7?

i*k^

^,.

''A

,*i

~.fA

Sf's

//

ft. / ' ft*

if

J&MW

. <

WE *■

'A &

7 «-

<3^

c

Zfa.*

^.^

*2&?&

^^m^
'^^i

:^^ ^

<Tc

5^QC^?

^ ,^

«?

53

:Xv <3J£=>

' fr- "-

' ^>.

.'V

sai

~<r^ *-*^

S^Sfe^s <S?

«sc

PROCEEDINGS

OF THE "

AMERICAN ACADEMY OF ARTS AND SCIENCES.

PROCEEDINGS

OF THE

AMERICAN ACADEMY

OF

ARTS AND SCIENCES.

NEW SERIES.
Vol. XVI.

WHOLE SERIES.
Vol. XXIV.

FROM MAY, 1888, TO MAY, 1889.

SELECTED FROM THE RECORDS.

seo-

BOSTON:
UNIVERSITY PRESS: JOHN WILSON AND SON.

1889.

CONTENTS.

Page
I. On the Action of Sodium Malonic Ester on Tribromdinitro-

benzol. By C. Loring Jackson and W. S. Robinson 1

II. On Sodic Zincates. By Arthur M. Comey and C.

Loring Jackson 14

III. A Preliminary Account of the Development and Histology of

the Eyes in the Lobster. By G. H. Fakker . . . . 24

IV. Atmospheric Economy of Solar Radiation. By Arthur

Searle 126

V. The Crystalline Structure of the Coahuila Irons. By

Oliver W. Huntington 30

<h~- VI. Contributions to American Botany. By Sereno Watson

36

VII. The Determination of Chromium in Chrome Iron Ore. By

Leonard P. Kinnicutt and George W. Patterson 88

VIII. The Strength of the Microphone Current as influenced by
Variations in Normal Pressure and Mass of the Electrodes.
By Annie W. Sabine 90

IX. Researches on Microphone Currents. By Charles R.

Cross and Annie W. Sabine 94

X. On Pentamidobenzol. By A. W. Palmer and C. Loring

Jackson 105

XL The Strength of the Induced Current with a Magneto Tele-
phone Transmitter as influenced by the Strength of the
Magnet. By Charles R. Cross and Arthur S.
Williams 113

XII. An Account of a New Thermograph, and of some Measures
in Lunar Radiation. By C. C. Hutchins, assisted by
Daniel Edward Owen 125

b'Gl]

VJ CONTENTS.

Page

XIII. On the Charging of Condensers by Galvanic Batteries. By

15. 0. Peikce and It. \Y. Willson 146

XIV. Classification of the Atomic Weights in Two Ascending

Series, corresponding to the Groups of Aiiiads and
Perissads. By W. R. Livermoke 164

XV. Neutralization of Induction. By John Trowbridge and

Samuel Sheldon . 176

XVI. The Magnetism of Nickel and Tungsten Alloys. By John

Trowbridge and Samuel Sheldon 181

XV 1 1. The Micanique G'leste of Laplace, and its Translation with

a Commentary by JJowditch. By JOSEPH Lovering . 185

XVIII. On a New Method of determining Gas Densities. By

Josiah Paksons Cooke 202

XIX. On the Action of Sodium Malonic Ester on Tribromdinitro-

benzol. By C. Loring Jackson and W. S. Robinson 234

XX. On some Nitro Derivatives of Metabromtoluol. By W. B.

Bentley and W. II. Warren 250

XXI. On the A el ion of Sodium Malonic Ester upon Tribromtrini-
trobenzol. By C. Loring Jackson and George Dun-
ning Moore 256

XXII. On the Action of Sodium Acetacetic Ester upon Tribromdi-
nitrobenzol. By C. Loring Jackson and George
Dunning Moore 271

XXIII. On Tetrabromdinitrobenzol By C. Loiung Jackson and

W. I). Bancroft 288

XXIV. General Considerations in regard to Certain Compounds

■prepared from Bromnitrobenzols. By C. Loring Jackson 306

XXV. Features of Crystalline Growth. By Oliver "Whipple

IIlNTINGTON 313

XXYI. I. On Chlorpyromucic Acids. By Henry B. Hill and

Louis !-■ Jackson 320

II. <>n Certain Derivatives of Furfuracrylic Arid. By

II. B. Gibson and C. F. Kahnweiler . . . 364

III. On the so called Dioxymaleic Acid. By W. S.

Hi n drixson ; . 376

CONTENTS. vii

Page
XXVI T. An Address delivered at the Meeting of April 10, when the

Rumford Medals were j)resented to Professor A. A.

Michelson. By Joseph Lovering 380

Proceedings 403

Memoirs : —

George Rumford Baldwin 429

Jonathan Ingersoll Bowditch 435

Samuel Kneeland 438

Frederick Augustus Porter Barnard 441

John Call Dalton 445

Ezekiel Brown Elliott . 447

Michel Eugene Chevreul 452

Rudolf Julius Emanuel Clausius 458

Franciscus Cornelius Donders 465

List of the Fellows and Foreign Honorary Members . . 471

Statutes and Standing Votes of the Academy .... 479

Rumford Premium 488

Index 439

PROCEEDINGS

OF THE

AMERICAN ACADEMY

OF

ARTS AND SCIENCES.

VOL. XXIV.
PAPERS READ BEFORE THE ACADEMY.

I.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF
HARVARD COLLEGE.

ON THE ACTION OF SODIUM MALONIC ESTER ON
TRIBROMDINITROBENZOL.

By C. Loring Jackson and W. S. Robinson.

Presented June 13, 1888.

In an earlier paper from this laboratory,* the preparation of tribrom-
trinitrobeuzol and its action with ammonia and aniline were described,
and the announcement was made that the study of the behavior of
this very reactive body would be continued. G. D. Moora and one of
us accordingly have taken up the action of sodium malonic ester upon
it, in the hope of substituting for each of the bromine atoms the radi-
cal of malonic ester, CH(COOC.2H.)2, and obtaining in this way a
substance which might have yielded interesting derivatives. As,
however, an entirely different product was the result of the reaction,
it seemed wiser to study it with the much more accessible tribrom-
dinitrobenzol, which acted in the same way, and to confine the work
on the trinitro compound, which at best can be obtained only with
difficulty, to those points in which the two substances showed a dif-
ferent behavior. In this paper, therefore, we give the results obtained
up to the present time from the action of sodium malonic ester on

* These Proceedings, Vol. XXIII. p. 138.

VOL. XXIV. (N. S. XVI.) 1

2 PROCEEDINGS OF THE AMERICAN ACADEMY

tribromdiuitrobenzol, which we are obliged to publish now, although
in some respects the work is not so finished as we could wish, because
the departure of one of us from Cambridge will prevent us from con-
tinuing the investigation together.

The most important of these results can be summarized briefly as
follows. Sodium malouic ester under the conditions used by us
removes from tribromdiuitrobenzol only two atoms of bromine, one
of which is replaced by the radical of malonic ester, the other by
hydrogen, so that the formula of the product is

C6H2Br(N02)2CH(COOC2fL)2.

This substance shows marked acid properties, and, although of a pale
yellow color itself, forms salts most of which are dark blood-red ; of
these the sodium salt was analyzed, and its formula proved to be

C6H2Br(N02)2CNa(COOC2H5)2.

The yellow copper salt was also analyzed, but proved to be a some-
what complex basic salt. The bromine contained in the substance is
easily removed by treatment with aniline, the product being

C6H2CcH5NH(N02)2CH(COOC2H5)2,

which also gives a blood-red sodium salt with a formula similar to
that of the salt derived from the original substance, but the acid prop-
erties of the anilido compound are much less pronounced. When
heated with strong hydrochloric acid the bromdinitrophenylmalonic
ester gives ethylchloride and a new substance melting at 170°, the
analysis of which led to the formula C(.H2Br(NO.)),,C;,H502, but the
determination of its constitution must be postponed till a later paper.

Action of Sodium Malonic Ester on Tribromdiuitrobenzol.

If .in alcoholic solution of sodium malonic ester is added to tribrom-
diuitrobenzol (melting point 192°) dissolved in ether, the first drop
imparts to the ethereal solution a blood-red color, which increases in
intensity as more of the solution of sodium malonic ester is added, and
the liquid contains the sodium salt of a new substance, which we pre-
pared in the following way. To an ethereal solution of a weighed
quantity of tribromdinitrobenzol we added an alcoholic solution of
sodium malonic ester in the proportion of three molecules of the ester
to one of the tribromdiuitrobenzol. The following were found to be

OF ARTS AND SCIENCES. 3

convenient proportions : 20 grm. of tribromdinitrobenzol, 16 grm. of
raalonic ester dissolved in 100 to 150 c.c. of absolute alcohol and
treated with 2.4 grm. of sodium as sodic ethylate. It was not neces-
sary to add the ester gradually, and there was no perceptible evolution
of heat during the reaction. The mixture was allowed to stand in a
corked flask at ordinary temperatures for three or more hours, and
the dark blood-red liquid thus obtained filtered from the precipitate
of sodic bromide, which had been deposited, acidified with dilute sul-
phuric acid, which changed the dark red color to pale yellow, and
filtered once more to remove the sodic sulphate formed. The filtrate
was then distilled on the water bath until most of the ether had passed
over, when the residue in the flask deposited, as it cooled, crystals of
the new substance mixed with a large quantity of unaltered tribrom-
dinitrobenzol. The mother liquor from these crystals was allowed to
evaporate spontaneously, and left a red or yellow oil mixed with a
small additional quantity of the crystals, which were separated from
it, and the oil allowed to stand in an, open dish for some weeks, when
it deposited an additional quantity of the crystals. The crystals,
whether obtained by cooling the hot alcoholic solution, or from the
oil by long standing in the cold, were purified by dissolving them in
a little hot alcohol, and adding a moderately strong aqueous solution of
sodic hydrate (the common laboratory solution diluted with its own
volume of water), which converted the new substance into its soluble
sodium salt, but left the tribromdinitrobenzol unaltered, the red solu-
tion was then poured into a large quantity of water, and the tribrom-
dinitrobenzol removed by filtration. If the solid crystals were treated
directly with the aqueous solution of sodic hydrate, the tribromdinitro-
benzol was left in a form which clogged the filter very badly, whereas
when an alcoholic solution was used, as directed above, no difficulty
was encountered in the filtration. The red filtrate was acidified with
dilute sulphuric acid, and the precipitate purified by crystallization
from hot alcohol until it showed the constant melting point 75°-76°.
After being dried in vacuo, its composition was determined by the
following analyses.

I. 0.2061 grm. of the substance gave on combustion* 0.2910 grm.
of carbonic dioxide and 0.0602 £rm. of water.
II. 0.2013 grm. gave 0.2827 grm. of carbonic dioxide and 0.0511
grm. of water.

* Care must be taken to heat the substance very gradually, as it shows a
tendency to explode in the combustion tube.

4 PROCEEDINGS OP THE AMERICAN ACADEMY

III. 0.1075 grm. gave 0.2780 grm. of carbonic dioxide

IV. 0.1926 grm. of substance gave 12.4 c.c. of nitrogen at a temper-

ature of 20° and a pressure of 740 mm.
V. 0.221 1 grm. of substance gave by the method of Carius 0.103C
grm. of argentic bromide.
VI 0.2005 grm. gave 0.0937 grm. of argentic bromide.

Found.
I. II. III. IV. V. VI.

Carbon 38.51 38.30 38.38

Hydrogen 3.24 2.83

Nitrogen 7.17

Bromine 19.91 19.88

These results agree about equally well with the numbers calculated
from two different formulas, as appears by the following comparison.

Calculated for

Found

Calculated for

C0HBr(NO, ).,('( C02C2H5)2.

Mean.

CGH2Br( N02)2CH( C02C2H;)2.

Carbon

38.71

38.40

38.51

Hydrogen

2.72

3.03

3.21

Nitrogen

6.94

7.17

0.91

Bromine

19.85

19.89

19.75

The first of these formulas, in which the two atoms of bromine
removed from the tribromdinitrobenzol have been replaced by the
bivalent radical =C(C02C2H.)2, seems at first sight the most probable,
especially since W. II. Perkin, dr. and others have observed the
strong tendency of sodium malonic ester to react in this way ; but
although this is the only formula we have been able to find which ex-
plains easily the removal of two atoms of bromine in this reaction, we
are inclined to reject it in favor of the second one, in which one atom
of liromine has been replaced by the univalent radical -CH(C02C2H5)2,
and the other by hydrogen, for the following reasons: —

First, the formation of a sodium salt is hard to reconcile with the
lirst formula, unless, indeed, sodic hydrate was added to it, t giving

CTI15r(N011)liOHCNa(C02C2TI,)2;

but our analyses of the sodium salt, given later in this paper, prove
that it contains no hydroxy], and to make this proof more convincing
we prepared the suit, for one of our analyses from sodic ethylate, as

* Tin* water was lost in tins analysis,
t Wislicenus, Ann. Chem., ccxlii. 23.

OF ARTS AND SCIENCES. 5

the radical CJLO, which in this case would be added instead of
hydroxyl, would make a very great difference in the percentage com-
position of the salt; but this preparation, like the others, gave on
analysis numbers corresponding to the formula

C0H2Br(NO2)2CNa(COOC2H5)2.

Secondly, the ease with which the sodium salt is formed, and with
which it is decomposed by an acid giving the original substance, mili-
tates against the first formula, as we can hardly suppose that a ring of
carbon atoms, such as must be assumed in this formula, would break,
when treated with an alkali, and reunite under the influence of an acid
in dilute solutions. Finally, we may add that the meta position of the
bromine atoms replaced deprives the first formula of any support from
analogy.* The principal objection to the formula

CGH2Br(NO:,)2CH(COOC2H;i)2,

which we adopt, is that we cannot account as yet for the formation
of such a substance by the reaction of sodium malonic ester on tri-
bromdinitrobenzol. We hope, however, that a careful study of the
quantitative conditions of the reaction, and of the oil which is the
secondary product, will throw light on this part of the subject ; but
owing to the departure of one of us from Cambridge, its further
investigation must be postponed till next autumn.

The yield was far from satisfactory, 14 grm. of tribromdinitroben-
zol after treatment with 12 grm. of malonic ester and 1.7 grm. of
sodium, as described above, giving only 4.3 grm. of the bromdinitro-
phenybnalonic ester instead of the 14 grm. required, if the whole of
the tribromdinitrobenzol was converted into the new substance, that
is, 30.7 per cent of the theoretical yield. A large amount of the tri-
bromdinitrobenzol was recovered, however, amounting to 5.5 grm., or
39.3 per cent. This leaves 30 per cent unaccounted for, which prob-
ably remains dissolved in the oil deposited by the alcoholic mother
liquors ; but that the dissolved substance is bromdinitrophenylmalonic
ester admits of great doubt, as we have not succeeded in obtaining any
considerable quantity of this substance from the oil by treatment with
aqueous sodic hydrate. This point will be investigated more thor-
oughly hereafter. We have tried several modifications of the process,
in hopes of improving the yield, but none of them have given satisfac-
tory results ; if, for instance, the proportion of sodium malonic ester is

* Compare J. Stanley Kipping, Ber. d. ch. G. 1888, p. 32.

6 PEOCEEDINGS OF THE AMERICAN ACADEMY

increased in order to bring more of the tribromdinitrobenzol into the
reaction, any gain from this cause is more than compensated by the
larger amount of the product taken up by the greater quantity of oil,
which is only in part deposited on long standing. Nor was the yield
increased by longer standing, or by boiling the materials with a re-
verse condenser ; in this latter case the product came down in an oily
form, and with, if anything, rather a smaller yield was much harder
to purify.

Properties of the Bromdinitrophenylmalonic Ester,
CGH2Br(NO,)2CH( COOC2H5)2.

The substance crystallizes in pale yellow flattened needles, or long
plates terminated by a single plane, which forms with the sides angles
not very far from 90°, but distinctly obtuse and acute. It melts be-
tween 75° and 76°, and shows a tendency to explode at high tem-
peratures, so that the heat must be applied carefully when making a
combustion of the substance. It is not very soluble in cold alcohol,
but freely in hot, more soluble in methyl than in ethyl alcohol, freely
soluble in ether, benzol, or glacial acetic acid, even more so in chlo-
roform, moderately soluble in acetone or carbonic disulphide, nearly
insoluble in ligroine or water. Strong sulphuric' acid in the cold has
little action on it, although perhaps a small quantity dissolves ; on
warming the mixture, complete solution takes place, and addition of
water precipitates a solid, which seems to consist principally of un-
altered bromdinitrophenylmalonic ester mixed with a small amount
of a substance with a higher melting point, probably the compound
formed by the action of concentrated hydrochloric acid in sealed tubes
(see page 11). Nitric acid, either strong or fuming, dissolves the
bromdinitrophenylmalonic ester in the cold, but does not produce any
change of color. In this respect, it differs markedly from the corre-
sponding trinitro compound. If the acid solution is warmed, decom-
position seems to take place.

The bromdinitrophenylmalonic ester possesses well-marked acid
properties, as was to be expected, since it contains a hydrogen atom
standing between the two carboxyls of the malonic ester and a phenyl,
which, as Victor Meyer* has shown, frequently exercises a negative
influence, much heightened in this case by the presence of the two
nitro groups. Sodic hydrate in either aqueous or alcoholic solution at
once converts it into a soluble red sodium salt, and the same salt is

* Bcr. d. ch. G. 1887, p. 534, 1888, p. 1291 et seq.

OP ARTS AND SCIENCES. 7

formed by an alcoholic solution of sodium malonic ester, by sodic car-
bonate dissolved in water, or by acid sodic carbonate mixed with very
dilute alcohol. The aqueous solution of the salt is, however, entirely
decomposed by carbonic dioxide, the original substance being precipi-
tated. An aqueous solution of the sodium salt prepared by treating
an excess of the original substance with a solution of sodic hydrate
gave precipitates with salts of all the common metals except those of
the alkalis. Most of these precipitates were red, and among them the
following were especially characteristic : —

Baric, strontic, or calcic chloride, dark brick-red.

Magnesic sulphate, rust color.

Zincic acetate, dark chrome-orange.

Basic plumbic acetate, very dark brick-red.

The neutral plumbic acetate gave a yellow precipitate, which turned
white when in contact with an excess of lead salt.

Cupric sulphate, dark golden yellow.

Ammonic hydrate also dissolved the substance with a red color,
but the salt was decomposed on trying to evaporate off the excess of
ammonia on the water-bath, leaving a whitish residue, which seemed
to consist principally of the original substance, with a small amount of
decomposition product. The substance is in fact easily decomposed
by either ammonic or sodic hydrate, and our rather unsatisfactory ob-
servations on these reactions will be found on page 11. The sodium
salt boiled under a reverse condenser with ethylbromide in alcoholic
solution, passed from red through purple, bluish green, green, and
brown to reddish yellow ; sodic bromide was formed, and the other
product was an oily substance, which solidified after some time. It
will be studied later. The bromine atom in the bromdinitrophenyl-
malonic ester seems to be removed with comparative ease. Up to
this time, we have studied its behavior in this direction only with
aniline (see page 9).

The Sodium Salt CGIT2Br(X02)2CXa(COOC,II,V,

This substance was prepared by dissolving some of the bromdini-
trophenylmalonic ester in alcohol and adding a little sodic hydrate,
care being taken that the hydrate was not in excess; a little ether
was then added, and the mixture evaporated rapidly in a Bmall beaker
sunk throughout its whole length in the water-bath. The ether vapor
kept the air from coming in contact with the liquid, while the alcohol
was being heated to its boiling point. The residue, after all the alcohol
had evaporated, was extracted with benzol to remove the excess of the

8 PROCEEDINGS OF THE AMERICAN ACADEMY

ester, and then dried by gentle warming, and standing in vacuo over
paraffine and sulphuric acid. These precautions were adopted to
avoid the contamination of the substance with the carbonic dioxide of
the air. The same substance was obtained when sodic ethylate was
used instead of sodic hydrate, and Analysis III. was made with a
sample prepared in this way.

I. 0.1950 grm. of the substance ignited with sulphuric acid gave

0.0324 grm. of sodic sulphate.
II. 0.1354 grm. gave 0.0216 grm. of sodic sulphate.

III. 0.1943 grm. gave by the method of Carius 0.0838 grm. of argen-

tic bromide and the filtrate gave 0.0330 grm. of sodic sulphate.

IV. 0.2148 grm. gave 14.3 c.c. nitrogen at a temperature of 25° and

a pressure of 759.6 mm.

Sodium

Calculated for
,Br(N02),CNa(C02C2H5)2.

5.39

i.
5.38

Found.
II. III.

5.16 5.50

IV.

Bromine

18.74

18.35

Nitrogen

6.55

7.42

The results of these analyses show that the substance is the normal
salt, containing but one atom of sodium, and the important bearing of
this point on the determination of the formula of the bromdinitro-
phenylmalonic ester has been discussed already. (See page 4.)

Properties. — The salt forms a dark blood-red powder, easily sol-
uble in water, alcohol, or ether. From a solution containing much
sodic hydrate, the salt can be extracted with ether ; but pure water,
on the other hand, extracts the salt from its ethereal solution. Its
aqueous solution is decolorized by carbonic dioxide bromdiuitrophenyl-
malonic ester being precipitated.

The copper salt seemed from its appearance to promise excellent
results on investigation, and was accordingly prepared as follows.
An alcoholic solution of the bromdinitrophenylmalonic ester was
treated with a small quantity of sodic hydrate taking care that the
ester was in large excess, and then poured into a volume of water
sufficient to precipitate all the unaltered ester; after filtering, an alco-
holic solution of cupric chloride was added, and the precipitate washed
first with 50 per cent alcohol, then with water, and finally once with
hot alcohol ; it was then dried in vacuo. It is not necessary to pro-
tect the salt from contact with the air during its preparation, as it is
not perceptibly affected by carbonic dioxide. It gave the following
results on analysis.

OF ARTS AND SCIENCES. 9

I. 0.2052 grm. of the salt was treated according to the method of
Carius and gave 0.0844 grm. of argentic bromide. The fil-
trate was freed from argentic nitrate, and the copper precipi-
tated by electricity from a sulphuric of nitric acid solution
giving 0.0197 grm. of copper.
II. 0.2073 grm. treated in the same way gave 0.0864 grm. of

argentic bromide and 0.0198 grm. of copper.
III. 0.1455 grm. gave 0.0138 grm. of copper.

Calculated for

HOCu2[C6H2Br(NO,)2C(COoC2H5)2]3.

Copper 9.36

i.
9.60

Found.
II.

9.55

in.
9.48

Bromine 17.70

17.50

17.74

From these results it would appear that the substance is a basic
salt, or, at least, that it is not the normal salt, and in spite of its very
promising appearance is not fitted to throw light on the nature of the
original substance. We have not thought it worth while, therefore, to
examine it more carefully. We may add, that the salt changes color
from a lighter to a darker yellow soon after it is prepared, which we
are inclined to ascribe to the conversion of

ClCu2[C6H2Br(N02)2C(C02C2Hs)2]3

into the corresponding hydroxyl compound by the action of the water
in which it is suspended.

Properties. — The copper salt forms an obscurely crystalline pre-
cipitate of a golden yellow color, with a very slight greenish tinge.
It is essentially insoluble in water, alcohol, or ether, and is a compar-
atively stable substance, as it seems to be unaltered by exposure to
the air, or by heating to 100°. Strong nitric acid, even in excess, de-
composes it, setting free the bromdinitrophenylmalonic ester.

Anilidodinitrophenylmalonic Ester,
C6H5NHC6H2(N02)2CH(COOC2H5)2.

This substance was made by mixing the bromdinitrophenylmalonic
ester with aniline in the proportion of one molecule of the ester to
two of the base. The solid becomes yellow almost immediately, and
the action has come essentially to an end after the mixture has stood
for a few minutes in the cold. As a measure of precaution, however.
we always heated it gently for a few minutes to make certain that
all the bromine had been removed. The product was purified by
cystallization from alcohol, until it showed the constant melting point
118°, dried at 100°, and analyzed.

10 PROCEEDINGS OF THE AMERICAN ACADEMY

I. 0.2104 grm. of the substance gave on combustion 0.4262 grm. of

carbonic dioxide and 0.0903 grm. of water.
II. 0.1886 grm. of the substance gave 16.9 c.c. of nitrogen at a tem-
perature of 22° and a pressure of 766.6 mm.

Calculated for Found.

C6H2(NHCGH5)(N'0,)2CH(C02C2H5)2. I. II.

Carbon 54.68 55.24

Hydrogen 4.56 4.77

Nitrogen 10.07 10.24

Properties. — The anilidodinitrophenylmalonic ester forms flattened
needles terminated by one or two planes, which form much more
oblique angles with the sides than those which appear in the bromdi-
nitrophenylmalonic ester. It is also found occasionally in star-shaped
groups of needles, has a bright yellow color, and melts at 118°. It is
only slightly soluble in cold alcohol, much more soluble in hot, but
not freely even in this. More soluble in methyl than in ethyl alcohol ;
soluble in ether, or in carbonic disulphide ; freely soluble in benzol,
glacial acetic acid, or acetone, very freely in chloroform ; essentially
insoluble in water or ligroine. The best solvent for it is alcohol.
Concentrated hydrochloric acid does not dissolve, or act on it in open
vessels, either hot or cold. An aqueous solution of sodic hydrate
acts on it in the cold only very slowly and imperfectly, but gives a
red solution when boiled with it., In alcoholic solution sodic hydrate
converts it into a red soluble salt ; the anilido compound therefore still
possesses acid properties, but these are much less marked than in the
corresponding bromine compound ; the negative influence of the
dinitrophenyl therefore has been partly neutralized by the introduc-
tion of the basic aniline radical CrH.NH.

The Sodium Salt C6H2(C6H5NH)(NOJ2CNa(COOC2H5)a.

This substance was formed in the same way as the sodium salt of
the bromdinitrophenylmalonic ester ; that is, by treating an alcoholic
solution of the substance with alcoholic sodic hydrate not in excess,
evaporating to dryness with a little ether on the water-bath, and
extracting the excess of the ester with benzol. It was dried at 100°,
and analyzed.

0.1180 grm. of substance gave 0.0194 grm. of sodic sulphate.

Calculated for Found.

C0H2(CCH3N H X NO A,C \a( C02C2H5)2.

lium 5.24 5.32

OF ARTS AND SCIENCES. 11

Properties. — It resembles the sodium salt of the bromdinitrophenyl-
malonic ester closely in appearance and solubility, but is of a some-
what lighter red. Its solution is decomposed, when treated with
carbonic dioxide.

Experiments on the Saponification of BromdinitrophcnyhnaJonic

Ester.

In taking up this subject we considered it necessary to study the
action of alkalis on the ester, although, owing to the ease with which
the nitro groups are attacked, we had little expectation of reaching
satisfactory results. In this we were not deceived, but we think it
best to give a brief account of these experiments before describing
our more successful work with strong hydrochloric acid in sealed
tubes.

A solution of sodic hydrate in water, if boiled with the brom- or
anilidodinitromalonic ester, gave a brownish red solution, from which
acids precipitated nothing ; but after acidification, and extraction with
ether, an unpromising oil was obtained in very small quantity. Cold
alcoholic sodic hydrate with the anilido compound gave a somewhat
more promising result, but in this case also most of the substance
seemed to be decomposed. The bromine compound after standing for
two weeks with ammonic hydrate at ordinary temperatures was con-
verted into a similar brown solution, which on acidification gave a
resinous brown precipitate and a yellow filtrate, from which ether ex-
tracted a substance melting above 200° , ammonic bromide being
formed during the process. This is the only one of these products
which seems worth further investigation.

o

Action of Hydrochloric Acid.*

"When the bromdinitrophenylmalonic ester was heated with strong
hydrochloric acid in a sealed tube to 140°-145°, it was decomposed,
and upon opening the tube a gas was given off in tolerable quantity,
which burnt with a green-bordered flame, and was without doubt
ethylchloride. About one gram of the substance and :><> c.c. of
common strong hydrochloric acid were used in each tube, and the
temperature should not be allowed to go above 150°, as in this r:i--
an impure product was the result. If proper care was used in the
heating, the tube contained crystals, or a clear liquid, from which
crystals were deposited after the pressure was relieved; these were

* All our work on this part of the subject must lie regarded as preliminary.

12 PROCEEDINGS OF THE AMERICAN ACADEMY

filtered out and the filtrate thrown away, as it gave almost no residue
on evaporation, and no extract when shaken with ether.

The crystals were purified by crystallization from dilute alcohol (three
parts of water to one part of alcohol). If the temperature was kept
below 150°, the product was nearly pure, as it came from the tube ; but
if it had risen above this point, a small quantity of a lower melting
substance was formed, which it was almost impossible to remove with
dilute alcohol. In this case the best plan was to dissolve the substance
in chloroform, when the impurity separated at first as a pasty mass,
leaving the principal product in solution, which after one crystalliza-
tion from dilute alcohol showed the constant melting point 170°. It
was dried at 100°, and analyzed with the following results.

I. 0.1307 grm. of substance gave on combustion 0.1634 grm. of
carbonic dioxide and 0.0278 grm. of water.
II. 0.2162 grm. of substance gave 0.2688 grm. of carbonic dioxide
and 0.0495 grm. of water.

III. 0.2136 grm. of substance gave 18.3 c.c. of nitrogen at a temper-

ature of 22° and a pressure of 760 mm.

IV. 0.1197 grm. gave 10.2 c.c. of nitrogen at a temperature of 28°

and a pressure of 762.9 mm.
V. 0.2212 grm. of substance gave by the method of Carius 0.1365
grm. of argentic bromide.
VI. 0.2120 grm. gave 0.1275 grm. of argentic bromide.

VI.

26.29 25.59

These numbers agree tolerably well with those required for the
formula CcH2Br(N02)2C3H502.

Calculated.

Carbon 33.86

Hydrogen 2.19

Nitrogen 8.77

Bromine 25 08

But the formula cannot be considered definitely established, until we
have supported our analytical results by a careful study of the deriva-
tives of the substance.

Found.

I.

ii.

III. IV

Carbon

34.10

33.91

Hydrogen

2.37

2.54

Nitrogen

9.48 9.3

Bromine

OF ARTS AND SCIENCES. 13

Properties. — The substance crystallizes from dilute alcohol in yel-
lowish white narrow plates, sometimes half a centimeter in length
which are made up of needles attached to another needle acting as a
midrib at very acute angles, giving an exact imitation of a feather or
more commonly of one side of a feather ; this form is very character-
istic, and, when not so well developed, the substance can be recognized
by the formation of narrow plates, usually smooth on one' side and
rather irregularly serrated on the other, or upon the ends. The plates
often occur in radiating groups, the members of which form very acute
angles with each other. From methyl alcohol or ether it crystallizes
in very slender needles, much branched at very sharp angles, often
forming circular groups resembling certain seaweeds. It melts at 1 70° ,
is not very soluble in cold alcohol, freely in hot, more soluble in methyl
than in ethyl alcohol, freely soluble in glacial acetic acid or acetone,
tolerably soluble in ether, slightly soluble in benzol either cold or hot,
or in carbonic disulphide, slightly soluble in cold chloroform, more
soluble but not freely in hot, nearly insoluble but not quite so in cold
water, more soluble but still very slightly in hot, essentially insoluble
in ligroine. The best solvent for it is chloroform, or a mixture of
alcohol and water. Its behavior with alkalis is very characteristic.
If a drop of an aqueous solution of sodic hydrate is added to the sub-
stance dissolved in alcohol, the solution takes on a dark but brilliant
Prussian green color. This solution becomes colorless on addition of
hydrochloric acid, but turns green again on addition of sodic hydrate.
The green alcoholic solution, if allowed to evaporate spontaneously,
leaves a brownish yellow residue, and a solution of the same brown-
ish yellow color is obtained if an excess of sodic hydrate is added to
the original substance. If the yellow solution, in whichever way pre-
pared, is acidified with hydrochloric acid, it loses its color, and oil-drops
are precipitated, which solidify in time. These dissolve in sodic hy-
drate with a brownish claret * or pale magenta color, which is unal-
tered by dilution with water, but turns green on addition of alcohol.
The substance melting at 170° is not affected by an aqueous solution of
sodic carbonate. The further study of these interesting reactions has
been broken off by the summer vacation, but will be continued in thi-
laboratory in the autnmn.

* A somewhat similar change of color was observed during the action "i
ethylbromide on the sodium salt of the bromdinitrophenylmalonic ester.

14 PROCEEDINGS OF THE AMERICAN ACADEMY

II.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF
HARVARD COLLEGE.

ON SODIC ZINCATES.
By Arthur M. Comey and C. Loring Jackson.

Presented June 13, 1838.

Although the solubility of zincic hydrate in alkalis is one of the
most familiar facts in chemistry, very few attempts, so far as we
can find, have been made to determine what compounds exist in such
a solution. The reason for this is not far to seek, as the results ob-
tained by the few who have studied this subject are so decidedly
unsatisfactory, that they are not likely to induce others to take up
this line of work.

The first important papers in this field appeared in the Annalen
der C/iemie unci Pharmacie in 1834, having been brought out by a
prize offered from the Hagen-Bucholz foundation for work on zincic
oxide. Of the various competitors Laux,* who was most successful
in his work on this part of the subject, found that a solution of zincic
oxide in caustic alkali, if covered witli a layer of alcohol, gradually
deposited little shining crystals, which were easily soluble in water,
and, he states, contained equal molecules of zincic oxide and the alka-
line oxide (Zn02K2). These crystals were decomposed by heat, giv-
ing a white powder, which contained one molecule of alkaline oxide
to two of zincic oxide, but no analyses of either of these substances
are given.

Two other competitors for this Hagen-Bucholz prize, Bonnet t
and Sander,J also touch on this subject, but their results are of less
importance. Sander, in fact, comes to the conclusion that no defi-
nite compound is formed when zincic oxide is dissolved in a caustic
alkali.

The next paper on the subject was published in 1842, by Fremy, §

* Ann. Ohcm. Pliarm., ix. 183. \ Ibid., ix. 181

t Ibid., ix. 177. § Comptes Rendus, xv. 1106.

OP ARTS AND SCIENCES. lo

who states that the compounds formed by zincic oxide and an alkali
were in general deliquescent and amorphous, but by using a potassic
hydrate solution of zincic oxide, and adding to it a small quantity of
alcohol, long needles were obtained, which he considered a " bi-
zincate of potassa." It was immediately decomposed by water into
zincic oxide and potassic hydrate. A few years later, in his article
in the Annales de Chimie et de Physique,* Fremy states that he has
met with great difficulties in obtaining the crystals, and hopes to re-
turn to the subject when he has determined the conditions under
which they are formed. Neither of his papers contains analyses, and
he has never published anything more on the subject, the reason for
which we can well understand after our unsuccessful attempts to pre-
pare potassic zincates.

The work on the ammonic zincates has led to about the same un-
satisfactory results, no analyses of laboratory products having been
published; but Malaguti t gives an analysis of an incrustation found
upon a brick in the vault of a privy, which had the composition
3ZnO. 4NH3. 12H20.

The only other research which has approached this subject is one
by Prescott,* published in 1880, on the solubility of zincic oxide in
caustic alkalis. He found that more alkali is needed for complete
solution of the zincic oxide than would be required by the following
reaction,

ZnS04 + 4 KOH = Zn(OK)2 + K2S04 + 2 IT,0,

but this excess was smaller in the case of sodic or ammonic hydrate
than with potassic hydrate; also that the excess of alkali could be
neutralized in such a solution without precipitating zincic hydrate,
until the amount was reached indicated by the reaction given above.
Addition of a large quantity of water precipitated the zincic hydrate
from the solution, even when it contained the excess of alkali. The
effect of dilution and temperature on the solubility were also studied,
but the determination of the composition of the substances formed did
not come within the scheme of his work.

As, therefore, we could find in the chemical literature no analyses
of a zincate, with the exception of Malaguti's ammonic zincate, we
decided to take up the subject, and fortunately began our work with
the sodic zincate, since this proved to be the one of them all which
can be most easily prepared.

* Ann. Chim. Phys., Ser. 3, xii. \ Cliem. News, xlii. 30.

f Comptes Kendus, Ixii. 413.

16 PROCEEDINGS OF THE AMERICAN ACADEMY

The principal results which we have obtained can be briefly sum-
marized as follows. From a solution of zincic oxide in aqueous
sodic hydrate two crystalline sodic zincates can be separated by the
action of alcohol; the principal product, which fuses below 100°, has
the formula Zn30(.Na4H2 . 17 H20, the other, which does not fuse at
30UJ, is (Zn02NaH)2 . 7 H20. Both substances are decomposed into
ziucic oxide and sodic hydrate by water or alcohol. We had hoped
at first to extend our work to other ziucates, but, after a number of
experiments, have abandoned this intention, as with potassic hydrate
no crystalline or definite compound could be obtained, and, although
amnionic hydrate gave a product crystallizing in needles, it was ob-
tained with difficulty, and proved to be of varying composition. We
give, however, at the end of this paper, a brief account of the negative
results of these experiments, and of some others in similar directions.

Preparation of the Sodic Zincates.

The sodic zincates can be prepared by the action of sodic hydrate
on either metallic zinc, zincicJoxide, or zincic hydrate, and the jiroduct
seemed to be the same in every case. It is most convenient, therefore,
to prepare them from zincic oxide, as the metallic zinc dissolves very
slowly, long boiling with aqueous sodic hydrate being necessary to pre-
pare a sufficiently strong solution. On the other hand, we have not
succeeded in preparing the sodic zincates from sodic carbonate, as a
mixture of this substance with zincic oxide showed no loss of weight
even when kept at a white heat for some time. The method adopted
by us consisted in dissolving ziucic oxide in a strong aqueous solution
of sodic hydrate,* with the aid of heat, in a flask, which was usually
closed with a cork fitted with a set of potash bulbs, to prevent the
absorption of carbonic dioxide by the alkaline liquid, but this precau-
tion was not absolutely necessary. The solution, after it had cooled,
was treated with twice or three times its volume of alcohol, and the
mixture, after being thoroughly shaken, allowed to stand securely
corked for about twenty-four hours. During this standing two layers
were formed, a heavy aqueous solution, and a lighter alcoholic liquid.
These were separated, the aqueous solution treated again with alcohol
in the same way, and this extraction with alcohol repeated until the
heavier portion solidified shortly after it had been removed from the
alcohol, which happened usually after the third or fourth extraction.
The crystalline mass thus obtained we have called fusihle sodic ziuc-

* The sodic hydrate used by us had been purified by alcohol.

OF ARTS AND SCIENCES. 17

ate, because it melts below 100°. The second sodic zincate was ob-
tained from the alcoholic washings, which were mixed, and allowed
to stand in a corked flask, when after some time, which varied from
a few hours to several days, white crystals were deposited on the sides
of the flask, which gradually increased iu quantity. The formation of
these crystals did not occur invariably, but they could often be made
to appear by adding more alcohol to the washings. This compound
deposited from the alcoholic washings does not fuse even at 300°, anil
we shall call it infusible sodic zincate.

Analysis of the Fusible Sodic Zincate, Zn306Na4H., . 1 7 II20.

This substance is the principal product of the reaction, forming, as
nearly as we could estimate, over 90 per cent of the total product.
Since both water and alcohol decompose it, a further purification
seemed impossible, and we were obliged to analyze it as it was obtained
by the precipitation with alcohol, simply drying it by pressure between
filter paper under heavy weights.

The method of analysis, which we adopted after trying several
others, consisted in dissolving the substance in dilute sulphuric acid,
and precipitating the zinc by means of a measured quautity of a stand-
ard solution of sodic carbonate. The zinc was weighed as oxide.
The filtrate was acidified with sulphuric acid, evaporated to dryness,
weighed, and then the amount of sodium calculated after subtracting
from the total weight of the sodic sulphate the weight of sodic sulphate
corresponding to the sodic carbonate added.

I. 0.4720 grm. of the substance gave 0.1744 grm. of zincic oxide

and 0.2080 grm. of sodic sulphate.
II. 0.6420 grm. of the substance gave 0.2302 grm. of zincic oxide
and 0.2718 grm. of sodic sulphate. 0.8802 grm. of substance
from the same preparation lost at 240° 0.3920 grm. of water,
and the dried product gave on heating with chromic oxide in
a stream of dry oxygen 0.0342 grm. of water.
III. 0.9682 grm. of the substance gave 0.3370 grm. of zincic oxide
and 0.3992 grm. of sodic sulphate. 0.8770 grm. of substance
from the same preparation lost at 240° 0.4021 grm. of water.

i.

ii.

TIT.

Zinc

29.66

28.7. B

27.93

Sodium

14.28

13.72

13.36

Water

44.:. i

15.88

Constitutional "Water

3.88

VOL. XXIV. (N. S. XVI.)

18 PROCEEDINGS OP THE AMERICAN ACADEMY

Each of these three analyses was made with a sample from an en-
tirely separate preparation, in order to determine whether the substance
had a definite composition, and when it is remembered that it could
not be purified by crystallization, and was dried only imperfectly by
pressing with filter paper, the agreement is as close as could be ex-
pected ; but it is obvious that in a case like this the atomic ratios are
of much more importance than the percentages, and we have accord-
ingly calculated them as follows.

Atomic Ratio of Zinc to Sodium in Fusible Sodic Zincate.

Zinc

Sodium.

T.

1.

1.36

II.

1.

1.35

III.

1.

: 1.35

The agreement between these numbers proves conclusively that the
substance is a definite compound, and that the proportion of zinc to
sodium is as three to four. In trying to calculate a formula for it, we
have felt much doubt in regard to the amount of water of crystalliza-
tion for the reasons given above ; but the one finally adopted by us
gives percentages which agree so well with those found, that it cannot
be far from the truth, especially as the amounts of water found are
about as much higher than those calculated as was to be expected.
The constitutional water, on the other hand, comes very high, but we
do not have much faith in this determination.

Calculated for
Zn306Na4H., 17 H20.

Zinc 28.28

Sodium 13.30

Water 44.24

Constitutional Water 2.60

To prove that the substance contained water, and not alcohol of
crystallization, a combustion of it was made in the usual manner, but
only a mere trace of carbonic dioxide was obtained.

The formula of this substance, Zn30GNa4H2 . 17 II20, agrees, except
in water of crystallization, with that of the substance analyzed by
Malaguti (see page 15), which, written in the same way, becomes
Zn.,Oi;(NH4)4H2 . 9 H20. This, as has been mentioned already, is the
only zincate of which an analysis has been published.

We have made only one analysis of another preparation of this sub-
stance, which is comparable to those given above ; this gave about the

Found.

I.

II.

in.

29.66

28.78

27.93

14.28

13.72

13.36

44.54

45.83

3.88

OF ARTS AND SCIENCES. 19

same percentage of zinc, 29.45, but a much larger amount of sodium,
19.56, which was undoubtedly due to a small quantity of sodic hydrate
left adhering to the crystals, as this was the first preparation we made,
and therefore was not pressed out so thoroughly as the later ones,
after we had become more familiar with the manipulation.

Properties of the Fusible Sodic Zincate.

The substance as precipitated by alcohol from its solution in aqueous
sodic hydrate forms a white mass, made up of radiating crystals, often
of considerable size. It fuses at about 70°, but, owing to the diffi-
culty of drying it, no attempts were made to determine the melting
point accurately. Water decomposes it rapidly and completely, con-
verting it into a white insoluble powder, which gave on analysis the
following result.

o

0.1990 grm. of the substance gave 0.1910 grm. of zincic oxide and
no sodic sulphate.

Calculated for Found.

Zd02H2. ZnO.

Zinc 65.72 80.28 77.03

The substance therefore seems to be principally zincic oxide mixed
with a slight impurity of the hydrate. Alcohol decomposes it in a
similar way, but more slowly, as is shown by the following percentages
obtained from the analyses of two preparations (IV., V.), which we had
attempted to purify by washing with alcohol.

in.

IV.

v.

Zinc

27.93

29.96

33.36

Sodium

13.36

12.27

10.48

III. is one of the analyses already given. The substance for this
analysis was only pressed with filter paper. IV. was a sample which
had been washed three times with common alcohol, while the substance
which had the composition given under V. had been washed four times
with absolute alcohol. These numbers show that the action of the
alcohol consists in gradually removing the sodium as sodic hydrate,
and this point is brought out even more sharply by a comparison of
the atomic ratios.

Ziuc : Sodium.

nr. i. 1.35

IV. 1. 1.16

V. 1. 0.89

20 PROCEEDINGS OF THE AMERICAN ACADEMY

It follows from what has been said on this subject, that care must be
taken not to continue longer than is absolutely necessary the treat-
ment with alcohol during the preparation of the sodic zincate. Al-
though the substance is decomposed by both alcohol and water, it
seems to exist during the process of mauufacture dissolved in a mix-
ture of these solvents in presence of an excess of sodic hydrate, an
observation which is confirmed by that of Prescott (see page 15). It
is insoluble in ether, and absorbs carbonic dioxide rapidly from the
air. At 100° it loses only 12 molecules of its water of crystallization,
as is shown by the following analysis.

0.4720 grm. of the substance analyzed under I., when heated to 100°
in a stream of pure dry air, lost 0.1510 grm. of water.

Calculated for

a loss of 12H20.

Found.

31.23

32.00

Water

Analyses and Properties of the Infusible Sodic Zincate,
(ZuOaNaH),.. 7H20.

This substance, which crystallizes from the alcoholic washings
obtained in the preparation of the fusible compound just described, is
formed in comparatively small quantities. As nearly as we could
estimate, only a few per cent of the total product consisted of it. As
like the fusible compound, it is decomposed by both water and alcohol,
no attempt was made to purify it further, but it was analyzed after
drying at ordinary temperatures by the method already described (see
page 17).

I. 0.7470 grm. of the substance gave 0.3124 grm. of zincic oxide

and 0.2913 grm. of sodic sulphate; 0.4762 grm. of substance
from the same preparation lost at 220° 0.1637 grm. of water.

II. 0.6044 grm. of the substance lost at 300° 0.2066 grm. of water,

and gave 0.2700 grm. of zincic oxide and 0.2355 grm. of
sodic sulphate.

Calculated for Found.

(Zn0,NaH),7H20. I. II.

Zinc 35.32 36.78 35.86

Sodium 12.50 12.62 12.62

Water 34.24 34.37 34.19

The considerable difference between the percentages of zinc is not
greater than was to be expected, when the impossibility of purifying

OF ARTS AND SCIENCES. 21

the .substance is remembered. Each of these analyses was made with
a sample from a separate preparation, and we add the following analy-
sis, made early in the work simply to determine the ratio between the
zinc and sodium, no care having been taken to dry the substance.

III. 0.7025 grm. of slightly moist substance gave 0.2260 grm. of
ziucic oxide and 0.1951 grm. of sodic sulphate.

The percentages derived from this analysis are of course of no
value, but it gives an atomic ratio which agrees with those from the
preceding analyses so closely that there can be no doubt that this
substance possesses a definite composition.

Atomic Ratio of Zinc to Sodium in the Infusible Sodic Zincaie.

Zinc : Sodium.

I. 1.03 1.

II. 1.03 1.

III. 1.02 1.

A combustion of the substance showed that it contained no alcohol.

This substance undoubtedly corresponds to the crystalline body
obtained by Fremy (see page 15), who assigned to it the same ratio
between the zinc and potassium. Laux (see page 14), on the other
hand, ascribed to his crystals, which were soluble in water, the formula
Zn02K2, but stated that upon heating these crystals he obtained an
amorphous powder, which showed the ratio of 1 : 1 between zinc and
potassium ; that is, the same obtained in the crystals by Fremy and us.
We have met with no substance corresponding to the soluble crystals
of Laux.

Properties. — The infusible sodic zincate crystallizes from a .solution
in dilute alcohol containing an excess of sodic hydrate in white needles,
sometimes over a centimeter long, forming loose radiating groups
usually of a conical shape, but occasionally circular or spherical. It
does not melt even at 300°, and is decomposed by alcohol, or water,
if these solvents are free from sodic hydrate, absorbs carbonic dioxide
from the air, but much less rapidly than the fusible compound, and
does not lose the whole of its water of crystallization until heated
above 200°.

"We have not succeeded in finding any other definite compound
among the products of the action of sodic hydrate on zincic oxide,
and, if any other exists, it can be only in very small quantity.

22 PROCEEDINGS OF THE AMERICAN ACADEMY

After we had settled the composition of the sodic zincates, we turned
our attention to the study of other zincates, and also tried some ex-
periments with magnesic oxide. The results of all this work were
negative, but nevertheless we think it best to give a brief statement
of what we have done.

In all our attempts to prepare potassic zincate we encountered the
difficulties already mentioned by Fremy ; in fact, we have not suc-
ceeded in obtaining in any experiment the crystals described by him,
although we have modified the process employed in a number of dif-
ferent ways ; but the solution of zincic oxide in potassic hydrate, when
treated with alcohol according to the method which had yielded such
good results with sodic zincate, gave only amorphous precipitates,
which looked like zincic oxide, but were not wholly free from potas-
sium ; as the absence of crystalline form left us no means by which
we could judge of the purity of this substance, we did not think it
worth analysis.

With amnionic hydrate the results looked more promising at first,
as the solution of zincic hydrate in amnionic hydrate occasionally gave
a small quantity of crystals, after it had been mixed with alcohol and
allowed to stand ; but in this case there was no separation of the liquid
into two lavers. These crystals looked very much like the infusible
sodic zincate described above, but the analyses of four preparations
showed that they had no constant composition, the following percent-
ages being obtained.

I. II. III. IV.

Zinc 48.66 45.90 59.96

Ammonia 7.35 4.58 3.67 5.28

In view of these results, it did not seem worth while to continue the
investigation.

