PROCEEDINGS
OF THE
AMERICAN ACADEMY
OF
ARTS AND SCIENCES.
Vout. LVII.
FROM MAY 1921, TO MAY 1922.
BOSTON:
PUBLISHED BY THE ACADEMY.
1922.
Tse Cosmos PREss
CAMBRIDGE, MASS.
CWA . (Lr ¢ i 4 ‘ i} g A ha ,. =
Sy, ,
=
Ly Ifans
II.
ITT.
IV.
VI.
Vil.
VIII.
IX.
XI.
XII.
XIII.
XIV.
XV.
XVI.
CONTENTS.
Paae.
The Grid Structure in Echelon — Lines. By N. A. Kent
AND L. B. Taytor . ,
The General Conditions of _— a the Principe 1 Le Chatelier.
By A. J. LoTKa .
The Effect of Tension on the Electrical Resistance of Certain Ab-
normal Metals. By P. W. BripGMan PCr ae
Notes on the Early Evolution of the Reflector. By Louis Bre. .
The Effect of Pressure on the Thermal —- - Metals. By
P. W. BRIDGMAN
The Failure of Ohm’s Law in Gold and Silver at — Current
Densities. By P. W. BRIDGMAN
A Table and Method of Computation of Electric Wave eiieaiaatais
Transmission Line Phenomena, Optical Refraction, and Inverse
Hyperbolic Functions of a Complex Variable. By G. W.
PIERCE . Fe a eae aS ae ES AS oe TT
Artificial Electric Lines with Mutual Inductance between —
Series Elements. By G. W. Pierce pes R
The Parasitic Worms of the Animals i Bermuda. I. Trematodes.
By F. D. BARKER , ‘ ;
Additions to the H _— Fauna af the Bermudas. By Rvupo.Fr
BENNITT ‘ j ;
Some Hymenopterous Parasites of Lignicolous Itonididae. By
C. T. Bruges Peay es eterna ote 8, es os eae
A Revision of the Endogoneae. By RoLAND THAXTER ‘
The Echinoderms ¢ the ibrar Bank, Bermuda. By H. L.
CLARK re age anger ag
Atmospheric Attenuation - Ultra-Violet — By E. R.
SCHAEFFER We Pee er eal SO
The Ratio of the Calorie at 73° to that at 20°. By ARNOLD Rom-
BERG pegs rata ee ae ig eS a
Studies on Insect Spermatogenesis. IV. The Phenomenon of Poly-
megaly in the _— Cells — the gees Pentatomide. By R.
H. BowEn . , A ke apos”
421175
1
19
39
67
75
129
173
193
213
239
261
289
351
363
375
388
CONTENTS.
XVII. Note on Two Remarkable Ascomycetes. By RoLaAnp THAXTER.
XVIII. Recorps or MEETINGS
BIOGRAPHICAL NOTICES
OFFICERS AND COMMITTEES FOR 1922-23
List OF FELLOWS AND FoREIGN HONORARY MEMBERS
STATUTES AND STANDING VOTES .
RUMFORD PREMIUM
INDEX
ae
ik ae
ae
pray os
Pee
57 —1
Proceedings of the American Academy of Arts and Sciences.
“Vou. 57. No. 1.— Decemser, 1921.
THE GRID STRUCTURE IN ECHELON SPECTRUM LINES.
By Norton A. KEntT AND Lucien B. TAYLor.
INVESTIGATIONS ON Licnr AND HEAT MADE AND PUBLISHED WITH AID FROM THE
RumFrorp Funp.
(Continued from page 3 of cover. )
VOLUME 57.
i. Kent, Norton A. and Taytor, Lucren B.— The Grid Structure in Echelon Spectrum
Lines. pp. 1-18. December, 1921. $.75.
bi
u
eke
d ay
er Ma Fe
Proceedings of the American Academy of Arts and Sciences.
Vou. 57. No. 1.— Drecemser, 1921.
