February 9, 1922] 



NATURE 



179 



rere more frequent than specific variations. Hitherto 

 exception has been found to the required rule. 



>ne group of fungi tested (Hymenomycetineae) gave 

 line with a slope very little exceeding unity (i-o8), 



It the figures found for flowering plants lie between 



le narrow limits 1-38 and 1-64, with an average of 

 ibout 1-43. Snakes and lizards both give a figure 



;ry near 1-50, and the Chrysomelidae about 1-37. 

 The development of a more com- 



)lete theory may in some degree 



lodify conceptions and interpreta- 



ions, but the results so far ob- 



lined suggest that the basic prin- 



iple put forward is correct. 



Inasmuch as all families, both 

 )f plants and animals, show the 



ime type of curve, whether 



graphic or logarithmic, it would a 

 Appear that in general the manner g 

 which evolution has unfolded ^ 



self has been relatively little « 



Fected by the various vital and E 



ther factors, these only causing 5 

 deviations this way and that from ^ 

 the dominant plan. And since, 

 assuming that genera " throw " 

 other genera and species, it was 

 predicted that the logarithmic 

 curves would be straight lines, 

 and it was then discovered 1 



that they actually were so, it is 

 probable that the assumption was 

 rorrect. But if this be so, then not only 

 must evolution have been by mutation, but 

 it must also have been, as one of us has 

 contended for many years, by mutations that were 

 at times of rank sufficient to give rise to Linnean 

 species, genera, or even families. Not only so, but 

 evolution must have proceeded on the lines of 



Guppy's theory of differentiation, the larger genera, 

 and the species of larger area, being the parents of 

 the smaller : i.e. it must have proceeded on the whole 

 in the reverse direction to that postulated by the 

 Darwinian theory, as one of us has long maintained. 

 Finally, it is clear that geographical ^distribution 

 has been largely mechanical, the general effect 

 of the many factors that are operative being to cause 



lot? (Number of species) 

 5 ^ 10 ^ ^ 20 30 100 



Number of species 

 Fig. 4. — Log. curve for chrysomelid beetles. 



species to spread at a fairly regular speed (differing 

 for each), so that spread forms a measure of age. 



Space does not permit of detailed argument, which 

 must be left for forthcoming books ; but a couple of 

 hours' work at statistics of genera (by sizes) will 

 suffice to make clear the general position taken 

 up. 



Some Problems of Long 

 By Dr. J. A. 

 II. 



ANOTHER cause operating to effect a separation 

 of the positively and negatively charged dust 

 is found in the viscosity of the atmosphere. Roughly 

 speaking, the viscosity of a gas is that quality of it 

 in virtue of which fine particles experience a resist- 

 ance in moving through it. Maxwell showed long 

 ago that the viscosity of a gas is independent of the 

 pressure over wide limits. Crookes continued these 

 researches and demonstrated that between atmo- 

 spheric pressure and a pressure of about one ten- 

 thousandth of an atmosphere the viscosity remains 

 constant, but that when the pressure falls below this 

 last figure the viscosity very rapidly decreases to 

 zero. Again, both Maxwell and Crookes found that 

 the viscosity of hydrogen is about half that of 

 oxygen or nitrogen. The viscosity of air at 760 mm. 

 is 000018 C.G.S. units. 



Sir George Stokes proved that if a small sphere 

 of diameter d and density o- is falling through a 



1 Continued from p. 143. 



NO. 2728, VOL. 109] 



distance Radio-telegraphy.^ 



Fleming, F.R.S. 

 gas of density p and viscosity /x under the action 

 of gravity it will attain a final velocity v such that 



where g is the acceleration of gravity. This ex- 

 plains the extremely slow rate of fall of water 

 particles constituting clouds, and also the very slow 

 settlement of fine dust particles through air. 



The positively-charged solar dust particles are 

 probably larger than the negatively-charged par- 

 ticles, as the latter consist of electrons having con- 

 densed round them molecules of gases, probably 

 hydrogen and helium, gathered from the solar 

 chromosphere. Accordingly the negative ions will 

 be brought to rest before the positively-charged 

 particles and gas viscosity will assist the separation. 



But Stokes's expressions apply to smooth spheres 

 and not to irregularly shaped particles. Also, if 

 the diameter of the particle is much less than the 

 mean free path of a gas molecule, the expression 



