John Elihu Hall.

The American journal of science online

. (page 11 of 61)
Online LibraryJohn Elihu HallThe American journal of science → online text (page 11 of 61)
Font size
QR-code for this ebook

orange-red color, which is the peculiarity of the isomorphous

This bright red mixture was first observed in connection
with the fractional crystallization of about 600 g. of the salt
CSjTeCle. The object of this operation was to find if in this
way any separation of tellurium into different elements could
be effected, and it is sufficient to say that the results of a very
extensive fractionation were entirely negative in regard to any
such separation.

The salt used for this systematic crystallization was prepared
from crude tellurium, as it was considered best to purify care-
fully the tellurium of the end-products, rather than the whole
of it. When the fractionation had been carried on to a con-
siderable extent, it was observed that the products at the solu-
ble end showed a bright red color, and since this substance was
not recognized as any known caesium double chloride, it
attracted attention as a possible indication of the presence of
the much-sought impurity in tellurium. However, a qualita-

• WeUs aod Metzger, Amer. Chem. Jour., xxvi, 268, 1901.
t H. L. WeUs, this Journal (3), xlvi, 180, 1893.
JH. L. Wheeler, ibid., xlv, 267, 1893.

Digitized by VjOOQ IC

104 n. L. WeUs — Color-Effect of Isomorphous Mixture.

tive examination of the red product showed nothing unuBual
in it except the presence of much lead. Then it was found
that it could be readily prepared by dissolving csesium, lead
and tellurium salts in hot aqua regia and cooling or evaporat-
ing to crystallization. It should be explained that the pres-
ence of the lead tetrachloride corapouna in the mother-liquor
from the fractional crystallization was due to the presence in
the hydrochloric acid solution of a considerable amount of
nitric acid which had been used in dissolving the crude tellu-
rium. The lead was an impurity in the latter.

The separate salts Cs^PbCl^ and Cs^TeCl, were prepared
repeatedly from aqua regia solution, but they gave invariably
pure yellow products. Dnder precisely the same conditions
when both tellurium and lead were present the products were
always red, and there was no very marked variation in this
red color when the proportions of lead and tellurium were
changed considerably. The red products formed octahedral
crystals like the yellow salts, and they were of similar size.

it was suspected that the red suDstance might be a triple
salt of ceesium, tellurium and lead, but analyses of several
crops showed that there was no constant relation between the
lead and tellurium, and that recrystallization from aqiM, regia
changed the composition of a product very much by increas-
ing the proportion of the lead compound. Therefore it must
be concluded that the products were isomorphous mixtures.

The following analyses were made of separate crops of the
red mixture, where v was obtained by a single recrystalliza-
tion of IV :

I II lu IV V

C8,PbCl. 41-2 65-3 47-7 57-0 884

C8,TeCI, 57-6 44-3 61-3 48-6 11-4

It is to be noticed that when the two yellow salts in separate,
very small crystals are mixed, either dry or under hydrochloric
acid, there is no development of any red color, so that it
appears that light in passing from one kind of crystal to the
other kind gives no unusual effect. Hence it is evident that
the effect under consideration is due to the crystallization of the
two things together.

Sheffield Laboratory, New Haveu, Conn.,
December, 1911.

Digitized by VjOOQ IC

Rogers — Lorandite from the Rambler Miney Wyoming. 105

Abt. XIV. — Lorandite from the Rambler Mime^ Wyoming ;
by Austin F. Kogees.

I AM indebted to Mr. Berger, of Placerville, California, for
an interesting specimen from the Rambler mine, near Encamp-
ment, in southern Wyoming. This specimen consists of dark
fine-grained massive pyrite, upon which are implanted barite
crystals and well-formed crystals of orpiment. With the orpi-
ment and barite are associated several orange-red realgar crys-
tals and a single deep red crystal of what proved to be
lorandite or thallium metasulfarsenite, TlAsS,. This is the

second known occurrence of lorandite, the original locality
being Allchar in Macedonia.*

The crystal mentioned is an imperfect one, of about 4""
size, bounded by the faces of a rhombic prism with interfacial
angles of about 90° (calc, 93°) and by tnree cleavages in one
zone, which is at right angles to the prism zone. Using
Goldschmidt's orientation,t the prism faces constitute the
jllOf form and the three cleavages are parallel to |100],
j 001 1, and }I01}. The accompanying figure (plan and siae
elevation) gives an idea of the crystal. All the faces but
1 110 1 are cleavages. The following angles measured on the
reflection goniometer prove that the crystal is lorandite. The
first mentioned angle was measured on a detached fragment,

