THE INFLUENCE

OF A

MAGNETIC FIELD UPON THE SPARK SPECTRA

OF IRON AND TITANIUM

BY

ARTHUR S. KING

WASHINGTON, D. C.
Published by the Carnegie Institution of Washington

1912

CARNEGIE INSTITUTION OF WASHINGTON

Publication No. 153

Papers of the Mount Wilson Solar Observatory, Vol. II, Part I

George E. Hale, Director

PRESS OF J. H. LIPPINCOTT COMPANY
PHILADELPHIA, PA.

TABLE OF CONTENTS.

PAGE

Introduction i

Theory and Former Inn'estigations.

1. General 3

2. Possible Relation between Zeeman Separation and Pressure Displacement S

3. Former Investigations of the Zeeman Effect for Iron 7

4. Former Investigations op the Zeeman Effect for Titanium 8

Apparatus and Methods.

1. Spark .\pparatus 9

2. The Electro-Magnet 11

3. The Spectrograph 13

4. Photographic Methods 16

S- Measurement of Magnetic Field 16

6. Methods of Measurement and Reduction 17

Explanation of the Tables.

1. Wave-Lengths 19

2. Intensity 19

3. Character of Separation 19

4. Weight 20

5. Values of AX 21

6. Values of AX/X^ 21

Table i, Measurements of Zeeman Effect for Iron 22

Table 2, Measurements of Zeeman Effect for Titanium • • 35

Types of Separation.

1. Un.\ffected Lines 44

2. Triplets 44

3. Quadruplets 45

4. Quintuplets 45

S- Sextuplets 45

6. Septuplets 46

7. octuplets 46

8. Nonets 46

9. More Complex Types 46

Relation of Separations to the Normal Interval.

1. Summaries for Various Types 47

2. Discussion of Relations to Normal Interval 49

Possible Relations Betw'een Lines as Indicated by the Zeeman Effect 50

Cases of Dissymmetry 51

Law of Change of the A\-erage Separation of the A'-Components with the Wave-length ^ 53

The Effect of the Magnetic Field lt>on Enhanced Lines 54

COMP.\RISON OF the ReSLT-TS FOR THE ZeEMAN EfFECT AND FOR PRESSURE DISPLACEMENT 56

Summary of Results 64

Bibliographical References ds

iii

4111

INTRODUCTION.

The investigation of which an account is given in the following pages was carried out during the year
1910 in the Pasadena laboratory of the Solar Observatory. The object was to obtain as complete data as
possible concerning the influence of a magnetic field on the spectra of iron and titanium through a con-
siderable range of wave-length, and to present this in such form as would be useful for reference in con-
nection with questions concerning the effect of a magnetic field on the spectrum Hnes, such as those arising
in investigations on sun-spots, as well as for comparison with the known phenomena of the Zeeman efTect for
spectra other than those of iron and titanium. The tables are designed to give an accurate description
of all lines between X3700 and X6700, so far as it has been possible to photograph them. The measure-
ments of magnetic separations for each spectrum through this range show clearly the degree in which
the separation changes with the wave-length. The complex types as well as the simpler are studied with
reference to the prevalence of a fundamental interval between the components. Numerous cases are
noted of the recurrence of certain types of separation, and while the search for series relations in these
many-lined spectra has not proved fruitful, the descriptions of the type of separation show whether
certain lines are possibly connected, or whether they unquestionably arise from different radiating par-
ticles. A few cases of dissymmetry among components are given in the tables. It has been possible,
by reason of the large amount of material collected, to make a detailed comparison between the Zeeman
separation and the displacement of lines produced by pressure around a light source, and it is shown to
what degree a correspondence exists. The reproductions of spectra which are given are of selected regions
showing the various types of magnetic separation and the behavior of groups of lines which are of special
interest in other investigations on these spectra.

The desirabihty of making the material as complete as possible has necessitated photographing
the weaker hnes in these two spectra so far as they were obtainable, a condition which has added to the
labor and altered to some extent the experimental methods that would have been used for the stronger
Hnes alone. The tables for titanium contain all but the weakest of those hnes given in the regular lists
of arc and spark lines. As much can not be claimed for iron, however, as numerous lines, fairly strong
in the arc, are not brought out by the spark in the magnetic field even with an exposure of many hours.
This is especially true of lines of diffuse appearance, which are particularly numerous in the iron
spectrum.

The results of a number of investigations on the Zeeman effect for certain parts of the iron spectrum
have been published, and will be spoken of in the historical summary to follow. These are fragmentary,
however, with some discordances, and it is beheved that there is little real duphcation in the present
paper, even for those parts of the spectrum which have been treated to some extent by others.

THEORY AND FORMER INVESTIGATIONS. ^^ *

I. General.

It is not the purpose of the author to give here in any detail the development of the theory of the
Zeeman effect or to summarize at length the many investigations which have led to the present state
of knowledge regarding the phenomenon. Several such accounts have appeared in publications which
are usually accessible. Among these may be mentioned the memoir of Cotton (r)* (1899), the chapter
by Runge in Kayser's Handbiich der Spectroscopie (2) (1902), the detailed discussion by Voigt (3) (1908)
in connection with the related optical phenomena, and the brief treatment by Lorentz (4) (1909) in his
Columbia Lectures. Of these the second is by far the most complete, covering fully the historical devel-
opment, methods of investigation, and the theory and spectroscopic results contained in the literature
up to that time. For the purposes of the present paper, we shall consider the points in the theory which
apply closely to the results of this investigation, and summarize the work of other investigators in so
far as their results relate direct!}- to those of the present research.

The later work on the Zeeman phenomenon has been concerned largely with the study of complex and
unusual types of separation. It was shown during the earlier investigations by Zeeman (5) , Michelson (6),
Preston (7), Cornu (8), Becquerel and Deslandres (9), (10), Ames, Earhart and Reese (n), Reese (12) , and
Kent (13) that a large proportion of the spectrum hnes of any of the elements that have been examined
are split into more than three components. This involved an extension of the original theory of Lorentz,
which satisfactorily explained the triplet separation, in which two components are given by the light
vibrations in a plane perpendicular to the lines of magnetic force, these showing respectively a right-
handed and a left-handed circular polarization, and a central component by the light vibrations paral-
lel to the magnetic force-lines. Since the phenomenon in its simplest form justified taking the electron
theory as the basis of all conceptions of the action of the magnetic field upon spectra, a series of investi-
gations, among which those of Lorentz (14), Larmor (15), Voigt (16), and Robb (17) may be mentioned, have
greatly extended the mathematical theory, both for radiation in general and for the explanation of the
more complex forms of magnetic separation. Voigt and Robb have based their theory on the idea of
mutually connected systems of electrons, and have thus been able to account for many of the more com-
plicated types of Zeeman separation. However, both the nature of the connections and the way the
magnetic field efTects such systems are but imperfectly explained.

The proportionality of separation of components to field-strength has been worked on by Reese (i'),
Kent (13), Runge and Paschen (i8),Farber {19), Weiss and Cotton (2o),Paschen (21), and Stettenheimer (22),
and established to a very close approximation. The law enunciated by Preston (23 1 that the character of
separation and chstance between components (measured in terms of change of vibration frequency) is the
same for corresponding lines in the series of Balmer, Rydberg, and Kayser and Runge has been investigated
by Reese (12), Kent (13), Runge and Paschen (24), Runge and Precht (25), Miller (26), and Lohmann (2:).
The last two have found some exceptions, though Runge and Paschen observed very close agreement for
the series lines of a number of elements. This relation has frequently been used, recently by Moore (28),
in an attempt to find series among spectra containing many lines.

There has been considerable work in recent years on the commensurability of the separations of
spectrum lines, that is, on the existence of a fundamental interval of which the separations of all com-
plex lines are multiples, and on the extent to wliich this applies to the separations of triplets in which

* Numbers in parentheses indicate references to the Uterature on p. 65.

4 INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

there is always a great diversity for the lines of the same element. As this point will receive a good deal
of attention in the consideration of the results for the spectra of iron and titanium, it may be well to go
briefly into this portion of the theory.

If the Zeeman phenomenon were in full accord with the simplest form of the electron theory as given
by Lorentz, all lines would show the separation of the "normal triplet," in which the distance of each
side component from the central Hne would be given by the relation

.- e HX=
AX =

m 4irv

where e/m is the ratio of charge to mass of the electron, H the field-strength, and v the velocity of light.
This is derived (2«) from the fact that the change of period of the light producing one side com-
ponent is eH/2m in 27r seconds, or eR/^-Kni vibrations in one second. The number of vibrations per
second is n = v/\. The change of frequency is then

, _vdX_ eH
X' 47rw
from which

^^ e HX'
AX =

m ^TTV

If I' be expressed in centimeters per second, the change in frequency per cm length is

AX _ e H
X^ m 47rii

The factor e/m is here expressed in electro-magnetic units. This value of AX/X^ for a given field deter-
mines the separation of the side, components of the "normal triplet" from the central line, and a con-
siderable number of lines in a spectrum will usually give a value of e/m in close agreement with that
obtained for cathode rays. The separation of the majority of triplets, however, differs from the normal
type, though sometimes by even multiples. This means either that there are real differences in the values
of e/m for different negative electrons, or that the relation derived from the elementary theory is not
sufficiently general. Lorentz inclines to the latter view (4a). In discussing this question, Voigt (3a)
observes that it is by no means certain that the field acting upon a given electron is the same as that
which we measure by one of our regular methods. The field due to the movement of charged parts of
the molecule itself must be recognized as possibly superposed on the external field due to the magnet.

The elementary theory does not provide for the more complicated types of separation, nor does any
extension so far worked out cover them satisfactorily. However, an examination of the results of Runge
and Paschen (18) (24) for several elements and of Lohmann (27) for the spectrum of neon (with the
echelon spectroscope) enabled Runge (29) to enunciate the following:

Die bisher beobachteten komplizierten Zerlegungen von Spektrallinien im magnetischen Felde zeigen die folgende Eigentiim-
lichkeit: Die Abstande der Komponenten von der Mitte sind Vielfache eines aliquoten Teil des normalen Abstandes

X'- m ^nv
Sicher beobachtet sind bisher die Teile a/2, a/j,, a/4, a/^, 0/6, 0/7, a/ii, a/12.

This work of Runge is regarded by Voigt as showing that the internal field acting on the electron can
have little effect, that the electrons within the molecule have the same value of e/m as that of cathode rays.

Such a relation between the separation for individual fines and that of the normal triplet is of high
interest when appHed to spectra containing many lines. It has been examined by Moore (28) for the spectra
of barium, yttrium, zirconium, osmium, and thorium, and relations similar to those observed by Runge
have been obtained. The objection can be raised to this method that, by choosing small fractions of the
interval a and correspondingly large multiples, the difference between the calculated and observed values

THEORY AND FORMER INVESTIGATIONS. 5

can be made as small as we please and brought within the errors of measurement. Runge gives a cri-
terion as to how far it is allowable to go in such calculations. This question of commensurability will
receive attention in the following study of the iron and titanium spectra.

Dissymmetry in the separation and in the intensity of components on the red and violet sides has
been observed many times in Zeeman investigations. Voigt (3/') arrived at the conclusion that light
observed at right angles to the force-lines should give a triplet whose red component is slightly closer to
the central hne and stronger than the violet component. Observations by Zeeman (30) on the iron
spectrum gave a number of cases where such a dissymmetry seemed to exist. Reese (12) also found triplets
and lines of higher separation for several elements which appeared to show the effect. More recently
a series of papers has been published by Zeeman (30 comparing the mercury triplets XS770 and XS791
by various optical methods. The latter line is distinctly shown to have its red component nearer the
central line than is the violet component, while X5770 remains perfectly symmetrical. The amount of
dissymmetry appeared to vary as the square of the field-strength. This confirmed a measurement made
about the same time by Gmelin (32) with the echelon grating. A dissymmetry of this sort is always small
and difficult of detection. Large dissymmetries are to be classified as abnormal separations. A few lines
of such a character occur in the iron and titanium spectra, which will be noted later. Lines of very pro-
nounced dissymmetry were measured by Jack ( 33) in the spectra of tungsten and molybdenum. Chromium
also shows a great number of unsymmetrical separations. Some striking cases were observed by Dufour (34) ,
and many others have been photographed in this laboratory. The theory of coupled electrons, by which
Voigt (35) has sought to explain complex separations in general, allows for the occurrence of such dissym-
metries.

The magnetic separation of absorption hues, or the "inverse Zeeman effect," has been investigated
by a number of observers, as a rule for only a few hues. In such experiments white light is passed through
the vapor of a luminous source placed between the poles of a magnet. It was shown by Konig (36) and
Cotton (37) that there is a full correspondence between the effects of the magnetic field for both emission
and absorption lines. The splitting of lines in the spectra of sun-spots observed by Hale (38) was thus
proved to be due to the action of magnetism by comparing the Zeeman effect for the same lines as pro-
duced in the laboratory. The peculiarities in separations of sun-spot lines can thus be studied, as is being
done in this laboratory and by Zeeman and Winawer (39) in their investigation of special polarization
effects for absorption lines, especially when the light passes at different angles to the magnetic force-lines.

2. Possible Relation Between Zeeman Separation and Pressure Displacement.

A preliminary paper on this subject has been published by the author (40). In the discussion of the
present results material will be offered for an extended study to test the hypothesis of a direct connection
between the Zeeman effect and the pressure displacement for spectrum lines. That such a relation exists
has been strongly advocated by Humphreys (41) in a series of papers which have been summarized (42) by
him, together with all other pressure investigations up to the year 1908. Humphreys's hypothesis, briefly
stated, is that the part of the atom to which the light impulse is due is a ring of electrons, rotating with
a period of the order of the light vibration. Each of the electron rings will then set up a magnetic field
of its own. The luminous gas will be in a condition of minimum potential energy when the planes of
the rings are parallel and the electrons revolving in the same direction. We must, however, in view of
the Zeeman effect, consider that different rings may rotate in opposite directions, and assume merely
that the regular condition is a rotation of the electrons in orbits approximately circular, with a tendency
for the planes of these to become parallel. The effect of pressure in the surrounding medium will be to
bring the rings closer together, thereby altering their mutual induction. If two rings rotating in the same
direction are made to approach, the current in each ring will decrease, which means a retardation of the
rotating electrons and an increase of period in the corresponding light vibration, resulting in a shift of
the spectrum hnes toward the red. ^

6 INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

If rings of opposite rotation are forced closer together, their motion will be accelerated, resulting
in a shift of the spectrum lines to the violet. Assuming that both directions of rotation are present for
electrons producing each spectrum line, the general result will be a widening of all hues as the pressure
increases, with a prevailing shift of the maximum of each line toward the red. This last is due to the
fact that the condensing action of the pressure on rings rotating in the same direction is assisted by the
effort of these rings to get into the strongest part of their mutual field; while for oppositely rotating
rings the approach is opposed by the magnetic action, so that on the whole the retardation of the period
for a given line is greater than the acceleration, and the line, while being widened toward both red and
violet, has its ma.ximum intensity moved toward the red.

Another theory, by Richardson (43), opposes the connection of pressure displacement with Zeeman
effect. Instead of basing his reasoning on magnetic perturbations, Richardson considers the electron
as an oscillator which sets up an alternating electrostatic field in its neighborhood. This field would
produce forced vibrations in the electrons belonging to neighboring atoms, an effect increased by pres-
sure in the medium. The electric field produced by the forced vibrations would then react on that of
the radiating electrons. The mathematical development gives a change of wave-length proportional to
the pressure and toward the red. Worked out numerically with the available data, the electrostatic
resonance theory requires values for the pressure displacement many times greater than those observed
experimentally. A modified conception of the equilibrium conditions might account for this discrepancy.

Richardson objects to Humphreys's theory largely on the ground that the magnetic disturbances
of period would be far too small to account for the observed displacements of lines unless the magnetic
field for any atom is greater than that corresponding to saturated iron, which Richardson holds to be
an upper limit. This is replied to by Humphreys in a later paper (41*, in which he questions the right to
base the possible magnetic intensity of iron atoms upon the properties of iron in large masses, since the
permeabihty and saturation point depend upon many factors of composition and physical condition.
Going farther, Humphreys considers an ideal electron ring and deduces an expression for the change of
rotation frequency brought about by an external magnetic field H, such as that due to a neighboring
electron ring. This is found to give an expression for the change of wave-length AX in the ether vibra-
tions of original wave-length X which reduces to AX/HX- = C, a constant, which is Preston's law for the
Zeeman phenomenon, indicating that the ideal electron ring is very similar in structure to the actual
radiating particle. If this similarity be admitted, Humphreys is justified in his next step, which is the
substitution of known values in the expression for the change of wave-length of ether vibrations pro-
duced by a change in the period of the electron ring. This gives a field-intensity for the rotating ring
of 45 X 10^, which is about ten thousand times that of the strongest fields used in spectroscopic work.
The change in mutual induction by pressing together electron rings having fields of this magnitude may
be expected to give sliifts of spectrum lines of the order of those measured.

A third theory is that presented by Larmor (44), who treats the electron as a Hertzian doublet in a field
of electric force. This field would be altered by any change in the distribution of material particles in
the medium such as would result from increased pressure. A molecule approaching a vibrating electron
would decrease the rigidity of the ether at that point. A lowering of the ether strain would tend to increase
the period of the electron, and it is shown that this might give displacements of the magnitude observed
for spectrum Unes. A note by Humphreys (4' c) points out that several consequences of Larmor's theory
agree only to a hmited degree with observed facts, although his claim that Larmor's equations should
give the amount of displacement inversely proportional to the wave-length is incorrect.

The interacting magnetic atoms of Humphreys seem to provide a very plausible theory, but experi-
mental data have been lacking to show the probability of a connection between the effects of pressure
and magnetic field on spectrum lines. Humphreys considers that, in general, lines of large Zeeman
separation are strongly displaced by pressure, but admits that there is scanty material on which to

THEORY AND FORMER INVESTIGATIONS. 7

base this conclusion. The refusal of banded spectra, notably that of carbon, to show either Zeeman effect
or displacement has often been cited as probably resulting from a connection between the two phenom-
ena, and interesting developments on this point have recently been presented. Dufour (45) obtained
Zeeman separations for the component lines of the band spectra of the chlorides and fluorides of the
alkaline earths, the magnitude of separation being about the same as for line spectra. A short time
after, Rossi 146) selected three of these, the fluorides of calcium, strontium, and barium, and obtained
distinct pressure shifts for the bands, the shift being of the same order as for line spectra. Comparing
his results with those of Dufour, Rossi did not find any general relation between the magnitude of
the two effects. Numerous investigations on the Zeeman effect for banded spectra have been made
during the past two years, part of which are summarized by Dufour (47), but corresponding results
for pressure have not been obtained.

A detailed comparison of Zeeman separation and pressure displacement for the line spectra of iron
and titanium will be made in the present paper.

3. Former Investigations of the Zeeman Effect for Iron.

Passing to special investigations on the iron spectrum in which the magnetic separations for certain
lines have been described and measured, the first to be mentioned is that of Becquerel and Deslandres (9).
In this, 10 lines are given from X3821 to X3873, most of them of complex separation. Shortly after,
these writers used a stronger field and covered a larger region. This pubhcation (10) gives no measure-
ments, being confined to a description of a few interesting t^pes of lines.

A note by Ames, Earhart, and Reese (n) speaks of the general characteristics of the iron lines between
X3500 and X4400, with special mention of the type of separation for a few lines. Reese (n) gives measure-
ments of the separation for 23 of the stronger lines in this region, the source being a carbon spark with
iron as an impurity. Kent (13) continued the investigation with better equipment, measuring about 90
iron lines between X 3550 and X 4550. Special attention was paid to a number of complex lines. Reese
had observed that the fines on his plates could be classified as to amount of separation in about the same
way that they were classified as to pressure displacement. Kent, with more material available for com-
parison, found that this relation was not verified.

The paper by Zeeman (30) was concerned chiefly with the question of a dissymmetry of the side compo-
nents of triplets, as measured from the central line. Hartmann (48) investigated the structure of a num-
ber of iron fines with the echelon spectroscope. He did not, however, obtain as good resolution of com-
plex tj-pes as was given by the grating method in the present investigation. The most extensive set of
measurements thus far pubfished on the iron spectrum is given in the thesis of Mrs. van Bilderbeek (49).
These are from photographs made with a concave grating for a magnetic field of 32,040 gausses. Measure-
ments are given for 137 fines between X 2382 andX4529. Of these fines 55 (40 per cent) are to the violet
of the region covered by my photographs; the others are the stronger fines among those given in my
tables, and have been of great service in determirung the field-strength. As will be noted later, there is
an excellent agreement between the two sets of measures for all fines whose components are sharp enough
to give measurements of high weight. Besides checking my standard field, the agreement between Mrs.
van Bilderbeek's field-value and that which I had obtained by other methods supports the contention
in her paper that the field-strengths pubfished by Kent and by Hartmann are both low.

It wiU thus be seen that several investigations of special regions have been carried out for the iron
spectrum with regard to the Zeeman effect. The region covered, however, has not extended beyond
about X 4500, with the exception of a few lines in the green examined by Hartmann, lea\'ing nearly three-
fourths of the range included in this paper as new territory. For the region from X3700 to X4500, which
has been covered to some extent by others, the previous investigators have measured only the stronger
fines, the description of the character of separation is usually brief or lacking, and the complex separa-

8 INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

tions are but incompletely considered. The range of spectrum covered previously has not been sufficient
to draw any conclusions regarding the variation of separation with wave-length, the comparison with
pressure effects and other changes of physical condition has not been carried out, and no application
has been made of Runge's rule for the commensurability of the distances between components. These
points will be handled in the present paper as fully as the material will permit.*

4. Former Investigations of the Zeeman Effect for Titanium.

A set of measurements was pubHshed by Purvis (50) for many of the stronger lines of titanium from
X 2S00 to X 5000. The majority of these are in the ultra-violet, 86 lines being measured in the region
covered by my tables. Three violet triplets were measured by Reese (12). A former paper by the author (51)
gave descriptions and measurements for 291 lines between X3900 and X6600. These were made from
the first set of plates taken in this laboratory, the first and second orders of the 13-foot (4 m) spectro-
graph being used, with a field of 12,500 gausses. The data for the present paper were compiled from a
much more extensive set of plates, taken with higher dispersion and stronger field, the gain in all points
being so great that these measures may be taken as superseding the previous ones. A still earlier paper
by the author (52) gave preliminary measures of some titanium and iron lines in a discussion of the charac-
ter of their separation in the laboratory as compared to that observed in sun-spot spectra.

* Note added January, 191 2: A dissertation by Immina Maria Graftdijli on Magnetische SpUlsing van h:l Nikkei- en Kobalt-
Spectrum en van liet Ijzer-Speclriim (Amsterdam, igii) has just been received. Measurements are given for 38 of the stronger iron
lines between / 4300 and > 6500 for a field of 32,040 gausses. The measured separations of triplet lines agree in general very closely
with those presented in this paper. The only notable discrepancies are for a few comple.x lines where a large difference in field
necessarily alters the appearance of the components which are measured.

APPARATUS AND METHODS.
I. Spark Apparatus.

The source of light used in all of the work was a spark discharge from a 5-kw transformer made
according to special design by the Peerless Electric Company, of Warren, Ohio. The coils of this trans-
former are immersed in the best moisture-free oil and contained in a cyUndrical iron tank 83 cm in diameter
and 125 cm high. The primary and secondary leads are passed through the flat top of the transformer,
on which is a large knife switch for the regulation of the secondary voltage. The bar of this switch forms
the radius of a circle, one end being pivoted, while the other end fits into any one of a series of jaws along
the circumference of the circle. The connections with the transformer coils are such that the secondary
voltage may be 10, 20, 40, 80, 160, 320, or 640 times the impressed primary voltage, according to which
jaw the bar of the switch is titted into. Thus with 100 volts on the primary the secondary voltage is
1,000, 2,000, 4,000, 8,000, 16,000, 32,000, or 64,000, according to the connection. The use of a rheostat
in the primary circuit to regulate the impressed voltage will obviously give any secondary potential
desired up to 64,000 volts. The adjustable rheostat used is one capable of carrying heavy currents con-
tinuously. It is composed of sheets of tin cut into strips i cm wide by cutting almost across the sheet
first from one side and then from the other. The sheets of strips thus made are mounted vertically against
strips of asbestos fastened to a wooden frame, the distance between successive sheets being sufficient
to provide air circulation for coohng. Copper wires soldered to the tin strips at the proper intervals
lead to knife switches on the top of the rheostat frame. Various combinations of these switches place
parts of the tin resistance in series or parallel, and permit the resistance to be reduced by short steps
until all is out. One switch may be connected to an external resistance, thus allowing the latter to be
connected in series with any part of the tin resistance for fine adjustment of the rheostat. A bank of
twenty-four 32-cp incandescent lamps in parallel is usually used in this branch.

The primary current is supphed at about 104 volts from one side of the three-phase connection of a
1 5-kw transformer. This transformer and one similar to it are mounted in the transformer room of the
laboratorj-, fed by 2200 volts from the lines of the Southern California Edison Company, and are used
together to supply the 208-volt three-phase current for the D.C. motor-generator set which furnishes
current to the electro-magnet.

Two glass-plate condensers were used for the spark circuit during the series of experiments. The
more efficient one, used in taking the later photographs, is built up of 16 sheets of plate glass, of area
61 X 66 cm, and thickness 5.5 to 6.0 mm, laid horizontally in a strong, copper-Hned wooden tank.
Between the glass plates and at the top and bottom of the pile are sheets of copper, 17 in number, each
0.9 mm in thickness and with an area of 3330 sq cm, one side of each sheet having a tongue 2.5 cm long
projecting beyond the glass plates for the connection, while the plates immediately above and below are
cut away so as not to reduce the insulation at this point. Around the other three sides the copper is
cut so as to come 2.5 cm inside the edge of the glass plates. This arrangement, together with the form
in which the copper is cut on the fourth side where the tongues project, insures a distance of 5.7 cm
along the glass from the edge of one copper plate to the edge of the next. The condenser plates are sepa-
rated from the copper fining of the tank by a wood flooring 2.5 cm thick and held in place by a wooden
box inside the tank. A thick copper wire is soldered to each of the tongues coming from the copper plates
and the other end of the wire connected to a bincfing post set in a plate of fiber extending across the width
of the tank, 7.5 cm below the top. This fiber plate was at first placed level with the top, as shown in the
photograph of the laboratory (Plate I). This condenser is entirely immersed in the best transformer oil,
which fills the tank up to about 5 mm above the fiber plate, thus insulating the condenser plates and also

9

lO

INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTR.\ OF IRON AND TITANIUM.

O'C

^

^

_^_

the binding posts, the screw tops of the latter projecting from the oil to receive the wires connecting them
in any desired combination to the discharge circuit, so that the whole or any part of the condenser may
be used. An adjustable spark-gap between the nearer binding posts on each side protects the condenser
against too long a spark in the circuit which might cause the glass to be punctured. Connecting wires
from the two central binding posts are inclosed in thick glass tubes which pass through a second fiber
plate directly above the first and level with the top of. the tank to the high tension wires supported b}'
glass insulators and extending across the laboratory below the ceihng. A wooden cover fits into the top
of the tank and protects all parts of the condenser from dust. The leads from the transformer pass to
the overhead wires and other leads drop down to any piece of apparatus under the wires, so that the
heavy condenser can remain permanently in its place.

In addition to the condenser, the circuit from the transformer contains a self-induction spool and
spark gap in series with the spark under observation. Several self-induction coils are available, but the
one regularly used consists of 207 turns of insulated copper
wire wound on a wooden cyUnder 132 cm long and 13 cm
in diameter. A sUding contact may be moved to any point
along the spool so as to include any desired portion of the
self-induction.

The terminals for the spark on which the magnetic
field acts require different handUng according as the sub-
stance under examination is magnetic or not. In the
experiments with titanium, small pieces of the substance
known as "cast titanium," obtained from Eimer and
Amend of New York, were held in small brass clamps, the
vertical rods of which passed through larger horizontal
brass pieces set in a thick piece of fiber, through the
middle of which a brass rod passed and fitted into a
clamp, movable up and down on a support attached to
the base of the electro-magnet.

When iron terminals were used it was necessary that
they be held rigidly in place on account of the attraction
of the magnet. In all cases small cylinders of Norway
iron were screwed on the end of brass rods. The size of
the iron tips varied somewhat according to the kind of
spark desired and the width of the magnetic gap used.

Those most generally used with a strong field were 3.5 mm in diameter and about 10 mm long. In the
earher work the iron-tipped brass rods were held in a hard-wood frame composed of two vertical rings
held apart by four horizontal pieces. The wooden rings fitted over the magnet core, against the face of
each coil, while the brass rods passed with some friction through two of the horizontal wood pieces at
opposite ends of the diameter of the rings. A better holder for iron terminals was devised later. This is
shown in Fig. i and is a modification of that used for non-magnetic substances, the parts being much
more rigid. The rod of 6 mm diameter to which the iron tip is screwed passes through a square brass
rod 16 mm in thickness, having a saw-cut from the hole out to the end. A screw at right angles to this
saw-cut, worked with a bar, serves to clamp the rod so firmly that the magnet does not move it. As
the column supporting the holder is screwed to the base of the magnet, all parts could be clamped so
firmly that the iron tips were held exactly in place.

The spark length for both iron and titanium was usually short on account of the proximity of the
magnet poles and the tendency of the spark to jump to these. With iron terminals, particles were given

3^

"^^

APPARATUS AND METHODS. 1 1

off rapidly by the strong transformer discharge and it was necessary to clean these off every few minutes
and also to tile off the oxide from the iron tip. Titanium terminals wore away rapidly, owing to disin-
tegration of the metal, and the oxide also needed to be removed frequently if the brightest discharge was
to be obtained. The short spark gap necessitated an auxiliary gap in series, as otherwise the discharge
was not sufficiently disruptive to avoid melting the terminals. This auxiliary gap was a simple affair
of brass mounted on fiber.

When using the spark, the various parts of the secondary circuit, as well as the step-up connection
and the current in the primary, were adjusted to give the sort of spark desired. In this investigation
self-induction has been used in the spark circuit somewhat sparingly, since on the majority of the photo-
graphs it was necessary to obtain the fainter lines of sufficient strength for accurate measurement. Self-
induction in the spark circuit sharpens the Zeeman components in about the same degree that it sharpens
the lines of the regular spark spectrum, but the brightness of the spark is greatly diminished at the same
time, an eflect onl)- partially due to the decrease in intensity of the enhanced Unes. The weaker lines as
a whole, especially the faint and diffuse lines of iron, are so reduced by self-induction that very long
exposures are required to bring them out. A compromise must be made, since in exposures running
many hours, especially for more than one day, there is risk of instrumental disturbances. The method
followed was to use the spark with rather high self-induction for one or more photographs of any region
containing strong lines, and especially enhanced Knes, for which moderate exposure time was sufficient,
then to use small self-induction for photographs in which as many of the weak lines as possible were
desired. The loss of sharpness in such cases was counteracted as far as possible by the use of a narrow
slit and by selecting the kind of plate and developer which would give the sharpest definition and at the

same time show the Hues.

2. The Electro-Magnet.

This apparatus is of the Du Bois half-ring type, made by Hartmann and Braun of Frankfort. It is
shown (in its present state, after being rewound) in the photograph of the laboratory (Plate I). The
coils, as used until recently, were each wound with 1250 turns of No. 9 wire (diameter =3.0 mm). They
are clamped to a horizontal iron base which completes the magnetic circuit. The magnetic gap is varied
by mo\nng the coils upon this base, which is itself supported by three legs on an iron plate. A hole in
the center of this plate fits over a pivot in the middle of a round iron table, the ends of the plate resting
on a planed ring which forms the rim of the table. The magnet can thus be turned in any desired direc-
tion by rotating the base-plate upon the planed ring of the table. The magnet rests upon a cement pier
60 cm square and 82 cm high. The core of each magnet coil is pierced by a horizontal hole 17.5 mm in
diameter for the transmission of light along the lines of magnetic force. These holes are filled with cylin-
drical iron rods when such an axial opening is not needed.

