March 6, 1902] 



NA TURE 



415 



weights " have been calculated, but presumably the ordinary 

 formula has been employed. It is important to note that in 

 seven out of the nine boiling-point measurements carried out 

 with magnesium sulphate, the concentration of the solution 

 varies from about 15 per cent, to over 40 per cent. In other 

 experiments with solutions of cane sugar, concentrations varying 

 from 20 grams to 290 grams of sugar per 100 grams of water 

 have been employed, the calculated "molecular weight" 

 (theory = 342) decreasing from 360 to 212 at the highest 

 concentration. 



To attribute to the numbers calculated by means of the 

 simple boiling-point formula for solutions of such concentra- 

 tions the significance of molecular weight values can scarcely 

 be regarded as justifiable. In so far as the ordinary freezing- 

 point and boiling-point formulx rest on a thermodynamical 

 basis, they only hold good for ideal solutions ; for such solu- 

 tions the formula in question will give molecular weight 

 values, but it has yet to be shown that the numbers calculated 

 by Kahlenberg from his experiments can be taken as repre- 

 senting molecular weights. 



For moderately concentrated solutions which no longer 

 satisfy the requisites of an ideal solution, distinguished by the 

 properties that no heat is evolved or absorbed and no change 

 of volume takes place when it is diluted, Ewan, amongst 

 others, has deduced an expression connecting the osmotic 

 pressure with the lowering of the freezing point, the formula 

 resting on a thermodynamical basis. This formula, when used 

 for the calculation of molecular weights, gives, even with 

 solutions- containing as much as 40 per cent, of cane sugar, 

 values scarcely differing from theory (342), whereas the simple 

 freezing-point formula for a solution of the concentration 

 mentioned gives 275. 



Another reason advanced by Kahlenberg to prove the inad- 

 missibility of the ionic theory is the lack of agreement between 

 the numbers representing the degree of dissociation as calculated 

 from the conductivity on the one hand and the freezing- or 

 boiling point on the other. If for the reasons previously stated 

 the calculations based on the boiling-point and freezing-point 

 measurements have little significance so far as the ionic theory 

 is concerned, it is obviously impossible to effect the required 

 comparison. Furthermore, it seems questionable svhether the 



numbers calculated by the formula o = '^^— really represent de- 



grees of dissociation. The formula involves the as yet unproved 

 and scarcely probable assumption that the ionic velocities are the 

 same in solutions of all possible concentrations. The develop- 

 ment of the ionic theory is by no means conditioned by the 

 validity of such a formula. So far as Kahlenberg's measure- 

 ments are concerned, the comparison between the results of the 

 boiling point and conductivity measurements is moreover, im- 

 possible, except in the case of the binary salts, since the range 

 of concentrations employed is quite different. 



Although, therefore, the publication contains a large number 

 of valuable empirical data, yet it cannot be allowed for one 

 moment that the ionic theory has been shown to be untenable. 

 It is far from the wish of the writer to minimise the difficulties 

 which do admittedly confront the theory of electrolytic dissocia- 

 tion. It must not, however, be supposed that the theory has 

 received its final and complete form ; the possibilities of its 

 rational expansion and development to explain existing irre- 

 gularities are far from being exhausted. A warning note may 

 be sounded against a too ready assumption that new experi- 

 mental data prove the untenability of the theory without very 

 careful consideration of what exactly is, and is not, stipulated by 

 the theor)'- H. M. Dawson. 



The Yorkshire College, Leeds. 



Birds attacking Butterflies and Moths. 



I WAS much interested in the letter in Nature of January 16 

 on the frequent capture of butterflies in India by the King Crow, 

 as some years ago I experimented with a captive bird of this 

 species, and found that it avoided " warningly-coloured " 

 butterflies when possible, and was deceived by mimicry 

 (/. A. S. B., ii. 1897, p. 651). 



With regard to the capture of butterflies by bush-haunting 

 birds which do not take thein on the wing, I pointed out as long 

 ago as 1895 (J. A. S. B., ii. 344) that the common Babbler 

 Craleropus canorus was likely to meet with butterflies in re- 

 pose, and proved experimentally that it dislikes the " warningly- 



NO. 1688, VOL. 65] 



coloured " species. But I could then give no positive evi- 

 dence that it does, as a matter of fact, attack butterflies on its 

 own account, though it will take them if thrown in its way. 