In beginning this research we had hoped, by acting on the sodic
zincate with a cobaltous salt, to obtain Rinman's green, and in this
way throw some light on the composition of this pigment. These
hopes have not been fulfilled, as the zincates were decomposed by
alcohol or water, as already stated, and therefore the action could not
be carried on in solution, and the melted fusible zincate, when treated
with cobaltous chloride, cave only a blackish precipitate with no shade
of green. The same result was'obtained when an ethereal or absolute
alcohol solution of cobaltous chloride was allowed to act on the zincate.
We have also made many attempts to purify Rinman's green, prepared
according to the usual method, in order to fit it for analysis, but none

OF ARTS AND SCIFNCES. 23

of these have succeeded, the substance being decomposed by solutions
of alkalis or acids, iu fact even by a solution of carbonic dioxide
under pressure, which dissolved both zincic and cobaltous carbonates,
and finally left a blackish gray residue.

Although magnesic hydrate does not dissolve in sodic hydrate under
ordinary conditions, we thought that possibly a very strong solution of
sodic hydrate might have some solvent action at its boiling point, and,
upon trying the experiment, obtained long prismatic crystals, which,
however, we found contained only a mere trace of magnesium ( 0.39
per cent), and consisted of the crystallized sodic hydrate recently
obtained by Cripps,* as shown by the following analytical results.

I. 0.5620 grm. of the substance lost 0.2150 grm. of water.
II. 0.4115 grm. of the substance gave 0.4585 grm. of sodic sulphate.

Found.

II.

36.11

The curious point about this observation, which alone makes it
worth recording, is that the solution of sodic hydrate used could not
be made to crystallize before it was treated with magnesic oxide, but
after such treatment crystallized so rapidly that it was impossible to
filter it. The experiment was repeated several times, each time with
the same result. What the cause of this difference in behavior may
be we have been unable to determine, but think that possibly a small
quantity of sodic carbonate in our solution of sodic hydrate may have
prevented the crystallization, which took place as soon as this was
converted into hydrate by the magnesic oxide.

Calculated for

]

(NaOH)3 4 H.,0

i.

Water

37.50

38.26

Sodium

35.93

* Pharm. Jour. Trans , Slt. 3, xiv. 833 (1884).

24 PROCEEDINGS OF THE AMERICAN ACADEMY

III.

STUDIES FROM THE NEWPORT MARINE LABORATORY.

XXI. — A PRELIMINARY ACCOUNT OF THE DEVEL-
OPMENT AND HISTOLOGY OF THE EYES IN THE
LOBSTER.

By G. II. Parker.

Presented by Alexander Agassiz, October 10, 1888.

The following is a brief statement of the results obtained frotn study-
ing the development and histology of the eyes in lobsters. The method
in which the optic nerve appears to terminate is so exceptional, that,
before making a final publication on this subject, it seems desirable to
seek confirmation in the structure of the eyes in other Crustacea. As
this will delay the appearance of the paper, and since iu other direc-
tions definite conclusions have been reached, it seems advisable to
publish now an account of my present conclusions.

The first indication of the optic apparatus in the young lobster is a
pair of ectodermic thickenings on either side, and slightly in front of
where the mouth is to appear. The superficial part of each of these
thickenings gives rise to the retina, and the deep part to the optic
ganglion. The ganglionic portion is cut off from the retinal portion
by the ingrowth of the basement membrane. In certain regions, how-
ever, the basement membrane does not cut the connection between the
retina and ganglion. These primitive connections persist in the adult
as optic nerve fibres.

In the eye of an adult lobster each ommatidium consists of at least
sixteen cells. Directly under each corneal facet are found two flat
lentigeuous cells (corneal hypodermis). Under these are four retino-
phoroe, one for each angle of the corneal facet. The retinophora?
are extremely elongated, and extend from the deep face of the corneal
hypodermis to the basement membrane. From the corneal hypoder-
mis to the spindle the four retinophora? are closely applied to one
another. At the distal end of the spindle they separate, passing
around that structure as fibres. As they approach the basement
membrane they converge slightly, and terminate on the retinal surface

OF ARTS AND SCIENCES.

of that membrane. Under the centre of each ommatidium the base-
ment membrane is considerably thickened, and it is on this thickening
that the four retinophorse terminate.

Each ommatidium has ten pigment cells, — two distal and eight
proximal. The distal cells surround the retinophora; in the region of
the crystalline cones, and from this region they are continued inward
as fibres till they pass through the basement membrane. The eight
proximal pigment cells are closely applied to the spindle, the fibres of
the four retinophorse passing between them. Seven of these are
deeply pigmented ; one is without pigment. The eight cells extend
only a short distance in front of the spindle ; the seven pigment cells
proper are continued inward as large fibres through the "basement mem-
brane. In addition to the sixteen cells just described, each omma-
tidium has two or three irregular cells filled with a pigment, brownish
by transmitted, white by reflected light. These cells envelop the
proximal half of the spindle, and extend to the basement membrane.
The spindles themselves do not reach the basement membrane.

The sixteen cells already described are ectodermic in origin. The
two or three additional cells may be from either an ectodermic or me-
sodermic source, but the evidence thus far gathered points decidedly to
their ectodermic origin.

The basement membrane, as was previously mentioned, has a thick-
ening in it under each ommatidium. Around a given thickening there
are four openings through the membrane. Each opening, however, is
placed between two thickenings, so that in reality only half of each
cluster of four openings belongs to a given thickening. There are
two classes of openings, one with a single small and four large fibres,
and another with one small and three large fibres, passing through.
Each thickening has two of each class accompanying it. Of the
fourteen large and four small fibres passing through the four open-
ings, only one half, or seven large and two small fibres, belong to a
given ommatidium. These represent the seven deep and two super-
ficial pigment cells. After passing through the basement membrane,
these fibrous ends of the pigment cells thicken considerably, and,
having grouped themselves in bundles, pass inward, constantly di-
minishing in calibre, till they reach the optic ganglion. The optic
nerve between the retina and first optic ganglion is composed oi
these fibres bound together by a small amount of connective tissue.
All attempts at isolating any other form of fibres have failed, and it
would therefore seem that the fibres of the optic nerve terminal'
these nine pigment cells.

26 PROCEEDINGS OP THE AMERICAN ACADEMY

Investigations on Light and Heat, made axd published wholly or in pakt with
Appropriation from the Ruiiford Fund.

IV.
ATMOSPHERIC ECONOMY OF SOLAR RADIATION.

By Arthur Skarle.

Presented October 10, 1888.

The terrestrial atmosphere acquires energy from the solar radiation
by direct absorption, by the absorption of terrestrial radiation, and by
conduction from terrestrial solids and liquids. It loses energy in three
corresponding ways ; by radiation into space, by downward radiation,
and by conduction.

As some time must elapse between the acquisition and the loss of
any given amount of energy, the air always contains a certain accumu-
lated store of activity resulting from the solar radiation, and manifested
in warmth, expansion, and movement. It is a general and apparently
well founded belief, the reasons for which need not here be repeated,
that terrestrial temperatures are maintained to a great extent by the
aid of this atmospheric accumulation of energy ; so that a far lower
temperature would prevail in the absence of the air. The hypothesis
which has been current until recently with regard to this protective
action of the atmosphere depended upon a supposed effect of selective
absorption, which has now been largely, if not entirely, disproved by
Langley's experiments. The supposition, indeed, was always some-
what difficult to reconcile with the familiar fact that celestial bodies
appear redder at a small than at a great altitude ; since, so far as the
visible spectrum is concerned, this proved that among the constituents
of the atmosphere there were some, abounding in its lower strata,
which absorbed radiations of small wave-length more readily than the
others. Hence it did not seem probable that the radiation from ter-
restrial substances the temperature of which was far below red heat
would be absorbed by the air with peculiar readiness, and thus pre-
vented from escaping into space. This reasoning, however, could not
be conclusive, and actual experiment was required to overthrow the
assumption that the air was much more transparent to solar than to
terrestrial radiation.

As we are now obliged to abandon this assumption, it is natural to

OF ARTS AND SCIENCES. 27

inquire whether the known phenomena of conduction will suggest a
better explanation of the protective action of the atmosphere than can
be afforded by the observed laws of absorption and radiation. By con-
duction, fluids acquire heat most readily when hot bodies are applied
to their lower surfaces, in consequence of the convection currents thus
established. On the other hand, the application of cold bodies to their
lower surfaces, as it does not originate such currents, withdraws their
heat only by the much slower process of conduction through their own
substance.

Hence, an undisturbed atmosphere will acquire heat more readily
by contact with warm ground than* it loses heat by contact with cold
ground. Part of the heat thus acquired might have been conducted to
adjacent portions of the ground in the absence of the atmosphere, but
another portion would have been directly radiated into space. It now
becomes a question whether the atmosphere thus heated will discharge
its recently acquired energy by radiation into space as readily as the
ground would have done in the absence of an atmosphere.

Without undertaking to decide this question, it will here be assumed
that the heat acquired by the atmosphere from warm ground will not
be radiated as readily as it would have been radiated by the ground
itself. Since it will not be readily lost by conduction, in the absence
of violent agitation of the air, for the reason already given, the hypoth-
esis seems admissible that it tends to accumulate, and to increase the
stock of energy contained in the atmosphere much more efficiently
than can be done by the processes of absorption and radiation.

If we admit the existence of this tendency, we have next to consider
what natural provision can be suggested for checking its effects when
they have attained a certain magnitude ; for it is obvious that they do
not increase indefinitely. If we suppose atmospheric energy to be
manifested only as heat, its accumulation would ultimately be checked
by an increasing radiation from terrestrial solids and liquids ; if mani-
fested only by expansion, it is perhaps possible that portions of the
atmosphere would be driven off into space, carrying away the energy
mechanically ; but a more immediate check is afforded by that portion
of the accumulated energy which displays itself as atmospheric move-
ment. When the winds have attained a certain degree of violence,
they disturb the portions of the air which would otherwise remain
stagnant over the colder parts of the ground, and the process of heat-
ing the atmosphere from beneath gradually ceases to retain sufficient
advantage over that of cooling it from beneath to permit a furtbej
accumulation of energy.

28 PROCEEDINGS OF THE AMERICAN ACADEMY

If a permanent increase should take place in the amouut of solar
radiation, it is clear, upon the present hypothesis, that some time would
elapse before the increasing agitation of the atmosphere put an end to
the increase of its energy. The stock of atmospheric energy in gen-
eral, and that part of it manifested as heat, would thus be permanently
increased. The terrestrial temperature would be raised, tempests
would be more frequent and severe, and the entire atmosphere would
probably occupy more space.

On the other hand, a permanent diminution in solar radiation would
tend to diminish the agitation of the air, and, although the terrestrial
temperature would decline, this loss of temperature would not be so
great as that which would have occurred if the winds had maintained
their previous force. The atmosphere, accordingly, acts as a check
upon extreme variations of heat and cold ; when little heat is received,
it will be better economized than when the supply of heat is excessive,
although particular regions may have, in the former case, a very
severe climate.

The observed association of extreme cold with still air, and the
greater violence of tempests in the heated portions of the world, on the
whole, are among the facts tending to support the hypothesis above
explained.

In the present discussion, the consequences resulting from the com-
pressibility of the atmosphere have thus far been neglected, and what
has been said above would be equally applicable to an atmosphere
wholly incompressible. But in such an atmosphere the distribution of
heat would materially differ from that actually observed. As a con-
vection current rises, the air composing it expands, from the removal
of pressure, and its energy largely ceases to exhibit itself as heat.
Under these circumstances, the solid and liquid particles carried up
with it are reduced in temperature, and made less capable of radiation
into space than before. It may likewise be supposed, indeed, that the
expanded air itself will have its previously small capacity for the dis-
charge of its energy into space still further diminished. The addi-
tional tendency to retain energy, thus suggested, would demand more
consideration if the discharge were effected by any process of the
nature of conduction ; that is, if the outer surface of the atmosphere
were chiefly instrumental in the process. In this case, the expansion
of the ascending air would be a highly important means of delaying
the escape of energy received by conduction from warm ground ;
and an incompressible atmosphere might not in any considerable
degree protect the planet which it surrounded.

OF ARTS AND SCIENCES.

However slightly the expansion of ascending currents may cluck
the escape of energy, it is certain that the subsequent descent of the
air composing them must exhibit more and more of its remaining
energy in the form of heat. This phenomenon is generally recognizul
by recent writers upon meteorology, and there can be little doubt that
it powerfully affects the relative climates of places at different alti-
tudes. The climate of an elevated region is colder than that of the
sea level, because a smaller part of the atmospheric energy appears
there as heat, and a larger part as expansion. Whether this is a suffi-
cient explanation of the observed difference of climate can scarcely be
determined until we have more knowledge with regard to the actual
extent and velocity of the convection currents of the atmosphere.

The considerations above set forth indicate the conclusion, that the
effect of conduction, aided by convection currents, is probably an im-
portant means of maintaining the present terrestrial temperature, as
well as the present distribution of warmth in different latitudes and at
different elevations. Jf this conclusion should hereafter find a more
satisfactory basis in observation and experiment, it would have some
interesting applications to the climate of the larger planets. Their
extensive atmospheres, subjected to a powerful force of gravitation,
may perhaps enable them to economize very efficiently the compara-
tivelysmall quantity of solar radiation which they receive. Another
branch of inquiry connected with the same general subject relates to
the conditions of temperature in different parts of the ocean. The
atmospheric and oceanic currents prevailing in former times are also
frequently discussed by geologists. If the atmospheric convection
currents have the effects here attributed to them, they may help to
account for some of the unexplained phenomena of prehistoric climates,
the evidences of which have remained to the present day.

30

PROCEEDINGS OF THE AMERICAN ACADEMY

V.

THE CRYSTALLINE STRUCTURE OF THE COAHUILA

IRONS.

By Oliver W. Huntington.

«

Presented October 10, 1888.

In a previous paper on the crystalline structure of iron meteorites,*
the author described two cleavage crystals broken from a specimen of
the Butcher meteorite (Coahuila), but, from the compact nature aud
softness of the iron, no further examples of cleavage were at that
time obtained. Recently, however, on examining a large number of
small sawed slabs of the same iron in the Harvard Collection, one slab
was found to be intersected by an angular crack, as shown of actual
size in Fig. 1. On taking the slab in the hand, it was found that the
two portions could be readily separated by a slight pressure, and the

Fig. 1.

surfaces, though considerably oxidized, showed sharply defined crystal
faces extending through the entire thickness of the plate (about six
millimeters) and forming angles of about 132° and 90°.

As this cleavage was so very striking, an attempt was made to
break another slab artificially. It was mounted in a vice for the
purpose, and the projecting portion struck with repeated blows of the

* American Journal of Science, 3d series, vol. xxxii. pp. 281-303.

OP ARTS AND SCIENCES.

31

hammer. The slab broke readily along the jaws of the vice, exhibit-
ing a superb crystalline cleavage, the faces of the crystals having a
most brilliant lustre, almost like antimony, and, though the iron was
soft enough to cut with a knife, yet some of the single cleavage faces
were nine or ten millimeters in extent. These crystal faces usually
formed angles of about 132°, or else right angles, though in some
parts there were sharp projecting points, formed by the meeting of
two rectangular planes with a third set obliquely. Furthermore, all
the faces were most beautifully striated by numerous sets of fine
parallel lines, easily distinguished by the eye, generally making angles
with each other of 127°, 90°, and 53° respectively, a few of the lines
appearing to be parallel to the intersections of the faces.

Fig. 2 shows enlarged one set of crystal faces as they appeared on
the surface of fracture. The planes were absolutely perfect, and lar^e
enough to be readily measured with an application goniometer. Thus

Fig. 2.

Fig. 3.

it was found that the plane B made with C a right angle, B and A an
angle of about 132°, A and C an angle of about 109°; and the small
plane D, set obliquely on the solid angle thus formed, made an angle
of about 125° with B and C, and 164° with A. The plane E and the
one parallel to it were the sawed faces of the slab. All the other
planes forming the surface of fracture were parallel to the ones shewn
in the figure, and they could all be referred to faces of the twin cube,
as shown in Fig. 3, where the corresponding faces are lettered the
same and placed in a parallel position. The striations already men-
tioned appeared most markedly on the face A, that being the largest,
and therefore the one most easily examined.

Fig. 4 shows, on a large scale, this plane referred to the face of
a cube, only the most prominent of the striations being represented,
in order to preserve clearness in the figure. It will be seen by the

32

PROCEEDINGS OF THE AMERICAN ACADEMY

diagram, that all the lines follow the intersections of the different
members of an ordinary cube twinned on all of the trigonal axes, and

'OV\d IVlj

Mesb p^o-Tl
y Y»ounceaJ

Fig. 1.

there were no markings on the face which could not be thus referred.
However, on examining the triangular face D of Fig. 2, it appeared
in marked contrast to the face just described, as all the markings were
parallel to the octahedral edges, thus forming simple triangles.

A large number of specimens of the Butcher iron were afterwards
examined, and they all exhibited the same crystalline cleavage on a
fresh fracture. Moreover, it made no difference in what direction the
slab was broken, the planes always showed a prevalence of the 132°
angle (theoretical angle 131° 48' 37"), with here and there the sharp
projecting points already referred to, but, contrary to expectation, the
simple cube angles seldom appeared, though in some parts they were
evident.

It now appeared important to examine the cleavage of the Sancha
Estate or Saltillo iron (Santa Rosa), which has been considered part
of the same meteorite as the Butcher specimens. For this purpose
Mr. S. C. H. Bailey sent us a wedge-shaped slab which had been
sawed to a thin edge and then broken, giving a surface of fracture
about seventy millimeters long and two millimeters wide in the thick-
est part, diminishing towards the two ends. The crystal faces over
this surface were very small, and appeared to have a grayer color than
those of the other irons. When this surface was examined under a
low power microscope, the crystals were found to be all simple cubes,
looking exactly like a specimen of galena, and, instead of the faces
being striated as in the Butcher irons, they exhibited the little crys-
talline projections and depressions which are so characteristic of some
alloys. However, on the face of the slab there was the suggestion of
two cracks crossing at angles of about 132° and 53°, and an attemnt

OF ARTS AND SCIENCES. 33

was made to break the specimen parallel to one of these cracks, with
the expectation of getting the planes already described in the previous
iron.

When the slab was mounted in the vice and struck with the hammer,
there at once appeared on the surface numerous very fine cracks
parallel to the two directions just mentioned, and at once suffjrestine

l DO C

the two sets of fine parallel lines which are first brought out on an
etched surface of this iron but are obliterated by the continued action
of the acid. Finally, the slab broke, but exhibited only a siucde plane
through the entire extent of the fracture (33 mm. in length by G mm.
in breadth), and, instead of being striated, there was developed a very
striking flaky surface as if little thin layers had resisted the cleavage.
parting at last with irregular edges more or less separated from the
main surface. An attempt was next made to break the slab at ri-ht
angles to the first direction, in order to get other crystal faces; but,
instead of breaking along the edge of the vice, as a Butcher specimen
readily would have done, the Saltillo iron broke again with the same
single cleavage plane, at right angles to the desired direction and
running directly down into the jaws of the vice, entirely regardless of
the way it had been clamped and the blows of the hammer applied.

Every attempt to break the slab gave the same result. It would
only break in the two directions indicated by the first fine cracks, and
corresponding to the angle that a cube face makes with the adjacent
face of a twin cube, and always broke along a single plane exhibiting the
flaky surface already mentioned. The only way of getting a different
fracture was to saw the specimen to a thin edge, and so force it to
break in the desired direction, and then it would present a surfaa
fine cubes wholly different in character from the large striated faces
of the Butcher irons.

Just at this time we received from Ward and Howell a large slab
of a new meteorite found in Allen County, Kentucky, and the whole
character of this iron, including the etched surface, so closely resembled
the Coahuila specimens that we were interested to see whether it would
show any striking cleavage. Ward and Howell kindly Furnished ua
with some small slabs, and, on breaking them in the way already
described, they showed exactly the same characters as the Saltillo iron.
The fine parallel cracks appeared under the first blow- of the hammer,
and then the slab broke regardless of the way it was clamped in only
the two directions at angles of about 132 and ■'>'■'> , presenting a single
cleavage plane with the marked flaky appearance characteristic of the
Saltillo iron.

vol. xxiv. (n. s. xvi.) 3

34 PROCEEDINGS OF THE AMERICAN ACADEMY

Exactly the same result was obtained in breaking slabs of the
Maverick County* iron, and also the iron from Chattooga County,
Georgia,! both of which were described as independent falls, although
closely resembling the Coahuila specimens.

Now it will be borne in mind that the Coahuila irons form a group
by themselves, their etched surfaces appearing so markedly different
from any other known meteorite, and for this reason, together with a
resemblance in composition, they have commonly all been considered
to belong to one fall. But, after examining the structure of the
Butcher irons and that of the Saltillo or Saucha Estate as exhibited
by the cleavage, it is impossible to class them as portions of a single
original mass. And, on the other hand, when we compare in the
same way the Saltillo iron with those of the Allen County, Chattooga
County, and Maverick County meteorites, it seems equally impossible
to believe that they did not at one time form a single meteorite, espe-
cially when they also resemble each other in composition as shown
in the following analyses, though to be sure it is hard to tell how far
such analyses can be depended upon as a means for comparison or
distinction of irons.

% Allen Co.

§ Santa Rosa.

|| Maverick Co.

IT Chattooga Co,

Iron

94.32

95.82

94.90

94.63

Nickel
Cobalt

5.01
trace

3.18)
.35 \

4.97
.21

Sulphur

.34

• . . .

• > • .

....

Phosphorus

.16

.24

....

.21

Carbon

.12

.23

99.95 99.59 10U.U0 99.99

That these masses were found in places so remote from each other
does not seem to preclude their having belonged to one individual,
since the Rochester meteorite was seen to pass over the States of
Kansas, Missouri, Illinois, Indiana, Ohio, and is supposed to have
passed over Pennsylvania and New York, and thence out to sea, drop-
ping fragments in its course. It therefore is possible that at some
remote period an enormous iron meteorite may have passed over the

* Amer. Jour. Sci., 3d series, vol. xxxii. p. 304, October, 1886.

t Ibid., vol. xxxiv. p. 471, December, 1887.

t Ibid., vol. xxxiii. p. 500, 1887.

§ Ibid , 2d series, vol. xix. pp. 160, 101, May, 1855.

|| Ibid., 3d series, vol. xxxii. p. 304, October, 188G.

IT Ibid., vol. xxxiv p. 472, December, 1887.

OF AETS AND SCIENCES. 35

entire breadth of the Uuited States, the main mass reaching Mexico,
but large fragments breaking off and falling during its passage across
the country.

Two irons commonly classed under the general head of" Coaliuila''
have not been included in the above consideration. They are labelled
in the Harvard College collection, " Hacienda de Concepcion," * and
" San Gregorio Iron, Chihuahua, Mexico," f both bearing the signa-
ture of Dr. H. B. Butcher. The specimen of the Hacienda de Con-
cepcion was small, and had been hammered, so that its structure could
not be conclusively studied, but it appeared to be markedly different
from any of the other Mexican irons.

The San Gregorio iron, of which there were numerous specimens, ap-
peared to be made up of very striking octahedral plates, and therefore
must be placed among a very different class of irons from the Coahuila
group, which it has been the purpose of this paper to describe.

* Amer. Jour. Sci., 2il series, vol. xix. p. 1G3, 1855.
t Ibid., 3d series, vol. ii. p. 336, 1871.

86 PROCEEDINGS OP THE AMERICAN ACADEMY

VI.

CONTRIBUTIONS TO AMERICAN BOTANY.

By Skreno Watson.
Presented October 10, 1888.

1. Upon a Collection of Plants made by Dr. E. Palmer, in
1887, about Guaymas, Mexico, at Muleje and Los Ayigeles Bay
in Lower California, and on the Island of San Pedro Martin
in the Gulf of California.

The peninsula of Lower California and that portion of the Mexi-
can mainland which borders the intervening Gulf, though reputed a
sterile land, have always wherever they have been explored yielded
a rich harvest of novelties to the botanical collector. Much therefore
was expected from so keen and careful a collector as Dr. Palmer, when
he undertook to spend a season at Guaymas, and from that point to
explore such other places as might be accessible to him. Though the
season of 1887 proved very unfavorable on account of its dryness,
the result has nevertheless been very satisfactory. Of the 415 native
species collected, 89 species, or more than one fifth, are wholly new,
and many others are of great interest in various respects.

The larger part of the collection was made about Guaymas itself,
which town lies on the eastern side of the Gulf of California, in the
State of Sonora, in lat. 28° N., and 250 miles south of the United
States boundary. It is hemmed in closely by very rocky hills and low
mountains (of 1200 to 1500 feet altitude), intersected by narrow val-
leys. The artificially watered gardens, with their irrigating ditches and
brush fences, protecting and favoring the growth of numerous native
plants, the rocky islands in the harbor, and the valleys and mountains
around, wore all alike searched. Dr. Palmer remained here from the
middle of dune to the middle of November, during which time there
were only occasional slight showers, which commenced in August.
The sp< cies obtained here numbered 283, of which 40 were also

OF ARTS AND SCIENCES. :,7

found in other localities.* Muleje, upon the western side of the Gulf,
90 miles from Guaymas, is described as in a dry, barren, and moun-
tainous region, where except in the very short rainy season the only
green vegetation to be seen is along the banks of a small crei k.
This place was visited early in dune, and again late in December.
Of the 49 species collected here, 24 occurred elsewhere, mostly at
Guaymas. Los Angeles Bay, also on the peninsula, about 200 miles
northwest from Guaymas, was visited at an unusually favorable time,
after a rain which was the first that had fallen in twenty-two months
and when vegetation was in full bloom. The surrounding country is
very mountainous, some of the ridges having an altitude of 2,000 to
3,000 feet. About a month was spent here (from November 22 to
December 20) and 112 species were collected, of which 23 had been
found previously. The remaining station was the island of San Pedro
Martin, lying about 80 miles northwest from Guaymas. This island
has a circumference of about 4i- miles and an altitude of 1,200 feet,
and is exceedingly rough and rocky, intersected by canons and largely
covered with guano. A very few small fig trees were found, but the
only useful wood is furnished by the Ccreus Pringlei, which forms
almost a forest over the summit. During an eight days' stay (Octo-
ber 24 to November 5) only 19 species were collected, of which seven
were not peculiar to the island.

The characteristics of the flora of the region bordering the Gulf of
California, so far as shown by this collection, are for the most part
those common to the flora of the whole arid region of the interior,
from southeastern California, Arizona, and New Mexico southward
into Mexico, distinct in a great measure from that of California
proper on the one side, and that of the Gulf States on the other
Nearly or quite two thirds of the sjiecies range northward beyond the
Mexican boundary. In the mountains about Guaymas we find a <
siderable number that are identical with or allied to species thai have
recently been collected by Pringle and Palmer in the mountains
Chihuahua. We have here also probably the northern limit on the
Pacific coast of the tropical or subtropical genera Rkizophora, fftema-
toxyloiu Portlandia, Oitharexylum, Pedilanthus, Ficus, etc. The pro

* Cultivated and introduced plants are not included in the numbers al
piven. Sixteen species of the collection are considered of this character, viz. :
Oligomeria glaucescens, Portulaca i, Gossypium herbaceum, Trip)

foliuta, Melilotus parviflorus, Tamarindus Tndicus, Capsicum ca . C. an-

mtum, Crescentia alata, B'tu vulgaris, Panicum sanguinale, Sorghum II
El usine Mgyptiaca, E. Indica, Eragrostis major, and Lolium

38 PROCEEDINGS OF THE AMERICAN ACADEMY

portion in which the several orders are represented in the collection
is somewhat remarkable. Of the 415 species, one fourth are equally
divided between the Gru mi 'nece (50) and the Compositce (50). Another
fourth includes only the four orders Leguminosce (44), Euphorbiacece
(32), Malvaceae (17), and Solanacece (15). These are followed by the
Ngctaginaceoe (15), Convolvidacece (13), Asclepiadacece (10). and 53 other
orders with still fewer species. The important orders lianunculacece,
Rosacece, Saxi'fragacece, Umbelliferce, Ericaceae, Ouptdiferce, Coniferce,
and Orchidacece are wholly unrepresented. Excluding the Cypera-
ccce and Graminece, there are only five endogenous species in the entire
collection.

For the determination of the species the Cyperaceae were referred
to Dr. N. L. Britton, Curator of the Torrey Herbarium, the Gra-
mine<e to Dr. George Vasey, Botanist of the Agricultural Department
at Washington, and the Filices to Prof. Daniel C. Eaton. Special
acknowledgment is also due to Prof. Daniel Oliver, of the Kew
Herbarium, for suggestions respecting the relations of a few of the
more difficult species in other orders.

Cocculus diversifolius, DC. Guaymas. (60.)

Argemone albiflora, Hornem. Differing from A. Mexicana in
the narrower, less distinctly mottled, and less deeply sinuate leaves,
more naked peduncles, white flowers, narrower buds and capsules, the
capsules narrowed upward from near the base, and the much smaller
seeds. " Cardo "; the dried juice used as a remedy f.»r inflamed eyes.
Muleje. (7.)

Argf.mone Mfxicana, var. (?), with obovate-spatulate petals, few
stamens, narrow buds, narrow capsule a'tenuate to a distinct style,
and irregularly pitted seeds much smaller than usual. About Guay-
mas. (105.)

Eschscholtzia c^espitosa, Benth., var., with more coarsely di-
vided leaves. Rocky ridges, Los Angeles Bay. (590.)

Cardamine Palmeri. Annual, erect, smooth and somewhat glau-
cous, branching above, 1 or 2 feet high : leaves simple, thin, round-
ovate, cordate at base, obtuse, usually rather deeply sinuate-toothed
or -lobed, the larger 2 inches long; petioles a little shorter: flowers
purplish, 3 or 4 lines long: pods slender, 1 to \}, inches long by g of
a line wide, beaked with a slender style, ascending on petioles 3 or 4
lines long: seeds small. — Resembling C. cordifolia of the Rocky
Mountains. Under bushes near salt water, Muleje. (421.)

OF ARTS AND SCIENCES. 39

Cardamine angelorum. Animal, glabrous and glaucous, Bub-
erect, branching above, leafy to the base of the short raceme : lea\ ( -
pinnately divided, the straight divaricate lobes (5 or 6 pairs) linear,
obtuse, entire or rarely with a lateral lobe at base, 9 lines long or
less: flowe.s purplish, 2 lines long: pods slender, 1| inches long by
| of a line wide, ascending ; style short : seeds small. — On stony
ridges near Los Angeles Bay. (594.)

Nastcrtium (?) laxl'm. Annual, glabrous, slender, ascending,
the elongated branches very lax : leaves pinnately divided, the nar-
rowly linear straight divaricate segments (2 or 3 pairs) entire, acute,
8 lines long or less : raceme lax, elongated, the soon distant flowers
small (!■£ lines), purplish; on slender spreading pedicels: immature
pods reflexed, narrow, 3 to 6 lines long, about equalling the pedicels,
the cells about G-seeded ; style stout. — The purplish flowers and
immature pods leave the genus somewhat doubtful. It may be a
Thelypodium, though differing in habit from other species of that
genus. On sandy plains near Los Angeles Bay. (598.)

Sisymbrium canescens. Nutt. Mountains, Los Angeles Bay. ( 579.)

Lf.pidium Palmeri. Annual, decumbent or procumbent, branch-
ing from the base, hispid throughout: leaves pinnately divided, the
narrow lobes with usually a single tooth on the upper margin :
racemes sessile, short, rather dense ; pedicels short, broad and flat :
flowers minute, apetalous : stamens 2 : pods thinly pubescent, round-
ovate with short blunt wings, lh to nearly 2 lines long, on spreading
or somewhat recurved pedicels ; stigma sessile. — At Los Angeles
Bay, in the shade of rocks and bushes. (560.)

Lyrocarpa Coulteri, Hook. & Am. The typical form. At
Los Angeles Bay, among rocks and under bushes. (540.)

Cleome tenuis. Annual, erect, slender, branching, 1 to 3 feet
high or more, glabrous or the petioles pubescent : leaves 5-foliolate,
the leaflets narrowly to linear-oblanceolate, obtuse, £ to 1 inch long.
equalling the petioles: racemes loose, few-flowered; flowers yellow,
2 lines long, on slender pedicels: pods sessile, narrow, spreading, 1 to
H inches long, including the style: seeds transversely rugose. -At
Guaymas, on hillsides. (214.)

Wisltzexia Palmeri, Gray. Three to five feet high ; leaves
mostly 3-foliolate. At Guaymas. ("4.)

Atamisquea emarginata, Miers. A stiff brittle shrub, about 6
feet high, with white flowers. Muleje. (28.)

Oligomers glaucescevs. Camb. At Los Angeles Day, in saline
soil. (547.)

CO PROCEEDINGS OF THE AMERICAN ACADEMY

Ionidium polygal/EFOLIum, Vent. Among rocks in dried creek-
beds ; Guaymas. (253.)

Amoreuxia palmatifida, DC. "Sayas"; the roots have the
taste of the parsnip and carrot, and are eaten by the Yaqui Indians,
and made into a preserve by the Mexicans. Guaymas. (176.)

Krameria canf.scens, Gray, var. ; slightly pubescent, and leaves
small and distant. Guaymas and Los Angeles Bay. (151, 565.)

Krameria parvifolia, Benth. The typical form. At Guaymas.
(248.)

Frankenia Palmeri, "Watson. A compact shrub, 1 to 3 feet
high. At Los Angeles Bay. (541.)

Drymaria crassifolia, Benth. At Los Angeles Bay. (570.)

Portulaca oleracea, Linn. " Verdulaga ; common all over
the country." Guaymas. (116.)

Anoda pentaschista, Gray. Guaymas. (661.)

Horsfordia NEWBERRYr, Gray. Flowers orange. On rocky
hills at Guaymas. (314.)

Horsfordia rotundifolia. Annual, erect, slender, 2 feet high
or more, finely and softly pubescent ; branches short : leaves round-
cordate, obtuse or retuse, creuate, an inch broad or usually much less,
white-velutinous beneath, greener above : inflorescence elongated and
paniculate, nearly naked ; peduncles slender, 4 to 9 lines long, jointed
above the middle : flowers orange to sulphur-yellow, 9 lines broad ;
calyx 2 lines long, the lobes acute or short-acuminate : carpels 6 to 8,
4 lines long. 2-ovuled, 1 -seeded, the thin sides of the cell strongly
reticulate and becoming clathrate, the erect wings lanceolate: seed
pubescent. — On an island in the harbor of Guaymas. (351.)

Horsfordia Palmeri. Erect, 3 to 6 feet high, stout and branch-
ing, densely stellate-pubesrent throughout : leaves lanceolate, obtusish,
cordate at base, entire or subcrenate, 3 inches long or less : flowers ax-
illary, solitary or clustered, on slender peduncles 6 or 8 lines long and
jointed above the middle: calyx 3 lines broad ; corolla " light pink."
turning blue, an inch broad: carpels 10 to 12, 4 lines long, 1-seeded,
the basal portion strongly' reticulated, the wing obliquely deltoid:
seed puberulent. — On broken stony ground near Los Angeles Bay.
(558.)

Sida carpinifolta, Linn. Guaymas. (316.)

Abutilon crispum, Don. Guaymas. (95.)

Abutilon incanum, Don. (A. Texense, Gray.) Guaymas.
(104, 288.)

Abutilon Palmeri, Gray. Guaymas. (239.)

OF ARTS AND SCIENCES. 41

Abutilon aurantiacuji, Watson. San Pedro "Martin Island.
(401.)

Abutilon scabrum. Erect, 3 feet high, scabrous throughout
with very fine stellate pubescence, scarcely at all canescent : leaves
thin, cordate to cordate-lanceolate, acuminate, rather coarsely and
unequally toothed, 3 inches long or less, on slender petioles : flowers
bright orange, an inch broad, axillary, solitary, on slender peduncles
exceeding the petioles and jointed near the top ; calyx as long as the
petals, the ovate acute lobes apparently cordate, enlarged in fruit :
carpels 8, shortly acuminate, very finely pubescent. 5 lines long. —
Near A. Jacquini and A. Berlandieri, which are softly pubescent and
not at all scabrous. Among bushes in ravines and by garden fences
near Guaymas. (97, 662.)

Spfleralcea ambigua, Gray. In mountain canons near Los An-
geles Bay (537), and a small-flowered form from S. Pedro Martin
Island (405). Also what is probably a form of this species from
garden fences at Guaymas. (90.)

SpiJuERALCEa axillaris. Stellate-pubescent ; stems long and
slender : leaves ovate to lanceolate, acutish, subcuneate at base, finely
crenate, 2^ inches long or less, the petioles half as long: flowers
yellowish, an inch broad, in short axillary solitary or clustered ra-
cemes or panicles ; calyx 3 lines long, the lobes acuminate: carpels
about 15, 1-seeded, 1^ lines long, obtuse, the base very strongly
reticulated, the broader empty cell thin and scarious. — Muleje; com-
mon in shady places. (17.)

Spil/ERALCEa (?), sp. Imperfect specimens from S. Pedro Martin
Island. (404.)

Kosteletzkya Coulteri, Gray? A rather stout branching plant.
3 feet high, with deeply 5-7-lobed leaves, and " white " flowers drying
to greenish yellow. Coulter's original specimens are much reduced,
with only the upper leaves deeply lobed, the flowers apparently yel-
low, and the authors few. The present plant is probably the fully
developed form. In waste places about Guaymas. (236 )

Hibiscus denudatus, Benth. Los Angeles Bay. (523.)

Hibiscus Coulteri, Gray. River-bed, Guaymas. (668, 66

Gosstpium iiERiiACEU.u, Linn. In garden fences at Guaymas ;
the plants 8 feet high and said to be forty years old. (11" I

Gossypium Davidsoni, Kell. A loose sbruh :; to ."» feet high, the
larger leaves 3-lobed ; bractlets ample, deeply laciniate, persisteni
about the ovate capsule, which is half an inch long and 3- or rarely 1-
celled; seeds 6 or 8 in each cell, in two rows, not at all lanat ■ the

42 PROCEEDINGS OP THE AMERICAN ACADEMY

testa showing only a rudimentary lanate condition. The flowers are
sulphur-yellow with a brown spot at the base of each petal. This
appears to be a true Gossypium, notwithstanding the naked seeds.
In mountain ravines near Guaymas. (244.)

Hermannia pauciflora, Watson. Among rocks in the moun-
tains near Guaymas. (227.)

Melochia tomentosa, Linn. In fences about Guaymas. (148 )

Melochia speciosa. Closely resembling M. pyramidata in foli-
age and fruit, but more shrubby and branching, less virgate, and finely
stellate-pubescent throughout : flowers in loose axillary and terminal
cymes, large, rose-color turning purple ; sepals 3 lines long ; petals 8
lines long, spreading : capsule finely pubescent, shortly stipitate, re-
sembling that of M. pyramidata, but the margins of the cells rounded
(not acuminate). — In mountain canons about Guaymas. (650.)

Waltheria detonsa, Gray. Guaymas. (251.)

Atenia Berlandieri, Watson, var. ? Less pubescent, etc. ; the
same as 83 Palmer (1885) from Batopilas. In ravines near Guay-
mas. (243.)

Ayenia filiformis. Annual, erect, the slender stems finely pubes-
cent : leaves thin and glabrous or nearly so, from narrowly ovate and
obtuse to narrowly lanceolate and acute, 2£ inches long or less, on
slender petioles, subtruncate at base, coarsely toothed : pedicels clus-
tered or solitary in the axils : sepals narrow, a line long ; petals pur-
ple, the blade rhombic, attenuate to the filiform claw, with a deep
sinus at the apex embracing the filament, and a filiform dorsal append-
age : fruit 2 lines in diameter, beset with numerous slender green
processes: seeds surrounded by 2 or 3 stout irregular ruga?, the inter-
vals more or less pitted. — Resembling and probably including some
western forms that have been referred to A. pusilla, but certainly
distinct from the typical form of that species as it is found in the West
Indies and Brazil. In the shade of rocks high in the mountains about
Guaymas. (292.)

Bunchosia parvifolia. A shrub 3 or 4 feet high, intricately
much branched : leaves sparsely appressed-hairy, round-ovate to ovate-
oblong, acute, mostly obtuse at base, glandless, 15 lines long or less,
on very short petioles: racemes few-flowered, pubescent: calyx 10-
glandular ; petals pink, 2 lines long : ovary glabrous ; styles nearly
distinct : drupe red, 5 to 6 lines in diameter. — On a high rocky peak
near Guaymas. (336.)

Galphimia angustifolia, Benth., var. oblongifolia, Gray.
Mountains about Guaymas. (217.)

OF ARTS AND SCIENCES. 43

Echinofterys Lappula, Juss. A shrub 3 to 5 feet hi»h, with
an abundance of golden yellow flowers. The plant differs from Jus-
sieu's description only in the acute or acutish instead of acuminate
leaves, and perhaps in the indumentum of the ovary. The cocci of
the mature fruit, which was not known by him, are 2 lines long, ovate,
acute, densely pubescent, carinate with a row of stout subsninulose
pubescent bristles, and the whole dorsal surface covered more or less
densely with similar bristles. The seed is oblong-ovate, with a thin
membranous testa, the cotyledons curved and somewhat unequal, and
the radicle short and straight. Rocky mesas near Guaymas. (181.)
Hir.ea macroptera, DC. " Matanene " ; the leaves used in poul-
tices tor bruises and sores. Muleje. (19.)

Janusia Californica, Benth. In mountain ravines about Guay-
mas. (263.)

Tribultjs grandiflorus, Benth. & Hook. "Mai de ojos " ; the
pollen is said to be injurious to the eyes. Common about Guaymas.
(177, 225.)

Tribulus Californicus, Watson. Guaymas. (G51.)
Tribultjs maximus, Linn., var. ? Scarcely or not at all hispid,
the pubescence short and mostly appressed : flowers smaller ; calyx
(1^ lines long) usually deciduous: fruit smaller; nutlets about 8,
more strongly tuberculate. — About Guaymas. (1^7.)

Fagoxia Californica, Benth. Los Angeles Bay. (546.)
Guaiacum Coulteri, Gray, — at least as to Thurber's specimens
from Sonora. It is probably also the G. parvifolium of Planchon
(Gray, PI. Wright, 1. 29; G. Planchoni, Gray, Proc. Am. Acad. 22.
306, where this name is proposed on account of a supposed earlier
" G. parvifolium, Engelm.," by which G. angustifolium, Engelm., must
have been meant), of which there is a fragment in Herb. Gray. A
shrub or small tree, 10 to 15 feet high and 15 inches in diameter, on
hills and mesas about Guaymas, where it is known as " guayacan."
The flowers are dark purplish blue and very fragrant ; filaments
naked. (113.)

Viscainoa geniculata, Greene. A weak shrub, 5 to 8 feet
high: flowers white. The genus is nearly allied to Cliitonia. Muleje
and ( riiavmas. (27.)

Eroimum Tkxanum, Gray. Lo-; Angeles liny. ("'66.)
Triphasia trifoliata, DC. Cultivated at Guaymas. (64
Burskra microphylla. Gray. " Torote bianco"; 6 to 12 feet
high ; the bark used for tanning. Common on the mesas about
Guaymas. ('1 63.)

44 PROCEEDINGS OF THE AMERICAN ACADEMY

Bursera Hixdsiana, Benth. & Hook. A scrubby bush, 8 feet
high, the leaves simple or 3-foliolate upon the same plant: seed black
at base, covered above with a hard salmon-colored arillus. The bark
is much used for tanning, giving a reddish brown color. Los Angeles
Bay. (572.)

Bursera laxiflora. A small tree with a trunk from 6 inches
to a foot in diameter : leaves pinnate, sparingly pubescent, the pubes-
cence short and mostly spreading ; rhaehis narrowly winged ; leaflets
2 to 4 pairs, thin, ovate to oblong, mostly obtuse, entire or usually
with 1 to 4 (rarely 6) blunt teeth, 3 to 9 lines long : peduncles axil-
lary, solitary, very slender, 1 to 4 inches long. 1-3-flowererl. and often
with 1 to 3 simple foliacpous bracts; pedicels about 6 lines long:
sepals narrow, nearly a line long : fruit obovate, narrowed at base,
somewhat compressed, 2-valved, 4 lines long. — " Copal " ; in ravines
about Guaymas. (280.)

Bursera fubescens. A tree 10 to 15 feet high, with stout trunk
and branches : leaves finely pubescent, pinnate, the rhaehis not winged,
1 or 2 inches long; leaflets 4 or 5 pairs, rather thick, oblong-elliptic
and entire, 2 to 4 lines long, obtuse, the terminal one cuneate-obovate
and crenately toothed or lobed : flowers and fruit unknown. — Evi-
dently a Bursera, but unlike any known species in its foliage. "Wood
brittle and bark deciduous in broad thin flakes. Plains and stony
ridges about Los Angeles Bay. The bark is exported to Europe for
tanning. (585.)

Maytenus phtllanthoides, Benth. A low bush or small tree
in sandy and alkaline soils. Muleje and Guaymas. (30.)

Ziztphus ltcioides, Gray, var. canescens, Gray. A small tree,
10 feet high. In sandy bottoms about Muleje and Guaymas. (29.)

Ziztphus Sonorensis. A small tree with a trunk 5 to 8 feet
high and G to 10 inches in diameter, with smooth grayish bark and a
compact top, the short rigid branches very thorny: leaves nearly
glabrous, thin, ovate to ovate-elliptical, obtuse or retuse, subcuneate
to slightly cordate at base, entire or slightly toothed, 1 or 2 inches
long with petioles 2 lines lon^: panicles short, finely pubescent, axil-
lary and terminal: fruit subglobose and reddish brown when ripe,
about 5 lines long, edible, the pedicel 2 lines long : putamen oblong-
obovate. 2-celled, 2-seeded, 4 lines long. — Common about Guaymas,
near brackish water. (124, 659.)

Colubrina GLABRA. A compact shrub, 5 or G feet high, unarmed,
glabrous or very nearly so: stipules subulate, rigid, deciduous; leaves
mostly fascicled, oblong-ovate to elliptical or obovate, obtuse or retuse,

OF ARTS AND SCIENCES. 4.",

subcuneate at base upon a sbort slender petiole, 2 to 10 lines loi^ :
pedicels very slender, 2 to 4 (or in fruit 6 to 8) lines long: p<
linear, equalling the filaments: fruit depressed-globose, 2 or 3 lines
broad. — Ravines about Guaymas. (200.)

Serjania (Phacococcus) Palmkri. A stout climber, pubes-
cent throughout with short soft spreading hairs : leaves twice pinnate,
the lowest pinna? o-foliolate, the next 3-foliolate, and the upper of a
simple leaflet ; leaflets entire or rarely toothed, mostly oblong, acute
or acutish, cuneate at base, 2 to G or sometimes 12 lines long: pedun-
cles solitary, equalling or exceeding the leaves, the narrow panicles
much longer: sepals finely pubescent, a line long: filaments pubes-
cent: fruit 7 to 11 lines long, finely pubescent, the cells subglobose,
2 lines or more broad, faintly reticulate, thin and crustaceous, narrowly
crested on the back, the wing broadly dilated to a rounded base (3 or
4 lines broad) ; dissepimeut narrow : seed attached below the middle,
globose and filling the cavity. — Iu foliage and pubescence closely re-
sembling S. sphenocarpa, as described, but the fruit is very different ;
flowers very fragrant. Common about Guaymas. (61.)

Cardiospermum Halicacabum, Linn. Guaymas. (126.)

Paullinia Soxorexsis. A compact shrub 2 or 3 feet high, with
flexuous branches, very finely tomentose when young: leaves glabrous,
piunately 5-foliolate, the rhachis narrowly winged above; leaflets
ovate to oblong-ovate, 3-o-toothed or -lobed, the lower often temate,
cuneate at base or shortly winged-petiolate, 6 to 8 lines long : pedun-
cles about equalling the leaves (2 inches long or less), few-flowered :
sepals oblong, obtuse, a line long or more : capsule depressed-globose,
abruptly stipitate, finely tomentose, 3 lines long, on a pedicel 1 or 2
lines long. — On rocks in the mountains about Guaymas. (238.)

Dodox.ea VISCOSA, Linn. Guaymas. (290.)

Lunxrs Arizonicus, Watson. Sandy bottoms at Los Angeles
Bay. (595.)

JLupixus, sp. near L. concinnus, Ag. Mountains near Los Ang<
Bay. (58';.)

Melilotus parviflorus, Desv. Muleje. (12.)

IlosvCKiA rigida, Benth. Mountains: Los Angeles Bay. (582 )

Hosackia strigosa, Nutt. Abroad-leaved and Bmall-flowered
form. Rocky ridges at Los Angeles Bay. (602.)

Dalea megacarp a, Watson. Leaflets often glabrous on the up-
per surface; 2 or 3 feet high, with very fragrant yellow llov.
Rocky ravines near Los Angeles Bay. (522.)

DALEA PRINGLEI, Gray. Guaymas. (GOT.)

46 PROCEEDINGS OF THE AMERICAN ACADEMY

Dalea mollis, Benth. Guaymas and Los Angeles Bay. (333,
512, and 550.)

Dalea Parryi, Gray. Guaymas and Los Angeles Bay. (315,
666.)

Dalea Emoryi, Gray. Leaves mostly oblong-obovate, occasion-
ally obtusely toothed. A rough bush, 2 to 5 feet high ; used as a
yellow dye by the Indians. Los Angeles Bay. (542.)

Indigofera Anil, Linn. Used by the Indians for dyeing palm
leaves. Guaymas. (102.)