A variety of pole-pieces was used for the magnet according to the way in which the spark terminals
were arranged and the directions in which the light was to be taken. Into each vertical face of the magnet
core is screwed the first section of the pole-piece, a truncated cone of soft iron 16.5 mm thick, whose
double angle is 112°. The small end of this cone is a circular plane surface 39 mm in diameter. To this
circular face was fastened a pole tip of one of the following forms, each of which has a double angle equal
to that of the truncated cone just described.

(a) For the observation of the fight from the iron spark parallel to the lines of force, the magnet
poles themselves were used as spark terminals in some of the earlier experiments. In this arrangement
the faces of the tips were circular, of 6 mm diameter. One pole was left solid and the other pierced with
a hole 3 mm in diameter, the spark being viewed through the tubular hole in the core. The pole-tips
were each insulated from the core by mica plates and held in place by fiber screws. The method gave
trouble, not only from the occasional breaking down of the insulation, but from the fact that the spark
did not stay in front of the hole in the pole-piece. It had the advantage, however, that the field was not
affected by the introduction of extra iron as spark terminals.

12

INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

Fig. 2.

(b) A stronger light from the spark was obtained for observation along the axis by not insulating the
pole-tips, using one solid and the other pierced as in (a), with the spark between iron tips at the ends
of brass rods held vertically between the magnet poles by means of a wooden frame or the brass and fiber
holder described on p. lo. Titanium terminals were held in the simple clamp described above. This
worked well for getting the "longitudinal effect" (^-component) without the introduction of a Nicol
prism in the optical system. Such an end-on arrangement is of course necessary for the study of the
circular polarization of Zeeman components. However, for general work in measuring the separation
of components, this method has the disadvantage that there is a considerable increase of field-intensity
close to the magnet poles, amounting with some gaps to 25 per cent, as well as an inequality at the two
poles resulting from one of them being pierced, so that the sharpness of the Zeeman components is not
all that could be wished.

(c) The most useful method, and that used (with varying shapes of the pole-tips) for almost all of
the best observations, was to set the magnet at right angles to the direction at which the Hght was
observed, use both pole-pieces solid, and separate the light by means of a Nicol prism over the sUt into that
vibrating in a plane at right angles to the magnetic force-lines, or parallel to these. This arrangement
made it possible to photograph successively the Zeeman components given respectively by vibrations
perpendicular and parallel to the force-lines by turning the Nicol prism through 90°, leaving the magnet
unchanged. Furthermore, by projecting the image so that only the hght from that part of the spark
midway between the magnet pole-pieces falls upon the slit of the spectrograph, the
change of field near the pole-pieces does not disturb the definition of the Zeeman com-
ponents. Even if the sht is long enough so that parts of the image come from regions
of different field-strength, the spectrograph, not being astigmatic, shows merely a
wider separation toward the ends of the components, the sharpness not being affected,
so that accurate measurements may be made by selecting the narrowest portion of the
separation.

Three forms of magnet pole-tips were used with this arrangement. In the first,
the conical tips ended in circular faces 6 mm in diameter. Tliis was used for most of
the work on iron and for the earlier work on titanium. With titanium, however, the
pieces of metal were irregular in shape and often rather large, so that with a short
magnetic gap it was difficult to bring the terminals close enough together to avoid sparking to the
magnet pole-pieces. The later and best set of titanium plates was taken with pole-pieces somewhat
chisel-shaped, made by miffing out opposite sides of a conical tip of 12 mm face to a thickness of 1.5 mm.
The thin ends were then placed parallel to each other and in a line with the beam of hght passing to the
slit. This gave a very uniform field for the hght of the thick spark, part of the vapor of which might
otherwise have gotten into weak portions of the field. Probably the best design is a modification of that
just described, in which the chisel edge was left 3 mm in thickness and 12 mm long, and not so deeply
milled as before. This form of tip gave a very strong field and a gap of 6 mm could be used without diffi-
culty. The drawing in Fig. 2 shows this design, with which a number of the later iron spectra were taken.

A current of 10 to 12 amperes from a 1 2. 5 -kw generator was generally used for the magnet circuit.
15 amperes could be used for runs of two or three hours, but the magnet rapidly became heated. This
current was almost sufficient to saturate the core and a larger current gave but a small increase of field.
The heating of the core by long-continued runs, even at 10 amperes, was considerable in warm weather,
when the two electric fans used to blow the sparks, and which also played on the magnet, e.xerted httle
cooling effect. Almost at the close of this investigation a very efficient means of cooHng the core was
devised. Injuries to the insulation of the wire made it necessary to rewind both magnet coils. When
the cores were laid bare, a spiral of soft copper tubing of 6 mm outside diameter and 4 mm bore was
wound around each core next to the iron. Strips of "5000-volt hnen" were wound over the spiral as

ArrAR.'XTUS AND METHODS.

13

insulating material, the face-plates at the ends of the coil being protected by ebonite sheets, and 1300
turns of wire were wound on each coil, the extra 100 turns on the two coils more than compensating for
the magnetic leakage caused by introducing the copper spiral.
With a stream of water flowing through the spiral, the core remains
perfectly cool and a current of 14 amperes may be used without
serious heating of the wire. This improvement has given an
increase of field of about 25 per cent over what could previously
be used for long runs with the same magnetic gap.

The current is controlled by means of two Ruhstrat sliding
resistances in parallel and is read to o.i ampere by a Weston
millivoltmeter with shunt used as an ammeter.

3. The Spectrograph.

The spectrograph which was used in this investigation was
described briefly in the general account of the Pasadena labora-
tory published in 1908 (53) . It is of the Littrow or autocollintating
t)-pe, placed vertically in a well 30 feet (9.1 m) deep. The design
of tliis spectrograph was worked out during the early solar inves-
tigations on Mount Wilson and the first instrument in the obser-
vatory equipment was made by William Gaertner of Chicago, and
has been in use for over three years as a part of the 60-foot tower
telescope on Mount Wilson. A description was published in
Contributions from the Mount Wilson Solar Observatory (54) . When
the physical laboratory in Pasadena was equipped in 1908, an
exact duplicate of the mountain spectrograph was obtained from
Gaertner, with the addition of holders for lens and plane grating
to give a focal length of 13 feet (4 m) when desired, as well as
the full focal length of 30 feet (9.1 m).

The details of the mounting of the spectrograph can be seen
from the drawing in Fig. 3 and from the photograph of the upper
end (Plate II). The well is made water-proof with a lining of
brick, several layers of tarred building paper, and cement plaster,
the dimensions being 30 feet (9.1 m) below the floor of the labora-
tory and 8.5 feet (2.6 m) in diameter. Since the well was thor-
oughly dried out, no moisture has appeared to come through the
walls. The cover of the well is of reinforced concrete, with two
openings. A circular opening at the east side is inclosed by a
cement ring 70 cm high and no cm outside diameter, which
supports the metal top of the spectrograph. Entrance to the
well is provided for by an opening at the south side closed by a
wooden cover, from which a vertical iron ladder leads to the
bottom. Attached to the iron ladder is a stout wood platform, at
such height that the parts of the spectrograph for the 13 feet
(4 m) focus can be conveniently adjusted.

The spectrograph consists essentially of a skeleton steel frame
50 cm square, at the top of which is a circular cast-iron plate on which is the slit and holder for the
photographic plate, while below, the objectives and gratings are supported in the steel frame at the

1 4 INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

proper levels for the focal lengths desired. The weight of the frame is supported by a concrete pier
placed at the bottom of the well. This pier carries an iron plate with a spherical cavity, into which
fits a lubricated hemisphere on the lower end of the spectrograph frame. The iron plate at the upper
end of the instrument fits loosely inside a circular iron casting imbedded in the concrete ring already
described. The whole spectrograph turns easily about a vertical axis by means of a gear and pinion in
the outer casting. A simple clamping device holds the instrument against accidental turning when in use.

The sUt of the spectrograph, 51 mm long, is placed on the end of a brass tube sliding within another
tube attached to the iron top. The divided head regulating the width of slit is graduated to read 0.025 mm.
For strong light sources, a slit width of 0.075 m™ ^^.s regularly used. When the 30-foot arrangement
is in use, the light passes from the sht to an 8-inch (20.3 cm) visually corrected objective by Brashear,
which lies horizontally in a holder capable of being moved vertically for focusing by turning a rod pass-
ing to the top of the spectrograph and rotated by a hand-wheel. A metal box to hold a plane grating
is just below the lens. A rod, geared to the grating box and passing above to a second wheel at the
top of the instrument, permits the rotation of the grating about a horizontal axis to obtain the order
or region of spectrum desired. Scales which show the position of the lens and the inclination of the grat-
ing can be read by a small telescope at the top of the instrument when illuminated by incandescent
lamps turned on from above. The fight reflected by the grating passes again through the lens and the
spectrum is brought to a focus above, the middle of the photographic plate lying in the same plane as
the slit. The holder carrying the plate rests in an iron frame supported at its center so that by tilting
the plate-holder good focus can usually be obtained over the whole of the plate, which is 17 inches (43 cm)
long and 3.62 inches (9.2 cm) wide. The plate-holder can also be moved parallel to itself by means
of a rack and pinion to permit the photographing of successive spectra. Two shutters, sfiding horizon-
tally, are placed 7.5 mm below the plate and can be adjusted to shut out all fight except the strip of
spectrum, the width of which is regulated by the length of sfit used. Light reflected from the lens sur-
faces would reach the plate were it not for these shutters and for the fact that a narrow bar is laid across
the center of the lens so as to cut off the reflected rays which would enter through the opening of the
shutters. With the 30-foot focal length, a slight inclination of the objective removed the reflections
without appreciably affecting the definition.

The arrangement of lens and grating to give the spectrograph a focal length of 13 feet (4 m) follows
the plan of that for the longer focus. The movements of lens and grating are regulated by the same
rods wliich control those below. The grating-holder may be moved over to the side of the steel frame
and the lens-holder swung back out of the way when the 30-foot arrangement is desired.

The two plane gratings used during the investigation were a Rowland grating 12.5 cm long and g cm
wide, having 568 fines to the milfimeter and a Michelson grating 19 cm long by 7.2 cm wide, having 500
fines per mm. The former was used with the 13-foot arrangement for the majority of the plates. The
Michelson grating was obtained near the end of the investigation and a number of the later plates were
taken with this, which was adjusted for the 30-foot focus. While longer exposure must be used with the
longer focus, the large scale is very desirable and the field is much flatter, so that as a rule the whole
length of spectrum over a 17-inch (43 cm) plate can be obtained in fair focus, even in the first order.
For very weak light-sources, however, the 13-foot arrangement often gives better results, as there may
be unavoidable changes in either the source or the spectrograph if the exposure is greatly prolonged.

The scales of the photographs for the two focal lengths and the several orders used in this work are
approximately as shown in the smafi table on the following page, there being a variation in the second
decimal place according to the part of the spectrum observed.

Other important features of the spectrograph are the occulting plate of the sfit, the mirror support,
and the polarizing apparatus. Plate II shows the form of the occulting plate. It is of brass, dull silver-
plated, and supported on four pins screwed into the top of the spectrograph, so that it is entirely free

ArrARATUS AND METHODS.

IS

Focus.

Order.

0
Angstroms

PER MM.

13 foot
13 foot
30 foot
30 foot
30 foot
30 foot

Second

Third

First

Second

Third

Fourth

2.05

I -35
1.92

0.9s
0.60

from the slit, and about 2 mm above the latter. By moving the V-shaped opening a, by means of the
rack and pinion, any length of slit up to 11 mm may be obtained. A further movement brings the
double opening b, whose size may be adjusted by the sHding plate c, over the slit. By a proper setting
of the scale d, a double comparison spectrum can thus be placed outside that made with the opening a
without risk of instrumental displacement, since the plate-holder and all essential parts of the spectro-
graph are left untouched.

Plate II shows the arrangement of the mirror by which the hght coming horizontally from any piece
of apparatus in the laboratory is reflected to the slit of the spectrograph. The holder for the mirror, which
is of plate glass 12.5 cm in diameter, silvered on its front surface, can be turned about a horizontal axis,
and is supported at the lower end of a brass cyhnder. This cylinder
can either rotate or move up and down inside a stationary cylinder held
in position by three curved iron supports which are screwed to the top
of the spectrograph. The mirror may thus be placed in any position
necessary to direct the beam into the instrument. As the mirror can be
turned in any direction independently of the spectrograph, we may have
any desired orientation of the slit with respect to the Hght source, which
is usually out of the cjuestion with a spectrograph mounted horizontally.
This is a very great advantage in an instrument free from astigmatism.

For the Zeeman photographs the slit was regularly used parallel to the lines of force of the magnet.
In photographing arc and spark spectra in general, it is desirable to use the slit sometimes parallel,
sometimes perpendicular to the direction of discharge in the image projected upon it.

The Nicol prism, by which the Hght polarized in one plane is transmitted to the sHt, is held on a metal
platform 3.5 cm above the sHt. The Nicol prism which has been used thus far was loaned by Director
Stratton of the National Bureau of Standards. The diagonals of the face are 25 and 30 mm and the prism
is 6.5 cm long. It is held in a brass cyhnder having a graduated circle by which the Nicol can be set at any
desired angle to the plane of polarization of the incident Hght. A second platform can be placed above
the Nicol to hold a Fresnel rhomb when this is desired for the study of circular polarization.

Since the beam passing through the Nicol is displaced parallel to itself, when the prism is rotated 90°
to transmit the other Zeeman component the image does not remain on the sUt. The image is then brought
back by moving the focusing lens, a simple glass lens of 58.4 cm focal length and 10 cm diameter. After
such a change, it was always noted whether the grating was well centered in the beam of Hght, which
usually had at least three times the diameter of the spectrograph objective. Although small movements
of a focusing lens of the focal length used produce very slight changes in the direction of the beam to the
grating, still care was taken never to move the lens when an instrumental displacement of the spectrum
Hues could have any disturbing effect. After an exposure with the magnetic field, the only change before
starting the exposure for the spark without field was to move the occulting plate above the sHt, so that
the comparison spectrum would be on each side of the spectrum taken with the field. The Hght source
thus remained unchanged in position, and all parts of the optical system as well as the photographic
plate were left untouched.

The spectrograph remains in adjustment for longer periods probably than with any mounting other
than the vertical arrangement in a well. The temperature change at the bottom of the well is entirely
negHgible during short periods of time. A recent test showed that during three months in which tem-
perature variations of over 15° C were experienced in the laboratory, a thermometer placed beside the
grating rose very gradually from i8°6 to 19°© C. During this time the Hghts were frequently turned
on to read the adjustment scales, and there were occasional visits by observers to the bottom of the well.
Mechanical vibrations are more disturbing. It has been necessary to close the driveway beside the
laboratory during e.xposures with the spectrograph, and to take care that no machinery be used which
would transmit a vibration to the spectrograph mounting.

i 6 influence of a magnetic field upon the spark spectra of iron and titanium.

4. Photographic Methods.

The requirements as to photographic plates in an investigation of this sort are to some extent con-
flicting. Speed, a fairly tine grain, good latitude, so that weak lines may be obtained without serious
over-exposure of the stronger ones, together with enough contrast to give sharply defined lines, are ele-
ments not easily combined in one plate. A number of plates have been tried, including the Lumiere
"Sigma," the Seed "Gilt Edge 27," "23," and "Process," the Cramer "Crown" and "Inst. Isochro-
matic." Each kind of plate will give superior effects for a certain t^-pe of line; but in general I have
obtained the best results for the work from the Seed "Gilt Edge 27" for the blue end of the spectrum as
far as about X4600, and from there on into the red from the same plate bathed with the solution of pina-
cyanol, pinaverdol, and homocol recommended by Wallace (55). This plate is the best adapted of those
I have tried in regard to doing justice to all classes of hnes. It is a fast plate without an objectionably
coarse grain. The latitude is good. In the case of hnes of complex Zeeman separation, a plate with more
contrast will often fail to show weak components very close to stronger ones.

A properly chosen developer will sharpen the hnes to a great extent, avoiding troublesome shad-
ing off from the central maximum. After trying several solutions, I have preferred a hydroquinone
developer giving strong contrast, due to Mr. Wallace, but not pubhshed so far as I know. The propor-
tions are as follows, using equal parts of A and B :

Solution A: Solution B:

Water. 48 oz. Water 48 oz.

Hydroquinone 640 grains Carbonate soda (anhydrous) i oz.

Sulphite soda (anhydrous) i oz. Carbonate potassium (anhydrous) 4 oz.

Sulphuric acid (cone.) 30 drops Bromide potassium >^ oz.

This developer does not stain the plates, even when warm. Development was usually carried to the
point where chemical fog sets in. This comes on slowly, and the solution is as efficient in bringing up
weak images as any I have tried. When used at 20° C a bathed plate is usually fully developed in 6 to
7 minutes. Some very good photographs were obtained for the region X5200 to X 5500 by the use of the
Cramer "Inst. Isochromatic"; but it was found best to soften its contrast by the use of a metol-hydro-
quinone developer. For the region X4800 to X5100, where the "Isochromatic" is weak, as well as for
the whole of the orange and red, the action of the bathed "27" has been unsurpassed by any plate used
in these experiments.

5. Measurement of Magnetic Field.

The accurate measurement of field-strength presented some difficulties in the case of iron on account
of the use of metallic terminals for the spark. The field for titanium was more easily obtained, and was
based on direct measurements by a bismuth spiral. This instrument was obtained from Hartmann and
Braun, but instead of using the regular formula for temperature correction, the spiral was sent to the
National Bureau of Standards and there calibrated to provide a series of curves for the variation of field-
strength with change of resistance for temperatures of 15°, 20°, 25°, 30,° and 35° C. When used at inter-
mediate temperatures the interpolation was simple. The resistance in and out of the field was measured
with a Kohlrausch bridge.

A set of plates of the titanium spectrum, extending over the whole region investigated, was taken
with the magnetic field as nearly the same as possible. All parts of the magnet were left unchanged
and the same current was used throughout. By check measurements with the bismuth spiral and by
comparison of plates which overlapped enough to measure some of the same lines on both, it appeared
that a field-strength of 17,500 gausses was maintained for this set with a variation no greater than 200
gausses. Other photographs taken to supplement the measurement of certain regions had their values
reduced to correspond to a field-strength of 17,500 by comparison of the separationsof sharply defined lines.

For the iron spectrum it is well known that indirect methods must be used to determine the field-
strength, since the use of iron spark terminals distorts the field to such an extent that any object as large

APPARATUS AND METHODS. I 7

as the bismuth spiral or an exploring coil for the ballistic method will not give true values for the field
to which the spark vapor is subjected. It may be that the iron vapor, when sufficiently dense, has an
appreciable permeability of its own. There is, however, no e\ddence on this point.

The plates for the iron spectrum were taken at intervals extending over a year, during which various
changes were made in the experimental arrangements which involved changes in the magnetic field.
However, a considerable region in the blue and violet was photographed with the same field, and the pub-
lication of Mrs. van Bilderbeek (49) gave an opportunity to make a comparison with her values. In her
work some photographs were taken using a spark with one iron and one zinc terminal, thus obtaining the
zinc triplet X 4680.3 17, as well as some iron lines in that region. Weiss and Cotton (20) by a series of very
careful measurements obtained the relation iAX/HX-'= 1.875 ^ 1° '* for the separation of the outer compo-
nents of this triplet, from which Mrs. van Bilderbeek deduced the value 32,040 gausses for the standard
field which she used when iron terminals alone were employed. I was able to select from my fist 33 lines
between the limits X3700 and X4400, which are also given in Mrs. van Bilderbeek's table, in nearly all
cases clear triplets, for which my measurements are of high weight. The ratio between Mrs. van Bilder-
beek's values and mine for these fines was in every case verj^ close to 2, the greatest deviation being given
by the value 2.14. The mean ratio for the lines is 2.01, giving a value of 15,940 gausses for the field used
by me in photographing the iron spectrum. This is in very satisfactory agreement with a value which
I had already determined by photographing the strong fine X 4383.720 as given by a spark between car-
bon terminals on which a little iron solution was placed in a field measured by the bismuth spiral as
17,600, and comparing the separation with that of the components of the same line very sharply
photographed with iron terminals used in the standard field. Exactly the same \-alue was given by com-
paring the separation of X4383.720 in two photographs, one with iron poles, the other in wliich the line
came up as an impurity in a titanium photograph taken with the standard titanium field of 17,500.
Assuming that the value of the field for iron was estabhshed by the other measurements, this last test
gave an excellent check on the standard field for titanium, which would otherwise depend on the meas-
urement with the bismuth spiral. It would seem then that the value of 16,000 gausses can be safely
taken for the standard iron field with an error less than i per cent. A considerable number of photo-
graphs for both iron and titanium were made with fields close to 20,000 gausses, sometimes slightly
higher, but the measurements were reduced to correspond to fields of 16,000 and 17,500, respectively.

A similar system of checking field-strengths was appfied for the region to the red of X4400. A spark
was used with one terminal of iron and the other of brass. Two photographs were taken in which the
zinc triplet X 4680.3 17 appeared as well as a number of iron lines, among them the wide and sharp
triplet X4878.407. Using the value of Weiss and Cotton, the field-strength for the measured separation
of this iron fine (20,360 gausses for AX=i.389 A) was deduced. Spark terminals of the same kind with
all parts of the magnet unchanged were then used for a series of photographs covering the iron spec-
trum as far as X6700. The field was thus kept as nearly constant as possible, and by comparing the
separations of iron lines with this known field with those on former plates taken with various fields, it was
possible to reduce all values for the iron spectrum to the standard field of 16,000.

6. Methods of Measurement and Reduction.

The measurement of the earlier plates was carried out by Miss Wickham, while the later plates were
measured by Miss Griffin. The machine used was a small Gaertner comparator having a range of 8 cm,
the divided head reading to o.ooi mm. The process of measurement included the identification of Unes,
the determination of the reduction factor for the portion of the plate under examination and the measure-
ment of the separation of the Zeeman components.

Various tables were used in the identification of lines. For the iron spectrum the tables of Kayser
and Runge (56) for the iron arc were supplemented by those of Exner and Haschek (57) for the spark.

I S INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

also by the list of enhanced lines given by Lockyer (58) and by plates of the arc and spark spectra of iron
taken in this laboratory. For the titanium spectrum the tables and charts of Hasselberg (59) were use-
ful as far as Xsgoo. This was supplemented for the red end by the measures of arc hues by Fiebig (60).
The spark tables of Exner and Haschek and of Lockyer were used as for iron. The identifications of
solar lines in Rowland's Tables are in most cases so close to the values in the tables of arc and spark
spectra that there is no doubt of the correspondence of the lines. The wave-lengths given in this
pubhcation are entirely on the Rowland system.

The chart of the iron arc spectrum by Buisson and Fabry (61) was of great assistance in the approxi-
mate identification of lines, the scale being almost the same as that of my plates taken in the third order
with the 13-foot focal length. In addition to using this chart for the iron spectrum, it served also for
titanium when used in conjunction with a set of plates which I made of the spectra of the titanium spark
and iron arc side by side.

The definitive identification of lines was in the usual way by measurement from neighboring lines
whose identity was certain. On account of the incompleteness of the general tables of spectra for the
red region, a few lines are entered in my titanium table which may belong to other substances. Some
of these, in all probability, are lines given stronger in the spark than in the arc, which explains their
absence from Fiebig's fist. The doubtful origin of such lines is indicated in the column headed " Remarks."

The spectrum given by the plane grating spectrograph not being quite normal, the reduction factor
of the plate, expressed in Angstrom units per millimeter, was determined at intervals usually of 2 to 3 cm.
The change in the factor between successive determinations was thus almost always less than 5 in the
third decimal place. This factor was multiphed by the distance in miUimeters between the Zeeman
components, which was the mean of at least four differential measurements taken alternately right and
left, setting first on one, then on the other of the components whose separation was desired. The accu-
racy of setting on first-class lines was usually well within 0.005 ^m- From such lines there are all grada-
tions up to those for which the measurements recorded can be taken only as indicating the order of mag-
nitude of the separation. Frequently a fine has its components on one side blended with those of an
adjacent line. In such a case it is usually possible to make a more or less accurate measurement of half
the separation by measuring from the clear component to the no-field line which was always photo-
graphed in juxtaposition. The accuracy of measurement will be discussed further in the explanation
of the tables when the weight of measurements is considered.

After measurement by a member of the Computing Division each plate was carefully gone over by
the author. In this examination the identification of fines was checked, the character of the separation
and weight of the measurement as determined by the quality of the fine were decided upon, and many
check measurements with the machine were made, including all measures for determination of the mag-
netic field by a comparison of the separation of lines on different plates.

EXPLANATION OF THE TABLES.
I. Wave-lengths.

The wave-lengths given in the first column are on the Rowland system. The methods of identifi-
cation and the tables used have been treated in the preceding section.

2. Intensity.

This column is intended to give an approximate value of the intensity of the fines in the spark spec-
trum. The numbers are taken (with occasional modifications) from the tables of Exner and Haschek
for the spark spectrum as far as X4700, beyond which the intensities were estimated on the same scale
from my plates. Weak fines are graded " i " on this scale, but there is considerable variation in thestrength
of fines which are given this value. For the purposes of this paper, this grading of intensities is sufficient.

3. Character of Separation.

In this column is described the type of separation of each line when the n- and /(-components* are
combined, as is the case when the light of the spark is observed at right angles to the magnetic force-
fines without Nicol or other apparatus to separate the light vibrating in the two directions. Thus in
the reproductions the two portions of each spectrum showing the effect of the magnetic field should be
superposed to give the appearance of the fine as described in this column.

The description gives the best judgment of the t}'pe of separation that can be made from the photo-
graphs. It must be considered in connection with the measured separation and widening of components
given in the columns for AX of the ti- and /(-components, and is usually made clear by these. Frequently
a supplementary remark is needed in the case of complex fines.

A fine designated as triple has its one p- and two ?2-components of sufficient sharpness to give no indi-
cation that any of them are compound. Since the Zeeman components follow to some extent the general
character of the spectrum fine, when a fine is itself wide and diffuse, its components may be simple and
still not so sharp as those of fines which do not tend to diffuseness. The pro.ximity of the no-field fine on
the plate aids in the judgment of such cases, but some of them are uncertain at the best. The tendency
of some fines to reverse is very disturbing in this connection, since it is very difficult to obtain such fines
with really sharp components. Several iron lines between X3700 andX3900, which give wide reversals
in the arc and spark between iron terminals, can be made to show the Zeeman components also reversed,
by the use of a strongly condensed spark, so that a triplet appears as a sextuplet. To decide such cases
it was necessary to make special photographs, using much self-induction and also with carbon terminals
containing a little iron. The titanium photographs were also useful in this connection, since the titanium
used contained enough iron to give the stronger iron lines which appear with sharp components under
such conditions.

The interrogation point is very freely used to indicate that the line is probably of the character given,
though not clearl}' shown to be so on the plates. The reason for doubt is usually given in the columns
for AX. Thus "triple?" means that the /(-component is sfightly widened so that it may not be simple,
but stiU the widening may be explained by the strength of the component or by the fact that the fine

*n and /> are used throughout this paper as abbreviations of "normal" and "parallel." n denotes the Zeeman components
given by light vibrations in a plane at right angles to the lines of magnetic force, and /> those given by vibrations parallel to the force-
lines. The symbols correspond to the letters 5 and p regularly used in German publications.

19

20 INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

itself is slightly diffuse, which may account for the lack of sharpness in the components. "Quadruple?"
means that the two ^-components are fairly sharp, but the /i-component is probably double. A doubtful
quintuplet will usually have five components measurable, with indications that others are possibly present.
Doubtful sextuplets are very common. As a rule such a line has its two H-components widened so that
there are probably two pairs, while the /^-component is either distinctly double, or unresolved and con-
siderably widened. The decision between doubtful sextuplets and septuplets is frequently difficult and
often quite uncertain. The /(-component in such cases is not resolved, but the character of its widening
will often show whether it is double or triple. A widening with strong central maximum means usually
three /(-components, but there may be five. Such a line, if it has two widened ^-components, is classed
as a probable septuplet. Octuplets and hues of higher separation are classified in a similar way, the widen-
ing given in the two AX columns, together with the remarks, showing in what respect the given char-
acter may be doubtful. Lines whose H-components are "fringed" are difficult to classify. Such "fringes"
indicate very close, unresolved components, and these may be numerous. A field double that available
here would probably show the full structure. Many hnes were fully resolved by a field of 20,000 which
had to be described as "fringed" for a field of 16,000. The degree of widening due to the fringes is given
in the AX column and a remark tells whether the fringes are toward the center or outwards. The number
of components is estimated as closely as possible from the width of the fringes, but when the structure
is very complex, an interrogation point is used without any attempt to give the number of components.
Although the doubtful elements which have been mentioned come into the estimates as to the char-
acter of hnes, the large number of plates from which the material was taken gave an opportunity to study
each fine under various conditions of intensity and degree of separation, so that the classification as to
character is probably as accurate as can be made without very much greater field-strength combined
with as high resolving power as was here used.

4. Weight.

Under this heading, each fine for wliich measurement was possible is given the weight 3, 2, or i,
according as the quahty of the Zeeman components for measurement is good, fair, or poor. This grading
should be of much service in any use which is made of these tables. In attempts which the author has
made to compare his measurements with those of others, the discordances were nearly always found to
occur in the case of lines of such character that one or both sets of measurements were poor. If hnes
of high weight in each set are compared, a good check on the observations is obtained.

Lines of weight 3 have sharply defined components, and for such hnes measurements of the same
plate by different observers or different sets by the same observer usually give differences in the third
decimal place only, while for many lines of this class the probable error is not greater than two or three
thousandths of an Angstrom. Only hnes of weight 3 should be used in comparisons of field-strength.

Lines are weighted 2, when the fine is reasonably strong, because the components are widened and
probably compound, fringed, or perhaps single and poorly defined for some reason, so that the measure-
ment is not so close as for lines weighted 3. Measurements of weight 2 have usually a probable error
not greater than 10 per cent and may be used for quantitative comparisons where a high degree of pre-
cision is not required. When a component is measured from the no-field fine, it is never weighted higher
than 2. A fine whose components are uniformly widened, each consisting of two or more components
of about equal intensity, gives a better measurement than a line whose components are fringed, since
in the latter case photographic conditions affect the distinctness of the maximum of each shaded com-
ponent, this maximum being the part measured.

Weight I is given to hnes which are very faint, much disturbed by blends, or of such complex struc-
ture that the components are extremely diffuse. The error of measurement for such hnes may be large
and the three decimal places are entered only for the sake of uniformity. However, the figures given

EXPLANATION OF THK TABLES. 21

show whether the line is to be classed as having small, medium, or large separation, and for this reason
the inclusion of such lines is justified.

When measurements are given for both the n- and the /^-components, the weight for each is given,
separated by a comma. In case only the /"-component is measured, a dash before the comma indicates
the omission of the weight for the K-component.

5. V.\LUES OF AX.

The fifth and sixth columns of the tables contain the separation in Angstrom units of the components
given by light vibrating respectively perpendicular and parallel to the lines of magnetic force. (See
foot-note, p. 19.) When there is an even number of components for the same polarization, measurements
are made between the members of each pair which presents itself. A single value in one of these columns
means that one pair of components is present. When there are two or more pairs, the largest separa-
tion is given first, but the innermost pair is designated "Pair I." When there is an odd number of com-
ponents, any outer ones that may appear are measured from the central component, instead of being
treated as pairs, and the values are listed beginning with the outermost on the violet side, the presence
of a central component being indicated by 0.000. No attempt is made to give the relative intensity of
the n- and /^-components, as this depends largely upon the optical system. However, if there are more
than two components for the same polarization, the relative intensity of the pairs (or of each component
when there is an odd number) is given in parentheses after the value of the separation.

