824 
was described in 1883 from the underground waters 
of the Canterbury Plains, New Zealand, and other 
species of the genus and of allied genera were sub- 
sequently described from the surface and under- 
ground waters of Australia. In 1914, 
recorded a species of Phreatoicus from the mountain 
streams of Cape Colony. The characters and dis- 
tribution show that the family is an ancient one, 
and this was proved also by the discovery of a 
fossil species, from the Triassic beds of New South 
Wales, not very different from some of the existing 
species, 
NEw SQUALODONTS FROM THE MIOCENE OF MArRy- 
LAND.—R. Kellogg describes and figures the remains 
of two Squalodonts recently discovered in the 
Calvert Cliffs (Miocene), Maryland (Proc. U.S. Nat: 
Mus., vol. lxii., Art. 16). Of the two, one is definitely 
referred to a new species, Sgualodon calvertensis, the 
other is at present indeterminate. The introduction 
to the paper gives a good summary of the history 
of the Squalodonts generally, and is followed, under 
the misleading and erroneous title of ‘“‘ Nomenclature,” 
by a descriptive list of the various species and a 
“e key.” 
the U.S. Geological Survey.(1922),; H. Ries and other 
Eastern United States.’ Of all rocks, clays probably 
offer the greatest difficulties to petrographers. W. 
Maynard Hutchings (Geol. Mag., 1890, p. 266) prepared 
thin films from clays for microscopic examination, 
retaining the particles in their relative positions. 
Except where lamination has to be studied, no great 
advantage arises from this method, and H. C. Sorby 
(Quart. Journ. Geol. Soc., Proc., 1880) did good work 
on separated grains, which could be pressed down or 
rolled over under the cover-glass. Allan B. Dick’s 
“smudge ”’ -method (‘ Kaolin, China-clay, etc.,” 
Mus. of Pract. Geology, p. 261, 1914) keeps the 
constituents matted together for comparison of their 
optical characters, and something similar seems to be 
effected by the squeezing-out process adopted in the | 
American researches (Bull. 708, p. 293) by R. C. 
Somers. His photographs give the impression of 
coherent sections; but probably only a massing of 
the mineral particles can be inferred. Chlorite is 
omitted from the list of minerals observed, and 
Hutchings failed to recognise it even where altered 
biotite formed an important constituent of his clays. 
He found it, however, abundantly in slates derived 
from the decay of basic igneous rocks, such as his | 
“ash-slates,’’ and H. B. Milner, in his recently pub- 
lished “‘ Introduction to Sedimentary Petrography,” 
regards its appearance in loose sediments as due to 
the breaking down of slates and schists. It would 
seem that the green hydrated products from ferro- 
magnesian silicates, wisely called by various authors 
“ chloritic matter,’’ should find a considerable place 
as constituents of many clays and sandstones, though 
probably in a highly comminuted form. R. F. 
Somers. recognises halloysite and diaspore in the 
American clays, in addition to the prevalent flakes of 
kaolin. The white material known as “ indianite ”’ 
from south-central Indiana is discussed by W. N. | nici 
| structure is thus obtained. Besides its rust-resisting 
Logan (Bull. 708, p. 147), who finds it to consist 
mainly of the hydrous aluminium silicates, halloysite, 
and allophane, In the field it is associated and inter- 
locked with a sandstone of Pennsylvanian age, blocks | 
of which graduate into the clay, and H. Ries (p. 161) 
comes to the very interesting conclusion that the 
indianite has arisen from replacement of quartz 
spreading from the underlying pyritous shales. 
NO. 2798, VOL. 111] 
Barnard | 
NATURE 
| 

| g7o° C. 
pebbles, through the action of aluminium sulphate | 
[JUNE 16, 1923 
Stokes’s Law or Fatt oF A SPHERE.—In the 
issue of the Proceedings of the United States Academy 
of Sciences for March 15 Prof. Millikan, now of the 
California Institute of Technology, gives a summary 
of a theoretical and experimental investigation of 
the law of fall of a sphere in a viscous fluid which will 
be published later in full in the Physical Review. On 
theoretical grounds he shows that the viscous resist- 
ance to the motion of a sphere of radius a with 
velocity v in a fluid of viscosity x and mean free path / 
must be 6zanv/(1 +A‘’l/a) where A’ is a constant 
which must decrease as the density of the fluid de- 
creases. In order to express this decrease he writes 
A’=A +Be-@!/ where c is a constant. He finds that 
his experiments with drops of different materials 
falling in gases of various constitutions and densities 
are all reproduced by the complete formula with 
values of A+B which differ by not more than 3 per 
cent. from each other. 
CurvE Firtinc.—In many physical problems, 
experimental data give the numerical values of a 
function at regular or irregular intervals of a variable 
on which the function depends. Often it happens 
that these experimental values indicate that the 
: | function starts at zero, rises to a maximum and then 
THE CONSTITUTION OF CLAYS.—In Bulletin 708 of | 
falls again to zero. Frequency distributions, river 
authors describe “‘ The High-grade Clays of the | gange readings, end) certain ‘phys oe 
data define functions of this type. In such cases it 
is often necessary to obtain a theoretical function 
which gives a reasonable fit to the data and can at 
the same time be integrated. The constants occurring 
in the theoretical equation should also be capable 
of calculation without such laborious computation as 
to make the work impracticable. In a pamphlet, 
‘“ A Method of Curve Fitting,” issued by the Physical 
| Department of the Egyptian Ministry of Public 
Works (Cairo, Government Publications Office, 1922, 
P.T. 5), Mr. S. Krichewsky explains an equation 
which has been found to include a wide range of 
observations, including frequency distributions. The 
equation is 
z= f(x) =dy/dx=ky™(a -y)", 
ol. x 
wherein f(I,)=f(ls)=0, a= | 4(x)dx, y= / f(x)dx, so 
Jl i 
that z is the ordinate at x, and y the area from /, 
to x. Mr. Krichewsky’s method certainly possesses 
limitations due to the small number of free constants 
contained in his equation. Much of the pamphlet 
is concerned with the calculation of the constants to 
fit the equation to sets of observed data. 
HARDENING STAINLESS STEEL.—Messrs. Auto- 
matic and Electric Furnaces, Ltd., of Farringdon 
Road, London, manufacture a special electric furnace 
for hardening ‘“‘stainless’’ cutlery. The demand 
for stainless cutlery has been affected adversely by 
the fact that in many cases a permanent edge cannot 
be obtained by its use. To get over this difficulty 
the steel is treated as follows: it is first heated in — 
the furnace to a temperature of 970° C. (as shown 
by the pyrometer) and then cooled in air. It is 
next reheated to the same temperature and quenched 
in water. Finally, it is tempered to 220° C. in an 
electrically heated muffie. A very fine micro- 
qualities, stainless steel has a thermal conducting 
coefficient less than one-third that of pure iron. 
It is an excellent material for making permanent 
magnets for use in positions where freedom from 
corrosion is an advantage, as when quenched at 
it has a large coercive force and great 
remanence. It is also useful for making mirrors 
of all kinds. 