Measured Calculated

100(clv.) :001(clv.)= 52°49' 52°27'

001(clv.) : 101(clv.)=: hVir 51°49'

while the other angle was measured by mounting the matrix
specimen on the goniometer, as it was feared that the crystal
would go to pieces if detached from its matrix. The cleavage
parallel to |100| is very perfect, that parallel to |001} goo^S,
and that parallel to |I01} fair. The luster is adamantine on

♦Krenner, abstract in Zeitschr. Kryst. Min., vol. xxvii, p. 98, 1897.
fZeitachr. Kryst. Min., vol. xxx, pp. 272-294, 1899.

Av. JouB. Sci.— Fourth Series, Vol. XXXIII, No. 194.— February, 1912.

Digitized by VjOOQ IC

106 Rogers — Lorandite from the Ramhler Mine, Wyoming.

the cleavaffe faces, but the prism faces |110| are dull. Even
if bright tnev could not be measured on account of the close
proximity oi the matrix.

Fragments are prismatic, non-pleochroic, and have parallel
extinction. Lorandite is monochnic, but the cleavages are in
the zone of the ortho-axis and so have parallel extinction.

On charcoal lorandite fuses easily to a black globule, color-
ing the flame bright green. It gives a green flame when fused
on platinum wire and alloys with the platinum. In the closed
tube it fuses to black globules, giving a black and red subli-
mate and also minute colorless adamantine crystals of As,0,.

The lorandite is soluble in nitric acid, turning yellow. With
chloroplatinic acid the solution gives a light yellow precipitate
(TljPtClJ. After evaporating off the nitric acid, potassium
iodide gives a yellow precipitate (Til). The nitric acid solu-
tion with hydrochloric acid gives a white precipitate (TlCl).

With the spectroscope this white precipitate of thallium
chloride gave a single bright line in the green. With a pure
thallium salt the green line appeared at exactly the same

Although the blowpipe and chemical tests were made with
a very limited amount of material consisting of minute
detached fragments, the identity of the mineral with lorandite
is well established. The spectroscope proves it to be a thallium
mineral and the goniometrical measurements prove it to have
the crystal form of lorandite. The blowpipe and chemical
tests are confirmatory.

Stanford Universityy California,
Oct. 1911.

Digitized by VjOOQ IC

C. Barm — Rate of Decay of Nuclei.


Abt. XV. — The Rate of Decay of Different Sizes of Nuclei^
Determined hy Aid of the Coronas of Cloudy Uondenaor
tion ;* by C. Babus.

A SERIES of experiments into which I entered at some length
and which, though from the natnre of the experiment they
cannot lay claim to a high order of pi*eci8ion, nevertheless lead
to very definite results, are contained in the following table,
which is an example of many similar results. These nuclei
were produced by X-rays (of moderate intensity) in an alumi-
num-covered fog chamber, with short exposures to the radia-
tion. The data show that no persistent nuclei of the usual
large type appreciably occur, since these, if present, require
almost no supersaturation of moist air, for condensation.
Moreover, the vapor nuclei of dust-free wet air are not caught
in my fog chamber, in the presence of ions, or else their num-
ber is specially determined and small in comparison with the
nuclei here obtained by the X-rays. This premised, the table
gives the relative drop of pressure ijp/p from p^ the lapse of

Table. —Decay of different sizes of nnolei (radios r\ the number n present
being determined by the coronas of cloudy condensation, dn/dt = — bn*.



Lapse, sec.








































^ ^

^ _

, ,









D. f. air











_ .



„ _

. .



. ,





^ ,


i). f. air




'^* Abridged from the Report to the Carnegie Institution of Washington,
D. C.

Digitized by VjOOQ IC

108 C. Barns — Rate^ of Decay of Nuclei.

time, i^ between the instant the X-radiation is cut off and the
instant of the exhaustion made to catch the remaining ions,
together with the number of nuclei, n, per cu. cm. caught, as
estimated from the apertures of the corresponding coronas.
The column t is an estimate of the size of nuclei, obtained
from the drop of pressure Sp, and h is the rate of decay. In
the table, the nucleation discovered in dust-free wet air at any
drop of pressure 8p/p, in the absence of X-radiation is always
relatively small, even at the highest exhaustions made. Never-
theless, the persistence of these nuclei, surviving as much as
even ten minutes after exposure, is remarkably large, as if all
nuclei (vapor nuclei and ions) were eventually caught together.
The results, in fact, often point to an almost indeiinite persist-
ence. On the other hand, at the lower exhaustion, the number
of nuclei soon vanishes. The rates of decay are determined by
the usual equation

A = h^ or dn Idt = — bn^

Briefly, therefore, at a suflSciently high drop in pressure,
hp/p = -35, the nuclei produced in the presence of ions in dust-
free moist air by the (moderate) X-rays, decay at phenomenally
small rates, or are almost indefinitely persistent, whereas larger
nuclei decay faster in proportion to their size.