If either AX column is blank for a certain hne, this indicates that a single, sharp component appears
for this polarization. Thus for all clear triplets, the /^-component column is blank. If, however, the
^-component is unresolved, but widened so as to indicate that a higher field would separate it into two
or more, the letter "w" with subscript i, 2, or 3 is used to show the degree of widening. Components
marked "W2" or "ws" as a rule are certainly compound. A slight widening, which probably means more
than one component, is indicated by "w,," but this may in some cases result from the diffuse character
of the no-field line.

There are many cases, especially in the ^-component column, where a measurement is given, followed
by "w" with a subscript. This means that a pair is measured, but each member of the pair is widened and
probably compound. If the widening is uniform, there are probably two or more components of equal
intensity. If the constituents of the widened component are of ditTerent intensity the component is
shaded toward one side. Such a line has the degree of widening given and in addition is denoted in
the "Remarks" column as "fringed" when each component shades off from the center, or as having
"inner fringes" when the shading is toward the center.

The letters "n.m." indicate that a separation exists but is not measurable, usually by reason of the
faintness of the components. In such cases it is possible, as a rule, to tell the character of the separation
with fair certainty and the line is included on this account. Thus a faint but sharp /)-component com-
bined with traces of two sharp w-components is given as a triplet. The designation "n.m.w." is used
when the components are hazy as well as faint.

6. Valxjes of AX/X^

Since in most points relating to the theory of the Zeeman phenomenon, the values of AX/X- rather
than of AX are considered (p. 4), the former quantity is entered in the seventh and eighth columns,
the positions of the numbers in the column being the same as that of the corresponding values of AX.
When X is in Angstrom units, the values given for AX/X- are to be multiplied by 10 *.

22

INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

Table i. — Measurements of Zeem.\n Effect for Iron.

Chakacter

H

AX

A\/X=

X

W

Remarks.

SEPARATION.

a
H

^

JI-COMP.

p-COVLP.

«-COMP.

/l-COMP.

3659.663

I

Triple

2

0.176

1-314

3669.666

2

Triple?

2

0. 176

Wl

1.307

3670.240

I

Quadruple?

2

0 . 26 1 \Vl

W2

1.938

3676.457

I

Triple

3

0.236

1.746

3677-457

I

Triple

I

0.144

1.06s

Faint in spark

3677-764

2

Triple

3

0.167

1.234

3679.002

I

Triple

I

0.268

1.980

Faint in spark

3680.069

3

Sextuple?

3,3

0 . 296 W2

O.III

2.186

0.820

3682.382

3

Triple

2

0. 230

1.696

3683.229

2

Sextuple?

2

0.480 Wl

W2

3-S39

«-comps. have inner fringes

3684.258

2

Triple?

3

0. 170

Wl

1.252

^-comps. almost resolved

3686.141

2

Sextuple?

2

0. 194 W'>

Wo

1.428

»- and /i-comps. diffuse

3687.610

4

Triple

3

0.3II

2.286

3689.614

2

Sextuple?

2,1

0.373 Wl

0.097

2-739

0.712

3690.870

I

Sextuple?

2,1

0.268 Wl

0.096

1.967

0.70s

3694.164

2

?

2

0.305 Wo

W3

2-235

n-comps. fringed. Probably 4
/)-comps. blended

3695.194

2

Triple

3

0. 261

1 .912

3697 567

I

Sextuple?

-,i

n.m. ws

0.345

2.522

«-comps. very diffuse. Probably
more than 4

3701 • 234

2

?

2

0.236 Wa

Wa

1-723

«-comps. fringed. Probably 4
/>-comps. blended

3702.170

I

Triple

2

0-315

2.298

3702.629

I

Triple?

2

0-333

Wl

2.428

3703.962

I

Triple

n.m.

Faint

3704.603

2

Triple?

3

0.319

Wl

2.324

3705-708

4

8 or 10
comps.

2,3

0 . 294 W2

0.147

2. 141

I .070

Probably 3 pairs K-comps. May
be 2 pairs /i-comps.

3707.186

I

Quadruple

2,1

0.282

0.250

2.052

I.S19

3707-959

I

Quadruple

2,1

0.627

0.306

4.560

2.225

Blend with 3708.068

3708.068

4

Sextuple?

2

0.315 Wl

Wl

2.291

K-comps. fringed

3709-389

4

Triple

3

0.312

2.268

3711-364

I

Triple

3

0.216

1.568

3716.054

I

Quadruple

1,1

0.290

0. 146

2.100

I -057

3716.591

2

Triple

3

0-394

2-853

3718.554

I

Quintuple?

III

0.271 (i)
0.000 (2)
?

0.286

1 .960
0.000

?

2.068

Unsymmetrical. Probably 3 »;-,
2 /)-comps. Red H-comp. not
measurable. Red />-comp. half
as strong as violet

3720.084

10

Triple?

2

0.268

1.936

All comps. may be compound.
Line reverses readily and
comps. are never sharp

3721.418

I

Triple

I

0.342

2.470

Faint in spark

3722.071

I

Triple

I

0. 266

1.920

Faint in spark

3722.729

4

1 2 comps.

2,2

Pair IV, 0.415 (2)
Pair III, 0. 311 (3)
Pair II, 0.211 (3)
Pair I, 0. 103 (2)

Pair II, 0.195 (5)
Pair I, n.m. (i)

2.994
2.244
1.522
0-743

1.407
?

3724-526

2

Triple

3

0. 256

1.845

3727-778

5

Septuple?

2

0.318 Wl

W2

2.288

«-comps have inner fringes.
Probably 3 /i-comps.

3730-534

I

Triple?

I

0.341 Wl

Wl

2-451

Blend with faint lines

3731-093

I

Triple

I

0.176

1.264

3732-545

2 1 Triple

3

0-399

2.86s

3733-469

3 , Quintuple

3,3

O.IS7 (2)

0.322

1. 127

2. 3"

0.000 (3)

0.000

0.158(2)

1. 134

3735-014

10

Triple

2

0.310

2.222

3735-485

I

Quadruple?

2,2

0.472 Wl

0.166

3.384

1 .190

3737-281

^

Septuple?

2

0.254 Wl

W2

1. 818

H-comps. fringed. Probably i
/i-comps.

3738-454

2

Triple

3

0.207

1.481

3743.508

6

Octuple

3.3

0.208(4)

0.107 (2)

1.484

0.763

Comps. of faint line to red (com-

1

0.104 (2)

0.000 (3)

0.742

0.000

puted 3 743 .6 1 5 ) are superposed

0.000 (2)

0.106 (2)

0.000

0-756

on central and outer red 11-

0.112 (2)

0.799

comps. giving apparent dis-

0.213 (5)

1.520

s>iimietry. Compare 378S.046

MEASUKKMKNTS OF ZEEMAN EFFECT FOR IRON.

T.\ni.E I. — Mkasuremknts of Zeeman Effect for Ikon — Continued.

23

Character

H

AX

A\/\-

\

"?

Remarks.

Z

SEPARATION.

0
1

»l-COMP.

^-COMP.

n-COMP.

/"-COMP.

3744.251

I

Septuple?

1,1

0.401 W2

0.283(1)
0.000 (l)
0.415 (l)

2.861

2.018
0.000
2.960

Probably 4 n-comps. Apparent-
ly 3 /i-comps, but faint. Com-
pare 3747

3745-717

5

Septuple?

2

0. 228 Wi

W2

1.624

K-comps fringed. Probably 3
/i-comps.

3746.05S

4

Unaffected

3747-065

I

Septuple?

I.I

0.413 W2

0.306 (l)
0.000 (2)
0.341 (l)

2.941

2.180
0.000
2.429

Apparently same type as 3744.
Comps. more distinct

3748.408

4

9 comps.?

2

Pair III, 0.316 (4)
Pair II, 0.226 (3)
Pair I, o.ioi (i)

W3

2.250
1 .609
0.719

Probably 3 />-comps. almost
resolved

3749-049

I

Quadruple?

2

0.240 Wl

W2

1.708

3749-631

10

Triple

2

0.289

2-055

3753-732

2

Triple

3

0395

2.803

3756.213

I

Triple?

2

0.300

Wl

2.126

3757-081

I

Triple?

I

0.197

Wl

1.396

3757-597

I

Triple

I

0.388

2.747

Faint in spark

3758-375

8

Triple

3

0. 269

1-905

3760.196

2

Triple

3

0-235

1.662

3760.679

I

Quintuple?

2

0.146 (i)
0.000 (i)
0-179 (i)

W2

1.032
0.000
1.265

n-comp. appears as unsymmet-
rical triplet. No evidence of
blend

3763-945

6

Triple

3

0.218

1.538

3765.689

3

Triple

3

0.228

1.608

3767-341

■>

Unaffected

3768.173

I

Triple

2

0.600

4-225

Faint in spark

3770.446

I

Triple

2

0.191

1.343

3773 -S03

I

Unaffected

3774-971

I

Se.xtuple

2,1

Pair II, 0.545 (i)
Pair I, 0. 274 (i)

0. 240

3 824
1.922

1.684

/i-comps. faint

3776.600

I

Triple?

2

0.229

Wl

1.605

3777-593

I

Quadruple?

n.m.

n.m.

Very faint. Wide separation of
/>-comp.

3778.652

I

Triple

I

0.344

2.408

3779.569

I

Triple?

I

0.277

1.938

3781.330

I

?

n.m.

W2

Many comps. n diffuse. Not
resolved

3786.092

2

Triple

3

0. 220

1. 535

3786.314

2

Triple?

n.m.wa

W2

n-comps. diffuse

3786.820

2

Unaffected

3788.046

4

Octuple

3,3

0.214 (4)
0.108 {2)
0.000 (i)
O.III (2)
0.219 (4)

O.III (2)
0.000 (3)
0.109 (2)

1. 491

0-7S3
0.000
0.774
1.526

0.774
0.000
0.760

Magnetic duplicate of 3743 . 508

3790.238

2

Sextuple?

2

0. 164 W2

W2

1 . 142

3794.48s

I

Triple

3

0.197

1.369

3795-147

5

Septuple?

2

0-325

W2

2.257

n-comps. have inner fringes.
Probably 3 />-comps.

3797.659

3

Triple

3

0.261

1.809

3798.655

4

Triple

3

0.326

2.259

3799.693

5

Triple

3

0.326

2.258

3801.820

I

Quadruple?

I

0.190

W2

1. 314

3805.486

3

Triple

3

0. 204

1.409

3806.865

3

Triple

3

0.226

1-559

3807.681

2

7 or g
comps.

2,2

O.IlS W3

0.109 (l)
0.000 (2)
O.IOO (l)

0.814

0.751
0.000
0.689

Sharp inner pair of «-comps.
measured. Wide fringes prob-
ably indicate 2 outer pairs

3808.423

Quadruple?

I

0.227

W2

1-565

3808.873

Triple

2

0.288

1-737

3810.901

Triple

3

0.255

1.756

3813.100

Septuple?

2

0 . 203 Wl

W2

1.396

H-comps. fringed. Probably 3
/i-comps.

3813-781

Triple

I

0.266

Wl

1.828

3814.671

Quintuple

I.I

?

?

Faint. Blend makes «-comp.

0.000 (2)

0.324

0.000

2.226

difficult

0.182 (l)

1.250

2 4 INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

Table i. — Me.asurements of Zeeman Effect for Iron — Continued.

X

H

!a

Character

•9

AX

AX/X^

SEPARATION.

1

M-COMP.

p-COUP.

W-COMP.

p-COUP.

38IS-987

10

Triple

2

0. 264

1. 813

3816.490 I i Sextuple? 1

I

0.258 Wo

W2

1.772

3817-786

I

Triple

I

0. 169

I -159

Difficult

3820.586

10

Triple

2

0.282

1.932

Line reverses easily. Comps.
never very sharp

3821.328

3

Triple

3

0.218

1-493

3821.981

I

Triple

I

0.I4I

0-965

3824.441

2

Triple

2

0.184

1.258

Not given by Rowland as Fe.
Computed X = 3824.463

3824-591

s

Triple

3

0-345

2-359

3826.027

8

Septuple?

2

0.274 W2

W2

1.872

M-comps. fringed. Probably 3
p-comps.

3827.980

7

Triple

2

0.225

I -535

3830-896

I

Triple

n.m.

Faint

3831.002 I

Triple

I

0.222

1. 512

3833-458

I

Sextuple?

2

0.25s Wl

Wl

1-736

3834.364

6

Septuple?

2

0 . 248 Wi

W2

1.687

»-comps. fringed. Probably 3
/>-comps.

3836.476

I

Triple

3

0. 266

1.808

3837.768 1 I

Quadruple?

2

0.198

W!

1-344

3S39.405 ! 2

Triple

3

0.222

1 .506

3839.762 ! I 1 Triple?

I

0.257

Wl

1.748

Enhanced line, diffuse

3840.580 4 9 comps.

2,1

Pair III, 0.337

(I)

0.061 (2)

2.28s

0.414

Difficult. Comps. not fully re-

Pair II, 0. 220

(2)

0.000 (3)

1-492

0.000

solved

Pair I, 0.106 (4)

0.059 (2)

0.720

0.400

3841-195

S

Triple

2

0.164

1. 112

3843-404

2

Triple

3

0.220

1.490

3845-310

I

Sextuple?

2,2

0.234 Wj

0.197

1.582

1-332

K-comps. almost resolved

3846 -554

I

Triple

2

0.247

1.670

Enhanced line

3846.943 1 2

Triple

3

0.300

2.027

1

3850.118 1 4 Unaffected

3850.962

2

Sextuple?

2,3

0.298 VV3

0.218

2.009

1.470

n-comps. almost resolved

3852.714

2

Quadruple?

2

0.268

Wl

1.805

3856-524

5

Triple

3

0.341

2-293

3859355

2

Triple

3

0-243

1.632

3860.05s

6

Triple

3

0.341

2.289

3861.479

I ' Triple

I

0.322

2.160

3863. 888

I

Triple

I

0.291

1.949

Very faint. Enhanced line

3865-674

4

Quintuple

3,3

0.172 (i)
0.000 (2)
0.171 (i)

0-340

1. 151
0.000
1.144

2.27s

3867-356

2

Triple

3

0-339

2.267

3869.692

I

Triple

2

0.406

2. 711

3871-963

I

Quadruple

2,2

0.272

0.125

1. 814

0.834

Enhanced line

3872-639

4

12 comps.

2,3

Pair IV, 0.452
Pair III, 0.344
Pair 11,0.225

(I)
(2)
(2)

Pair 11,0.231 (6)
Pair I, n.m. (i)

3-013
2.293
1 .500

I -54°

Pair I, 0.116 (i)

0.773

3873-903

2

Triple

3

0.226

1.506

3S78.152

4

10 comps.?

2>3

O.3II W3

0 . 1 5 I Wl

2.067

1.004

Probably 6 h-, 4 /i-comps.

3878.720

5

Triple

3

0.346

2.300

3883.426

I

Triple

2

0.312

2.069

3885-657

I 1 Se.xtuple?

2

0.234 Wl

Wl

i-SSo

3886.434

5 1 Triple

3

0.348

2.304

3887.196

3 Se.xtuple?

2,2

0.335 W-2

0.117

2.217

0.774

3888.671 '4 II comps.

2,2

0. 190 (2)

Pair 11,0.23s

1-256

I -554

Red n-comps. disturbed by blend

0.128(3)

Pair I, n.m.

0.846

0.072 (3)

0.476

0.000 (l)

0.000

0-077 (3)

0.509

0.134 (3)

0.886

0. 190 (2)

1.256

3888.971

I

Triple

2

0.355

2.347

3890.986

I

Triple

2

0.358

2.364

3892.069

I

Quadruple?

2

0.237 Wl

W2

1-564

Red «-comp. stronger. Violet
^-comp. stronger

3893.542 2 Ouadruple?

3

0.269

Wl

1-775

MKASUREMENTS OF ZEEMAN EFFECT FOR IRON.

Tabi.k I. — Measurements of Zeem.\n Effect for Iron — Continued.

25

g

Character

a

AX

AX/X=

X

z

OF

3)
0

Remarks.

H
Z

t-H

SEPARATION.

1

W-COMP.

#-COMP.

»-COMP.

p-COMP.

3894.057

I

Triple

2

0.3II

2.051

3895-803

3

Triple

3

0.347

2.286

3897.596

I

Triple

2

0.249

1.638

3S98.032

2

Triple

2

0.376

2.474

3898.231

2

Quadruple

2,2

0.707

0.352

4-653

2-317

3899.850

4

Triple

3

0 349

2.294

3903.090

5

10 comps.?

2,2

0.278 W3

0.152 Wl

1.82s

0.998

Probably 6 n-, 4 /i-comps.

3904.052

I

Sextuple?

2,1

0.233 Wi

0.098

1-529

0.643

3906.169

I

Triple

n.m.

Enhanced line

3906.628

3

Triple

3

0.347

2.273

3908.077

1

Quadruple?

2

0.254

W2

1.663

3909.976

1

Triple

I

0.350

2.289

Difficult blend with 3909.802

3913-775

I

Triple

2

0.326

2.128

3916.879

2

Triple?

2

0.311

Wl

2.027

3917-324

2

Septuple?

2

O.S54 W2

Wj

3.610

n-comps. have inner fringes.
Probably 4 «-, 3 /"-comps.

3918.464

Triple

n.m.

3918.563

Triple

2

0.333

2.169

3918.789

Triple

2

0.175

I -139

3919.208

I

Quadruple?

I

0.251

W2

1-634

3920.410

Triple

3

0.349

2.271

3923.054

Triple

3

0.351

2.281

3925-790

Quadruple?

2

0.300

W2

1.946

3926.086

Triple

2

0.376

2.439

3928.07s

Triple

3

0.352

2.281

3928.231

Triple

2

0.354

2.282

3930.450

Triple

3

0.352

2-279

3932.785

Quadruple?

I

0.413

W2

2.670

3935-965

Sextuple?

2,2

0.319 Wi

0.227

2.059

1.465

Enhanced line

3937-479

Triple

2

0.3S4

2.477

3939.288

Triple?

n.m. wi

Wl

Enhanced line. Diffuse

3941.025

Triple?

2

0.460 \Vi

Wl

2.962

3942.586

I

Triple

2

0.275

1.770

3947.142

Quadruple?

I

0.243 Wi

W2

1.560

3947 675

^

Quadruple?

I

0.383 w,

Wl

2-457

3948 . 246

Triple

I

0.247

1-585

3948.92s

2

Triple

3

0234

1.500

3950.102

2

Triple

3

0.34S

2.230

3951. 3"

2

Triple?

2

0.288 Wi

Wl

1.844

3952.754

I

Sextuple?

2,1

0.2S7 w,

0.145

1.837

0.927

Red />-comp. twice as strong as
violet

3953 303

I

Triple

2

0.291

1.862

3956 . 603

2

Triple

3

0303

1-935

3956.819

3

Triple

3

0.289

1.846

3957.177

I

Triple

I

0.283 Wl

Wl

1.807

Comps. hazy

3963 25 2

I

?

W3

Wl

«-comps. not resolved

3964.663

I

Triple

I

0.420

2.672

Faint

3966.212

2

Septuple?

2

0.474 W2

W3

3-013

Measurement is for wide pair
«-comps. which have inner
fringes, /i-comps. not resolved,
probably triple

3966.778

2

Sextuple?

2

0.338 W2

W2

2.147

3967.570

2

Triple

3

0. 198

1.258

3968.114

I

Triple?

n.m.

Wl

3969.413

5

Septuple?

2

0.354 Wl

W2

2.247

Probably 4 «-, 3 /'-comps.

3970.540

I

Triple

2

0.348

2.207

3971-475

I

Sextuple?

2

0.212 \Vi

Wl

1-344

3976.532
3976.692

I
I

Triple?
Triple

I
I

0.330
0.319

Wl

2.088
2.019

\

i

Difficult blend

3977.891

2

Triple

3

0.441

2.787

3981.917

I

Sextuple?

2,2

0 . 240 W2

0. 127

1-513

0.800

3984.113

2

Triple

3

0.216

1. 361

3985-539

I

Triple

2

0.282

I -775

3986.321

I

Sextuple?

2

0. 196 W2

W2

1-234

3990.011

I

Triple?

I

0.293 Wi

Wl

1.840

3990.525

I

Triple?

2

0.251

Wl

1-577

3994.265

I

Triple

2

0.283

1-774

26 INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

Table i. — Measurements of Zeem.\n Effect for Iron — Continued.

3996.140
3997 -IIS
3997-547
3998.205
4001 .814
4005 . 40S

4006 . 464
4006.776
4007.429
4009 . 864

4013
4014,
4017
4018
4022
4024
4029
4030
4032
4040
4044
4044

4045
4062
4063
4067
4068
4070
4071

4073
4074
4076
4078

4079
40S0
4084
4085
4085
4096
4098
4100
4107
4109

964

677
308
420
018
881
796
646
117
792
056
766
975
599
759
139
137
930
908
921

947
792

515
996
368
647
,161
467
, 129

■335
.901

■649
■953

Character
or

SEPARATION.

4114.606
4118.708
4120.368

4121.963
4122.673

4123.907

4126.040

4126.344

4126.798

4127.767
4130.196

4132.235

Triple

Triple

Sextuple?

Triple

13 comps.?

Triple?
Triple
Triple
Septuple

Triple?

Triple

Sextuple?

Triple

Sextuple?

Triple?

Quadruple?

Sextuple?

Triple

Triple

?
Triple?
Triple
Sextuple?
Triple
Sextuple?
Quadruple?
Triple?
Triple
Triple
Octuple?
Triple
Triple
Triple
Triple?
Triple?
Triple?
Triple?
Triple?
Sextuple?
Triple
Triple
Sextuple

Sextuple?

Triple

Triple

Triple?

Quadruple?

Triple?

Triple?

Triple?

Triple

Triple

Triple

13 comps.?

AX

2
2,2

2,2
2
2
2

2
I
I
2
2
I
3
2,3

2)2

3
2

W3

0.2S9
o. 266
0.226 Wl

0.415

0.461 W3

O. 211
0.383

o. 176
Pair II, 0.470 (i)
Pair 1,0.284(3)

O. 236 Wl

0.250

0.397 Wl

0.2S8

0.272 VVl

0 . 209 Wl

0.3II

0.274 Wl
O. 211

0.254

n.m. W2

0-319
0.298
o.

418 W2

o. 269

0.402 W2

0.418

0.366

o. 170
0.360

0.302 W3
0.386
0.184
0.479

n.m.

0.300

0.311

0.400

0.237

0.383 Wl

0.302

0-397
Pair 11,0.382 (i)
Pair 1,0.188(1)

0.376 Wl

0.271
0.244
n.m.
n.m.

0.402 Wl

0.370

0-335
o. 2S4
o. 196
n.m.

0.510 W3

^-COMP.

W2

Wl
W3

Wl

O.0S5 (1)

0.000 (2)

0.089 (l)

Wl

W2

Wl
Wl
Wl
Wl

W3
Wl

0.097

0.1«(
W2
Wl

0.149

Wl
Wl
W2
W2
Wl
W2

O.I9I
0.13s

W2
W2

W2
W2
Wl

Wl

Ws

AX/X2

1. 621
1.664
1.414
2.591
2.874

1.3151
2.3851
1 .096
2.923
1.766

1-465
1. 551
2.460

1.784
1. 681
I. 290

I-915
1.685
1.298
1. 555

950
820
532
628
430
525
208
025
169
819
322
106
877

1.798)

1.864

2.397J

1. 413

2.281

1.796

2.352

2.261

1. 113

2. 220

1-597
1.437

p-COMP.

2.364
2.173]
1.968 >
I.667J
1.150

2.987

0.529
0.000

0-553

0.588
I -143

0.897

1. 131

0.797

Remarks

7s-comp. not resolved
Faint

Measurement is for outer pair
«-comps. Wide inner fringes,
probably at least 8 «-comps.
and 5 /)-comps. Compare
4132.235

Blend

Very faint

«-comps. scarcely resolved

Comps. diffuse

Probably 6 H-comps.

Very faint

Blend makes measurement dif-
ficult

Faint

H-comps. hardly separated, p-

comps. almost resolved
Comps. faint and diffuse

Blend makes measurement dif-
ficult

Faint, ((-comps. rather widely

separated
Measurement is for outer ;/-

comps. Wide inner fringes, indi-
ting4pairs. 5/)-comps. almost

ca. --„-,,
rescfh'cd

Similar to 4005.408

MEASUREMENTS OF ZEEMAN EFFECT FOR IRON.
Table i. — Measurements of Zeeman Eki-ect for Iron — Continued.

27

g

Character

a

AX

AX/X^

X

Z
H
H

z

Remarks.

SEPARATION.

s

W-COMP.

p-COlSP.

tl-COiiP.

p-COUF.

4133.062

2

Sextuple?

2

0.273 w,

Wl

I -598

4134

840

2

Sextuple?

2

0.303 w,

Wl

1.772

4136

678

I

Triple?

2

0.252

w,

1.472

4137

156

I

Triple

2

0-314

1-834

4140

089

I

Triple

n.m.

Very faint

4142

025

I

Triple?

2

0.392 Wi

W2

2.285

4143

572

3

Quadruple?

2

0.280

Wl

1 .630

4144

038

S

Septuple?

2

0.393 Wi

W2

2.288

«-comps. have inner fringes.

Probably 3 /)-comps.

4147 836

I

Sextuple?

2

0.340 W2

W3

1.976

Diffuse p-comp. appears stronger
on violet side. Possibly blend

4149 533

I

Triple?

2

0.397 Wl

w,

2.30s

4154.071

2

Tnple

I

0.379

2.196

4154.667

2

Triple

2

0.379

2-195

4154.976

2

Triple

I

0.38s

2.230

4156.970

2

Sextuple?

2,2

0.367 Wi

O.I2I

2.123

0.700

4157948

I

Sextuple?

2

0.415 W2

W2

2.400

415S.959

I

?

I

0.589 Wi

Ws

3 405

/>-comp. very diffuse

4171 .068

I

Sextuple?

2,2

0.390 W2

O.II7

2.242

0.672

4172.296

I

Triple

2

0-315

1. 810

417^923

I

Triple

n.m.

4173.480

I

Quadruple

1,1

0.470

0.18s

2.698

I .062

Blend with next line makes
measurement difEcult

4173.624

I

Triple

n.m.

Enhanced line

4175.082

I

Triple?

2

0.374

Wl

2.146

4175.806

2

Triple

2

0.296

1.697

4176.739

I

Triple

I

0.420

2.407

4179.025

I

?

W3

W2

Comps. very diffuse, not re-
solved. Enhanced line

4181.919

4

Triple

3

0 339

1.938

4182.548

I

Quadruple?

I

0.272 Wi

W2

i.SSS

Faint

4185.058

2

Triple

3

0.390

2.227

4187.204

4

Septuple?

2

0.395 Wi

W2

2.253

«-comps. fringed. Probably 3
p-comps.

4187.943

4

Triple

3

0.402

2.292

4191-595

3

Septuple

2,2

Pair II, 0.540 (i)
Pair I, 0.264 (4)

O.I3S (i)
0.000 (2)

0.143 (l)

3-073
1.502

0.768
0.000
0.813

4195.492

I

Triple

2

0.320

1. 818

4196.372

I

Triple

I

0-3S9

2-039

Faint

4198.494

3

Triple

3

0-383

2.172

4199.267

5

Triple

3

0.276

1.565

4200.148

I

Sextuple?

I

0.364 W3

W3

2.062

Faint and diffuse

4201 .089

I

Triple

I

0.438

2.482

Faint

4202. 198

6

10 comps.?

2,2

0.323 W3

0.147 Wl

1.829

0.832

Probably 6 «-, 4 /i-comps.

4204. lOI

I

Triple

3

0.373

2. no

4206.862

I

Triple

I

0.338

1. 910

Faint

4207.291

I

Triple

I

0-317

1.791

Faint

4208.766

I

Quadruple?

I

0-413

W2

2.332

Faint

4210.494

3

Triple

3

0.806

4.S47

4210.561

I

Triple

I

0.411

2.319

Enhanced line Fei' Not identi-
fied by Rowland

4213.812

I

Triple

2

0.392

2.207

4216.351

I

Sextuple?

2,2

0.457 W2

0.236

2.571

1.328

4217.720

I

Sextuple?

I

0.402 W2

W3

2.259

Comps. very diffuse

4219.516

3

Triple

3

0.284

1. 594

4220.509

I

Triple

2

. 0.368

2.066

4222.382

2

Triple

3

0.475

2.66s

4224.337

I

Sextuple?

I

0.439 Wl

Wl

2.460

4225.619

I

Sextuple?

2

0.448 Wl

W2

2.508

4226.116

I

Triple

n.m.

Blend with adjacent lines

4226.584

I

Triple?

I

0.381

W2

2.133

4227.606

4

Triple

2

0.309

1.728

4233 328

2

Sextuple?

2

0.282 Wl

W,

1. 574

Enhanced line

4233 772

2

9 comps.

2,2

Pair III, 0.780(1)
Pair II, 0.550 (2)
Pair I, 0.265 (5)

0.140 (2)
0.000 (3)
0.140 (2)

4.351
3.068
1.478

0.780
0.000
0.780

28

INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

T.\BLE I, — Me.\sueemen'ts OF Zeem..\n EFFECT FOR Ieox — Continued.

g

Character

n

AX

AX/X^

X

1

H

Remarks.

SEPARATION.

0

W-COMP.

/l-COMP.

B-COMP.

p-COMP.

4236.112

Triple

3

0.452

2.519

4238.970

Triple

2

0.320

1.781

4240.014

Triple

2

0.440

2-447

4245.422

Triple

3

0 493

2.736

4246-251

Triple

I

0.273

1.514

4247 -59^

Triple

3

0-356

1.973

4248.384

Triple

2

0.377

2.089

4250.287

8

Septuple?

2

0.382 Wi

W3

2. 115

K-comps. fringed. Probably 3
p-comps,

4250.945

9

12 comps.?

2,2

0 . 246 W3

0 . 2 1 1 Wl

I. 361

1. 168

Probably 8 n-, 4 /)-comps.

4260.640

10

Triple

3

0-423

2-330

4267.122

I

Triple

I

0.300

1.648

4267.985

I

Triple

I

0.528

2.899

4268.915

I

Triple?

I

0.462

Wi

2-535

4271-325

9

Triple

3

0394

2 -160

4271-934

10

Triple

3

0.341

1.868

4282.565

I

Septuple?

2

0.310 W2

W3

1 .691

>j-comps. fringed. Probably 3
/i-comps.

4285.605

I

Triple

3

O.3IS

1.715

4294.301

5

Sextuple.''

2,2

0.319 W2

0.138

1.730

0.748

»!-comps. almost resolved

4298.195

I

Triple?

2

0.457

Wl

2.474

4299 410

5

Triple

3

O-406

2.197

4302-353

I

Triple

2

0.316

1.707

Enhanced line

4303-337

I

Sextuple?

2,2

0.415 W2

0.265

2.241

1. 431

Enhanced line

4305-614

I

Quadruple?

I

0.328 Wi

W2

1.769

4308- 081

15

J"'!?!^

3

0.320

1.724

4309-541

I

Triple

2

0.325

1-750

4315.262

3

Sextuple?

3,3

0.517 W,

0.090

2-777

0-483

4325-939

15

Triple

3

0-245

1-309

4327-274

I

Triple

2

0-313

I-672

432S.080

I

Triple

2

0. 246

1. 313

4337-216

2

Sextuple?

2,3

0 . 264 W2

0.154

1-404

0.819

4346-725

I

Triple?

I

0.292

W2?

1-545

Blend with air lines

4351-930

I

Septuple?

I

0.31 1 W2

W2

1.642

Probably 3 /i-comps. May be
C>, but given by Lockyer as en-
hanced line Fc

4352-908

2

Septuple?

2.3

0.416 W2

0.075 (l)
0.000 (2)
0-075 (l)

2-195

0.396
0.000
0.396

K-comps. fringed

4367-749

I

Triple

2

O.3II

1.630

4369.941

2

Triple

3

0.282

1-477

4376.107

2

Triple

3

0.424

2.214

4383-720

20

Triple

3

0-332

1.727

4385 • 548

I

Quadruple

2,2

0.367

0-391

1. 910

2.032

Enlianced line

4388-057

I

Triple

n.m.

