NATURE 



[July 14, 19 10 



uncles and aunts, but not his cousins, are generally con- 

 sidered ; but the results of the present paper show that 

 his cousins, usually more numerous, give just as good a 

 knowledge of his constitutional tendencies as do his uncles 

 and aunts. — C. J. T. Sevwell : The propagation of sound 

 In a fog. This paper is intended as a sequel to the author's 

 previous paper on " The Extinction of Sound in a Viscous 

 Atmosphere by Small Obstacles of Cylindrical and Spherical 

 torm," in which the loss of energy from the primary 

 waves owing to viscosity was investigated. In the present 

 paper the author has included the additional loss of energy 

 due to heat conduction. The work proceeds on much the 

 same lines as before, and the results obtained are of the 

 same order of magnitude. The chief interest consists in 

 the application of the results to the effect of atmospheric 

 log upon the propagation and audibility of sound. Waves 

 of high frequency suffer most. If the diameter of the 

 drops of water in a dense fog is assumed to be 002 mm., 

 and the density of the fog amounts to 45 grams per cubic 

 metre, the intensity of sound of wave-length 100 cm. is 

 reduced in the ratio of i to « before the sound has travelled 

 a distance of 100 metres. If the wave-length is 1000 cm., 

 this distance is increased to about 350 metres. In any 

 case, the results seem to show that the presence of fog 

 at sea must diminish quite appreciably the audibility of 

 sound. — L. Southerns : A determination of the ratio of 

 mass to weight for a radio-active substance. A determina- 

 tion has been made of the ratio of mass to weight for 

 uranium oxide by comparison with the known value for 

 a normal substance (lead oxide). It had been supposed 

 by Sir J. J. Thomson that a radio-active substance might 

 possess greater mass than the same weight of a non- 

 radio-active substance, on account of the greater store of 

 potential energy which is associated with the former. In 

 the case of uranium oxide, the increase in the ratio of 

 mass to weight would be about i in 16,000. The investi- 

 gation has been made by means of a rigid pendulum fitted 

 with two knife-edges and a hollow bob, into which could 

 be packed either of the substances used. Special means 

 have been employed in order to eliminate errors due to 

 slight variations in the position of the centre of gravity 

 of the pendulum, and to other causes. The results show 

 that the ratio for the uranium oxide does not differ from 

 the normal value by more than i in 200,000, and thus 

 that the contemplated effect is absent. — F. P. Burt and 

 F. L. Usher : The relative atomic weights of nitrogen 

 and sulphur. The object of the research was to determine 

 the combining weights of nitrogen and sulphur by the 

 analysis of nitrogen sulphide. The method adopted was 

 briefly as follows : — A weighed quantity of nitrogen 

 sulphide, purified by sublimation in -vacuo over silver at 

 100° C, was decomposed by subliming over red-hot quartz 

 wool contained in a quartz tube. The sulphur was de- 

 posited a few inches beyond the wool, and the nitrogen 

 was pumped off and estimated in a constant-volume gas 

 burette. .Assuming the density of nitrogen, the relative 

 weights of nitrogen and sulphur could be calculated, the 

 sulphur being obtained by difference. The problem was 

 complicated by the impossibility of starting an experiment 

 with_ the quartz wool in the reaction tube in a gas-free 

 condition. The difficulty was overcome by measuring in 

 blank experiments the quantity of air removable from the 

 hot wool in vacuo, and by estimating traces of more con- 

 densable gas present by exposing the nitrogen to potash 

 and re-mcasuring it. the final corrected N/S ratios are 

 as follows, the weight of a " normal litre " of nitrogen in 

 London being taken as 1-25144 grm. : — 0-436847, 0-436875, 

 0-436839, 0.436S57, 0-436S97, 0-436878, 0-436898. The mean 

 is 0-436870, and the greatest deviation from the mean is 

 I in 14,000. From this ratio the atomic weight of sulphur 

 becomes 2067. if nitrogen be given the very probable value 

 14-009.-— Dr. F. W. Edridge-Green : The 'relation of light 

 perception to colour perception. It may be easilv shown 

 that light perception and colour perception are quite 

 distinct. In fact, we can divide cases of colour-blindness 

 into two classes, according as the defect is (a) one of light 

 perception, or (b) one of colour perception or differentia- 

 tion without any defect in light perception. Of course, 

 both defects may be present in the same individual. The in- 

 vestigation of these two classes of defective vision is much 

 facilitated by the use of a spectrometer which the author 



NO. 2124, VOL. 84] 