Similar facts were brought out by all the other tests with
coronas. As ip/p decreases, or r the radius of the nucleus
increases, h increases at a rapidly accelerated rate, from very
small values b = 10 "^ to enormous values. The small coeffi-
cient b is less than 1/10 the normal value obtained for ions
with the electrometer (of the order of 10~"), while the large
values are nearly fifty times the normal value.

The interpretation of these results is made difficult by the
variability of the X-ray bulb. It is safe to assert, however,
that the large ions or nuclei produced by the X-rays in dust-
free wet air vanish with relatively enormous rapidity, whereas
the very small nuclei are almost indefinitely persistent, and
that there is a definite relation between the rate of decay and
the size of nucleus. If, therefore, we regard these nuclei as
water droplets of different sizes, evaporation is rapid until a
limiting diameter depending on the intensity of radiation is
reached, after which evaporation nearly ceases. It is also
probable, that the limiting diameter increases with the intens-
ity of radiation, so that with strong X-rays almost no super-
saturation is required. If, therefore, the X-rays produce any
chemical body which may go into solution, as has been sup-
posed, the greater or less abundance of this body, supplied by
f greater or less intensity of X-radiation, would account for
arger or smaller persistent nuclei.

Digitized by VjOOQ IC

C. Barus — Displacement Interferometer. 109

Art. XVI. — A Displacement Interferometer Adapted for
High Temperature Measurement^ Adiabatic Transformor
tions of a OaSy etc, ;* by C. Babus.

1. EUiptic Interferences. — Interferometl'y by displacement
has an advantage inasmuch as the observer never loses the
ellipses, even wnen the displacement is sudden. Their center
may always be brouffht back again to a given spectrum line, by
the micrometer. Moreover, since

N^ = 6/AC08 R p -^ = e cos i2 {fi-\-2B/coB* B)

cos ./v clA

where JV^ is the reduced micrometer reading, e the thickness,
fft the index of refraction of the glass plate of the grating, for
the wave length X, B the angle oi refraction, and where
fft = A+B/X*, the sensitiveness, may be regulated by decreas-
ing the thickness of the grating, «, by aid of a compensator of
tliickness e'y for the virtual thickness is now e—e\ Hence,
since for radial motion the sensitiveness per fringe across any
given Fraunhofer line is


this may be combined with the shift of ellipses controlled by
iTc, in any ratio. The limit of this procedure is conditioned
by the size of the ellipses or the available size of the field of
the telescope, since when e—e' approaches zero the ellipses
become enormous.

Furthermore it has been shownf that the quantity d^i/dX
occurring in the value of N^ may be computed preliminarily
from observations of ^JVc = ^c - Jf^\ between definite Fraun-
hofer lines, particularly when the angles of incidence I, and
of refraction i?, are small. In such a case the constant 3eB = fi
nearly, where (7 = 0)

A^=J5(l/V-l/V.),AiV; = i8(l/V-l/VK).
Finally if AiT is the motion of the micrometer to bring the
center of ellipses back to a given line
where e' is the thickness of the compensator. If e^ is large, one
may expect to distinguish between the indices of refraction of
a birefringent crystal, when the source of light is polarized.
Again when the arms or the interfering beams of light are
long, the refraction of a gas and its relation to temperature and
pressure are determinable.

* Abridged from the Report to the Carnegie InstitTition of Washington.
Reprinted by permission,
f Carnegie Publications, No. 149, 1911, chap, v, §44.

Digitized by VjOOQ IC

110 G. Bams — DUplacement InterferomeUr.

In view of these advantages among others, I have con-
structed a definite form of apparatus for displacement inter-
ferometry, specially adapted for general observations, such for
instance as I have m view with fog particles. The apparatus
is to be light, portable, rigid, with relatively long distance
between the opaque orthogonal mirrors M and N^ and the
oblique mirror or grating, as well as height of mirrors above
the arms, and with an easy adi'ustment for different angles of
incidence /, large and small. l!n the following apparatus, figs.
1, 2, and 3, the distance between the center and either remote
mirror is about 35*^. It may easily be enlarged many times.
The angles of incidence / = 15°, 45 , and 75°, are available at
once for the given bi*aces, though of course other angles may
be used.