H-comps. blended with adjacent
lines

4388.571

I

Triple

2

0.432

2.243

4391-123

I

Triple

n.m.

Faint

4404.927

15

Triple

3

0-334

1.720

4407.871

I

Triple?

2

0.651

Wl

3-247

4408.582

I

Triple?

2

0.488

Wl

2. 511

4415-293

10

Septuple?

2

0.338 W2

W2

1.734

K-comps. have inner fringes.
Probably 3 p-comps.

4422-741

I

Sextuple

2,3

Pair II, 0.432 (i)
Pair I, 0.154 (i)

0.280

2.208
0.787

1. 431

4427 -4S2

2

Triple

3

0-430

2.194

4430-785

I

Triple

2

0.719

3.662

4433-390

I

Triple?

n.m. W2

W2

Comps. diffuse and blended with
air band

4442-510

2

10 comps.?

2,2

0.485 V/3

0.184 Wl

2.458

0.932

Probably 6 11-, 4 /)-comps.

4443-365

2

Triple

2

0. 170

0.8O1

4447.892

2

Sextuple

2,3

Pair II. 0.721 (i)
Pair I, 0.449 (i)

0.307

3.644
2.269

1.552

Close to air line

4454-552

I

Sextuple?

2,2

0.445 Wl

0-I73

2.243

0.872

4459.301

2

Sextuple?

2,2

0 . 449 W2

0.127

2.258

0.639

4461.818

I

Triple

3

0-435

2.185

4466.727

2

Sextuple?

2

0-343 Wl

Wl

1.719

MEASUREMENTS OF ZEEMAN EFFECT FOR IRON.

Table i. — Measiirements of Zeeman Ekfect for Iron — Continued.

29

X

n

z

Character

OF
SEPARATION.

AX

AX/X2

Remarks.

2

H-COUP.

p-COMl'.

»-COMP.

/l-COMP.

4469.545

I

Triple?

2

0.438 Wi

Wl

2.192

4476.185

2

Septuple?

2

0.306 Wi

W2

1.527

«-comps. fringed. Probably 3
/>-comps.

4482.338

I

Sextuple?

1,1

0.401 Wi

0.146

1.996

0.727

«-comps. very diffuse

44S2.438

2

Quadruple?

1,1

0-139

0.229

0.692

1. 140

Probably also outer pair »j-comps .
Close blend with preceding

4484 . 392

I

Sextuple?

2

0 3S5 Wl

Wl

1. 915

4489.351

I

Triple

n.m.

Enhanced line

4491 57°

I

Triple?

Wo

«-comps. close, not resolved.
Enhanced line

4494-738

2

Septuple?

2

0.302 W2

W3

1-495

n-comps. fringed. Probably 3
^-comps.

4508.45s

I

Triple

I

0.184

0 90s

Enhanced line. Comps. diffuse

4515-508

Triple

2

0332

1.628

Enhanced line

4520-397

Triple

2

0.472

2.310

Enhanced line

4522.802

Triple

2

0.274

1-339

Enhanced line

4525-314

Sextuple?

1,2

0.457 W2

0.164

2.232

0.801

4528.798

Septuple?

2

0.358 W2

W3

1-745

»-comps. fringed. Probably 3
/,-comps.

4531-327

Sextuple?

2.2

0.398 Wl

0. 104

1.939

0.506

4548.024

Triple

2

0,311

1.504

4549.642

Triple

2

0.319

1.541

Enhanced line

4556.063

Triple

I

O.3II

1.498

Blend with 4556.306. Enhanced
line

4556.306

Triple?

n.m.

Wl

4584.018

2 Triple

2

0-376

1.789

Enhanced line

4592.840

I Se.xtuple?

2,2

0.416 Wl

0.126

1.972

0-597

4603.126

I Sextuple?

2

0.566 Wl

W2

2.671

461 1 .469

I , Triple

2

0.652

3.067

4619.468

I 1 Quadruple?

2

0.581 Wl

W2

2.723

4629.521

Triple

2

0.398

1.857

Very close to air line. Enhanced
line

4637-685

I ' Triple?

n.m.

Wl

n-comps. diffuse. Close to air line

4638.193

I Triple

n.m.

4647.617

I Triple

2

0-392

1. 814

4654.800

I 1 Triple?

I

0.543 Wl

Wl

2.506

n- and /"-comps. diffuse

4667.626

I Triple

2

0.481

2.207

4668.331

I Sextuple?

I

0.362 W"

Wl

1. 661

4679.027

I 1 Triple

2

0.465 Wl

2.124

4691 .602

I Triple

2

0-358

1.626

4707.457

I Triple?

2

0.365

Wl

1.647

4710.471

I Triple

I

0.242

1. 091

Blend with air line

4736-031

I Triple?

I

0.405 Wl

1. 80s

Weak, rather diffuse

4736.963

I 1 Sextuple?

2

0.426 W2

Wl

1.898

4741.718

I 1 Triple

n.m.

Faint

4745-992

I Triple?

n.m. W2

Wl

Comps. weak and diffuse

4787.003 I Triple

2

0.409

1-785

4788.952 I Triple?

n.m.

«-comps. diffuse

4789.849 i I Sextuple?

2

0.352 Wl

W2

1-534

4839-734 ! I Triple

n.m.

Too weak to measure

4859.928

2

Octuple

2,2

n.m.

Strong central «-comp. Trace of

0.275 (i)

0.271 (2)

1. 166

I. 147

faint outer pair

0.000 (4)

0.000 (3)

0.000

0.000

0.289 (i)

0.269 (2)

1.222

I -139

4871.512

3

II comps.?

1,2

0.336 W-2

0.193 (l)

1-414

0.813

»j-comps. fringed. Probably 3

0-093 (2)

0-392

pairs

0.000 (3)

0.000

0.087 (2j

0.366

O.181 (l)

0.762

4872.332

2 ; Sextuple

2,3

Pair II, 1.044 (3)
Pair 1,0.515 (2)

0-538

4.400
2. 171

2.264

4878.407

I Triple

3

1.092

4.574

Violet comp. i/2 stronger than

red

4890.948

3

10 comps.?

2,2

0.635 W3

0-341

2.656

1.424

n-comps. uniformly widened,

!

probably 3 pairs. Trace of in-

ner pair /,-comps.

30 INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

Table i. — Measurements of Zeem.an Effect for Iron — Continued.

^ Char.acter

H

AX

AX/X2

X

z^ r*i?

Remarks.

SEPAR.4TI0N.

«-COMP.

p-COMP.

W-COMP.

p-COUP.

4891.683

s

?

2

0.388 W2

Wo

1.620

«-comps. fringed. Probably 3

^-comps.

4903 . 502 I Sextuple?

2,1

0.866 Wi

0. 152

3.600

0.632

4911.963 I Triple

2

0-397

1.644

4919.174 5 Sextuple?

2,2

0.591 W3

0. 270

2.441

I. IIS

4920.68s 1 8 Triple?

2

0.447 W-.

W2

1-845

Probably complex, but widening
may be due to strength

4924.107 10 Septuple?

2

0.615 W2

Wa

2-537

K-comps. fringed, .i^t least 3 p-
comps. Enhanced line

4924.956 i I Quadruple?

n.m.

n.m.

Comps. weak and diffuse

4938.997 I Quadruple?

2

0.737 Wl

Wj

3-023

4939.868 I Triple

2

0.580

2-378

4946.568 I Triple

2

0.481

1.968

4957.480 2 , Sextuple?

2,2

0 . 630 Wl

0. 190

2.565

0.773

4957-785 5 Tnple?

2

0.501 Wl

W2

2.307

Widening may be due to strength

4966.270 I Triple

2

0.588

2-134

4973.281

I

Quadruple?

I

0.248 Wl

Wl

1.003

4978.785

I

Unaffected?

W2

Only narrow /;-comp. visible.
Faintness of line may prevent
appearance of others. Possibly
similar to 4859.928

4982.682

I

Triple?

2

0.469

Wl

1.890

49S3.433 ! I

Triple

I

0-587

2.362

Blend with adjacent lines

4984.028 I Quadruple?

2

0 549

W2

2.209

4985.432 , I Triple

2

0.452

2-337

4985.730 1 I ! Sextuple?

2

0.717 Wl

W2

2-885

4991 452 I Triple?

2

0.504 Wl

Wl

2.023

Faint

4994316 I , Triple

2

0.581

2.332

5002.044 I Triple

2

0.415

1 .660

Close to air line

5005 . 8g6 2 Triple

I

0.490

1-953

»-comps. blend with adjacent

line

5006 . 306

I

Sextuple?

2

0.295 Wl

0.182

1. 1 76

0.724

5012. 252

I

Triple

3

0.537

2.136

5015123

I 1 Quadruple?

2

0.374

W2

1.488

5018.629 3 Quadruple?

3

0.742

Wj

2.944

Enhanced line

5022.414 [ I 1 Triple

2

0.263

1-043

5027.305 I

Triple?

I

0-505

W2 ?

1.998

Diffuse comps. due to blend with
5027.939

5028.308 I

Triple

2

0.353

1.396

5039-428

I

Triple

2

0.600

2.363

5041-255

I

Triple

2

0.498

W2 ?

1-959

Widening of /i-comp. probably
due to blend with 5041 .069

5041.936 1 ' Triple?

2

0.517

Wl

2.034

5050.008 , 1 Triple?

3

0-439

Wl

1-723

5051-825

I 1 Triple

3

0.521

2.042

5065 . 207

I Triple?

I

0.341 W2

W2

1-329

Comps. diffuse, disturbed by

blend

5068.944

I

Triple

3

0.684

2.664

5074-932

I

Sextuple?

2

0.408 Wl

W2

1-585

5079.409 I

?

I

0.457 W-:

Wa

1. 771

Blend with adjacent lines. Prob-

ably at least 7-comps.

5079.921 I Septuple?

1,1

0-737

0 . 206 ( I )

2.S56 0.798 , Faint. Probably weak inner |

0 000 (2)

0.000

pair »i-comps.

0. 210 (i)

0.814

5083.518 I Triple

2

0-475

1.840

5090.954 I Sextuple?

n.m. W2

W2

Comps. weak and diffuse

5097.175 I Sextuple?

2

0.513 Wl

W2

1-974

5098.885 I Quadruple?

2

0.607

W2

2-333

5107.619 I Triple

I

0.404

1-5491

Blend

5107.823 I Sextuple?

I

0.625 Wl

W2

2-394^

5110.574 I Quadruple

1,1

0-S39

0.226

2.066

0.866

Measurement difhcult owing to
blend with 5109.827

5123.899 1 I 1 Unaffected

5125.300 1 I Sextuple?

2

0.519 Wl

W2

1.976

5127.533 I Triple

2

0.676

2.572

5131.642 1 Triple

2

0 957

3-634

Very faint

5133.870 I Sextuple?

2

0.459 Wl

W2

1 . 743 ' ' p-comp. almost resolved 1

MEASUREMENTS OF ZEEMAN EFFECT FOR IRON.

Table i. — Measurements of Zeem.w Effect for Ir(in — Continued.

31

g

Character

H

AX

AX/X2

X

cn

H

T? F M A R k* t:

OF
SEPARATION.

0

«-COMP.

/)-COMP.

K-COMP.

/i-COMP.

X\, Hi M. >\. K. i\, 3 .

5137558

Sextuple?

2

0.567 Wi

W2

2.148

5139-427

Triple?

2

0. 716

Wl

2-7i3\
2.622/

Close blend makes judgment of

5139.644

Quadruple?

2

0-693

W2

/)-comps. difficult

SI43II1

Triple?

I

0.578

Wl

2.184

Blend with adjacent lines

5151 .020

Quadruple?

I

0.629 Wi

wa

2.371

Very faint

5152.087

Octuple?

n.m.

n.m.

Probably 5 n-, 3 /i-comps. Very
faint

5159-231

Triple?

I

0.442 W2

W2

1.662'

Comps. very diffuse

5162.449

Triple?

2

0.5S6

W2

2.200

5167.678

8

Triple

2

0.462

1-730

5169.220

10

7 or 9
comps.?

2

0.563 W2

W3

2.106

Enhanced line. «-comps. fringed
and probably compound, p-
comp. much widened with
strong center. Blend with
5169.069.

5171-778

Triple

3

0.521

1.949

5191.629

Septuple?

2

0. 702 W.j

W3

2.606

H-comps. fringed. Probably 3
/"-comps.

5192-523

Sextuple?

2,2

0.749 Wl

0.213

2.780

0.790

5I95-II3

Triple

3

0-457

I -69s

5195.647

Quadruple?

n.m.

W2

Comps. very diffuse

5197-743

Triple

2

0.304

1. 125

Enhanced line

5 198. 888

Quadruple?

n.m.

W2

Comps. very diffuse

5202.516

Triple

3

0.683

2.525

5208.776

Triple?

0.623

Wl

2.294

5215-353

Triple?

2

0.625

Wl

2.296

5216.437

Triple

2

0.305

1. 1 20

5217-552

Triple?

2

0.615

Wl

2.259

5225-695

J.

Quadruple?

n.m.

n.m.

Very faint, /"-comp. apparently

wide doublet

5227-043

Sextuple?

2,2

0 . 949 W.J

0.281

3-472

1.030

Probably 4 w-comps.

5227.362

Triple

3

0.413

1. 512

5230.030

Triple?

2

0.615

Wl

2.248

5233-122

,

Septuple?

2

0.507 Wa

W3

1. 851

K-comps. fringed. Probably 3
/>-comps.

5234-791

Triple

2

0.385

2.147

Enhanced line

5242.658

Triple

2

0.385

1.400

5250.817

Triple?

2

0.618

Wl

2-243

5263.486

Triple

2

0.651

2.352

5266.738

3

7 or 9
comps?

2

0.502 W3

W3

1. 810

;j-comps. widely fringed. Prob-
ably 3 /i-comps.

5269.723

8

Triple?

3

0.501

Wl

1.804

5270.558

5

Triple

3

0.299

1.076

5273-339

I

Triple

2

0.651

2-343

5276.169

I

Sextuple?

I

0.431 Wo

W2

1-547

Enhanced line

5281.971

2

?

I

0.311 Wj

W3

I. IIS

K-comps. strongly fringed. Prob-
ably s /i-comps.

5283.802

3

Triple

3

0.623

2.232

5302.480

2

Triple

3

0.632

2.272

5316.790

4

Triple?

2

0.455 Wl

W2

1. 610

Enhanced line. Diffuse comps.

may be due to character of
line

5324-373

s

Triple

3

0.648

2.286

Red comp. slightly stronger than

5328.236

7

Septuple?

2

0.470 VVi

Ws

1.656

«-comps. fringed. Probably 3
/>-comps.

5328.696

3

Sextuple?

2,2

0.488 W2

0.275

1. 718

0.968

5340.121

Triple

2

0.664

2.329

5341-213

3

>

1)2

0.486 W3

0.429

I 703

1.502

K-comps. blurred, probably at
least 6

5353-571

I

Triple

..

n.m.

Very faint

5365.069

I

Sextuple?

I

0.354 W2

W2

1.229

Comps. very diffuse

5365 596

I

Triple

I

0.471

1.636

Blend with preceding line

5367.669

I

Triple?

2

0.414

Wl

1-437

Diffuse

5370.166

2

Triple?

2

0.456

Wl

1.581

Diffuse

5371-734

6

9 comps.?

2

0.413 w.

W2

1. 431

K-comps. fringed. Probably 6

H-, 3 /"-comps.

3 2 INFLUENCE OF A M.AGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

T.1BLE I. — Measurements of Zeem.an Effect for Iron — Continued.

s

Character

H

AX

AX/X-

X

2

OF

0

Remarks.

H

4

separation.

«-COMP.

^-COMP.

«-COMP.

^-COMP.

5383.578

Triple?

3

0.480 Wi

Wi

1.656

5393-375

3

Triple

2

0.673

2.3U

5397-344

6 '

Sextuple?

2-3

0 . 630 W2

0.222

2.163

0.762

5404-357

5

Triple?

2

0.467 Wi

Wl

1-599

Diffuse

5405 989

6

9 comps.?

2,2

Pair 11,0.461 (i)

0.II7 (2)

1-577

0.400 Probably third pair »-comps.

Pair I, 0. 222 (4)

0.000 (3)

0. 121 (2)

0. 760

0 . 000 outside
0.414

5411.124

3

Triple?

2

0.43s wi

Wl

1.486

Diffuse

5415-416

■;

Triple?

2

0.510 Wi

Wl

1-739

Diffuse

5424.290

5

Triple?

2

0.498 Wi

Wi

1.693

Diffuse

5429.911

6

10 comps.?

2,3

0 . 607 W3

0.300

2-059

1. 01 7 ,3 or possibly 4 pairs »-romps.
Probably weak inner pair p-

5434 -740

5

Unaffected

comps.

5445-259

2

Triple?

2

0.415

Wi

1-399

Diffuse

5447-130

5

12 comps.

2.3

Pair IV, 0.874 (i)

Pair II, 0.447 (6)

2.946

1 . 507 K-comps. barely resolved

Pair III, 0.701 (2)

Pair 1,0.226(1)

2-363

0.762

Pair 11,0.477 (2)

1.608

Pair I, 0.219 (i)

0.738

5455-834

4

Quintuple

2,3

0.347 (i)
0.000 (2)

0.345 (i)

0.680

1.163
0.000
1. 160

2 . 283 Central line of n triplet dis-

placed 0 . 009 A toward red

from no-field line

5463-494

Triple

I

0565

1.893

Very faint

5474-113

Triple

I

0.686

2.289

Very faint

5476.500

Triple

n.m.

5476.778

Triple

2

0.647

2. 157

5487-959

Quadruple?

n.m.

W2

Very faint

5497-735

3

Octuple

2,2

0.683 (4^
0.352 (2)
0.000 (i)
0.340 (2)
0.706 (4)

0.346 (2)
0.000 (3)
0.341 (2)

2.260
1. 165
0.000
1. 127
2 338

1. 144
0.000
1. 128

5501.683

3

Sextuple?

2

I .001 W2

W2

3-307

Appears as diffuse triplet. All
comps. doubtless compound

5507.000

3

9 comps.?

2

I .026 W3

Wa

3 383

Probably 6 «-, 3 />-comps. Outer
»-comps. strongest

5535-644

I

Triple

I

0.437

1-427

Blend with air line

5555-122

I

Sextuple?

I

0.502 Wi

W2

1.628

Weak and diffuse

5563-824

I

Triple

2

0.652

2.107

Faint

5565 931

I

Sextuple?

I

0.478 Wi

Wl

I -542

Comps. diffuse

5569.848

2

Septuple?

2

0.335 W2

W3

1.080

«-comps. fringed. At least 3
^-comps.

5573 -07s

3

Septuple?

2

0 . 469 W2

W3

i.Sii

«-comps. fringed. At least 3
/i-comps.

5576.320

I

Unaffected

5586.991

5

Septuple?

2

0.510 W2

Wa

1-634

«-comps. fringed. Probably 3
close />-comps.

5598-524

I

Triple

n.m.

Very faint

5603.186

I

Quintiple

2,2

0.372 (i)
0.000 (2)
0-343 (i)

0.728

1. 186
0.000
1.092

2.318

Compare 5455 834

5615-877

6 Triple

2

0.586

1-859

5624.769

I Se-xtuple.'

1,2

0.664 W2

0.481

2.098

1.520

5638.488

I Sextuple.'

n.m. W2

W2

Faint and diffuse

5659 052

I Sextuple?

2,2

0.724 W2

0.307 Wl

2.259

0.960

n-comps. diffuse

5662.744

I , Triple.'

2

0.596

Wl

1-858

Faint

5693 865

4 i Triple

n.m.

Very faint

5701.772

I : Triple

2

0.607

1.867

5706.215

I Triple?

2

0.605 Wi

Wl

1-858

5709.601

I Quadruple?

2

0.785 Wi

W3

2.408

^-comp. almost resolved

5718.055

I ; Triple

2

0.383

1. 172

5731-984

I Sextuple?

I

0.631 Wi

W2

1.920

Comps. diffuse

5752.254

I Sextuple?

n.m. W2

W2

Comps. diffuse

5753-344

( I : Triple

2

0-57S

1.737

5763-218

I Triple

3

0.631

1.900

5775-304

I Sextuple?

I

0.762 Wi

W3

2.279

Comps. diffuse

5816.601

I ?

I

0.442 W2

Wj

1.306

Probably numerous «-comps.

\'cn,- diffuse

MEASUREMENTS OF ZEEMAN EFFECT FOR IRON.
Tadi.k I. — Measurements of Zeem.\n Effect for Iron — Continued.

33

X

Character

OF
SEPARATION.

s

AX

AX/X^

Remarks.

0

1

n-coMP.

/l-COMP.

n-cOMP.

p-COiSP.

5856.312

I

Triple

n.m.

Very faint

5S59.809 ; I 1 Sextuple?

2

o.6g8 \vi

W2

2.033

5862.582

I Triple

2

0.684

1.990

5884.028

I Triple

2

0.436

1. 259

5905-895

I ?

Ws

W2

»-comps. not resolved

5914-335

I Quadruple?

2

0.658 Wi

W2

1.880

5930.406 ! I ! Triple?

2

0.587 w,

Wl

1 .669

5934-881 I 1 Triple

2

0.589

1.672

5952-943 I 1 Triple?

n.m.

Blend with air line

5975-575 i ?

n.m.

W2

M-comps. very diffuse

5977.007 I Tnple

2

0.635

1.777

5983.908 j I Triple?

I

0 . 60S Wi

Wl

1.698

5985.040 1 I Triple

2

0.662

1.848

5987.290 I 1 Triple?

I

0.624

Wl

1.740

6003.239 ' I 1 Triple

2

0.772

2.142

6008.785 1 I 1 Triple

2

0.652

1.806

6020.401 I Triple?

2

0.845 Wl

Wl

2.332

Diffuse

6024.281 I Triple

2

0.649

1.788

6027.274 I Triple

2

0.568

1.564

6042.315 III ?

n.m.

W2

n-comp. blurred

6056. 227

I 1 Sextuple?

I

0 . 499 W2

W2

1. 361

»-comps. diffuse, p almost re-
solved

6065 . 709

3 Triple

3

0.403

I 095

6078.710

I i Triple?

2

0.635 Wi

Wl

1. 718

Faint

6102.392

I 1 Triple

n.m.

Faint

6103.400

I ! Sextuple?

-2

n.m. VV2

0.583

1-565

re-comps. very diffuse

6128.124

I Sextuple?

I

0.414 W2

W2

1. 102

Faint and diffuse

6136.829

5 Triple

3

0.515

1.367

6137-915 ' 5 j Triple

3

0.654

1.736

6141.938

Triple?

2

1 .017

Wl

2.696

Faint

6148.040

Sextuple?

I

0 . 649 W2

Wl

1. 717

Very diffuse

6149-458

Quadruple

I.I

0,812

0.740

2.148

I -950

Enhanced line

6157-945

Triple?

2

0.719

Wl

1.896

6165.577

Sextuple?

n.m. W2

W2

»-comps. diffuse

6170.730

Sextuple?

2

0.725 Wj

W3

1.904

/)-comp. almost resolved

6173553

I ! Triple

2

'■590

4.171

Difficult. Blend with air band

6180.420

I : Triple?

n.m.

Wl

Faint

6188.210

Triple

n.m.

Faint

6191.779 ; 5

Triple

3

0.541

1. 41 1

6200.527 1 I

Sextuple?

I

I .026 W2

W3

2.669

Comps. diffused

6213.644

I

Sextuple

2,3

Pair II, 1.603 (i)
Pair 1,0.798(1)

0.572

4.151
2.067

1. 481

6215.360

I

Triple

2

0.583

1.509

6219.494

I

Sextuple?

2,2

0.991 W2

0.385

2.562

0.995

6230.943 6

Triple

3

0.767

1.976

6232.856 j I

Sextuple?

2

1 . 205 Wi

W2

3.102

6238.598 j I

Sextuple?

1,1

1 .042 Wl

0.512

2.677

I-3IS

Enhanced line

6246.535

2

Sextuple?

2,2

0.958 Wl

0.278

2.456

0-713

6247.774

I

Septuple?

I

0.670 W2

Ws

1. 716

Enhanced line. K-comps. very
diffuse. Probably 3 /)-comps.

6252.773

3

Triple

3

0.582

1.488

6254.456

Triple

3

0.95^

2.434

6256.572

Sextuple?

-2

n.m. W3

o.6g6

1.778

«-comps. ver>' diffuse

6265.348

Sextuple?

2,2

0.969 W2

0.278

2.469

0.708

6270.442

I ! Triple?

n.m.

Wl

Very faint

6291 . 184

I 1 Sextuple?

I

1 .014

W2

2.562

6298.007

I Sextuple?

n'm. W2

Ws

6301 .718

3 ' Sextuple?

2,2

1.063 W2

0.392

2.676

0.987

?i-comps. probably double

6302 . 709

Triple

2

1. 618

4.073

6315-517

Sextuple?

I

0.906 Wl

W2

2.271

Faint, ^-comp. almost resolved.
Enhanced line

6318.239

3

Triple

2

0.452

1. 132

6322.907 I I

Triple

2

0.965

2.414

6331-067 i I

Sextuple?

I

0.820 W2

Wl

2.046

Very faint

6335-554 I 3 i Septuple?

2

0.665 ^^'2

Wa

1.656

Probably 3 />-comps.

6337.048 1 3 Sextuple

2,2

Pair II, 1. 641 (i)

0.632

4.086

1-574

Pair I, 0.94O ( i)

2.356

1

3 4 INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

T.\BLE I. — Me.^surements OF Zeem.vn Effect for Iron — Continued.

g

Character

H

AX

AX

/X=

X

OF

separation.

H

Remarks.

H
Z
1— (

H
^

«-COMP.

^-COIIP.

«-COMP.

#-COMP.

6339.096

Triple?

n.m.

Very faint

6344371

Sextuple?

I

0.771 Wi

n.m.

1. 916

Very faint. />-comp. double

6355-246

Triple

I

0-73°

1.808

Very faint

6358 . 898

Sextuple?

I

0.746 W2

Wi

1.84s

Very faint

6380.958

Triple

I

0.464

I .140

Enhanced line

6393.820

8

Triple

2

O.S93

1.450

6400.217

9

Septuple?

2

0.802 Wi

W2

1.958

»-comps. slightly fringed. Prob-
ably 3 close /)-comps.

640S . 233

I

Septuple?

-,2

n.m.

0-343 (l)
0.000 (2)
0.349 (l)

0-837
0.000
0.852

Apparently 4 weak H-comps.
about equally spaced

6411.865

5

Septuple?

2

0.686 W2

W3

1.668

«-comps. fringed, probably 3
/"-comps .

6417-133

I

?

I

I .018 W2

Wa

2.472

n- and /i-comps. very diffuse.
Enhanced line

6420.169

I

Sextuple?

I

0.743 Wo

W2

1.802

«-comps. diffuse. Enhanced line

6421.570

2

Triple

3

0-993

2.408

6431 .066

3

Sextuple?

2

0.775 Wl

Wl

1.873

6436.630

I

Triple?

I

0 . 840 Wi

Wi

2.027

Diffuse

6456,603

I

Sextuple?

2

0.781 Wi

W2

1.873

Enhanced line. Possibly diffuse
triplet

6462.965

I

Sextuple?

-,2

n.m.

0.585

1.400

H-comps. faint and diffuse

6409 . 408

I

?

n.m.

W2

Very faint

6475.846

I

?

n.m.

W2

Very faint

6495-213

3

Triple

3

0.682

1. 617

6518.599

I

Triple

I

0.837

1.970

Very faint

6546.479

I

Triple

2

0.584

1-363

6569.460

I

Triple

I

0.921

2.134

Close to air line. Enhanced line

6575-270

I

Sextuple?

n.m. W2

n.m.

/>-comp. apparently double.
Very faint

6593.161

3

Triple

3

0.699

1.608

6594.121

I

Sextuple?

-,2

n.m. wi

OS79

1-332

Very faint

6627.797

I

Triple?

n.m.

Wl

Very faint. Enhanced line

6633-995

I

Triple

n.m.

Very faint

6663. 701

I

Triple

2

1.088

2.443

Enhanced line

6678.23s

3

Triple

3

0.787

1.76s

MEASUREMENTS OF ZEEM.AN EFFECT FOR TFTANIUM.

TAni,i-; 2. — Measurements of Zeeman Effect for Titaniim.

35

X

to

Character

AX

AX/X^

Remarks.

^

0

H

5

SEPARATION.

[4

n-coMP.

p-COlSP.

M-COMP.

p-COitP.

3659.901

10 Triple

2

0.243

1. 814

3660.774

2 Quadruple?

-2

n.m.

0.188

1-403

3662.378

10

Triple

2

0.199

1.484

3669.106

2

Triple

n.m.

3671.819

3 Quadruple? |

-.2

n.m.

0. 161

1. 194

3679.821

3

Quadruple?

n.m.

W2

3685.339

20

Triple

2

0.232

1. 70S

Enhanced line

3690.053

2

Triple

n.m.

3706.363

8

Quadruple?

2.3

0.201 Wi

0.139

1.463

1 .012

Enhanced line

3710.094

2

Quadruple?

n.m. w-2

W2

3717-539

2

Quadruple?

n.m. W2

W2

3721-779

5

Quadruple?

3

0.344

W2

2.483

Enhanced line

3722.729

3 1 Triple?

I

0-145

Wl

1.046

«-comp5. faint

3724.716

3 Triple

2

0.243

1.751

3725-300

3 Triple?

I

0.261

Wl

1. 881

«-comps. faint

3729-952

4 Triple

3

0.142

I.02I

3741-205

2 Triple

3

0.248

1.772

3741-791

10 Triple

3

0.263

1.878

Enhanced line

3748-232

6 Triple

2

0.190

1.352

Enhanced line

3753-003

5 Triple

3

0.277

1.967

3753-732

3 ' Quadruple?

2

0.374

W2

2.654

3757-824

6 Triple

2

0.199

1.409

Enhanced line

3759.447

20 1 Triple

2

0.288

2.038

Enhanced line

3761.464

10

Triple

2

0.207

1.463

Enhanced line

3762.012

3

Triple

3

0.253

1.788

Enhanced line

3771-798

3

Triple?

2

0.356

Wl

2.502

3776.198

4

Triple?

2

0.336

Wl

2.357

Enhanced line

3786.181

3 1 Triple

3

0.229

1.598

3813-537

3

Quadruple?

2

0.288 Wi

Wi

1.980

Enhanced line

3814.671

3

Sextuple?

2

0.319 W2

W2

2.192

Enhanced line

3836.229

3

Triple?

2

0.341

Wi

2.317

Enhanced line

3853-872

2

Triple?

2

0.267

Wl

1.798

3858.262

2 j Triple?

2

0.278

Wl

1.868

3866.577

2 1 Triple

2

0.284

1.900

3868.539

2 ' Triple?

2

0-343

Wl

2.292

3875-425

2 , Triple

2

0.322

2.144

3882.309

2 Triple

2

0-325

2.157

38S2.439

4 Se.xtuple?

2

0.399 Wl

W2

2.648

3883.033

3 Triple

2

0.279

1.850

3895-377

2 Triple?

2

0.309

Wl

2.037

3900.681

50 Triple

3

0.272

1.787

Enhanced line

3904.926

5 : Triple

3

0.240

1. 574

3913.609

20 ! Triple

3

0.219

1.430

Enhanced line

3914.477

2 1 Se.xtuple?

2,3

0.352 W2

0.206

2.298

1. 345

3921-563

2 II comps.?

I-

0.340 (l)
0.224 (3)

n.m.

2.210
1.456

Probably two pairs />-comps.
Compare X 3930.022

0. 100 (2)

0.650

0.000 (3)

0.000

0.097 (2)

0.631

0.198 (3)

1.287

0.320 (l)

2.080

3924-673

3

Octuple?

2,3

0.292 Wa

0. 162

1.895

1.052

Probably 3 pairs «-comps.