THE GRID STRUCTURE IN ECHELON SPECTRUM LINES.
By Norton A. Kent anp Lucien B. Taytor.
INVESTIGATIONS ON LIGHT AND HEAT MADE AND PUBLISHED WITH AID FROM THE
Rumrerp Funp.
THE GRID STRUCTURE IN ECHELON SPECTRUM LINES.
Norton A. KEntT AND LuciEN B. TAyLor.
Received July 7, 1921. Presented October 19, 1921.
SoME years ago Nutting?! noted a peculiar, complex structure,
termed by him the “fluting” or “grid,” which appeared in many
echelon spectrum lines, and consisted of several fine components of
different and often changing intensity. Later one of us ? independ-
ently noted this structure. Nutting crossed the 12’ Lummer plate
of the Bureau of Standards with his echelon and was apparently
forced to the conclusion that the structure was real — that is, that it
indicated an actual discontinuity of emission in the source.
Proceeding on the assumption of reality, the writers attempted a
solution of the problem using Lid 6104 which, although known to be a
spectroscopic doublet, offered peculiar advantages in that the grid
was extremely brilliant, well-marked and persistent.
APPARATUS.
The apparatus used consisted of :—
Two echelons: No. 1, made by Porter, 30 plates, each 14.76 mm.
thick, step 1 mm., aperture 31.0 by 33.0 mm.; No. 2, made by Petit-
didier, 30 plates, each 23.29 mm. thick, step 1 mm., aperture 31.0 by
35.5 mm. |
The Bureau of Standards 12’ Lummer plate kindly loaned by Dr.
Stratton.
A Hilger Lummer plate — length 131 mm., width 14.5 mm., depth
4.827 mm.
A Hilger constant deviation prism spectroscope combined with an
echelon as in Figure la; also a separate Hilger spectroscope with
another echelon spectroscope as in Figure 1b. The achromatic lenses
of both echelon spectroscopes are of about 50 cm. focal length and 5 cm.
1 Astrophys. Jour. 23, pp. 64 and 220. 1906.
2 Kent, Proc. Am. Acad. XLVIII, No. 5. Aug. 1912.
4
KENT AND TAYLOR.
aperture; each echelon bed rotates on an axis at its center; the Hilger
micrometer is fitted with one fixed and two movable cross-hairs as
shown in Figure 2.
Fic. la.
vt
°}
va
m
LSo
Fia. lib.
, L
s
ee
s &
/ ;
\
mM M
®
Ss
Fic. 2.
FicurREs laand lb. S, slit; L, L, lenses; E, echelon; P, prism;
O, ocular.
Figure 2. SS’, fixed crosshair; MM’, MM’ movable system;
LL’, spectrum line.
A Littrow mount spectroscope consisting of a Petitdidier
achromat — focal length 30 feet, aperture 6”; and an An-
derson grating — aperture 33” vertical by 5” horizontal,
15,000 lines per inch, used in the third order.
A5K. W. transformer, 110 to 30,000 volts ratio of trans-
formation, fed by a 60 cycle Holtzer-Cabot 4.5 K. W. gen-
erator. Various large induction coils capable of giving 6”
sparks and operated by a rotary mercury break and an
electrolytic interrupter, had proven insufficient.
A vacuum arc of construction as indicated in Figure 3.
THE GRID STRUCTURE IN ECHELON SPECTRUM LINES. ~ 5
a
D
O
x |
SS S
S008:7
»_ A™® JAN
Nex * AA
NZ)
N
Ni
N
N
N
SS .
\ N
\
\ T
A,
e 4
° ,
‘
W Ww
1
‘ \
e \
| ‘
‘
~
ae
\
Ni
NI
J N
NJ
NJ
N
N
:
OCR EE OROR EE Ee
==G@
Iz
Ficure 3. One fourth original size. D, D, fibre disks; O, O, water outlet;
P, P, fibre plugs; A, A, asbestos; J, water jacket; W,W, W, windows;
I, I, I, water inlet. ‘
6 ; KENT AND TAYLOR.