Indigofera mucronata, Spreng. Mountains about Guaymas, in
the shade. (296.)

Tephrosia tenella, Gray. Mountains about Guaymas. (249.)

Tephrosia Palmeri. A tall slender much branched perennial,
very finely appressed-pubescent throughout : stipules narrow, 1 or 2
lines long ; leaflets 3 to 6 pairs, linear to linear-oblong, 8 to 20 lines
long, obtuse or acutish, the rhachis 1| to. 3 inches long: peduncles
and racemes elongated : flowers scattered on short pedicels, light
yellow, 9 lines long; calyx 3 lines long, the teeth as long as the
broad tube ; banner pubescent : pod narrow, nearly straight, 1|- to 1\
inches long or more, pubescent, spreading. — In the mountains near
Guaymas. (246.)

Tephrosia constricta. Annual, slender, erect, glabrous or
nearly so, a foot high : stipules linear-subulate; leaves 1-foliolate, the
iijaflet nearly sessile, linear, 1 to 2^ inches long by a line wide or
les%; flowers solitaiy, opposite or lateral to the axil ; calyx-lobes
acurrAaate, exceeding the campanulate tube ; corolla reddish and
white,l^$ines long: pod glabrous, linear, an inch long, spreading or
reflexed upon the short pedicel, 6-10-seeded, with a thin septum be-
tween the seeds and usually a partial one at the stricture : seeds
quadrate-oblong, sharply constricted in the middle. — In habit much
resembling T. tenuis of India. The seed is very peculiar. In ravines
in the mountains about Guaymas, sometimes very abundant. (222.
261.)

Diphysa sknnoides, Benth. On stony mesas; Guaymas. (198.)

Cracca Edwardsii, Gray. Mountain slopes ; Guaymas. (218.)

Sesbania macrocarpa, Muhl., var. ficta. Wings orange ; ban-
ner strongly marked with dark brown. " Bequilla " ; the older roots
were formerly used for tinder by the Yaqui Indians. Guaymas.
(286.)

Astragalus Nuttallianus, Gray. Los Angeles Bay. (571.)

Nissolia Scbottii, Gray. Guaymas. (170.)

OP ARTS AND SCIENCES.

Desmodium (Heteroloma) scopulorum. Annual, erect,
blanched, 2 feet high, the stem scabrous with very short stiff un-
cinate hairs; leaves 3-foliolate, the stipules and stipels setaceous;
petioles ^ to 1 inch long ; leaflets thin, linear-lanceolate, 1 to 2£
inches long by 1 to 5 lines broad, acutish, pubescent on the margin
and nerves : racemes terminal and axillary, very slender, loosely
flowered: flowers whitish, minute (scarcely 1^ lines long), on spread-
in"1 pedicels, solitary or in pairs : fruit 4-lobed, only the terminal lobe
usually fertile ; lobes 4 or 5 lines long, nearly semicircular, the
ventral side straight, finely uncinate-bristly on the margin, the sides
nearly glabrous. — On rocky ledges in the mountains about Guay-
mas. (258.)

Phaseolus filiformis, Benth. Mountains about Guaymas. (250.)

Phaseolus atropurpureus, DC., var. sebiceus, Gray. Muleje
and Guaymas. (8.)

Rhynchosia phaseoloides, DC. Seeds wholly red. Muleje. (1.)

C^esalpinia gracilis, Benth. A shrub 3 to G feet high, with
bijugate leaves, 10 stamens, and pods (1 to 1\ inches long by 9 lines
broad) mostly 1 -seeded : seeds 5 lines in diameter. — Hills and moun-
tains about Guaymas. (147, 147^.)

GffiSALPlNTA (Pomaria) Palmeri. An unarmed glabrous shrub
8 to 10 feet high: jugae 1 to 4 pairs and an odd one. on a slender
rhachis from \ to 3 inches long ; leaflets 2 or 3 pairs, oblong-elliptical,
obtuse, 1 to 3 lines long: racemes terminal or axillary, sessile, shcjt,
few-flowered, glabrous or subpubescent : calyx-tube oblique; jtfbea
nearly equal, obtuse, 3 lines long; petals brownish orange, gpvered
more or less thickly on the back and claw with white rabstipitate
glands: filaments and ovary pubescent: pod sessile, thin-coriaceous,
glandular-dotted, falcate, H inches long by nearly & broad, attenuate
downward, usually 2-seeded: seeds 3 lines broad.— Guaymas. (70.)

ILematoxylon boreale, AVatson. " Brazil " ; used as a dye, and
the young wood as a remedy for jaundice. Hillsides and stony ravin.-.
about Guaymas. (125.)

HOFFMANSEGOtA MICROPHYLLA, Torr., and Var. GLABBA. On

stony ridges at Los Angeles Bay. (543.)

Cassia Covesii, Gray. " Ojacen " ; used in infusion for colds and

fevers. Guaymas and Los Angeles Bay. (313, 575.)
Cassia nictitans, Linn. About Guaymas. (242.)
Tamauindus Indica, Linn. From three small trees in abandoned

gardens at Guaymas, said to have been planted many yean ago by

Chinamen from Manilla. (56.)

48 PROCEEDINGS OF THE AMERICAN ACADEMY

Prosopis (?) heterophylla, Benth. Without flowers or fruit,
but the peculiar foliage leaves the species unmistakable. A slender
tree 10 to 15 feet high, with two or three drooping branches and
thinly laminated bark, which cleaves off in large pieces. Guaymas.
(628.)

Prosopis (Algarobia) articulata. An upright shrub 8 to 12
feet high, somewhat finely pubescent, armed with straight stipular
spines |- inch long or less : pinnae a single pair on a rhachis 2 or 3
lines long; leaflets 6 to 15 pairs, narrowly oblong, somewhat reticu-
lated, H to 3 lines long: flowers greenish yellow, pedicellate in loose
short-pedunculate spikes an inch or two long : petals distinct, villous
at the apex, nearly 1^ lines long: anthers glanduliferous : ovary
stipitate, pubescent ; pod glabrous, thin-coriaceous, compressed, moni-
liform, 5-12-jointed, the joints about 2^ lines broad. — On stony
mesas and foot-hills about Guaymas. (197.)

Prosopis (Algarobia) Palmeri. A loosely branching shrub 4
feet high, finely pubescent, armed with straight stipular spines an inch
long or less : pinnae a single pair on a very short rhachis ; leaflets 3
to 6 pairs, oblong-elliptic, 2 lines long or less : flowers minutely silky,
sessile in a loose short-pedunculate spike an inch long; corolla tabular
and somewhat angled, deeply 5-toothed, not villous within, 2 lines
long: style and filaments twice longer, yellow, the anthers bearing a
large deciduous gland : ovary tomentose, stipitate, straight, many-
ovuled. — Without fruit, but there can be little doubt as to the genus,
though the petals are connate to near the top and without viHosit}'.
The flowers are said to be of a bright golden yellow and very
fragrant. Muleje. (2.)

Desjianthus virgatus, Benth. Shrubby, about 4 feet high.
Guaymas. (86.)

Desmanthus Jamesii, Torr. & Gray, var. Glabrous; the pods
3 to nearly 5 inches long; said to be a shrub. Common at Guaymas.
(641, 642.)

Mimosa laxiflora, Benth. A shrub 5 to 8 feet high ; flowers
rose-color turning to white. About Guaymas. (149, 169.)

Acacia Wrightii, Benth. A compact shrub 5 or 6 feet high,
armed with stout recurved spines. This is occasionally the case in
more eastern specimens, though usually they are wholly unarmed. In
a deep canon, Los Angeles Bay. (534.)

Acacia flexicaulis, Benth. A rigid shrub, 5 feet high. Deep
canons at Los Angeles Bay. (548.)

Acacia filicina, Willd. Guaymas, in mountain ravines. (25 1. )

OF ARTS AND SCIENCES. 49

Acacia cochliacantha, Humb. & Bonpl. A small tree, 10 feet
high and 2 inches thick, with fragrant flowers, and straight or curved
spines 3 to 5 inches long. A decoction of the thorns is used as a
remedy for vesical inflammation, etc. On mesas and hillsides at
Guaymas. (101.)

Acacia Farnesiana, Willd. " Binorama" ; the pods are used for
dyeing black and for making ink, and an ointment is made of the
dowers for the cure of headache. About Guaymas. (305.)

Lysiloma microphylla, Benth. A rather stiff shrub, 6 feet
high. In mountain canons at Guaymas. (640.)

Calliandra Coulteri, Watson. A loose shrub, 5 feet high.
Mountains about Guaymas. (297.)

Calliandra eriophylla, Benth. A stiff shrub, in high moun-
tain ravines. Guaymas. (293.)

Pithecolobium (Ortholobium) Sonor^:. A shrub or small
tree (15 or 20 feet high), armed with short stout recurved stipular
spines, the foliage, inflorescence, aud younger branches canescent with
very short spreading pubescence: pinna? 1 or 2 pairs on a short (1
or 2 lines) or very short rhachis ; leaflets 10 to 15 pairs, oblong-
elliptical, about a line long: peduncles mostly solitary (1 to 3) in the
axils, 6 lines long or less; heads loose: flowers white, fragrant, finely
pubescent, nearly 2 lines long : pods rather thin, short-stipitate, flat,
straight, dehiscent, puberulent, 2 to 4 inches long by 6 to 9 lines
wide, 3-6-seeded. — "Una de gato"; wood very hard and taking a
fine polish. Common at Guaymas, and at Loreto and Muleje. (58.)

Rhizophora Mangle, Linn. On a low island in Guaymas har-
bor. (342.)

Jusslea octonervis, Lam. Guaymas. (257.)

CExotfiera cjespitosa, Nutt. On stony ridges near Los Angeles
Bay. (582.)

CExothera (Spflerostigma) angelorum. Annual, from nearly
acauleseent to a foot high or more ; stems slender, glabrous : lea
more or less finely pubescent, narrow, pinnatifid nearly to the mid-
nerve, the linear divaricate lobes mostly entire: flowers axillary,
bright yellow, about 9 lines broad: capsule very slender ( 1 to \). inches
long or more, by half a line wide), becoming more or less curved. —
Of the (E. bistorta group. Abundant in sandy valleys and on ridg< -
near Los Angeles Bay. (519.)

Oenothera cardiophylla, Torr. A form with very Bmall flow-
ers. S. Pedro Martin Island and Los Angeles Bay. ( 103, 520.)

GaURA PARVIFLORA. Dougl. At Mllleje. (11.)

VOL. XXIV. (N. S. XVI.) 4

50 PROCEEDINGS OF THE AMERICAN ACADEMY

Petalonyx linearis, Greene. On S. Pedro Martin Island and
at Los Angeles Bay. (JH-)

Sympetaleia rupestris, Gray. (Loasella rupestris, Baill.) With
the habit of Eucnide : leaves round-cordate, sinuately toothed and
rather obscurely 5-7-lobed, 3 inches broad or less : corolla cylindrical,
4 to 6 lines long, yellow below, dark green above, the erect lobes
broadly oblong, the tube appendaged within near the base by a dense
ring of haii*s ; calyx-lobes narrowly oblong, shorter than the tube of
the corolla : stamens very numerous, covering the upper half of the
tube, the uppermost filaments stout, green and erect, the lower shorter
and slender: capsule subglobose, truncate, 3 lines long, opening by 5
valves at the summit: seeds minute, linear-oblong. — Doubtless the
species from Guaymas described by Baillon. The corolla-lobes are
imbricate and the anthers 1-celled as in S. aurea, which differs espe-
cially in its yellow corolla with narrow tube and the lobes of the
spreading limb rounded. In the shade of rocks near the shore on
islands in Guaymas harbor, " and at Los Angeles Bay." (325.)

Eucnide cordata, Kell. Described as 5 feet high, reclining
upon rocks, with white flowers. Mountains about Guaymas and Los
Angeles Bay. (341.)

Mentzelia multiflora, Nutt. Los Angeles Bay. (591.)

Mentzelia adherens, Benth. Summit of S. Pedro Martin
Island and at Los Angeles Bay. " Commonly called pega-pega ; it
sticks to everything." (402.)

Passiflora fcetida, Linn. Fruit edible. Guaymas. (91.)

Cucurbita cordata. Root long and large : pubescent through-
out with a dense rigid strigose pubescence, mostly reflexed except on
the leaves : leaves broadly cordate in outline, 2 to 4 inches wide, 3-5-
lobed nearly to the middle with narrow sinuses, the lobes coarsely and
obtusely toothed ; tendrils short, glandular-tipped : flowers solitary, on
peduncles If inches long, the corolla 2 inches long: ovary pubescent,
subglobose : fruit globose, 3 inches in diameter, smooth, green with
about 10 whitish longitudinal lines: seeds white, 4 or 5 lines long. —
The fruit resembles that of C. palmata. Sandy plains near Los
Angeles Bay. (584.)

Apodantiiera (?) Palmeri. Stems prostrate, from large long
white roots, glabrous or slightly scabrous on the angles : leaves thin,
rather thinly strigose-scabrous on both sides, round-cordate in outline,
If; to 24- inches broad, 5-lobed to the middle or deeper, with open
sinuses, the lobes acutely toothed ; tendrils mostly bifid: male flowers
racemose on long slender pedicels ; calyx-lobes linear, 3 lines long,

OP ARTS AND SCIENCES. 51

equalling the tube; petals 15 lines long; anthers much curved:
fertile flowers on slender peduncles 1 to 14- inches long: ovary oblong,
thinly pubescent ; fruit H inches long, broadly ellipsoidal, abruptly
attenuate at base, very shortly beaked, smooth, with green marking ;
rind thin; placentas 4 (or sometimes 5?), distinct, pulpy, many-
seeded: seeds brown with whitish margin, 3 lines long. — Apparently
dioecious ; differing from the genus in the number and distinctness of
the placentas. Fruit edible, with the taste nearly of muskmelon,
ripening in September. Plains about Guaymas. (282.)

Maximowiczia Sonor.e. Climbing, glabrous throughout, the
large root projecting above ground : leaves 4 inches broad or less,
twice 3-cleft nearly to the base, with broad sinuses, the lobes coarsely
sinuate-toothed : male flowers racemose, short-pedicellate, the calyx-
tube cylindrical, 3 lines long; petals pubescent, villous within, bifid:
fertile flowers on peduncles 2 or 3 lines long: ovary ovate, long-atten-
uate above; fruit ovate, abruptly stout-beaked, "amber-colored," 1]
to 1J inches long, smooth with a thick fleshy rind; placentas many-
(about 14-) seeded: seeds "covered with a red pulp," compressed,
rough-coated excepting the smooth margin, obovate or oblong-obovate
with a broad base, 3 lines long or more. — This species much resem-
bles M. Lindheimeri, but the leaves are more dissected and the ovary
and fruit more attenuate above, and the seeds though decidedly tur-
gid in the young fruit become compressed and are peculiar in their
generally very rugose-tuberculate surface. The green fruit is de-
scribed as having about 10 longitudinal rows of white dots. A decoc-
tion of the root is used as a cathartic. About Guaymas, and known
as " guarequi." (233.)

Echinopepon ixsularts. Annual, finely and rather softly pubes-
cent, or glabrate : petioles not pilose; leaves broadly cordate, 11 to 8
inches broad, 5-lobed nearly to the middle, the short triangular lobes
acute, finely and acutely toothed, the sinuses acute : male flowers in
narrow panicles mostly exceeding the leaves ; corolla 5 -cleft, 1' to 2
lines broad ; anthers sigmoid, horizontal ; pistillate flowers solitary in
the same axils, on slender peduncles 3 or 4 lines long : ovary obliquely
oblong-ovate, subpubescent, echinate, attenuate and naked above;
fruit subjjlobose after the falling of the beak, about I lines in di-
ameter, covered with stiff straight spines \ lint' long, firm and Bpongy
within, 1-celled, 1-seeded: seed ascending, smooth, oblong-obovate,
subcompressed with a broad flat base, 2 lines long. dark. — The ovary
appears to have an upper cell with an abortive' ovule. Common on
S. Pedro Martin Island. (409.)

52 PROCEEDINGS OF THE AMERICAN ACADEMY

Echinopepon Palmeri. Annual ; stems glabrous : petioles not
pilose ; leaves thin, very minutely scabrous above, nearly glabrous
beneath, broadly cordate, 1 to 2 inches broad, 5-lobed to beyond the
middle with rounded sinuses, the subrhombic lobes acute and some-
what setosely apiculate, sparingly denticulate : male flowers in narrow
panicles mostly exceeding the leaves ; corolla 5-lobed to below the
middle, 1| to 2 lines broad ; anthers sigmoid ; pistillate flowers soli-
tary in the same axils, on filiform peduncles 3 or 4 lines long ; ovary
very oblique, glabrous, the naked slender beak longer than the
echinate base : fruit without the beak about 2j lines long, covered
with stout straight spines :| line long, 1 -celled, 1-seeded : seed ascend-
ing, compressed, 2 lines long. — In the shade of overhanging rocks in
the mountains about Guaymas. (304.) This and the preceding
species, with E. Bigelovii, E. parviflora, and E. minima, form a well
marked group in Naudin's genus Ec/unopepon, which genus should in
my opinion be restored.

Cereus Pringlei, Watson. " Cardon " ; forming a forest on the
summit of S. Pedro Martin Island. The dead wood is much used for
fuel and other purposes, and the seemly fruit is an article of food. It
attains an average height of 25 feet (sometimes reaching 35 feet),
with a circumference below the branches of 6 or 7 feet. (418.)

Opuntia ? Of the 0. proh'fera g'oup, but the specimens are

without flowers or fruit. About 4 feet high, much branched and
very spiny, the joints of the specimens 1 to 1| inches long, with the
areohe of the tubercles terminated above by a dense tuft of spiny
bristles 2 lines long ; spines 3 to 7, vaginate, the longer 9 to 12 lines
long. — " Choyer"; common on S. Pedro Martin Island. (419.)

Trianthema monogyna, Linn. Glabrous and stems red : sepals
very shortly cuspidate : stamens 10 : crest of the capsule 2-parted. —
Common in some gardens at Guaymas. (155.)

Trianthema monogyna, Linn., var. (?) Glabrous and green
throughout or pubescent : sepals long-cuspidate : stamens usually 5 :
concave crest of the capsule entire. — Apparently distinct from the
last, but the species as represented in the Gray herbarium is very
variable. Very common in gardens at Guaymas. (153, 156.)

Mollugo vkrticillata, Linn. In the mountains about Guay-
mas. (255.)

Mollugo Cerviana, Ser. Common in sand and on high gravelly
mesas, Guaymas. (184.)

Portlandia ptebosperma. A shrub or small tree. 2 to 10 feet
high : leaves thin, rhombic-ovate to oblong, acute or subacuminate,

OF ARTS AND SCIENCES. 53

cuneate at base, glabrous or usually pubescent on the margin and
rather short slender petiole, 2 or 3 inches long; stipules broadly tri-
angular: peduncles solitary in the axils, bracteolate, G to 9 lines Ion-:
calyx-lobes very narrow, 6 to 9 lines long; corolla white, funnel-fun,,.
2| inches long, the tube pubescent within, the limb nearly two inches
broad; filaments 6, pilose only near the base, the anthers (8 lines
long) equalling the corolla: capsule oblong, angled, 9 lines long:
seeds ascending, very thin and broadly winged, 4 lines long. — Pecu-
liar in the thin leaves and very broadly winged seeds : flowers large
and handsome, abundant, pendent, and very fragrant. The capsule
is tardily dehiscent septicidally, as in other species. In deep canons
near Guaymas. (298.)

Randia Thurberi. An upright shrub, 6 or 8 feet high, armed
with slender spines, the young branches canescent with fine spreading
pubescence : leaves obovate, attenuate to a short winged petiole, obtuse
or retuse, 6 to 15 lines long, finely pubescent or glabrate : calyx less
than a line long : fruit globose or ellipsoidal, axillary, sessile or nearly
so, 8 or 9 lines long, the pericarp thin-crustaceous : seeds numerous
(about 20) in black pulp. — " Papachi " ; the fruit eaten by the
Indians. In ravines about Guaymas; first collected by Thurber on
the hills between Rayon and Ures, Oct., 1851. (229.)

Randia obcordata. An upright bush, 6 or 8 feet high, armed
with stout spines and very leafy: leaves obovate or mostly obcordate,
attenuate to a very short petiole, finely puberulent or glabrate, 9 lines
long or less, much exceeding the internodes : calyx deciduous: fruit
sessile in the axils, globose, 3 or 4 lines in diameter, black, 2-1-
seeded. — On high gravelly mesas near Guaymas. (6 18.)

Genipa echinocarpa, Gray. A thorny shrub, 8 feet high ;
flowers " orange-yellow," fragrant. About Guaymas. (106.)

Galium stellatum, Kell. Mountain-sides, Los Angeles Bay.
(525, pistillate; 526, staminate.)

Hofmeisteria CRAssiFOLiA. Suffruticose (?) with brittle branch"
glahrous : leaves alternate, very fleshy, linear, once or twice pinnatifid
with 1 to 3 pairs of linear divisions, 1 or 2 inches long: peduncles
long, naked, mostly solitary ; involucre broadly campanulate, of very
numerous linear acuminate bracts; receptacle broad, strongly eon
naked: flowers very numerous ; corolla white, very narrow. I . lines
long: style-branches elongated, much exserted : pappus of 5 scabrous
setae nearly equalling the corolla and .'5 or 4 intermediate on each
side, short, unequal, and rigid. — In crevices of rocks in the high
m untains about Guaymas; flowers very fragrant. (309.)

54 PROCEEDINGS OP THE AMERICAN ACADEMY

Hofmeisteria pubescens. Herbaceous, rather fleshy, pubescent
with short spreading somewhat scurfy haiis, the stout branches oppo-
site or in threes : leaves alternate, broadly ovate in outline, 3-parted
with segments deeply lobed or toothed, 6 or 8 lines long or less, the
stout petiole much longer: peduncles solitary, elongated (6 to 10
inches long), bracteate with several scattered very narrowly linear
bracts ; heads campanulate, very many-flowered, ^ inch high, with
numerous linear setaceously acuminate bracts and a convex naked
receptacle : corolla white, very narrow, the cylindrical tliroat as long
as and little broader than the tube : pappus of 2 hispidulous seta?
longer than the corolla and 2 broad irregularly fimbriate paleae. —
At Muleje, in rounded masses on hillsides and stony ridges. (422.)

Malperia* tenuis. Slender, a foot high with ascending branches,
slightly scabrous : leaves narrowly linear, entire or occasionally with
a few teeth, subacuminate, 1 or 2 inches long or less : heads ^ inch
high, narrowed at base, the short outer bracts descending upon the
peduncle: corolla (apparently white) very slender, rather contracted
at the throat : the short stigmas scarcely at all exserted : achenes
slightly hispid on the angles. — Abundant on stony ridges near Los
Angeles Bay. (567.)

Eupatorium sagittatum, Gray. Growing 8 feet high, over
bushes and hedges ; flowers lilac. Guaymas. (87.)

Brickellia Coulteri, Gray. Guaymas, by garden fences. (62.)

Brickellia floribunda, Gray, var., with the inflorescence re-
duced to a terminal sessile compound cyme, and the achenes more
pubescent. Shrubby, 3 to 5 feet high, under overhanging rocks on
an island in Guaymas harbor. (322 ; 323, with entire leaves.)

Gutierrezia Euthamue, Ton*. & Gray. Los Angeles Bay.
(533.)

Aplopapptjs spinulosus, DC. Foothills ; Los Angeles Bay.
(539.)

Bigelovia diffusa, Gray. Compact, 3 feet high. In saline
localities near the beach at Guaymas. (212, 212.V.)

* MALPERIA, a new genus of Arjerateie. He.nl several-flowered. Invo-
lucre narrowly turbinate, of several rows of very unequal thin narrow nerved
bracts. Receptacle naked, flat. Corolla narrow, the throat not dilated, the
short lobes erect. Anthers with ovate obtuse appendages. Style-branches
short, filiform, truncate, included. Achenes pentagonal, slender. Pappus of 3
hispidulous setae as long as the corolla, wing-margined at base, and 3 or 4 broad
truncate erose paleae. — An erect annual, with narrow sessile and mostly entire
leaves, and the heads solitary or subcorymbose on short peduncles. Near
Hofmeisteria, but very different in habit.

OF ARTS AND SCIENCES. 55

Bigelovia venkta, Gray, var. (?), with small heads and the leaves
mostly entire, somewhat pubescent. On an island in Guaymas har-
bor. (334.)

Psilactis Coulteei, Gray. In waste places at Guaymas. (99.)

Aster frutescens. Shrubby " with very compact rounded top,"
with gray stem and branches, the herbaceous portion glandular-
scabrous: leaves oblong-obovate to oblanceolate, an inch long or less,
the lower attenuate to a narrow base, the upper sessile by a broader
or somewhat clasping or even subcordate base, coarsely spinulose-
dentate, rather rigid and conspicuously reticulate, sparingly glandular-
scabrous : heads solitary and terminal on the leafy branchlets, large
(5 or 6 lines high) ; bracts of the involucre loosely imbricated, linear,
narrowly acuminate, the green tips somewhat spreading; rays "lilac,"
about 20, rather short ; achenes canescent with subappressed hairs,
nerved, subcompressed ; pappus whitish. — A species of peculiar habit.
most nearly related to the section Me gala strum, but the heads smaller
and not naked-pedunculate. On a stony ridge near Los Angeles Bay.
(513.)

Baccharis sarothroides, Gray. "Yerba del pasmo " ; the
twigs are used as a remedy for toothache. Rarely over 2 feet high.
San Pedro Martin Island. (414, 415.)

Pelucha* trifida. A much branched and leafy shrub, about 3
feet high, canescent with a thin fine close tomentum : leaves linear
with a broader obtusely 3-toothed or usually 3-cleft summit, G to 9
lines long: heads about 2 lines high, mostly 10-15-flowered : achene
hispid. — On San Pedro Martin Island, in crevices of rocks near the
sea. (407.)

Lagascea decipiens, Hemsl. A compact shrub, 4 feet high.
Mountain ravines above Guaymas. (256.)

Franseria ambrosioides, Cav. At Muleje and Guaymas. (21.)

Franseria ilicifolia, Gray. Stems herbaceous, li feet high.
Ravines near Los Angeles Bay. (578.)

* PELUCHA, a new genus of Plucheinece. Beads discoid, rather few-

flowered, homogamous, the flowers all perfect or some Bterile. [nvolucre of two
rows of nearly equal linear herbaceous bract-. Receptacle naked. Corolla
slender, slightly dilated upward, very shortly toothed. Anthers very narrowly
sagittate at base. Style-branches narrow, or the Btyle Bometimes undid
Achene small ; pappus rigid, scabrous. — Shrubby, very heavy — nt< d, toraen
tuloso, with alternate 3-fid leaves and small heads in terminal cymes. Closelj
allied to Blumea and Pluchea. The specimens .-ire not in very good condition,
but a careful study of the material leaves little doubt of the distinctneM and
tribal position of the genus.

56 PROCEEDINGS OF THE AMERICAN ACADEMY

Franseria dtjmosa, Nutt. Common on foothills about Los Angeles
Bay, growing H to 2 feet high. (559, 564.)

Franseria tenuifolia, Gray. ''Istafiate"; a remedy for liver
complaints, used by filling the shoes with its leaves. Muleje. (10.)

Eclipta alba, Hassk. High in the mountains above Guaymas.
(220.)

Vigciera laciniata, Gray. Compact, suffrutescent. 2 feet high.
Abundant on the mountains about Los Angeles Bay. (529 )

Viguiera Parishii, Greene, — which may be a variety of V.
deltoidea. A form with more numerous heads. About 3 feet
high, forming large bunches in the canons above Los Angeles Bay.
(530 )

Helianthus annuus. Muleje. (6.)

Encelia farinosa, Gray. The stems exude a resin which is used
as incense in the churches. Ravines about Guaymas. (111.)

Verbesina (Pterophyton) Palmeri. Sufi'i uticose, forming
large bunches 3 or 4 feet high, scabrous throughout ; stems not
winged : leaves opposite, the upper alternate, triangular or rhombic-
ovate, more or less abruptly contracted to a broadly winged auriculately
clasping petiole, acute, repandly few-toothed or entire, 3 inches long
or less : heads loosely panicled, on slender peduncles, about h inch
high ; inner bracts thin, erect, linear, acute or acuminate, the outer
herbaceous, spatulate-oblong, obtuse, somewhat spreading: rays about
10, neutral, 4 or 5 lines long: acheues obovate, pubescent or some
only glandular-scabrous, the thin wing ciliate ; awns 2, erect or
spreading, equalling the corolla. — In canons above Los Angeles
Bay. (528.)

Leptosyne parthenioides, Gray. With the foliage of Coreo-
carpus parthenioides, Benth., as figured in Bot. Sulphur, but the
achenes mostly narrow and nearly wingless, the lobed margin only
rarely distinctly developed. L. heterocarpa is with little doubt a form
of this species with the lobed margin of the achene in most cases
prominently developed. Near a waterfall high in the mountains
above Guaymas. (299.)

Leptosyne parthenioides, var. dissecta. (L. dissecta, Gray;
Acoma dissecta, Benth. in Bot. Sulphur, t. 17.) With the more dis-
sected foliage of Bentham's figure, the achenes rather broider than in
the last, the margin either entire or lobed. It would seem that the
three species that have been described should probably be considered
as forms of one, with very polymorphous achenes. — Shady mountain-
sides near Los Angeles Bay. (6G0.)

OF ARTS AND SCIENCES. 57

Laphamia (?), sp. Foliage only of a shrubby composite, 2 i
high, much resembling L. insularis, but the pubescence different
Common over the summit of San Pedro Martin Island. ( LOG

Perityle California, Benth. Abundant about Los Angeles
Bay and around the old French fort at Guaymas. (562.)

Perityle deltoidea. Annual, low and slender, minutely [albes-
cent : leaves ovate-deltoid, coarsely and somewhat irregularly to. abed,
an inch long or less, exceeding tbe petioles: peduncles an inch long or
less ; heads 2 or 3 lines high, the bracts linear, acuminate : flowers
yellow; ligule linear-oblong, 2 lines long: style-appendages linear,
flattened, obtuse, not hispidulous: acbenes a line long, shortly ciliate,
the crown of very short bristles ; awn solitary, as long as the disk-
corolla, strongly hispidulous. — With the habit of P. microglossa.
Common in the shade on mountains and hillsides about Los Angeles
Bay. (568.)

Perityle Palmeki. Of the P. Jcptoglossa group; a low peren-
nial (?) with herbaceous succulent brittle stems, gland ular-villous when
young, becoming finely pubescent or subglabrate below : leaves thin,
broadly ovate, acute, subcordate at base, very acutely lobed and
toothed, the petiole equalling or exceeding the blade (1 or 2 inches
long): peduncles H to 3 inches long; beads 4 or 5 lines long, the
bracts very narrow, acuminate: flowers golden yellow ; ray 1 lines
long : acbenes strongly ciliate, a line long, tbe prominent crown deeply
fimbriate, the hispidulous awn equalling the disk-corolla. — In crevices
of bare but shaded rocks ; Guaymas. (308.)

Riddellia Cooperi, Gray. Mountains about Los Angeles Bay.
(538.)

Palafoxia linearis, Lag. Sandy plains ; Los Angeles Bay.
(581.)

Porophyllum gracile, Benth. Hillsides near Los Angeles
Bay. {b')b.)

Porophylluii Seemanni, Schultz Bip.. var., with larger heads of
white flowers. "Mara villa"; used for headache. Crevices <>l ex-
posed rocks high in the mountains about Guaymas. (-"'.*.)

POROPHYLLUM CRASSIFOLIUM. SuffYuticose, much branched and
compact: leaves somewhat fleshy, linear, attenuate al base, gland-
tipped and often setosely apiculate, \ to 1 inch loug or often less:
heads terminal on the short stout brauchleu : bracts broad, obtuse, 8
or 4 lines long, the yellowish flowers nearly twice longer: ach
lines long. — Apparently near P. tridentatum, Benth., from Magdalena
Bay, but the leaves are wholly entire and the flowers are much Ion

58 PROCEEDINGS OP THE AMERICAN ACADEMY

than the involucre. Muleje, on rocky hills and on sandy beaches
near salt water; also collected by Dr. Palmer in 1870 on Carmen
Island. (423.)

Nicolletia Edwakdsit, Gray. Annual; ray white with a red-
dish stripe on the outer side. Common on sandy plains and stony
ridges at Los Angeles Bay. (4G'J.)

Dtsodia porophtlloides, Gray. Mountain-sides, Los Angeles
Bay. (531.)

Pectis prostrata, Cav. In gardens at Guaymas. (145.)

Pectis angustifolia, Torr. The awnless form. Ravines near
the beach; Los Angeles Bay. (G57.)

Pectis (Pectothrix) Palmeri. Closely resembling P. papposa,
but the solitary setas at the base of the leaves more slender, the
peduncles longer (1 or 2 inches), the longer hairs upon the achenes
straight and slightly thickened upward (not at all uncinate or capi-
tate), and the pappus merely scabrous (not barbellate). — Hillsides
near Guaymas. (652, 653.) What may perhaps be a form of this
with the pappus reduced to an irregular subfimbriate margin was col-
lected on a rocky island in Guaymas harbor. (655.)

Pectis Coulteri, Gray. Low and diffuse, somewhat viscid-
pubescent, with leaves an inch long or less. — In gardens at Guaymas
and on islands in the harbor. (143, 654.)

Pectis punctata, Jacq. In sandy ravines and high upon the
mountains about Guaymas. (224. 656.)

Triciioptilidji incisum, Gray. Stony ridges at Los Angeles
Bay. (563.)

Bebbia juncea, Greene. Flowers yellow, very fragrant. Sandy
ravines at Muleje and Guaymas. ('^•)

Peucephyllum Schottii, Gray. A shrub, 3 or 4 feet high.
Stony ridges at Los Angeles Bay. (587.)

Senecio Lemmoni, Gray. Mountain-sides near Los Angeles Bay.
(577.)

Perezia Palmeri. Tall and branching, green and minutely
glandular-pubescent throughout: leaves thin, oblong-lanceolate, acu-
minate, sessile and cordate at base, very acutely dentate and denticu-
late, 6 inches long by 2 wide or smaller: heads few on the branchlets,
pedunculate, ^ inch long, many-flowered; involucre campanulate with
thin acute bracts, the lower ovate and subacuminate : flowers "lilac":
achenes glandular-puberulent. — To be grouped with P. patens and
P. carp/to/epis. On mountainsides near Los Angeles Bay ; rare.
(527.)

OF ARTS AND SCIENCES. 59

Tkixis angustifolia, DC, var. latiuscdla, Gray. A very
compact shrub. 2 or 3 feet high, common upon the summit of S.
Pedro Martin Island. (408.)

Lobelia splendens, Willi Near a waterfall in the mountains
above Guaymas. (301.)

Jacquinia pungens, Gray. "San Juanico " ; a small evergreen
tree, 12 to 15 feet high, the wood useless even for fuel. The globose
fruit is 8 or 9 lines in diameter, the seeds several, oblong-peltate, im-
bedded or enclosed in the resinous-waxy placenta. The flowers are
used by the Indians to give a durable yellow color to the palm-leaves
used in making baskets, etc., and they are also strung like beads and
worn for ornament. About Guaymas. (G9.)

Sideroxylon leucopiiyllum. A small tree, 5 to 8 feet high
and sometimes a foot in diameter : leaves finely white-tomentose on
both sides, oblong-elliptic or -lanceolate, obtuse or acutish, subcuneate
at base, short-petiolate, 2 inches long or less: flowers rather crowded
in the axils on short pedicels, 5-merous ; calyx tomentose, campanu-
late ; corolla greenish yellow, 2\ lines long : staminodia petaloid, en-
tire, broadly lanceolate, equalling the corolla: ovary densely villous.
— The fruit was not collected. The wood is hard, burning with
much smoke. In deep canons near Los Angeles Pay. (516.)

Vallesia dichotoma, Ruiz & Pavon. A large evergreen bush
with white fruit, which is eaten by children, and its juice is used for
inflammation of the eyes. This specific name is adopted as the oldest
in the genus, in preference to the later V. glabra of Link. The spe-
cies so named that is found in Florida and the West Indies appears to
have longer pedicels, a larger calyx, a longer corolla with longer nar-
rowly oblong lobes, and the leaves oblong rather than lanceolate. — In
garden hedges and sandy ravines about Guaymas and Muleje. (32.)

Haplopiiyton cimicidum, A. DC. High in the mountains about
Guaymas. (228.)

Philibertia linearis, Gray, var. heterophylla, I rray. Flow-
ers creamy white; fruit 3 inches long. At Muleje and Guaymas. (5.)

Philibertia Pavoni, Ilemsl. Climbing. 5 to 0 feet high : lea
very white-tomentose beneath; flowers white. The same as :; l<>
Palmer, 1886. In ravines at Guaymas. (195.)

Asclepias subulata, Decaisne. " Vumete"; til'' juice an active
emetic. Common in dry arroyos about Guaymas. (■><.)

Asclepias albicans. Erect, 4 feet high, white-puberulent : lea

verticillate in threes, very narrowly linear, mostly decid B : umb-h

short-pedunculate, many-flowered, pube at, the pedicels half an inch

60 PROCEEDINGS OF THE AMERICAN ACADEMY

long : lobes of the greenish corolla tinged with brown, narrowly ob-
long, acute, nearly 3 lines long ; column short ; hoods yellowish,
shorter than the anthers, somewhat thickened dorsally and bluntly
beaked, the sides subquadrate and the nearly straight subulate spur
scarcely exserted : follicle erect on the erect pedicel, 4 inches long,
smooth. — In rocky ravines near Los Angeles Bay. (588.)

Metastelma Pkinglei, Gray, var. (?), with longer and acuter
calyx-lobes and the filiform lobes of the crown somewhat shorter ;
flowers yellow. — Hillsides near Guaymas. (626.)

Metastelma albiflora. Glabrous or nearly so ; stems twining,
4 feet long : leaves petiolate, narrowly oblong to linear, short-acumi-
nate, 6 to 10 lines long: flowers white, in sessile fascicles, the pedi-
cels very short: calyx-lobes narrowly linear, acuminate ; corolla a half
longer (nearly 2 lines), densely villous within, the lobes narrowly
lanceolate ; crown of 5 very narrow and attenuate scales attached to
the very short column and equalling the distinct corneous cuneate-
obcordate anthers ; appendage of the anthers oblong. — Exposed
cliffs near Guaymas. (223.)

Pattalias* Palmeri. Stems slender, several from a rather thick
branching rootstalk, slightly puberulent or glabrate, a foot long or
more : leaves very narrow, 2 inches long or less, sessile, acute : flow-
ers (2 to 6) yellow, on pedicels 1 or 2 lines long or less : corolla 2
lines long, twice longer than the calyx: lobes of corona ovate, attached
to the column at the base of the anthers, lanceolate, acute, half the
length of the stout conical beak : follicle 4 inches long by 3 lines
broad, narrowed at base and long-attenuate above. — Near a sandy
beach at Muleje. (424). — A second species of this genus is P.
angustifolius, a Sonora plant doubtfully referred by Dr. Torrey in
the Mexican Boundary Report to Metastelma, and more recently
by Dr. Gray to the extra-tropical South American genus Melinia.
It is of similar habit, but has petiolate leaves, a longer calyx, the
crown at the base of the column, the anther-tips much more conspicu-
ous, and the beak of the stigma narrow and columnar.

* PATTALIAS, a new genus of the Asclepiadacece, of the Metastelma group.
Calyx 5-parted, the lohes narrow, acuminate, eglandular. Corolla becoming
opcn-campanulate, 5-parted, narrowly convolute in the bud, naked, the lobes
lanceolate. Crown of 5 distinct fleshy entire lobes inserted upon the short
column, exceeding the short anthers. Stigma surmounted by a prominent
conical or columnar entire beak, exceeding the membranous erect tips of the
anthers. Follicles smooth, slender. — Low twining herbaceous perennials, with
narrowly linear opposite leaves, and small flowers shortly pedicellate in an axil-
lary nearly sessile umbel.

OF ARTS AND SCIENCES. CI

Rothrockia cordifolia, Gray. In shady ravines, etc., about
Guaymas. (213.)

Himantostemma Pringlei, Gray. In an orange grove at Guay-
mas. (318.)

Marsdenia edulis. Stem frutescent, with corky bark when old,
climbing: leaves rather thick, uudulate, ovate-lanceolate, acuminate,
rounded at base, glabrous, 1 to 4 inches long, petiolate : flowers in a
nearly sessile simple or compound umbel, the calyx and short pedicels
pubescent: calyx-lobes ovate, acutish ; corolla cream-color, rather
fleshy, 2£ lines long, cleft to the middle, the throat very villous: lobes
of the crown subulate-acuminate, a little shorter than the tall conical
bifid summit of the stigma : fruit ovate, obtuse, smooth, 3 inches
long; seeds suborbicular, 3 lines long. — "Talajote"; the green
fruit is eaten. On sandy saline mesas near salt water, Guaymas.
(150, 658).

Gilia (Eugilta) Palmeri. Annual, erect and branched, some-
what pubescent with floccose tomentum especially in the axils, a foot
high : leaves alternate, very narrowly linear, entire, pungent, 1 \
inches long or less: pedicels scattered, about an inch long: corolla
" violet," 5-parted, the spreading lobes oblanceolate, acute and den-
ticulate, 5 lines long, the calyx half as long: filaments very slender
and anthers small : capsule oblong, equalling the calyx ; cells several-
seeded. — Near G. rigidula. On stony ridges near Los Angeles Bay.
(593.)

Ellisia chrtsanthemifolia, Benth. Shady mountain-sides near
Los Angeles Bay. (580.)

Phacelta crenulata, Torr. A small-flowered form. Stony
ridges at Los Angeles Bay. (592.)

Piiacelia (Eutoca) pauciflora. Annual, setosely hispid, lax
and very slender, a foot high or more : leaves (except the uppermi
pinnately divided, the oblong segments pinnatifid: dowers few, shortly
pedicellate in a short loose simple or geminate raceme: calyx-lobes
setosely hispid, linear-oblanceolate, 2 becoming 4 lines long, equalling
the corolla (scarcely 3 lines long) ; appendages narrow and united at
base over the filament: stamens and style included, the style cleft to
the middle: capsule ellipsoidal, obtuse, 2 lines long, 8-seeded. — Tech-
nically of the Eutoca section, hut closely related to P. hispida. On
mountains near Los Angeles Bay. ( 583. )

Cordia Grf.ggii, Torr., var. (?) Palmeri. Leaves somewhat
larger (5 to 9 lines Ion?), ovate-oblorig, acute to obtuse, more abruptly
cuneate at base on a slender petiole 2 to 1 lines long, and with m

62 PROCEEDINGS OF THE AMERICAN ACADEMY

numerous serratures : corolla more broadly funnel-form, the limb 12 to
15 lines broad; calyx-teeth slender, nearly equalling the tube. — A
shrub, 5 to 8 feet high, in gravelly ravines about Guaymas. (112.")

Cordia (Myxa § Spiciformes) Palmeki. A shrub, 4 or 5 feet
high, the young branches and lower side of the leaves finely pubescent :
leaves oblong-lanceolate, attenuate at base, very shortly petiolate, acut-
ish, rather coarsely toothed, scabrous above, 1 to 2^ inches long by 3
to 6 lines broad: spikes pedunculate, loose, tomentulose, 1 to 1A inches
long: corolla 3 lines long, twice longer than the calyx: fruit red, ex-
serted, nearly 3 lines long. — " Yerba del pasmo." In ravines in the
high mountains above Guaymas. (281.)

Bourreria Sonorje. A shrub, about 10 feet high, with stout
very leaf}' branches : leaves very roughly papillose-scabrous above,
soft-pubescent beneath, obovate to oblong-spatulate, obtuse or retuse,
attenuate to a very short petiole, entire, 1^ inches long or less : flow-
ers few on slender pubescent pedicels 2 to 5 lines long : calyx cleft
to the middle, the deltoid lobes acute, half the length of the corolla-
tube ; corolla greenish yellow, puberulent, 5 lines long, the lobes
oblong, acute; filaments glabrous, exserted : fruit black, depressed-
globose, 2| lines broad. — Near a waterfall in the mountains above
Guaymas. (289.)

Coldenia* angelica. Annual, prostrate, wide-spreading, some-
what hispid : leaves ovate to rhombic, acute, strigose-hispid on the
margin and finely appressed-pubescent, entire or the margin sinuate,
irregularly 2-3-nerved each side of the midvein : calyx nearly 2 lines
long, the narrowly acuminate hispid lobes nearly twice longer than
the tube ; corolla " mauve," 3 lines long ; filaments slender, adnate
nearly to the middle, with a short obtuse prominent appendage at

* Two other species of this genus may be defined as follows : —
C. brevicalyx. White with a close fine pubescence and scarcely at all
hispid : leaves very like those of C. angelica in form and veining: calyx very
short (1 line), the broadish acute lobes shorter than the tuhe; corolla 2 lines
long; filaments slender, slightly dilated below the insertion: nutlet subglo-
bose, much smaller; embryo similar. — Confused with the true C. Palmeri
as to characters of flower and fruit. Ou the lower Colorado (Palmer, 1869;
C. Palmeri, Gray, Proc. Am. Acad. 7. 292, and Syn. Fl., in part) ; San Ber-
nardino (147 W. G. Wright, 1880).

C. Palmeki, Gray. Limiting the species to the one of Palmer's original
specimens which has the leaves " plieate-lineate by about G pairs of straight
and strong veins," it may be otherwise distinguished by the 5-parted calyx (1^
lines long), the lobes linear and acute. The corolla is 2 \ lines long, with the
rather stout fi'ament somewhat dilated below the insertion. The fruit is
unknown.

OF ARTS AND SCIENCES. 63

base: nutlet globose, nearly £ line broad: cotyledons thick, orbicular.
— Very common on sandy bottom* near Los Angeles Bay. (517.)
A flowering specimen collected by M. E. Jones at San Queutin is
probably the same, but is more woody and apparently perennial.
Heliotropium Curassavicum, Linn. Muleje. (15.)
Heliotropium phyllostachyum, Toit. On rocky mesas at
Guaymas. (232.)

Krynitzkia axgustifolia, Gray. Tlie root was formerly used as
a purple dye by the Indians. On stony ridges at Los Angeles Bay.
(606.)

KitYNiTZKiA ramosissima, Greene. Stony ridges, Los Angel
Lay. (551.)

Ipomcea coccinea, Linn. Mountain ravines near Guaymas. | :!10.)
Ipomcea hederacea, Jacq. In ravines near Guaymas. (295.)
Ipomcea leptotoma, Toit. Very common about Guaymas. (231 .)
Ipomcea Bona-nox, Linn. River-banks at Muleje. (33.)
Ipomcea triloba, Linn., var., with glabrous calyx. Collected also
by Pringle in 1884 in Santa Cruz Valley, Arizona, and by Palmer
(213) in 1885 in Chihuahua. In hedges and ravines about Guaymas.
(306.)

Ipomcea Palmeri. A vigorous climber, glabrous : leaves digi-
tately divided, on slender petioles, the 5 segments linear-lanceolate,
attenuate to each end, obtusish, 1 to 4 inches long: peduncles 1-
flowered, 2 to 4 inches long: calyx glabrous, becoming 1 }, inches
long, the sepals oblong, obtuse, chartaceous in fruit; corolla white, 2
inches long, with broad tube and rather narrow limb : anthers much
twisted, a little exserted : stigma biglobose : capsule globose, ! inch
broad, 4-valved, 4-seeded ; seeds very finely pubescent. — "Flowers
with the odor of Stramonium." Margin of a dry creek-bed near
Guaymas. (75.)

Jacqtjemontia Prtnglel Gray, var. glabrescens, Gray. Flow-
ers "pale blue with white lines." Hills near Guaymas. ('-"•' I.)

Jacquemontia Palmeri. An erect slender annual, or at length
somewhat climbing, simple or branched, a foot high or more, sparingly
soft-pubescent: leaves ovate, usually cordate at base, acute, punctate,
15 lines long or much less, rather shortly petiolate : flower- few (1 to
5), scattered on slender peduncles : calyx-lobes ovate-lanceolate, 2 to
2^ lines long; corolla blue, 3 or 4 lines lung: capsule globose, equal
the calyx, the four valves splitting equally to the base : seeds Bomewhat
roughened. — In shade in the mountains about Guaymas. (221 )
EVOLVULUS LINIFOLIUS, Linn. Mountains about (iu.iyn: 'I.)

64 PROCEEDINGS OP THE AMERICAN ACADEMY

Cressa Cretica, Linn. Guaymas, near salt water. (303.)

Cuscuta umbellata, HBK. About Guaymas, upon species of
Boerhaavia. (173.)

Cuscuta Americana, Linn. With small flowers and entire scales,
as in C congesta, Benth., from Acapulco, which is referred to this
species. About Guaymas. (331.)

Cuscuta (Eugrammica) Paliieri. Very slender: umbels soli-
tary or in loose fascicles, few-flowered, the pedicels about equalling
the flowers : calyx subcampanulate, acutely lobed, about half as long
as the corolla-tube ; corolla white, persistent, 1^ lines long, the lobes
equalling the tube, lanceolate, subacuminate, becoming reflexed : sta-
mens half as long, the filaments rather stout ; scales very short, del-
toid, fimbriate : styles slender, the longer equalling the stamens :
capsule globose, small (% line broad), breaking away at base. — At
Los Angeles Bay, very common, covering prostrate Euphorbias, etc.,
and forming sheets several yards in extent. (544.)