3926.465

2 Triple

3

0.247

1.602

3930.022

3 II comps?

2,3

0.264 (i)
0.187 (3)

0.292

1.709

I. 2X1

1.890

Trace of inner pair />-coraps.
Compare X 3921.563

0.09s (2)

0.615

0.000 (3)

0.000

0.094 (2)

0.609

0.178(3)

I. 152

0.267 (l)

1.729

3932.161

4 ' ?

2

0 . 406 W2

W2

2.626

Enhanced line, n-comps. have
strong inner fringes

3947-918

3 ?

2

0.098 W,

W2

0.629

«-comps. strongly fringed

3948. S18

4 Triple

3

0.186

I -193

3956-476

4 Se.xtuple?

2

0.229 Wl

Wl

1-463

n-comps. fringed

3958-355

5 Triple

3

0.287

1.832

3962.995

3

2

0.461 W2

W2

2.935

M-comps. have strong inner
fringes, similar to X 3932. 161

u

36

INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

Table 2. — ME.vsrREMEXTS of Zeem.an Effect for Tit.vnium — Continued.

B Character

d

AX

AX/X=

X

Remarks.

z
3

SEPARATION.

«-COMP.

p-COMS.

«-COMP.

^-COMP.

3964.416

Sextuple?

2

0.359^1

Wl

2-285

Probably inner pair H-comps.

3981.917

3 Triple

2

0.188

1. 186

3982.142

2 Sextuple

2,2

Pair II, 0.390 (i)

0.311

2.460

1 .961

Pair I, 0. 151 (i)

0.952

3982.630

3

10 comps.?

2,1

Pair III, 0.784 (i)
Pair II, 0.475 (2)
Pair 1,0.148(4)

0.565

4.942
2.994

0-933

3-561

Probably weak inner pair p-
comps. Difficult blend with
two preceding lines

3987.755

I Triple

n.m.

Enhanced line

3989.912

6 Triple

3

0-275

1.727

3998.790

6 Triple

3

0.317

1.720

Unsymmetrical. »;-comps. have
inner fringes, broader tor violet
comp. /"-comp. fringed toward

violet

4009.079

4

Sextuple?

2

0.347 W-2

W2

2. 159

4009 . 807

2

?

2

0.086

?

0-535

Unsymmetrical. Violet )i-comp.
3 times strength of red. p-
comp. hazy, displaced toward
violet

4012.541

4 Octuple?

2,3

0.198 w.

0.169

1.230

I .050

Enhanced line. Probably 3 pairs
H-comps.

4021.893

2 ?

n.m.

M-comps. diffuse, narrowly sep-
arated. />-comp. fairly sharp

4024.726

3 Sextuple?

2

0.394 VVl

Wl

2-432

«-comps. have inner fringes

4025.286

3 Sextuple?

2,3

0. 263 W2

0.129

1.623

0.796

Enhanced line

4026.691

2 1 Triple

2

0.220

1-357

4028.497

5 ■ Triple

3

0.269

I-65S

Enhanced line

4030.646

2 Triple?

2

0.243 Wi

Wl

I -495

4035-976

2 Triple

2

0-354

2.173

4053-981

5 Triple

3

0.232

1. 41 2

Enhanced line

4055-189

3 Triple

3

0-39S

2.402

Enhanced line

4060.415

3 Triple

3

0 395

2.396

4064.362

2 Triple

3

0.396

2.398

4065 . 239

3 Triple

3

0-39S

2.390

4078.631

4 Triple

3

0.395

2-374

4082.589

3 , Triple

3

0.398

2.388

4112.869

2 1 Sextuple?

I>2

0.301 W2

0.236

1-779

I -395

n-comps. very diffuse

4122.306

2 : Triple

2

0.261

^■5i(>

4123-713

2 1 Triple

2

0.264

I -552

4127.689

3 ' Triple

2

0.291

1.708

4137.428

2 Triple

2

0-365

2.133

4I5I.I29

3 Triple

3

0-305

1.770

4159.805

2 Triple

2

0.263

1.520

4I6I .682

3 Sextuple?

2

0.495 W-2

Wj

2.858

Enhanced line. »-comps. have
inner fringes

4163.818

20 Triple

3

0.294

1.696

Enhanced line

4I7I.2I3

2 Triple

2

0.210

1 .207

4172.066

1 5 Triple

3

0. 251

1.442

Enhanced line

4173.710

3 Sextuple?

2,2

0.361 Wi

0.096

2.072

0.551 i Enhanced line 1

4184.472

I ' ?

n.m. w-2

W3

, Enhanced line, all comps. diffuse |

4186.280

3 Triple

3

0.282

1.609

4200.946

2 Triple?

I

0.463

Wl

2.623

Faint in spark

4203.620

2 Triple

I

0-457

2.586

Faint in spark

4238.050

2 Triple

2

0.286

1-592

4256. 760

2 ' Sextuple?

2

0.396 VVl

2.186

4261 .748

2 Triple?

2

0.324 Wi

Wl

1.784

4263.290

4 Triple

3

0.331

1. 821

4270.329

2 Sextuple?

2,2

0.378 w-2

0.248

2.073

1.360

4272.701

2 Septuple?

2

0.364 VV2

W2

1.994

X given by Fiebig as 4272.581
agrees better with solar fine
X4272.590. »^-comps.have inner
fringes. Probably 3 /)-comps.

4274.746

4 Triple

3

0. 291

1-572

4276.587

2 Sextuple?

2

0-443 Wl

W-2

2.422

M-comps. have inner fringes

4278.390

2 Triple

3

0.304

1. 661

1

MEASUREMENTS OF ZEEMAN EFFECT FOR TITANIUM.
T..\BLF, 2. — Me.vsurements OF Zeem.^n Effect for Tit.\nium — Continued.

37

>•

Character

H

AX

AX/X^

\

H

RfM4RKS

HH

SEPARATION.

0
Id

W-COMP.

/l-COMP.

K-COMP.

^-COMP.

4281.530

2

Octuple

3.3

0.444 (4)
0.218 (2)
0.000 (l)
0.218 (2)
0.448 (4)

0.222 (2)
0.000 (3)
0.224 (2)

2.422
1. 189
0.000
I. 189
2-443

I. 211
0.000
1.222

4282.860

s

Triple

3

0.244

1-330

4285. 164

,■;

Quadruple?

3

0.566

Wi

3-083

4286.168

4

Sextuple?

2.3

0 . 400 W.2

0.166

2.177

0.904

4287.566

4

Sextuple?

2.3 •

0,421 W2

0. 146

2.290

0 794

42S8.038

2

Septuple?

2

0.516 Wj

W3

2.806

n-coraps. have inner fringes.
Probably 3 /.-comps.

4289.237

4

12 Comps.

2.3

Pair IV, 0.586 (i)
Pair III, 0.443 (2)
Pair 11,0.306 (2)
Pair I, 0. 144 (i)

Pair II, 0.288 (6)
Pair I, 0. 140 (i)

3.186
2.408
1.663
0.783

1.566
0.761

4290.377

10

?

2

0. 284 W2

W3

1-543

«-comps. strongly fringed. 3 or
more /.-comps. Enhanced line

4291. 114

2

Quintuple

3,i

0.221 (i)
0.000 (2)
0. 220 (i)

0.445

1.200
0.000
I- 19s

2.417

4291.375

2

Triple

I

0.230

1.249

Difficult blend with 4291. 114

4294.204

10

Triple

3

0.361

1-958

Enhanced line

4295.914

4

Unaffected

4298.828

4

Septuple

2.3

Pair II, 0.292 (i)

0.060 (2)

1.580

0-325

/i-comps. distinctly unsymmet-

Pair 1,0.145 (s)

0.000 (3)
0.086 (2)

0.784

0.000
0-465

rical

4299.410

3

Quadruple?

I

0,430

W2

2.327

4299.803

2

Triple?

I

0.356

Wl

1.925

4300.211

8

?

2

0.367 Wi

W2

1.98s

«-comps. fringed. 3 or more p-
comps. Enhanced line

4300.732

2

Septuple?

2

0.265 Wi

wa

1-432

n-comps fringed, probably 3 p-
comps.

4301 . 158

3

Triple?

3

0.350

Wl

1.892

4302.085

5

Sextuple

3.3

Pair 11,0.58s (i)
Pair I, 0.151 (3)

0.216

3. 161
0.816

I. 167

Enhanced line

4306.078

8

Septuple?

2

0.367 Wi

Wa

1.979

n-comps. fringed, probably 3 p-
comps.

4308.081

8

Octuple

2,2

Pair III, 0.588 (i)
Pair II, 0.442 (2)
Pair I, 0. 291 (3)

0.236

3-168
2.382
1.568

1.272

Blend with iron impurity line
probably gives apparent dis-
symmetry in intensity of 11-
comps. Enhanced line

4311.880

I

Sextuple?

2

0.147

Wl

0.791

Outer pair H-comps. not measur-
able. Possibly 3 /■-comps.
Blend with faint lines

4313 034

8

Sextuple?

2,2

0 . 449 W2

0 159

2.414

0-855

Enhanced line

4314.964

3

Triple

3

0.424

2.277

4315.138

5

Quadruple

3.3

0.392

0-349

2.10S

1-874

Enhanced line

4316.962

2

Triple

3

0.207

I. Ill

Enhanced line

4318.817

3

Triple

3

0337

1.807

4321. 119

3

Sextuple

3,3

Pair 11,0.78s (2)
Pair I, 0.257 (i)

0.261

4.204
1-376

1-398

Enhanced line

4321.813

3

Triple

3

0.310

1.660

4323 531

I

Triple

2

0.464

2.482

4325 306

3

Triple

3

0.301

1.609

4326.520

2

Triple

3

0.403

2.152

4330.405

^

Sextuple?

2

0.415 Wi

W2

2.213

Enhanced line

4330.866

3

?

2

0.654 Wj

Wj

3-487

Enhanced line. «-comps. have
inner fringes

4338.084

10

Triple

3

0.247

1. 312

Enhanced line

4341.530

3

?

W3

W2

Enhanced line. Probably nu-
merous close K-comps, not
resolved, center strong

4344.451

3

Sextuple?

2

0.474 W2

W!

2.512

Enhanced line

4346.278

2

Sextuple?

2

0.453 Wl

W2

2.398

4351.000

2

Triple

3

0.387

2.044

Enhanced line

4354.228

2

Quadruple?

2

0.300 Wi

Wl

1-583

4360.644

2

Triple

3

0.348

1.830

1

3 8 INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

T.\BLE 2. — Measurements of Zeeman Effect for Titanium — Continued.

Character

H

AX

A\/\'

\

K

Remarks.

H

1— I

SEPARATION.

0

W-COMP.

p-COUP.

W-COMP.

#-COMP.

4367 ■ 839

6

Triple

3

0332

1.740

Enhanced line

4369-873

2

Triple

3

0.306

I .602

4372.498 2

Triple

3

0.314

1.643

4374 981

^

Triple

3

0-334

1-745

Enhanced line

4387.007

s

Triple

3

0.294

1.528

Enhanced line

4391-192

2

Septuple?

2

0 . 304 Wo

W3

I-S77

Enhanced line. Probably 4 n-,
3 /i-comps.

4394.093

2

Triple

2

0.325

1.683

4394-225

2

Triple

2

0.420

2-175

4395-201

20

Triple

3

0.347

1.796

Enhanced line

439b. 008

2

Triple?

3

0.374

Wi

1-935

Enhanced line

4398.460

I

Quadruple?

2,3

0.144

0.224

0-744

I. 158

Possibly faint outer H-comps, but
not visible on strong photograph

4399-935

6

Triple

3

0-431

2.226

Enhanced line

4404-433

4

Triple

2

O.4II

2. 119

4405.082

I

Triple

I

0.316

1.628

4405.896

I

Se.xtuple?

-3

n.m. W2

0.314

1.617

4409 . 408

I

Se.xtuple?

1,1

0.523 Wi

0. 171

2.690

0.880

4409 . 683

I

Se.xtuple?

-,i

n.m. W2

0.248

1-275

441 I 240

5

Triple

3

0.291

1.496

Enhanced line

4417-450

2

Triple

3

0.381

I -953

4417.884

6

10 comps.?

2,2

Pair 11,0.288 (i)

Pair II, 0. 240 (2)

1.476

1.230

Probably weak pair «-comps.

Pair I, 0.120 (2)

Pair I, 0.072 (3)

0.615

0.369

outside. Enhanced line

4418.499

2

Sextuple?

2

0.402 Wi

Wl

2.060

4421.928

2

Triple

2

0.289

1-478

Enhanced line, blend

4422. 104

2

Quadruple?

2,1

0.358 Wi

0.107

1-831

0.547

4422.98s

2

Triple

3

0.377

1.927

4424.531

I

Triple

2

0.304

I -553

4426.201

1 ^

Triple

3

0.318

1-623

4427. 266

4

Triple

3

0.312

I -592

4430.524

2

Sextuple?

2

0.468 Wj

W2

2-384

n-comps have inner fringes

4431-453

I

Triple

2

0.154

0.784

4433-742 I

Triple

3

0.187

0.951

4434.168 3

Triple

3

0.259

1. 317

4436.750 2

Sextuple?

2

0.46b Wi

W2

2.367

4438-359 1 I

Sextuple?

1.2

0.441 W2

0. 180

2.239

0.914

4440.515 2

Sextuple?

2.3

0.270 Wi

0.168

1.369

0.852

4441.433 I : Quadruple?

I

0.413 VVi

W2

2.094

4443-976 15

Triple

3

0. 298

I -509

Enhanced line

4444.728 ! I

Sextuple?

2,2

0.317 W2

0.245

1.604

1.240

4449-313 S

Triple

3

0.388

1.960

4450.654 4

10 comps.?

2,3

0.388 W3

0.264 Wl

1.958

1-333

Probably 6 n-, 4 /)-comps. En-
hanced line

4451.087 1 3

Triple

3

0.340

1. 716

4453.486 3 : Triple

3

0. 210

1. 059

4453-876 3 1 Quadruple?

3

0. 263

Wl

1-326

4455-485 4

Triple

3

0 351

1.768

4457.600 5

Triple

3

0.400

2.013

4463-569 ; I

Quadruple ?

I

0. 509 Wi

W2

2-554

4463.843 : I

Triple ?

I

0.509

Wl

2 -554

4464,617 ! 2

' Quintuple

2,3

0.285 (l)

0.000 (5)
0. 262 (l)

0.287

1-430
0.000
1-314

1.440

Enhanced line

4465.975 3 i Quadruple?

3

0.481 Wi

Wl

2.412

4468.663 15 1 Triple

3

0.340

1.702

Enhanced line

4469.316 ' I 1 Triple

2

0.426

2-133

4471.017 1 2 1 10 comps.?

2,2

Pair III, 0.724 (i)
Pair 11,0.386 (i)
Pair I, 0.126 (i)

0.458

3.622

1.931
0.630

2.292

Trace of inner pair p-comps.

4471.408 2 9 comps.

2,2

Pair III, 0.826 (i)

0.113(2)

4.132

0.565

Pair II, 0.613 (2)

0.000 (3)

3,066

0.000

Pair 1,0.364 (4)

O.I16 (2J

1. 821

0.580

4475.026 2 Septuple?

2

0 . 509 \Vi

Wa

2 -542

Probably 4 n-, 3 />-comps.

4479.879 2 Triple

3

0.829

4-130

4480.752 I 1 Triple

2

0.611

3-043

4481.438 3 Triple

3

0.548

2.729

4482.904 2 Sextuple?

2

0 . 498 Wi

W2

2-478

1

MK.VSUREMENTS OF ZEEMAN EFFECT FOR TITANIUM.

Table 2. — Me.\surements of Zeem.\n Effect for Tit.anium — Continued.

39

b Character

H

AX

AX/\2

X

a

Remarks.

H SEPARATION.

t-H

0

n-coMP.

p-COMP.

«-COMP.

p-COVP.

4488.493

6

Triple

3

0.355

1.762

Enhanced line

4489 . 262

3

9 comps.

2,2

Pair 111,0.858 (1)
Pair II, 0.597 (2)
Pair 1,0.382(4)

0. Ill (2)
0.000 (3)
0. no (2)

4-258
2-963
1.896

o-SSi
0.000

0.546

Compare X 4471 .408

4495.182

Triple

I

0.353

1.747 1

4496.318

3 Triple

3

0-493

2-439

4497.842

I Triple

2

0-534

2639

4501-445

15 Triple

3

0. 298

1 .471 1 Enhanced Ime

4512.906

4 Triple

3

0.501

2.460

45IS.I98

4

Quadruple?

3

0.498

W:

2.440

4518.866

I

Triple

2

0. 220

1.077

4522-974

Sextuple?

3

0.502 w,

W2

2.454

4527-490

Octuple

3.3

0.324 (7)
0.162 (2)

0.164 (i)
0.000 (2)

1.581
0.790

0.800
0.000

0.000 (i)

0.165 (i)

0.000 0.805

0. 166 (4)

0.810

0.338(7)

I ■ 649

4529656

2

Sextuple?

2,2

0.358 W2

0. 278 Wi

1-745 1-355

Enhanced line

4533-419

5

Triple

3

0.469

2.282

4534.139

6

Triple?

2

0.360 Wi

Wi

I -751

n-comps. slightly fringed. En-

4534-953

4

Triple

3

0.449

2.183 !

4535-741

3

Triple

3

0-424

2.061

4536-094

3

Triple

2

0-323

I 570 >

4536.222

3

Unaffected?

i

1

No resolution. Blend with 36.094
may conceal slight widening of

«-comp.

4537.389

I

Triple

I

0-355

1-725

4544-190

I

Quadruple?

I

0.308

W2

1-492 \

4544-864

3

Octuple

3.3

0.334 (7)

O.I7I (l)

I. 617 : 0.828

Comps. in all respects similar to

o.i68 (2)

0.000 (2)

0.813 0.000

X 4427.490

0.000 (i)

0.166 (l)

0 . 000 0 . 804

0.170 (4)

0823

0.332(7)

1.607

4548 -93S

3

Septuple?

2

0.560 Wi

W2

2.706

H-comps. have inner fringes.
Probably 3 />-comps.

4549.808

20

Triple

3

0.440

2.125

Enhanced line

4552

632

4

Quadruple?

3

0.510

Wi

2.460

4555

662

3

Triple

3

0.506

2.438

4560

102

I

Triple

2

0.446

2.14s

4562

814

I

Triple

3

0.424

2.036

4563

939

10

Triple

3

0.276

1-325 i

Enhanced line

4568

409

I

Quintuple?

-.2

n.m.

0.293

.. .. ; 1.404

Only central »-comp. visible.

Line probably similar to

X 4464.617

4571.095

I

Triple

2

0.221

1.058

4572-156

20

Triple

3

0-319

1.526

Enhanced line

4590. 126

3

Octuple

2,3

Pair III, 0.549 (2)
Pair 11,0.360(3)
Pair 1,0.165(2)

0.288

2.606
1.709
0.783

1.367

6 «-comps. not completely re-
solved
Enhanced line

4599-408

I

Triple

3

0.423

2.000

4617.452

4

Sei)tuple?

2

0.404 Wi

W2

1-895

«-comps. fringed. Probably 3
/>-comps

4623.279

3

Septuple?

2

0.379 W2

W3

1-773

M-comps. fringed. Probably 3
/>-comps.

4629.521

2

9 comps.

3.3

Pair III, 0.873 (i)
Pair 11,0.535 (2)
Pair 1,0.173 (4)

0.172 (2)
0.000 (3)
0.180 (2)

4.072
2.496
0.807

0.802
0.000
0.840

4638.050

I

Triple

3

0.528

2-455

4639 538

2

Octuple?

2,2

0.594 W3

0.224

2.759 1.040 i Probably 3 pairs «-comps. |

4639.846

2

Quadruple?

2

0.549

n.m.

2 . 550

Blend prevents measurement of
p-comps,.

4640.119

2

Sextuple

3.3

Pair 11,0.864 (i)
Pair 1,0.535(1)

0-353

4.013 1.640
2.485

4645.368

2

Triple

3

0.880

4079

Violet comp. 3/2 stronger than

4650.193

2

Sextuple?

3

0.733 Wl

W2

3-390

40 INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

T.\BLE 2. — Measurements of Zeeman Effect for Titanium — Continued.

Character

H

AX

AX/\=

\

H

separation.

0

«-<;oMP.

p-COlSP.

B-COMP.

^OMP.

4656.644

^

Triple

3

0.29s

1.360

4657.380

I

?

W2

ws

Not resolved. «-comp.has strong

center with fringes. Enhanced

4667.768

3 Triple

3

0-349

1.602

line

4675 294

I Triple

3

0.506

2-315

4682.088

3 Triple

3

0.399

1.820

4688.554

I Triple

2

0.346

1-574

4691 523

2 Triple?

3

0.441

Wl

2.003

4697.101

I ! Triple

2

0. 196

0.888

4698 . 946

2 j Sextuple?

2

0,367 W-2

W2

1.662

n-comps. fringed

4710.368

3 Quintuple

3

Pair 11,0.486 (i)
Pair 1,0.183(1)

2. 191
0.82s

Single sharp ;>-comp. Only line
of type in spectrum

4722.797

2 Sextuple

2,3

Pair 11,0.555 (i)
Pair I, 0.224 (i)

0.390

2.488
1.004

1.748

4723.359

2 Sextuple?

2,2

0.453 W2

0.226

2.031

1. 013

«-comps. fringed

4731 356

2 Triple?

3

0.412

Wl

1. 841

4733 604

2 Triple

2

0.350

1.562

4742.979

5 Triple

3

0.29s

1. 311

4758.308

8 Triple

3

0.382

1.687

4759.463

8 Triple

3

0.430

1.899

4764.108

I , Triple?

W3

Enhanced line. Unresolved n-

comp. Diffuse

4769.991

I

Sextuple?

2

0.583 W2

W2

2.562

4778.441

3

Sextuple?

2,2

0.338 W2

0.184

1. 481

0.806

4780.169

5 Quadruple

3>2

0.498

0.243

2.180

1.064

Enhanced line

4781.913

2 10 comps.?

-I

n.m. W3

0.314 W2

1-373

All comps. wide and hazy.
Probably 3 pairs «-, 2 pairs p-
comps.

4792.702

3 1 Sextuple?

2

0.370 W2

W2

1. 611

4796-373

I 1 Triple?

2

0. 195 Wi

0.848

4798.169

I Sextuple

1,2

Pair 11,0.594 (2)
Pair 1,0.218(1)

0.388

2.580
0.947

I -68s

4798.293

I j ?

W3

W2

n-comps. diffuse, not resolved

4799 984

3 1 Sextuple?

2,2

0.329 W2

0.182

1.428

0.790

4805.285

10 ' Sextuple

2,2

Pair 11,0.643 (i)

0.149

2-785

0.645 1 Enhanced line |

Pair 1,0.364(3)

1-577

4805 . 606

5 Triple?

2

0.406

Wl

1-758

4808.733

2 ' Triple?

2

0.383 Wi

Wl

1-656

4811.235

I Triple

2

0.398

1.720

4820.593

4 Triple

3

0.390

1.678

4827.804

I Triple?

I

0.405

Wi

1-737

4836.313

2 Triple

3

0.364

1-556

4841.074

6 Triple

3

0.390

1.664

4848 . 60s

2 Triple

3

0.516

2 -195

4856. 203

7 , Triple

3

0.416

1.764

4864.362

I ! Triple

2

0.441

1.864

4865.798

I 1 Sextuple?

2

0.534 W2

W2

2.255

Titanium?

4868.451

5 ' Triple

3

0.317

1-338

4870.323

5 Triple

3

0.389

1 .640

4874.196

3 Triple

3

0.336

1. 414

Enhanced line

4881.128

2 Triple

I

0. 198

0.831

4885 . 264

8 Triple

3

0.425

1. 781

4900.095

6 Triple

3

0-395

1-645

49" -374

6 Triple?

3

0.434 w,

Wl

1.799

Enhanced line

4913 803

8 Triple

3

0.34S

1. 441

4915-414

I 1 Sextuple?

2,2

0.415 W2

0.234

1. 718

0.969

4920.047

3 Triple

3

0.387

1-599

4921-963

3 Triple

3

0-439

1. 812

4925-594

I 1 Sextuple?

-I

n.m. W2

0.36s

1-504

4926.334

I Triple

I

0.509

2.098

4928.511

3 Sextuple?

2

0 . 268 \V2

W2

1. 103

4938-467

3 Triple

3

0.392

1.608

4968.769

I Sextuple?

-I

n.m. W2

0.358

1-450

4975-530

I Triple

3

0.414

1-673

4978.372

I Quadruple?

2

0.238 \Vi

Wl

0.960

4981 .912

10 Triple

3

0.481

1-938

4989.325

2 Triple

3

0.329

1.322

MEASUREMENTS OF ZEEMAN EFFECT FOR TITANIUM.
Table 2.— Me.\surements of Zeem.^n Effect for Tit.xnrtm — Continued.

41

H

Character

(H

A\

^\|\^

\

s

(d

s

SEPARATION.

0

;i-COMP.

/>-COMP.

n-Qcms.

/>-COMP.

4991 . 247

10

Triple

3

0.458

1.839

4997 . 283

2

?

-.3

n.m.

0.415

1.662

Numerous K-comps. blurred.
Red n- and />-comps. stronger
than violet

4999.689

10

Triple

3

0.413

1-652

5001.165

3

Triple?

3

0.405

VVi

1. 619

5007.398 10

Triple

3

0-339

I-3S2

5008.632 I

Triple

2

0.414

1.650

5009.829

I

?

-,2

n.m.

0.279

I .112

Numerous »-comp5. blurred

5010.396

I

Triple?

2

0-354

Wi

1. 410

5013-479

.S

Triple

3

0-4SS

1. 811

5014.236

4

Triple

I

0.177

0.704

Titanium?

5014.369 5

Triple

2

0.217

0.863

5016.340 7

Sextuple?

2.3

0-S43

0.214

2.158

0.851

5020.208 , 8

Octuple?

2,3

0.507 W3

0.276

2.012

I -09s

Probably 3 pairs H-comps.

5023.052 8

12 comps.?

2,3

0.466 W3

0.370 w,

1.847

1.466

Probably 4 pairs H-comps.

/>-comps. have inner fringes

5025.027

7

10 comps.

2,2

Pair III, 0.684 (i)
Pair 11,0.416 (2)
Pair 1,0.133(4)

Pair 11,0.546 (6)
Pair I, 0.269 (i)

2.709
1.647
0.527

2.162
1.065

5025-749

S

Triple

3

0.471

1.865

5036.089 10

Triple

3

0.4S5

1.794

5036.645 8

Triple

3

0.436

1. 718

5038.579 8

Triple

3

0.340

1-339

5040.138 8

Triple

3

0.404

I 590

5053.056 3

Triple?

2

0.449

Wl

1-759

5062.285 : 3

Triple?

2

0.412

Wl

1.608

5064 244

I

Triple

n.m.

Very faint

5064 . 836

8

Triple

3

0.463

1.805

5066.174

I

Sextuple?

-,i

n.m. Ws

0.407

1-586

5069.592

2

Triple

2

0.235

0.914

S071 .666

4

Sextuple?

1,1

0 . 470 W2

0.275

1.827

I .069

5072.479

6

Triple

3

0.502

I-9SI

Enhanced line

5087.239

4

Triple

3

0.329

1. 271

5113-617

5

Triple

3

0.431

1.648

5120.592

7

Triple

3

0.434

I -655

5129.336

8

Triple

3

0.478

1. 812

Enhanced line

5145-636

6

Triple

3

0-493

1.862

5147-652

5

Septuple?

2

0.805 Wo

W3

3 038

Probably more than 7 comps.
^-comps. almost resolved

5152.361

S

Quadruple?

3

0.671 Wi

Wl

2.528

5154-244

4

Sextuple?

2

0.666 W2

W2

2.507

Enhanced line

S173-917

10

Triple

2

0. 292

1.091

5186.073

8

Triple

3

0.385

1-432

Enhanced line

5188.863

12

Triple

3

0.512

1.902

Enhanced line

5193-139

10

Triple

3

0.468

I -735

5201.260

3

Triple?

2

0.648 Wi

Wl

1.396

5206. 215

4

Triple

2

0.470

1.734

5210.555

TO

Triple

3

0.547

2.014

5219-875

4

Sextuple?

2

0.326 (2)
normal
0.264 (i)
0.400 (i)

W2

1. 191

0.969
1.468

Unsymmetrical. Probably 4 n-
comps., 2 violet blended, 2 red
separated. /)-comp. displaced
0.062 to violet from normal.
All comps. measured from nor-
mal

5222.849

3

Unaffected

5223.791

3

Triple

2

0.437

1.887

5224-471

8

Triple

3

0.631

2.312

5224.712

5

Triple

2

0.608

2.227

5225.198 6

Triple

2

0.548

2.007

5226.707 10

Triple

3

0-349

1-277

Enhanced line

5238-742

3

Triple?

2

0.379 Wi

Wl

1. 381

5247.466

2

Sextuple?

I

0.773 W2

W2

2.808

5252.276

3

Sextuple?

2

0 . 686 W2

W2

2.487

5255-973

3

Triple?

2

0.647 Wi

Wl

, 2 342

5260.142

I

Triple

2

0.430

1 I -554

1

5262.321

I

Sextuple?

-

0.739 "■!

W3

2.671

Enhanced line

42

INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

Table 2. — Measurements of Zeeman Effect for Titanium — Continued.

>

Character

H

AX

AX/X^

X

z

OF

g

Remarks.

z

SEPARATION.

1

K-COMP.

p-COUP.

n-COMP.

p-COUP.

5263.669

I

Triple?

2

0.701 Wi

2 -530

5266.141

6

Triple

3

0.498

1.796

5282.576

2 Sextuple?

n.m. W2

W2

5283-613

6 Triple

3

0.469

1.680

5284.281

I 1 Triple

n.m.

5295 -955

3 ' Triple?

2

0.562 Wi

Wl

2.004

5297 407

5 Triple

3

0.389

1.386

5298.672

4 Triple

3

0.475

1.692

5336.974

10 Triple

3

0.495

1.738

Enhanced line

5351.261

4 Triple?

2

0.487 Wi

Wl

1. 701

5369.782

5

Triple

3

0.480

1.665

5381.221

7

Triple

3

0.440

1.520

Enhanced line

5390.203

■>.

Triple

I

0.422

1.452

5397.271

4

Triple

2

0.488

1.675

X by Fiebig

5404,216

2 Triple?

I

0.735 Wl

Wi

2.517

X by Fiebig

5409.823

7 Triple

3

0.501

1. 712

5418.979

3 , 10 comps.

1,2

0.538 W3

0.346 W2

1.832

1. 178

Probably 6 H-, .t />-comps. En-
hanced line

5429.349

4 Triple

3

0.753

2.555

5474.436

4 Triple

I

0.512

1.709

Blend makes measurement dif-
ficult

5477-901

6 Triple

2

0.620

2.066

5481.652

4 ' Tnple

2

0.549

1.827

5482.078

4 1 ?

n.m.

n.m.

Numerous n- and /!-comps.
blurred

5488.374

2

Triple?

I

0.338 Wl

Wl

1. 122

5490.367

7

Sextuple?

2

0.390 W2

Ws

1.294

K-comps. fringed

5504-117

8

Triple

3

0.5IS

1.700

5512.741

12

Triple

3

0.567

1.869

5514.563

12

Triple

3

0.363

1. 193

5514.753

12

Triple

3

0.466

1-532

5565-700

8 Sextuple?

2,2

0.475 W2

0.287

1-533

0.926

5644-365

12 1 Triple

3

0.508

1-595

5648.796

4 1 Triple?

2

0.575

Wl

1.802

5662.374

10 ' Sextuple?

2

0.676 Wl

Wl

2.109

5663-155

3 Triple

2

0.462

1. 441

5675 647

8 Sextuple?

2

0.564 Wl

Wl

I.7SI

5689 . 694

6 Sextuple?

2

0.515 W2

W2

I -591

5702.876

4 ?