This was also adapted to pressures of several atmospheres as the glass
windows and fibre plugs were held in place by threaded rings.
Quartz vacuum tubes — even pyrex glass having proven unsuit-
Water intake
yo
To vacuum pump
Water intake
|
R R
| . e Water outlet § HF
(ZZ TET
Vater outlet
Figure 4. C,C, cork stoppers; R, R, rubber sponges; T, T, terminals.
| ~
Ay Abr ~dh- nan
f NP ‘ F wee,
Jo vacuum
| pumps
To H, rator
- ky M
drying apparatus
Ficure 5. TT, tube; M, manometer; B, bulb.
able — of various forms, the most successful of which, for salts such as
lithium chloride, proved to be that shown in Figure 4 in which fine
brass wire, often in helical form, was fitted into brass caps, 6 mm. in
THE GRID STRUCTURE IN ECHELON SPECTRUM LINES. 7
diameter, and sealed in with De Khotinsky cement, each joint being
cooled by a water jacket. The salt is shoved into the capillary by a
wire and the tube will run many hours without refilling. It may be
used end on as well as side on. The capillaries varied from 2 to 0.5
mm.
Auxiliary apparatus as shown schematically in Figare 5. The bulb
B, prevented too rapid changes in pressure. The system was washed
out with hydrogen from a Kipp generator, dried by sulphuric acid and
a calcium chloride tower. The mercury manometer, M, indicated the
pressure — generally from 8 cm. to a fraction of a millimeter.
PROCEDURE AND CERTAIN RESULTs.
Both Lummer plates were each in succession crossed with echelon
No. 1. In each case, with a carbon are soaked with lithium chloride,
both at atmospheric pressure and in a moderate vacuum, there .
appeared a pattern which, at this stage of the investigation, seemed
to indicate that the grid was real. The following facts, (1) to (6), are,
however, clearly not in accord with this conclusion, and prove con-
clusively that this curious structure is due to the phenomenon of
“secondary maxima” observed by Stansfield * and resulting from
successive reflections from the surfaces of the echelon plates, producing
a Fabry and Perot system in the region of the primary light of the
echelon. (1) to (3) deal with some of the criteria of echelon secondary
maxima given by Stansfield. These criteria are, in essence, indicated
below by italics.
(1) The width of Lid 6104, given by an open carbon arc at atmos-
pheric pressure, as seen in the Littrow grating, using a narrow slit,
was found to be about 0.25 t. m. when echelon No. 2 showed the grid
plainly. The suspicion, therefore, was confirmed that the line was
too wide for the echelon, the difference between the adjacent orders,
being about 0.26 t. m. In the case of Janicki’s observation * of
Hg. \ 5461, Nutting’s work on lines of many elements, and the work
of one of us on the zinc lines as given by arc and spark, the indications
are that with all lines for which the echelon shows the grid, their
breadth is so great that the use of this instrument is not at all justifi-
able.
The writers then proceeded to study the structure from this new
3 Phil. Mag. (6) 18. 383. 1909.
4 An. der Phys. Vol. XIX, p. 36. 1906.
8 KENT AND TAYLOR.
standpoint, considering the primary line of width approximately
0.25 t. m., and not as formerly, one of the grid components itself.
These components are indeed, in this sense, each narrower than the
primary maximum — 0.25 t. m.— the grid components, all of them
now regarded as secondary maxima, being only about 0.05 t. m. in
width in echelon No. 2.
(2) The curvature of one of the mercury yellow lines was compared
with that of a grid component in A6104. By stopping down the
echelon spectroscope slit, a line of definite length was observed, and
by setting the stationary cross-hair of the filar micrometer upon
the ends of the image, and the movable system upon its center, the
horizontal distance, d, Figure 2, from the ends of each line to its
center were measured. It was found that the curvature of the com-
ponent is about 25%, greater than that of the primary line.
