account of the 

 variations in the direction of the earth's motion rela- 

 tive to the stars ; if all larger disturbances were 

 eliminated this could be detected, but observation of 

 it could only determine the variations in the velocltv 

 of the earth relative to the stars, not its motion in 

 aether or "'space." The same applies to the immer- 

 sion in water of the Michelson and Morlev apparatus ; 

 none of these methods could tell us anything we do 

 not already know more accurately by other means. 



H.-iROLD Jeffreys. 



Conical Refraction in Biaxial Crystals. 



An arrangement for demonstrating conical refrac- 

 tion usually found in laboratories is a piece of 

 aragonite crystal mounted inside a little tube which 

 has one end covered with a metal foil pierced bv a 

 number of pin-holes, and an eye-lens in a focussing 

 mount at the other end. When the tube is directed 

 against a luminous object and the eye-lens focussed 

 on the pin-ho'es through the crystal suitably oriented 

 they are seen as luminous rings of light. \\'riters on 

 physical optics who describe this experiment refer to 

 it as illustrating internal conical refraction — that is, 

 as due to the fact that the Fresnel wave-surface has 

 a tangent-plane which touches it along a circle. I 

 wish to point out that this is really an error. A little 

 consideration will show that as the eye-lens is 

 focussed on the pin-holes, which may be as 

 small as we please, we are concerned here with the 

 waves diverging from them in all directions within 

 the crystal, and the observed effect is due to the fact 

 that the two sheets of the wave-front intersect at a 

 conical point. In other words, the experiment really 

 illustrates external conical refraction. This is con- 

 firmed by the fact that an extended source of light 

 may be used without interfering with the success of 

 the experiment. 



A remarkable effect is observed if, with the tube 

 pointed towards an open window, the eye-piece is 

 steadilv drawn back from the crystal. It will be 

 noticed that a well-defined image of each pin-hole 

 mav be traced behind the crystal for a distance of 

 several centimetres. The formation of this continuous 

 image by a crystalline plate with par^illel faces can- 

 not be explained on geometrical principles, and 

 is of great interest. The effect appears^ to be due to 

 the dimp'ed form of the wave-front within the crystal, 

 and is being further investigated by Mr. V. S. Tamma 

 and mvself. C. V. Raman. 



22 Oxford Road, Putnev, S.W.15, August 4. 



Coiling of Underground Shoots of Convolvulus arvensis. 



The shoots ascending from the rootstock of Con- 

 volvulus arvensis, before they reach the surface of the 

 ground, are frequently found to be coiled. The coils 

 vary in diameter from one to two inches or more, 

 NO. 2702, VOL. 107] 



and lie closely adpressed upon one another. A con- 

 siderable length of shoot, in some cases three or four 

 feet, is thus compressed into a small ^ace. No object 

 has been found enclosed by the coils which would serve 

 as a stimulus ; the soil contains very few stones to 

 obstruct the straight upward growth of the shoots. In 

 a few cases a similar coiling has been observed in the 

 ascending shoots of Cnrduus arveyisis. One of the 

 "popular " names of Convolvulus arvensis is "Devil's 

 Corkscrew." These white corkscrew coils of the 

 shoots underground seem more likely to be the origin 



I Fig. I. — Devil's Corkscrtw (Convolvulus arvensis). 



of the name than the less noticeable above-ground 

 coiling portion. 



In the accompanying illustration (Fig. i) A-B is 

 the rootstock ; the ascending shoot, originating at B, 

 is coiled at C, and terminates in the leafy above- 

 ground portion at D. (The coils were slightly pulled 

 out before taking the photograph.) 



J. E. H. Blake. 



Bees and Scarlet-Runner Beans. 



I SHOULD like to add to the remarks on bee visitors 

 to the flower of the scarlet-runner bean contained in 

 my letter in Nature of July 28, p. 684, the following 

 further observations. Some ten days from the time of 

 making the original notes a complete change was found 

 in the insect visitors to the flowers and in their be- 

 haviour towards them. Instead of the smaller black 

 and black with grey humble-bees busy over the blooms 

 in what I termed the legitimate way, there were num- 

 bers of a larger, yellow-banded species of humble-bee 

 that had bitten every newly developed bloom and 

 were searching the nectaries through the perforations 

 made in the base of the flower. They all unhesi- 

 tatingly scrambled to the underneath- part of the 

 blooms, which In every instance had been bitten before 

 the observation was made. Many honey-bees were 

 following In their wake, busily draining the exposed 

 nectaries of every particle of the sweet liquid that had 

 been left or had newly formed. 



The results of the two ways of visiting the flowers 

 are very marked and distinguishable. The earlier 

 flowers and lowest on the racemes that were first 

 visited without injury are replaced with a good show 

 of pods, while the later bitten blooms drop off very 

 quickly, with only barren pedicels remaining. 



The season being so unusually forward gave oppor- 

 tunity to the earlier Insect workers, which made some 

 return to the plant for its sweet gifts, while the 

 later humble-bees are mere depredators that only rob 

 and injure the plant. Harford J. Lowe. 



The Museum, Torquay, August 3. 