Solancji nigrum, Linn., var. nodiflorum, Gray. " Yerba
mora"; the young leaves and tops are much used by the Indians in
cooking. At Muleje and Guaymas. (9.)

Solanum Hindsianum, Benth. Erect, very thorny, 3 to 6 feet
high ; flowers light to dark purple. " Perhaps distinct from S. elce-
agnifolium" (Gray). Guaymas. (109.)

Physalis pubescens, Linn. Muleje and Guaymas. (14.)

Physalis angulata, Linn. ? Flowers yellow with brown base;
fruit edible. About gardens at Guaymas. (G22.)

Physalis angulata, Linn., var. Linkiana, Gray. Stem succu-
lent, 3 feet high ; flowers " white with green centre" ; fruit fleshy
and edible. At Guaymas. (175.)

Physalis ? Resembling an entire-leaved P. pubescens, but

the pedicels elongated ; fruit tinged with purple when ripe, sometimes
eaten and used in cooking. Rocky ridges, in shade, at Los Angeles
Bay. (561.)

Physalis ? Prostrate, very finely pubescent, the upper side

of the small ovate to rhombic-lanceolate irregularly blunt-toothed
leaves nearly glabrous: calyx-teeth deltoid; corolla 6 lines broad,
yellow with dark brown centre : anthers yellow : fruiting calyx
puberulent, subglobose with open orifice, 6 to 8 lines long, equalling
the pedicel. — Guaymas. (94.)

Physalis ? Flowers in the dried specimens purplish blue,

but described as " drab with wood-colored centre." Under shaded
cliffs in the mountains about Guaymas. (G21.)

OF ARTS AND SCIENCES. 65

Capsicum baccatum, Linn. A perennial, 5 feet high, with small
red fruit which is sent to San Francisco as " chiltepin peppers."
Guaymas. (136.)

Capsicum cordiforme, Mill., var. globosum, Dun. ? A culti-
vated perennial, 6 feet high, with globose orange-colored fruit. Guay-
mas. (135.)

Capsicum annuum, Linn. Various cultivated forms. Guaymas.
(137-140.)

Lycium Richii, Torr. Unarmed, 6 to 8 feet high ; flowers lilac ;
fruit edible. Like the type of L. Palmeri, Gray. At Muleje. (4.)

Lycium Ricnn, Torr., var. A very thorny shrub, sometimes 15
feet high ; pedicels shorter ; flowers " violet" ; fruit edible. In alka-
line bottoms about Guaymas. (71.)

Lycium Andersoni, Gray, var. pubkscexs. A shrub, 4 or 5
feet high, resembling var. Wrightii, but finely pubescent, the calyx
half a line and the corolla 3 lines long with a very narrow tube ; flow-
ers " lavender," tetramerous, the filaments glabrous : berries red. —
In stony ravines near Los Angeles Bay. (559.)

Lycium barbinodum, Micrs? A loose shrub, 5 feet high, with
black wood, small white flowers, and scarlet fruit. Hillsides and
ravines at Guaymas. (230.)

Lycium carinatum. A thorny glabrous shrub, 1 or 2 feet high :
leaves narrowly spatulate, 3 lines long or less, glaucous : pedicels
clavate, compressed, 2 or 3 lines long: calyx bifid, the broad acutish
lobes carinate ; corolla 4-lobed, white, 2 lines long, the tube very
short and the throat scarcely as long as the lobes : stamens -4, villous-
tufted at the point of insertion, equalling the corolla : fruit said to be
red. — " Sal sieso." In large patches near Guaymas. (178.)

Lycium ? A shrub, 4 feet high, with woolly nodes, very broadly
spatulate leaves, and " white flowers." The specimens are without
flowers or fruit, and the reference to this genus is only a conjecture.
Guaymas. (337.)

IS'icoTiANA trigonophylla, Dun. At Muleje and on San Pedro
Martin Island. (18. 4'.0.)

Nicotiana Clevelandi, Gray. Common at Los Angeles Bay.
and used for smoking by the Indians. (55 G.)

Mohavea viscida, Gray. Flowers lemon-color with brown dots,
the lower stamens wholly wanting in the flower examined. On stony
ridges at Los Angeles Bay. (597.)

Antirrhinum cyatiiiferum, Benth. {A. chytrospcr mv >» , Gray.)
In gardens at Guaymas. (152.)

\OL. XXIV. (S. b. XVI.) 5

66 PROCEEDINGS OP THE AMERICAN ACADEMY

Antirrhinum Kingii, Watson, var., with longer pedicels. Col-
lected also in Sonora by Pringle in 1884. Among rocks near Los
Angeles Bay. (589.)

Stemodia durantifolia, Sw. Flowers white. Muleje. (25.)

Conobea intermedia, Gray. In dried roiky river-bottoms in
the mountains about Guaymas. (241.)

Mimulus moschatus, Dougl. On wet rocks near Guaymas,
(664.)

Crescentia alata, HBK. The gourd-like fruit is subglobose,
about 4 inches in diameter, 1 -celled (as also the ovary), filled by the
pulpy parietal placentae and numerous flattened obcordate seeds (4
lines long). The species appears to be in every respect a Crescentia.
It is cultivated at Guaymas, under the name of " ayal," for shade and
for the medical properties of the fruit, which is filled with water and
the liquid afterwards taken as a remedy for contusions and " internal
bruises." (85.)

Martynia althe^efolia, Benth. Flowers yellow, lined with
orange and dotted below with brown. Guaymas. (114)

Martynia fragrans, Lindl. Flowers honey-scented. In low
moist places at Guaymas. (326.)

Martynia Palmeri. Stems herbaceous, prostrate, from a large
yellow fusiform root : leaves opposite (the upper alternate), long-
petiolate, ovate-cordate, obtuse, the margin sinuate, 14; inches long or
less : inflorescence floccose-pubescent and viscid ; pedicels 2 or 3
inches long ; bracts at base of calyx ovate, short, becoming thickened
and spongy : calyx campanulate, 5-toothed, 4 to 6 lines long, the
throat oblicme ; corolla 1| inches long, buff with orange and red
stripes and the throat dotted with red : stamens 4 ; seeds irregularly
oblong, angled and more or less prominently tuberculate, 3 lines long.
— Root resembling a carrot and often weighing 3 or 4 pounds ; flow-
.ers carnation-scented. The green fruit and the seeds are used for food
by the Yaqui Indians. Sandy places at Los Angeles Bay. (599.)

Elytraria tridentata, Vahl. " Cordoncillo" ; used as a remedy
for fevers, venereal diseases, etc. In shaded places at Guaymas.
(285.)

Ruellia tuberosa, Linn. Under hedges at Guaymas. (98.)

Ruellia ? A shrub, 4 feet high, with glandular-pubescent

narrowly ovate leaves (6 to 10 lines long), and 1 to 3 flowers in the
axils upon a very short peduncle ; bracts very small : calyx-lobes
linear, acuminate, 5 or 6 lines long ; corolla light purple, 2 inches
long, the tube equalling the calyx and the dilated throat longer than

OF ARTS AND SCIENCES. 67

the lobes : stamens unequal : capsule 6 to 9 lines Ion"-, acuti
seeded. — In moist shaded places on the mesas and in ravines about
Guaymas. (196.)

Beloperone Californica, Benth. Shaded spots in the moun-
tains, at Muleje and Guaymas. (16.)

Dianthera Sonor^e. Perennial (?), the herbaceous stems very
finely roughish-puberulent, about a foot high : leaves ovate to lance-
olate, shortly petiolate, acutish, sparsely pubescent on the margin and
veins beneath, 1 or 2 inches long; spikes dense, terminal, sessile, soli-
tary, 1 \ inches long or less ; bracts imbricated, oblanceolate, acute
and mucronate, strongly 3-nerved and ciliate, 4 to 6 lines long: calyx
4-parted, the very narrow attenuate ciliate segments 2 lines long;
corolla cream-color, 8 lines long, the straight narrow tube somewhat
longer than the limb: anther-cells parallel, nearly equal, separated by
a rather narrow connective. — Near a creek in the mountains about
Guaymas. (240.)

Jacobinia ovata, Gray, var. subglabra. Branches glabrous
and leaves nearly so; flowers in short spikes; otherwise apparently
identical with the original of the species. Shrubby, 4 to 6 feet high,
the flowers bright scarlet. Mountain ravines near Guaymas. (264.)

Lantana Camara, Linn. A nearly glabrous form. " Confiturea";
used as a preventive of hydrophobia. Plains about Guaymas. c2~- 1. 1

Lippia (Zapania) Palmeri. Near L. graveolens ; shrubby, 3 to
5 feet high : leaves ovate to elliptical, the blade decurrent upon the
short petiole, obtuse or acutish, rugose, subcrenately toothed, finely
substrigose-pubescent, an inch long or often much less : peduncles
solitary or in pairs in the axils, very short; heads often few-flowered,

2 to 6 lines long; bracts decussate, the lower united to the middle,
the upper distinct: calyx thin, not carinate ; corolla salverform. —
"Origano"; with a strong sage-like odor and used as a potherb. In
arroyos about Guaymas. Flowers white or cream-color (277, 644),
or rose-color (643).

ClTKAREXYLU.Yl (CaCOCALYX) FLABELLIFOLIUM. A tall shrub,

the younger portions more or less pubescent with shorl spreading
hairs: leaves thin, flabelliform, truncate or rounded above and cre-
nately few-toothed, abruptly contracted to a short petiole, \ inch lung
or less : racemes terminal or sometimes axillary, sessile, loosely I
ered; pedicels very short : calyx 5 nerved and -angled, acutely toothed,

3 lines Ions, becoming thin and dilated and loosely enclosing the
fruit; corolla dark violet. 6 lines long, the yellowish tube shorter than
the calyx, tomentose within, the broadly expanded limb with D<

68 PROCEEDINGS OF THE AMERICAN ACADEMY

equal rounded lobes : fifth stamen anantherous : fruit black, 2 or 3
Hues long. — Differing from Citharexylum in foliage, in the large
violet flowers, and in the calyx enveloping the fruit. The juice of
the fruit stains a persistent black. Mountain ravines about Guay-
mas. (237.)

Bouchea dissecta. Annual, slender, erect, branched, very finely
puberulent or glabrate, 2 feet high : leaves thin, petiolate, ovate in
outline, pinnately cleft nearly to the rhachis, the narrow segments
entire or 1-3-toothed .: spikes slender, elongated, nearly sessile, the
appressed flowers about equalling the internodes ; bracts very small,
subulate: fruiting calyx 3 or 4 lines long, narrowly and acutely
toothed ; corolla white, scarcely exserted : fruit linear, 6 lines long,
conspicuously long-beaked. — Peculiar in its dissected leaves. On
rocky ridges about Guaymas. (259.)

Vitex mollis, HBK. "Uvalama"; a small tree with fragrant
violet-colored flowers and black edible fruit. At Muleje. (3.)

Hyptis Emorti, Torr., var., with leaves less tomentose. A shrub,
3 to 6 feet high, known as " salvia " and used for seasoning. Sandy
bottoms near the beach at Los Angeles Bay. (573.)

Hyptis Palmeri. Shrubby, 6 to 8 feet high, more or less hoarv-
puberulent throughout: leaves petiolate, lanceolate, rounded or cuneate
at base, acute, crenately denticulate or the larger more coarsely toothed,
1 or 2 inches long: flowers in loose umbel-like very shortly pedunculate
axillary cymes, the upper nearly naked ; pedicels slender, 1 or 2 lines
long : calyx narrowly turbinate-cylindrical, white furfuraceous-puberu-
lent (like the pedicel), the slender teeth more than half the length of
the tube ; corolla lilac. — " Salvia " ; used for rheumatism. Common
in arroyos about Guaymas. (278.)

Salvia privoides, Benth. At Guaymas. (320.)

Stachys coccinea, Jacq. Near a waterfall in the mountains
above Guaymas. (300.)

Plaxtago Patagonica, Jacq. On stony ridges at Los Angeles
Bay. (524.)

Mirabilis tenuiloba, "Watson. A foot high or more, with flesh v
leaves and white flowers ; fruit globose, 2 lines in diameter. Moun-
tains near Los Angeles Bay. (600.)

Mirabilis Califormca, Gray. Flowers purple. Mountains,
Los Angeles Bay. (001.)

Allionia incarnata, Linn. Guaymas. (100.)

Bqerhaavia paniculata, Rich. Rocky ledges at Guaymas.
(003.)

OF ARTS AND SCIENCES. 69

Boerhaavia ? A doubtful species, not in good fruit, between

the last and B. hirsuta. Common iu gardens at Guaymas, widely
procumbent (172) or smaller and upright (172.1,).

Boerhaavia erecta, Linn. Bloom white, becoming pink. High
mesas at Guaymas. (182.)

Boerhaavia erecta, Linn., var., with narrow leaves ; a common
western form. Guaymas. (678-680, 684-687.) Also what is prob-
ably an abnormal form of the same. (682.)

Boerhaavia alata. Apparently erect, 2 feet high, much branched,
puberulent, reddish: leaves narrowly oblong to linear, or the lower
broader, H inches long or less: inflorescence loosely paniculate, the
slender pedicels (1 to 3 lines long) solitary or subumbellate : perianth
pink, 2 lines long: stamens (5) and style included: fruit nearly 2 lines
long, truncate, more or less broadly 5-winged, the wings often broader
than the body, narrower below. — On a small rocky island in Guay-
mas harbor. (332.)

Boerhaavia triqdetra. Similar to the last, puberulent ami
viscid, procumbent : leaves lanceolate, acute or oblusish : pedicels
very short (about ^ line) : perianth pink, a line long: stamens (2)
and style included: fruit a line long, truncate, broadly obpyramidal,
very acutely 3-4-angIed, the sides rugose. — Allied to B. pterocarpa,
which has still smaller flowers (^ line long) with 2 or 3 stamens, and
a broader-bodied wing-angled fruit attenuate to a stipe-like base.
Sandy plains and stony ridges near Los Angeles Bay. (521.).

Boerhaavia Wrightii, Gray. (B. bracteosa, Watson.) Flowers
"waxy white'': distinguished from the following triandrous spicate
species by the usually conspicuous thin bracts and the broader fruit,
which is obtuse and acutely angled. Los Angeles Bay. (603 I

Boerhaavia Xante Annual ; stems sparingly branched, Bome-
what floccose-pubescent and puberulent, or glabrous above, occasionally
with viscid spots, 2 feet high or less: leaves ovate to oblong-ovate or
the upper narrowly lanceolate, sinuate, obtuse or acute: flowers large,
in a rather loose raceme an inch long or less ; bracts lanceolate and
bractlets linear, acuminate: perianth " white,*' spreading, 11 lin<
stamens (4) and style long-exserted: fruit obloug-obovate, obtuse, ;•
line long, rather acutely 5- (4-6-) angled, rugose in the intervals.
First collected by Xautus (n. 93) at Cape Saint Lucas. Sands m<
Guaymas. (681.)

Boerhaavia Palmeri. Resembling tie' last ; stems procumbent
or ascending, glandular-pubescent nearly tbrougl t : spike-like ra-
ceme short and dense in flower, 4 lines long or less, about 6 lines long

70 PROCEEDINGS OF THE AMERICAN ACADEMY

in fruit ; bracts obovate and bractlets lanceolate, acuminate, ciliate :
perianth " white," f line long : stamens (3) and style included : fruit
a line long, clavate, obtuse and obtusely-angled, the rather shallow
intervals transversely rugose. — Dry sandy soil near Guaymas. (683.)
It is also n. 209 of Palmer's collection in 1885, from Hacienda San
Miguel near Batopilas, Chihuahua. — B. Coulteri (Senkenbergia
Coulteri, Hook.) is a similar species, but less glutinous and the loose
slender spikes more elongated ; bracts and bractlets very narrow ;
perianth a line long; fruit clavate, 1 to 1^ lines long, truncate, the
angles acutish and the channels extending to the very apex. Coul-
ter's specimen (n. 1425) is labelled as from " Mexico." Arizona
specimens of Palmer's early collections, 378 Rothrock from Camp
Grant, and specimens of Pringle's collection in 1881 from the foot-
hills of the Santa Catalina Mountains appear to be the same.

Boerhaavia spicata, Choisy, var. (?) Palmeri. Similar to the
preceding group ; stems procumbent or ascending, finely pubescent :
leaves thin, nearly glabrous : flowering racemes open and very slen-
der, becoming 1 or 2 inches long ; bracts lanceolate or linear, acu-
minate ; bractlets none : perianth pink, ^ line long : stamens (2) and
style included : fruit clavate-oblong, a line long, obtuse and mostly
obtusely angled, the channel very narrow and sinuate. — Sandy mesas
about Guaymas. (141.) — Typical B. spicata, as shown by a flow-
ering and fruiting spike from Pavon's original specimen in Herb.
Boissier, which was very kindly loaned me for comparison by M.
Barbey, is peculiar in the ovate or ovate-lanceolate acute dark-
colored bracts (with a pair of very narrow bractlets at the base of the
young fruit), and in the conspicuous brown nerves of the nearly
truncate perianth (^ line long). The fruit in the specimen is still
young. The stamens are described as solitary, and but one was
detected in the first flower examined, though there were two in the
second one. I have seen no specimens that correspond to the type.
The form of Texas, New Mexico, and Arizona that has been referred
to this species is usually more glandular than the var. Palmeri, the
leaves thickish and scabrous, and the perianth about a line long. It
may be distinguished as var. (?) Torreyana.

Boerhaavia scandens, Linn. An " evergreen, its many weak
stems hanging upon bushes and fences, with greenish yellow flowers,
and the fruits adherent to everything they touch." Near Guaymas.
(146.)

Abronia umbellata, Lam. Sandy plains near Los Angeles
Bay. (604.)

OF ARTS AND SCIENCES. 71

Cryptocarpus (?) capitatus. A scraggy shrub, 10-15 feet high,
with rigid divaricate branches aud branehlets, the younger portions
finely pubesceut throughout: leaves alternate, entire, mostly broadly
obovate, obtuse or acutish, cuneate at base, 1 or 2 inches long includ-
ing the slender petiole (3 to 6 lines long) : peduncles solitary,
axillary, naked, becoming 6 to 9 lines long, bearing a globose head
(4 to 6 lines broad) of cymosely clustered flowers on very short
minutely bracteate pedicels ; involucre none : perianth cream-colored,
turbinate-campanulate, very shortly and obtusely 5-toothed, a line
lone : stamens 5, distinct, the slender filaments exserted ; anthers
rounded : ovary obliquely linear-oblong, acute, shorter than the slen-
der included style, attenuate to the base; ovule solitary, sessile, erect;
fruit unknown. — A single head of flowers was collected in August,
but it appears to be a winter bloomer, as the January specimens show
an abundance of imperfectly developed flowers. In most of its char-
acters, so far as known, the plant is certainly closely allied to Cryp-
tocarpus. The fruit and fuller flowering specimens may, however,
modify the reference. Near Guaymas. (6-47.)

Achyronychia Cooperi, Torr. & Gray. Los Angeles Bay. (545.)

Amarantus fimbriatus, Benth. Sandy plains near Los Angeles
Bay (515), and in gardens at Guaymas (154, in part).

Amarantus venulosus, Watson. Gardens at Guaymas. (154,
in part.)

Amarantus Palmeri, Watson. Varying in habit from procum-
bent or ascending to erect and 5 or G feet high, and in the more or
less slender or compact spikes, which are often much elongated. It
is one of the commonest plants in Sonora and Lower California
after the setting in of the rains, in gardens aud cultivated fields and
in all bottom lands. It is valuable as a forage plant, and the seeds
are largely gathered and sold in the markets for making bread and
"attole." The collection includes numerous forms. (76-78, 127-
134, 312, 675, 676, 688.)

Amarantus Palmeri, Watson, var. (?), with very thin-acarious
broad perianth-segments, the midvein scarcely reaching to the apes j
seeds rather smaller. — On an island in Guaymas harbor, near salt
water, prostrate. (3 11 4-.)

Cladothrix lanuginosa, Nutt. Gardens at Guaymas. (157.)

Gomphrena Sonora, Torr. Rocky mountain-sidea above Guay-
mas. (252.)

Frozlichia alata, Watson. By irrigating ditches near Guaj mas.

(245.)

72 PROCEEDINGS OF THE AMERICAN ACADEMY

Iresine alternifolia. A shrub, 6 feet high, the young branches
finely white-tomentose : leaves often or mostly alternate, thin, ovate
or sometimes oblong-ovate, obtuse or acute, truncate or cuneate at
base, tomentulose beneath, nearly glabrous or glabrate above, an inch
long or less: flowers dioecious, in mostly broad rather open and nearly
sessile panicles 2 to 4 inches long, scattered or subapproximate along
the ultimate branches, § line long, the staminate pubescent, with very
short bracts, the pistillate with smooth and shining bracts. — In the
mountains about Guaymas. (276.)

Chenopodium album, Linn. Guaymas. (72, 73.)

Chenopodidm ambrosioides, Linn. "Hipasote"; Guaymas.
(171.)

Beta vulgaris, Linn. "Acelga"; Guaymas, in old gardens.
(115.)

Atriplex elegans, Dietr. Upright, 3 feet high. Guaymas.
(117, 122.)

Atriplex Barclayi, Dietr. Though varying somewhat in habit,
as well as in fruit, the specimens are apparently all to be referred to
one species, which is very probably the same as the little known A.
Barclayi from Magdalena Bay on the Pacific side of the peninsula.
The specimens from about the gardens at Guaymas are noted as
growing 2 or 3 feet high. Those from the beach and islands of the
harbor are low, and evidently sometimes perennial. (118, 119, 670,
677, pistillate ; 123, 673, 674, staminate.)

Atriplex linearis. Dioecious, woody at base, 1 to 8 feet high,
much branched, the branches slender : leaves canescent, narrowly
linear, -| to 2 inches long, attenuate to the base : staminate flowers in
mostly small globose clusters in slender spikes bracteate below ; pistil-
late flowers in similar spikes, sessile or shortly pedicellate, solitary or
somewhat clustered, the bracts 2 or 3 lines long in fruit, lanceolate
with broad free and more or less spreading tips, the sides irregularly
tuberculate and developing 4 broad wings, which are more or less
deeply toothed or lacerate. — Near forms of A. canescens, which is
very variable in its fruit but appears never to have the tips of the
bracts so large and conspicuous. In alkaline soil about Guaymas.
(120, 121, 235.)

Su^da Torreyana, Watson. Very common on sea beaches and
in alkaline soil, in large bunches 2 to 6 feet high. The ashes are
used in soap-making. Muleje and Guaymas. (13, 329.)

Spirostachys occidentalis, Watson. Sea-beach at Guaymas.
(330.)

OF ARTS AND SCIENCES. 73

Phaulothamnus spinescens, Gray. A thorny shrub, 10 feet
high ; fruit a drupe with white fleshy pericarp, the embryo surround-
ing copious mealy albumen. About Guaymas. (68.)

Stegnosperma halimifolium, Benth. An upright evergreen, 5
or 6 feet high, with fleshy leaves, white flowers, and red pulpy fruit.
At Muleje, on the island of San Pedro Martin, and at Guaymas.
(31, 400.)

Eriogonum inflatum, Torr. "Tivinagua"; much used as a
remedy for fevers. Muleje and Los Angeles Bay. (24, 574.)

Eriogonum fasciculatum, Benth. Los Angeles Bay, abundant ;
1 or 2 feet high. (549.)

Eriogonum insigne, Watson, var., with the small leaves round-
ovate instead of reniform, and the sepals broader. On mountains
near Los Angeles Bay. (585.)

Polygonum Persicaria, Linn. High mountains above Guay-
mas. (211.)

Antigonon leptopus, Hook. & Arn. " San Miguelito " ; its
long black wiry roots develop large tubers at intervals, which have
a pleasant nutty flavor and are an article of food with the Yaqui
Indians. About Guaymas. (59.)

Aristolochia brevipes, Benth., var. acuminata, Watson.
" Yerba del Indio " ; the root used for aches and pains. About
Guaymas. (174.)

Loranthus Spirosttlis, DC. ? Leaves narrower than described
(2 to 3| inches long by 2 to 5 lines broad), and sessile or subsessile.
Flowers white and very fragrant ; fruit ovoid, 3 lines long, said to
be red. About Guaymas, growing on fig, lime, mesquite, and other
trees, the stems elongated, rooting or hanging in long tangled masses,
(199.)

Loranthus (Psittacanthus) Sonor^e. Glaucous and glabrous,
the slender stems 3 or 4 inches long: leaves alternate, very narrowly
linear or terete, 1 or 2 inches long and less than a line broad : flowers
solitary or in pairs, terminating the branchlets, on pedicels 3 lines long
or less; ovary a line long, subtended by a very shallow cupule and a
short lateral bractlet; perianth scarlet, an inch long, the petals dis-
tinct : fruit unknown. — Hillsides and edges of ravines about Guaymas,
growing upon Bursera microphylla. (287.)

Phoradendron flavescens, Isutt. Guaymas, on Maytenus
phyllanthoides. (88.)

Phoradendron Californicum, Nutt. About Guaymas, in lai
bunches, chiefly upon the thorny mimosas and acacias. (665*.)

74: PROCEEDINGS OP THE AMERICAN ACADEMY

Pedilanthus macrocarpus, Benth. Growing in large masses,
the flowers and fruit bright crimson. A decoction of the plant is used
as an active cathartic. Mountains and hills about Los Angeles Bay.
(605.)

Euphorbia capitellata, Engehn., var. laxiflora. Differing
from the type of the species in the looser cymes, the mostly smaller
gland-appendages, the glabrous capsules, and more sparingly toothed
leaves. In gardens (83, 142) and high in the mountains (210) about
Guaymas. Specimens from a sandy spit in the harbor (317) vary also
in having the narrow and narrowly acuminate leaves nearly all entire,
the involucres few in the cymes, and the stipules of the floral leaves
shorter and less dissected. Pringle's n. 699 of his 1885 collection in
Chihuahua, referred doubtfully to E. pycnanthemum, belongs rather to
this variety of E. capitellata.

Euphorbia Brasiliensis, Lam. At Guaymas. (81, 82.)

Euphorbia albomarginata, Torr. & Gray. At Guaymas. (93.)

Euphorbia (Cham^esyceye) intermixta. Annual, prostrate,
glabrous : stipules broadly triangular, entire ; leaves more or less
broadly oblong or obloug-ovate, obtuse or acutish, oblicpuely truncate
or subcuneate at base, 3 lines long or less, several times shorter than
the internodes : involucres solitary in the forks and axils, pedunculate,
turbinate-campanulate, \ line long; glands dark brown, reniform, with
a usually conspicuous white or pinkish appendage : capsule glabrous,
rather acutely lobed : seeds (immature) pinkish, oblong, angled,
smooth. — Near E. cordifolia among the Leiospermce. Guaymas.
(187, in part; distributed with E. ylyptosperma.)

Euphorbia trachysperma, Eugelm. High mesas about Guay-
mas. (183, 319.)

Euphorbia Magdalene, Benth. ? A low glabrous shrub (the
branches finely tomentulose), with slender elongated virgate branches:
leaves opposite, very shortly petiolate, oblong (2 to 4 lines long), ob-
tuse at both ends, shorter than the internodes (upon the short spurs
shorter, broadly elliptical, and exceeding the internodes) : involucres
solitary, terminal on the spurs, w line long, the broad subreniform
glands and appendages entire: styles long, 2-parted : fruit unknown.
With little doubt the same as Bentham's species from Magdalena Bay,
though not fully agreeing with the description. At Muleje, in dry
ravines. (26.) Found also by Palmer in 1870 on Carmen Island.

Euphorbia tomentulosa, "Watson. With white glands and ap-
pendages and green capsules, or pinkish, or dark purple ; U to 3
feet high". Mountains; Los Angeles Bay and Guaymas. (216, 536.)

OF ARTS AND SCIENCES. 7,",

Euphorbia serpyllifolia, Pers. On gravelly ridges near Guav-
mas. (187.)

Euphorbia setiloba, Engelm. Near the bea;h, Los Angeles
Bay. (629.) — About Guaymas. (185, 1851 634.) The latter a
more open form with longer internodes and larger leaves, and the
appendages less deeply 3-4-lobed.

Euphorbia petrina. Near E. setiloba ; annual, prostrate, much
branched and slender, more or less densely pubescent throughout with
short spreading hairs: leaves very shortly petiolate, entire, obliquely
ovate or oblong-ovate, obtuse, somewhat cordate at base, 3 lines long
or usually much less: involucres solitary in the axils, shortly' pedun-
culate, minute, turbinate-campanulate ; glands greenish, reniform, un-
appendaged : capsule (and pedicel) pubescent, \ line long : seeds
oblong-ovate, quadrangular, strongly and irregularly 3-4-rugose or
-pitted. — Common among rocks at summit of San Pedro Martin
Island. (412.)

Euphorbia portulana. Of the same group ; annual, prostrate,
much branched and slender, the minute and sparse pubescence spread-
ing : leaves oblong to broadly elliptical, obtuse at both ends, entire,
petiolate, 4 lines long or less: involucres oblong, f- line long, nearly
glabrous, somewhat contracted above ; lobes subulate, slightly ciliate ;
glands dark purple, the appendages variable in form, often unequal
(1 or 2 larger), entire or nearly so, white or pinkish: stamens very
few : ovary villous ; styles bifid nearly to the middle : capsule sub-
acutely lobed, nearly a line long : seed oblong, tetragonal, very strongly
4-5-rugose. — Island in Guaymas harbor. (321.)

Euphorbia gi.yptosperma, Engelm. High gravelly ridges at
Guaymas. (187, in part.)

Euphorbia polycarpa, Benth. Several forms, belonging chiefly
to the variety with scarcely appendiculate glands {E. micromera,
Boiss.). By garden fences at Guaymas and on islands in the harbor
(96, 324, 335), and at Los Angeles Bay on sandy bottoms, sometimes
covering many yards of surface (632, 633). — Also var. hirtella,
Boiss., in like localities at Los Angeles Bay. (630.)

Euphorbia Florida, Engelm. In rocky places high in the moun-
tains above Guaymas. (20'J.)

Euphorbia maculata, Linn., or near it; the glands unappendaged
and capsules glabrous. Abundant in low sandy places about Guaymas.
(186.)

Euphorbia pediculifera, Engelm. Stony ridges near I

Angeles Bay. (631.)

76 PROCEEDINGS OF THE AMERICAN ACADEMY

Euphorbia pediculifera, Engelm., var. linearifolia. Closely
resembling the ordinary form in the characters of the inflorescence and
fruit, but erect and branching above the base, the stem and leaves
mostly glabrous ; leaves thin, glaucous beneath, linear, acutish at both
ends, an inch long or less. In both forms the appendages are unequal,
the upper pair larger and oblique. — Hills and mountains about Guay-
mas. (215, 627.)

Euphorbia Californica, Benth. ? Probably not only this spe-
cies but also E. Hindsiana, Benth. A shrub 3 or 4 feet high, with
slender branches ; leaves round-obovate, 3 to 6 lines long, obtuse or
retuse, on very slender petioles longer than the blade: peduncles ter-
minal, mostly solitary : appendages lobed or merely crenate even in
the same involucre : lobes of the capsule (2 lines long) scarcely cari-
nate. — High mountain ravines above Guaymas. (260.)

Euphorbia misera, Benth. A shrub 3 to 5 feet high. On stony
mountain-sides above Los Angeles Bay. (514.)

Euphorbia eriantha, Benth. Rocky ridges about Guaymas and
Los Angeles Bay. (84, 518.)

Simmondsia Californica, Nutt. Common about Guaymas. (343,
344.)

Jatropha canescens, Mull. " Sangre en grado." A shrub, 3 to
8 feet high, with rose-colored flowers, and thick clustered roots. A
decoction of the plant is used as a mordant in dyeing, and the juice as
a remedy for warts, diseased gums, sore throats, etc. Common on the
mesas about Guaymas. (103.)

Jatropha spathulata, Mull., var. sessiliflora, Mull. " Tocote
prieto." A shrub, 5 to 15 feet high, common about Guaymas and
Los Angeles Bay. The fruit of this variety appears to be always 1-
coccous. The bark is used as a mordant and for tanning, and is ex-
ported for these purposes. It is also a dye, giving a reddish brown
color, but is injurious to cloth. (164, 576.)

Jatropha Palmeri. A shrub, the branches softly pubescent:
leaves round-ovate, coarsely and very acutely sinuate-dentate (the
teeth glandular-tipped), densely and subtomentosely pubescent on both
sides, 1-J inches long by 2 wide or smaller: panicles shorter than the
leaves; flowers apetalous, the white cylindrical pubescent calyx 4 or
5 lines long, with round-ovate lobes: stamens of male flowers 10, in
two unequal ranks, the filaments connate and disk densely tomentose :
ovary densely pubescent (becoming 3 or 4 lines long), the styles twice
bifid. — Apparently referable to Midler's subsection Calyptrosolen, but
it is more shrubby in habit and the calyx is less deeply lobed. Only

OF ARTS AND SCIENCES. 77

a single plant found at a high elevation in the mountains above Gun-
mas. (302.)

Croton Pringlei, Watson. With a minty odor. Ravines and
mesas about Guayrnas. (180.)

Argythamnia sericophylla, Gray. In gravelly waste places
about Guayrnas. (108.)

Argythamnia Neo-Mexicana, Mull. Several forms, all of which
appear to belong to this species. Plains and mountain-sides about
Guayrnas. (80, 624, 625.)

Argythamnia Palmeri. Stems numerous, erect, H to 2 feet
high, rather sparsely appressed-villous except on the young branches :
leaves oblong to lanceolate, acute, more or less attenuate to a very
short petiole, entire, 1| or 2 inches long or smaller : flowers apparently
dioecious, the pistillate solitary or in pairs in the axils, on recurved
pedicels 1 or 2 lines long; sepals 2 lines long, becoming 4 lines in
fruit; petals pilose, ovate-lanceolate, shortly acuminate: styles bifid,
the stout branches dilated upward and hispid on the inner side : seeds
ovate-globose, over a line long, coarsely reticulate-pitted. — High in
the mountains above Guayrnas. (247.)

Manihot angustiloba, Mull. Growing 2 or 3 feet high. In the
shade on high mountains above Guayrnas. (233.)

Acalypha Pringlei, Watson, var., with the pubescence scarcely
or not at all glandular, and the teeth of the bracts more numerous
(often 13). Mountains above Guayrnas. (219, 262.)

Tragi a n.epet.efolia, Cav. Hedges about Guayrnas. (63.) —
Also var. amblyodonta, Mull. (623.)

Sebastiania (?) bilocularis, Watson. A shrub or small tic-'.

sometimes 15 feet high. " Yerba flcche " ; the juice is au exceedingly

active and violent cathartic, and the fresh bark is used by the Indians

' to stupefy fish. Common on the shores and in the hills and mountains

about Guayrnas. (234.)

Celtis pallida, Torr. A thorny shrub, 6 feet high or more, with
orange-colored fruit. Guayrnas. (89.)

Ficus (Urostigma) Palmeri. A tree 8 to 12 feet high, l)r.m< li-
ing near the ground ; young branchletswhite-villous-pul)( >■■• lit : leaves
at first densely white-tomentose beneath, becoming Dearly equally
green on both sides and finely pubescent or suhglabrous above, rather
thick, ovate, with a somewhat cordate or roundel base, acute, '■> inches
long by 2 or 2^ broad, on petioles an inch long: fruit in pairs in the
axils, on stout peduncles 6 lines long, globose thick and fleshy, <'■ lines
in diameter, subtended by an irregular disk like involucre '■> lines

78 PROCEEDINGS OF THE AMERICAN ACADEMY

broad : fertile flowers sessile, the perianth of three distinct sepals ex-
ceeding the compressed ovary, the lateral ovate and somewhat concave,
the other carinate ; abortive pistillate flowers very similar, but upon
a short stout pedicel and the lateral sepals narrower ; staminate
flowers also short-pedicellate, the perianth of three large oblong-
obovate sepals ; anther sessile, broad, obtuse. — Fruit edible. On
San Pedro Martin Island; rare. (413.)

Ficus (Urostigma) fasciculata. A tree with a large trunk
(8 feet high) bearing a widely spreading top, the lower limbs horizon-
tal, sending down and supported by aerial roots ; foliage and rather
slender branchlets wholly glabrous : leaves somewhat crowded at the
ends of the branchlets, thin, oblong-lanceolate, short-acuminate or
acute, 3-nerved at the rounded or acutish base, puncticulate above,
finely reticulated beneath, 1^-2^ inches long, on petioles \ inch long :
fruit solitary in the axils, on slender peduncles 2 lines long, sub-
tended by a 4-lobed involucre, depressed-globose with a sunken orifice,
5 lines in diameter (immature) ; flowers all on short stout pedicels,
somewhat oblique, the perianth short, gamophyllous, unequally and
irregularly 2-3-lobed, the lobes short and obtuse ; style of some of the
pistillate flowers filiform and entire, of others bifid ; anther sessile,
ovate, acute, the cells rather long-apiculate. — Cultivated at Guaymas,
but said to be native to the region. (646.)

Ficus (Urostigma) Sonor^e. A tree 15 to 40 feet high, with-
out aerial roots, glabrous throughout : leaves as in the last, but more
scattered on the branchlets and more cuneate at base, the larger
nearly 4 inches long: fruit solitary, on slender peduncles 4 or 5 lines
long, subtended by an irregular disk-like involucre, globose, 5 lines in
diameter : fertile flowers sessile, the perianth cleft to the middle with
broad acutish lobes, shorter than the globose nutlet, which develops
much mucilage when wetted; abortive pistillate flowers with a stipi-
tate oblong-obovate empty ovary (not mucilaginous) surrounded by a
tubular-funnelform 3-cleft perianth ; st iminate perianth similar, but
shorter and somewhat broader, the anther ovate, acute. — Fruit
black, edible, and known as "Nacapuli." Cultivated at Guaymas
(92), and also found wild in the ravines of the mountains (645).
In its foliage it closely resembles F. turbinate/.

Brodi.ea Palmeri. Stem H to 2 feet high, bearing numerous
bulblets at and below the surface; leaves linear, a little shorter than
the scape, 8 lines broad or less : pedicels numerous, J to 1 inch long or
more : corolla purple, 6 to 8 lines long, funnel-form from a narrow
base, cleft to the middle, the throat coronate with a row of very short

OF ARTS AND SCIENCES. 70

thin laminae between the principal veins : filaments 6, naked, scabrous,
free from near the top of the tube and nearly equalling the corolla;
anthers basifixed, liuear, 1£ lines long: capsule equalling the peri-
anth, ovate, narrowed to a short stipe, the cells 2-seeded. — Of the
Seubertia section, though with the basifixed anthers of Eubrodicea.
Abundant in low places on the sandy plains about Los Angeles Bay.
(596.)

Washingtonia Sonoiue. A tree reaching 25 feet in heieht and
a foot in diameter : leaves 3 or 4 feet in diameter, somewhat irlaucous.
very filiferous, upon rather slender petioles which are armed with stout
curved spines: spadix slender, 5 or 6 feet long: fruit about 3 lines
long, the flattened-globose seed 2 to 2| lines in the longest diameter. —
Apparently distinct from W. filifera and W. robusta in its more slender
petioles, more glaucous leaves, and smaller fruit. The seeds are used
for food by the Indians. It may here be noted that in the " Genera
Plantarum " the genus Waslu'ngtonia is placed in the group having
basilar styles, though it is correctly described as having the style ter-
minal. In secluded canons in the mountains about Guaymas. (311.)

Potamogeton pectinatus, Linn. " Grama " ; used in decoction
" to purify the blood." In a small lake near Muleje. (23.)

Naias major, All. " Saragossa." Abundant in a shallow lake
near Muleje. (20.)

Junctts* robustus, "Watson. In brackish water. Muleje. (34.)
Cyperus speciostjs, Vahl. Margin of a fresh-water lake near
Muleje (36), and high in the mountains above Guaymas (010). —
C. LjEvigatus, Linn. Near Muleje and at Los Angeles Bay. (39,
509.) — C. aristatus, Rottb. At Guaymas, near the harbor ami also
high in the mountains (193, 637), and on the summit of San Pedro
Martin Island (417). — C. esculentds, Linn. At Guaymas. (208.)
— C. articdlatus, Linn. High in the mountains above Guaymas.
(636.) — C. ferax, Rich. At the same locality. (639.)
Eleocharis capitata, R. Br. Same locality. (635, 635.1.)
Fimbristylis laxa. With pale glumes. Same locality. (638.)
Scirpus Olxeti, Gray. At Muleje. (37.)
Paspalum distichum, Linn. High in the mountains abo
Guaymas. (693.) — P. pchiflorum, Ruprecht. At Muleje and
Guaymas. (45, 79.)

* The following Juncacea. and Cyperacece were determined by Dr. N I-
Britton, and the Graminew l>y Dr. George Vaset.

80 PROCEEDINGS OP THE AMERICAN ACADEMY

Eriochloa punctata, Desv. At Muleje (44), and at Guaymas.

Panicum sanguinale, Linn. At Muleje (48), and the var. cili-
are near Guaymas (269, 695). — P. Hallii, Vasey. About Guay-
mas. (206.) — P. dissitiflokum, Vasey, n. sp. In deep ravines
and by irrigating ditches about Guaymas. (159, 190.) — P. fascicu-
latum, Sw. Plains and foothills about Guaymas. (207.) Also var.
majus, by ditches. (158.) — P. paspaloides, Pers. Near Guaymas.
(690.) ■ — P. colonum, Linn. At Muleje (46), and about Guaymas.
(51, 202.) — P. lachnanthum, Torr. Not eaten even by goats.
About Guaymas. (54, 348.) — P. capillare, Linn., var. At Guay-
mas. (208.)

Setaria composita, HBK. About Guaymas. (53.) — S. cau-
data, E. & S. At Guaymas and upon a small sandy island in the
harbor. (340.) Also var. pauciflora. At Guaymas. (191.)

Cenchrus echinatus, Linn. At Guaymas. (168.) — C. myo-
suroides, HBK. Near salt water on an island in Guaymas harbor.
(327.) — C. tribuloides, Linn. On the ruins of the old French fort
at Guaymas. (349.) — C. Palmeri, Vasey, n. sp. Near Guaymas
and at Los Angeles Bay. (271, 689.)

Hilaria cenchroides, HBK., var. longifolia. Islands in
Guaymas harbor. (347.)

Cathesticum erectum, Vasey & Hack., and var. Abundant on
mesas and hillsides and in gardens about Guaymas, and on islands in
the harbor. (161,345.)

Heteropogon contortus, R. & S. Hillsides and mountains
about Guaymas. (267.)

Sorghum Halepense, Pers. Gardens and fields about Guaymas.
(64.)

Aristida dispersa, Trin. Gardens and fields about Guaymas.
(66.) — A. bromoides, HBK., and var. On rocky ridges at Los
Angeles Bay. (503, 504.) — A. Schiedeana, Trin. & Rupr. Hedges
and rocky ledges about Guaymas. (55, 268.) — A. fugitiva, Vasey,
n. sp. On sandy beaches at Los Angeles Bay. (501.)

Stipa Californica, Vasey, n: sp. In mountain canons about
Los Angeles Bay. (505.)

Muhlenbergia spiciformis, Trin. Rocky hills near Guaymas.
(272.) — M. tenella, Trin. Island of San Pedro Martin, common
among rocks; the only grass. (416.) — M. debilis, Trin. Plains
and mountains about Los Angeles Bay. (510.)

Sporobolus cryptandrcs, Gray. Abundant about Guaymas.
(65.) — S. humifusus, Kunth. Abundant about Guaymas in the

OF ARTS AND SCIENCES. s]

rainy season. (188.) — S. Virginicus, Kuntli. On islands in Guay-
mas harbor. (338.) — S. Domingensis, Kunth, and vars. Common
about Guaymas. (160, 165, 696.)

Agrostis verticillata, Vill. At Muleje. (41.)

Chloris elegans, Kunth. Near Los Angeles Bay. (506.)
. Bouteloua aristidoides, Thurb. About Guaymas (162, 194),
and at Los Angeles Bay (507). — B. Rothrockii, Vasey. Gardens
and fields at Guaymas. (166.) — B. arenosa, Vasey, n. sp. Plains
and ridges about Guaymas. (189.) — B. bromoides, Lag. Moun-
tains and ledges about Guaymas. (201.) — B. polystaciiya, Torr.
Sandy bottoms near the sea at Los Angeles Bay. (346 ?, 508.) Also
var. major. Common about Guaymas. (204, 205.)

Eleusine Indica, Gaertn. Common at Muleje. (35.) — E.
JEgyi'Tiaca, Pers. About Guaymas. (275, 328.)

Leptochloa mucronata, Kunth, and var. Common about
Guaymas. (50, 192, 694.)

Pappophorum apertum, Munro. On the old fort at Guaymas
and on rocky mountain-sides. (274. 350.) — P. Wrightii, Watson.
Rocky ridges near Bay of Los Angeles. (51ft)

Cottea pappophoroides, Kunth. Rocky island in Guaymas
harbor. (339.)

Triodia pulchella, HBK. On a high rocky point above Los
Angeles Bay. (500.)

Diplachne imbricata, Vasey. Common at Muleje (47), and at
Guaymas. — D. dubia, Benth. High in the mountains above Guay-
mas. (270, 273.) — D. Tracyi, Vasey. Same locality. (691.) —
D. viscida, Scribner. Same locality. (692.)

Phragmites communis, Trin. At Muleje. (38.)

Eragrostis major, Host; forms. At Muleje (40. 401), and about
Guaymas. — E. Pursiiii, Schrad. At Muleje. (42, 49.) Also var.
diffusa. . Common about Guaymas. (167.)

Distichlis maritima, Raf. At Muleje. common. (43.)

Lolium perenne, Linn. In a garden at Guaymas. (52.)

Notholvena* Lemmont, Eaton. In high mountains Dear Guay-
mas. (266.) — N. cretacea, Liebm. (JV! Californica, Baton.)
Rocky mountain-sides above Los Angeles Bay. (552.)

Cheilanthes Prixglet, Davenport. Rooky mountain ]<■'!
near Guaymas. (265.) — C. myriopiiylla, Desv. .Mountain- at
Los Angeles Bay. (553.)

* The Ferns were determined l>y Prof. D. C. Eatox.
vol. xxiv. (nt. s. xvi ) 6

82 PROCEEDINGS OF THE AMERICAN ACADEMY

Pell^a Seemanni, Hook. Shaded rocky ledges in the mountains
about Guaymas. (226.) — P. Wrightiana, Hook. The form with
numerous pinnules (P. longimucronata, Hook.). Los Angeles Bay.
(554.)

Indeterminable Species.

179. Flowering specimens from a tree growing on sardy plains
about Guaymas, described as 12 to 15 feet high and 1 to 5 feet in
diameter, with very green bark and a dense symmetrical top. Its
wood is said to be useless even for fuel. The young branches are
finely pubescent, and the leaves are alternate, coriaceous, entire, linear
with a cordate base, obtuse, glabrous and veiny above, very strongly
reticulate-veined beneath, with a stout midnerve and the margins
revolute, about 3 inches long by 2 lines broad, the very short petiole
(a line long) jointed upon the stem ; stipules none. The flowers as
collected are in loose simple naked racemes 2 inches long, the scattered
spreading pedicels 2 or 3 lines long. They are probably dioecious, as
all those collected have only an imperfectly or scarcely at all developed
ovary. The calyx is .very small, 6-8-parted, apparently valvate ;
petals none; stamens about 20 (18-22) upon a prominent lobed
hypogynous disk (nearly equalling the calyx), the filaments distinct,
slender, 3 lines long, and the anthers short, basifixed, 2-celled, and
longitudinally dehiscent. The most fully developed ovaries are
smooth, oblong, somewhat obcompressed, 2-celled, with a nearly
sessile thick 2-lobed stigma, and apparently one or more (?) pairs of
collateral rudimentary ovules upon the axis. The relations of this
remarkable species are very obscure, and must await fuller material
for their determination. It will probably be found to belong to the
Tiliacece.

307. " Yerba del ayre " ; foliage only of an unrecognized shrub,
found in ravines about Guaymas. A decoction of the leaves#is used as
a remedy for paralysis.

2. Descriptions of some New Species of Plants, chiefly
Californian, with miscellaneous Notes.

Silene Bernardina. Finely glandular-pubescent throughout ;
stems slender from slender rootstocks, a foot high, few-flowered :
leaves very narrowly linear-oblanceolate, 1 or 2 inches long: pedun-
cles slender, 1-3-flowered : calyx cylindrical, J- inch long, with oblong-
ovate teeth ; petals greenish, 8 lines long, the blade cleft to below the

OF ARTS AND SCIENCES. 83

middle into four equal narrow lobes; coronal appendages nearly half
the length of the blade, thin, 2-parted, the outer segment linear, entire,
the inner oblong and lacerate ; claw naked, broadly auriculate : stamens
included, unequal: capsule oblong, shortly stipitate. — Of the S. Lem-
moni and S. occidentalis group. On shady slopes in Long Meadow,
Tulare Co., California; Dr. E. Palmer, June, 1888 (n. 185).

Erigeron sanctaru:m. Perennial and dwarfish, the branches of
the rootstock very slender and naked, bearing one to several erect or
subdecumbent 1-flowered stems, leafy except toward the top. a span
high or less, rough-pubescent : leaves entire, narrowly oblanceolate,
becoming linear above, the lower an inch long (or sometimes longer
and with a slender petiole): heads medium-sized; involucre of nu-
merous very narrow acuminate bracts in two series, the outer hispid,
densely so at base: rays many, purple or fading to white; pappus
simple. — Not closely allied to any other of the Euerigeron section.
Collected by Mr. T. S. Brandegee in the Santa Inez Mountains, Cali-
fornia, also near Santa Barbara, and on the island of Santa Rosa,
1888.