(3) With a small mirror, set at 45°, over the lower half of the echelon
spectroscope slit an argon vacuum tube and the lithium are were
observed at the same time. The relative motion of the grid compo-
nents in \6104 and a nearby argon line were then studied as the
echelon was rotated. The primary argon line moves about one-half as
fast as the grid components.
Quantitative measurements of the relative displacements were later
made with Zn 44810. A quartz vacuum tube was fitted with coiled
brass wire leads and brass terminals, exhausted, filled with hydrogen
to 10 or more cm. pressure and then gradually exhausted to 1 mm. or
less. The zinc lines given by the brass wire leads appeared very
sharp, steady and brilliant. With \4810, as thus produced, was com-
pared the “gridded” line of a cored carbon arc at atmospheric pres-
sure, in which small pieces of zinc had been placed, the small mirror
arrangement allowing simultaneous observation of both sources.
Upon rotation of the echelon the grid components rushed by the
narrow tube line. To measure the relative speed a plane mirror was
attached to a side of the echelon case. The image of an illuminated
slit in a piece of cardboard was formed by a lens upon a distant scale
after reflection from the mirror. The echelon was set near the 6 = 0
position. A reading of the position of the slit image on the scale was
taken when the tube line lay upon the fixed hair of the filar micrometer.
The echelon was then rotated until the slit image moved about 2 cm.
The displacement of the tube line was then measured by the movable
cross-hair system. A similar series was then taken with a grid line.
The ratio of the displacements was 3.6 : 6.4 or about 1 : 2.
(4) The echelon was removed and the ocular focussed on the prism
THE GRID STRUCTURE IN ECHELON SPECTRUM LINES. 9
image of a line. Replacing the echelon shortened the focus for a true
narrow echelon image by about 0.6 mm. The focus for the grid com-
ponents of the same line was 0.7 mm. shorter yet — the light forming
the grid had traversed the echelon plates more than once.
(5) Although the grid components are generally very well defined
(the minimum being “deep”’), it is a difficult matter, with a fluctuating
source such as an open arc, to obtain accurate measurements. The
grid spacings appear to vary slightly at different stages. When the
grid is complete the spacing is regular and, within the limits of error of
measurement is equal to one fifth the distance between the orders.
This was proven as follows:— A quartz tube having merely coils of
fine brass wire as terminals gave extremely fine zinc lines. Echelon
No. 2 was set in double order condition and Ao, the difference between
the two orders, measured for \4810 (see Table I). Then the grid was
measured as given by a 3 ampere open carbon arc. Three distinct
series of readings were taken. Then the tube was again used. ‘The
accuracy of an individual setting was about 0.2% in Aoand about 5%
in Ag. It thus appears that in this region, at least within 2%, Ao =
5Ag. The focus of the instrument was, of course, not changed, the
difference between that for primary and secondary maxima being so
slight that distances between the components of the grid are not
appreciably affected. |
TABLE I.
Distances are measured in divisions of the{micrometer head. Each Ao die.
tance given is the mean as calculated from four settings; each Ag from two.
Settings were made on the six centrally situated grid components.
Ao for Zn \ 4810 | Ag for six grid components.
22.60 1-2 | 2-3 | 3-4 | 45 | 5-6 |Mean/| Mean
Mean Motes
4.3 4.8 4.6 4.8 4.7 4.6
to
bo
qr
QO
4.8 | 4.4 | 46 | 48 | 4.1 |] 4.5 | 46
22.50 4.3.] 5.0 | 5.0.|.4.3 | 4.7 | 4.6
22.58 +4.6=4.9 or Ao = 4.9 dg
A similar series for Zn \6362. gave Ao = 27.65 and Ag = 5.6, 5.9,
5.5, 5.6, 5.4: mean = 5.5. Hence Ao = 5.0 Ag. |
10 KENT AND TAYLOR.
(6) The structure given by both echelons is the same. That is,
there are five secondary maxima for every primary maximum. Ag for
Hydrogen 6563 is 0.061 t. m. for instrument No. 2, while for No. 1
it is about 0.096 t: m., which again is another fact fatally inconsistent
with the existence of a definite discontinuous emission in the source.