has devised for the purpose. This instrument is a spectro- 

 meter so arranged as to make it possible to expose to 

 view in the eye-piece the portion of a spectrum between 

 any two desired wave-lengths. Tested with this instru- 

 ment, a normal individual will, as a rule, name six 

 distinct colours, namely, red, orange, yellow, green, blue, 

 and violet, and will mark out by means of the shutters 

 about eighteen monochromatic patches. Occasionally we 

 come across individuals with a greater power of differ- 

 entiating hues, to whom, as to Newton, there is a distinct 

 colour between the blue and violet, which Newton called 

 indigo. Such individuals will mark out a greater number 

 of monochromatic patches, from twenty-two up to twenty- 

 nine. Those who have defective light perception for 

 certain rays, with normal hue perception, behave exactly 

 in the same way as a normal-sighted person with those 

 rays removed or reduced to the same intensity, and not as 

 if a light-perceiving substance which was sensitive to rays 

 from a considerable range of the spectrum had been re- 

 moved. Those with defective hue perception mark out 

 with the spectrometer a smaller number of monochromatic 

 patches than the normal, and say that there are five, four, 

 three, two, or one colour instead of the normal six. They 

 behave in every way as if their colour sensations were 

 correspondingly limited. Therefore, if the normal be 

 designated hexachromic, then pentachromic, tetrachromic, 

 trichromic, dichromic, or monochromic correctly describes 

 their colour-vision. — M. G. Sykes : The anatomy and 

 morphology of the leaves and inflorescences of Weliuitschia 

 mirahilis. An account is given of the anatomy of the 

 leaves, and of the inflorescence axes, cones, bracts, and 

 flowers of both sexes. It is shown that the male and 

 female inflorescences are essentially similar in their method 

 of vascular supply and in their detailed anatomy, and it is 

 concluded that they are homologous. Various characters 

 suggest comparison with the Cycads and the MeduUoseae. 

 From the position of the embryo-sac relatively to the two 

 coverings of the ovule at various stages of development, 

 they are regarded as two integuments. The seed can be 

 closely compared with that of Lagenostoma ; its differences 

 from this primitive type are referable to changes dependent 

 on the evolution of siphonogamy and possibly insect 

 fertilisation. In both these seeds the free outer integu- 

 ment is regarded as a primitive character, in contrast with 

 the fused integuments of Cycas and Cardiocarpus. In 

 all these cases the entire vascular system appears to be 

 integumental. The connection between Welwitschia and 

 the Cycads, the Bennettitales and the Angiosperms, is 

 discussed. — Colonel Sir David Bruce, C.B., Captains 

 A. E. Hamerton and H. R. Bateman, and Captain 

 F. P. Mackie : (i) The natural food of Glossina palpalis ; 

 (2) mechanical transmission of sleeping sickness by the 

 tsetse-fly. — V. H. Veley and Prof. A. D. Waller : The 

 comparative toxicity of theobromine and caffeine as 

 measured by their direct effects upon the contractility of 

 isolated muscle. It is shown by measurements of the 

 contractility of isolated muscle that the toxicity of theo- 

 bromine, the base of cocoa, is greater than that of caffeine, 

 the base of coffee and tea, in the proportion of 3 : 2. The 

 introduction of a second methyl group into the oxy-purine 

 residue, namely, the formation of caffeine from theo- 

 bromine, thus diminishes the toxic value, a result which is 

 the converse of that observed in the case of the parafiinoid 

 alcohols. The toxic effects of coffee and tea extracts are 

 also studied, and it is shown that the effect of the former 

 is mainly due to the caffeine contents, and not to the 

 tannic acid. — Prof. W. B. Bottomley : The assimilation 

 of nitrogen by certain nitrogen-fixing bacteria in the soil. 

 — Prof. A. B. Macallum : The inorganic composition of 

 the blood in vertebrates and invertebrates and its origin. 

 • — Mary T. Fraser and J. A. Gardner : The origin and 

 destiny of cholesterol in the animal organism. Part vii. — 

 The quantity of cholesterol and cholesterol esters in the 

 blood of rabbits fed on diets containing varying amounts 

 of cholesterol. 



Cambridge. 

 Philosophical Societv. lunc 6. — Mr. W. Pateson, presi- 

 dent, in the chair. — H. H. Brindley : Further notes on 

 the procession of Cnethocampa pinivora. Lantern-slides 

 were shown illustrating observations on the procession of 

 the caterpillar of this Eupterotid moth, which infests the 