The long arms and feet of the apparatus, which in general
form is naturally much like a spectrometer^ are made of 1/4
inch gas pipe, and the braces are neavy strips of tin plate, bent
so as to be U-shaped in cross section, mnch like umbrella
steel, with the ends bolted down. In the drawing (of which
fig. 1 is the plan and fig. 2 the elevation) the axles are cylindric
or slightly conical. In my own apparatus sufficient rotation,
180°, of the parts was secured by ordinary well-cut gas pipe
screws. The long arms of gas pipe a, c, d^ e are not only con-
venient for the attachment of objects to be examined, by ordi-
nary clamps, but they admit of a circulation of cold water, so
that their lengths remain invariable whatever be the tempera-
ture of the environment of air.

The tripod, figs. 1 and 2, carries a standard Q of 3/8 inch

fis pipe, which is secured snngly by the cross-coupling B.
rom this the horizontal rigid arms, a and c, lead respectively
to the collimator A and to the slide micrometer O and they are
screwed into B parallel to the plane of fig. 1. The arm e
which carries the telescope E must be revolvable around Q^
a wide axle PP' and braces V diverging as they approach Q
sufficing for the purpose. The telescope is used for reading
only, and need not be clamped. It must, however, be quite
firm so as not to shake the instrument.

The standard Q is prolonged above the cross-coupling B as
shown at j and the graduated plate at K (for measuring the
angle of incidence 1) rotates around Q prolonged. The plate
may be clamped by the set screwy''. Kadially to K^ the lat-
eral arm d is bolted below the plate at /'. D carries the
opaque mirror J/, which thus rotates around Q with jff".

The adjustment chosen is such that the parts J/, iT, G (the
grating), the telescope E and the collimator A may be dis-
placed upward several inches, in the clear. It is thus possible,
for instance, to place the fog chamber between MG or NG.

Digitized by VjOOQ IC

C. BaruB — DUplac&inent Interferometer.


The inside of j or Q prolonged is ground and receives the
hollow cylindrical plug i of the table / and this may be
clamped by the set screws i\

Upon 1 stands the grating O secured by the screw g to the
ring-shaped support SU whicn reposes in an annular gutter in I
on wiree leveling screws. Moreover a spiral spring (not shown)

Figs. 1, 2, 8.

in the- inside of the hollow conical tube i^ pulls down the ring
H firmly upon /, so that nothing is liable to fall on transporta-
tion. I'he ruled surface of the grating G is toward the light
or collimator A and in the axis of rotation. The adjustment
need not be very accurate. The rulings are parallel to the

Certain details of the parts of the apparatus may now be
pointed out. The collimator A may be raised or lowered on
its vertical stem or clamped in any position by aid of the wing
nut, fl^', on a longitudinally split tube. It may also be slightly

Digitized by VjOOQ IC

112 C. Baru8 — Dieplacement Intefferometer.

inclined ta the horizontal, either by the hinge indicated in the
figure, or by the special device shown in fig. 3, where the tele-
scope or collimator reposes on Y*s made of strips of elastic
brass aa. These are so adjusted that the end of T at a' is
naturally higher than at a. The ring b and the thumb screw c
then lowers this end against the upward pressure of all the
springs. Reading telescopes so mounted are firm and the
device is very convenient if a slight inclination is to be
imparted. They are removed by loosening <?, and slipping T
out of the ring, b.

The tube a ends on the left in the cross-coupling, which
also adniits the adjustable standard a" and affords an attach-
ment for the braces bb (U-shaped in section), the other ends
of which are bolted down to the nearer feet of the tripod.
Thus A is held suflSciently rigid by the braced system aa bb.
An inch or a 3/4 inch objective and a 6 inch focus is sufficient
and by reason of its lightness perhaps preferable to a larger
and heavier tube. The slit may usually be opened about 1/2™'°.