B^eria (Edb^eria) Parishii. Low and slender, much branched,
pubescent with loose woolly hairs : leaves narrowly linear, acute,
mostly pinnately cleft with one to several pairs of lobes : heads
small, 2 lines high or rarely more ; involucral bracts 10 to 12, thin and
lax, the midvein not prominent : ligules short : style-tips truncate,
with an oblique apiculation on the inner side : achenes slender, with-
out pappus. — Resembling epappose small-headed forms of /!. gracilis,
differing in the pubescence and cleft leaves and in the apiculate style-
tips, and approaching the B. platycarpha group. On the foothills of
the San Bernardino Mountains, at 1500 feet altitude; collected by
Mr. S. B. Parish, in May, 1888.

Bahia Palmeri. Low and slender (not 6 inches high), somewhat
soft-pubescent and resinous-atomiferous : leaves mostly alternate, en-
tire, very narrowly linear, an inch long or less: heads sessile and
somewhat clustered, narrowly turbinate, 2-6-flowered ; involucral
bracts 3 or 4, narrowly oblanceolate, acutish, whollj green and her-
baceous: rays none; corolla-lobes ">, more or less unequal : achenea
rather broad, obtusely quadrangular, finely striate and gland ular-
puberulent ; pappus of 8 to 1<) thin obtuse oblong-spatulate denticu-
late chaff, narrowed and thickened toward the base, half tin- length
of the corolla-tube. — An anomalous species, intermediate between
Schkuhria and Bahia as defined by Dr. Gray, and tending t'> invali-
date the distinction drawn by him between the genera. It is placed

84 PROCEEDINGS OP THE AMERICAN ACADEMY

in Bahia as nearer to the section Achyropappus of that genus than to
Schkuhria § Hopkirkia, especially in the characters of the involucre
and achene. Found by Dr. E. Palmer (n. 1 G8) on low slopes aud
ridges in Long Meadow, Tulare Co., California, June, 1888.

Cacalia tussilaginoides, HBK. In the notes made by Dr.
Gray (as given in these Proceedings, vol xxii. p. 432) upon the speci-
mens collected by Dr. Palmer in Jalisco (n. 1G8, of 1886), which he
referred with some uncertainty to this species, the words, "ex char."
should have been erased in the proof-reading. These were in his
notes as sent from the Kew Herbarium, where he had found only the
Coulter specimen to which he alludes. Later, he examined at Paris
Humboldt's original specimens upon which the species was based, with
the result that he directed the retention of the reference, though leav-
ing it still questionable.

Microseris anomala, Watson, in Proc. Amer. Acad., 22. 474.
This species was described under a misapprehension of the character
of the pappus. It is without a doubt a form of M. meter ochceta,
Gray. The mature fruit of this species is still unknown.

Collinsia Wrighth. Low and divaricately branching (3 or 4
inches high) , glandular-pubescent and viscid : leaves opposite, linear-
oblong or -lanceolate, entire, about an inch long : flowers pedicellate,

1 or 2 in the axils, the uppermost in a naked umbel ; calyx 2 lines
long, the narrow lobes obtusish ; corolla but little longer, the tube and
throat shorter than the broad obtuse lobes of the limb, the lower lip
blue, the upper yellowish : capsule equalling or a little shorter than
the calyx, 2-seeded : seeds oblong, nearly terete, with a deep ventral
sulcus. — Near C. Torreyi, with the habit and flowers of O. parvi-
flora. On the Greenhorn Mountains, Kern County, California, at

6,000 to 7,000 feet altitude ; collected by Mr. W. G. Wright aud
Dr. E. Palmer in June, 1888.

Mimulus (Eumimulus) deplexus. A dwarf slender annual, 2 or
3 inches high, sparingly viscid-puberulent : leaves linear to linear-
oblong, entire, less than \ inch long : pedicels exceeding the leaves,
soon widely spreading or usually reflexed : calyx glabrous, narrow,
not strongly nerved or angled, nearly equally and obtusely toothed,

2 to 2^ lines long ; corolla 6 or 7 lines long, the narrow tube nearly
twice the length of the calyx, the broad limb deeply bilabiate;
upper lip deep purple, with emarginate lobes, the lower deep yellow
with bifid lobes, somewhat hairy and spotted below: capsule \\
lines long : seeds brownish, irregularly ovate to oblong-ovate, some-
what rugulose longitudinally. — A very pretty little species, of the

OF ARTS AND SCIENCES. 85

r

same group with M. bicolor, montioides, aud rubellus. The placenta
splits at the apex. Collected by Dr. E. Palmer (n. 17G) in June,
L888, on the drier edges of low wet places in Long Meadow, Tulare
Co., California.

Louteridium Donnell-Smithii, Watson, Proc. Amer. Acad. 23.
283. In the description of this curious Guatemalan genus the calyx
is inadvertently described as inverted. It is of course the three upper
sepals that are united into one, the two lower remaining distinct.
The same plant has since been detected in my own collection made in
Guatemala in March, 1885, — found in flower at the plantation on the
Chocou River and noted as growing from 3 to 9 feet high.

Eriogonum: (Ganysma) Es.meraldense. Leaves all radical,
hispid, especially on the margins, round-obovate, cuneate at base,
small (an inch long including the petiole), the rest of the plant wholly
glabrous and glaucous : stem slender, repeatedly dichotomous above ;
bracts small, deltoid to oblong ; pedicels very slender, 2 or 3 lines
long, mostly strongly reflexed : involucres narrowly turbinate, .1 line
long, cleft to the middle : flowers glabrous, white or pinkish, the
oblong to oblong-spatulate segments nearly ecpial, retuse or obtuse,
^ line long. — Near E. glandulosum, but wholly destitute of glands.
Found by Mr. W. H. Shockley, July, 1888, in Esmeralda Co., Nevada,
at Candelaiia, and on Miller Mountain at 7,000 feet altitude.

Eriogonum (Oregoxium) gracilipes. A dwarf cespitose per-
ennial, with compactly branched caudex and crowded oblanceolate
tomentose leaves, ^ inch long or less : peduncles slender, % to 2 inches
long, glandular-puberulent : involucres few in the small solitary head,
turbinate, tomentose : flower glabrous, more or less deep rose-color,
exserted. — Of the E. Kennedyi group, distinguished most conspicu-
ously by the glandular-puberulent peduncle. On the White .Moun-
tains, Mono Co., California, at 13,000 feet altitude; W. II. Shockley,
August, 1888.

Ne.uastylis Prixglet. Stem a span high, terete, usually simple,
1-leaved near or above the middle: radical leaves shorter than tli<'
stem, 1 or 2 lines broad, 2-3-nerved (the nerves somewhat winged
alternately on the two sides) ; spathe-bracts unequal, the larger
nearly equalling the solitary pedicel, 1 \ inches long: flowers very
fragrant, of a delieate pale blue, the segments oblong-oblanceolate,
obtusish, the outer 15 lines and the inner 13 lines Ion-, minutely
apiculate : stamineal column nearly 2 lines long, half the length «-t
the linear yellow anthers: divergent Btigma-lobes 1! line, long:
capsule oblong, 5 to 10 lines long. — Collected by -Mr. ('. <;. Pringle

86 PROCEEDINGS OF THE AMERICAN ACADEMY

in 1887 (n. 1378) in the mountains of Chihuahua in fruit, and de-
scribed from plants in flower at the Cambridge Botanic Garden in
July, 1888.

Nemastylis Dugesii. The plant described by myself as a
Tigridia in Proc. Am. Acad. 20. 375, based upon drawings by Prof.
Duges, should doubtless have been referred rather to Nemastylis, as
the style-branches are plainly represented to be alternate with the
anthers.

Calathea crotalifera. Low; leaves oval, abruptly acute at
each end, 1J to 2 feet long by 10 or 12 inches broad, yellowish green,
the midvein with a narrow whitish margin, paler beneath ; petiole 2
or 3 feet long, at base sheathing the one or two peduncles, which are
8 or 10 inches high : spike erect, distichous, G to 10 inches long, the
bracts (20 to 40) closely imbricated, duplicate-reniform, subtruncate,
1^ inches long or less, yellowish green tinged with red ; bractlets
doubly winged on the back : flowers mostly included, yellowish ;
sepals scarious, 4 to 6 lines long, about half the length of the petals.
— In wet shaded places at the base of the El Mico Mountains near
Yzabal, Guatemala, April, 1885 ; also received in flower from Messrs.
Reasoner Brothers, of Manatee, Florida, cultivated from Guatemala
roots. The resemblance of the spike to the rattle of a Crotalus sug-
gests the name.

Sisyrinchium anceps, Cav. Having had occasion to investigate
the nomenclature and synonymy, as well as validity, of the species
of Sisyrinchium found within the limits of Gray's Manual, I would
again propose the restoration of the above name for one of our forms.
The error made by Linnaeus in uniting the Tournefortian Bermudiana
from Bermuda with the species that had come into cultivation in Eu-
rope from this country was recognized by the botanists generally of
that day, and the two species were kept distinct afterward, as they had
been before. Miller (1768) was the first to give a definite binomial
name to the American plant, calling it S. angustifolium, and the form
intended by him is easily recognized from his description, and from the
figure given by Dillenius, which represent a plant with a simple stem
bearing a single spathe. In 1783, Lamarck published the name
S. gramineum for what as described by him and afterward, figured by
Redoute would appear to be the same thing. In 1788, Cavanilles
gave the name S. anceps to what might also be considered the same,
except that his figure shows a plant with a branching stem. Michaux
afterwards in his Flora described two species (S. Bermudiana and
S. mucronatum) as found by him in the Atlantic States. Of these,

OF ARTS AND SCIENCES. 87

S. mucronatum is the old species, i. e. S. angustifolium of Miller. His
S. Bermudiana is distinguished as having a broadly winged branching
stem and equally bracted spathes, and he cites S. anceps, Cav., as
probably a variety. Nuttall in his "Genera" simply mentions three
species, S. anceps, S. mucronatum, and S. Bermudiana, and Sweet is
cited (" Brit. Fl. Gard., 2 ed., 498 ") as having given to the last the
name S. Nuttallii. Later Americau botanists have endeavored to dis-
tinguish two species, S. anceps and S. mucronatum, by the characters
of the perianth, spathes, and stems, but with so little satisfaction that
more recently Dr. Torrey and Dr. Gray have united them as varieties
under the Bermudan species S. Bermudiana. A comparison of such
fruiting specimens as have come in my way have shown an evident
difference in the seeds of our common species, some being globose,
quite strongly pitted, and less than \ a line in diameter, others more
angled, rather less pitted, and about twice larger (rather more than
^ line broad). This difference of seed appears always to accompany
a difference of habit, the larger seed belonging to the simple stemmed
form (S. angustifolium), while the plant with smaller seeds is always
branching, bearing several spathes. This seems to justify the reten-
tion of the two species of our early botanists, and, restoring Miller's
name for one, there seems to me no good objection to keeping up
Cavanilles's name for the other. Both species are very commonly dis-
tributed through the eastern United States, from Canada to the Gulf.
The species of the Gulf region still need investigation.

Allium hyalinum, Curran in herb. Bulb-coats gray, the areolae
of the peculiar reticulation narrowly linear and much contorted :
leaves several, very narrow : scape slender, a foot high or less ; spathe
bifid, unilateral ; pedicels rather few, an inch long or less : perianth
white to purplish, broad, 3 to A\ lines long, the acute segments broadly
lanceolate to ovate, the inner somewhat narrower : stamens a third
shorter, the filaments dilated at base: ovary not crested.— First col-
lected at Salmon Falls in Eldorado Co., California, by Mis- M. K.
Curran, June, 1881 ; also recently found by Mr. T. S. Brandegee on
Santa Cruz Island. The flowers much resemble those of A. cam-
panulalum, which is a stouter plant with broader leaves, the reticula-
tion in the bulb-coats minute and extremely sinuous, the filaments
more slender, and the ovary crested.

88 PROCEEDINGS OF THE AMERICAN ACADEMY

VII.

THE DETERMINATION OF CHROMIUM IN CHROME

IRON ORE.

By Leonard P. Kinnicutt and George W. Patterson.

Presented January 9, 1889.

The complete decomposition of chrome iron ore, and the determination
of the amount of chromium, is one of the difficult problems of quanti-
tative analysis. Of the very many methods that have from time to
time been published, one given last year by Donath * using barium
dioxide as the oxidizing and fusing agent, seemed the most expeditious
and simple.

The method as given by him was as follows. One part of the very
finely divided mineral was mixed with five times its weight of barium
dioxide, and heated for one half-hour in a porcelain crucible over one
Bunsen lamp. The semifused yellowish green mass thus obtained
was treated with dilute hydrochloric acid. After a few hours, all the
chromium was found in solution as the chromate of barium. The
barium was precipitated with sulphuric acid, and the filtrate from
the barium sulphate neutralized with sodium carbonate, and a few
drops of an alkaline solution of permanganate of potassium added to
oxidize any chromium that might have been reduced. The excess of
permanganate was then removed by the very careful addition of fer-
rous sulphate, and the amount of chromium determined volumetrically.

Donath, however, did not publish any results to show the accuracy
of the method, and it was on this account that we were led to repeat
his work.

A few experiments were sufficient to satisfy us that the perfect de-
composition of chrome iron ore from Pennsylvania could not be
accomplished in the manner described by Donath. A number of
fusions of the very finely powdered and sifted mineral with five or
even ten times its weight of barium dioxide were made, and in every
case the residue contained a greater or less amount of the black unde-
composed mineral. Also we were unable to obtain any porcelain cru-
cibles in which more than one fusion could be safely accomplished ;
while if the fusion was made in platinum crucibles, we obtained, as

* Dingler's Polytechnisches Journal, vol. cclxiii. p. 245.

OF ARTS AND SCIENCES. 89

was also noticed by Donath, a comparatively large amount of the
sesquioxide of chromium.

Another serious difficulty was met with later in the process ; namely,
in determining the exact point where sufficient ferrous sulphate had
been added to reduce the excess of permanganate of potassium, used
in oxidizing the small amount of reduced chromic acid. The change
of color of the solution being so slight as to be almost useless as an
indicator.

We were therefore unable, with the most careful manipulation, to
obtain satisfactory results with the above process. In our study of the
process we made a great many experiments with different fusing mix-
tures, and also tried various ways of oxidizing the small amount of
sesquioxide of chromium we always found present in the solution. The
result of these experiments have enabled us to formulate a very quick
and simple process, which, as can be seen by the results given below,
is also, as regards accuracy, very satisfactory.

About three tenths of the very finely divided mineral is mixed with
twenty times its weight of a mixture containing equal parts of dry
sodium carbonate and barium dioxide, and heated in a platinum cru-
cible with the full flame of one Bunsen lamp for one half-hour. At
the end of this time a quiet fusion is obtained and the decomposition
is completed. The crucible is then placed in a beaker, covered with
water, and hydrochloric acid added, a little at a time, till the mass is
completely disintegrated. The crucible is then removed, the solution
made strongly alkaline with caustic potash, and five or six cubic centi-
meters of a five per cent solution of hydrogen dioxide added to
oxidize the small amount of reduced chromic acid that may be present.
The solution is now boiled for twenty minutes to remove any excess of
hydrogen dioxide, made acid with hydrochloric acid, and the amount
of chromic acid determined by the aid of a standardized solution of
ferrous chloride.

Six analyses of a sample of chrome iron gave the following per
cents of chromium sesquioxide.

1. 2. 3. 4. 5. 6.

49.87 49.97 49.84 49.85 49.88 49.87

The solution of ferrous chloride used was kep( under oil and titrated
before each analysis. One cubic centimeter equalled aboul 0.015
gram of chromium sesquioxide.

Salisbury Laboratories,

Worcester Polytechnic Institute.

90 PROCEEDINGS OP THE AMERICAN ACADEMY

VIII.

CONTRIBUTIONS FROM THE PHYSICAL LABORATORY OF THE
MASSACHUSETTS INSTITUTE OF TECHNOLOGY.

XXXI.— THE STRENGTH OF THE MICROPHONE CURRENT
AS INFLUENCED BY VARIATIONS IN NORMAL PRESS-
URE AND MASS OF THE ELECTRODES.

By Annie W. Sabine.

Presented by Charles R. Cross, Norember 14, 1888.

The experiments described in the present paper, undertaken at the
suggestion of Professor Cross, constitute a study of the variation of
the current in the secondary circuit of a microphone transmitter, as
related to variations in the normal pressure and in the mass of the
electrodes of the microphone. They form a continuation of work of
a similar character which has been prosecuted in the Rogers Labo-
ratory during the past few years, and a portion of which has already
been presented to the Academy.*

The instruments used were similar to those previously employed by
Messrs. Patterson and Tucker, and described in the latter of the two
articles just referred to. The electrodynamometer was calibrated by
means of a Hartmann galvanometer, using reversals in the ordinary
manner so as by four readings to eliminate the effect of the earth's
magnetism on the suspended coil.

The microphone contact was set into vibration by the sound of a
stopped organ-pipe (C4 of 512 vibrations) kept as constant as possible
by means of an air-blast furnished with a regulating air-chamber.
Weights were gradually added to the upper (anvil) electrode, so that
the mass of this and its pressure on the lower electrode were thereby
increased by measured amounts. The weights added were usually in
the form of thin copper washers weighing t8q of a gram each, though
fractions of this weight were used in some cases. One chromic acid
cell was used as a battery.

* See Proc. Am. Acad., Vol. XXI. p. 248 ; Ibid., Vol. XXIII. p. 228.

OP ARTS AND SCIENCES. M

The character of the results obtained will be seen by an inspection
of Figure 1, which gives the curve showing the relation between the
normal pressure and the resulting current, with electrodes of the ma-
terials employed in the Blake transmitter, viz. a platinum hammer
and a hard carbon anvil electrode. The abscissas represent the normal
pressures, i. e. the added weights, and the ordinates the currents pro-
duced in the secondary. The data from which this curve is construct d
will be found in the following table. The load is given in terms of
the arbitrary unit employed ; the currents are given in milliamperes.

Table.

Remarks.

Load.

Current.

0

0

.5

.18

1.0

.46

Loud, goo

1.5

.48

ti u

2.5

.56

«( it

3.0

.49

3.5

.45

Faint, go<

4.0

.33

a a

4.5

.21

5.0

.19

The curve shows a rapid rise at first, as the mass of the anvil elec-
trode, and with it the pressure between the electrodes, is increased,
which rise soon reaches a maximum, the curve then falling off rapidly
at first, more slowly afterwards.

The nature of the curve is interesting, and requires explanation.
In his article already referred to, Mr. G. W. Patterson, who obtained
similar results to my own, considers the curve to be composed of two
separate branches, the rising portion of the curve corresponding to a
motion of the electrodes sufficient to break the circuit, and the falling
portion to the case when the pressure is too great to allow this to
occur. The curve (these Proceedings, Vol. XXIII. p. 235, Fig. h
constructed upon this hypothesis greatly resembles one of the experi-
mental curves shown in Fig. 3 of the same paper. There is less
resemblance, however, to the lower curve in Fig. •'!, and but littl<- to
the curve shown in Fig. 1 of this article, which was obtained under
exceptional conditions of quietness of the piers upon which the appa-
ratus was placed. The sound in a receiving telephone placed in ih<'
secondary circuit is so harsh for this portion of the curve, thai one
might well infer that actual breaks i ccurred ; but this is very doubtful,
and such breaking is certainly not essential to the production of the

92

PROCEEDINGS OF THE AMERICAN ACADEMY

results which are obtained. In fact, the article referred to assumes
that the varying pressure on the contact due to the action of the given
sound-waves will always have the same maximum value, ± S. This
would be approximately true were the normal pressure between the
two electrodes alone to be varied, but the effect of the addition of
weights, as in the method of experiment adopted, is to increase the

Fig. 2.

0

%
\
\

3 4 5 6
Fig. 3.

mass at the same time that the normal pressure is increased, and
under these circumstances the effect of a sound-wave of given intensity
will necessarily be to give to the corresponding pressure-variation a va-
riable value, increasing with the added mass, and hence, with the form
of apparatus used, as the normal pressure is greater. The effect of this
will be to cause at first a gradual increase of current in the secondary,

OF ARTS AND SCIENCES. 93

which increase is succeeded by a diminution of current when the mass
is still further increased. This will be apparent from an inspection of
the curve shown in Figure 2, which illustrates in a general manner
the relation of the pressure between the electrodes to the current in
the primary circuit.

In this the momentary changes in pressure, Ap, Ap', Ap", etc , due
to the sound-waves, and corresponding to loads and normal pressures
p, p\ p", etc., have increasing values within certain working limits,
owin£f to the increasing mass of the anvil electrode. The currents in
the primary also increase, though at a gradually diminishing rate, as
the pressure between the electrodes is increased, so that the increments
of current, Ac, Ac', Ac", corresponding to the pressure-changes Ap,
Ap', Ap", have increasing values up to some point, as a, after which
they decrease. This being the case, it is evident that the current in
the secondary will at first increase to a maximum, and afterwards
diminish, since the currents in the secondary corresponding to press-
ures p, p', p", etc. will be proportioned to Ac, Ac', Ac", etc., and this
is precisely the curve which is obtained in the experiments. The
explanation just offered seems therefore to be the' true one.

The matter was still further tested by carrying out a set of experi-
ments similar to those already described, except that the variations in
normal pressure were brought about by means of a spring instead of
by adding weights. In such a case the successive values of Ap in
the curve (Fig. 2) would be of the same magnitude, while Ac would
continually diminish. The current in the secondary should therefore
have its maximum value when the initial normal pressure is least,
and continually diminish as that pressure becomes greater.

The experimental results verified this conclusion, as will be seen by
reference to Figure 3, in which the normal pressures are represented
by abscissas, and the currents in the secondary by ordinates, as before.
The curve is approximately a straight line. It is possible that the
deviations from this are due to instrumental imperfections, as the
apparatus used did not allow of more than an approximate determi-
nation of the pressure applied by the spring.

Eogers Laboratory of Physics,

June, 1888.

94 PROCEEDINGS OF THE AMERICAN ACADEMY

IX.

CONTRIBUTIONS FROM THE PHYSICAL LABORATORY OF THE
MASSACHUSETTS INSTITUTE OF TECHNOLOGY.

XXXII. — RESEARCHES ON MICROPHONE CURRENTS.

By Charles R. Cross and Annie W. Sabine.
Presented November 14, 1888.

In continuation of studies described in former papers from the Rogers
Laboratory, the authors have observed the variations in the secondary-
current which occur under different circumstances as to mass and
normal pressure when the material of the electrodes of a microphone
is varied. The methods and apparatus employed were identical with
those described in the immediately preceding paper in the?e Proceed-
ings (page 89).

In the first series of experiments both electrodes were made of the
same material, the hammer electrode being a very small button, and
the anvil electrode a large one, as in the Blake transmitter. As the
mass and pressure upon the anvil electrode were varied, the currents
in the secondary at first increased up to a maximum, and afterwards
diminished, as shown in the article just referred to, as well as in the
curves given with the present paper. Tables I. to XVI. give the
numerical results obtained, and Figures 1 to 14 are curves showing
the relations between the variables, the loads being abscissas, and the
corresponding currents in milliamperes the ordinates. The unit of
weight adopted was -j85- of a gram. The remarks in the third column
of the tables indicate the nature of the sound heard in a receiving tele-
phone placed in the secondary circuit. The current was in all cases
zero when the load was zero.

TABLE I.

Electrodes. — Iron, Iron.

Load.

Current.

Remarks.

1

.19

Sound rough and unsteady.

2

.24

Slirill, clear, metallic, unsteady,

3

.16

u a it it

Load.

1
2
3

4
5
6

7

Load.
1

2
3

4
5

6

OF ARTS AND SCIENCES. 95

TABLE II.

Electrodes. — Platinum, Platinum.
Load. Current. Remarks.

Quality good ; fluttering,
metallic.

1

.43

2

.02 Qualii

3

.73

4

.23

5

.10

TABLE III.

fluttering.

Electrodes. — Carbon, Carbon.

Current.

Remarks.

.44

Harsli

; loud.

.63

IC

It

.71

Quali

ty good ;

fluttering.

.67

II

n

slight fluttering,

.27

i t

tt

faint fluttering.

.18

It

u

a a

.12

tt

tt

t€ ft

TABLE IV.

Electrodes. — Platinum, Carbon.
Load. Current. Remarks.

Harsli and loud.

1

.43 Hars

2

.59

3

.67 Qual

4

.59

5

.12 Fain

6

.09

TABLE V.

Quality good and clear, some fluttering.

U it it tt te

Faint, fluttering.

Electrodes. — Carbon, Platinum.

Current.

Remarks.

.445

Harsh and loud.

.61

Clear : good quality.

.66

Clear; some fluttering.

.63

Faint; good quality ; some flutt<

.29

n n it tt >>

.20

■( ii ii "

9Q

PROCEEDINGS OF THE AMERICAN ACADEMY

.4

*■©..

js>

n-Fe

V -/V

/ 2 3 4

Fig. 1.

/ <? 3 4 5 6
Fig. 2.

OF ARTS AND SCIENCES.

97

.6

/ 1

• 6

/ '

.5

.5

/ \

/ 1 Pl

C

/ \ r. />

.4

A

/ \

•3

■3

<A •

.2

2

/

./

\ ®\

1

m

/

/ 2 3 4 5 C
Fig. 4.

y <? 3 4 5 6 7
Fig. 5.

vol. xxiv. (x. s. XVI.)

98 PROCEEDINGS OF THE AMERICAN ACA.DEMY

TABLE VI.

Electrodes.—

■Iron, Carbon.

Load.

Current.

Remarks.

1

.115

Harsh, irregular noise.

2

.260

a u a

3

.310

<< n (i

4

.290

Clear, good quality ; faint.

5

.200

Faint ; slight fluttering.

6

.095

Faint.

7

.040

n

TABLE

VII.

Electrodes. —

Carbon, Iron.

Load.

Current.

Remarks.

1

.395

Harsh and loud.

2

.530

Clear, shrill, good quality.

3

.570

Clear ; fluttering.

4

.590

Quality good ; slightly harsh,

5

.460

ci tt a cc

6

.290

it it it a

7

.100

Very faint.

8

.040

tt

TABLE VIII.

Electrodes. — Platinum, Platinum.
Load. Current. Remarks.

1 .450 Harsh and loud.

2 .500

3 .670 Uncertain ; sometimes good quality.

4 .230 Harsh; fluttering.

TABLE IX.

Electrodes. — Carbon, Carbon.
Load. Current. Remarks.

1 .300 Harsh and loud.

2 .560 Somewhat shrill ; good quality.

3 .620 Good quality, somewhat rough.

4 .545 " " fluttering.
4.5 .225 Faint and rough.

5 .140 Faint.
5.5 .050

OF ARTS AND SCIENCES.

TABLE X.

Electrodes. -

- Iron, Iron.

Load.

Current.

Remarks.

1

.140

2

.310

Quality good.

3

.270

4

.180

TARLE

XI.

Electrodes. — Platinum, Carbon.

Load.

Current.

Remarks.

1

.210

Rough.

2

.420

m

o
O

.540

Quality good ; loud.

4

.630

" " slight fluttering,

5

.120

" '• faint fluttering.

6

.070

si .< t* <(

TABLE

XII.

Electrodes. — C

arbon, Platinum.

Load.

Current.

Remarks.

1

.260

2

.440

Shrill and loud.

3

.530

Clear, loud, good quality.

4

.620

It U II it

5

.555

Fainter and rougher.

TABLE

XIII.

Electrodes —

Iron, Carbon.

Load.

Current.

Remarks.

1

.240

Sound uncertain.

2

.300

3

.310

Clear ; good quality.

4

.290

ii i> «

5

.180

faint.

6

.099

11 11 << ><

TABLE

XIV.

Electrodes. —

on, Iron.

Load.

Current.

Remarks.

1

.345

Rough, steady.

2

.515

3

.615

Good quality.

4

.560

<i •<

5

.235

Paint, rough.

G

.140

Faint ; clear.

100

PROCEEDINGS OF THE AMERICAN ACADEMY

r>t- 7V

,7

i

.5
A

7~ a 3 j 5
Fig. 8.

Fig. 9.

An

i
i

'0

\ Ft - n

X

' 2- 3 4 5
Fig. 10.

/ <2 3 4 5 6

Fig. 11.

OF ARTS AND SCIENCES.

llll

-7V

ft -C

I 2 3 4 5 6
Fig. 12.

i 2 3 4 5 6
Fro. 13.

' 2 3 4 5 6 7

Fig. 14.

102 PROCEEDINGS OF THE AMERICAN ACADEMY

TABLE XV.

Electrodes. -

- Copper, Carbon.

Load.

Current.

Remarks

1

.390

Rough and loud.

1.6

.490

« <<

2

.590

(i «

2 5

.550

<< (£

3

.250

« H

3.5

.130

<C <£

TABLE XVI.

Electrodes. — Carbon, Copper.

Remarks.
Good quality ; fluttering.

Load.

Current

1

.240

2

.520

3

.600

4

.660

5

.240

6

.090

Figures 1 to 3 show very well the characteristics of the substances
iron, platinum, carbon. Comparing the curves for carbon and plati-
num (Figs. 2, 3), it will be seen that the maximum current obtained
under the conditions of the experiment was about the same in both
cases. But the current falls off far more rapidly with platinum as
the pressure is further increased, thus giviDg to carbon a grpater
working range of pressure-variation within which it can be practically
used in a microphone transmitter, a fact long since recognized in
practice. Iron (Fig. 1) has a considerable range, but the current
produced is not great.

Tables IV. to VIII., and Figures 4 to 8, illustrate the results ob-
tained when the electrodes were of different materials, the substance
of which the anvil electrode is made being the first mentioned. In the
experiments now particularly referred to, the two electrodes were
made of the customary size, the hammer small and the anvil large.

An inspection of the curves shows that the character of the result
obtained under these circumstances was determined chiefly by the
material of the anvil rather than of the hammer electrode. This is
particularly well illustrated in the case of iron and carbon, as shown
in Figures 6 and 7.

This difference persisted even when the hammer and anvil elec-
trodes were made of the same size and shape, as will be seen by an

OF ARTS AND SCIENCES.

103

inspection of Tables IX. to XVI. and Figures 9 to 14. We have
stijl to determine whether the same peculiarity would be observed
with other modes of mounting the electrodes.

Besides the foregoing results, we have made a determination of the
actual strength of the working currents employed on telephone lines.
The values previously obtained in the Laboratory had been ascertained
from experiments on very short experimental circuits. But through
the courtesy of the Long Distance Telephone Company, we were able
to put our measuring apparatus in circuit both with the city telephone
lines and with one of the long lines to New York.

Several interesting results were thus reached, as shown in Table
XVII. The first column gives the name of the transmitter employed,
the second the nature of the sound transmitted, the third the location
of the transmitter, the fourth the strength of the telephone current in
milliamperes. The length of the line from 95 Milk Street to the
Institute was about two miles ; that of the line to New York was two
hundred and sixty miles.

TABLE XVII.

Transmitter

Sound.

Locality of Transmitter.

Current.

Blake

Talking

95 Milk Street

.is:,

n

Singing

(t

.52

Hunnings

Talking

a

.28

ct

Singing

tt

.78

tt

Talking

New York

.02

t<

Counting

a

.02-

ft

Organ Pipe

it

.01+

€i

Counting

Rogers Laboratory

2.05

t<

Talking

it

2.20

ft

Organ Pipe

a

1.24

The speakers were Dr. W. W. Jacques of Boston and Mr. F. A.
Pickernell of New York, who kindly aided us in our work. They
are both experts in the use of the telephone, and accustomed to
work with each other. The pitch, as well as the loudness of the
sounds used, was kept as nearly as possible the same. The vocal
sounds transmitted were spoken in a very loud tone, and close to
the transmitter.

104 PROCEEDINGS OF THE AMERICAN ACADEMY

The figures obtained with the long line to New York are very
instructive, as they give some knowledge of the loss of current which
is sustained in long distance telephony. When the transmitter was
at the Institute, near to the dynamometer, the full current produced
by the former passed through the latter instrument, while, when the
transmitter in New York was the one used, it is clear that only what
was left after all leakage, etc. passed through the measuring instru-
ment. Assuming the sounds as produced at the two stations to be of
approximately the same pitch and loudness, it appears that only about
one one-hundredth of the original current produced at the transmitting
station is finally utilized at the receiving station. It further appears
from these figures that about 13 per cent of the current produced at
the transmitting station was utilized in ordinary telephonic transmission
over the local lines from 95 Milk Street to the Laboratory.

Rogers Laboratory of Physics,
June, 1S88.

OP ARTS AND SCIENCES. 1Q5

X.

ON PENTAMIDOBENZOL.
By A. W. Palmer and C. Loring Jackson.

Presented February 13, 1889.

Nearly two years ago, in a paper upon Tribromtrinitrobenzol,* one of
us in conjunction with J. F. Wing described some attempts to reduce
triamidotrinitrobenzol, and announced that the work would be contin-
ued. Before doing this, however, it seemed advisable to try similar
experiments with the triamidodinitrobenzol, as tbis substance can be
obtained much more easily, and the experience gained in this way
would be useful when we returned to the much less accessible trinitio
compound. Accordingly, we have undertaken together the work on
the dinitro compound, which has led, as we had hoped, to the prepara-
tion of the chloride of pentamidobenzol ; but, shortly after we had
sent to the Bericlde der deutschen chemischen Gesellschaft a preliminary
noticef of this substance, a number of the same journal reached us con-
taining a paper by Andrew Barr j in which the chloride of pentamido-
benzol was also described, and, as this paper, although it came into
our hands after our preliminary notice had been sent, was presented
to the German Chemical Society a fortnight before ours was written,
we have dropped this line of work in order to avoid interfering with
Barr, or Merz, in whose laboratory the research of Barr was carried
on, and publish in this paper the results which had been already ob-
tained by us when we received Barr's paper. § The method of making
pentamidobenzol used by us is different from that of Barr, who pre-

* These Proceedings, Vol. XXIII. p. 138.

t Ber. d. ch. G., 1888, p. 1706.

\ Ber. d. ch. G., 1888, p. 1541.

§ The puhlication of this paper has been delayed until the present time, be-
cause the change of residence of one of us prevented us from finishing more
promptly the study of trianilidotlinitrobenzol, a substance which does not come
into Barr's field of work.

106 PROCEEDINGS OF THE AMERICAN ACADEMY

pared the diamidotrinitrobenzol from the ethylether of trinitroresor-
cine and alcoholic ammonia, and then reduced this product with tin
and hydrochloric acid, whereas we submitted to a similar reduction the
triamidodinitrobenzol made by the action of alcoholic ammonia on tri-
bromdinitrobenzol. Both this triamido and the corresponding triani-
lido compound are new, and an account of them is therefore included
in this paper.

Triamidodinitrobenzol, C0H (NHn^fNO^.

This substance was prepared by treating tribromdinitrobenzol,*
melting point 192°, with alcoholic ammonia ; but this action takes place
much less easily than in the case of the corresponding trinitro com-
pound, for whereas that was attacked by alcoholic ammonia in the cold,
and the reaction was rendered complete by short boiling in a flask
with a return condenser, the dinitro compound gave no action with
alcoholic ammonia in open vessels, but it was necessary to heat the
two substances under pressure in the water bath to bring about the
reaction.

As we had no digester fit for this purpose, we were obliged to heat
the mixture in soda-water bottles, which were kept in a boiling water
bath for about eight hours. The contents of the bottles at the end of
this time consisted principally of a mixture of dark reddish yellow
needles with nearly black amorphous masses. There was also a colored
alcoholic solution, which on evaporation left a red amorphous sub-
stance with a comparatively low melting point ; but this was formed in
such small quantities that we did not attempt to study it beyond ascer-
taining the fact that it is not altered by further heating with alcoholic
ammonia. The tribromtrinitrobenzol, when treated with alcoholic
ammonia, forms a similar by-product,f soluble in alcohol, which is now
under investigation. The two products of the reaction which were
not dissolved in the alcohol were separated mechanically, washed
thoroughly with various solvents, and analyzed without further purifi-
cation, as they both showed such a slight solubility in all the common

* The tribromhenzol used in making this substance must be purified by crys-
tallization from alcohol after it has been distilled, as, if this crystallization is
omitted, a substance melting in the neighborhood of 95° remains mixed with
the tribromhenzol, and gives rise to very impure products after the treatment
with nitric acid. The nature of this impurity will be considered in another
paper.

t These Proceedings, Vol. XXIII. p. 145.

OP ARTS AND SCIENCES. 107

solvents that a crystallization of them in any quantity was a matter of
very great difficulty. In spite of the great difference in their appear-
ance they both had the same composition, that of the expected triami-
dodinitrobenzol, as is shown by the following analytical data, in which
Analyses I. and II. were made with the reddish yellow needles, Anal-
ysis III. with the black amorphous body.

I. 0.3000 grm. of the substance crystallized in yellow needles gave
on combustion 0.3710 grm. of carbonic dioxide and 0.1 085 grm.
of water.
II. 0.1510 grm. of the crystalline substance like that used in I. gave
42.4 c.c. of nitrogen at a temperature of 8°. 6 and a pressure of
748.5 mm.
III. 0.3028 gi-m. of the substance deposited in black amorphous masses
gave on combustion 0.3778 grm. of carbonic dioxide and 0.0925
grm. of water.

Calculated for

Found.

C6II(XH2),(N02)2.

i.

II.

III.

Carbon

33.80

33.73

34.02

Hydrogen

3.28

4.02

3.39

Nitrogen

32.87

OO.Ol

Properties. — The triamidodinitrobenzol, as deposited by the action
of alcoholic ammonia on tribromdiuitrobenzol, forms either reddish
yellow needles, or nearly black amorphous masses. When the sub-
stance is crystallized from a mixture of chloroform and alcohol, which
can be done only wdth great difficulty on account of its slight solubility.
it forms microscopic yellow plates, or broad flattened prisms, and, if
the treatment with the mixed solvents is carried on in a continuous
extracter, these crystals collect into characteristic yellow masses shaped
like an S or U. It does not melt even above 300°, but is decomposed
if heated more intensely. It is very slightly soluble in alcohol or
chloroform, although it imparts a perceptible yellow color to these boI-
vents, more soluble in nitrobcnzol, but cannot be obtained in crystals
easily from this solution, so that a mixture of alcohol and chloroform
is the best solvent for it. In the other common solvents it is either
entirely insoluble or very slightly soluble. In this respect it reseml
the corresponding trinitro compound, but it is, if there is any dill', rence,
slightly more soluble than that. The substance does not form definite
salts, although it dissolves easily in concentrated hydrochloric acid
forming a dark red liquid. It shows therefore a greater tendency to

108 PROCEEDINGS OF THE AMERICAN ACADEMY

form salts than the triarnidotrinitrobenzol, which is insoluble in strong
hydrochloric acid.

Trichloride of Pentamidobenzol, C6H(NH2)2(NH,C1)3.

To prepare this substance, some of the triamidodinitrobenzol (either
the crystalline or amorphous) was heated on the water bath in a flask
with strong hydrochloric acid, a very strong solution of stannous chlo-
ride, granulated tin, and some pieces of platinum foil. After some
hours the very deep red solution had lost most of its color, but still
showed a dirty brownish yellow tint ; it was then allowed to cool, when
crystals were deposited, which increased in quantity, if the liquid was
saturated with hydrochloric acid gas. The crystals were filtered out,
and washed with concentrated hydrochloric acid, then, as we did not
succeed in removing the salts of tin completely by precipitation from
an aqueous solution with hydrochloric acid gas,* the substance was
dissolved in very little water and the solution saturated with sulphu-
retted hydrogen, the stannic sulphide formed was filtered out, and the
filtrate, after freeing it from sulphuretted hydrogen with a stream of
air, once more saturated with hydrochloric acid gas, which threw down
the substance in an essentially pure state and nearly white. In all
this work care must be taken to avoid heating the solutions, as the
chloride of pentamidobenzol, after the reducing agents have been re-
moved, is easily decomposed by heat into a black tar ; the solutions,
therefore, should never be concentrated by evaporation on the water
bath. "We obtained a less pure substance also, if the solution in water
and precipitation with hydrochloric acid gas was repeated more than
once. The crystals were dried at first in a desiccator over calcic oxide,
calcic chloride, and potassic hydrate, and finally heated to 70°-80° for
about five hours in an air bath. On analysis they gave the following
results.

I. 0.2372 grm. of the substance gave on combustion 0.2445 grm. of
carbonic dioxide and 0.1157 grm. of water.
II. 0.1827 grm. of the substance gave 0.1880 grm. of carbonic dioxide
and 0.0880 grm. of water.

III. 0.1533 grm. of the substance gave 35.8 c.c. of nitrogen at a temper-

nature of 14°. 2 and a pressure of 750 mm.

IV. 0.1 457 grm. of the substance gave according to the method of Carius

0.2377 grm. of argentic chloride.

* Barr succeeded in removing all the tin from his preparation in this way.

OF ARTS AND SCIENCES. 10'.)

Calculated for Found

C6U(N11.|,(M1JC1)3. I. II. ill. iv.

Carbon 27.43 28.11 28.06

Hydrogen 5.33 5.12 5.35

Nitrogen 26. G7 27.08

Chlorine 40.58 40.33

Considering the ease with which the substance is decomposed, the
agreement between these results and those calculated from the formula
is as close as can be expected. The method of purification described
above, however, does not always give such satisfactory results. Sam-
ples from two other preparations gave numbers which differed by about
one per cent from those corresponding to the formula, the chlorine
being too low and the other constituents too high by this amount ; but we
ascribe this to partial decomposition of the substance during the purifi-
cation, as both of these specimens were much darker in color than that
which gave the more satisfactory analyses. The objection which might
be founded on these results, that our substance was not a definite ■■'im-
pound, but a mixture of amnionic chloride with some organic body less
rich in nitrogen than the chloride of pentamidobenzol, is removed by
the microscopic examination of the substance, which showed that it
consisted of well developed rhombic crystals without any admixture
whatever, to which may be added the fact that chlorplatinic acid gave
no precipitate with a solution of the substance.

Properties. — The trichloride of pentamidobenzol, as precipitated by
hydrochloric acid from its aqueous solution, is seen under the micro-
scope to consist of rather thick well developed transparent rhombic
plates, with the sharper angles truncated by a single plane. It is color-
less when first prepared, but soon turns pink or gray, and on 1
standing passes through dark brown to black, a change which is much
accelerated by heat; yet even after it has become nearly black, if ex-
amined with the microscope, it is found to retain unaltered its crystal-
Ike form, and the individual crystals are still transparent, although
very much darkened, so that this change in color seems to have taken
place only superficially. It is very easily soluble in water, but only
slightly soluble in strong hydrochloric acid, a property which i- useful
in the purification of the substance. The aqueous solution decomp
quickly when heated, a black tar being formed, and a similar change
takes place more slowly at ordinary temperatures, the solution becom-
ing at first clear pink, and then darkening through shades «-t' red to
black. The action of more violent oxidizing agents i^ described below,
and resembles that of the air. The -alt when dry or moistened with

110 PROCEEDINGS OF THE AMERICAN ACADEMY

strong hydrochloric acid is much more stable than a solution, or the
wet salt. The substance is sparingly soluble in alcohol and essentially
insoluble in ether, benzol, or chloroform.

If amnionic hydrate is added with caution to a concentrated solution
of the salt, a black jelly is precipitated ; if the solution is dilute, no
precipitate is formed, but the liquid takes on a pink color, which in a
few minutes passes through purple and violet blue to a yellowish green ;
from this solution hydrochloric acid precipitates black flocks, which
redissolve in an excess of the acid, or on the addition of amnionic hy-
drate. Sodic hydrate acts on the chloride in the same way. The free
base seems to be insoluble in ether; at any rate, nothing was extracted
by this solvent from an aqueous solution of the salt, to which an excess
of sodic hydrate had been added. It is evident from the observations
just described, that the isolation of the free base would be a matter of
great difficulty.

The foregoing results correspond to those of Barr in every respect
except the crystalline form of the chloride, which he describes as white
needles as fine as hairs. He also made the pentacetyl compound
C6H(NHC.2H30)5.

Chlorplatinic acid with a solution of the trichloride of pentamido-
benzol gives no precipitate at first, but after some time a slimy brown
substance is deposited on the sides of the test-tube, evidently a product
of oxidation. We are inclined to think, however, that the oxidizing
agent is the oxygen of the air rather than the chlorplatinic acid, as a
similar result was obtained by the action of picric acid. Benzaldehyd
added to a solution of the salt also gave a tarry product.

When the trichloride of pentamidobenzol is treated with nitric acid
and ferric chloride, a deep purple solution* is formed looking very
much like potassic permanganate, and when an excess of the ferric
chloride has been added, a brown flocculent precipitate is deposited,
which when dried forms a hard mass of a brown color without crystal-
line form, since under the microscope it appears to be made up of
sandy granules. It is slightly soluble in water, imparting to it a dark
brown color. Preliminary analyses 'gave the following results.

0.1600 grm. of the substance on combustion gave 0.2592 grm. of car-
bonic dioxide and 0.0650 grm. of water.

0.1685 grm. of the substance gave according to the method of Carius
0.0797 grm. of argentic chloride.

* Compare this with the action of air on a solution of the salt or of the free
base described earlier in this paper.

OF ARTS AND SCIENCES. HI

Found.

Carbon 44.1 S

Hydrogen 4.51

Chlorine 11. GO

We were engaged on the study of this substance when we received
Barr's paper, and accordingly stopped our work on this subject, and
we publish these analyses only because they may be of use to any one
who may continue the investigation of this substance, as it would be
idle to try to derive a formula from them until they have been sup-
ported by further work. Barr also observed the formation of this sub-
stance, but did not attempt to study it carefully, or analyze it.

It seems probable that the pentamidobenzol, in addition to the tri-
chloride, can form a pentachloride, as one of our first preparations, in
which the reduction was carried on at a lower temperature than usual,
yielded a substance which, dried at first in a desiccator over calcic
chloride without any alkaline absorbent, and finally for a short time
in an air bath at 90°, gave the following result ou analysis.

0.1738 grm. of the substance gave 30.9 c.c. of nitrogen at a tempera-
ture of 6° and a pressure of 750 mm.

Calculated for Cr,H(NH3Cl)6. Found.

Nitrogen 2U.»0 21.37

The appearance of Barr's paper has prevented us from making an-
other attempt to prepare this substance. If it is a definite compound,
it is a very unstable one, as it gave off a distinct smell of hydrochloric
acid.

to

Trianilidodinitrobenzol, CCH( C6H5NH)3(NOJ )8.

This substance can be made by heating for a short time aniline with
tribromdinitrobenzol in the proportion of six molecules of the base to
one of the bromine compound. It seems that the two substances do
not act on each other in the cold, or, if there is any action, it is a very
slow one, and in this respect the dinitro compound differs from the tri-
bromtrinitrobenzol, which acts on aniline quickly at ordinary tempera-
tares. The product was purified by crystallizing it from alcohol till
it showed the melting point 179°, dried at 100°, and analyzed with
the following results.

I. 0.1820 grm. of the substance gave on combustion 0.4847 grm. of
carbonic dioxide and 0.0760 grm. of water.
II. 0.1906 grm. of the substance gave 26 c.c. of nitrogen at a tempera-
ture of 12° and a pressure of 762 mm.

112 PROCEEDINGS OF THE AMERICAN ACADEMY

III. 0.1815 gr. of the substance gave 26 c.c. of nitrogen at a tempera-
ture of 17° and a pressure of 743 mm.

Calculated for Found.

C6H(NHC8H5)3(N02)2. I. II. III.

Carbon 65.31 65.13

Hydrogen 4.30 4.64

Nitrogen 15.88 16.24 16.24

Properties. — The trianilidodinitrobenzol, when crystallized from
alcohol, forms vivid orange-red needles, which under the microscope
appear as long prisms terminated by a single plane at angles not very
far from right angles to the sides, occasionally groups of radiating
needles are observed also. When crystallized from ether, it forms
irregularly spherical crowded groups of short stout radiating prisms
with blunt ends made up of several planes, a form which is decidedly
characteristic. It melts at 179°, and is essentially insoluble in either
cold or boiling water, or in ligroine ; slightly soluble in cold alcohol,
freely in hot ; moderately soluble in ether, carbonic disulphide, glacial
acetic acid, or acetone ; freely soluble in benzol, or chloroform, but
the substance is deposited from the two solvents last mentioned in an
amorphous form, so that hot alcohol is the best solvent for it. Strong
hydrochloric acid, either hot or cold, has no perceptible action on it.
Strong nitric acid dissolves a very little, with a pale yellow color ;
strong sulphuric acid dissolves rather more of the substance, with a
stronger yellow color. Upon heating it with either of these last two
acids it is charred. It is therefore, as was to be expected, less basic
in its properties than the triamidodinitrobenzol.

OF ARTS AND SCIENCES. 113

XL

CONTRIBUTIONS FROM THE PHYSICAL LABORATORY OF THE
MASSACHUSETTS INSTITUTE OF TECHNOLOGY.

XXXIII. —THE STRENGTH OF THE INDUCED CURRENT
WITH A MAGNETO TELEPHONE TRANSMITTER AS IN-
FLUENCED BY THE STRENGTH OF THE MAGNET.

By Charles R. Cross and Arthur S. Williams.

Presented November 14, 1888.

It is a well-known fact in practice, as well as an evident consequence
of theoretical considerations, that the effectiveness of a magneto tele-
phone when used either as a transmitter or as a receiver varies with
the strength of the magnetism of the core. But the relation of the
one to the other has never been studied, so far as we are aware.