I

0,361 Wj

W2

I. no

«-comps. fringed and diffuse

5708.435

2 Sextuple?

2.1

0.846 W2

0.373

2.596

1. 145

Xby Fiebig

5712.098

3 Sextuple?

2,1

0 . 790 W3

0.360

2.421

1. 103

5714.120

2 Unaffected

5715.308

6 Triple

3

0.623

1.907

.....

5716.671

3 Sextuple?

2,2

0.673 W3

0.502

2.059

1.536

5720.666

3 Quintuple

I>2

0.473 (l)
0.000 (2)
0.476 (l)

0.882

I -445
0.000

1.454

2.695

Difficult

5739-698

5 1 Triple

3

0.555

1.684

5740.195

3 1 Triple

2

0.508

1.542

5762.479

I

Sextuple?

n.m. W2

W2

;j-comps. diffuse, barely sep-
arated

5766.550

I

Triple

2

0.53s

1.609

5774.250

2

Triple

2

0.569

1.707

5781.130

I

Sextuple?

n.m. W2

W2

5786.193

2

Triple

2

0-035

1.897

5804.479

2

Triple?

2

0 . G34 Wl

Wi

1.882

5823.910

2

Sextuple?

2,2

0.480 W2

0.268

1-415

0,790

5866.675

10

Triple

3

0 674

1.958

5S80 490

3

Triple

3

0.830

2.401

5899.518

7

Triple

3

0,656

1.88s

5903-555

2

Triple

2

0.876

2.513

Red comp. strongest?

5918.773

3

Triple

3

0,882

2.S18

5922.334

5

Triple

2

0.312

0.890

5938.035

2

Sextuple?

2il

0 . 840 W2

0.226

2.382

0.641

/i-comp. scarcely resolved

5941.985

5

Sextuple

2,2

Pair II, 0.905 (2)
Pair 1, 0.295 (3)

0.547

2.563
0.836

1.549

5953.386

8

Triple

3

0.637

1.798

MEASUREMENTS OF ZEEMAN EFFECT FOR TITANIUM.

Table r. — MK.\srREMF.NTS of Zeem.^n Effect for Tit.anium — Continued.

43

H

Character

s

AX

AX/X^

\

z

7

SEPARATION.

0

K-COMP.

/'-COUP.

»-COMP.

p-COiSP.

5966. OSS

Triple

3

0.590

1. 617

5978.768

7

Triple

3

0.533

1. 491

S999-920

2

Quadruple?

2

0.823 Wi

W2

2.286

6064.833

2

Triple

3

I-I59

3-151

6085.470

4

Sextuple?

2,2

I .001 W2

0.307

2.705 0.829

6091.39s

6

Triple

3

0.679

1.830

6093.030

I

Triple

2

0-7S5

2.034

6098.870

2

Triple?

2

0.563 Wi

Wi

1.514

6121.21S

I

Triple

2

0.622

1.660

6126.43s

s

Septuple?

2

0 . 706 Wi

W3

1. 881

»-comps. fringed. Probably 3
p-comps.

6146.445

I

Triple

2

0.476

1.260

6149.950

I

Triple

2

0.698

1-845

6215.630

3

Triple

3

0.697

1.804

6220.700

2

Triple

2

0.618

I -597

6221.552

I

Triple

2

0.511

1-320 ! 1

6258.322

10

Triple

2

0.661

1.688)

1 Red «-comp. of first line blended |

6258.927

12

Triple

2

0 714

1.823)

1

with violet comp. of second

6261 .316

10

Triple

3

0.556

I 418

6303 985

S

Octuple?

2,2

0.565 W3

O.SS4

1.422

1-394

»-comps. very wide and diffuse.
Probably at least 3 pairs

6312.456

5

Octuple?

2,2

0.766 W3

0.463 Wi

1-923

1. 162

M-comps. wide, not so diffuse as
6304 . Probably 3 pairs. Pos-
sibly 4 /)-comps.

6318.239

2

Triple

2

0.529

I -335

6336.329

4

Triple

2

0.606

1.510

6366 . 564

4

Triple?

2

0.671 Wi

Wi

1-655

1

6419.329

I

Sextuple?

I

0.758 W2

Wl

1.840

\ by Fiebig

6491 .800

5

Triple

2

0.752

1.78s

Enhanced line Ti? Not given by
Fiebig for arc

6497.840

2

Triple

2

I - 2og

2.863

Titanium?

6508.380

2

?

I

I. 318 W3

W3

3-"2

n-comps. very wide and diffuse

6513-300

2

Triple

2

0.610

1-438

Enhanced line Ti? Not given by
Fiebig for arc

6546.479

8

Triple

2

0.459

1. 071

6554.470

9

Triple

3

0.79S

1.851

6556.308

10

Triple

3

o.8g6

2.08s

6559-815

3

Triple

2

0.594

1.380

Ti? Not given by Fiebig

6565-783

2

Se.xtuple?

I

I . 000 Wj

Wj

2.320

Ti? Not given by Fiebig

6575-437

2

Triple?

I

0.644 Wi

1.489

Xby Fiebig

6599 -353

S

Triple

3

0.7II

1-633

6606 . 160

2

Quadruple?

2

0.881 Wi

Wl

2.019

Ti? Not given by Fiebig

6666.714

2

Triple?

I

0.753 Wl

Wl

1.694

6667.998

2

Triple?

n.m. Wi

Wl

Xby Fiebig. Blended with flut-

6716.922

2

Triple

I

0.772

1. 711

ing lines

6717.964

2

Triple

I

0.862

1.910J

6743.381

6

Triple

2

0.759

1 .669

TYPES OF SEPARATION.

The number of lines for each type of separation, including both the clear and the doubtful cases, is
given in Table 3. For the quadruplets, sextuplets, and septuplets, the questioned lines greatly out-
number the clear cases. For example, iron shows only two clear septuplets and titanium two clear quad-
ruplets. A strong field will probably show that these doubtful lines have usually been correctly classi-
fied as to number of components, but actual measurements for the unresolved components are at present
lacking.

Table 3. — Summary of Types of Separation.

Separation.

Iron.

TiTANItrai.

XTnafiFected . ....

9

393

49

.7

118

37

6

9

7
2

4

2

19

4

291

28

S

77

12

II

3

7

2

2

0

16

Triple

Quadruple

Quintuple

Se.xtuple

Septuple .... ....

Octuple

9 components

10 components

12 components

13 components

Unclassified

Total

662

458

I. Unaffected Lines.

A number of fines in each spectrum show no tendency toward separation or even widening by a
magnetic field as high as 20,000 gausses. The fight giving such fines is unpolarized so that a single
sharp line appears in the magnetic field spectrum, whatever the optical system may be.
ber of these fines is not large, the undoubted cases being as herewith:

The num-

Iron.

Titanium.

X3746.058
3767-341
3773-803
3786.820

X38S0.118
5123-899
5434-740
5576.320

X429S.914
5222.849
5714.120

2. Triplets.

The number of triplets is larger than that of any other one type, the number of clear cases, i.e., those
whose components show no widening which would indicate that they are compound, being 297 for iron
and 247 for titanium. The relation of the separation of these to the "normal interval" will be treated
in another part of this paper.

A rather curious mistake has found its way into the literature based on some lines in the iron spec-
trum. Becquerel and Deslandres in their first pubfication (g) gave X3865.674 as an "inverted triplet,"
having but a single ^-component and two /(-components. This evidently arose from under-e.xposure of their
photographs for the w-component, as in their next paper (10) they gave the correct character of this fine,
44

TYPES OF SEPARATION. 45

it being a quintuplet with three n- and two /)-coniponents, the central «-component being strongest.
Reese (12) made the same error concerning both this line and X 3643.469, together with another farther
to the vdolet. Kent (13) followed with a publication in which the lines are correctly described. Cotton (i-O
calls attention to the confusion which has come about and gives the correct structure. Runge (2/') cites
the first paper of Becquerel and Deslandres and that of Reese concerning the inverted triplet, without
noting that the error had been corrected in each case by later publications, though Runge later (2<r) repro-
duced the diagram of Becquerel and Deslandres from their second paper, in which the correct structure
of these lines is given. Other works on spectroscopy speak of the inverted triplet, the basis for this
being the publications which have been mentioned. No real case of the inverted triplet has presented
itself in the iron or titanium spectrum, nor, so far as the author is aware, does such a type exist in any other
spectrum. Apparent examples of such inversion are hkely to appear on plates not fully exposed, since
some quintuplets, X4464.6i7 of titanium for example, show a central M-component very much stronger
than the two outer ones, so that the central component may easily appear alone.

The tendency of Zeeman components to follow the appearance of the no-field line as regards sharp-
ness or diffuseness frequently makes it difficult to judge whether a line is a true triplet or not. If the
lack of sharpness is not due to the character of the no-field line, a doubtful triplet may be (i) a quadruplet
if the difi'use /^-component is really double, (2) a sextuplet if each of the three widened n- and ^-com-
ponents are double, (3) a septuplet if the two M-components are double and the /^-component has three
close constituents. Still higher separations for doubtful triplets are not impossible, but probably there
are very few such. The criterion for distinguishing between these possible types is given on p. 19.

3. Quadruplets.

The unquestioned quadruplet is somewhat rare. The great majority of lines having two n- and two
/(-components have their w-components widened to some extent and are usually classed as doubtful
sextuplets, since so many of these have been resolved by the strongest fields used that it seems probable
that a still stronger field would show four ^-components for such lines in every case. Occasionally, how-
ever, the two M-components are sharp. The relative separation of n- and /^-components varies greatly
for different lines, but the /^-components are almost always closer together. The most decided excep-
tion is the titanium line X 4398.460, apparently a quadruplet, whose ^-components show only two-thirds
the separation of the p pair.

4. Quintuplets.

The quintuplet appears least often of any of the less complex types. As a rule this separation gives
three n- and two /J-components, the distance between the /^-components being the same as between the
outer w-components. The central ^-component is the strongest of the three, and the effect when the light
is observed at right angles to the lines of force without a Nicol prism is to give a triplet, the components
of which are of about equal intensity, caused by the superposition of the p doublet on the two outer
M-components. Good examples are XX3733.469 and 3865.674 of iron, and 4291. 114 of titanium. The
first two were originally mistaken for "inverted triplets" on account of the strong central M-component.
Dissymmetry is sometimes present, as in X 5455.834 of iron. A different type is presented in X 4710.368
of titanium, which shows four w-components and a single sharp />-component. No similar line has been
observed in either of these spectra.

5. Sextuplets.

This t>pe usually has the two pairs of «-components of equal intensity, shown by a uniform widening
in cases where the pairs are blended. As has been previously noted, the sextuplet is a very common

46 INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

type, a great number of lines being classed as probable sextuplets which show as diffuse triplets with this
field. A field of considerably greater intensity will probably show these to be similar to most of the sex-
tuplets which are fully resolved here. The /"-components of those sextuplets which have been measured
are usually rather narrowly separated, while the two pairs of w-components are frequently almost blended.

6. Septuplets.

The prevailing type of septuplet has four w- and three /^-components, the two pairs of w-components
not usually being of the same intensity. When blended, the w-components give the "fringed" appear-
ance often noted in the "Remarks" column, in which case the weaker pair may be either inside or
outside. When the /^-components are not resolved, it is often difficult to distinguish this type from the
sextuplet, the difference depending on the existence of a central maximum in the widened />-component.

7. OCTUPLETS.

The typical octuplet has five n- and three /)-components, equally spaced. The outer w-components
are usually the stronger and the central one quite weak, so that when the three /^-components, if the
central one is the stronger, are superposed, as when the light is viewed across the lines of force without a
Nicol, the effect is to show five components of about equal intensity. Examples of such lines are
XX 3743.508, 3788.046, 5497.735, of iron, and 4281.530, 4527.490, 4544.864, of titanium. The last two
were given as septuplets in my former paper (51) on account of the weakness of the central «-com-
ponent. Another arrangement is presented by the titanium line X 4308.081 which has three pairs of
M-components and two /j-components.

8. Nonets.

Good examples of lines having nine components are found in XX 3840.580, 4233.772 of iron, and 447 1 .408
4489.262, 4629.521 of titanium. These have each three pairs of H-components, the innermost pair being
strongest, and three /^-components. The type is probably rather common in both spectra, since many
fines classed as doubtful septuplets may have a weak outer pair of w-components, making a total of nine.

9. More Complex Types.

Lines having ten components are represented by XX 44 17. 884, 4471.017, and 5025.027 of titanium.
These are made up in each case of three pairs of n- and two pairs of /^-components. Eleven components
are shown by X 3888.671 of iron, which has a central w-component in adcUtion to the pairs of the ten-
component type. Several good examples of twelve-component lines are given by XX 3722.729, 3872.639,
5447.130 of iron and 4289.237 of titanium. These are all of similar structure, having four pairs of
w-components, the two inner pairs having the same separation as the two pairs of /^-components. While
twelve is the highest number of components which is measurable on my plates, the iron lines XX 4005.408
and 4132.235 are given as probably having thirteen components each. Five /^-components are almost
resolved in each case and the wide inner fringes for the w-components are estimated to consist of four
pairs. Many of the lines whose type is questioned without attempt to estimate the number of compo-
nents have probably as many as the most complex of those measured, and some of them possibly more.

Good examples of almost all of these types of separation are present among the violet iron fines shown
in Plate III, which has the advantage of showing the w- and /^-components both separate and in combi-
nation, the latter spectrum being taken at right angles to the force-fines without the use of a Nicol prism.
Polarization by the grating reduced the intensity of the /(-component for this region of the spectrum,
as is shown by the relative weakness of the central component of triplets in the spectra lettered b, for
which the Nicol prism was not used.

RELATION OF SEPARATIONS TO THE NORMAL INTERVAL.

I. Summaries for Various Types.

The study of how generally the separations observed show a simple relation to the fundamental
interval, the theory of which was summarized on p. 4, has been gone into in some detail. The relation

H

a = -

gives a value for a of 0.753 ^or H= 16,000, and of 0.812 for H= 17,500, if e/m be taken equal to 1.75 X 10^.
The "normal triplets" for iron and titanium, with the standard field-strengths used in this work, should
accordingly show values of AX/X- for the distance between the side components of about 1.500 and 1.600
respectively.

In the following summaries an attempt has been made to show to what extent the separations for
various classes of lines may be considered as multiples of the interval a. In Table 4 the clear triplets
for iron and titanium are thus classified, those triplets given in Tables 1 and 2 as doubtful not being
included. The allowable deviation for any line from the exact multiple was estimated as closely as pos-
sible according to the weight of the measurement, knowing the probable error for each weight. Lines
not falling into any class are placed in the "Odd" column. In the case of titanium a large proportion
of such lines appeared to be definite odd multiples of a/4, while the regular classes consider only multiples
of a/2. As in all of the following work relating to the interval a, greater field strength is desirable, as the
accuracy of the classification increases with the numerical value of a; but Table 4 shows in a general way
how the magnitudes of the separations may be grouped.

Table 4. — Separation op Triplets as Related to the Normal Interval a.

a

30/2

2a

S0/2

3a

70/2

40

Sa

60

Odd

Remarks.

Element, Iron:

Wt. ^

0
2
I

S
II

s

25
19
10

17

27
II

35
27
10

6
2
3

4
2

I

3
0
0

4
0
0

16

20

6

"Odd" includes 70/4, 14 lines;

ga/4, 32 lines
"Odd" includes 70/4, 7 lines;

90/4, 8 lines

Wt. 2

Wt. I . . .

Total

Element, Titanium:
Wt. ?

3

21

54

S5

72

II

7

3

4

42

0

S
2

5
S
I

40

27

6

IS

13

2

14
2

I

I
2
0

I
I
0

2
0
0

0
0
0

76

21

0

Wt. 2

Wt. I

Total

7

II

73

30

17

3

2

2

0

97

The relation of the separation to the normal interval was also studied for those fines which appear
on my plates as quadruplets with components in many cases diffuse, indicating a compound structure.
The two ^-components are usually fairly sharp, but the w-components are often formed of two or more
pairs blended. Close agreement with exact multiples of the normal interval can not be expected for
fines of this class, but in the majority of cases the distance between the components of the n and p pairs
could be expressed as multiples of a or a/2 closely enough to show a real relation; 66 fines of iron and

47

48

INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

49 of titanium were thus treated, and 80 per cent of the former and 75 per cent of the latter were found
to show separations related to a in this way. Table 5 gives the number of lines corresponding to each
ratio of separation for n- and /"-components when these could be expressed in terms of a or a/2.

Table

5. — Separations

OF apparent Quadruplets

in terms of Normal Interval.

Separation.

No. OP Lines.

Separation.

No. of Lines.

Ratio.

Ratio.

B-COMP.

p-COUP.

Iron.

Titanium.

n-coMP.

p-COUP.

Iron.

Titanium.

2a

2a

1:1

2

3

31/2

a

3:2

3

. .

51/2

S1/2

1:1

I

70/2

a

7:2

I

2a

a

2:1

8

7

a

30/2

2:3

I

6a

30

2:1

2

2a

30/2

4:3

3

31

30/2

2:1

4

2

Sa/2

31/2

5:3

I

30

a

3:1

18

6

70/2

30/2

7:3

I

90/2

30/2

3:1

I

S0/2

2a

5:4

I

50

a

S:i

I

n.m.

31/2

I

3

3a

2a

3:2

S

I

n.m.

20

4

6

30/a

a

3:2

I

n.m.

30

...

I

Fourteen lines of iron and 13 of titanium gave separations not to be expressed in terms of a/2, but most
of these showed a simple ratio between the n and p separations. For iron, 2 hnes showed a ratio of 1:1,
6 lines of 2:1, 2 hnes of 3 :i, 2 lines of 4:1. The odd hnes of titanium had among them 5 of ratio 2:1,
and I of ratio 4:1.

The more complex hnes, so far as they were resolved by the field employed, have been arranged in
the following summary. In each case, below the wave-length, the first column gives the interval in
terms of a from either the red or the violet component to the center of separation. (The triplets and
quadruplets in Tables 4 and 5 have their separations given as the total distance from the red to the violet
component.) The second column shows whether it is the n- or /^-component which has this separation.
If both letters are present for a given interval, there is a superposition of components having the two
polarizations. The last column gives the ratio of the successive intervals.

Iron:

>>37i8.SS4

±50/4 n 2

Sa/8 p I

Quintuplets.

f^?' 3733 469 3865.674"]

3760.679 5455.834 have the arrangement
(_ 3814.671 5603. i86j

f ±30/2 n,p
\ o »

Titanium :

Iron: X 3774.971

±50/2 n 10

5^/4 « 5

a p 4

Titanium: X 3982. 142

±33/2 « 3
a p 2
a/2 n I

X 4302.085

±2a « 8

30/4 p 3

a/2 n 2

X 4291 . 114

±30/2 n,p -
o n —

X 4109.953

±33/2 K

3^/4 >',P

X 4422.741

±30/2 n 3
a p 2
a/2 n I

X 4464.617
±70/4 n 2

^a/% p I

o no

Sextuplets.
X 4447.892
±2a » 4
33/2 n 3
a p 2

X 4710.368

±40/3 n 8

a/2 n 3

o p o

X 5720.666

±30/2 n,p -

X 4872.332
±30 n 2
30/2 n,p I

X 6213.644

±30 K 6

3«/2 " 3

a p 2

X4321.119
±2ia/8 n 3
70/8 n,p I

X 4640. 119

±53/2 n 5

30/2 n 3

a pi

X 4722.797

±30/2 » 12

90/8 p 9

5<'/8 « 5

X 4798.169

±30/2 II 3

a p 2

a/2 It I

X 4805 . 285

±30/2 n 3

a n 2

a/2 p I

X 6337.048

±33 n 6

33/2 » 3

a p 2

X 5941.985

±30/2 n 3

a p 2

a/2 n I

Septuplets.

Iron:

X 4009 . 864
±23 « 8

a n 4
3a/4 P 3

o p o

X 4191

±23
3
O

595
n 2
n,p I
p o

X 4352.908

±(23 « 4)?

33/2 n 3

3/2 p I

o p o

X 5079.921

±28 K 2

3 n?,p I

Op O

Titanium:

X 4298.828
±3 « 2
3/2 n,p I
o p o

RELATION OF SEPARATIONS TO THE NORMAL INTERVAL.

49

OCTUPLETS.

Iron: X 3743.308 X 3788.046 X 4859.928

Titanium :

±20 «
a n,p
o n,p

X4281.S30
±30 n 2
30/2 n,p I
o n,p o

± 2a « 2
a n,p I
o n,p o

±(30 n 2)?
30/2 n,p I
o «,/> o

^5497-735
±Sa n 2
3a/2 n,p I
o «,/> o

X 4308.081
±2a » 8

30/2 n 6
a n 4

3a/4 P 3

X 4527.490

±23 ?I 2
a H,/) I
O H,/! O

X 4544.864

±2a n 2

a n,p I

O K,/! o

X 4590.126
±33/2 » 12

o n 8
70/8 p 7

a/2 » 4

X 3748.408

±33/2 K 3

a n 2
a/2 n,p?i
o p o

X 3840.580

±3"/^ " 3
a « 2
a/2 ti,p I
o p o

^ 4233-772

±3" « 3

2a » 2

o «,/> I

o /> o

Nonets.

X 5405 . 989

±(30/2 n 3)?
a n 2
a/2 n,p I
o p o

Titanium:

Titanium:

Ten-Component Lines.
X 4417.884 X 4471.017

; ? « ?

a n 8
31/4 /> 6
3a/8 n 3

a/4 p 2

±93/4 « 6

30 A/" 4
90/8 )»,/>? 3
30/8 « I

X 4471 .408

±2ia/8 n 7

150/8 K 5

9a/8 » 3

30/4 p 2

o p o

X 5025.027
±33/2 « 24(5)
53/4 p 20(4)

3 » 16(3)

53/8 P 10(2)
53/16 M S(l)

X 4489.262

±2ia/8 n 7

153/8 » 5

93/8 n 3

33/4 ^ 2

o /> o

X 4629.521

±53/2 n 5

33/2 » 3

3 ^2

0/2 « I

o p o

The numbers in parentheses for X 5025.027 give a simpler relation between the intervals than the exact
ratio of the multiples of parts of a. Another probable ten-component line isX3982.630, for which the
measurements are poor. Its n-components are in the ratio 5:3:1.

Iron: X 3888.671

±3<i/2 n 3

a n,p 2

a/ 2 ti,p? I

o « o

Eleven-Component Lines.
X4871.S12
±(33/2 n 3)?
a n,p 2
a/2 n?,p I
o p o

Titanium : X 3930 .022

±93/4 n,p 3

33/2 n,p? 2

33/4 H I

O H O

X 487 1.5 1 2 has its M-components blended, but the structure indicates the above arrangement.
The titanium line X 392 1.563 has probably the same structure as X 3930.02 2. The «-components
have the ratio 3:2:1:0, but the measurements are not good enough to be sure of the relation to a.

Twelve-Component Lines.

Iron:

X3722.729

±2a » 4

33/2 « 3

a n,p 2

a/2 n,p?i

X 3872.639
±23 n 4
33/2 n 3
a n,p 2

0/2 n,p}i

X 5447.130
±23 n 4

33/2 K 3

a n,p 2

3/2 n,p I

Titanium:

X 4289.237

±23 n 4

33/2 n 3

a n,p 2

a/2 n,p I

2. Discussion of Relations to Normal Interval.

It is shown in Table 4 that for iron two-thirds and for titanium over one-half of the clear triplets
are separated by the intervals 2a, 5^/2 and 3a. For both elements, however, a very large majority have
separations of this order of magnitude, since almost all of the hues classified as "odd" give intervals
within this range, the numbers corresponding to 70/4 and ga/4 of weights i and 2 being given for titanium
in the "Remarks" column. A more precise classification, in which smaller fractional parts of a can be
used, must await an investigation with greater field-strength, which will also decide the structure of
most of the doubtful triplets, the separation of which is not included in any of these summaries.

Table 5 shows how generally the separations of those lines showing two n- and two /(-components
can be expressed in terms of the interval a, also the wide variety of separations wliich prevails. The ratios
of 2 : 1 and 3 : i predominate for both elements. As has been previously noted, the /^-components almost
always show the narrower separation.
4

50 INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

The ease with which the separations of the complex lines both in iron and titanium can be expressed
in terms of a affords a confirmation of Runge's law, since failure to give approximation to exact multiples
of a appears to occur only in the case of measurements of small weight. It has been necessary only in
a very few cases to use multiples of any quantity smaller than a/4, so that errors of measurement are
seldom large enough to influence the ratios found. This question will become of more importance when
very close components are resolved by a stronger field.

The presence of "magnetic duphcates," fines exactly similar in structure, with the same intervals
between components, furnishes a means of selecting fines which may be connected by series relations.
Such dupficates occur for almost every type of separation. Six quintuplets of iron and two of titanium
show the same structure and intervals. These are XX 3733.469, 3760.679, 3814.671, 3865.674, 5455.834,
5603.186 of iron and 4291. 114, 5720.666 of titanium. Several types of sextuplets appear. The red fines
of ironXX62i3.644 and 6337.048 are dupficates, also the titanium fines XX3982.142, 4798.169, and 5491.985.
Dupficate septuplets of iron are XX4191.585 and 5079.921. The only titanium septuplet fully resolved,
X 4298.828, has the same structure. The four iron octuplets are of the same appearance but have dif-
ferent spacing, XX3743.508 and 3788.046 being afike, as are probably also XX4859.928 and 5497-735i
though the former was not fully measurable. The blue octuplets of titanium XX 45 2 7. 490 and 4544.864
are also dupficates. The iron nine-component fines XX3748.408 and 3840.580 are alike, and X 5405.989
has probably the same intervals. Another spacing is shown by the titanium dupficates XX 447 1.408 and
4489.262. The fines of iron which probably have ten components are not fuUy resolved, while the three
titanium fines show diverse arrangements. Perhaps the finest examples of spacing in multiples of a are
the twelve-component fines XX3722.729, 3872.639, 5447.130 of iron, which are exact dupficates, while
X4289.237 of titanium is in all respects similar.

POSSIBLE RELATIONS BETWEEN LINES AS INDICATED BY THE ZEEMAN EFFECT.

It is hoped that the measurements presented in this paper, especially the summary of complex sepa-
rations given on pp. 48 and 49, may eventuaUy aid in finding definite relations among the fines of these
spectra. At present, nothing conclusive along this fine is to be offered. Numerous cases of magnetic
dupficates have been shown to exist in both spectra. Such fines, especially if they are in the same part
of the spectrum, are often affected in the same way as to change of intensity in various fight sources
and show a similar magnitude of displacement by pressure. The same vibrating particle probably pro-
duces them.

The differences in wave-number (i/X) have been formed for the various pairs of magnetic dupficates.
Only one case was found where two pairs of magnetic dupficates have the same difference of wave-number.
The iron octuplets XX3743. 508 and 3788.046 have exactly the same difference in wave-number (314) as
the sextuplets XX 6313.644 and 6337.048. No case was found where two pairs of magnetic dupficates
of the same t^-pe have the same difference, though this was tried wherever promising, both between known
dupficates and as a means of finding new pairs. The differences between duplicates were found to vary
greatly for each element and to bear no simple relation to one another; so that as yet no clue has been
found which will serv'e in building up series relations.

CASES OF DISSYMMETRY.

Iron.

Titanium.

^3718.554

X 3998. 790

3760.679

4009.807

3892.069

4298.828

3952-754

4645 368

4878.407

4997-283

5324373

5219-875

S4SS-834

S9°3-555

There are but few striking examples of dissymmetry in the iron and titanium spectra, either in spacing
of the components or in the intensities of the violet and red components. However, fourteen lines show-
ing distinct dissymmetry may be listed as shown herewith:

The nature of the dissymmetry is covered in each case in the
"Remarks" column. Several triplets show either the red or the violet
component decidedly stronger. Quintuplets are hkely to show irregular
spacing or intensity, or both, as in the cases of XX3718.554, 3760.679 and
5455.834, of iron. The last hne has its central w-component moved dis-
tinctly to the red from the position of the no-field line (see Plate IV).
The titanium septuplet X 4298.828 shows three ^-components, the interval
between the central and violet components being about two-thirds that

between the central and red. This line appears on Plate V. Several of the other lines are of complex
type and highly unsymmetrical.

The plates taken in this investigation are for the most part not suitable for the detection of a differ-
ence in the spacing from the central line of the violet and red component of triplets, since a Nicol was
almost always used to separate the n- and />-components. However, two of the best plates in the set
were taken without a Nicol for the iron spectrum in the blue and violet regions and include most of the
lines mentioned by Zeeman (30) as showing a difference in the intensity or in the spacing of the violet and
red components. These plates were taken with a field-strength of 19,500 gausses. A set of measure-
ments was made for the sharpest triplets occurring in this region to test the question of a difi'erence in
the spacing of the violet and red components from the central line. The method was to make settings
successively on the violet, central, and red components, and then repeat in the inverse direction, con-
tinuing until four sets of readings were obtained from which the mean distance to each side component
was computed. The measurements given in Table 6 are the mean of two independent sets taken in this
way, which in general agreed closely. Thus each value of AX is the mean of eight determinations of the
interval in question. The values of AX are not reduced to the standard field. Differences in favor of the
violet interval are -f , those in favor of the red interval — .

Table 6. — Spacing of Violet and Red Components of Iron Triplets from the Central Component.

X

AX

Differenck.

X

AX

Difference.

Center to

Center to

Center to

Center to

Violet.

Red.

Violet.

Red.

3687.610

0.204

0.198

-f-0.006

3920.410

0.231

0.218

+0.013

3709-389

0. 200

0.201

— O.OOI

3923-054

0.228

0.223

+ 0.005

3758.375

0.179

0.165

-ho. 014

3928.075

0.227

0.223

+ 0.004

3763-945

0.148

0.136

+0.012

3930.450

0.230

0.22I

+ 0 . 009

3765.689

0-153

0.142

+ O.OII

3997-547

0.173

0.168

+ 0.005

3798-655

0.212

0.209

+0.003

4063 . 759

0.179

0.177

+ 0.002

3799-693

0.213

0.211

+0.002

4236.112

0.285

O.2S0

+ 0.005

3827-980

0-IS3

0-139

+0.014

4260.640

0.278

0. 272

+ 0 . 006

3856-524

0.222

0.213

+0.009

4271-934

0.218

0. 216

+ 0.002

3860.055

0.217

0. 219

—0.002

4308.081

0.200

0.194

+ 0 . 006

3886.434

0. 225

0.214

+ O.OII

4325-939

0.170

C.161

+ o.oog

3895-803

0.223

0.223

0.000

4383.720

0.217

0. 210

+0.007

3899.850

0.22s

0.221

+0.004

4404.927

0.212

0.208

+0.004

51

5 2 INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

These measurements are intended only as a preliminary test of the reality of the difference in triplet
spacings. The evidence, however, points strongly to the existence of a true difference for many, if not
all triplets. Only 3 out of 26 lines fail to show a larger interval for the violet component. Although the
settings on a component seldom show a range greater than 0.004 A, which would indicate a very small
probable error in the mean of 8 determinations, it is likely that the actual probable error of the indi-
vidual diiTerences shown in Table 6 may amount to 0.003 or 0-004 A as a result of systematic errors in
the settings due to the character of the lines. The mean of all the differences is + 0.006 A, with a calcu-
lated probable error of ± o.ooi A, which can scarcely leave any doubt as to the reality of the difference.

The measurements show that the magnitude of the difference can hardly be the same for all of the
lines. The true probable error will then be somewhat smaller than that given above, which would only
make the e\'idence for the reahty of the dissymmetry predicted by Voigt the stronger. The Unes from
X 3930.450 toward the violet, 17 in number, are with one exception either normal triplets or have the separ-
ation 3a, usually the latter. Of the 9 lines showing a difference greater than 0.008 A, 3 are normal trip-
lets and 4 have a separation of 2,(i- The question of dissymmetry seems worthy of investigation through
a long range of field-strengths for these hues, especially to test the generality of the change of spacing
with the square of the field-strength observed for one of the Hues in the mercury spectrum (see p. 5).