With this evidence at hand the writers then attempted to clear up
the results of the crossed dispersions. The source previously used was
hardly adequate. By removing the soft core of the lower carbon it
was possible to feed copiously into the are a strong LiCl] solution.
Greater brilliancy and steadiness were obtained. The results were
unmistakably in accord with the facts given in (1) to (6) above.
Figures 6a and 6b indicate the structure observed. These will be
%
Fic. 6a. Fic. 6b.
Ficure 6a. Li 46104 with crossed Lummer plate and echelon. Grid not
indicated. , and dz in single and double order condition respectively.
Ficure 6b. Li \6104 as in Fig. 6a. Grid shown. A, and Az both between
single and double order condition.
discussed in full below (see page 16). When one component of the
spectroscopic doublet is in double and the other in single order condi-
tion three lines appear; when both are in a condition between single
THE GRID STRUCTURE IN ECHELON SPECTRUM LINES. 11
and double order there are four lines. It is probable that these four
lines, under conditions of inferior illumination, were interpreted as
four separate and true lines. It is unfortunate that at first the only
line available for study was a doublet. With this latter and better
source a zinc chloride solution gave 4810 sufficiently strong. The
crossed dispersions prove it to be a simple, though broad, single line
when the echelon alone shows the grid.
FuRTHER RESULTS: CHARACTERISTICS OF THE GRID.
(a) From numerous observations upon Li \6104 and Zn 4810, as
developed by various sources, such as vacuum tubes and arcs (on 110
and 220 volt D.C. circuits and from 1 to 20 amperes) under high,
normal and low pressure, in which the cross-hairs of the filar micro-
meter were set successively upon the true, narrow, lines given by the
tube and the grid components given by the are, it is quite certain that
the grid is built up approximately as follows:— Suppose that in a
hypothetical grating of resolving power and dispersion equal to that
of the echelon, a line which is at first very narrow, e.g., 0.025 t. m.,
gradually becomes less monochromatic, owing to changing conditions
in the source, and appears as represented diagrammatically by the
small letters a to e, Figure 7. Four cases must be discussed as shown
in Figures 7 to 10, respectively.
CasE I:— The echelon in double order condition gives successively
images A to E. When the line is very narrow the echelon shows it as
such, in A. Similarly for a line of width, Ag — the width of a grid
component or an intergrid distance — it is shown as in B. When of
width 3Ag, the echelon shows no change, C appearing as B; for, at m,
the primary and secondary action together give a decided minimum.
When the line has a width, as in d, the echelon shows a triplet, D, and
when of width as in e, or greater, the grid is complete — five grid
maxima, 1 to 5 and 6 to 10, for each maximum, such as 3 and 8, which
a narrow line would give; four maxima, 4 to 7, between the double
order positions, 3 and 8, of such a narrow line. For a given position
of the echelon these grid components do not, in forming, move very
much, if at all: they come up in situ. There exists an apparent
motion, in and out, which is probably due to the changing width of
the primary line, which may not at all times be such as to complete
the entire width of a grid component.
CasE II:— When the position of the echelon, its temperature and
the wave length of the line observed, result in the central grid minimum
KENT AND TAYLOR.
(23945678 90N
224756789
Fic. 8. Fic. 9.
Ficure 7. Case I: Echelon in double order
condition and a grid maximum coincident with
the primary maximum.
FicureE 8. Case II: Echelon in double order
condition and a grid minimum coincident with the
primary maximum.
Ficure 9. Case III: Echelon in single order
condition and a grid maximum coincident with
the primary maximum.
THE GRID STRUCTURE IN ECHELON SPECTRUM LINES. 13
occurring in the position of the narrow tube line, as in Figure 8, for a
double order condition of the echelon, there are nine or even eleven
components when the grid is strong. Note that the grid components
2, 3, 7, 8, which at first are very brilliant when the grid is “young,”
grow weaker, 2 and 8 often being so faint that it is difficult to make
accurate micrometer settings upon them.