In a similar way steadiness, elevation and inclination of the
telescope £ is secured, the tube e and e" (adjustable foot) and
the braces b" terminating in the cross-couphng e^ as has been
suggested. An inch objective and a 6 incn focus is adequate.
Cross hairs are convenient but not necessary, as the spectrum
lines are available when sunlight is used. If the arc light is
used, strong sodium lines are usually in the field with the

The opaque mirror M is controlled by three leveling screws
(horizontal and vertical axes) and a suitable spring in the cap-
sule D. It is adjusted vertically like the tetescope and kept
firm by the tubes d and d'' (adjustable foot) and braces S' b\ all
parts meeting at the cross-coupling d\ The braces S' V are of
equal length. Hence they may be bolted down to two of the
feet of the tripod in succession, while the tube d together
with the plate IC take the three positions at 30°, 90°, and 150°
to the rod a. The grating G does not turn with ^ but must
be specially adjusted to corresponding angles of 15°, 45°, and
75°, as easily determined by the refiected rays.

Finally the slide micrometer is sustained by the tubes c and
c'^ (adjustable foot) and the braces bb, all parts meeting in the
cross-coupling </. The latter carries the table Z, to which the
slide micrometer C, with its drum at i^, is bolted down. ^ is
the opaque mirror adjusted by three leveling screws and a
spring (horizontal and vertical axes) within the capsule JS. The
slide should have from 1 inch to 2 inches of clear play and its
displacements should be determinable to about -00005"". The
opaque mirrors M and iT may both be silvered on the back
and thus last indefinitely.

Digitized by VjOOQ IC

C. Baf'us — Displacement Interferometer*. 118

Since the telescope E rotates both around its own axis e" and
around the standard Q^ elaborate centering of the grating O
is not usually necessary. The latter is mounted between strips
of cardboard or wood and secured by the screw g^ the brass
clutclies being about twice as far apart as the thickness of the
grating. In other words, the grating may be slightly moved in
a direction normal to itself.

To adjust the parts, sunlight (preferably) or arc light is
passed into the widened slit of the collimator, in a dark room,
so that the spots falling on the mirrors M and N (tiie grating
being suitably turned) and on the objective of the telescope E
are seen and the diflferent reflected images brought nearly into
coincidence. A further adjustment is then made through the
telescope jff, two of the usual four images of the slit (now nar-
rowed) being placed in coincidence horizontally and vertically
by manipulating the leveling screws on B. Specks of dust, or
nicks in the slit, greatly facilitate this adjustment. The tele-
scope is then turned to the diffmction spectrum, preferably of
the first order, and the drum actuated till the interferences
appear. Naturally the distances GN and GM are to be
approximately equal to begin with. The solitary ellipses are
best for general purposes and they usually correspond to unde-
viated yellowish and bluish single slit images. The multiple
slit image is to be avoided. If the rings are not quite centered
in the spectrum, they may be made so by cautiously adjusting
the screws at jff, which tip the mirror about a horizontal axis.
The telescope may be moved with its foot sliding on a plane.
The three possible positions of the mirror N (positive uncom-
pensated, self-compensated, negatively uncompensated) are
about l""" apart on the micrometer, for a plate of glass -68*^"
thick. When the arc lamp is used, the accentuated sodium
lines in the spectrum may be used in place of the white unde-
viated images of the slit, both for adjustment of the two
spectra for coincidence and as a fiducial mark, in place of the
cross hairs in the telescope. For a small angle of incidence the
sodium lines of higher orders of spectra are also liable to be

To measure the angle of incidence /, the table /is turned
in its socket, until the reflected image of the slit coincides with
the slit itself. A hole is cut in the top or side of the collima-
tor tube near the slit (not shown), for this purpose. There-
upon the table /is turned back again until the images coincide
in the telescope. The angle read off on the graduated plate K
is 7, the reflected ray travelling over 2Z The table I is pro-
vided with an index and vernier (also omitted in the diagram).

The apparatus described being made virtually of hollow
parts is light enough to be carried about with convenience.

Digitized by VjOOQ IC

114 (7. Barus^Displacement Interferometer.

In the case of the figure where the angle of incidence /is
small, the distance from Jf to ^ is about W°. It may easily
be increased to several times this, by inserting longer gas pipes at
e and d with appropriate braces. The fringes are very stable
even when the mstrument stands on a table fastened to a wall
bracket. They naturally quiver when the observer is manipu-
lating the micrometer screw, but they return at once to quies-
cence when the hands are removed. To obviate quivering, i. e.
to follow the motion of individual rings, the usual tangent
screw method may be employed.

2. Other Interferences, — The same apparatus may be adapted
for observing tne linear diffraction-reflection interferences
described by Mr. M. Barns and myself.* The equations here

Online LibraryJohn Elihu HallThe American journal of science → online text (page 11 of 61)