I therefore give here the results of a few experiments 

 which, in my opinion, show that this bird also naturally preys 

 on butterflies. 



In March last year I gave to a wild-caught bird of this species 

 a Danais limniacc together with a Junonia. The bird took and 

 ate the latter ; I then removed the Danais. 



I have just now been offering three specimens of Danais 

 gciiulia, together with three plain brown butterflies, to three 

 wild-caught adult Babblers placed in separate cages. 



Two of the birds disregarded the Danaids until they had eaten 

 the other butterflies, and then did not attack them eagerly or eat 

 them (except the abdomen in one case), although they had no 

 food in their cages at the time. 



I conclude, therefore, that they were last year's birds, which 

 knew and disliked D. limniace, and the present two D. genutia, 

 from previous experience in catching and tasting butterflies 

 when wild. 



The third bird experimented with to-day attacked its specimen 

 of D. genutia first, but soon left it to eat the other butterfly 

 given ; nor did it tear the Danais to pieces as did the others, 

 although, like them, it had no other food in its cage. 



Either, then, this bird had forgotten its wild experience, or, 

 what is more likely, it had never happened to catch D. genutia, 

 and so knew nothing about this species, which it evidently dis- 

 liked, from what has been said above, although it was not im- 

 pressed by the *' warning colours." 



In my previous experiments with this Babbler I did not ob- 

 serve the same precautions, when first offering the butterflies to 

 the birds, as I did in these later experiments, so that the results 

 I obtained, although sufficiently demonstrative of the preferences 

 of the species, threw no light on the individual experience of the 

 specimens experimented with. F. Finn. 



Indian Museum, Calcutta, February 6. 



Si cela peut interesser vos lecteurs : . . . dans une traversee 

 de la Mer des Caraibes sur le steam, angl. Mariner, en Mai, 

 1 886, nous fumes, par un temps calme, assaillis par un grand 

 nombre de tout petits oiseaux, bien qu'a une assez grande distance 

 de la terre, invisible. lis poursuiv.aient de petits papillons 

 qu'ils venaient happer au vol jusques sur mes genoux. J'etais 

 assis tres fatigue et un peu inerte sur la dunette. Je ne pouvais 

 songer a determiner oiseaux ni insectes. Les matelots laissaient 

 faire. La brise fraichit et tout ce petit monde disparut en un 

 clin d'ceil. Que sont-ils devenus ? Ai). Nicolas. 



Angers (M.-et-L.), le 22 Fevrier. 



On Prof. Arrhenius* Theory of Cometary Tails and 

 Aurorae. 



In the more or less popular accounts which have recently 

 been given of Prof. Arrhenius' theory of cometary tails and the 

 aurorce, it is generally stated that the smaller the diameter of 

 the corpuscle upon which the light is falling the greater the 

 excess of light-pressure over gravitational force. This explana- 

 tion, however, holds only so long as the diameter is greater than 

 the wave-length of light. If the diameter becomes of the same 

 order as the wave-length, the ratio between light-pressure and 

 gravitation follows an entirely different law. This has recently 

 been demonstrated by Prof. Schwarzschild by an exhaustive 

 mathematical treatment of the question in a paper entitled " Der 

 Druck des Lichtes auf kleine Kugeln und die Arrhenius'sche 

 Theorie der Cometenschweife '' (Sitzungsbcrichte der k. b. 

 .■\cademie der Wissenschaften zu Miinchen, 1901, Heft iii.). 

 The conclusions arrived at in this paper are of considerable im- 

 portance in so far as they show that the effect of gravitation is 

 exceeded by that of the pressure of light only so long as the 

 diameter of the corpuscle is greater than about 0'07(U. For this 

 limiting value the two forces are exactly balanced ; but for 

 smaller values of the diameter the light-pressure becomes rapidly 

 less, so that it is then always exceeded by gravitation. It would 

 appear from Prof. Schwarzschild's computations that the globular 

 corpuscles thrown off in the tails of comets should have diameters 

 not smaller than o'O?^ and not exceeding i'5;c/, supposing the 

 specific gravity of the corpuscle to be that of water. Now these 

 values far exceed the limits assigned to the dimensions of the 

 molecules. According to our present knowledge, based on 