CasE III:— The treatment is the same for a single order condition
of the echelon, as in Figure 9 which shows a triplet, quintuplet, or,
with neighboring parts of adjacent orders, even as many as eleven
components.
Cast IV:—Here a grid minimum coincides
with the primary maximum and the grid com-
ponents are as shown in Figure 10.
The above statements explain why an origi-
nally narrow line, as its width increases, may ap-
pear, as it actually does, a triplet or quintuplet, as |
in Figures 7 and 9, or may, as it were, “reverse”
and then quadruple, asin Figures8 and 10. Actual
reversal as shown by the grating probably occurs
much later in the history of the line. (See page
15.)
Further, if a line be intrinsically unsymmetri- |
cal, shading off to the red for instance, the sec-
ondary action masks an early stage of broadening,
and the left grid line, 2, forms as in A, Figure 11. |
Line 3, as in A’, then comes up as 2 strengthens.
(b) The grid begins to disappear and the line
gradually becomes broad and structureless when
the primary line exceeds 2Ao in width, Ao being ie
the distance between two adjacent orders. This
was determined as follows:— Using
as narrow a slit as possible, a low
power ocular and a mm. scale, an
A i eye estimate was made of the
breadths of various portions of an’
$23456789H
arc line shown by the grating.
| | These were reduced tot. m. The Fic. 10.
same source was viewed simultane-
a46
Figure 10. Case IV: Echelon in single order condition and
Fig. 11, a grid minimum coincident with the primary maximum.
14 KENT AND TAYLOR.
ously by echelon No. 2. For Zn \4810 three components of the grid
exist when the grating shows a line 0.12 t. m. broad. Ao for \4810 =
0.155 t. m. $ X 0.155 = 0.09 t. m. which compares favorably with 0.12
t.m. The complete grid exists when the line is 0.3 t. m. or 2Ao
broad and the image begins to pass into a structureless line at 3Ao.
Similarly for Li \6104 a full and well-marked grid exists at a line
width about 0.2 to0.5 t.m. or Ao to 2Ao (as here Ao = 0.25t. m.).
The grid is poorly marked above about 2Ao and is gone at 3Ao.
(c) Numerous lines in the spectra of Na, Hg, Fe, Mg, Cd, Ca, Sn,
Pb, and Bi, developed by an open carbon arc, show the grid whenever
the line is sufficiently broad — rendered so by introducing more of
the substance or increasing the current; also by increasing the capac-
ity in the case of a spark.
(d) Li AA6708 and 6104, Zn \d4810, 4722 and 4680, also Hg \5461
(mercury being fed into the lower cored carbon) show by their behavior
that a line which is too broad will appear structureless in the echelon,
that the center of the core of an arc may show the grid complete while
light from the wings of the image gives a simple structure of but one
to three components. With a sufficient amount of vapor the com-
plete grid may be obtained even at low pressure.
(e) A study of Zn 44810, from an arc in the vacuum or pressure tank,
at pressures from 2 cm. of mercury to about three atmospheres,
showed that moderate changes of pressure do not produce measurable
displacements in the grid components, but merely alter somewhat
their relative intensities, shifting the maximum over one or two com-
ponents or even bringing up new ones. This of course means that,
as long as a grid exists, the components do not change appreciably
their position with changes of wavelength as small as 0.015 or 0.020
t.m.° Their position is affected more strongly by the position of the
echelon and its temperature. Similarly, the grid components of the
spectroscopic doublets Li AA6708 and 6104 developed in vacuum tubes
show intensity shifts with changes of pressure over the range of one
atmosphere.
(f) The “end on”’ position of a vacuum tube will generally show a
more complete grid than that “side on.”’