Our investigations include a study of the changes in strength of the
current produced by a magneto transmitter under varying conditions
of magnetization, and of the magnitude of the momentary changes in
the magnetic condition of the core of the receiving telephone when
subjected to the action of undulatory or other brief currents, as influ-
enced by the strength of the primitive permanent magnetization of the
core. The present paper contains only the results of a series of ex-
periments relating to the first of these, that is, to the effect of varying
strength in the magnet of the transmitter, the study of the allied
problem of the receiver being still in progress.

The apparatus employed consisted of a cylindrical bar of soft iron
about 4| inches in length and £ of an inch in diameter, around one
end of which was placed a coil of fine wire similar to thai used in ordi-
nary telephonic practice. The resistance of this coil was 100 ohms.
It was placed in circuit with a ballistic mirror galvanometer, from
whose deflection the momentary current produced in the coil by any
variation in the strength of the core could be determined. The dia-
phragm, which was in all cases 2-& inches in diameter, was in its usual
place opposite the end of the magnet about which the wire coil WM
wound, and about Tjfr of an inch from that end. By means of a rod

VOL. XXIV. (N. S. XVI.) 8

114

PROCEEDINGS OP THE AMERICAN ACADEMY

carrying a cam moved by a weight, a rapid inward push of definite
amount was given to the diaphragm, thereby inducing a current in the
coil already referred to, and so deflecting the needle of the ballistic
galvanometer. The soft iron bar was also surrounded by a second
helix, through which was passed a current from a storage battery,
serving to magnetize the core. A tangent galvanometer inserted in
this circuit gave the strength of the magnetizing current. A mag-
netometer placed in the prolongation of the axis of the core, which last
occupied an east and west position, made known the relative strengths
of the field produced by the core under different conditions of magnet-
ization.

Corresponding observations of the magnetometer reading, and of
the current induced when the diaphragm was moved by the cam, were
made throughout a widely varying range of strength of field, and the
results were represented graphically by constructing a series of curves
in which ordinates represent the relative strength of field, and abscis-
sas the current due to a given predetermined throw of the diaphragm
(about T^ of an inch), as ascertained from the readings of the ballistic
galvanometer.

o

s

FERROTYPE IRON.

K «> i» 60 70

INDUCED CUfiRCHT.

Fig. 1.

OF ARTS AND SCIENCES. 11. "»

One of these curves is shown at 1, Figure 1, the core in this cum-
being a cylindrical bar of Norway iron 4| inches long and \ of an inch
in diameter, and the diaphragm an ordinary disk of ferrotype iron •_'1'J
inches in diameter and yj^ of an inch thick (No. 31 I». W. G.).

Table I. gives the data from which Figure 1 was constructed. The
strength of field is given in terms of the tangents of the angles of de-
flection of the magnetometer needle. The induced current is given in
arbitrary units, as only relative values are needed. A determination
of the value of the deflections was made by observing the excursion
due to the discharge of a condenser through the ballistic galvanometer,
and it was found that the abscissa 100 on the curves corresponds to a
sudden discharge of approximately 0.00000097 of a coulomb through
the coils of the galvanometer.

31.

TABLE I.

i, Norway

Ikox. — Diaphragm, Disk of Ferrotype Iron, N

Strength of

Induced Strength of

Induced

Field.

Current. Field.

Current.

. . .

0.7 .211

20.5

.016

3.3 .229

19.8

.044

12.0 .248

19.2

.058

19.3 .270

18.7

.089

23.3 .302

18.0

.118

27.0 .342

16.9

.141

26.8 .390

16.1

.132

26.6 .454

14.7

.146

25.3 .530

13.8

.164

23.8 .625

12.7

.182

22.5 .773

12.0

.196

21.6 1.014

11.5

Cores of Bessemer steel and of untempered soft steel were also used,
with results given in Tables II. and III.

TABLE

II.

ESSEMER

Steel. — Diaphragm,

Disk

of Ferrotype Iron,

•ength of

Induced

Strength of Induced

Field.

Current.

Field. Ciirri'iit

.005

2.0

.279 L7.8

.030

7.7

.:;:::: 16.7

.082

20 7

.390 1-"'"

.137

26.7

.507 18.6

.160

24.8

.<;.'.". 11.7

.191

21.2

.748 10.8

.213

19.3

.'jut 9.7

.248

18.6

1 099 8.7

116 PROCEEDINGS OF THE AMERICAN ACADEMY

TABLE III.

Core, Untempeeed Soft Steel. — Diaphragm, Disk of Ferrotype Iron,

No. 31.

Strength of
Field.

Induced
Current.

Strength of
Field.

Induced
Current.

.031

8.2

.275

18.0

.049

12.0

.321

16.5

.089

22.0

.388

15.2

.137

25.9

.473

13.8

.157

24.9

.618

12.7

.171

22.9

.776

11.0

.194

21.1

1.011

9.5

.216

19.7

An examination of Figure I, as well as of the various curves fol-
lowing it, will show that the effect of increasing the strength of the
magnet of the transmitter is in all cases to cause at first a rather rapid
increase of the strength of the induced current, which later increases
less rapidly, rising soon to a maximum value, from which it falls off,
at first rapidly, and afterwards more and more slowly as the strength of
the field is further increased. We proceed to consider the explana-
tion of these results.

It is evident that three distinct sources of variation exist to affect
the current furnished hy a magneto transmitter as the strength of the
magnet is increased. First, the direct effect of the increased strength
of field in which the diaphragm moves is to increase proportionally the
strength of the induced current, since it increases correspondingly the
rate of change in the number of lines of force enclosed by the coil of
the instrument ; second, an approach toward saturation of the magnet,
so far as it alone is concerned, will tend to diminish the induced current,
on account of the smaller variation in the strength of the pole due to a
given motion of the diaphragm ; and, third, the nearer approach toward
saturation of the diaphragm will have the same tendency.

The rapid rise in the induced current at the beginning is of course
due to the predominating influence of the increasing strength of the
field in which the diaphragm moves, as both core and diaphragm are
then but slightly magnetized. The subsequent changes in the current
must be explained by a consideration of the increasing magnetization
of either the core or the diaphragm, or both.

It will be seen by comparing Tables I., II., and III., that the value
of the maximum induced current for a given excursion of the diaphragm

OF ARTS AND SCIENCES. 117

is approximately the same with all the three cores used, and that the
same is true as to the strength of held corresponding to this maxi-
mum current. Moreover, saturation curves constructed for the several
cores showed that in all cases the magnet was still very far removed
even from half-saturation when the maximum induced current was ob-
tained. From these facts it appears that the degree of saturation of
the magnet is practically unimportant, so far as the general results
shown in Figure 1 are concerned.

It remains to observe the part played by the increasing magnetiza-
tion of the diaphragm. Other things remaining the same, as this ap-
proaches more and more closely towards saturation the increase in the
number of lines of force passing between it and the magnet on the
approach of the diaphragm to the magnet must become smaller and
smaller, and this change will tend to oppose the effect of the increased
absolute strength of the magnetizing force. The small mass of the
diaphragm will evidently cause it to show the effect of an approach to
saturation while the core is far below that condition. And such an ac-
tion will clearly explain the observed changes in the current strength.

In order to test this matter still further, the experiment was tried of
varying the mass and material of the diaphragm.

The results are shown by the curves in Figures 1, 2, and 3, which
are constructed from the data given in Tables I. to X. ; further results
of the same character are given in Tables XI. and XII.

TABLE IV.

Core, Norway Iron. — Diaphragm, Two superposed Disks of Ferrotype

Iron, No. 31.

Strength of

Induced

Strength of

Induced

Field.

Current.

Field.

Curreut,

.016

5.0

.184

49.3

.054

19.3

.205

50.2

.070

25.7

.229

48.0

.081

28.5

.266

40.3

.089

30.2

.306

31.8

.103

35.0

.342

27.7

.119

38.5

!483

18.7

.133

40.7

.530

11.9

.144

42.7

.607

8.3

.155

45.7

.830

7.0

.169

47.0

1.018

5.0

118

PROCEEDINGS OF THE AMERICAN ACADEMY

moN

Ul U> N «• M 60 .0 UU 90 HU *m 120

vi/ducco cunm/7
Fig. 2.

C
ft;

V3 ytQ

iO at, ~$Z ip 60 60 70 60 so

induced current.
Fig. 3.

OF ARTS AND SCIENCES.

119

TABLE V.

Core, Norway Iron. — Diaphragm, Three superposed Disks of Ferrotype

Iron, No. 31.

Strength of
Field.

Induced
Current.

.000

0.5

.018

3.2

.047

8.0

.079

14.2

.084

14.7

.096

17.0

.110

19.7

.118

21.8

.135

25.5

.151

28.3

.175

33.0

.213

39.0

.249

46.3

.277

49.8

.313

55.3

.329

58.7

TABLE VI,

Strength of

Induced

Field.

Current.

.346

59.7

.366

60.7

.390

61.3

.416

63.0

.445

62.3

.479

61.0

.514

58.5

.550

55.7

.591

52.4

.637

48.7

.659

47.3

.765

41.7

.854

36.7

.983

30.0

1.180

22.8

Core, Norway Iron. — Diaphragm, Sheet Iron, No. 21.

:ength of
Field.

Induced
Current.

Strength of
Field.

Induced
Current.

• > •

0.3

.456

96.7

.016

3.1

.477

94.0

.058

13.7

.499

92.5

.082

20.3

.521

90.5

.102

25.2

.536

88.3

.125

31.2

.573

86.7

.150

37.8

.584

84.0

.178

45.5

.637

81.0

.213

54.7

.675

76.7

.249

63.8

.724

72.8

.296

75.5

.784

.348

86.2

.851

.378

92.3

.949

67.8

.416

96.7

1.001

49.1

.437

97.8

120 PROCEEDINGS OP THE AMERICAN ACADEMY

TABLE VH.

Coke, Norway Iron. — Diaphragm, Sheet Iron, No. 22.

Strength of

Induced

Strength of

Induced

Field.

Current.

Field.

Current

1.7

.410

942

.018

4.7

.431

92.0

.047

11.7

.452

88.3

.075

19.7

.473

85.0

.088

22.3

.494

81.2

.100

26.0

.516

77.5

.123

32.5

.541

74.7

.146

39.5

.566

71.3

.171

48.2

.601

67.7

.194

54.2

.635

64.0

.218

61.3

.659

60.0

.240

66.9

.719

56.7

.262

73.2

.779

53.0

.292

81.3

.851

48.0

.313

87.7

.946

44.0

.342

92.7

1.072

39.3

.372

94.3

1.091

38.3

TABLE

VIII.

Core, Norway Iron. — Diaphragm, Sheet Iron, No. 23.

Strength of
Field.

Induced
Current.

Strength of
Field.

Induced
Current.

...

1.7

.272

109.3

.026

10.0

.289

113.8

.068

26.8

.310

119.3

.086

33.0

.339

118.6

.096

39.0

.382

114.1

.116

47.0

.420

102.9

.139

66.3

.458

82.2

.162

65.8

.499

68.7

.176

72.3

.629

67.6

.194

78.3

.582

48.0

.216

86.4

.652

39.9

.229

90.0

.751

30.8

.242

96.3

904

23.7

.255

101.8

1.011

19.0

OP ARTS AND SCIENCES.

121

TABLE

IX.

3RE, Norway

Iron. — Diaphragm,

Steel, No. 26,

Untempere

Strength of
Field.

.004

Induced
Current.

7.7

Strength of
Field.

.281

Induced
Current.

66.7

.019

11.5

.313

G3.8

.054

20.5

.342

60.4

.095

32.2

.380

56.3

.105

34.5

.422

50.5

.128

42.0

.483

44.7

.153

48.3

.530

39.0

.178

55.7

.5JB4

35.3

.200

G0.3

.652

32.0

.227

64.3

.810

28.2

.236

67.0

TABLE

X.

1.043

24.3

Core, Norway Iron. —

Diaphragm, Steel, No. 26,

, Tempered,

Strength of
Field.

Induced
Current.

Strength of
Field.

Induced
Current.

.002

7.8

.344

36.3

.019

10.4

.400

32.0

.008

18.5

.406

26.9

.096

23.8

.545

21.7

.128

30.3

.642

17.7

.151

32.4

.662

17.3

.180

37.0

.732

15.2

.214

40.2

.816

13.8

.246

40.0

.854

13.0

.272

39.5

1.025

10.0

.304

37.9

TABLE

XI.

ire, Norway

Iron. — D

IAPHRAGM,

Steel, No. 22,

Untemperei

Strength of
Field.

Induced
Current.

Strength of
Field.

Induced
Current.

.007

6.2

.272

86.7

.023

8.0

.302

34.8

.070

17.3

.340

33.0

.103

23.0

.392

29 7

.121

26.4

.458

•J';.;:

.139

30.1

.538

28 2

.157

32.8

.666

28.0

.176

35.0

.HIT

21.7

.202

37.1

.77::

L9.8

.224

34.9

1.046

17.6

.246

37.0

122 PROCEEDINGS OP THE AMERICAN ACADEMY

TABLE XII.
Core, Norway Iron. — Diaphragm Steel, No. 30, Tempered.

Strength of Induced Strength of Induced

Field. Current. Field. Current.

.007 6.2 .272 36.7

.023 8.0 .302 34.8

.070 17.3 .340 33.0

.103 23.0 .392 29.7

.121 26.4 .458 26.3

.139 30.1 .538 23.2

.157 32.8 .566 23.0

.176 35.0 .647 21.7

.202 37.1 .773 19.8

.224 ... 1.046 17.6

.246 37.0

TABLE XIII.
Core, Norway Iron. — Diaphragm, Steel, No. 30, Untempered.

Strength of Induced Strength of Induced

Field. Current. Field. Current.

.005 5.2 .331 31.2

.031 11.7 .376 28.9

.082 24.2 .437 26.0

.103 29.7 .512 23.3

.130 37.7 .632 20.4

.160 41.9 .735 19.7

.200 38.3 .819 18.0

.231 36.2 .956 16.3

.272 35.2

Curve 2, Figure 1, represents the results when two of the ordinary
ferrotype diaphragms were superposed, forming a diaphragm of double
thickness, and Curve 3 of the same figure shows the results when three
such diaphragms were superposed. Each diaphragm was 0.01 of an
inch thick. The curves of Figure 2, marked 21, 22, 23, respectively,
show the results of similar experiments with diaphragms of sheet iron
whose thickness was 0.030, 0.027, 0.024 of an inch, respectively (Nos.
21, 22, 23, B. W. G.). Figure 3 shows the results when a steel dia-
phragm 0.017 of an inch thick (No. 2G) was used, the curve U being
that for untempered steel, and T that for tempered steel. Steel dia-
phragms respectively 0.026 and 0.012 of an inch thick (Nos. 22
and 30) gave similar results, as will be seen from Tables XL, XII.,
and XIII.

OF ARTS AND SCIENCES. 123

An inspection of these curves shows immediately that the greater
the strength of the field required to saturate the diaphragm, the <reater
is the strength of the field at which the maximum current occurs.
Thus in Figure 1 the maximum current with Curve 1 corresponds to
a strength of field of about 12 units of the scale used, while with Curves
2 and 3 the corresponding strengths of field are 20 and 43 units re-
spectively. Also in Figure 2 the maximum currents will be seen to
correspond to greater strengths of field in proportion to the thickness
of the diaphragm, and in Figure 3 similar though less marked results
hold for tempered as compared with untempered steel of the same
thickness.

It would also be expected that the value of the maximum current
would be greater with a thick than with a thin diaphragm. This was
usually the case in our experiments. Thus the Curves 1, 2, 3, of Fig-
ure 1 give maximum currents of 27.5, 50.2, and G2.G units respectively.
Results of a similar nature are shown by Figure 3, the maximum cur-
rent with the untempered diaphragm being far greater than with the
tempered one. Curve No. 23 of Figure 2 is apparently an exception.
It is probable, however, that this is in appearance only, and that the
three curves of that figure are not strictly comparable with one another.
The rigidity of the diaphragms here used, especially of the thicker
ones, is considerable, and any slight yielding of the supports of the rod
which carried the cam would prevent the actual throw of the diaphragm
from being as great when this had considerable thickness, and would
greatly diminish the strength of the current produced.

The peculiarity of Curve 2, as compared with 1 and 3, is probably
caused by the want of both magnetic and mechanical continuity in the
material of the multiple plate formed by the several diaphragms used.

In those cases where steel diaphragms were employed, there was
always a notable induced current, even when the reading of the mag-
netometer was zero. This was probably due to a slight residual mag-
netization of the diaphragms.

The results stated in this paper may serve to explain a phenomenon
which has seemed somewhat obscure. Frequent attempts have been
made to increase the efficiency of a magneto transmitter by polarizing
the diaphragm as well as the magnet, a common way of doing this
being to employ a horseshoe magnet one leg of which >- in con
with the edge of the diaphragm, while the other, about which the coil
is wound, is placed in its usual position opposite the centre. Bui as :i

general rule little or no gain has seemed to result therefr • ■■■ far U

can be judged by the performance of such instruments in actual prac-

124 PROCEEDINGS OF THE AMERICAN ACADEMY

tice. It is quite probable in tbis case tbat the increased approach to
saturation of the diaphragm may have so great an effect as entirely to
prevent the expected improvement-
It will also be seen from our results, that an increase in the thickness
of the diaphragm of a magneto transmitter tends to allow of the use of
a stronger magnet, and for a given amplitude of vibration to produce a
stronger current. But it must be remembered, on the other hand,
that the greater rigidity of the thick diaphragm will diminish this range
of vibration under the action of the voice, a difficulty which may to a
certain extent be remedied by using a diaphragm of large diameter.

Rogers Laboratory of Physics,
October, 1888.

OP ARTS AND SCIENCES. 125

Investigations on Light and Heat, made and published wholly oe in paet with
Appeopeiation feom the Rumfobd Fond.

XII.

AN ACCOUNT OF A NEW THERMOGRAPH, AND OF
SOME MEASURES IN LUNAR RADIATION.

By C. C. Hutchins, assisted by Daniel Edward Owen.

Presented by Professor Trowbridge, March 13, 1389.

During the winter of 1886-87 the writer carried out at Cambridge
a series of experiments with the bolometer and other sensitive heat-
measuring devices, the intention being, if the apparatus could be made
to perform satisfactorily, to attempt the solution of some problem in
radiation. One in particular in mind was that of the radiation from
rocks, as important from a geological standpoint. It was also thought
possible that the work might be made a valuable supplement to Lang-
ley's investigations in lunar radiation.

After two or three months of preliminary experimentation the
matter was abandoned, and the further pursuit of the subject was not
undertaken until October, 1887. In designing a working apparatus
for use in the research thus resumed, some modification of the thermo-
pile suggested itself, and a contrivance was finally adopted constructed
upon the principle embodied in that familiar instrument. This new
device possesses the sensitiveness necessary for accurate and delicate
work, and at the same time is very simple in operation. This last is
an important feature, for in all instruments intended for measuring
small quantities of heat the matter of simplicity is a very grave one.
It hardly needs saying, that the number of the parts in the instrument
should be limited to as few as possible, and that the whole should be
as simple as it can be made. Multiplication of parts leads to inaccu-
racy in performance with almost absolute certainty. It is the element
of simplicity that has, until recently, rendered the ordinary thermopile
the most successful instrument possible to be employed in dealing with
small quantities of heat. The failings of the thermopile, however,
are so numerous and so fatal that it becomes an instrument of pn
ion in the hands of a very few, who must serve a long apprenticeship
in its use.

126 PROCEEDINGS OF THE AMERICAN ACADEMY

Impressed with the number of the difficulties which beset the use of
the thermopile, the writer set about a systematic study of the condi-
tions necessary to their removal. It was soon observed that a great
advantage could be gained by employing a single thermal junction,
instead of many, and by condensing the heat rays upon this one by
means of a concave mirror. All of the advantages of the thermopile
are thus retained, while its most serious failing — the slowness with
which the large mass of metal presenting the sensitive surface parts
with its heat — is obviated altogether. The condensing mirror fulfils
the function of multiplied junctions, while the single union of bars
employed in the new instrument rapidly attains thermal equilibrium
under any series of conditions.

This preliminary proposition having been laid down, — that a single
junction of two metals, suitably supplemented by a concave mirror,
can be made as effective as a larger number of junctions without the
mirror, — various experiments were next undertaken with a view to
ascertaining the best form for the single junction, as well as the most
suitable materials to enter into it. With these ends in view a great
number of unions were made, including some very delicate combina-
tions into which tellurium was introduced. The bars of metal used in
these experiments were made very thin, and in attempting to form
certain junctions great difficulty was experienced ; for when very thin
bars of easily fusible metals are to be soldered together, it is some-
times almost impossible to perform the act properly, without melting
the bars. In such cases the electro-deposition of silver or copper in
the junction succeeds admirably.

As the outcome of these experiments it was learned that the best
results could be obtained from a junction formed by two of the stronger
metals, worked into very thin strips, the opposite ends of the metal
ribbon thus formed being attached to copper rods — the terminals —
with the point of junction midway between. The reasons for such a
construction and the advantages especially to be derived from it are
briefly these. It is plain that a thin strip of metal will not only
take up and part with its heat in a much shorter period of time than
will a thick bar, but that it will also, in a given interval, be raised to
a much higher temperature than the thick bar. The strong metals
are employed for the obvious reason that the brittle metals cannot be
made into very thin leaves and at the same time remain self-sustaining.
Further, thin strips of even such metals as iron and copper, the thermo-
electric force of which is comparatively feeble, brought into conjunc-
tion, give a greater electric current when excited by a feeble radiation

OF ARTS AND SCIENCES. 127

than is obtained from a junction of bars of antimony and bismuth
under similar circumstances. Finally, to employ a condensing mirror
having the junction at its focus is simpler than to multiply the number
of junctions, and, moreover, it adds nothing to the resistance of the
circuit.

Instruments.

The Thermograph.

There will now be described, in detail, the instrument which was
the outgrowth of the principles touched upon above. It is not beyond
the power of one possessed of very ordinary mechanical skill, and who
may be perforce his own instrument-maker in matters like the present.
The measures will be given in inches, since most workmen still employ
that unit.

A cvlinder is turned from vulcanite, or some hard wood. 3 in. lon<»
and 1 in. in diameter. For f in. at one end this cylinder is made of
less diameter by -^s in., in order that it may receive and hold firmly a
piece of brass tubing 3 in. long, 1 in. in external diameter, and ^.y in.
thick. The end of the cylinder carrying the tube will be designated
in what follows as the " front end." The cylinder is next pierced, in
the direction of its length, with two J in. holes lying upon the same
diameter and ^-e in. apart. Through these holes copper rods are firmly
driven, being left to project -J- in. above the front end of the cylinder.
They are allowed to project f in. from the back end of the cylinder,
and are afterwards bent apart somewhat to facilitate the attachment of
screw cups. The front ends of the rods are made somewhat smaller
than the other portion by filing.

The thermal junction is made as follows. Some bits of rather
wide watch-spring are procured, together with some rolled sheet nickel.
A strip of the same width as that of the watch-spring is cut from the
nickel, and, the ends of the spring and of the piece of nickel having
been scarfed with a file, the two are united by means of the strongest
hard solder obtainable. For convenience in handling, it is well to
leave the piece of watch-spring long. The compound strip can now
be made straight upon the edges, and worked thin with a line file.
Trial is occasionally made with a screw caliper, and when the thick-
ness has been reduced to .02 mm. it is best to discontinue. The
solder junction must be very perfect, or it will be pulled apart in the
filing. With great care and patience the thickness of the strip of
metal can be reduced to .01 mm., but there is no compensating ad-
vantage for the labor expended. When finished, the wide strip can

128 PROCEEDINGS OF THE AMERICAN ACADEMY

be cut with scissors into several narrow bands, and preserved in a dry
bottle.

To complete the thermograph, a strip is selected from the stock, laid
across the rods at the front of the cylinder, and secured in place with
a little soft wax. The wax permits the strip to be shifted, until it is
proved by measurement that the junction is exactly midway between
the rods, when the strip is permanently fixed by soldering it to the
rods. A small concave mirror of glass, silvered upon the first surface,
is now fixed to the front end of the cylinder, between the copper rods,
with glue or varnish. The mirror should be \ in. in diameter and of
| in. focus. When in place, the junction in the metal strip should be
exactly at its focus.

The brass tube is carefully blackened within, and is provided with
diaphragms, the openings in which are determined by the character of
the work for which the instrument is intended. The front end of the
brass tube is also furnished with a suitably perforated end-piece of the
same material as the cylinder. For some purposes the tube of the
instrument requires to be made longer than that of the sample de-
scribed above, and for other purposes it may be dispensed with. For
linear measurements, as in spectrum work, it is only necessary to re-
place one or more of the diaphragms by slits at right angles to the
direction of the strip. The apparatus is particularly successful in
dealing with small quantities of heat at a point. When employed for
this purpose, the condensing mirror is not used, the junction being
merely placed behind a small hole in a trif>le screen, formed by two
plates of metal separated by a thick plate of cork or other non-con-
ducting substance.

If for any particular purpose it should be desired to make use of a
strip thinner than can be produced by the methods given above, it can
be prepared after the following manner. A flat plate of copper is
rubbed with graphite and nickel-plated over one half its surface. It
is then dried, rubbed with graphite a second time, and the nickel-coated
end dipped in melted paraffine nearly to the junction of the nickel and
copper. The remaining half of the copper surface can now be plated
with iron from a solution of the citrate, the nickel already deposited
being protected by its paraffine coating. When the entire thin film is
afterward stripped from the copper backing, one half of it will be nickel,
the other half iron, and it can be cut into narrow strips as before.
The writer has to learn that there is any advantage to be gained from
employing strips made in this manner, and they are certainly difficult
of manufacture. In general, thin films of electrically deposited metals

OF ARTS AND SCIENCES. 129

are hard to manipulate, for they possess the elasticity and brittleness
of highly tempered steel, and usually require careful annealing.

The Galvanometer.

The galvanometer intended to accompany the heat-measuring device
should receive as careful attention as the device itself. The writer
had had some previous experience in the construction of sensitive
galvanometers before attempting the research of which the present
article is an account ; nevertheless, several different forms were tried
before one was found exactly adapted to the thermograph. The re-
quirements to be met were, in brief, first, considerable sensitiveness,
and secondly, rapidity of working, — qualities which are to some ex-
tent incompatible, but which have, as far as possible, been combined
in the instrument in use. It would have been possible to produce an
instrument much more sensitive than the one actually employed, had
the rate of working been left out of account; but nothing is more
annoying than to be obliged, when time is precious and opportunities
for observation are perhaps few, to wait for a galvanometer needle
to come to rest ; consequently the element of time has had the first
consideration.

The galvanometer adopted for use with the thermograph has four
coils, each containing about four feet of wire 0.063 in. in diameter.
The wire was twice shellacked and baked before beinjj wound. The
coils are contained between brass plates, which constitute the body of
the instrument, the inner faces of the two plates, which are in contact,
being grooved for the passage of the staff of the needle and its sup-
porting silk fibre. The needle thus swings in an entirely closed space
but little larger than itself, and needs no damping vanes to bring it
quickly to rest. Two minute scales of mica are attached to the needle
to prevent it from being turned completely round. The astatic system
has the magnets fixed upon a very slender aluminum staff two inches
in length. The magnets are from ] in. to g in. in length, and very
slender. The mirror is flat, and | in. in diameter. Much of the suc-
cess of the instrument depends upon the perfection of the mirror. It
is difficult to obtain perfectly plane mirrors, and. moreover, those to
be had in the market are too heavy. They can be made, of great per-
fection, by selecting a perfectly flat piece of thin plate glass, cementing
it to a thicker piece of glass for a support, and then grinding it down
to the thickness of stout letter paper. A very thin sheet of glass is
thus obtained, from which circles of the required size can be cut with
a diamond. The glass being now silvered upon the Burfaoe, the polish

VOL. XXIV. (\-. S. XVI.) 9

130 PROCEEDINGS OP THE AMERICAN ACADEMY

of which was not destroyed by the grinding, we have a first surface
reflector which will bear a high magnifying power if the mirror be not
strained in mounting.

The scale is read in a novel manner. The cell in which the mirror
swings is closed by a bit cut from the centre of an ordinary spectacle
lens of from 36 to 48 inches' focus. The scale is located as in the
ordinary arrangement of telescope and scale, but instead of the tele-
scope we have the following device. A tin or paper tube is provided,
and in it, at a distance of five inches from one end, is placed a lens,
three inches in diameter, and of six inches' focus. The end of the
large tube has a draw-tube, which is closed by a metal plate provided
with a peep-hole at its centre. The peep-hole may be made to occupy
the focus of the lens by sliding the draw-tube backward or forward as
required. A horse-hair or very fine wire is stretched across the tube
in the focus of the lens opposite the draw-tube, so that it may be
clearly seen by the eye applied to the peep-hole. This arrangement
is to be mounted in the place occupied by the ordinary telescope,
which it far surpasses both in definition and in the breadth of the field
of view. The whole apparatus when in position opposite to the gal-
vanometer constitutes a telescope, of which the lens in the galvanome-
ter is the object-glass, the large lens in the tube the eye-piece, and
the scale the object viewed. The definition is quite remarkable, the
fibre of the paper upon which the scale is ruled showing clearly, as
though seen through a magnifying-glass.

The details of construction of the entire apparatus having thus been
outlined, it remains to speak of the action of the whole. As an ex-
ample of the accuracy that may be expected in the performance of the
instrument, the following test may be taken.

October 23d, 1888, the sun was taken as a source of heat, the rays
being reduced by passage through a small opening. The following
galvanometer deflections were obtained : —

Divisions. Divisions.

First 162 Sixth 164

Second .... 162 Seventh .... 162

Third ..... 161 Eighth 161

Fourth .... 162 Ninth 161

Fifth 162 Tenth 162

The observations were taken rapidly, occupying less than ten min-
utes in all.

No great effort has been made to secure superior sensitiveness, yet
the heat from the face of a person at a distance of fifty feet is measur-

OF ARTS AND SCIENCES. 131

able. A comparison was made between tbe thermograph and a ther-
mopile of forty-eight couples. Using the same galvanometer for each,
the thermograph proved about twelve times as sensitive as the other
instrument. The limit of sensitiveness could be very much extended,
if it were desired, by the use of a more delicate galvanometer. As
actually adjusted, the galvanometer employed gives a deflection of
one scale division for 0.00000007 Ampere, the period being ten
seconds. The constant of Langley's galvanometer was 0.0000000013
Ampere, or fifty times less.

Radiation from Rocks.

As experiments upon the radiating power of rocks and minerals
were the first in which the new instrument was employed, they will be
first discussed.

The want of a standard of radiation which should be more satis-
factory than a lamp-black-coated surface was seriously felt in these
experiments. Lamp-black is a hygroscopic substance, and radiates
differently when in different states, unless it be applied with sufficient
varnish to make it water proof. But if it be mixed with the smallest
possible amount of varnish, and applied to a surface, then the radiating
power is different from that of the same surface coated by camphor
smoke. Also, it appears to radiate differently when applied in the
same manner to different substances. The writer has always found
ordinary white pumice-stone a better radiator than the same black-
ened. There is no other instance in which so variable and imperfect
a standard is tolerated. The writer has employed as a working
standard in these experiments a piece of pure quartz finely ground
with emery but unpolished. Such a standard seems to possess many
advantages. Quartz is readily obtainable in a pure state. It is im-
perishable and unalterable at any ordinary temperature. It has the
great merit of perfect definiteness of composition.

The measures of radiation from the rocks were made as follows.
Each specimen had a wooden handle adapted to it, for convenience in
handling when hot. The rocks were heated in an oven intended for
drying chemical precipitates. The oven was of heavy iron, and was
heated from beneath by numerous small gas jets. It was found that.
after the gas had been lighted about an hour, the temperature of the
oven became constant, and remained so for any length -.1' time.
rock under experiment and the quartz standard were placed in tl"'
oven together, side by side, and after the lap- of an hour they were
removed, first the quartz and then the rock being quickly taken from

132

PROCEEDINGS OF THE AMERICAN ACADEMY

the oven, presented before the opening in the thermograph, and the
deflection noted. The specimens were then returned to the oven, and
some time allowed to elapse before the experiment was repeated. In
this way as many measures were obtained from each specimen as
could be made in the course of a day, i. e. from ten to thirty. The
method seems far from perfect, but it was the best of several that
were tried.

To reduce the results to the lamp-black standard, a very careful
comparison was made between the radiating powers of blackened and
unblackened faces of the quartz. The temperature at which the
measures were made was near 100°. The results are given in the
following table, the radiation from a blackened surface of quartz being
taken as 100.

Name.

Locality.

Character.

Radiating
Power.

Lava

Lava

Lava

Lava

Lava

Pumice

Lava

Scoria

Pumice

Trap

Trap

Obsidian

New Red

Sandstone
Flexible

Sandstone
Slate
Slate
Feldspar
Gypsum
Iceland Spar
Mica

Serpentine
Quartz
Amphibnle
Tremaolite
Pyrite

White Marble
Mica Schist
Granite

Hawaii

Vesuvius

D'Auvergne

D'Auvergne

Hawaii

Vesuvius
Vesuvius
Vesuvius
Nova Scotia
Mt. Kineo

Pembroke

Georgia
Pennsylvania
Georgia
Topsliam, Me.

Hebron, Me.

New York

Imperfectly fused. Black
Dense. Gray
Basaltic. Dark
Scoria. Light red
Well fused. Red
Common. White
Vesicular. Black
Very light and porous. Black

Red
Dark

Light gray
Black

91.2
90 4

81.8

86.5

91.9

71.3

86.3

91.6

89.1

83.2

94

88.5

89.3

85.9

89.8

84 1

88.5

86

85

76.3

891

90.8

91.4

85.5

76.4

86.7

80.8

93.5

Red
From coal beds, with pyrite.

Red

Pure white
Pure white
Pure

Common. Green
Pure. Transparent
Common. Green

Pure white

Fine-grained.

An inspection of the table shows a remarkable uniformity in radiat-
ing power for materials so diverse in character. The large number of
volcanic rocks was selected, because it seems probable, from its physi-

OF ARTS AND SCIENCES. 133

cal features, that the rocks of which the surface of the moon is made
up are of volcanic character.

Lunar Radiation.

While the light received from the moon has been frequently meas-
ured by many different observers, the more important problem of its
dynamic radiation has been solved by but a single person. Observers
previous to Langley can pretend to so little accuracy in their results,
that our knowledge of lunar radiation may be said to rest upon his
authority alone. This condition does not arise from any lack of in-
terest in the subject, — on the contrary the problem has always been a
fascinating one, — but is due to the fact, that, previous to the invention
of the bolometer, no instrument existed capable of dealing accurately
with so small an amount of heat as the moon affords. Great interest,
therefore, cannot fail to attend the results of careful observation, made
with competent instruments, and tending to increase our knowledge of
the subject.

The first trial of the new thermograph on the moon was made on
January 27, 1888. The apparatus, tested at that time, has been regu-
larly employed since, and has performed so satisfactorily that it has
scarcely been altered from its original form. The arrangement of the
thermograph for the measurement of lunar radiation is this. An
oblong box of pine wood carries at one end a silvered glass mirror,
0.196 metre in diameter, and of one metre focus. This mirror is ad-
justable by screws passing through the end of the box and bearing
upon its back. Near the other end of the box is fixed a grooved
block, upon which the thermograph may be placed and secured. The
block is also provided with adjusting screws, so that the thermograph
may be made to point exactly at the centre of the mirror. The mirror
is inclined, so that the moon's image is thrown into the opening of tin-
thermograph, and brought to a focus upon the strip. The whole
arrangement is that of a Herschel's telescope with the thermograph in
place of an eye-piece.

To secure accuracy in pointing, the opening in the thermograph is
made of such a size as just to receive the cone of rays from the large
mirror, and, a rim of white paper being glued about the opening, the
slightest departure from a central position of the cone of rays is an-
nounced by the appearance of light upon this paper. A window oi
plate glass has its place in the side of the box, just in front of tin-
position occupied by the thermograph, and a person looking through
this can readily direct the image of the moon into the orifice of tin-

134 PROCEEDINGS OF THE AMERICAN ACADEMY

thermograph. A telescope, with cross wires, is also attached to the
box, and may be used to direct it, but the former method is much the
more satisfactory.

The eutire apparatus is mounted upon a massive telescope tripod.
The mounting has tangent screw motions, and is provided with a ver-
tical circle for reading the altitudes. An hour or two before it is in-
tended to observe, the apparatus is mounted in the open air, in order
that it may have time to acquire the external temperature. Two
strands of heavy electric light wire pass from the thermograph to the
laboratory, where they are attached to the galvanometer, which is
mounted upon a firm pier.

For some reasons the effect of phase should precede all other con-
siderations as a subject of investigation, but it has been postponed
until later. The subjects that have engaged our attention, up to the
present time, may be for the most part classed under two heads :
1. The comparative intensity of radiation of the full moon and of the
sun. 2. The absorptive effect of the earth's atmosphere upon the
radiation of the moon. These two topics will be considered in turn.

It is obviously impossible to confine the observations to the time of
full moon without giving undue time to the investigation ; hence the
effect of phase demands to be considered. To avoid the introduction
of errors from this source the following precautions have been taken :

1. To confine the observations to the three days nearest full moon.

2. To measure always the radiation from a constant area of the moon's
surface. To fulfil the conditions of the second jitrecaution, there is
placed, close to the strip of the thermograph, a diaphragm, the open-
ing in which is much smaller than the moon's image which falls upon
it. Thus there may be a large change of phase, and the opening still
remain full of the moon's image. This arrangement would completely
remove the difficulty, provided that the moon reflects like a flat disk,
as Zblner thinks, — which is not at all certain. It is believed, how-
ever, that, when taken in connection with the first precaution, the out-
standing error cannot be large. The method has the disadvantage of
reducing the available heat by about one half; but, as it is, the heat
is sufficient to drive the needle off the scale with the galvanometer in
its most sensitive condition. The controlling magnet is set so that a
deflection of between one and two hundred scale divisions is obtained,
corresponding to a period of between five and ten seconds' vibration
of the needle.

OF ARTS AND SCIENCES.

135

One observer being stationed at the galvanometer, and a second at
the instrument out of doors, at a signal from the first, the latter brings
the image of the moon into the opening of the thermograph, and the
deflection occasioned is noted and recorded by the observer at the gal-
vanometer. In this manner a set of ten readings is taken, and, the
altitude of the moon being measured at the beginning and the end of
the set, the mean of the ten deflections is taken as the deflection cor-
responding to the mean altitude. About six minutes are required to
make a set of the readings.

The following series, selected at random from the observation-book,
will serve as an example of the results obtained.

Date. — February 24, 1888.

Mean Altitude.

Mean Time.

Deflection.

Mean Deflection

62° 40'

7 h. 34 m.

....

• • •

• • • •

• ■ • •

171.5

• • *

• . . .

. .

172

• • •

• • • •

. .

168.5

• • •

• • • •

, .

166

. • •

• • • •

• •

171

. . .

• • • •

• .

174

• • •

• • • •

• • •

172

• • a

• • • •

• .

175

. a •

• • ■ •

» • •

173.2

a a •

• • *• •

. .

175

a . a

• • • •

• • ■

• a • •

171

8

The probable error of a single observation, when the moon is the
source of heat, is considerably greater than is the case with an arti-
ficial source, or even with the sun. The cause is, in part, due to in-
visible clouds of vapor floating in the atmosphere, as Langley thinks.
It has been observed that immediately after rain this effect is particu-
larly noticeable, even when the sky appears perfectly clear to the eye.
There is another cause, which doubtless exerts a large influence ;
that is, the irregular radiation of different portions of the moon's
surface. It has been found* that, photographically, the mean bright-
ness of the dark region is to that of the bright rejrion in the ratio of
55 to 100. We might be prepared to expect large differences in the
radiant energy of the respective regions, and the average radiation
may therefore be more closely attained from a series in which the in-
dividual readings vary considerably, than from a more perfectly con-
cordant set.

* Annals of Harvard College Observatory, Vol. XVIII. No. IV.

136 PROCEEDINGS OF THE AMERICAN ACADEMY

Having obtained a sufficient number of sets, as perfect as may be,
we have next to effect a comparison between the deflections produced
by the moon, and the deflections brought about by the sun under simi-
lar circumstances of observation, — such as altitude, height of barom-
eter, etc. The heat of the sun's rays must be reduced in some
manner in order to bring them within the range of the same instru-
ment employed in measuring the feeble radiation of the moon. Both
Lord Rosse and Langley accomplished the reduction by passing the
sun's rays through a small ojjening, and placing the heat-measuring
appliance in the diverging beam.

The enormous differences that appear in the results of observers
who have measured the light of the moon, and compared it with sun-
light, put us upon our guard against sources of error, and impressed
upon us the necessity of going over the ground by independent
methods. Let us consider the probable sources of error in the above
mentioned method of comparison : —

1. Errors of observation. These are small, and can be reduced to
any extent by repeating the observations.

2. Absorption by the mirror, used in condensing the lunar rays.
The mirror is always kept perfect by frequent silvering, and a flat
mirror, silvered by the same process, is employed in reflecting the
sun's rays to the small opening.

3. The plate of metal in which is the perforation becomes a
source of heat when the sun's rays are thrown upon it, and would
increase the heat indication. This effect is avoided by placing the
thermograph so far from the metal plate as to render its influence
practically imperceptible.

4. Effect of diffraction at the opening. Unknown.

As an example of the results obtained by this method, let us take
the observations of April 25 and 26, 1888.

On the night of April 25, the deflection produced by the moon at an
altitude of 42° was 184.8 scale divisions.

We have also :

Semidiameter of condensing mirror . . 95 mm.
Focus " " . . 1000 mm.

Semidiameter of moon, augmented . . 17'.

Hence we have, for the concentration of the moon's rays,

^_ opq 1

(1000 X sinlT')2 ~

OF ARTS AND SCIENCES. 137

Had the moon's rays, uncondensed, acted upon the thermograph,
the deflection would have been :

1S4 8

.„, ' = 0.5003 scale divisions.

o09.1

On the following morning, the altitude of the sun being 42°, its
rays were reflected through a small opening, and being received by
the thermograph, placed at a distance, a deflection of 147.16 scale
divisions was found to be the mean of 30 readings. The opening
through which the sun's rays were reflected was a carefully reamed
hole in a metal plate. Its diameter, being measured under the micro-
scope, with the aid of a stage micrometer which purported to be
divided into tenths and hundredths of a millimeter, was found to be
1.67 mm. A paper screen received the spot of light, formed by the
rays from this opening. The diameter of the spot was 40.6 mm.
The thermograph was thrust through an opening in the screen until
its strip lay in the plane of the paper. The spot of light could then
be thrown into it by turning one of the screws of the heliostat.

We have for the reduction of the sun's rays:

M, = 615.3.
1.672

Had the thermograph received the undiminished sunbeam, the de-
flection would have been :

615.3 x 147.16 = 90596.

The ratio of the sun's radiation to that of the moon is, therefore,

90596

0.5008

= 180,900.

Second Method.

In the second method, the sunbeam is employed without diminution,
but its effect is reduced by interposing a resistance in the galvanom-
eter circuit. This method is free from all objections, as far as can be
seen, for its accuracy depends upon the measurement of a resistance,
an operation which can be accomplished with great accuracy. More-
over, it removes any effect due to minor disturbing causes, such as the
heating of the metal plate in the first method, since, when a large
resistance is interposed, the apparatus becomes insensible to such
influences.

The resistance of the thermograph, galvanometer, and the connect-
ing wires, is 0.2095 ohm. On October 21, 1888, the sun's altitude

138 PROCEEDINGS OF THE AMERICAN ACADEMY

being 31°, 100 ohms were added to the resistance of the circuit, and
when the sun's rays were thrown directly into the thermograph a
deflection of 114.6 scale divisions was obtained. Without the 100
ohms of additional resistance, the deflection would have been :

100 X 1U.6 . .

■ — — = 54<00 scale divisions.

0.209a

That same evening the moon was observed at an altitude of 30° 15',
the deflection being 124.59 scale divisions. From its augmented semi-
diameter, 14' 52". 7, and the same constants as in the previous example,
we find the deflection for the uncondensed moonbeam to be 0.2945
of a scale division.

The ratio of 54700 to 0.2945 is 186,300. The results obtained by
the two methods are here tabulated.

Ratio of Sun Radiation to Moon Radiation.

First Method. Second Method.

March 30, 1888 . . . 181,000 October 18, 1888 . . . 187,200

April 25, 1888 .... 180,900 October 21, 1888 . . . 186,300

October 18, 1888 . . . 167,400

Mean of the five results 184,560

The results given by the two methods do not show the wide dis-
agreement so noticeable in the comparative light of the two bodies, as
found by different methods. This fact is strong evidence in favor of
the accuracy of the mean result. In reality, the problem of the total
comparative radiation of the sun and moon is in many respects simpler
than the problem of the comparative light intensity of the two bodies.
In the former case, there is freedom from all personal bias ; the meas-
urements are all given by the indications of the instruments directly,
and, in general, the chances for error seem to be fewer.

Lord Rosse found the ratio of solar to lunar radiation to be 80,000
to 1. Langley obtained the ratio 96,509 to 1. Provided the moon
were a flat disk, reflecting perfectly all of the sun's rays that fall upon
it, we could not receive more than one 97,000th part of the solar heat
from such a disk, as Zolner has shown. The close agreement of his
result with this number Langley considers to be largely a matter of
chance, or, rather, of constant errors, tending in an unknown degree
to increase the observed values.

It would be a priori quite improbable that we should, under the
circumstances of reflection from the lunar surface and subsequent

OF ARTS AND SCIENCES. 139

passage of the reflected rays to the surface of our planet, receive any-
thing like the maximum possible radiation. Our atmosphere cuts off
a considerable portion of the lunar rays. It has been observed that
aqueous vapor is particularly efficacious in stopping them. It may
not be uninstructive, as illustrating the minuteness of the quantity of
heat that we receive from the moon, to express it in terms of melting
ice. It will not be far from the truth if we say that the sun's rays
will thaw through an inch of ice in 100 minutes, whence it follows,
f'iom our determination, that the rays of the moon will melt the same
thickness of ice in 18,456,000 minutes, or 35 years.

That portion of the solar rays not reflected from the lunar surface
is absorbed by it, and radiated at different wave lengths. At what
wave lengths we do not know, and whether or not those rays are
capable of passing our atmosphere we cannot tell. An idea of the
general character of the lunar rays may be gained from experiments
with absorbing media. For example, on April 25, 1888, the per-
centage of lunar rays transmitted by a plate of quartz was determined.
The quartz plate was 5 mm. thick, cut perpendicularly to the axis of
the crystal. It was arranged to be drawn before the thermograph by a
thread, so that the observations could be made in pairs, alternately
with and without the quartz. The results were as follows : —

Moon's Altitude.

Deflection
through Quartz.

Deflection
without Quartz.

Per Cent

transmitted.

13°

38.4

137.2

34.6

19°

54 4

157.0

27.9

24° 30'

51.3

163.2

31.4

Mean of the whole (GO readings) . . . 31.3%

An attempt was then made to find some artificial source of heat
whose radiations would be transmitted by the plate of quartz to the
extent, approximately, of 31%. It was discovered that a coil of
platinum wire, in a Bunsen lamp, turned as low as possible, very
nearly fulfilled the requirement. We may say, then, that in general
character the rays from the moon resemble those from an incandescent
platinum wire. The plate of quartz used in the above experiments
transmits 93.3% of the sun's rays.

A study of the reflection of the sun's rays from rocks may serve, to
some extent, to aid in the lunar problem, and to this end a few experi-
ments have been made, after the following manner.

The thermograph is mounted upon a strip of board, which can ho
turned about a pin passing through it and into the table below. 'I he
rock under experiment is held in a clamp a short distance in front of

140 PROCEEDINGS OF THE AMERICAN ACADEMY

the instrument. The board bearing the thermograph is moved about
the pin by drawing upon a cord, and is brought back to place by an
opposed spiral spring. Should it not be desired that the thermograph
remain continuously exposed to the radiation from the rock, it may be
turned away from it by a pull upon the cord. When a reading is to be
made, the cord being released, the spiral spring draws back the board,
which is stopped at the proper position by a pin fixed in the table.
The rock under examination is given a flat surface by grinding it
upon a sheet of glass with emery. The sun's rays are directed upon
it by a heliostat, the diameter of the solar beam being limited to
26 mm. by a circular diaphragm.

On November 4, 1888, a slab of white marble was made the subject
of experiment. The distance from the slab to the thermograph strip
was 217 mm. The mean deflection obtained was 82.5 scale divisions.
Had the marble slab reflected perfectly in every direction the solar
rays that fell upon it, the ratio of the radiation received at the thermo-
graph strip to the solar radiation would have equalled the ratio of
the area of the cross section, of the solar beam employed to the area
of the hemisphere of which the distance from the marble to the ther-
mograph strip is the radius ; or,

82.5 7rl32

Solar Radiation o n 2172

From which we find, Solar Radiation = 45980 scale divisions, upon
the supposition of perfect reflection by the marble.

The thermograph was next placed in the direct solar beam, with a
resistance in the galvanometer circuit, and the true solar radiation was
found to be measured by 84170 scale divisions of deflection. Hence,

of the solar rays falling upon it, the marble slab reflected part,

or 54.6%. In the same manner, it is found that a surface of black
slate reflects 3>S% of the solar rays.