An element which might sometimes affect the spacing of Zeeman components is the apparent differ-
ence in the wave-lengths of arc and spark lines. The spark is made more disruptive by the magnetic
field, and a greater disruptiveness seems in general to cause the lines of the spark to be moved slightly
toward the red as compared with their positions in the arc spectrum. The reality of this effect is still
a disputed question, but evidence published by a number of observers, as well as some photographs of
the arc and spark which I have taken for this portion of the iron spectrum, indicate that measurements
taken in the regular way will give a sHghtly greater wave-length for the spark lines, the difference being
greatest for a very disruptive spark. If this effect has a part in the Zeeman phenomenon, we should
expect all components of the triplet to be displaced alike. The greater strength of the middle component,
however, would probably make the effect more perceptible for this, as the apparent displacement is more
or less combined with unsymmetrical widening and is usually more distinct for strong lines. However,
in the photographs from which the measurements of Table 6 were taken, triplets to the violet of X 4000
show the middle component only about as strong as either side component on account of the polariza-
tion given by the angle of the grating used, so that the conditions of the spark discharge would not seem
to be adequate to explain the difference in spacing, unless the direction of vibration of the electrons,
parallel or perpendicular to the hues of force, affects their susceptibility to the displacing action of the
spark discharge. On this point we have no evidence.

The other point of dissymmetry predicted by Voigt, a greater strength for the red component of the
triplet, is quite perceptible for many lines, especially in the iron spectrum. The difference is rarely greater
than 10 per cent., and, to be clearly detected, the two components must be distinct but not of full density,
since blackness of the components in the negative destroys so shght a difference. On account of this
necessity for just the right degree of exposure, it is difficult to say how general the phenomenon is, but it
is certainly present for many lines.

LAW OF CHANGE OF THE AVERAGE SEPARATION OF THE iV-COMPONENTS

WITH THE WAVE-LENGTH.

A glance through Tables i and 2 shows that for both iron and titanium the tendency is for the values
of AX gradually to increase as we pass to greater wave-lengths, while the values of AX/X^ remain of about
the same magnitude throughout. A statistical study of this apparent constancy of the averages AX/X'
has been made; and both the range of wave-length and the number of lines available are sufficient to
show clearly how the matter stands.

The method of treatment has been to obtain the mean value of AX/X^ for the M-components for each
500 A from X3700 to X6700. When there are two or more pairs of «-components the mean of the separa-
tions is taken. This is necessary for the sake of consistency if any Hues other than clear triplets or quad-
ruplets are to be considered, since the measurement of the widened «-components given by a great many
lines is merely the mean separation of two or more unresolved pairs.

The averages thus obtained are presented in Table 7. The means for the six groups of 500 A are
given first, then the means for the three groups of 1000 A. These latter are the means for the whole
number of Unes considered in the range, not the averages of the means for the 500-groups. Of course,
no account can be taken in this summary of the considerable number of Hnes which are described, but
whose w-components are not measurable.

Table 7. — Means of AX/X' («-components) for Successive Regions of Wave-length.

Iron.

Titanium.

No. of Lines.

Mean AX/X^.

No. of Lines.

Mean AX/X'.

3700-4200

267

2.003

80

1.909

4200-4700

lOI

2.051

152

2.027

4700-5200

74

2.125

81

1.684

5200-5700

62

I 932

47

1. 819

5700-6200

37

1-837

34

1.942

6200-6700

41

2-131

28

1.764

3700-4700

368

2.016

232

1.986

4700-5700

136

2.037

128

1-734

5700-6700

78

1.989

62

1.862

The close agreement of the means shows that there is a real relation, giving an approximate constancy
of the values of AX/X- for different parts of the spectrum. Taking the successive means of the 500-groups,
the average value for iron is 2.013, for titanium 1.858. The largest deviation from the mean for any
group is 8.7 per cent for iron and 9.4 per cent for titanium. For neither element is there any systematic
change in the means for successive groups.

The means for the groups of 1000 A show a still closer agreement, the largest deviation from the mean
of these groups being only 1.2 per cent for iron and 6.8 per cent for titanium.

It will be noticed that the mean values for titanium run smaller than those for iron, although the
titanium measurements correspond to the larger field-strength. A number of spectra will have to be
examined in this way and the measurements reduced to the same field-strength before we can say what
significance, if any, there is in this point. It may prove to be connected with certain properties of the
elements concerned.

It is not difficult to see that this constancy of the mean value of AX/X- depends on the general relation
of this quotient to the fundamental interval a, and that it results from the fact that the great majority

53

54

INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM

of the separations for the «-components range from the values of 2a to :^a and that the various values
of the multiples of the interval are more or less uniformly distributed throughout the spectrum. This
was shown for the triplets (p. 47), the greater number of which show a separation greater than 2a. The
exceptional large and small values for triplets, together with the mean separations of the complex lines,
combine to form a fairly definite mean which holds for the whole range of spectrum examined.

Since AK/\- is sJwun to be very nearly constant, it may be said that for iJie spectra of iron and titanium, and
probably for spectra in general, the mean separation oj the n-components varies as the square of the wave-length.

A similar rule must hold for the /^-components, since it was shown (pp. 48-49) that complex lines of
the same structure in different parts of the spectrum show the same relation to the interval a.

It is of interest to note that a computation along the hues of that carried out here, but different in
method and with comparatively Httle material at disposal, was made by Mr. Hale (38) in his comparison
of sun-spot doublets with the Zeeman separations on some prehminary plates made by the author. The
mean AX for a number of iron lines in the blue was divided by the square of the mean wave-length for
the region considered. Measurements for lines extending from the green into the red were treated simi-
larly. The quotients of the mean AX by the square of the mean X for the two regions agreed exactly.
While this result does not have the same significance as the comparison of the mean values of AX/X-, it
is clearly based on the same relation for the rate of increase of AX with X.

THE EFFECT OF THE MAGNETIC FIELD UPON ENHANCED LINES.

In my former paper (51) on the titanium spectrum, the behavior of the enhanced hnes was examined
to see if, as a class, they were affected by the magnetic field differently from the non-enhanced lines.
The various types of separation were found to occur in about the same proportion for the enhanced hnes
as for the spectrum in general. The same conclusion was arrived at by Mr. Babcock (62) for the enhanced
lines of chromium and of vanadium.

Table 8 gives the numbers of enhanced and non-enhanced lines considered both as to type and magni-
tude of separation. Here, as in Table 3, a given type includes both the clear and the questioned cases for
that type occurring in Tables i and 2.

Table 8. — Comparison or Types of Separation for Enhanced
AND Non-Enhanced Lines.

Character of
Separation.

Iron.

Titanium.

Enhanced.

Non-
Enhanced.

Enhanced.

Non-
Enhanced.

Unaffected

Triple

Quadruple

Quintuple

Sextuple

Septuple

0
25
4
0
8
3
3

9

368

45

7

no

34
46

0

49

5

I

13

I

13

4
242

23

4

64
II
28

Total

43

619

82

376

The enhanced hnes of each element are found to present a diversity of types. The enhanced and
non-enhanced triplets are in about the same ratio as the total number of enhanced and non-enhanced
lines, both for iron and titanium, this ratio being about 1:14 for iron and about 1:5 for titanium. Those
types for which the number is suflicient to give the comparison some weight are in the same ratios. There
seems to be no undue proportion of any one type among the enhanced lines, considered as a whole.

EFFECT OF THE MAGNETIC FIELD UPON ENHANCED LINES.

55

Since the triplets appear to be representative, and as their magnitudes of separation can be handled
most readily, Table 9 is arranged to compare the values of AX/X- for enhanced and non-enhanced triplets.
Triplets whose separation was not measurable are omitted, as are some non-enhanced triplets of very-
large separation, larger than is shown by any enhanced lines.

T.VBLE 9. — Values of A\/\- for Enhanced and Non-Enhanced Triplets.

Range of AX/X'.

Iron.

Titanium.

Enhanced.

Non-
Enhanced.

Enhanced.

Non-
Enhanced.

o-i .0
I. 0-1.4
I. 4-1. 8
1.8-2.2
2.2-2.6

I
3
7
S
3

2
40
94
93
84

0

6

26

II

4

9

30
99
66
26

On account of the small number of enhanced lines of iron, Table 9 serves to bring out little more than
the distribution of the values of AX/X-. More enhanced Hnes are available for titanium, and in the study
of these, two points are noteworthy: the absence of very small separations, and the disproportionately
large number of enhanced triplets giving values from 1.4 to 1.8. This range includes the normal triplet
at about 1.6, and the table shows that the separations of over half of the lines in question are close to
this value. This is due in part to a condition which appears to be the only respect in which the enhanced
lines are in a class by themselves as regards the Zeeman phenomenon. In the region from 3600 to 4600,
which is rich in enhanced lines for titanium, the strongest enhanced lines were selected, 22 in number.
These are lines showing a high degree of enhancement in the spark and are as a rule much stronger in
the spark than any of the lines characteristic of the arc. A short exposure with a strongly condensed
spark would show these lines almost alone. Of these 22 Hnes 17 are clear triplets; the remaining 5, with
one e.xception, the weakest in the list, are of more complex character, These Hnes, with their intensity on
the scale here used, their type of separation, and the values of AX/X- for the triplets, are given in Table 10.

Table 10.-

-Effect of the Magnetic Field upon the Stronger Enhanced
Lines of Titanium.

X

Intensity.

Separation.

AX/X^

X

Intensity.

Separation.

AX/X^

3685.339

20

Triple

1.708

4302.085

5

Sextuple

3741-791

10

Triple

1.878

4308.081

8

Octuple

3759-447

20

Triple

2.038

4313.034

8

Sextuple

3761.464

10

Triple

1-463

4338.084

10

Triple

1. 312

3900.681

5°

Triple

1.787

4395. 201

20

Triple

1.796

3913.609

20

Triple

1-43°

4443.976

IS

Triple

1-509

4103.818

20

Triple

1.696

4468.663

IS

Triple

1.702

4172.066

IS

Triple

1.442

4501.445

15

Triple

1. 471

-4290.377

10

?

4549.808

20

Triple

2. 125

4294.204

10

Triple

1.958

4563-939

10

Triple

1-325

4300.211

8

?

4572.156

20

Triple

1.526

The values of AX/X^ for the lines in Table 10 do not appear to be as closely related to the interval a
as is usual among a Hke number of triplets taken at random. The measurements are usually of high
weight, the photographs being made with self-induction in the spark circuit, and still there is a total
lack of normal triplets, the values of AX/X= being scattered rather uniformly from 1.3 to 2.1. The most
we can conclude is that for titanium the strongest enhanced lines tend toward the triplet type, but not
toward the simplest intervals of separation. When we extend the comparison to the weaker enhanced
Hnes, many of which are of considerable strength in the arc, a large variety of types appears, with none
predominating.

COMPARISON OF THE RESULTS FOR THE ZEEMAN EFFECT AND FOR PRESSURE

DISPLACEMENT.

A summary of the theories on the possible connection between magnetic separation and pressure
displacement is given on pp. 5-7. The data now at hand permit a considerable extension of the compari-
son made in my former paper {40). This is mainly in two directions. First, photographs of titanium arc
spectra under pressure made in this laboratory by Mr. H. G. Gale have materially added to pressure
measurements for this substance. Although this material has not yet been published by Mr. Gale, he
has kindly permitted me to use his values in this comparison. Second, spectra given by the electric
furnace under pressure have recently been obtained by me, and the preHminary results (63) bear on one
of the questions involved in the present discussion.

In Tables 11 and 12 the values of the magnetic separations in the second column are taken directly
from Tables i and 2 respectively. These values of AX are for the H-components, the mean being taken
when there are two or more pairs. Numerous changes have been made as compared to the former paper
on this subject, due to better photographs being available.

The measurements of pressure displacements expressed in Angstrom units are taken from the publi-
cations of Humphreys (41'-) and of Duffield (64) for the iron spectrum. For titanium, some measurements
are given by Humphreys, but most of the pressure values are from the photographs of Gale. The meas-
urements by Humphreys in the third column are for a pressure of 42 atmospheres, his other measurements,
for 69 and loi atmospheres, being for only a part of the lines. For the iron spectrum, the displacements
of Dufifield for 41 atmospheres are given in the fourth column. For titanium, the measurements of Gale
taken for 9 atmospheres total pressure were multiplied by 4.7 to bring them to the same order as those
of Humphreys, assuming a direct proportion between displacement and pressure. Occasionally a hne
was not obtained by these observers for the given pressures, in which case an approximate value was
deduced from the measurement for some other pressure and is accompanied by an interrogation point.

T;\BLE II. — Zeeman Separations and Pressure Displacements for Iron.

X

Sepa-
ration

H=
16,000.

Displacement.

a

i4

Sepa-
ration
H =
16,000.

Displacement.

2Q

Classes Sep.

AND DiSPL.

42 ATM.

(Humph-
reys.)

41 ATM.
(DUF-
FIELD.)

2Q

X

42 ATM.

(Humph-
reys.)

41 ATM.

(Duf-
field.)

3659.663

3669 , 666

3670.240
3676.457
3677.764
3680.069
3683.229
3684.258
3687.610
3689.614

3695.194
3704.603
3705.708
3709.389
3716.054
3720.084
3722.729
3724.526
3727.778
3733.469
3735.014
3737.281

0.176
0.176
0. 261
0.236
0.167
0.296
0.480
0.170
0.311

0.373
0.261
0.319
0.294
0.312
0.290
0.268
0. 260
0.256
0.318

0.31S
0.310

0.254

0.050
0.050
0.047
0.050
0.052
0.062
0.040
0.053
0.090
0.084
0.070
0.046
0.054
0.095
0.107
0.047
0.050
0.054

0. TOO
0.050
0.092
0.040

3
3
5
4
3
4
12

3
3
4
3
6

S
3
2

5
5
4
3
6

3
6

54
54
55
72
21

77
00
21
46
44
73
93
44
28

71
70
20

74
18
30

37
35

S:S

S:S

S:S

S:S

S:S

S:M

L:S

S:S

M:M

M:M

S:M

M:S

S:S

M:M

S:L

S:S

S:S

S:S

M:M

M:S

M:M

S:S

3738
3743
3745
3746
3748
3749
3758
3763
3765
3767
3788
3795
3798
3799
3805
3813
3815
3820

3824

3826

3827

i 3834

454
508
717
058
408
631
375
945
689

341
046

147
655
693
486
100

987
586

591

027
980
364

0.207
0.318
0.228

0.214
0. 289
0.269
0.218
0.228

0.326

0.325
0.326
0.326
0.204
0.203
0.264
0.282

0.34s
0.274
0.225
0.248

0.078

0. 100?

0.050

0.050

0.040

0.085

0.090

0.095

0.106

0.118

0.090

0.093
0.085
0.075
0.092
0.058
0. no
0.125
0.040
0.090
0.102

O.IIO

2.65
3.18
4.56

5.35
3.40
2.99
2.29
2.15

3.62
3.49
3.84
4. 35
2.22
3.50
2.40
2.26
8.63

3 04
2.20
2.25

S;M

M:M

S:S

0:S

S:S

S:M

S:M

S:M

S;L

0:L

M:M

M:M

M:M

M:M

S:M

S:S

S:L

S:L

M:S

S;M

S:L

S:L

56

COMPARISON OF RESULTS FOR ZEEMAN EFFECT AND FOR PRESSURE DISPLACEMENT.
Table ii. — Zeeman Separations and Pressure Displacements for Iron — Continued.

57

DlSPL.\CEMENT.

pi ^

Displacement. 1

p* ;

fi J

Sepa-

w ^

(fi S

Sepa-

2Q

t« m

X

ration
H=

42 ATM.

41 ATM.

2Q

X

ration
H =

42 ATM.

41 ATM.

!5 Q

16,000.

(Humph-
reys.)

(DUF-
FIELD.)

U

16,000.

(HuirpH-

REVS.)

(Duf-

field.)

^S

2§

3840.580

0.221

0.090

2.46

S:M

4233-772

0-532

0.240

0.370

2.22

L:L

3841.19.';

0.164

O.IOO

1 .64

S:M

4236.112

0.452

0.274

0.40s

i-6s

L:L

3850.118

0.082

0:M

4245-422

0.493

0.060

8.22

L:S

3856.524

0.341

0.038

8-97

M:S

4250.945

0. 246

0.089

0.082

2.76

S:M

3860.055

0.341

0.042

8.12

M:S

4260.640

0.423

0. 246

0.177

1.72

L:L

3865.674

0343

0.103

3-33

M:L

4271-934

0.341

0.083

0.069

4. II

M:M

3872.639

0.284

0.108

2.63

S:L

4282.566

0.310

0.043

0.056

7.21

M:S

3878.720

0.346

0.044?

7.86

M:S

4294.301

0.319

0.084

0.086

3-80

M:M

38S6.434

0.348

0.056

6.21

M:S

4299.410

0.406

0-313

1.30

L:L

3887.196

0.335

0.073

4-59

M:M

4308.081

0.320

0.090

0.060

3-56

M:M

3888.671

0.264

0.089

2.97

S:M

4315-262

0.517

0.036

0.041

14-36

L:S

3893-542

0, 269

0.072

3-74

S:M

4325-939

0.24s

0.097

2.52

S:M

3895.803

0.347

0.030

1. 16

M:S

4337-216

0.264

0.090

0.082

2.93

S:M

3899.850

0.349

0.036

9.69

M:S

4352-908

0.416

0.052

0.056

8.00

L:S

3903.090

0.278

0.095

2-93

S:M

4367 - 749

0.311

0.060

5.18

M:S

3904.052

0.233

0.056

4.16

S:S

4369.941

0.282

0.05s

0.060

S-13

S:S

3906.628

0.347

0.050

6.94

M:S

4376.107

0.424

0.039

0.047

10.87

L:S

3920.410

0.349

0.033

10.58

M:S

4383-720

0.332

0.125

0.060

2.66

M:L

3923 054

0.351

0.032

10.96

M:S

4404.927

0.334

O.IIO

0.056

304

M:L

3928.075

0-352

0.038

0.92

M:S

4407.871

0.631

0.180

3SI

L:L

3930.450

0.352

0.047

0.7s

M:S

4408.582

0.488

0.160

3-oS

L:L

3948.925

0.234

0.050

4.68

S:S

4415-293

0.338

0.087

0.078

389

M:M

3950.102

0.348

0.066

S-27

M:M

4422.741

0.293

0.065

0.046

4.51

M:M

3956.819

0.289

0.036

8.03

S:S

4427.482

0.430

0-055

0.043

7.82

L:S

3969-413

0.354

o.o8g

3-98

M:M

4430-785

0.719

0. 190

0-159

3-78

L:L

3977.891

0.441

0.042

10.98

L:S

4442.510

0.485

0. 190

0.164

2-SS

L:L

3981.917

0.240

0.060?

4.00

S:S

4443-365

0. 170

0.060

0.060

2.83

S:S

3984-113

0. 216

0.085

2-54

S:M

4447.892

0-585

0.180

0.172

3-25

L:L

3986.321

0.196

0.061

3.21

S:M

4454-552

0-44S

0.080

5.56

L:M

3997-547

0.266

0.048

5-54

S:S

4459-301

0.449

0. 160

0.172

2.81

L:L

3998-205

0.226

0.066

3-42

S:M

4461.818

0.435

0.060

0.039

7-2S

L:S

4005.408

0.461

0.103

4-48

L:L

4466.727

0.343

0.056

0.046

6.12

M;S

4009 . 864

0.377

0.040

9.42

M:S

4476.185

0.306

0.072

0.042

4-25

M:M

4014.677

0.250

0.050

5 00

S:S

4494-738

0.302

0.200

0.168

1-51

M:L

4017.308

0.397

0.062

6.40

M:M

4528.798

0-358

0. 172

0.172

2.08

M:L

4022.018

0.272

0.037

7-35

S:S

4531-327

0.400

0.07s

0.078

5-32

L:M

4045-975

0.298

0.103

0.082

2.89

M:L

4548-024

0.311

0.097

3.21

M:M

4063 - 759

0.269

0.107

0.082

251

S:L

4592.840

0.416

0. no

3-78

L:L

4071 .908

0.170

0.092

0.086

1. 8s

S:M

4603.126

0.566

0.093

6.09

L:M

4107.649

0.397

0.060

6.62

M:S

4647.617

0.392

0.070

S-60

M:M

4109-953

0.28s

0.062

4.60

S:M

4691.602

0.358

0.070

S-11

M:M

4118.708

0.271

0.085

0.099

3-19

S:M

4710.471

0.242

0.060

4 03

S:S

4127.767

0.196

0.082

2-39

S:M

4736.963

0.426

0.085

S-01

L:M

4132.23s

0.510

0.105

0. 108

4.86

L:L

4787-003

0.409

0.076

S-38

L:M

4134.840

0-303

0.05s?

0.086

5-51

M:S

4789.849

0.352

0.080

4.40

M:M

4143-572

0.280

0.09s?

S:M

4859.928

0.564

0.390

I -45

L:L

4144.038

0-393

0.116

0.099

3-97

M:L

4871.512

0-336

0.420

0.80

M:L

4154.667

0-379

0.086

4.41

M;M

4878.407

1 .092

0.400

2.73

L:L

4156.970

0-367

0.064?

0.065

5-73

M:M

4919.174

0.591

0.37s

1.58

L:L

4175.806

0.296

0.065

4-55

S:M

5171-778

0.521

0.075

6-95

L;M

41S1 .919

0-339

0.070?

4-84

M:M

S195-I13

0-4S7

0,080

S-71

L:M

4185.058

0-390

0.040

0.047

9-75

M:S

5269.723

0.501

0.083

6.04

L:M

4187.204

0.39s

0.190

2.08

M:L

5328.236

0.470

O.IOO

4.70

L:M

4187.943

0 . 402

0.431

093

L:L

5371-734

0.413

0.095

4-3S

L:M

4191-595

0.402

0.310

1.30

L:L

5397-344

0.630

0.080

7.88

L:M

4195-492

0.320

Large

M;L

5405.989

0.341

0. 100

3-41

M:M

4196.372

0.359

Large

M:L

5429.911

0.607

0.085

7.14

L:M

4198.494

0.383

Large

M:L

5434 - 740

0.120

0:L

4199.267

0.276

0-073

0.06s

3-78

S:M

5447-130

0.568

0.09s

5-98

L:M

4202.198

i 0.323

0.071

0.078

4.14

M:M

5455-834

0.692

0.105

6-59

L:L

4204.101

! 0-373

0 . 060

6.22

M:S

5497.735

1 .040

O.IIO

9-4S

L:L

4210.494

0.806

O.IS7

S-13

L:L

5501.683

1. 001

0.095

10.54

L:M

4219.516

0.284

0.074

i 0.078

384

S:M

5507.000

1 .026

0.120

8.SS

L:L

4222.382

0.47S

0-358

1-33

L:L

5615-S77

0.586

0.080

7-33

L:M

4227.606

0.309

0.431

0.72

M:L

58 INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

Table 12. — Zeeman Separations and Pressure Displacements for Titanium.

DlSPT.ACF.MFNT.

,

i^ .

J II.-~1'L.\CE1IE.NT.

c- .

Sepa-

w ^

"2 g!

Sepa-

1^

C/2 0-

ration

^ 53

X

ration

i

X

H =

17,500.

42 ATM.

(Humph-
reys.)

42 ATM.

(Gale.)

^8

Classe

AND D

H =

17.500.

24 ATM.

(Humph-
reys.)

42 ATM.

(Gale.)

2Q

3900.681

0.272

0.212

1.28

S:L

4318.817

0.337

0.042

8.02

M:S

3904.926

0. 240

0.073

0.085

2.82

S:M

4326.520

0.403

0.141

2.86

L:L

3913.609

0.219

0.174

1.26

S:L

4338.084

0.247

0.089

2.78

S:M

3914.477

0.352

0.028

12.57

M:S

4346.278

0.453

0.028

1.62

L:S

3921.563

0.426

0.019

22.42

L:S

4360.644

0.348

0.183

1 .90

M;L

3924.673

0.292

0.047

6.21

S:S

4394.093

0.325

0.066

4.92

M:M

3926.465

0.247

0-235

1-05

S:L

4395 ■ 201

0.347

0.118

2.94

M:M

3930.022

0.362

0.042

8.62

M:S

4417-450

0.381

0.127

3.00

M:L

3947.918

0.098

0.028

3-SO

S:S

4421.928

0.289

0.179

1. 61

S:L

3948.818

0.186

0.045

0.075

2.48

S:M

4422.985

0.377

0.113

3-34

M:M

3956-476

0. 229

0.030

0.047

4.87

S:S

4426. 201

0.318

0. 122

2.61

M:M

395S.35S

0.287

0.045

0 . oSo

3-59

S;M

4427. 266

0.312

0.024?

0.070

4.46

M:M

3962.995

0.461

0.042

10.98

L:S

4434.168

0.259

0.174

1-49

S:L

3964.416

0.359

0.03S

9-45

M:S

4440.515

0.270

0.141

1. 91

S:L

3981.917

0.18S

0.056

0.094

2.00

S:M

4443-976

0.298

0.103

2.89

S:M

3982.630

0.469

0.019

24.68

L:S

4449-313

0.388

0.118

3-29

M:M

3989.912

0.275

0.049

0.103

2.67

S:M

4451.087

0.340

0.122

2.79

M:M

3998.790

0.317

0.047

0.113

2.80

M:M

4453-486

0.210

0.183

I-15

S:L

4009.079

0.347

0-055

0.028

12.39

M:S

4453-876

0.263

0.108

2.43

S:M

4009 . 807

0.086

0.038

2. 26

S:S

4455-485

0-351

0.193

1.82

M:L

4012.541

0. 198

0.042

4.71

S:S

4457-600

0.400

0.193

2.07

L:L

4024.726

0.394

0.038

10.37

M:S

4465-975

0.481

0. 122

3-94

L:M

4028.497

0.269

0.08s

3.16

S:M

4468 . 663

0.340

0.216

1-57

M:L

4035-976

0.354

0.244

1-45

M:L

4471.408

0.601

0.089

6.75

L:M

4055-189

0.39s

0.085

4-65

M:M

4475.026

0.509

0.362

1. 41

L:L

4060.415

0.395

0.07s

S-27

M:M

4479.879

0.829

0.132

6.28

L:L

4064.362

0.396

0.094

4.21

M:M

4480.752

0.611

0.136

4-49

L:L

4065 • 239

0.395

0.047

8.40

M:S

4481.438

0.548

0.113

4-85

L:M

4078.631

0.39s

0.019

20.79

M:S

4489.262

0.612

0. 146

4.19

L;L

4082.589

0.398

0.061

6.52

M:M

4501.445

0.298

0.216

1.38

S:L

4 1 1 2 . 869

0.301

0.047

6.40

M:S

4512. 906

0.501

0.132

3 -So

L:L

4151.129

0.305

0.207

1-47

M:L

4518.198

0.498

0.136

3-66

L.L

4159-805

0.263

0.160

1.64

S:L

4518.866

0.220

0.094

2-34

S:M

4103.818

0.294

0.179

1.64

S:L

4522.974

0.502

0.146

3-44

L:L

4171-213

0. 210

0.146

1-44

S:L

4527.490

0.49S

0.132

3-75

L:L

4172. 066

0. 251

0.188

1-34

S:L

4533-419

0.469

0. 176

0. 150

i-^i

L:L

4186.280

0.282

0.056

504

S:S

4534-953

0.449

0.124

0. 160

2.81

L:L

4203.620

0-457

0.179

2-SS

L:L

4535-741

0.424

0.136

3-12

L:L

4272.701

0.364

0.07s

4-8s

M:M

4536.094

0-323

0.113

2.86

M:M

4276. 5S7

0.443

0.136

3-26

L:L

4536.222

0.160

0:L

4278.390

0.304

0.188

1.62

M:L

4544 . 864

0.502

o.oSo

0.136

6.27

L:L

4281.530

0.664

0.061

1 .09

L:M

4548.938

0.560

0.150

3-73

L:L

42S2.860

0.244

0.132

1.85

S:L

4549.808

0.440

0.226

1-95

L:L

4285.164

0.566

0. 160

3-54

L:L

4552.632

0.510

0.132

3-86

L:L

4286.168

0.400

0.103

0.099

4-04

L:M

4555-662

0.506

0.132

3-83

L:L

4287.566

0.421

0.0S7

o.n8

3-57

L:M

4562.814

0.424

0.038

1. 12

L:S

4289.237

0.370

0. 108

3-42

M:M

4563-939

0.276

0.150

1.84

S:L

4290.377

0.284

0.216

I-31

S:L

4572.156

0.319

0.235

1.36

M:L

4291. 114

0.441

o.iis

0.103

4.28

L:M

4617.452

0.404

0.136

2.97

L.L

4294.204

0.361

0.136

2.6s

M:L

4623.279

0.379

o.n8

3.21

M:M

4295.914

o.ioo

0.103

0:M

4629.521

0.527

0.169

3-12

L:L

4298.828

0.218

0.118

1-85

S:M

4682.088

0.399

0.077

5.18

M:M

4299.410

0.430

0.103

4-17

L:M

4691-523

0.441

0.080

S-Si

L:M

4299-803

0-356

0.103

346

M:M

4758-308

0.382

0.067

5-70

M:M

4300.211

0.367

0.136

2.70

M:L

4759-463

0.430

0.092

4.67

L:M

4300.732

0.265

0.104

0.099

2.68

S:M

4841.074

0.390

0.029

13-44

M:S

4301.158

0.350

O.IIO

0.113

3.10

M:M

4981.912

0.481

0.077

6.25

L:M

4302.085

0.368

0.160

2.30

M:L

4991.247

0.458

0-135

3-39

L:L

4306.078

0.367

0.104

0.113

3-25

M:M

4999.689

0.413

0.120

3-44

L:M

4313 034

0.449

0.216

2.08

L:L

SO07.398

0.339

0.150?

2.26

M:L

4314.964

0.424

0.146

2.90

L:L

5013-479

0.455

0.056

8.12

L:S

COMPARISON OF RESULTS FOR ZEEMAN EFFECT AND FOR PRESSURE DISPLACEMENT.

59

The fifth and sixth columns contain ratios of Zeeman separation to pressure displacement, the one
numerical, the other of letters denoting the order of magnitude. In the numerical ratios for iron the
values of Humphreys are used for the sake of uniformity, those of Dullield for an almost equal pressure
being taken when a Une was not measured by the former. In the case of titanium the values of Gale
are the more numerous and are used in the ratios when possible. The letters S, M and L in the sixth
column stand for small, medium and large values, respectively, of separation and displacement. The
limits covered by these classes are as follows:

Sep.vration.

Displacement.

Iron.

Titanium.

s

<o.3oo

0.300-0.400

>o.4oo

<o.o6o

0.060-0. 100

>O.IOO

<o.o6o
0.060-0.125

>O.I25

M

L

The reasons for this classification are given later.

The question as to whether there is a close proportionality between magnetic separation and pressure
shift is decided in a definite manner by the sixth column in Tables 11 and 12, giving the numerical
ratio of separation to displacement. The separations for each spectrum are taken for a constant field
and the displacements for a constant pressure. The probable errors in measurement can explain only
in a very small degree the larger differences in these ratios. For iron the ratio-values run from 0.72 to
14.36, for titanium from 1.05 to 22.42. The distribution between these limits is such that any range
which might reasonably be assumed as due to poor measurements covers but a fraction of the lines.
Thus in Table 11, ratios ranging from 2.00 to 5.00 take in 90 out of 173 lines, or 52 per cent; the same
range for titamum includes 67 out of 122 lines, or 55 per cent. The range from 3.00 to 5.00 in the two
spectra covers 35 and 34 per cent respectively.