(g) If a line broaden unsymmetrically with increase of current the
maximum of intensity will shift. Those components which are just
being formed show an apparent motion outward as the number of
5 According to Humphrey’s and Mohler’s results for Zn, the pressure shift
reduced to 4000 is 0.057 t. m. for twelve atmospheres.
THE GRID STRUCTURE IN ECHELON SPECTRUM LINES. 15
components increases, the first step resembling a narrow reversal as
in Figures 8 and 10 or a central fixed line with two moving wings as in
Figures 7 and 9. But the writers feel that this apparent motion is
due to the fact that each grid component is not formed 2n toto at once:
the part which lies nearest the center of the system is formed first.
Certain it is that this apparent motion ceases abruptly when the
component has reached a position which is one grid distance from
its neighbor. If the source be an arc, many rapid fluctuations in
intensity occur.
(h) Although the resolving power of the grating (225,000 in the third
order) is far below that of the echelon (about 750,000 for \6100 for
echelon No. 2) it is hard to reconcile the images given by the two
instruments on any other assumption than that the grid is due to
secondary action.
To throw further light on the problem, Li \6104, given by a vertical
carbon arc soaked with LiCl, was viewed simultaneously by echelon
and grating. Table II gives a summary obtained from various ar-
‘rangements.
TABLE II.
|| || indicates the grid; § a broad structureless line; | a narrow unreversed
line, or one very slightly reversed; jj a broad and strongly reversed line.
Arrangement when oF ree me . Echelon shows Grating shows
upper pole with solution
1 + + At + pole |||| |
: cee, oe eg iT
2 + ” a |
eo. «@ 4 il
3 ai Be oe. 2 s il
“—" B |
4 = + ae B il
a ae |
Therefore which pole is soaked makes no difference, nor does it
matter which pole is above. The region near the + pole generally
shows the grid in the echelon, that near the — pole a broad structure-
less line. The grating always gives a narrow unreversed line or one
very slightly reversed where the echelon shows the grid, and a strongly
reversed line where the echelon shows no structure. Thus the grid
does not result from conditions which produce a reversed grating line.
16 KENT AND TAYLOR.
With Li \6708, which usually appears widely reversed in the grat-
ing, the grid is more difficult to obtain in the echelon, while with
Na \4972 — given as an unreversed line by the grating at either edge
or centre of the arc image — the echelon shows the grid at both edge
and centre.
(i) We are now in a position to discuss in detail Figures 6a and 6b.
These were obtained with the 131 mm. Lummer plate set between the
collimator and prism of Figure 1b and crossed with echelon No. 2.
The source was that described on page 10: the are current being from
10 to 25 amps. The plate dispersed vertically, the echelon horizon-
tally. Both figures are drawings based on visual filar micrometer
measurements, a single cross hair being moved successively along the
axes, vv’ (vertical), hh’ (horizontal), aa’ (across the structure) pp’
(parallel to it), as shown below the two figures.
Two Lummer plate orders are shown in each figure, the primes
distinguishing these. ‘The numerals indicate the two components of
the spectroscopic doublet, the breadth along axis aa’ their approximate
relative intensity. A, is the weaker line, A». the stronger in both
figures — dz being the component of longer wavelength.
In Figure 6a d2 is in double order condition; in 6b both A; and dz
are between double and single order. The echelon grid structure is not
indicated in Figure 6a: in 6b its approximate position is shown. It
was difficult to observe at the ends of the lines and so is not there
indicated: it is slanted at an angle of about 2.4° (see gg’ in Figure 6b)
with the vertical. The slant of the lines themselves as well as that of
the grid changes with the positions of both plate and echelon: further,
the grid slant is not due to the curvature of the echelon image. This
may throw some light on the disappearance of the grid at a breadth
of line greater than 2Ao. For, as the echelon action alone is given by
the projection, on the pp’ axis, of the grids of the lines \; and dg, it is
evident that lack of coincidence owing to slant would tend to obliter-
ate the grid altogether, this indicating that two broad lines, the centers
of which lie as far as 0.1 t. m. apart (the Ad of the two components of
Li \6104), may not give coincident grid structures; or, in other words,
the grid maxima do not (for any one position and temperature of the
echelon) necessarily fall together. This is not inconsistent with shift
of intensity for small changes of wavelength (0.015 to 0.020 t. m.) as
noted on page 14. Shift of intensity and position probably both enter
with change of wavelength of the center of gravity of a primary echelon
image.