We learn, then, that of the solar rays falling upon rocks a large
portion is absorbed and conducted through the mass of the rock, the
remaining portion being reflected. In both of the cases given, the
sunlight was flashed upon the rock when the reading was to be made,
the rock remaining, practically, at the temperature of the room. It
appears, therefore, that the solar rays are about equally divided into
two portions, one of which is reflected directly, the other absorbed,
conducted through the mass of the rock, and radiated in long waves.
If the rocks were in large masses, and continuously acted upon by the

OF ARTS AND SCIENCES. 141

solar rays, their temperature would be considerably raised by the ab-
sorbed heat. The radiation of a body like the moon must consist, in
considerable part, of emanations from such a heated surface ; and it
becomes important to inquire whether our atmosphere is permeable
by such rays.

Lord Rosse found, at the time of total lunar eclipse, that with tl;e
disappearance of the last ray of light from the vanishing moon radia-
tion, to which his thermopile was sensitive, vanished also. On the
night of January 28, 1888, the first systematic observations were
made by the writer and his assistants with the previously described
apparatus. The moon rose eclipsed. A reduction of the observations
shows that

19 minutes before totality the deflection was 11.2 divisions.

o (t tt (( it a u n o (t

During first minute of totality " " 3.2 "

Mean of 30 readings taken during total phase 2.09 "

After the total phase had passed, but while the moon was still in
the penumbra, .254.4 was obtained as the mean of 20 readings.

No attempt was made to observe continuously. The evening was
so intensely cold, that one could not stand quietly in the open air for
any length of time without freezing. The inference to be drawn from
these observations is, that all but a minute portion of the rays from
the lunar soil and rock are cut off by our atmosphere, for it is impos-
sible to conceive that a surface like that of the moon, upon which the
sun has been shining continuously for many days, should suddenly
cease to radiate upon withdrawal from the sunshine. It is very
questionable, then, if at any time we receive any considerable portion
of the radiation from the lunar rocks.

The results given above are in substantial agreement with those
obtained by Dr. Boedicker during the same eclipse. He employed a
thermopile, and condensed the lunar rays with the three-foot speculum
of Lord Rosse. He says,* that, twenty minutes before totality com-
menced, the heat was reduced to less than five per cent of what it waa
before first contact with the penumbra. He also observes, that the heal
radiated by the moon commences to diminish long before first contact ;
and that the heat, after last contact with the penumbra, did not mount
immediately to what it had been before first contact with penumbra.

It was intended, upon the occasion of the total lunar eclipse of July,

* Nature, March 8, 1888.

142 PROCEEDINGS OF THE AMERICAN ACADEMY

1888, to make as complete an investigation as possible of these and
allied matters. With this purpose in view, two complete sets of
apparatus were prepared, and two parties of observers were in the
field. Upon the night of the eclipse, however, it rained, and observa-
tion was consequently impossible.

If we are to conclude that our atmosphere is opaque to radiations
from the lunar rocks, we must endeavor to explain the observed be-
havior of the lunar rays, with reference to absorbing media, upon the
hypothesis of selective reflection. We have found that 31% of the
lunar rays will pass through quartz ; and if it is observed that ordi-
nary rocks absorb and reflect selectively, taking up the short waves,
and reflecting the long, the absorption of the lunar rays may be as
effectively explained as upon the supposition that we receive radiations
from the lunar rocks.

We have arranged to reflect the solar rays from rocks, and, by
placing in the path of the reflected ray the same quartz plate that was
used in the lunar observations, have ascertained the amount of absorp-
tion. The results are shown in tabular form below.

.. . . Mean Deflection Mean Deflection Per Cent

Material. without Quartz. through Quartz, transmitted.

Red Lava, Porous .... 153 112 73

Quartz 131 116 88

Black Lava 82 65 80

White Marble 180 147 82

Slate 103 77.8 79

Direct rays of sun transmitted .... .... 93

The results show that selective absorption takes place, but to an
extent quite limited, and altogether insufficient to explain the great
absorption of the lunar rays. When we consider the great differences
in character of the specimens of rocks experimented upon, as compared
with the very inconsiderable differences in the transmitted rays, it
seems useless to undertake to find a material that shall absorb so large
a proportion of the short solar waves that only one third of the re-
flected rays will pass through the quartz plate.

The question is how to reconcile the results here obtained with the
eclipse observations, showing that our atmosphere is nearly opaque to
radiations from the lunar rocks. The materials for the solution of
this problem do not seem to be at hand. It is, of course, possible that
the materials of the moon's surface may absorb selectively to a much
greater extent than those that we have examined. On the other
hand, if the lunar surface were of very light and porous materials,

OP ARTS AND SCIENCES. 143

like our volcanic pumice rocks, a withdrawal of the solar rays might
be attended by so sudden a fall of temperature as to explain the eclipse
observations. Somewhere between the two the explanation probably
lies : but at present we can do little better than conjecture.

Atmospheric Transmission Curve.

An attempt has been made to construct a curve which would show
the transmission of the atmosphere for lunar rays at every altitude of
the moon. Observations for this purpose have been made, at every
opportunity, for nearly a year, and already amount to some thousands
in number. It has been customary to begin the observations at moon-
rise, or, if the moon were already in the sky, as soon as the evening
temperature had become sufficiently stationary, and to continue them
until culmination.

The working of the apparatus is attended with considerable fatigue,
and it is not possible for one person to operate it uninterruptedly.
Tbe readings, therefore, are made in sets of ten or twenty with a short
rest between the sets. Sometimes a third person has measured the
altitudes with a portable transit, while two gave their attention to
making the readings ; but usually the same person has directed the
thermograph-box and taken altitudes at .the beginning and the end of
each set, the mean reading of each set being regarded as the deflection
corresponding to the mean altitude.

Below is given the work of the night of February 28, 1888. Con-
siderable difficulty has been experienced in obtaining the deflections for
the high altitudes, as it very frequently happens that, late in the even-
ing, clouds arise from the ocean near at hand and obscure the sky. On
this particular evening the sky was very serene and beautiful until
10 h. 40 m., when the clouds arose. Because of this unusual seren-
ity, this series of observations shows more than ordinary regularity,
although less than the customary number of observations was made
to a set.

144

PROCEEDINGS OF THE AMERICAN ACADEMY

February 28, 1888.
Barometer 30.35. Thermometer — 13° C.

j

Altitude.

Time p. m.

Deflection.

Altitude.

Time p. m.

Deflection.

O '

h. m.

o »

h. m.

3 10

7 24

67.2

20 0

9 5

174.3

7 30

7 51

128.6

20 30

9 8

177.5

8 0

7 53

129.4

23 10

9 24

183.1

8 30

7 56

134.5

23 30

9 27

185.3

11 0

8 13

146.6

24 0

9 31

187.1

11 30

8 15

147.5

24 30

9 33

184.1

12 0

8 19

149.0

25 0

9 36

188.1

12 30

8 22

152.2

25 30

9 40

188.3

15 0

8 34

158.3

26 0

9 44

189.0

15 30

8 37

160.3

28 10

9 56

191.0

16 0

8 41

161.6

28 30

9 58

192.0

16 30

8 44

165.1

29 0

10 2

193.4

19 0

8 59

168.1

33 40

10 31

202.1

19 30

9 2

172.3

To illustrate the effect of the weather upon such ohservations as
these, we may compare with the above the following table, made up
from the results of October 21, 1888. On the former occasion the air
was very calm, and from its low temperature ( — 13° C), must have
been quite dry. The evening of the latter date was fine, after many
days of fog and rain, and, although clear to the eye save for the merest
visible milkiness, the drifting clouds of vapor were made manifest by
the irregular transmission of the lunar rays.

October 21,

1888.

Altitudes

Deflections.

Altitudes.

Deflections

O '

6 45

105.1

20 15

182.7

11 45

146.1

21 0

175.7

12 15

149.1

22 45

178.7

15 45

163.3

29 45

187.8

17 0

159.1

31 0

191.8

Following Langley * we may find the approximate transmission
coefficient for a column of air capable of supporting one decimeter of
mercury, by employing the formula

t = (M2/?2 - M^)

rf,

* American Journal of Science, vol. cxxv. p. 176.

OF ARTS AND SCIENCES.

145

The dates upon which this has been determined, together with the
principal quantities entering into the computation, are here given in
tabular form.

Table of Transmission Coefficients.

Date.

1888.

Altitude (1).

Altitude (2).

d2. dv

ft.

ft.

M2/3j — MjjS^

t.

O '

o '

Feb. 28

12 0

33 40

149.0

202.1

7.696

7.676

22.47

0.9866

28

15 0

29 0

158.3

193.4

7.696

7.696

13.52

0.9852

March 29

14 54

30 25

112.2

143.4

7.645

7.640

14.42

0.9832

Oct. 18

22 30

41 0

174.5

222.3

7.670

7.070

7.82

0.9708

The mean of the values of t is 0.9852, and t 7.6, the amount of
lunar radiation transmitted by the entire atmosphere at the ordinary
pressure, is 0.8925.

Summary.

I. We have shown how an instrument may be constructed, capable
of measuring very minute radiations.

II. We have measured the radiations of several rocks at near 100°,
and presented the results in a table.

III. We have found that the heat which our planet receives from
the moon is to that from the sun as 1 to 184,560.

IV. We have compared the lunar rays with solar rays reflected
from rocks, with reference to the absorption of each by quartz.

V. We have constructed a curve, representing the change of trans-
mission of lunar rays by our atmosphere with changes in altitude of
the moon.

VI. We have found that our atmosphere, at the ordinary pressure,
transmits 89.25% of the vertical lunar beam.

VOL. XXIV. (N. S. XVI.)

10

146 PEOCEEDINGS OP THE AMERICAN ACADEMY

XIII.

ON THE CHARGING OF CONDENSERS BY GALVANIC

BATTERIES. '

By B. O. Peirce and R. W. Willson.

Presented March 13, 1889.

I. The Use of Water Cells.

The different methods which are commonly used for comparing the
capacities of condensers often lead to widely different results.

Several condensers may be apparently of the same capacity when
tested by one process, while each differs from all the others under a
second set of conditions. Every condenser has its own character-
istics, and these must be specially studied before it can be used in
accurate work.

The present investigation grew out of the discordant results which
one of us obtained in attempting to compare condensers made in dif-
ferent ways. Our experiments have removed some of our own diffi-
culties, and we think that an account of our work may prove useful
to others.

We shall begin by considering the behavior of different batteries
when they are suddenly called on to furnish definite quantities of elec-
tricity in definite short times, and in this first paper we shall give some
results which we have obtained in using water cells.

These results are interesting, because the water battery possesses in
an exaggerated degree some properties which are common to all bat-
teries, and which may seriously affect* the quantity of electricity fur-
nished to a large condenser by a cell with which it is connected for a
short time only.

The cells that we used were made of ordinary " 2^ ounce wide mouth
flint glass bottles," filled to the neck with faucet water, and containing
each a strip of zinc and a strip of copper placed vertical and nearly par-
allel to each other, at a distance apart of about two centimeters. The

* Von Beetz, Wied. Ann., xxvi. p. 24, 1885.

OF ARTS AND SCIENCES. 147

cylindrical part of each bottle under the neck was about six centimeters
high, and four centimeters in diameter, and each strip of metal was of
such a size as to expose a surface of seven square centimeters on each of
its sides to the liquid. The resistance of each cell as measured by the
use of alternating currents* was nearly 1200 ohms. Of a large number
of such cells at our disposal, we found 240 sufficient for our purpose,
and a switch-board enabled us to arrange these easily in various com-
binations of groups of twenty cells each.

The condensers which we used in this part of our work were nine
in number, including one of a capacity of one microfarad subdi-
vided into fractional parts. This condenser, which was made by
Messrs. Elliott Brothers and marked by them No. 72, we took as a
standard.

As we shall show when we come to discuss the characteristics of
different condensers, the ratios of the capacities of the other condensers
to the capacity of this one varied somewhat with the time of charging,
but, when this remained constant, were the same, whatever the kind of
battery used.

The capacity of each of the condensers for each time of charging
was measured in terms of the capacity which No. 72 has when charged
for one second by a battery of small internal resistance. The capaci-
ties of five of the other condensers, as determined in the manner just
described, did not vary on the average by more than two per cent
when the time of charge was decreased from 2.0 seconds to 0.01
second, that is, in the ratio of 200 to 1. By combining the different
condensers, we were able to get within one twentieth of a microfarad
any capacity from 0 to 11 mf., a range which was amply sufficient for
our purposes.

Most of our work on water cells was done with the help of a four-
coil mirror galvanometer (A), of 3300 ohms' resistance. For some pur-
poses, however, only one of the coils was used. The period of the
complete swing of the galvanometer needles was 27 seconds, and the
ratio of two successive swings 1.058. Unlike some of our other gal-
vanometers, this one, when used ballistically, within the limits which
we set for our experiments, gave throws which seemed to be func-
tions only of the quantity of eleetricityf discharged through it. that is,
independent of the tension of this electricity at the instant of dis-
charge. A minute coil of wire, placed in the core of one of the

* F. Kohlrausch, Wied. Ann., xi. p. 663, 1

t See Lord Rayleigh, Phil. Trans., I't. 2, p. Gl!», 18S2.

148 PROCEEDINGS OF THE AMERICAN ACADEMY

galvanometer coils, through which by means of a double key the
current from a small thermopile could be sent in either direction,
served to damp the oscillations of the needle after a throw had been
given.

A few of our measurements were made with the help of another
four-coil galvanometer (B), of 3000 ohms' resistance. The period of a
complete swing of its needle was 52.7 seconds, and the ratio of succes-
sive swings 1.068. With scale placed at a distance of one meter from
the mirror, one microcoulomb when discharged through this galva-
nometer gave a throw of 45 mm.

For intervals between 0.01 sec. and 0.5 sec, the apparatus shown
in Figure 1 was used. It consists of a pendulum with an iron bob, A,
eight inches long, two inches in diameter, and weighing 7 lb., and
a brass rod ^ inch in diameter weighing 2.8 lb. The axis of suspen-
sion is formed by a brass rod 2^ inches long held between pointed
screws.

Upon this axis and concentric with it are placed two ebonite disks,
F, G inches in diameter and | inch thick. The face of each of these
disks is armored through about half of its circumference with a strip
of German silver ■£% inch in thickness, held by rivets passing through
the disk, and the edge of the armature is turned true with the edge of
the disk.

The disks are placed with their armored faces adjacent, but separated
by an ebonite washer 5 inches in diameter and ^V inch in thickness.
One of the disks is fixed to the pendulum rod, while the other can be
revolved about their common axis so that the two German silver strips
may overlap to any desired extent ; the arc a, attached to the movable
disk, serves to clamp the two disks firmly together, and also to deter-
mine approximately the amount of the overlapping. To each of the
German silver strips is joined a wire, b, which dips in a mercury cup, g,
connected with one of the two points between which the circuit is to
be made ; to avoid as far as possible any interference with the motion
of the pendulum, these mercury cups are placed very close to the axis
of suspension.

The pointed screws upon which the pendulum swings are held in a
frame, c c, of wrought iron, which is itself swung between two pointed
screws, d d, perpendicular to the axis of suspension of the pendulum.
By rotation of the frame c c upon these screws, the axis of the pen-
dulum is easily levelled.

To the frame c c is attached a pillar, e, carrying a copper wire, /
5% of an inch in diameter, in such a position that its end runs with

OF ARTS AND SCIENCES.

149

Fig. 1.

150 PROCEEDINGS OF THE AMERICAN ACADEMY'

light friction in the narrow groove between the two disks as the pen-
dulum swings ; both the end of the wire and the sharp edges of the
disks are somewhat rounded to prevent undue wear. If the two disks
are clamped together in a position such that the unarmored part of
one is opposite the armored part of the other, and that as the pendulum
swings the wire leaves one German silver strip just as it reaches the
other, it will be in contact with each armature successively, but will
establish metallic contact betweeu them and close the circuit between
the mercury cups at most for a very short interval ; if, however, the
armatures overlap for a certain number of degrees, contact will be
maintained while the pendulum swings through just that number of
degrees of its arc. The front disk, which is rigidly fixed to the pen-
dulum rod, is so placed that it comes in contact with the copper wire
just as the pendulum reaches its lowest point. If therefore the arma-
tures overlap, for instance, 10°, and if the pendulum is urawn out to
a known point upon one side of its arc so that the copper wire rests
upon the unarmored side of the " fixed " disk, then, when the pendu-
lum is allowed to fall, metallic connection will exist between the disks
while it is rising through 10° from its lowest point. From this
"arc of contact" the time of contact is easily determined. The time
of vibration of the pendulum is 1.181 sec, and the arc of initial
displacement 28° 56'; the decrease of arc in one complete vibration
is about 30'.

The length of the arc of contact may easily be determined within
one minute. This uncertainty may give rise to an error in the com-
puted time varying from 0.00022 sec. for an arc of 1°, to 0.00042 sec.
for an arc of 25°.

The errors arising from a wrongly assumed value of the initial
displacement, and from decrease of arc, are in no case greater than
one fourth of one per cent of the computed time, and are here
neglected.

Since the copper wire does not touch each of the armatures in
a single point, the arc of contact will obviously not be exactly equal
to the angle of overlap, but its value may be easily determined by
direct observation. By throwing the switch at H to the right, the
circuit through the bell and battery, J, is made complete, except for
the break at the disks. The pendulum is then lowered, and the
readings on the divided arc E of the microscope K are noted when
the bell begins and stops ringing, as the pendulum is moved through
its ascending arc ; the former must nearly correspond with the
reading when the pendulum is in its position of rest. To facilitate

OF ARTS AND SCIENCE- 151

I
this latter adjustment the wire / is made movable in the direction
of its length.

For releasing the pendulum and catching it at the end of its spr
the contrivance shown at D is used. The latch h is firmlv fastened
to the spring i, which serves the double purpose of giving the up-
ward force upon the latch-bar to cause it to ensa^e with the catch
upou the pendulum bob, and also of lessening the shock produced
upon the whole apparatus by the arrest oi the pendulum. The sprint
i is fixed inside the tube k, which, sliding inside the sleeve /, serves
as guide and stop for the motion of the latch. To operate the mech-
anism, the tube k is drawn out by the cord attached, and the pen-
dulum drawn with it by the latch ; at a certain point, however, the
pin m, which passes through the latch, is brought against the wed^e n,
and a slight further motion trips the latch, and the pendulum is
released. The point at which the pendulum is released is so adjusted,
by sliding the wedge >t. that the arc through which the bob is lifted
- slightly greater than the loss of the arc through friction, etc. in
a complete vibration. The latch is drawn up by a cord passing around
an axis. L. which is turned by a lever and cord pulled by the observer
at the galvanometer. This is to insure that the pendulum is dropped
from nearly the same height at each observation, the multiplying
arrangement serving to decrease the velocity with which the peudu-
lum is raised, so that it rises an exceedingly small distance beyond
the point where the latch is tripped. As soon as the pendulum tails,
the latch, being at once released by the observer, is drawn back to
its lowest position by a spring, not shown in the figure, and the
pendulum returning is caught by the latch, and thus, having made
a complete swing, is left in its original position ready for the next
observation.

Attached to the frame in which the pendulum swings is a balanced
rocking kev. /'. with two adjustable screws making contact in mercury
cups. When the pendulum hang- by the latch, the key r- sts in
contact with the left-hand mercury cup q. through the k fi-hand bo
but as the pendulum completes its half-vibration, the pin .«. attached
to the " fixed " ebonite disk for this purpose, throws the lever over
upon the right-hand contact, r. where it remains during the second
half-vibration of the pendulum, being replaced in its original position
by the pin t, as the pendulum is caught by the latch.

The cycle of operations performed by the pendulum is explained
bv the diagram below, which shows the connections as made for the
three different kinds of observation : —

152

PROCEEDINGS OF THE AMERICAN ACADEMY

(1.) Charge of the condenser for a short measurable interval
followed by discharge through the galvanometer, the discharge
occurring about 0.6 sec. after the commencement of the charge,
and the contact between condenser and galvanometer lasting abuut
1.2 sec.

(2.) Charge of the condenser through the galvanometer for a
short measurable interval; discharge (not measured) about 1.8 sec.
after commeucement of charge, the condenser remaining short-circuited
1.2 sec. This is the arrangement shown in Fig. 1.

(3.) Current flowing directly through the galvanometer for a
short measurable interval of time.

The switch H is always turned to the left except when the arc of
contact is to be measured as before described.

For all the observations on condensers referred to in this paper,
the first figure is applicable ; the action is then as follows.

When the pendulum rests in the latch, one terminal of the con-
denser, C, attached to the binding screw 3, is in connection, through
the rocking key, with the mercury cup, q, and therefore with one pole
of the battery, B, at 4. When the pendulum is released, contact
is made between the disks during a certain arc of swing, and the
current then flows from the binding screw 5, through the disks and
switch H, to the other terminal of the condenser at 2.

On the pendulum's arriving at the end of its swing, the key is
shifted from contact with 4 to contact with 1, causing the condenser
to be discharged through the galvanometer ; returning, the pendulum
returns the rocking key to its initial position, ready for the next
observation. It may be noted that the battery is permanently con-
nected with one pole of the condenser except during the return swing

OF ARTS AND SCIENCES. 153

of the pendulum, but we have found no trouble from leakage in these
experiments. The action in cases (2) and (3) will be sufficiently
evident from the above description.

The weight of the pendulum rod being considerable, we found that
it was somewhat bent by its own weight when the pendulum bob was
supported by the latch, and when released a vibration was set up in
the rod on its own account, which caused a rapid vibration of the
disks in addition to the motion due to the pendulum. The result of
this was to produce an uncertainty in the beginning and end of con-
tact, depending upon the relative phase of the two vibrations at the
beginning and end of the arc of contact ; the amount of this uncer-
tainty was less than 0.0005 sec, and was greatly lessened by adding
steel wire stays to the rod when it became desirable to use the pen-
dulum for times approaching 0.001 sec. in magnitude.

A similar effect, probably attributable to the same cause, was
noticed as producing a considerable error in a much larger pendulum
of similar construction, when used for times approaching 0.001 sec,
although the rod was of wood 1£ inches square in section, the bob
weighing 40 lb. and the time of swing being 1.8 sec, with an initial
displacement of 26°.

We have used a more simple apparatus, shown in Fig. 3, in cases
where it was desired to charge a condenser for a definite interval
of time, the absolute value of which is required to be known with a
higher degree of accuracy. With the pendulum already described,
the time is liable to considerable error when determined by an arc
of contact nearly equal to the half arc of swing ; that is, when the
time of contact is in the neighborhood of 0.5 sec.

Where different times are not to be experimented upon, a greater
degree of accuracy may be attained if the contact existing through
the upward swing be maintained through the downward swing as
well, beginning and ceasing nearly at the lowest point of the swing.
The time of contact is then equal to the time of vibration, affected
by a very small correction, which depends upon the departure of the
beginning and end of contact from the lowest point, and which may
be determined with considerable accuracy.

The pendulum bob is of iron, and is held displaced from its position
of rest by one pole of a horseshoe magnet b, in contact with it at its
middle point; this magnet is attached to the plate c, which turns on
the screw d, so that the magnet maybe moved in the plane of vi-
bration of the pendulum by a cord leading to the observer al the
galvanometer. The pin e acts as a stop limiting this motion of the

154

PROCEEDINGS OF THE AMERICAN ACADEMY

Fig. 3.

magnet towards the left. If the magnet is pulled out a certain
distance toward the right, the pendulum bob is brought against the
stop /, and any further motion of the magnet detaches the pendulum ;
the magnet is at once released by the observer, and falls back to a point
fixed by the stop e, so as to arrest the pendulum at the completion of
its first swing. The pendulum rod is screwed into the edge of an eb-
onite disk, g, three eighths of an inch thick and one inch in diameter,
through the centre of which passes the pin h, which forms the axis
of suspension. Attached to the disk g are two copper strips, j, of
which one is shown in the figure, and which, as the pendulum swings,
rub lightly upon the edge of the fixed disk k. Inserted in the latter
are four pieces of copper m, n. In the ordinary position of the
pendulum as shown in the figure, the copper strips j j, which are
connected with the condenser by the binding screws at p, rest upon
the opposite copper pieces m ?», and through them are in circuit with
the binding screws at q, which are connected with the galvanometer.
The condenser is thus short circuited through the latter.

When the pendulum swings, the strips ^y make contact at the lowest
point of the swing with n n, which are connected through the binding
screws at r with the battery ; and this connection is maintained until
the pendulum reaches its lowest point on the return swing, when the
strips j j pass to the copper pieces, and the condenser is discharged

OF ARTS AND SCIENCES. 155

through the galvanometer. It is hardly necessary to say, that the
width between the pieces m and n is made slightly greater than the
width of j, to prevent a direct current from the battery through
the galvanometer.

The method above described has been found very satisfactory for
the purpose of obtaining a single swing of a not too heavy pendulum;
it is essential that the stop e should be somewhat elastic ; a short
piece of soft rubber tubing slipped over the pin answers well.

The upper pole of the magnet is brought quite near the upper part
of the pendulum bob to strengthen the induction.

The motion of the magnet is so slight, that its direct effect upon
the galvanometer is negligible. No great care is necessary in the
release, as, even if the bob is brought up against the stop f, so as to
rebound, the time of describing the second and third quarters of the
complete vibration is altered by an inappreciable amount.

With the apparatus described a great many measurements were
made, and these all gave concordant results. The following tables
give, in microcoulombs, the charges which variously arranged batteries
of water cells gave, in the times named, to condensers of different
capacities. Each number in the body of a table is the mean of a set
of closely agreeing observations. In the cases of many of the com-
binations of cells we have used a much greater number of different
capacities than is indicated in the tables, but the results given below
are representative.

Whereas in studying certain other cells we have found it desirable
to use times shorter than the ten-thousandth of a second, it is evident,
from the results in Table L, that in the present instance it was not
worth while to use charging times smaller than the hundredth of
a second.

156

PROCEEDINGS OF THE AMERICAN ACADEMY

TABLE I.

Quantity of Electricity delivered in 0.01 seconds by different Batteries of Water-cells
to Condensers of various Capacities.

Battery.

0.1

Capacities of Condensers
03 0.5 1.0

in Microfarads.
1.87 3.06

9.58

mc.

mc.

mc.

mc.

mc.

mc.

mc.

20 cells in series ....
■1 parallel 20's ....
8 parallel 20's ....
12 parallel 20's ....

0.9

1.6

1.7

1.0

3.5

4.2

1.0

3.1
4.6

5.8

1.1
3.4
5.8
7.9

6.8

9.1

3.8

6.8
9.7

1.1

4.0

7.9

10.8

6 parallel 40's ....

2.6

4.6

5.2

5.3

5.5

5.5

5.8

60 cells in series ....
4 parallel 00's ....

1.1

1.2
3.6

3.5

1.3

3.8

4.1

4.4

1.8

4.4

3 parallel 80's ....

3.0

....

3.3

o.o

3.6

2 parallel 120's ....

2.4

26

2.7

2.8

3.2*

240 cells in series . . .

2.4

2.7

2.7

2.8

TABLE II.

Quantity of Electricity delivered in 0.19 seconds by different Batteries of Water-cells
to Condensers of various Capacities.

Battery.

Capacities of Condensers in Microfarads.

0.5

1.0

1.87

3.06 C.82

10.70

20 cells in series . .
3 parallel 20's . .
12 parallel 20's . .

mc.
5.4

7.2
8.2

mc.

7.3
12.3
15.9

mc.

8.7
17.1
28.7

mc.

8.9

20.1

42.7

mc.
9.6
23.8
69.6

mc

9.0

25.3

82.5

40 cells in series . .
6 parallel 40's . .'

7.1
14.2

8.4
25.5

8.8
37.1

9.6
44.1

53.6

9.8
55.9

00 cells in series . .
4 parallel 60's . .

7.0
18.5

89
26.6

9.1
31.0

9.0
35.7

38.5

9.6
40.1

3 parallel 80's . .

18.7

24.6

26.7

29.7

30.3

31.0

2 parallel 120's . .

17.3

19.0

20.1

20.7

21.5

240 cells in series

10.9

11.2

11.3

11.3

11.8

OF ARTS AND SCIENCES.

157

TABLE in.

Quantity of Electricity delivered in 0.37 seconds by different Batteries of Water-cells
to Condensers of various Capacities.

Capacities of Conden

sers in Microfarads.

Battery.

0.5

1.0

1.138

3.07

6.86

10.85

mc.

mc.

mc.

mc.

mc.

mc.

20 cells in series . .

7.1

10.9

13.0

15.1

17.4

17.4

2 parallel 20's . .

8.0

140

20.6

26.0

31.8

33.9

3 parallel 20's . .

8.1

146

24.1

31.7

42.3

46.8

4 parallel 20's . .

8.1

15.5

• . • •

38.0

55.6

64.5

6 parallel 20's . .

8.6

16.6

....

43.6

70.5

83.0

8 parallel 20's . .

9.0

16.7

30.3

46.8

80.4

98.9

12 parallel 20's . .

17.0

31.7

50.3

94.8

119.3

40 cells in series . .

10.8

13.3

16.4

17 5

18.9

20.0

2 parallel 40's . .

14.6

22.4

29.2

32 9

37.0

39.2

3 parallel 40's . .

15.0

25.1

36.2

43 2

50.3

62.6

4 parallel 40's . .

16.4

28 1

42.4

52.1

66.7

71.0

5 parallel 40's

....

29.6

46.4

60.3

78.2

85.5

6 parallel 40's . .

16.6

31.6

51.5

68.2

91.3

100.2

60 cells in series . .

12.0

14 9

19.2

2 parallel 60's . .

18.5

26.2

• ■ * •

■ • > •

• . . .

38.4

3 parallel 60's . .

20.8

32.9

• . . •

49.3

54.4

57.5

4 parallel 60's . .

22.8

38.3

52.9

63.2

73.0

77.0

80 cells in series . .

14.1

16.6

18.7

19.8

20.0

20.7

2 parallel 80's . .

22.1

29.0

33.9

3i-,.2

38.9

39.5

3 parallel 80's . .

26.1

39.1

47.4

53.5

,>.i;

62.1

120 cells in series .

16.4

17.8

19.1

20.2

20.4

20.7

2 parallel 120's . .

26.2

33.6

37.3

38.8

41.0

42.4

240 cells in series .

18.9

20.2

21.9

21.7

158

PROCEEDINGS OF THE AMERICAN ACADEMY

TABLE IV.

Quantity of Electricity delivered in 0.93 seconds by different Batteries of Water-cells
to Condensers of various Capacities.

Battery.

0.5

Capacities of Condensers iu Microfarads.
1.0 1 89 3.11 7 18

11.41

mc.

mc.

mc.

mc.

mc.

mc.

20 cells in series
8 parallel 20's . .
0 parallel 20's . .
12 parallel 20's . .

7.2
73
7.6

13.0
14.2

16. i

19.2

24.6
38.6
46.3
51.0

29 6

61.8

89 7

103.9

32 2
72.8

40 cells in series . .
6 parallel 40's . .

14.4

16.0

21.5
81.1

26.2
56.8

28 9
84.6

32.9

80 cells in series . .
3 parallel 80's . .

210
29.7

27.6
50.7

32.9
717

35 1

85.3

36.5

37.0

120 cells in series
2 parallel 120's . .

24.3
35.(3

30.0
61.1

33.5
61.9

34.1

68.2

72.3

36.3
74.1

240 cells in series

31.6

34.9

37.2

37.9

39.0

TABLE V.

Quantity of Electricity delivered in 2.0 seconds by different Batteries of Water-Cells
to Condensers of various Capacities.

Battery.

Capacities of Condensers in Microfarads.

0.5

1.0

1.89

3.11

7.19

11.74

mc.

mc.

mc

mc.

mc.

mc.

20 cells in series . .

7.3

14.4

24.9

35.1

52.9

57.7

2 parallel 20's . .

....

14.3

26.8

42.3

75.5

93.8

4 parallel 20's . .

15.2

48.8

96.1

40 cells in series . .

14.8

26.2

40.0

50.1

59.3

64.1

2 parallel 40's . .

....

30.1

514

72.6

4 parallel 40's . . .

31.0

567

87.6

60 cells in series . .

20 0

33.1

53.6

62.3

65.0

3 parallel 00's . ■ .

44.5

77.1

80 cells in series .

25 8

40.8

53.6

59.8

68.5

69.8

2 parallel 80's . .

28 5

53.0

78.7

97.4

3 parallel 80's . .

30.4

58.2

94.2

120 cells in series

33.8

474

57.8

63.9

71.9

75.6

2 parallel 120's . .

73.5

100.0

240 cells in series

49.8

63.6

71.0

78.0

83.8

OP ARTS AND SCIENCES.

159

The results given in Tables II., III., IV., and V. have been used
in plotting the subjoined curves. In each curve the ordinates rep-
resent the charges on a scale of 10 microcoulombs to a division, and
the abscissas represent the capacities of the charged condensers in
microfarads. Attached to each curve is a number (m), and at the
upper left-hand corner of the square in which the curve is drawn
is another number («). These denote that a battery made up of
m parallel groups, each consisting of n cells in series, was used in
obtaining the results from which the curve has been plotted.

Fig. 4.

J

■

■

'

h

)

a

y

/

j^

/

h

/

/

3

— /

/

/

''

*—.

:

/

1

1

1

'

,

Y

st

i:

£

2V

I

1

^

t

■

f

:

>

. —

Time of charge, 0.19 sec. Unit of abscissas, 1 raf. Unit of ordinates, 10 mc.

Fig

5.

-'.

i :.,,

■„

',.

1 ■

^r

6C

'

i

A'

.

><

t

/

•

■

>,

-

-'

',

/

■.

A

t

_i

i

/

^

i

I

A

;

■ '

l

/

£

£

i

!

V

.

/

1

k

i

I

—

\

si

]•

p

r

0

l

/

:

i

t—

— - *

- —

7

'-

1

p

-

i

I

-

(

'

, _

Time of charge, 0.37 sec. Unit of abscissas, 1 mf. Unit of ordinates, 10 mc.

160

PROCEEDINGS OF THE AMERICAN ACADEMY

Fig. G.

.'

*

♦1

It

/

*

'

V

/

3

1

/

/

,

1

>,

1

/

(.

/

y

1

1

'

i

1

1

1

r

/

§

'

/

1

,

A

1

'

n

9

2.

0

2

'

-1

1

"

'

,

Time of charge, 0.93 sec. Unit of abcissas, 1 mf. Unit of ordinates, 10 ma

Fig. 7.

:t

K <

-40

i

1

/

u-

4

S

s

1

/

J

/

J-

4

-~

•

'

/

,

•

/

r~

,

'4

■ -

1

■

t

/

»i

1

{

!Z

"1

ti

i

f

/■

(»

j

[/

1

_j_

\

•

x

/

/

/

t

,

/

1

/

1

/

1

Time of charge, 2 sec. Unit of abscissas, 1 mf. Unit of ordinates, 10 mc.

It will be noticed that all the curves in all those squares which
have any one number in their upper left-hand corners, have the same
slope at the origin. This slope is numerically equal to one tenth* of
the electromotive force in volts of the charging batteries, as computed
from measurements of such cells in open circuit made with a Quadrant
Electrometer.

* Each vertical unit corresponds to ten microcoulombs. '

OF ARTS AND SCIENCES.

161

So far as we can judge from our observations with limited ca-
pacities, each curve has as an asymptote a line parallel to the axis
of a?, and at a distance from it which represents the quantity of elec-
tricity which the charging battery would yield if joined up for the
time given in simple circuit with a trifling outside resistance. Direct
experiment shows that this quantity is almost exactly proportional to
the number of cells joined up parallel to each other.

The general effect of interposing a large resistance wound bifilar
between the battery and the condenser which it is charging is shown
by the following diagram, in which the curves are drawn in the same
way as in Figures 4, 5, 6, and 7. The two upper curves represent
observations taken with a battery of twelve parallel groups of 20
water cells each, first with no interposed resistance and then with

Fig. 8.

247,000 ohms inserted. The two lower curves represent similar
observations made with a single group of 20 cells joined up in series.

For purposes of comparison with the numbers in Tables I., II., III.,
IV., and V., we give some results obtained with a battery consisting
of 20 water cells arranged in series. The poles of this battery were
connected together, through various outside resistances, for given
short intervals of time, and the quantities of electricity which traversed
the circuit during these intervals were measured.

The three curves of Fig. 9 were plotted by using as abscissas the
times * during which the battery circuit was closed, and as ordinatea
the quantities of electricity which passed through the circuit in these
times. The observations represented by points in the upper curve
were made when the resistance in the circuit outside the battery was
3,000 ohms. The corresponding resistances for the Other two curves
were 102,000 ohms and 250,000 ohms respectively. It will be

* Each vertical unit corresponds to 1 microcoulomb. Each horizontal unit
corresponds to 0.01 sec.

vol. xxiv. (n. s. xvi.; 11

162

PROCEEDINGS OF THE AMERICAN ACADEMY

noticed that the curvature near the origin is considerable, but that
after about one tenth of a second each line becomes nearly, though not
quite, straight. If the internal resistance of the battery is assumed to
be 24,000 ohms, the effective electromotive forces in the circuit after
one tenth of a second are about 1.2 volts, 4.5 volts, and 7.4 volts,
respectively. The electromotive force of the open battery was a little
over 16 volts.

0.1 sec.

0.2 sec.

0.3 sec.

Fig. 9.

The results in Table VI. were obtained with another battery of 20
water cells which had been resting for one month. The numbers
in the body of the table give in microcoulombs the quantity of elec-
tricity which passes through the circuit when the poles of the battery
were connected for the times named through various external re-

sistances.

TABLE VI.

Duration of Current.

Resistance of Outside Circuit in Ohms.

3,000

102,000 250,000

sec.

mc.

mc.

mc.

.0025

0.34

0.21

0.12

.011

0.94

0.59

0.41

.034

2.50

1.73

1.22

.077

4.94

3.59

2.52

.1033

6.43

4.64

3.34

After the poles of a battery of these water cells joined up in series
have been connected together for a short interval only (say for one

OP ARTS AND SCIENCES. 163

second or less), the battery seems to regain almost instantly its origi-
nal electromotive force. Indeed, it is evident from the numbers given
in Tables L, II., III., IV., and V., that the polarization — which, as is
well known, increases and diminishes in the case of any given simple
element with the intensity of the currents which pass through the
cell — must almost instantly respond to any decrease in the density
of the current. Of course, if a battery of water cells is short-circuited
for a number of seconds or minutes, the polarization slowly increases,
and the battery becomes " tired," and after the circuit is broken it
must be allowed to rest for a number of minutes, or even hours,
before it gets back its lost power.

From the fact that the lines in Fig. 9 are not straight, it appears
that the sudden falling off in the electromotive force of the cells when
the circuit is closed is not wholly due to a polarization, which is at
every instant simply proportional to the density of the current then
flowing.

AVe shall not discuss further the subject of polarization in water
cells until we have published the results of some observations made
with other kinds of battery.

Jefferson Physical Laboratory,
Cambridge.

1G4 PROCEEDINGS OP THE AMERICAN ACADEMY

XIV.

CLASSIFICATION OF THE ATOMIC WEIGHTS IX TWO

ASCENDING SERIES, CORRESPONDING TO THE

GROUPS OF ARTIADS AND PERISSADS.

By W. R. Livermore, Major of Engineers, U. S. Army.

Presented by the Corresponding Secretary, April 10, 1889.

If the atomic weights of all chemical elements are arranged in a
single ascending series, it will take the following form if expressed in
whole numbers :

1,-7, 9, 11, 12, 14, 16, 19,-23, 24, 27, 28, 31,32, 35,-39,
40, 44, 48, 51, 52, 55, 56, 59, etc. ;

corresponding to the elements,

H, — Li, Be, B, C, N, O, F, — Na, Mg, Al, Si, P, S, CI, — K,
Ca, Sc, Ti, V, Cr, Mn, Fe, Co, etc.

The well known " Law of Octaves " calls attention to the fact that
each element bears a strong resemblance to that seven places from it
in this series, and the " Periodic Law " states that the properties of
elements stand in periodic relation to their atomic weights. These
laws have been enunciated and illustrated by Newlands, Mendeleeff,
Lothar Meyer, and others, and are now familiar to all students of
chemistry.

Referring to these laws, Wurtz in his "Atomic Theory " says :
" Though it may be generally true that the properties of bodies are
subject to periodic modifications with the increase of their atomic
weights, the law of these modifications escapes our observation, and
seems to be of a complicated nature ; for, on the one hand, the atomic
weights of successive elements vary within considerable limits without
displaying any regularity in these variations ; on the other hand, we
must confess that the gradations of properties, in other words the
greater or less divergencies between properties of successive elements,
do not appear to depend upon the degree of the differences between
the atomic weights. These are real difficulties."*

* The Atomic Theory, 1881, p. 162.

OF ARTS AND SCIENCES. 1G5

Lothar Meyer on the same general subject says, " There cau be no
doubt as to the fact that these differences are subject to law";* and,
" There is still much work for the hands and mind ; but it will be
thoroughly rewarded. The prize is a systematic inorganic chemistry
which will bear comparison with the thoroughly developed system
of organic chemistry." f

The object of this paper is to contribute to the removal of these
difficulties, by showing that the increase in the atomic weights of the
elements follows a law of the utmost simplicity in its general form,
and identical with that of the series of organic compounds.

Let us take the series above mentioned and consider whether any of
its terms can be classed in a series of constant increase. Beginning
with 2 as a modulus, we find that it applies to 7, 9, and 11, with no
recurrence of this difference except for the terms 12, 14, 16 • 3 applies
only to 9, 12, — 24,27, — 48, 51, and to 16, 19 * * 28, 31 **** * 52,
55; 4 however applies to 7, 11 * 19, 23, 27, 31, 35, 39 * * 51, 55,
59, and to 12, 16 * 24, 28, 32 * 40, 44, 48, 52, 56; thus embracing
in two parallel series all the terms of the first three periods of New-
lands and Mendeleeff, with the exception of 9 and 14, i. e. 22 out of
24. The probability that these numbers would so far coincide with
the regular series from the result of chance, is too slight to be enter-
tained. If now, in addition to this coincidence, any points of resem-
blance can be discovered among several of the elements in each series,
such resemblances are significant; but when it appears that all or
nearly all in each series fall into a distinct group that has already been
recognized as such by chemists, there is scarcely room for doubt that
the regularity of the series is connected with the intrinsic nature of
the elements.

Those of the first series are all Perissads and those of the secm.d
Artiads, or elements of uneven aud even quantivalence respectively,
with the single exception in one case of scandium, a newly discovert 1
element whose atomic weight has never been so accurately determin< d
as to throw it outof its group, and cobalt and nickel, which apparently
belong to another group. For the present, therefore, the law of regu-
lar increase in the atomic weight of the parallel series is spoken of as
the Law of Artiad and Perissad Increase, or more briefly aa the
Pkrissad Law.

By similar methods, we find that the numbers between 70 and 100

* Modern Theories of Chemistry, English Translation, L888, p. 100.

* Ibid., p. 170.

166 PROCEEDINGS OF THE AMERICAN ACADEMY

fall naturally into two series, with the common difference of 5 for
the perissads and perhaps 4 for the artiads. The latter series,
however, is somewhat indefinite.

Between ICO and 150 the perissads increase regularly, with a com-
mon difference of about G|, and the artiads less regularly, with the
same difference, as shown in Tables I. and II.

Too little is known of the atomic weights between 150 and 190, and
of those above 210, to justify their classification. Between 1 90 and 210,
we find that 4 applies to 195, 199 * 207, and to 196, 200, 204, 208.

All of the first series are artiads, and in the second all except 200,
or II g, are recognized as perissads, and this element is so similar in
many respects to Cu, Ag, and Au that for the present it is allowed to
remain in its place. There only remain unclassified 9, 14, 59, 104,
and 193. Of these numbers, 9 is the atomic weight of Be and ap-
pears to differ too much from 8, the serial number of this element, to
be recognized in its natural place.

The remaining numbers, 14, 59, 104 * 193, form a series by them-
selves, with a difference of a little less than 45. Ni and Co correspond
to 59, Ru and Rh to 104, and Ir to 193. These elements are always
classified together, and this regularity of increase has often been
noticed. But perhaps 14, or N, falls into this series accidentally.

In Tables I. and II. the first column shows the symbols for the ele-
ments ; the second column, the serial numbers derived from the for-
mula; the third column, the observed atomic weights; and the fourth
column, the deviations from the serial numbers. The next columns
show the specific gravity, the atomic volume, the fusibility, mallea-
bility, and place in an electro-chemical series for each element, and the
last column the group in the classification of Newlands and Mende-
leeff to which the element has been assigned. The horizontal lines
divide the series and periods of this system.

The observed atomic weights are taken from the lists of Clarke,*
L. Meyer,f and Van der Plaats,$ and in most cases by selecting for
each element that one of the values that differs least from the mean,
taking oxygen at 1G for a standard, and expressing but one place of
decimal fractions. The data for the next four columns are taken
from Meyer's " Modern Theories." To express their fusibility, the
elements are divided into seven classes, that represented by 1 being
the most fusible. The brittle and malleable elements are distinguished

* The Constants of Nature. A Recalculation of Atomic Weights, 1882.
t Modern Theories of Chemistry, English Translation, 1888.
t Verification of the Atomic Weights of M. Stas.

OF ARTS AND SCIENCES. ll'.T

by the initial letters B and M. In the electro-chemical series taken
from Barker's " Text-Book of Chemistry," * 1 is the most electro-posi-
tive in an acid solution, and 63 the most negative.

It appears from these tables that none of the deviations from the
serial numbers exceed 0.6 excepting that for Cb among the perissads,
and Zn and Mo among the artiads ; Zn differs by just a unit, and
6eems to foim an exception to the regularity.

In his account of the Discovery of the Periodic Law,f Newlands
points out the irregularity of the single series, illustrating it by several
tables, and the data are clearly set forth and tabulated in a recent
paper by Venable, on the Recalculation of the Atomic Weights.!

In the artiad grand group no atomic weight corresponds to 20, 36,
83, or 131, and a reference to the electro-chemical series will show
that these are the points at which the elements change from a negative
to a positive maximum, corresponding also to the maximum atomic
volumes, and to the end of the Mendeleeff periods.

These points are marked with asterisks in the table.

Analogy points to the possibility of elements with atomic weights
of 15, 43 or 44, 47, 60, and perhaps of 99, 100, and 143.5.

The newly discovered philippium and praseodymium may perhaps
correspond to 47 and 143.5, respectively, but there appears to be no
place in either series for neodymium or 140.3.

In each of the grand groups, perissads and artiads, the breaks be-
tween the minor groups of common difference occur at or near the
minimum points of the series of atomic volumes, and therefore mid-
way between the dividing lines above mentioned.

The first term of each of the groups of common difference, except
the first, viz. Ga, Ag, Au, Ge, Pd, and Pt, appears to be a heavy
malleable metal, electrically neutral, and with an atomic volume a
little above 9 in the artiad, and 10 in the perissad groups.

With the increase of atomic volume is generally associated an in-
crease in electro-negative properties, and, as a rule, the elements be-
come fusible, brittle, and diamagnetic.

With the maximum atomic volume, the elements change from the
extreme of negative to positive, from diamagnetic to magnetic, from
brittle to malleable ; and with the decrease of volume, they lose their
electro-positive properties, and, as a rule, become at last brittle and
infusible.

* Text-Book of Chemistry, p. 16.

t The Periodic Law, p. 27.

1 Journal of Analytical Chemistry, January, 1889, p. 48.

1G8

PROCEEDINGS OF THE AMERICAN ACADEMY

TABLE I. — Perissads.

Atomic Weights.

Physical Properties.

El.

Gr.

Ser. Obs. Dev.

Sp. Gr.

At. Vol.

Fus.

Mai.

Elec.

H

1 1.0 0.0

••

••

I

Li

7 7.0 0.0

0.59

11.9

1

JV1

5

I

B

11 11.0 0.0
15 ?

2.68

4.0

6

B

49

III

F

19 100 0.0

1

B

60

VII

Na

23 231 +0.1

0.97

23.7

1

M

4

I

Al

27 27.1 +0.1

2.56

10.6

3

M

13

III

P

31 31.0 0.0

2.30

13.5

2

B

55

V

CI

35 35.5 +0.5

1.38

25.6

1

B

59

VII

K

39 39 1 +0.1

0 80

45.4

1

M

3

I

Sc

43 44. ? ....
47 1

••

III

V

51 51.3 +0.3

5.50

9.3

7

B

52

V

Mn

55 55.0 0.0

8.00

6.9

5

B

19

VII

Co?
Cu

59 (59.0 0 0

8.50

6.9

5

M

23)

63 63.5 +0.5

8.80

7.2

3

M

31

I

+ 7 + n X 5

Ga

70.0 70.0 0.0

5.96

11.7

1

-2

?

III

As

75.0 75.0 0.0

5.67

13.2

2

B

54

V

Br

80.0 80.0 0.0

2.97

26.9

1

B

58

VII

Kb

85 0 85.4 +0 4

1.52

56 1

1

M

2

I

Y

90.0 89.8 —0.2

* * • •

, t

11

III

Cb

95.0 94.2 —0.8
100.0

+ 7 5 + n X 6.25

G.27

15.0

7

B

44

V

Ag

107.5 107.9 +0.4

10.50

10.2

3

M

32

I

In

113.7 113.7 0.0

7.43

15.3

1

M

27

III

Sb

120.0 120.1 +0.1

6.70

17.9

2

B

47

V

I

126.3 126.8 +0.5

4.94

25.6

1

B

57

VII

Cs

132.5 132.9 +0.3

1.88

706

1

M

1

I

La

138 8 138.8 0.0

6.20

22.3

3?

18

III

Di

145.0 145.0 0.0
? + n X 4

6.50

22.3

3?