The lack of constancy in the ratio being apparent, the question arises as to whether there is any real
connection between separation and displacement. A broad classification of the values in order of magni-
tude may be of service in this connection. For this purpose the separation and displacement values are
classified as small, medium and large, the range for each class being given above. The ratios showing
the comparative magnitudes of separation and displacement for each line are given in the sixth col-
umn of the tables. The displacement measures for titanium run in general larger than for iron, so that
a higher point of division between the medium and large classes is chosen. The following summary of
the data will show to what extent a general agreement exists between the Zeeman and pressure phenomena.

The ratios of classes from Tables 11 and 12 enable us to form Table 13, in which the 173 iron and 122
titanium lines are placed in three main groups. Group i consists of the ratios S : S, M : M, L: L, and
shows that the separation and displacement for the corresponding lines are relatively of the same order.
Group 2 contains those lines for which separation and displacement are not in the same, but in adjacent,
classes; while for Group 3 the separation and displacement are of very different magnitude, one small
and the other large. Those lines which show no Zeeman effect, but distinct pressure displacement, are also
in Group 3, the letter 0 being associated with S, M, or L according to the magnitude of the displacement.

It will be seen that 44 per cent of the iron lines are in good agreement as to order of magnitude,
44 per cent show a probable discordance, while 12 per cent strongly contradict the hj'pothesis of
equality of relative magnitude. Titanium shows a somewhat larger proportion of its lines in poor agree-
ment as to separation and displacement. This indicates clearly that the two phenomena are not very
closely related as regards size of one increasing with size of the other. The large number of lines in Group
2 renders any positive conclusion difficult on account of the possible influence of errors of measurement.

6o

INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

Trials with other limits for the small, medium and large classes have shown that the group percentages
are not materially altered, as this results in a transfer back and forth of lines near the limits chosen. An
attempt to reduce Group 2 was made by taking all those Hnes which had one or both values so near the
hmit of the class that the error of measurement, if in the favorable direction, might have put the two
values into the same class and so have brought the line into Group i. Lines of complex Zeeman separa-
tion were also treated in this way; 35 iron lines were thus selected, which when added to Group i as given
in Table 13 raised its total to 64 per cent of the whole. This number, then, may be in fair agreement as
to order of magnitude, while the remaining 36 per cent are divergent beyond the errors of measurement
and in some distances widely different. This last device is of course not a fair treatment of the data,
since the error of measurement is as likely to move the values wider apart as closer together, and if the
same treatment had been applied to the lines of Group i, some of them would have moved into Group 2.
However, giving the agreement hypothesis the benefit of the doubt, the proportions of 64 and 36 per
cent appear to be the most favorable that can be gotten out of the list of iron lines.

Table 13. — Summary of Classes.

Iron.

Titanium.

Ratio of Mag.

No. of
Lines.

Group
Total.

Group
Percentage.

Ratio of Mag.

No. OF
Lines.

Group
Total.

Group
Percentage.

Group I

S:S

M:M

L:L

Group 2

S:M

M: S

M:L

L:M

Group 3

S:L

L:S

0:S

0:M

0:L

24
29

23

27
22
13
15

8
8
I

I
2

■ 76

' 77

• 20

44
44

12

Group I

S:S

M:M

L:L

Group 2

S:M

M:S

M:L

L:M

Group 3

S:L

L;S

6
21
26

12
10
12
12

IS
6
0

I
I

S3
46

23

43
38

19

0:8

0:M

0:L

In Group 3 we have those lines for which either separation or displacement is small and the other
large, and in addition 4 lines of iron and 2 of titanium which appear to be unaffected by the magnetic
field, while they show a variety of displacements, in some cases large. These offer examples of ability to
respond to one displacing agency and not to the other.

A closer quantitative comparison is afforded by taking the average separations and displacements for
large groups of lines. This is done in Tables 14 and 15. The method in forming Table 14 was to make
a list of all pressure displacements classified as small, place opposite them the Zeeman separations for
the same lines, and take the mean of each list for comparison of the magnitude of the two effects. Means
were formed in the same way for lines of medium and large displacement. The ratios of mean separa-
tion to mean displacement can then be compared. In obtaining the results for each class, means were
formed for the lines in three groups according to wave-length. The whole table thus gives a comparison
of the means for the several groups, and also an indication as to how the means for both separation and
displacement change with the wave-length.

Table 15 was made in the same way as Table 14, except that here the class of Zeeman separation,
small, medium, or large, was taken as the basis, and the corresponding pressure displacements used for
a comparison of means.

COMPARISON OF RESULTS FOR ZEEMAN EFFECT AND FOR PRESSURE DISPLACEMENT.
Table 14. — Means of Separation and Displacement Classified according to Amount of Displacement.

61

Iron.

Titanium.

Range of
X.

No. of
Lines.

Means.
Sep. Displ.

Ratio

Sep.

Displ.

Range of
X.

No. OF

Lines.

Means.
Sep. Displ.

Ratio

Sep.
Displ.

Displacement: Small

3660-4000
4000-4500
4500-5600

35

18

I

0.290
0.361
0.242

0.046
0.051
0.060

6.30
7.08
4-03

3900-4000
4000-4500
4500-5000

1^
3

0.339
0.319
0.423

0.034
0.038
0.041

9.97

8.39

10.32

Total of lines and weighted

54

6.52

22

0.340

0.037

9.19

Displacement: Mediimi .... J

3660-4000
4000-4500
4500-5600

30

22

19

0.272 0.084
0.297 ' 0.080
0.478 0.085

■3-24
3.71
5-62

3900-4000
4000-4500
4500-5000

6

30

9

0.249
0.378
0.385

0.092
0.099
0.093

2.71
3-82
4.14

Total of lines and weighted
means

71

0.335 1 0.083

4.04

45

0.362

0.097

3-73

Displacement: Large J

3660-4000

4000-4500
4500-5600

8

24
9

0.221 0.109

0.452 0.207
0.679 0.245

2.03
2.18
2.77

3900-4000 3
4000-4500 31

4500-5000 I 9

0.246

0.379
0.446

0.207
O.I75
0.157

1. 19
2.16
2.84

Total of lines and weighted

•

41

0.462

0.196

2.36

53

0.396

0.170

2.33

Table 15. — Means of Separation and Displacement Classified according to .Amount of Separation.

Iron.

Titanium.

Range of
X.

No. of
Lines.

Means.

Ratio

Sep.

Displ.

Range of
X.

No. OF
Lines.

Means.

Ratio

Sep.

Displ.

Sep.

Displ.

Sep

Displ.

Separation: Small

3660-4000

4000-4500
4500-5600

42
iS

I

0.240
0.258
0.242

0.072
0.077
0.060

3.33
3.22
4.03

3900-4000
4000-4500
4500-5000

II

19
3

0.230
0.247
0.265

0.107
0.128
0.153

2.IS

1.93
1.73

Total of lines and weighted
means

61

0.246

0.073

3-37

33

0.243

0.123

1.98

Separation; Medium <

3660-4000
4000-4500
4500-5600

28

^4

S

0.337
0.346
0.356

0.065
0.098
0.136

5 18
3-53
2.62

3900-4000
4000-4500
4500-5000

4

34

7

0.34S
0.360
C.362

0.055
0.114
0.113

6.33
3.16
3.20

Total of lines and weighted
means

60

0.343

0.088

3.90

45

0-3S9

0.108

3-32

Separation : Large \

3660-4000
4000-4500
4500-5600

2

23
20

0.460

0.495
0.629

0.041
0.173
0.137

II .22
2.86
4. 59

3900-4000
4000-4500
4500-5000

3
18

21

0.452
0.519
0.471

0.027
0.138
0.127

1.67
3.76
3-71

Total of lines and weighted
means

45

0.553

0.151

3.66

42

0.490

0.125

3.92

62

INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTR.V OF IRON AND TITANIUM.

In Table 14 the ratios of classes given by the weighted means for the three magnitudes of displace-
ment are M : S, M : M, and L : L for both iron and titanium. Table 15 gives for the three magnitudes
of separation the ratios S : M, M : M, L : L, for both elements. There is thus good agreement as to magni-
tudes except for the first class in each table. A large proportion of the lines for this class come from
the region below X4000 and there is a sufficient scattering of high values for both separation and displace-
ment to put the means into dififerent classes when formed in this way. The behavior of the ratios of
weighted means in the two tables is interesting. Those in Table 15 decrease very nearly in the ratio
3:2:1 for the three classes in the iron table, and about 9 : 4 : 2 for titanium, showing that the displace-
ments increase in size much faster than the separations. The same material is used in Table 15, but here we
find an approximate constancy for iron and a gradual increase for titanium. It is probable that the
change as shown in Table 14 is a real one and that it is obscured in Table 15 by the large difference in
range of values of separations and displacements. The limits of this range are in the ratio of about i to
3 for the separations (omitting a few extreme values) and about i to 10 for the displacements. Thus,
in Table 14, when the displacements are grouped so as to increase in magnitude, there is a much smaller
variation among corresponding values of separation than we have among the displacement values when
the separations are graded as in Table 15. The widely divergent values of displacement scattered through
Table 15 would thus act to make the ratios of means more or less discordant.

A classification byDuffield (64(0 maybe used in comparing the displacements measured by him with
the corresponding Zeeman separations for iron. He forms three main groups according to amount of
displacement. Table 16 gives the mean separation and displacement for each of these groups, at first
singly, then combined so as to form two groups with more lines in each.

Table 16. — Means of Separation and Displacement for Duffield's Displacement Groups.

No. OF Lines.

Mean Sep.

Mean Displ.

Classes Sep.
and Displ.

Group I

Unreversed

26
13

6
10

39
16

0. ^^?

n nfi.i

M:M
M:M

L:L
L:L
M:M
L:L

Reversed

0.317 0.077
0.483 0,168

Group II

Group III

Total of Group I

Totals of Groups I and II

0.329
0.431

0.068
0.262

We see that separation and displacement are of the same order of magnitude throughout. In the
last two lines the larger number of values gives means of higher weight. These means show as before
that a much larger range is covered by the displacements than by the separations.

Two additional points are to be considered in this comparison. The first is the rate of increase of
the two effects with magnetic field and pressure, respectively. Duffield found that the displacements of
lines belonging to the three groups treated in Table 16 have very different rates of increase with increase
of pressure, the lines of Group III showing the most rapid change. A corresponding phenomenon in the
Zeeman effect would mean a different rate of increase of separation with field-strength for different lines.
We are not certain that this does not e.xist, since the proportionality of separation to field-strength has
been estabhshed by careful measurement for only a very few lines, but no evidence of a difference for
different sets of lines has thus far been presented.

The second point is the relation of the variation of separation and displacement with the wave-length.
In Tables 14 and 15 the division into regions of wave-lengths shows the distribution of magnitudes in
these regions. Following down the columns headed "No. of Lines" in each table, we see that the propor-
tion of small values for both separation and displacement is greater in the region of short wave-lengths.

COMPARISON OF RESULTS FOR ZEEMAN EFFECT AND FOR PRESSURE DISPLACEMENT. 63

For the medium and large values in each table, the proportion of lines increases in the region of greater
wave-length, this being very decided for the "large" group. Thus there is a clear increase in magnitude
of both separation and displacement as the wave-length increases. The lines here compared seem to be
representative of the spectrum, as the same relation holds in the complete Zeeman tables, which contain
a much larger number of lines for this range of wave-length.

When pressure measurements of high accuracy are available for an extended region of wave-length,
the rate of variation with the wave-length will appear, and the closeness of agreement with the relation
found for iron and titanium, namely, that the magnetic separation increases proportionally with the
square of the wave-length (p. 54), will afford strong evidence concerning the common physical basis
of the two phenomena. An attempt at a comparison of this sort has been made by the author in a recent
paper (63) on the effect of pressure upon electric-furnace spectra. The displacements of iron lines given
by the electric furnace for a pressure of 9 atmospheres were measured for two regions 1000 A apart, from
X4050 to X 4450 and from X505o to X5450. The list for the latter region did not include as many of the
weaker lines, whose displacements are often large, as was available for the blue region, so that a compari-
son of the means of all displacements would not have been fair. It seemed best to hmit this preliminary
comparison to those lines in each region which show the same general behavior in various light sources.
In the furnace they appear at low temperatures and show reversal with strong widening under pressure.
They are lines which, although not connected by series relations, show such similarity in their response
to the excitations of furnace, arc, and spark that the vibrating particles which produce them can be
assumed to have many points of similarity.

Fifteen lines of this character in the blue region were compared with nine similar hnes in the green.
The mean pressure displacement for the two sets was found to be almost identical, being 0.058 A for the
blue and 0.060 A for the green lines. The magnetic separations of the same lines, taken from Table i,
give mean values of 0.330 A and 0.520 A, respectively, for the blue and green regions, an increase of 60
per cent for a difference of wave-length of about 1000 A. The evidence from these selected lines is, there-
fore, against a close connection between the magnetic and pressure phenomena. Measurements for the
arc under pressure, however, show a more frequent occurrence of large displacements as we pass toward
greater wave-lengths, and more complete measurements will show the rate of change.

Summarizing the comparison here presented, it may be said that there is a fair agreement between
magnitude of magnetic separation and pressure displacement for the lines of iron and titanium when the
means of large groups are considered. The number and character of the lines not in agreement, however,
show that the correspondence is not close enough to justify preferring any one of the theories for the
pressure effect on this ground, or to predict the eflect upon a given line of one influence from that observed
for the other. The degree of concordance which we have could perhaps result entirely from the fact
that the magnitude of each effect increases with the wave-length. This does not prove a close physical
relation, since any theory of the pressure effect that might be offered would probably involve a change
with the wave-length. A comparison of the rates of change of the two effects appears to be a more prom-
ising line of investigation than an extension of the method followed for iron and titanium; as the number
of lines treated for those spectra is sufficient to show clearly the degree of correspondence.

SUMMARY OF RESULTS.

The leading features in this investigation may be summarized as follows:

1 . The effect of a magnetic field upon the spark spectra of iron and titanium has been studied for a
total number of 1120 lines between the limits X3660 and X 6743. The character of the magnetic separa-
tion is given, with weighted measurements as complete as was permitted by the magnetic fields available.

2. The types of resolution, ranging from lines unaffected by the magnetic field to those having thirteen
and possibly more components, have been classified and the important features of each class have been
discussed.

3. The relation of the measured separations to the "normal interval"

e H

m \-KV
has been studied for all t>'pes of resolution. A large majority of the separations of triplets and quadru-
plets show a close relation to this interval, while the generality with which the more complex tj-pes show
the spacing of their components to be simply related to this interval indicates a full confirmation of
Runge's law.

4. Many cases of "magnetic duplicates," i.e., lines exactly similar in resolution, with the same inter-
vals between components, have been found among the more complex types, indicating close similarity
in the light vibrations which give rise to these lines. Large groups of lines showing triplet separation
are similar in this respect.

5. The large range of wave-length covered has made it possible to observe the rate of increase of
magnetic separation with the wave-length. This increase is such that the mean value of AX/X^ for suc-
cessive intervals throughout this range shows a close approach to constancy for both iron and titanium,
with no systematic variation. The conclusion is that for these spectra the mean separation of Zeeman
components varies as the square of the wave-length.

6. Cases of unsymmetrical separation of Zeeman components, so distinct as to be classed as abnormal,
have been pointed out. The theory of Voigt concerning a slight dissymmetry in the intensity and spacing
of the components of triplets has been tested for a number of iron lines, with the result that this effect
appears to be real in many cases, although some lines fail to show such a difference.

7. The enhanced lines of the two elements have been compared with those showing no enhancement
in the spark, both as to type and magnitude of separation. The only difference between the behavior
of the two classes in the magnetic field appears to be that among the stronger enhanced fines of titanium
the triplet type strongly predominates, the separations usually being of medium amount and not closely
related to the interval a.

8. On account of a possible similarity between the actions of the magnetic field and of pressure around
the light source as displacing agencies, a detailed comparison has been made of the magnetic separations
and corresponding pressure displacements for these spectra. It was proved that a close correspondence
does not exist, but there is a general agreement as to magnitude of the two effects when the means for
large numbers of lines are considered.

In conclusion, I wish to acknowledge my great obligations to Mr. Hale for his unfailing support and
interest in the equipment and development of the physical laboratory and for much advice as to the
conduct of the investigations. A great deal of credit is due also to Miss Wickham and to Miss Griffin
for their careful and often diflficult work in the measurement and reduction of the photographs. The
large number of spectrograms required to do justice to the iron spectrum, in particular, increased the
work of measurement out of proportion to the total number of lines treated.
64

BIBLIOGRAPHICAL REFERENCES.

1. A. Cotton, " Le Phcnonicne de Zeeman," Sciciiliii, (s),

Paris, 1899; (a), p. 52.

2. H. Kayser, "Handbuch der Spectroscopic," 2, ix, 1902;

(a), pp. 63s, 636; (6), p. 620; (c), p. 672.

3. W. VoiGT, "Magneto- und Elektrooptik," Leipzig, 1908,

(a), p. 86; (6), Ch. m.

4. H. A. LoRENTZ, "The Theorj' of Electrons," Leipzig, 1909;

(a), p. 109.

5. P. Zeeman, "Over den Invloed eener Magnetisatie op den

aard van het door een Stof uitgezonden Licht," Ver-
slagen Koninklijke Akademie Amsterdam, 5. 181, 242,
1896; Philosophical Magazine, (5), 43. 226, 1897;
Astrophysical Journal, 5, 332, 1897.

"Over Doubletten en Tripletten in het Spectrum teweeg
gebracht door uitwendige Magnetische Krachten,"
Verslagen Koninklijke Akademie Amsterdam, 6, 12, 99,
260, 1897; Philosophical Magazine. (5), 44. 255, 1897.

"Metingen over Stralingsverschijnselen in het Magnetisch
Veld, "Verslagen Koninklijke Akademie Amsterdam, 6,
408, 1897; Philosophical Magazine, (5), 45. i97i 1898.

6. A. A. MiCHELSON, "Radiation in a Magnetic Field," Philo-

sophical Magazine, (5), 44. 109, 1897; 45. 348, 1898;
AslrophysicalJoiirnal, 6, 48, 1S97; 7, 131, 1898.

7. Th. Preston, "Radiation Phenomena in a Strong Magnetic

Field," Philosophical Transactions Royal Society Dub-
lin, (2) 6, 385, 1897.

8. .A. CoRNU, "Sur quelques Resultats nouveaux relatifs au

Phenomene decouvert par M. le Dr. Zeeman," Comptes
Rendus, 126, 181, 1898.

9. H. Becquerel et H. Desl.\ndres," Contribution a I'Etude du

Phenomene de Zeeman, "Com/i/es Rendus, 126, 997, 1898.

10. " Obser\'ations nouvelles sur le Phenomene de Zeeman,"

Comptes Rendus, 127, 18, 1898.

11. J. S. Ames, R. F. Earhart, and H. M. Reese, "Notes on the

Zeeman Effect," Astrophysical Journal, 8, 48, 1898.

12. H. M. Reese, ".\n Investigation of the Zeeman Effect,"

Astrophysical Journal, 12, 120, 1900.

13. N. A. Kent, "Notes on the Zeeman Effect," Astrophysical

Journal, 13. 289, 1901.

14. H. A. LoRENTZ, "Ueber den Einfluss magnetischer Krafte auf

Lichtemission," Annalcn der Physik, (3), 63, 278, 1897.
" Beschouwingen over den Invloed van een magnetisch
Veld op de Uitstraling van Licht," Verslagen Konink-
lijke Akademie Amsterdam, 7, 113, 1898; Astrophysical
Journal, 9, 37, 1899.

15. J. L.ARSiOR, "On the Theory of the Magnetic Influence on

Spectra," Philosophical Magazine, (5), 44. 503, 1897.

16. W. VoiGT, "Zur Theorie der magneto-optischen Erscheinun-

gen," Annalen der Physik, (3), 67, 345, 1899.
"Weitercs zur Theorie des Zeemaneffektes," Annalen der

Physik, (3), 68. 353, 1899.
5

19.

23

A. A. ROBB, "Beitriige zur Theorie des Zeemaneffektes,"
Annalen der Physik, (4), 15. 107, 1904.

C. Runge und F. Paschen, "Ueber die Strahlung des Queck-
silberlichts im magnetischen Felde," Abhandlungen
Akademie der Wissenschaften Berlin, 1902, .\nhang i.

A. Farber, "Ueber das Zeeman-Phanomen," Annalen der
Physik, (4), 9, 886, 1902.

P. Weiss et A. Cotton, "Mesure du Phenomene de Zeeman
pour les trois Raies Bleues du Zinc," Journal de Phys-
ique, (4), 6. 429, 1907.

F. Paschen, " Ueber die absolute Messung des Zeeman-
effektes," Physikalische Zeitschrift, 8, 522, 1907.

A. Stettenheimer, "Absolute Messungen des Zeeman-
Phanomens," Annalen der Physik, (4), 24, 384, 1907.

Th. Preston, "Radiation Phenomena in the Magnetic
Field," Nature, 59, 224, 1899.
"General Law of the Phenomena of Magnetic Perturba-
tions of Spectral Lines," Nature, 59, 248, 1899.
"Radiating Phenomena in a Strong Magnetic Field,"
Philosophical Transactions Royal Society Dublin, (2), 7. 7.
1899; Report British Association, 1899, p. 63; Nature,
61, II, 1899.

C. Runge und F. Paschen. " Ueber die Zerlegung einander
entsprechender Serienlinien im magnetischen Felde,"
Sitzungsberichte Akademie der Wissenschaften Berlin,
1902, 380, 720.
Runge und J. Precht, "Ueber die Magnetische Zerle-
gung der Radiumlinien," Sitzungsberichte Akademie der
Wissenschaften Berlin, 1904, 417.
26. W. Miller, "Zeemaneffect an Magnesium, Calcium,
Strontium, Zink, Cadmium, Mangan und Chrom,"
Annalen der Physik, (4), 24, 105, 1907.

W. LoHMANN, " Beitriige zur Kenntnis des Zeeman-Phano-
mens," Dissertation, Halle, 1907; Zeitschrift fiir Wis-
senschaftliche Photographie, 6, 41, 1908.

B. E. Moore, "Upon the Magnetic Separation of the lines
of Barium, Yttrium, Zirconium, and Osmium," Astro-
physical Journal, 28, i, 1908; Annalen der Physik, (4),
25. 309. 1908-

"Upon the Separation of the Spectral Lines of Thorium in
the Magnetic Field," Astrophysical Journal, 30, 144,
178, 1909.

C. Runge, "Ueber die Zerlegung von Spektrallinien im mag-
netischen Felde," Physikalische Zeitschrift, 8, 232,
1907.

P. Zeeman, "Some Observations concerning an Asymmetri-
cal Change of the Spectral Lines of Iron, radiating
in a Magnetic Field," Verslagen Koninklijke Akademie
Amsterdam, 8, 328, 1899.

65

24

25. C.

27

29

30

66

INFLUENCE OF A MAGNETIC FIELD UPON THE SPARK SPECTRA OF IRON AND TITANIUM.

31. P. Zeeman, "Magnetic Resolution of Spectral Lines and

Magnetic Force," Second Part, Verslagen Konkinklijke

Akademie Atnslcrdam, December, 1907.
"New Observations concerning Asymmetrical Triplets,"

Verslagen Koniiiktijke Akademie Amslerdam,M!irch,igoS.
" Change of Wave-Length of the Middle Line of Triplets,"

Verslagen Koninklijke Akademie Amsterdam, May, 1908.
"The Law of Shift of the Central Component of a Triplet

in a Magnetic Field," Verslagen Konijiklijke Akademie
Amsterdam, January, 1909.

32. P. Gmelin, "Ueber die unsymmetrische Zertegung der gelben

Zuecksilberlinie 5790 im magnetischen Felde,"
Physikalische Zeitsehrift, g, 212, igo8.

33. R. Jack, "Zeeman-Effekt an Wolfram und Molybdan,"

Dissertation Guttingen, 1908; Annalen der Physik, (4),

28, 1032, 1909; Proceedings Royal Society Edinburgh,

29, 75, 1908.

34. A. DuFOiTR, "Observation faite parallelemcnt aux Lignes de

Force des Dissymetries d'Intensites et de Positions des
Composantes Magnetiques de quelques Raies d'Emis-
sion," Le Radium, 6, 298, 1909; Comples Rendus, 148,
1594, 1909.
" Sur les Triplets Magnetiques Dissymetriques," Le Radium,
7, 74, 1910-

35. W. VoiGT, "Magneto- und Elektrooptik," iv Kap.

36. W. KoNlG. "Beobachtung des Zeeman'schen Phanomens,"

Annalen der Physik, (3), 62, 240, 1897.

37. A. Cotton, "Le Phenomene de Zeeman," p. 69.

38. G. E. Hale, "On the probable Existence of a Magnetic Field

in Sun-Spots," Contributions from the Mount Wilson
Solar Observatory, No. 30; Astrophysical Journal, 28,
315, 1908.
3Q. P. Zeeman and B. Winawee, "The Magnetic Separation of
Absorption Lines in connection with Sun-spot Spectra,"
Verslagen Koninklijke Akademie Amsterdam, 12, 584,
1910; i3,35,i62,i9io;.-l5/TO/'/m/(;a/yo!(r»i;/,32>329,i9io.

40. A. S. King, "The Correspondence between Zeeman Effect

and Pressure Displacement for the Spectra of Iron,
Chromium, and Titanium," Contributions frotn the
Mount Wilson Solar Observatory, No. 46; Astrophysical
Journal, 31, 433, 1910.

41. W. J. Humphreys, (a),".\n Attempt to Find the Cause of the

Width and of the Pressure-Shift of Spectrum Lines,"
Astrophysical Journal, 23, 233, 1906.

(6), "Arc Spectra under Heavy Pressure," Astrophysical
Journal, 26, 18, 1907.

(c), "Note on the Cause of the Pressure-Shift of Spectrum
Lines," Astrophysical Journal, 26, 297, 1907.

(d), "The Luminous Particle a strong Magnet, and the
consequent Pressure-Shift of Spectral Lines," Astro-
physical Journal, 27, 194, 1908.

42. W. J. Humphreys, "Bericht iiber die Verschiebung von

Spektrallinien durch Druck," Jahrbuch der Radio-
aktivitat und Elektronik, 5, 324, 1908.

43. O. W. Richardson, "A Theory of the Displacement of Spec-

tral Lines produced by Pressure," Philosophical Maga-
zine, (6), 14, S57> 1907-

44. J. Larmor, "Note on Displacement of Spectral Lines,"

Astrophysical Journal, 26, 120, 1907.

45. A. DuFOUR, "Sur un Cas Exceptional du Phenomene de

Zeeman," Comptes Rendus, 146, 118, igoS. "Modifi-
cations anomales, dans le Champ Magn^tique, des
Spectres de Bandes de divers Composes," Comptes
Rendus, 146, 229, 1908.

46. R. Rossi, " Pressure Effect on Band Spectra of Fluorides,' '

Proceedings Royal Society, 82, 518, 1909.

47. A. DupOUR, "Normale und anormale Veranderungen gewis-

ser Banden in den Emissionsspektren der Molekiile
verschiedencr Korper in gasformigem Zustande unter
dem Einfluss eines Magnetfeldes," Physikalische
Zeitsehrift, 10, 124, 1909.
Hartmann, "Das Zeeman Phanomen im sichtbaren
Spectrum von Kupfer, Eisen, Gold, und Chrom,"
Dissertation, Halle, 1907.

B. VAN Bilderbeek-van Meurs, "Magnetische Splitsing
van het Ultraviolette Ijzerspectrum," Dissertation,
Amsterdam, 1909.

E. Purvis, "The Influence of a Strong Magnetic Field on
the Spark Spectra of Titanium, Chromium, and Man-
ganese," Proceedings Cambridge Philosophical Society,
14, (i), 41, igo6.

S. King, "The Zeeman Effect for Titanium," Contribu-
tions from the Mount Wilson Solar Observatory, No. 39;
Astrophysical Journal, 3°, i, 1909.

S. King, "On the Separation in the Magnetic Field of
Some Lines occurring as Doublets and Triplets in Sun-
Spot Spectra," Contributions from the Mount Wilson
Solar Observatory, No. 34; Astrophysical Journal, 29,
76, 1909.

53. G. E. Hale, "The Pasadena Laboratory of the Mount Wilson

Solar Observatory," Contributions from the Mount Wil-
son Solar Observatory, No. 27; Astrophysical Journal, 28,
244, 1908.

54. G. E. Hale, "The Tower Telescope of the Mount Wilson

Solar Observatory," Contributions from the Mount Wil-
son Solar Observatory, No. 23; Astrophysical Journal,
1908.

in Sensitometry, II — Ortho-
Astrophysical Journal, 26, 299,

48.

W.

49.

H.

so.

J-

SI-

A.

52-

A.

5S-

R.

56.

H.

57-

F.

ss.

J.

S9-

B.

60.

P.

61.

H.

62.

H.

" Ueber die Spectren der Ele-
Akademie der Wissenchaften

63. A.

64. W

27, 204,

J. WALL.-iCE, " Studies
chromatism by Bathing,'
1907.

Kayser and C. Runge,
mente," Ahhandlungen
Berlin, 1888.

Exner und E. Haschek, " Wellenlangen-Tabellen der
Funkenspektren der Elemente," Leipzig and Vienna,
1902.

N. LocKYER, "Tables of Wave-Lengths of Enhanced
Lines," Solar Physics Committee, 1906.

Hasselberg, "Researches on the Arc Spectra of the
Metals," Astrophysical Journal, 4, 116, 212, 1896.

FiEBiG, " Untcrsuchungen iiber den langwelligen Teil des
Titanspektrums," Zeitsehrift Jiir W isscnschaftliche Pho-
tographic, 8, 73, 1910.

BuissoN et C. Fabry, "Spectre du Fer," Annates de la
Faculte des Sciences de Marseille, 17, igo8.

D. Babcock, "The Zeeman Effect for Chromium," Con-
tributions from the Mount Wilson Solar Observatory, No.
52; Astrophysical Journal, 33> 217, 1911.
'The Zeeman Effect for Vanadium," Contributions from
the Mount Wilson Solar Observatory, No. 55; Astrophysi-
cal Journal, 34, 209, igii.

S. King, "The Effect of Pressure upon Electric Furnace
Spectra," Contributions from the Mount Wilson Obser-
vatory, No. 53; Astrophysical Journal, 34, 37, igii.
, G. Duffield, "The Effect of Pressure upon Arc Spectra,
— No. I, Iron," Philosophical Transactions, Royal Soci-
ety, London, A, 208, in, igo8; (a), p. 160.

PLATE 2

A.— OCCULTING PLATE. B.— SPECTROGRAPH.

PLATE 3

O R. o.

ir.

-;,8so

-3720

-3860

p- p

9 -^

-3866

-372t

-3S73

-3S78

"3 744

O

bBh

o

-3S89

-3896

-3750

-3758

-3903

—3764

\rmm

— '".i/

-6.1 0 =

-6318

— 5,iM7

C/1

-6337

k M

-6iy2

o s

G o

■ ^B —6394

-6412

l^>3

-6253

-6431

IL

-5400

-541;

-5424

-5430

-5435

-4204

-4308

'5447

—5456

-4326

—433 7

-4353

Q Cs CT- O

Q r^ cr a

V. 3

i. 8 °

?- &

O cr -I

o ^ ^

3 5'"

o o *■

3 _ <»

o

o °

— 44S()

—4496

-4501

SB"

■Bm

IP

-4270

-4275

I -:

-4513

-4518

-4523

—4527

i

IW-.-v .> I

—4533

MMilii

-4545

-4549

■
■

-4556

■

I
I

-42S2

-4 2 86

-4291

-4296

-4299

I —4302

-430S

1—4313

-4319

—4325

-4338

^^^^_^^^

PLATE 6

(3- Q

5S67

5QOO

2. o

-5016

O

o

3

—5922

-5025

<-^ 3:

I o

-5942

-5°3(>

-5979

-5040

THE INFLUENCE

OF A

MAGNETIC FIELD UPON THE SPARK SPECTRA
OF IRON AND TITANIUM

BY

ARTHUR S. KING

WASHINGTON, D. C.
Published by the Carnegie Institution of Washington

1912

MBl, WHOI r.lBKAHY

UH IfiJfl H