These two figures show that the grid is unquestionably a secondary
®
THE GRID STRUCTURE IN ECHELON SPECTRUM LINES. 17
echelon action. Otherwise the regions between lines 1 and 2 would
have been filled in with a structure along axis aa’ similar to that
along pp’.
With an echelon alone we have obtained only the weaker component
of Li \6104 as a single narrow line. We plan to cool the tube with
liquid air, thus sharpening the stronger component so that it will
no longer suffer the secondary action, to which the small satellite is
probably due.
(j) We have no record of having observed in either echelon any
ungridded line of width greater than Ag. Either there exists (1) a
very narrow line, (2) an irregular series of such, as, for instance, in the
yellow mercury lines, (3) a line of width Ag, (4) a series of such (the
grid more or less complete) or (5) a broad, structureless image cover-
ing between one and two orders. And it appears extremely probable
that the “reversal” of the main component of Hg \5461, noted under
certain conditions by several observers and often noticed by us, may
be modified by the entrance of secondary action due to the excessive
breadth of this component.
(k) The retardation producing the primary maxima of a narrow line
is proportional to n—1, while that of the light undergoing secondary
action is proportional to 3n—1. Thus the difference in retardation
in case of the two actions bears the ratio to the retardation of the
2n
n—l1
varies from 5.50 for \6563 to 5.37 for \4341 in echelon No. 2; and from
5.48 to 5.35 respectively in No.1. Since echelons are generally made of
substantially the same kind of glass, any two having equal separation
of primary orders will have equal separation of secondary maxima,
because this separation is the same fractional part of the separation of
the orders; but the values of Ag in t. m., varying with the dispersion,
will, of course, differ in different instruments.
We cannot state just why Ao = 5Ag. The measurements given
above indicate that this is so within the limits of experimental error
for both the violet and red regions.
It would be interesting to assemble an echelon under water, press
the plates together and allow the superfluous water to drain off. This
process might vastly reduce the secondary action. If successful
Canada balsam might be substituted for water thus producing a
more permanent instrument. We plan to try this experiment shortly.
2n
primary of —y which is a function of nalone. The value of
KENT AND TAYLOR.
CONCLUSION.
Summarizing the above results, we may state that the evidence is
entirely against the existence of a discontinuity of emission in the
source. The grid is due to a secondary action of the echelon which
enters when the line under investigation is not sufficiently monochro-
matic. This means that the previous work of one of us ® must be
considered as of small value and also that an explanation of the
apparent complexity of structure obtained by Nutting’ can be
found in secondary action.
The results obtained emphasize the fact that when an echelon is
used to measure small wavelength differences, great care must be taken
to obtain the lines so narrow that their width is less than 3 Ao, else
secondary action may enter to cut off an edge of a line and thus give
a false intensity-maximum position.
We must record our appreciation of the help rendered by various
student assistants, especially Messrs. Greenleaf and Risga. We are
also indebted to Dr. Lucy Wilson for her skilful aid during part of this
research and to our assistants, Miss Pearson for mathematical work
in connection with the calculation of the constants of the echelons,
and Mr. Gilman for making the sketches accompanying this article.
We wish also to thank sincerely the Rumford Committee of the
American Academy for numerous grants which made possible the.
purchase of the main pieces of apparatus used in this investigation.
6 Proc. Am. Acad., Vol. XLVIII, No. 5. Aug. 1912.
7 Astrophys. Jour., XXIII, pp. 64 and 220. 1906.
PuysicaL Lasoratory, Boston UNIVERSITY,
May, 1921.