17

V

Au

190.0 196.6 +0.6

19.30

10.2

3

M

40

I

Hg

200.0 200.2 +0.2

13.59

14.7

1

M

33

II

Tl

204.0 204 2 +0 2

11. si;

18.1

2

M

24

III

Bi

208.0 208.0 0.0

9 82

21 1

2

B

29

V

OF ARTS AND SCIENCES.

1G9

TABLE II. — Artiads.

El.

Atomic Weights.

Physical Properties.

Gr.

Ser. Obs. Dev.

Sp. Gr.

A.t. Vol.

Fus.

Mai.

Elec.

12 +n X4

C

0

2

12.0 12.0 0.0
16.0 16.0 0.0
20.0 *

0.33

3.6

7
1

B
B

54
63

IV
VI

Mg

Si

S

24.0 24.0 0.0
28.0 28.1 +0.1
32.0 32.1 +0.1
36.0 *

1.74
2.49
2.04

13.8
11.2
15.7

3
6
1

M
B
B

9
42
62

II
IV

VI

Ca

Ti
Cr
Fe

40.0 40.0 0.0
44.0 ?

48.0 48.1 +0.1
52.0 52.3 +0.3
56.0 560 0.0
60.0 1

1.57

6.80
7.80

25.4

' 7.7
7.2

2

*5
5

iM

B
B
M

8

43
53

21

II

IV
VI
VIII

Zn

Ge
As
Se

64.0 65.0 +1.0
+ 7 + n X4

71.0 ?
75.0 75.0 0.0
70.0 79.0 +0.0
83.0 *

7.15

i
5.07
4 60

9.1

13.2
17.1

2

i

2
1

M

B
B

20

i

54
56

II

IV

V

VI

Sr
Zr
Mo

Pd

87.0 87.5 +0.5
91.0 90 6 —0.4
95.0 95.9 +0.9
99.0 ?

+ 7 + n X 6.25
106.0 1060 0.0

2.50
4.15
8.60

11.50

34.9
21.7
11.1

9.2

2
6
6

4

M
B
B

M

7
14
51

34

II

IV
VI

VIII

Cd
Sn
Te

112.2 112.2 0.0
118.5 118.0 +0.5
124.8 125.0 —0.2
131.0 *

8.65
7.29
6.25

12.9
16.1
20.2

2
2

2

M
M
B

25
28
46

II

IV
VI

Ba

Ce?

Pt
Os

137.2 137.1 —0.1
143.5 ? ?
? + n X 4

195.0 194.9 —0.1
199 0 198.9 -0.1

3.75
6.70

21.50
22.50

36.5
21.0

9.1

87

2
3?

5
5

M

M
B

(i
16

37
39

11

IV

VIII
VIII

Pb

203.0 ?

•_!n7.0 206.9 —0.1

11.38

18.1

2

M

26

IV

Ni
Co
Ru

Rh

Ir

14 + n X 44.8

58.8 58.7 —01
58 8 59.0 +0.2

103.6 104.0 +0.4

103.6 104.2 +0.6

148.4 1

193.2 193.0 —0.2

8.80

8.50

12.26

12.10

22.46

6.7
6.9
8.4
8.6

8.6

4
5
5
6

5

M
M
B

M

B

22
23
85
86

38

VIII

\ III

VIII
VIII

VIII

170 PROCEEDINGS OP THE AMERICAN ACADEMY

The 'study of the periodic law has directed attention to the relations
between the atomic weights and these properties, many of which were
previously treated as functions of the electro-chemical order.

When the exact atomic weights are all determined with sufficient
accuracy, it is natural to infer that the deviations from the regular
increase will show a law as significant as the approximate coinci-
dence with the serial numbers, and even now we see clear evidences
of this law. To make it more apparent, it will first be necessary so
to adjust the formula as to reduce these differences to a minimum, and
in the low atomic weights to take account of smaller fractions.

The formula a -\- n d then has the following values for the sev-
eral groups of common difference : —

6.99 + «X 4.02 12.00 + nx 4.01

70.25 + n X 4.85 74.88 + n X 4.12

107.85 + w x 6.20 105.98 + n X 6.22

Table III. shows the deviations, arranged in the groups and series
of the periodic law ; the lower division shows the electro-chemical
numbers in the same form.

Neglecting insignificant deviations, we find the positive results, or
those where the atomic weight is slightly in excess of the serial num-
ber, either near the extremes of the electric series, in elements of
maximum atomic volume, etc., or in the very centre of this series.
The negative results occur half-way between the centre and the ex-
tremes. Of the perissads, 1, 2, — 31, 32, — 57, 59, are positive, and
17, 18, 19, 21, — 44, 47, 54, 55, are negative; of the artiads, 7, —
46, 51, 53, are positive, and 8, 14, 21, 23, 28, are negative.

In the periodic arrangement, the first and second groups show posi-
tive deviations in the high series of this arrangement; the third, fourth,
and fifth, negative in the high series ; the sixth, positive in the high
series ; and the seventh, positive in two terms, and insignificant in
two. The seventh and eighth groups are negative in the third series.
The deviation of Zn is not included in these calculations, nor is the
cause of its deviation yet apparent. Perhaps the perissads of the
second group are formed from an artiad nucleus, and the artiads of
the second group from a perissad nucleus, thus : * G5, 70, 75, 80, etc.,
and 63 * 71, 75, 79, etc. The last line shows the algebraic mean of
the deviations in each group, which, as might be expected, corre-
sponds nearly to the relative atomic volumes.

To do justice to the bearing of these results on the physical ques-
tion of the ultimate nature of "atoms" would involve a mathematical

OF ARTS AND SCIENCES.

171

TABLE III. — Deviations.

Gr.

I.

II.

III.

IV.

v.

VI.

VII.

VIII.

Ser.

1

—.03

i

—.03

.00

?

—.01

—.03

2

—.02

—.02

.00

+ .02

—.08

+.01

+.32

3

—.02

—.08

?

.00

+.05

+ .10

(-.26)

—.11

4

+.06

■?

—.30

?

—.06

+ .03

+ .01

5

+ .64

+.28

+ .01

—.80

—.30

+ .47

+.02

6

+ .08

+.03

—.36

—.42

—.14

+.31

+.41

7

+.25

+ .02

—.18

?

—.05

+.14

+.05

—.14

—.25

—.09

+ .17

+ 18

—.05

Electro-chemical Order.

1

5

i

49

48

i

63

00

2

4

9

13

42

55

02

59

3

3

8

■?

43

52

53

(19)

21

4

31

20

?

1

54

56

58

5

2

7

11

14

44

51

34

6

32

24

27

28

47

46

57

7

1

6

18

16

17

discussion too prolonged for this paper ; but the subject cannot be com-
pleted without a brief statement of its relation to well known systems
and theories of Chemistry. A few extracts will serve as a basis to
show what the perissad law is intended to contribute to the solution
of the all-important problem.

In 1854, in a paper read before the American Academy on " The
Numerical Relation between the Atomic Weights," Professor Cooke
announced the character of the problem as follows : —

" Numerical relations between the atomic weights of the chemical
elements have been very frequently noticed by chemists. One of the
fullest expositions of these relations was that given by M. Dumas, of
Paris, before the British Association for the Advancement of Science,
at the meeting of 1851. This distinguished chemist at thai time
pointed out the fact, that many of the elements might be grouped in
triads, in which the atomic weight of one was the arithmetical mean
of those of the other two. Thus the atomic weight of Bromine is the
mean between those of Chlorine and Iodine; that of Selenium i- the
mean between those of Sulphur and Tellurium ; and thai of Solium,
the mean between those of Lithium and Potassium.

"M. Dumas also spoke of the remarkable analogies between the

172 PROCEEDINGS OF THE AMERICAN ACADEMY

properties of the members of these triads, comparing them with sim-
ilar analogies observed in Organic Chemistry, and drew, as is well
known, from these facts arguments to support the hypothesis of the
compound nature of many of the now received elements. Similar
views to those of Dumas have been advanced by other chemists.

"The doctrine of triads is, however, as I hope to be able to show
in the present memoir, a partial view of this subject, since these triads
are only parts of series similar in all respects to the series of homo-
logues of Organic Chemistry, in which the differences between the
atomic weights of the members is a multiple of some whole number.
All the elements may be classified into six series, in each of which
this number is different, and may be said to characterize its series.
In the first it is 9, in the second 8, in the third G, in the fourth 5, in the
fifth 4, and in the last 3." (See Table IV.)

The paper speaks of the properties of the elements which are func-
tions of the atomic weights, and says that it does not seem bold the-
orizing to suppose that the atoms of the members of the same series
are formed of a common nucleus, to which has been added one or more
groups of atoms, or perhaps one or more single atoms, to which the
corresponding element has not been discovered.

Referring to this subject in 1857,* M. Dumas said : "I have often
tried, as Mr. Josiah Cooke has on his part, to compare them, to com-
bine them, and to discuss them, with the hope of drawing some conclu-
sion from them with certainty, but I have been unable to draw from
them anything but doubt. The formula deduced from the above
simple progression" (a+?id) "would not account for the generation
of simple bodies, as Cooke had supposed, but organic radicles are not
always formed by addition, they are also produced by substitution, as
we see in the compound ammoniums." Dumas proposed to substitute
a formula like n a -f- n' d-\- n" d' ; a being the nucleus; d and d\ the
common differences ; and n, n', n", whole numbers. Table IV. shows
the series of Cooke aud Dumas.

Following in their footsteps, Professor Newlancls extended their prin-
ciples to include newly discovered elements. Employing Cannizzaro's
atomic weights instead of those of the old system, and collating his own
results and all discovered up to 1864, he arranged the groups in hori-
zontal lines instead of vertical columns, and then discovered the " Law
of Octaves," as shown in Table IV. The coincidence of the ordinal
numbers led him, in 18G5, to express the following opinion: "I will

* Comptes Rendus, 1857, p. 709.

OF ARTS AND SCIENCES.

173

TABLE IV.
Cooke, 1854.

Ser.
Gr.

in.
3. 2. 1.

IV
3, 2. 1.

v.

VI.

VIII.

IX.

IV.
2. 1. 3.

H

Li
Na
K
Ag

CaMg
Sr Zn
Ba Cd
Pb

Al Ti
Cr Pd
Mn Sn
Fe Pt
Co Ir
Ni Os
U Au
Cu
Hg

C
B

Si

0

N

P

As

Sb

Bi

0

s

Se
M

Te
V
W
Ta

0
F
Cy

3 CI

Br
I

Al Ti Cu
Cr Pd Hg
MnSn
Fe Pt
Coir
Ni Os
U Au

Cu Mn

Ti

Cr
V

As
Mn

Cr
Mn

Os

Au

Dumas, 1857.

Form.

7 + n .8

12 + n 8

25+ nSi

6 + ?i5

14 + n 17

8 + ?i8

19 + m 16.5 + «'28

L

Na
K

Mg

Ca

Sr

Ba

Pb

Ti
Sn
Ta

C
B
Si
Zr

N

P

As

Sb

Bi

Cr 0
Mo S
V Se
W Te

F
CI
Br
I

Mendeleeff, 1869.

Or.

-S..-r_

1

2

.3

4

5

6

7

8

9
10
11
12

I.

II.

III.

IV.

v.

VI.

VII.

\ in.

H

Li

Na
K

(Cu)
Rb

(Ag)
Cs

(-)

(Au)

Be

Mg
Ca

Zn
Sr

Cd
Ba

Hg

B

Al
Sc

Ga
?Y

In
?Di

?Er

Ti

C

Si
Ti

Zr

Sn
?Ce

?La

Pb
Th

N

P
V

As
Nb
Sb

Ta
Bi

0

s

Cr

Se
Mo

Te

W

u

F

CI
Mn

Br

I

IV Co Ni Cu
RuRhPdAg

Oa Ir Pt Au

174

PROCEEDINGS OP THE AMERICAN ACADEMY

TABLE IV. — Continued.
Newlands, 1865.

No.

No

No.

No.

No.

No.

No.

No.

H 1

F 8

CI 15

Co&N

22

Br 29

Pd 36

I

42

Pt&I

^50

Li 2

Na 9

K 1(5

Cu

23

Rb 30

Ag37

Cs

44

Tl

53

G 3

MglO

Ca 17

Zn

25

Sr 31

Cd 38

Ba&V45

Pb

54

Bo 4

Al 11

Cr 19

Y

24

Ce & La 33

U 40

Ta

46

Tb

56

C 5

Si 12

Ti 18

In

26

Zr 32

Sn 39

W

47

HS

52

N 6

P 13

Mn20

As

27

Di&Mo34

Sb 41

Nb

48

Bi

55

0 7

i

S 14

Fe 21

Se

28

Ro&Ru35

Te43

Au

49

Os

51

endeavor to show that all the numerical relations among the equivalents
pointed out by M. Dumas and others, including the well known triads,
are merely arithmetical results flowing from the existence of the 'law of
octaves,' taken in connection with the fact of the equivalents forming
a series of numbers approaching to the natural order." And in 1866 :
" The fact that such a simple relation exists now, affords a strong
presumptive proof that it will always continue to exist, even should
hundreds of new elements be discovered. For, although the differ-
ence in the numbers of analogous elements might, in that case, be
altered from 7, or a multiple of 7, to 8, 9, 10, 20, or any conceiv-
able figure, the existence of a simple relation among the number of
analogous elements would be none the less evident." * The law of
regular increase, however, points to a relationship far more significant
than the identity of their ordinal numbers.

Working independently of Newlands, Mendeleeff discovered the pe-
riodic law in 1869. It has offered a grand incentive to new discov-
eries and received well merited recognition throughout the world.
The extracts from Wurtz and Meyer, at the beginning of this paper,
show what it left to be desired.

A paper by Rev. Dr. Haughton, and an abstract of one by Mr.
Stoney, have recently appeared in " The Chemical News." Each
gives a clear enunciation of the nature of the problem. Dr. Haugh-
ton discovers some of the regular intervals noticed here. Unfortu-
nately, the paper of Stoney, on " The Logarithmic Law of Atomic
Weights " f is so abridged, that it is hard to determine how his results
may compare with those herein discussed. The abstract states, that,
in plotting the atomic volumes as ordinates of a diagram, the effort to

* The Periodic Law, p. 20.

t The Chemical News, vol. lvii. p. 163.

OF ARTS AND SCIENCES. 175

extract information from the resulting curves was a failure ; iu plot-
ting their cube roots, however, he found a logarithmic curve with per-
turbations which showed a distinct law of increase for perissads and
aitiads. The abstract was not discovered until most of this paper was
written, and, from its indefinite statements, it would require much time
to construct the curves and divine the inferences.

At the close of his paper of 1854, Professor Cooke said : " To my
conceptions Chemistry will then have become a perfect science, when
all substances have been classed in series of homologues, and when we
can make a table which shall contain, not only every known substance,
but also every possible one, and when by means of a few general
formula? we shall be able to express all the properties of matter, so
that, when the series of a substance and its place in the series are
given, we shall be able to calculate, nay predict, its properties with
absolute certainty. . . . Then the dreams of the ancient alchemist
will be realized, for the problem of the transmutation of metals will
have been theoretically, if not practically, solved."

176 PROCEEDINGS OP THE AMERICAN ACADEMY

XV.

CONTRIBUTIONS FROM THE JEFFERSON PHYSICAL
LABORATORY, HARVARD UNIVERSITY.

NEUTRALIZATION OF INDUCTION.
By John Trowbridge and Samuel Sheldon.

Presented May 28, 1889

The invention of the telephone drew attention to the extraordinary
sensitiveness of Faraday's electrotonic state, and immediate attempts
were made to construct induction balances, so called, which might serve
for quantitative measurements. Thus we have Hughes's induction
balance, which had its prototype in the balance described in Max-
well's " Electricity and Magnetism," Vol. II. § 636, due to Felici,*
and which differs from Hughes's balance merely in the employment
of a galvanometer instead of a telephone. By substituting the lat-
ter instrument, Hughes showed that great sensitiveness could be
obtained, and even proposed to adopt an instrument for measuring
minute amounts of impurities in coins arising from alloys.

The great difficulty, however, in the employment of Hughes's induc-
tion balance in quantitative work arises from the difficulty of getting
a good minimum of tone in the telephone. The method that Hughes
employed was, briefly, to employ four coils, — two in a circuit through
which an alternating current or an interrupted current was passed,
and two other coils placed contiguous to the coils which were in the
interrupted circuit, but in another circuit.' By interposing a telephone
in the last mentioned circuit, and by properly placing the coils in this
circuit with reference to those in the circuit through which the inter-
rupted current was passed, a balance could be obtained, or an imperfect
minimum of sound in the telephone, when the induction between the
sets of coils was neutralized. In order 'to obtain a standard, Hughes
employed a wedge of zinc, which was thrust between one of the coils
in the interrupted circuit and one of the coils in the telephone circuit,
in order that the mutual induction between these coils might balance

* Nuovo Cimento, vol. ix. p. 345, 1859.

OP ARTS AND SCIENCES. 177

that arising between the other two similarly placed coils whin a
coin or sheet of metal was placed between these last mentioned coils.
Other devices have also been employed by various investigators who
have endeavored to use the apparatus for quantitative measurement.
Alexander Graham Bell employed a modification of Hughes's induc-
tion balance for the detection of the presence of a bullet in the human
body. In the form employed by him, one coil, which was a closely
wound flat copper band, was made to slide over a similar one by
means of a screw, one coil being placed in the telephone circuit and
the other in a circuit containing a current-breaker. The induction
arising from a similar pair of coils moved over a mass of metal like a
bullet could thus be neutralized by this sliding coil arrangement. In
no form, however, of Hughes's induction apparatus can one obtain a
satisfactory minimum of tone in the telephone. There is never abso-
lute silence, and no two observers can obtain the same point at which
the sound seems to be a minimum. The failure to obtain this min-
imum is thus a radical defect in the instrument. It is doubtless very
sensitive, but it cannot be called a quantitative instrument.

To remedy this defect, A. Overbeck and J. Bergmann * substituted
an electro-dynamometer for the telephone, and worked out a method
of obtaining the resistances of metals when they are in the form of
thin circular plates. The standard of comparison they employed was
a thin layer of mercury between disks of glass in a cylindrical reser-
voir. Preliminary investigations had shown the authors that a cer-
tain relation existed between the thickness and specific resistance and
coefficient of induction of metals in the form of thin disks, which were
placed between the coils of the induction balance. In a subsequent
paper,f A. Overbeck gives the mathematical theory of the induction
balance, which in the main is Maxwell's theory of current sheets
applied to Arago's disk.$ In employing the instrument to measure the
effect of change of temperature on induction in copper plates, or, in
other words, temperature coefficients, in which we found thai Messrs.
Overbeck and Bergmann had anticipated us,§ we were led to adopt
the following form of the instrument, which differed entirely from that
of these authors. Four coils were employed, as in the Hughes lorn.
of instrument. One of the coils in the telephone circuit was fixed
upon a horizontal axis which was at right angles to the axis of the

* Annalen der Physik, xxxi., 1887, p. 792.
t Ibid., p. 812.

| Maxwell's Electricity and Magnetism, vol. ii. § 668 et seq.
§ Annalen der Physik, xxxvi., 1880, p. 783.
vol. xxiv. (h. s. xvi.) 12

178 PROCEEDINGS OP THE AMERICAN ACADEMY

coil. The coil could therefore be moved through all positions, from
perfect parallelism to its. neighboring coil in the interrupted circuit to
a position at right angles to this coil. The horizontal axis was pro-
vided wiih an index arm which moved over a graduated circle. Call-
ino- 6 the anjrle of inclination of the axis of the movable coil with the
axis of the fixed coil in the interrupted circuit, and N the strength of
the induction current in the movable coil, we have evidently, on the
supposition that the strength of the alternating current remains constant,

N = constant X cosine 6.

When the axes of the coils are at right angles, cosine 6 = 0, and we
should have silence in the telephone. Since adopting this arrange-
ment we have discovered that Dr. Bowditch, of the Harvard Medical
School,* has employed this arrangement of a movable coil placed in
front of a fixed coil as a modification of Du Bois Reymond's appa-
ratus for controlling induction currents so that they may be admin-
istered by known amounts for physiological purposes. In Du Bois
Reymond's apparatus one induction coil was simply moved away from
a fixed coil through which an interrupted current was passed, much
in the same manner as the coils in Wiedemann's form of galvanometer
are moved. Here no minimum could be obtained. In Dr. Bowditch's
form of this apparatus, theoretically a minimum should be obtained, that
is, when cosine 8 = 0, or when the axes of the coils were at right angles.
An indication of an electrical current is obtained even when the axes
of the coils are at right angles, on account of the windings of the coil
not being perfectly at right angles to those of the stationary coil.

That no minimum should be obtained when the axes of the coils
are at right angles, and when induction arises from all parts of the
circuit, is evident upon an elementary consideration of the subject.
We have to deal in this form of instrument with the mutual induction
which arises between the fixed coil and the movable one, and also with
the self-induction which arises between the spires of the movable coil
in the telephone circuit. The mutual induction can be reduced theo-
retically to zero by placing the movable coil of the telephone circuit
at right angles to the fixed coil. The self-induction can be estimated
as follows. Taking Maxwell's discussion for the induction between
parallel circuits of radii A and a, we have the coefficient of mutual
induction,

cos e ds dsf

M

-ff-

* Proc. Am. Acad., vol. xi. p. 281.

OP ARTS AND SCIENCES. 179

Projecting one circle upon the plane of the circle of greater radius, A,
we have

M _ C2n C2n A a cos (ft — 0') dcf> <lcj>'

Jo J> V^M-o2 -R>2 — *- Waco's (ft — ft') '

Making 6 the distance between the planes of the circles = 0, we pass
from the case of mutual induction to that of self-induction between
two spires of a coil which may be considered approximately circular.
The form of M adapted for calculation is then

M= 4tt \/Aa j(c — -) F+ ?.#]•,

where c = -—-. — -, and F and E are complete elliptic integrals to
(A-\-a) i t- s>

modulus c.

If we make A — a = B, or A = D + a, in which D is the distance
between the spires at which the self-induction becomes insensible, the
most perfect minimum can be attained. We have found that copper
wire of 2 mm. diameter, wound in a flat loose spiral, the spires of
which from centre to centre of the wire are 4 mm. apart, gives no
sensible self-induction for spirals of eight to ten spires. On turning
a movable coil of this form so that its axis may be perpendicular
to the axis of the fixed coil, a perfect minimum can be obtained.
A slight movement to the right or left of this position is quickly
made evident by the note of the interrupted circuit which is heard
in the telephone. It is evident that, if four coils are employed, as
in Hughes's form of induction balance, the two coils in the telephone
circuit should be wound in the manner we have indicated, to avoid
self-induction. On placing a plate of metal between one set of the
coils of this balance, the movable coil no longer gives ;i minimum at
the position where its axis is at right angles to that of the fixed coil,
but at some point removed a few degrees from this. By placing a
mirror upon the movable coil, and by observing its deflection with a
telescope, a greater refinement of reading is possible.

This instrument in its modified form suggests the possibility of
neutralizing induction upon telephone circuits. The extension of the
various systems for transmitting power by electricity, especially the
electric car system, has led to great disturbances in the telephone
circuits. These disturbances are due both to leakage from the power
circuit into the telephone circuit, since the earth is used partially by
the electric power companies in their return circuits, and to actual
induction. The best remedy for these disturbances is doubtless the

180 PROCEEDINGS OF THE AMERICAN ACADEMY

adoption by either the power companies or the telephone companies
of entire metallic circuits, in which the earth plays no part. If this is
not possible, a system of neutralization for the inductive disturbances
might be adopted as follows. Let a shunt circuit from the electric
light wire or the wire carrying the current for motors be led into
a station through which also passes the telephone wire. The resist-
ance of this shunt or derived circuit can be made suitable for the pur-
pose. In all cases it reduces the resistance of the main line, and is
therefore not prejudicial. On this shunt can be arranged a fixed coil,
and on a neighboring telephone wire a movable coil of no self-induc-
tion. Let this movable coil be placed in front of the fixed coil in the
motor circuit, and let it be turned until the. mutual induction between
it and the fixed coil neutralizes the induction produced at all points
along the telephone circuit. Each telephone wire would need its
movable coil, and to every movable coil would correspond a fixed coil
in the shunt of the motor circuit. The operator at the central sta-
tion could adjust the movable coils until the disturbances arising from
induction at various points along the line are neutralized.

OF ARTS AND SCIENCES. 181

XVI.

THE MAGNETISM OF NICKEL AND TUNGSTEN

ALLOYS.

By John Trowbridge and Samuel Sheldon.

Presented May 28, 1889.

Introductory.

The fact that different kinds of steel, alloyed in small proportions with
Tungsten or Wolfram, and magnetized to saturation, increase in spe-
cific magnetism * has long been known. Whether the same effect
would result from the use of Nickel alloyed with Tungsten has never
been investigated. This paper has for its object a partial answer to the
query. It was instigated by Mr. Wharton, proprietor of the American
Nickel Works, whose chemist, Mr. Riddle, kindly prepared the alloys
which have been employed. These alloys were in two groups. The
first, received in November, 1888, consisted of three bars of the same
shape, one being of pure nickel and the other two having respectively
3 and 4 per cent of tungsten in alloy. These bars were rolled from
cast ingots, which were toughened by the addition of magnesium
after Fleitmann's method, the magnesium being added just before
pouring. They were hot when rolled. The one of pure nickel was
afterwards planed into regular shape. Those containing tungsten were
too brittle to allow of this manipulation. They were, however, of
sufficient regularity to permit accurate measurements. This group
contained also an octagonally shaped bar with 8 per cent of tungsten,
which was prepared like the others, and was afterwards ground into
shape.

The second group, received in May, 1889, contained bara which
were simple castings, m;ule without the addition of magnesium, and
consisted of pure nickel and alloys with 1, 2, 3, and G per cent of
tungsten. All the bars in this group were extremely hard and brittle.
In making them, tungsten oxide, of weight calculated to yield the de-
sired percentage of tungsten in the resulting alloy, was placed with

* Jour. Chem. Soc, 1868, xxi. 284, says 300 per cent.

182 PROCEEDINGS OP THE AMERICAN ACADEMY

adequate carbon in the bottom of a graphite crucible and covered by
the proper weight of pure grain nickel. All was then covered with
borax, the lid of the crucible was placed on, and the crucible was
heated until reduction and fusion wrere completed.

Method.

As the suspected influence of the tungsten would be to affect the
magnetic moment of the bars, these were magnetized to saturation and
their specific magnetism then determined, i. e. the magnetic moment
for each gram of metal.

The magnetization was effected by placing the bars separately in a
hollow coil whose length was 15 cm. and outside and inside diameters
respectively 6 and 3 cm. It consisted of 6 layers of wire having 63
turns each. A dynamo current of 40 amperes was then sent through
the coil for one minute, and the circuit then broken and the bars
removed.

For the determination of the magnetic moment, use wTas made of a
reflecting magnetometer, and deflections were observed with a telescope
and scale at a scale-distance of 100 cm. Measurements of the hori-
zontal intensity, H, of the earth's magnetism were first made. The
results from these determinations by means of the first and second
Gauss arrangements were, respectively,

H= 0.1724 cm. g. s.
H= 0.1720

The freshly magnetized bars were then placed in the second Gause
position relative to the magnetometer, and the angular deflection de-
termined. The specific magnetism, S, was then calculated by the
formula

c rz H tan <f>

£) = - —

111

where

r = distance from bar to magnetometer = 72.68 cm.

H = earth's horizontal intensity = 0.1722.

m = mass of the bar.

<j> — angular deflection of magnetometer.

Results.

The mean results of two sets of observations on Group I., and also
upon a similar bar of soft tool steel are given in the following table.

OP ARTS AND SCIENCES.

183

GROUP I.

Composition.

Size in cms.

Mass in grams.

/S[cm.sg.-J sec.-'].

Pure Nickel

18 X 2.7 X 0.65

284.5

1.23

Ni + 3%W

it tt u

286.5

10.60

Ni + 4%W

it (i u

283.5

10.40

Tool Steel

15 X 2.5 X 05

159.5

7.46

Ni + 8%W

/ Octagonal \
U3 X 1.5/

144.0

5.25

Group II., of cast bars, gave the following results.

GROUP II.

Composition.

Size in cms.

Mass in grams.

S [cm. 5 g ~i sec.-1].

Pure Nickel

18 X 1.8 X 1.6

459

1.05

Ni + 1 % W

it tt tt

455

1.92

Ni + 2 % W

tt tt tt

454

1.70

Ni + 3%W

tt It tt

463

1.75

Ni + 6%W

tt tt tt

465

1.15

The bars of both groups were, subsequent to the above observations,
completely demagnetized, and then freshly magnetized. New deter-
minations save the same results as before. The demagnetization was
accomplished by placing the bars inside two coils, which were traversed
by currents from an alternating dynamo. The coils were then slowly
drawn apart, and the bars maintained at a position central between
them. After treatment in this manner, they showed no appreciable
deflection when placed in position relative to the magnetometer.

The results tabulated indicate that tungsten greatly increases the
magnetic moment of nickel, if the alloy he forged and rolled, bul on
the other hand has but small influence if they be simply east. Fur-
thermore, changes in the amount of tungsten do not appear to cause
corresponding changes in the magnetic properties.

To see whether the remarkable effect in bars 2 and 3, aa compared
with bar 1, of Group I. was owing to some molecular condition oi their

184 PROCEEDINGS OF THE AMERICAN ACADEMY

surfaces induced by rolling, two bars from the same steel, one rolled
and the other pressed, were magnetized and then measured. The
ratio of the specific magnetism of pressed to rolled was as 9 to 5, the
rolled having the smaller amount. The existing difference, in this
case, is probably owing to a difference in hardness, rather than to any
molecular condition of the surfaces.

The specific magnetisms of all the bars are small when compared
with good steel magnets. Kohlrausch says that good magnets, of
common form, should have S = 40. The bar of ordinary tool steel,
however, retained but 7.46. Still it was soft, and by tempering would
doubtless have doubled this value.

If forged nickel and tungsten can be made to maintain a specific
magnetism of 10, it will form a useful addition to the resources of
physical laboratories. From the high polish of which it is suscepti-
ble and its freedom from damaging atmospheric influences, it will be
most happily suited for the manufacture of mirror magnets where
magnetic damping is to be employed.

Jefferson Physical Laboratory.

OF ARTS AND SCIENCES. 185

XVII.

THE MECANIQUE CELESTE OF LAPLACE, AND ITS
TRANSLATION WITH A COMMENTARY BY BOW-
DITCH.

By Joseph Lovering.

Presented June 8, 1889.

Laplace was born in Beaumont-en-Auge on March 28, 1749. The
early years of his life are hidden in the obscurity of his humble
origin. It is only known that he was the son of a small farmer.
He first gained distinction in a theological controversy. After attend-
ing some of the classes in a military school in Beaumont, he taught
mathematics there. At the age of eighteen, he approached the great
D'Alembert with the hope of finding some career in Paris, and was
rebuffed. Afterwards, Laplace sent to him a letter on the Principles
of Mechanics. D'Alembert wrote back : " You see that I make little
enough of the matter of recommendations. You have no need of
them. You have done better : to know. This is enough for me : my
support is your due." Thus he mounted at once to the position of
Mathematical Professor in the Military School at Paris. Before en-
tering the Paris Academy, at the age of twenty -four, he began his career
of investigation by which he won the title of the Newton of France,
having made a capital discovery relating to the mean distances of
the planets from the sun. Laplace was also an investigator in phys-
ics and chemistry, working with Lavoisier and Berthollet. Only
fifteen days before his last sickness, he communicated to the Memoirs
of the Academy a paper on the oscillations of the garth's atmosphere,
which was printed in the same volume which contained Poisson's
funeral oration.

Of his three great works, the five volumes of the Micanique
Celeste were published between 1799 and 1827 ; the Exposition du
Systeme dn Monde in 1796; and the Theorie analytiqne des Proba-
Mites in 1812. Arago reported, in 1842, that not a copy of the
last work was in the libraries of Paris, although three editions had
been published, and that the two other works were almost out of

186 PROCEEDINGS OF THE AMERICAN ACADEMY

print, so that he feared that it would be necessary to order from
America Bowditch's translation. In this emergency Madame de
Laplace, always devoted to the reputation and memory of her dead
husband, was negotiating for the sale of a small farm in order to pro-
vide tbe means of republishing his works, when the government of
Louis Philippe took the matter in hand and appropriated forty thou-
sand francs for a new edition. This appeared, in seven quarto vol-
umes, between 1843 aud 1847. At a later period, when this edition
was nearly exhausted, General Laplace, the distinguished son of the
mathematician, and his granddaughter, the Marquise de Colbert, ex-
pended seventy thousand francs on a third edition, which appeared
between 1878 and 1886.

Laplace was the friend of young mathematicians and physicists,
such as Arago, Poisson, and Biot, who were called " the adopted
children of bis thought." At an early age, while still a pupil in the
Polytechnic School, Arago was attached to the Paris Observatory as
secretary. He has given in his Autobiography the following inter-
esting sketch of Laplace. He says : " I entered this establishment,
then, on the nomination of Poisson, my friend, and through the inter-
vention of Laplace. The latter loaded me with civilities. I was
happy and proud when I dined in the Rue de Tournon with the great
geometer. My mind and my heart were much disposed to admire all,
to respect all, that was associated with one who had discovered the
cause of the secular acceleration of the moon, had found in the move-
ment of this satellite the means of calculating the ellipticity of the
earth, had traced to the laws of gravitation the long inequality of
Jupiter and Saturn, etc., etc. But what was my disenchantment when,
one day, I heard Madame de Laplace approach her husband, and say
to him, 'Will you intrust to me the key of the sugar ?':

Another anecdote told by Arago is not so complimentary to La-
place. Delambre, the Perpetual Secretary of the Academy of Paris,
died on the 19th of August, 1822. A committee was appointed to
present candidates for the succession. The choice lay between Fou-
rier and Biot. The election was always by secret ballot, — Arago
frankly saying that no one wished to incur the disaffection of a suc-
cessful candidate against whom he had voted. Laplace wrote the
same name upon two ballots, put them in his hat, and shook them up.
He took one out and tore it, and put the other in the urn, pretend-
ing not to know for whom he had voted, though an indiscreet neighbor
did. No calculus of probabilities, Arago says, was needed for this
problem.

OF ARTS AND SCIENCES. 187

Twenty-three years after the death of Laplace, Biot confided to the
Academy of Paris these interesting reminiscences of him. " Every
one," he says, " will understand how valuable to a young man were
these familiar and intimate interviews with a genius so powerful and
so comprehensive. But what he might not imagine were the senti-
ments of affectionate and paternal delicacy which attended them,
unless he himself had been the object of them. Shortly after I was
permitted to approach him, I had the good fortune to take a step,
which seemed to me new and unexpected, in a field of mathematics
hitherto scarcely invaded. In 1766, Euler had treated a peculiar class
of geometrical questions by an indirect method in a paper entitled JJe
inslgni promotione methodi tangentium inversce.* Subsequently, he
attacked a more difficult problem of the same kind, to which he re-
peatedly returned with different solutions, but always indirect. The
singularity of the problem originated in the nature of the curve, the
geometrical characters being of different orders, — some belonging
to points infinitely near, others to distant points separated by finite
differences. I succeeded in rendering a complete statement of the
problem in analytical language, applying to each part appropriate
symbols. The realization of this idea surpassed my expectations. f

" After I had found the key to a solution of the problem, I spoke to
Laplace about it. He listened with attention and with some surprise,
and then said, ' It seems all very well ; come to-morrow morning
and bring your memoir ; I shall be glad to see it.' After hearing it,
discussing it, and suggesting certain omissions, he told me to present it
to the Academy the next day, and, after the session, come and dine
with him. 'Now,' said he, 'let us lunch.'"

Biot communicated his paper to the Academy in the presence of
Monge, Lagrange, Laplace, and of Bonaparte who had just returned
from Egypt. Bonaparte pretended to recognize the diagrams, though
no one had ever seen them before except Laplace. " I had more fear
of Lagrange," said Biot, "than of Bonaparte, with all his military
glory."

After Biot had read his paper, and received the congratulations of
the Academicians, he accompanied Laplace home. As soon as they
had entered the house, Laplace took Biot into his study, unlocked a
closet, took out a copy-book, yellow with age, and showed him all

* Nov. Comment. Petrop., x.

t Considerations sur les equations aux differences mdlfes j Mem. Saratlta

Etrangeres, i 296-328.

188 PROCEEDINGS OF THE AMERICAN ACADEMY

the problems resolved by the method which Biot had supposed to be
original with himself. When Biot read his paper to Laplace he was
cautioned against certain inferences at the end which Laplace thought
too remote. " Go not," he said, "beyond the results which you have
actually obtained. You will probably meet difficulties more serious
than you suppose : the actual state of analysis cannot furnish you the
means of surmounting them." And Biot struck out his hazardous
inferences. It now appeared that Laplace had been arrested by the
same obstacle which he had signalized to Biot. Hoping to surmount
it later, he had said nothing about it to any one, not even to Biot,
when he came to bring him his own work as if it were a novelty.
Biot does homage to the abnegation of Laplace in allowing him to
make his communication without the consciousness of having been
anticipated. Laplace would allow no allusion to be made to the trans-
action when Biot's paper was printed. No mention of it appears in
the records of the Academy. Not until after fifty years did Biot feel
at liberty to mention it. " This discovery, the first I had made, was
everything for me. It was little for him who had made so many.
Would he always have been so just? Would he be so generous to a
rival ? I have no occasion to examine that here. He was all this
for me, and for many others who began their career in this way."

Biot has described the country-seat of Laplace in Arcueil, adjoin-
ing that of the eminent chemist, Berthollet. " It was acquired by
him in 1806, two years after the Emperor, Bonaparte, had promoted
him to the first dignities of the Senate. He bought it without having
seen it, on the report of Madame Laplace, being contented to know
that it was adjacent to that of his friend Berthollet. A simple garden
wall separated them. Berthollet had cut an opening in it and placed
a gate there before Laplace arrived. Then he came to receive him
with ceremony at the boundary of their domains, bringing the keys
which would give them free access to each other. It was in this de-
licious retreat that Laplace spent every day, every moment, of liberty
that his business left to him, — not to give himself to indolent repose,
but to continue, with unwearied passion, his great labors on mathe-
matical physics and the system of the world, emerging from these
meditations only to converse with his friend on physics and chemistry.
Here he also welcomed a retinue of zealous young men, whom he
condescended afterwards to call his colleagues, and who always were
proud of this association. Around him, in a more elevated sphere,
were seen continually Berthollet, Lagrange, Cuvier, etc., to whom he
introduced his vouug proteges. Here Laplace died, on March 5, 1827.

OF ARTS AND SCIENCES. 189

His last words were, ' Ce que nous connaissons est peu de chose ; ce
que nous ignorous est immense.' This sanctuary of sciences has been
preserved, with a religious respect, by Madame Laplace, to whom it
belongs to-day (1850). The house, the gardens where he walked, are
as they were then. The study in which he composed so many noble
works remains untouched, with the same furniture and books that
served him iu the state in which he left them. He alone is wan tin sr,
to the profound regret of those who knew him and will never see his
equal."

If we turn from this attractive picture of the domestic and scientific
life of Laplace to his public career, the contrast is not pleasant. Liv-
ing in the most disturbed period of France, he managed to keep him-
self always iu the ascendency. After the coup d'etat of Napoleon,
in 1799, his republican ardor was replaced by devotion to the First
Consul. But his incapacity as Minister of the Interior led to his re-
moval in six weeks. It was charged against him that he brought into
his administration the principles of the infinitesimals. As a consola-
tion, he was given a place in the Senate, of which he was made Chan-
cellor in 1803. He was also Grand Officer of the Legion of Honor.
On the erection of the Imperial throne, in 1804, he was made a
Count. In 1814, he gave his voice for the Provisional Government,
the deposition of Napoleon, and the restoration of the Bourbons.
Louis XVIII. rewarded him by a seat in the Chamber of Peers, and
in 1817 by a Marquisate.

The first edition of the Exposition du Sjjsteme da Monde was in-
scribed to the Council of Five Hundred. The third volume of the
Mecanique Celeste, published in 1802, was a tribute to the Pacificator
of Europe. In the edition of the Theorie analytique des Probabiliies,
published after the restoration, the dedication to the Emperor in an
earlier edition was suppressed.

Bonaparte recognized the splendor which the great intellect of
Laplace shed upon his administration. On October 19, 1801, ha\
received a volume of the Mecanique Celeste, he wrote to the author:
"The first six months at my disposal will be employed on your beauti-
ful work." On November 26, 1802, after reading some chapters <>f a
new volume dedicated to himself, he refers to "the new occasion for
regret that the force of circumstances had directed him to a career
which led him away from that of science. At least, he added, I de-
sire ardently that future generations, reading the Mecanique Celeste,
should not forget the esteem and friendship I have borne to the au-
thor." On June G, 1805, the General having become Kmperor. he

190 PROCEEDINGS OF THE AMERICAN ACADEMY

wrote from Milan : " The Mecanique Celeste seems to me called to
give a new splendor to the age in which we live." On August 12,
1812, he wrote from Witepsk: "There was a time when I should
have read with interest your Traite da Calcul des Probabilites. To-
day, I must limit myself to testifying to you the satisfaction I experi-
ence every time I see you communicate new works which perfect and
extend the first of the sciences and contribute to the glory of the
nation. The advancement and perfectionatiou of mathematics are
bound to the prosperity of the state."

Arago did not imitate the political subserviency of Laplace, much as
he admired his genius. In early life he went to Spain with Biot, to
complete a geodesic operation. This expedition, though full of adven-
ture and danger, had a successful issue, by giving a more precise value
to the magnitude of the earth. On his return he was made a member
of the Paris Academy when scarcely twenty-three years old. His sci-
entific career was brilliant, and in 1830 he replaced Fourier as Perpet-
ual Secretary. In 1 848 he espoused the cause of the second republic,
with Louis Bonaparte as President. After the second coup d'etat, he
refused to take the oath of allegiance to the new government, required
of all officials. Nevertheless, he was allowed to retain the post of as-
tronomer in the Bureau de Longitude, the new Emperor " making an
exception in favor of a savant whose works had thrown lustre on
France, and whose life his government would regret to embitter." As
Minister of the Marine, Arago had originated and carried an act for
abolishing slavery in the French colonies. Being urged to delay the
change and make it gradually, he declared, " I will not postpone till
to-morrow an act which sets free the oppressed."

Poisson was of humble origin, but rose to be Peer of France. He
befriended Arago, who was five years younger, and who describes him
in his youth as of delicate complexion, slight figure, and childish man-
ner. At the age of eighteen, he submitted to his teacher, Lagrauge,
some amelioration in the method of demonstrating the binomial the-
orem, which his teacher read publicly in his lectures, and said he
should adopt. Laplace wished to know a geometer who had made
such a beginning. Poisson was one of the earliest and the most
powerful in applying Fourier's definite integrals and periodic series
to physical problems. He extended the conclusions of Lagrange and
Laplace on the stability of the solar system, adding millions of years
to its probable duration. As there was no vacancy in the French
Academy in his proper section, he was placed in that of Physics.
Afterwards much of his work was upon physical problems, such as

OP ARTS AND SCIENCES. 191

electricity, magnetism, and capillarity. By common consent, he stood
at the head of European analysts, after the death of Laplace, whom
he succeeded as geometer in the Bureau de Longitude.

The publication of Dr. Bowditch's great work, a translation of four
volumes of Laplace's Mecanique Celeste, with an ample commentary,
was heralded by an article which he wrote for the North American
Review in 1820. After describing the works of Olbers and Gauss in
mathematical astronomy, the three principal publications of Laplace,
and especially his discovery of the great equation in the motions of
Jupiter and Saturn, with its long period of 917 years, and the neglect
by men of science in Great Britain of the vigorous science upon the
Continent, he thus expresses his admiration of Laplace: "These dis-
coveries, together with a multitude of improvements in analysis and
in every branch of mathematical knowledge, place this immortal man
far above any of his contemporaries in the walks of science."* Fourier
had called Laplace's works the Almagest of the eighteenth century.

Professor John Playfair, having alluded to the absence of diagrams
and geometrical figures in the writings of the great mathematicians of
France, says : " If we come to works of still greater difficulty, such
as the Mecanique Celeste, we will venture to say that the number of
those in this island who can read that work with any tolerable facility
is small indeed. If we reckon two or three in London and the mili-
tary schools in its vicinity, the same number at each of the two English
Universities, and perhaps four in Scotland, we shall not hardly exceed
a dozen : and yet we are fully persuaded that our reckoning is beyond
the truth." t

On the appearance of the first volume of Bowditch's translation, a
writer in the London Quarterly Review expresses his admiration in
these words : " The idea of undertaking a translation of the whole
Mecanique Celeste, accompanied throughout with a copious running
commentary, is one which savors, at first sight, of gigantesque, and is
certainly one which, from what we have had hitherto reason to con-
ceive of the popularity and diffusion of mathematical knowledge on
the opposite shores of the Atlantic, we should never have expected
to have found originated, or at least carried into execution, in that
quarter. Meanwhile, the part actually completed (which contains
the first two books of Laplace's work) is, with few and slight excep-
tions, just what we could have wished to see, — an exact and careful

* North American Review, 1820, vol. x. pp. 200-272.
t Edinburgh Review, 1808, vol. xi. p. 281.

192 PROCEEDINGS OF THE AMERICAN ACADEMY

translation into very good English, — exceedingly well printed, and
accompanied with notes appended to each page which leave no step
in the text, of moment, unsupplied, and hardly any material difficulty
of conception or reasoning unelucidated." Referring to the continu-
ance of the work, the writer adds : " Should this unfortunately not
be the case, we shall deeply lament that the liberal offer of the Ameri-
can Academy of Arts and Sciences to print the whole at their expense,
was not accepted.." *

In 1888, after three volumes of the translation and commentary
had been published, a writer in the London Athenaeum expresses
himself in these words : " Dr. Bowditch rose, like Franklin, from
humble life, and had much to struggle with, and did struggle man-
fully, and did succeed. He was, iu other words, an illustrious instance
of a self-educated man. . . . No matter what a man's facilities may
be, without this (the possession and use of the appropriating and in-
corporating power) he can never be educated, any more than he can
be healthy without sound bodily organs.

" The lack of facilities is not, however, to be spoken of lightly,
though it must not be confounded, nor yet compared, with the lack
of the power that uses them: nor must it be supposed that such a
thing as an absolute lack of facilities can exist. Nature herself has
provided facilities, food for education, materials for self-making
men to rise up, in times and places when and where no other facilities
may be had. She opens the great school-room of creation for them.
She gives them homes, society, the world at large. Above all,
she gives them eyes to see, and ears to hear, and an all-availing and
available spirit within them : the intellectual, immortal instinct, the
thirst for knowledge, and the faculty of finding it, in earth, and air,
and sea. No loads of appliances can surfeit such a mind on the one
hand ; no lack of them can starve it on the other.

" Let us not be supposed, then, to use the words as abusing them,
when we call Dr. Bowditch an illustrious instance of a self-educated
man. He was, in other words, as we understand it, an educated
man : his intellect informed and trained, his character seasoned and
consummated, and this under those circumstances of extraordinary
self-dependence, which, it is well known, are so admirably suited to
bring out and give their finest play to minds of a high order, on the
very same principle that they prove fatal to weaker ones. . . .

" Dr. Bowditch's great scientific work — the one on which his

* Quarterly Review, vol. xlvii., 18o2, pp. 558, 559.

OP ARTS AND SCIENCES. 193

European reputation rests and will rest — was a translation of the
Mjfanique Celeste of Laplace, accompanied with an extensive explan-
atory commentary, making a work, which existed uue may say in
mere abstraction before, as accessible to all public and popular pur-
poses as its essential nature would permit." *

Dr. Bowditch's work was not that of a mere translator and com-
mentator. The subject was brought down to the date of publication,
and was illustrated by the labors of geometers and astronomers who
had succeeded Laplace. Such mathematicians as Lacroix, Legendre,
Bessel, and Puissant recognized the great value of these additions.
3Ir. Babbage, in a letter to Dr. Bowditeh, under date of August 5,
1832, wrote: "It is a proud circumstance for America that she has
preceded her parent country in such an undertaking ; and we in Eng-
land must be content that our language is made the vehicle of the
sublimest portion of human knowledge, and be grateful to you for
rendering it more accessible." Letters of similar import were received
by Dr. Bowditeh from Airy, Francis Baily, Herschel, the Bishop of
Cloyne (Dr. Brinkley), and Cacciatore.

The Council of the Royal Astronomical Society of London, f in
noticing Dr. Bowditch's commentary and notes upon Laplace, ex-
presses its appreciation in these terms : " An expert mathematician
would find most of them {the notes) useless ; but to the student who
has sufficient knowledge to understand, without the habit of previous
investigation which the reading of Laplace always requires, the work
of Dr. Bowditeh is invaluable. We see in it, not the performance
of a practised analyst, but the record of the steps by which the trans-
lator became one ; and a person more familiar with the most modern
form of analysis than he appears to have been at the time when he
wrote would probably have filled his commentary with difficulties of
the same order as those of his author. This, however, was not done ;
and the work, as it stands, is most unquestionably fitted to bring the
Mtcaniqne Celeste within the grasp of a number of students exceed-
ing five times, at least, that of those who could master Laplace by
themselves.

"The name of Dr. Bowditeh must be long remembered in tin'
United States by the impulse which such a work as his commentary
cannot fail to give to analysis in that country. The undertaking re-
quired sound knowledge, power of combining brevity and clearness,

* London Athenaeum, 1838, pp. 451, 452.
t Memoir?, vol. xi. pp. 300, 301.
vol. xxiv. (k. s. xvi.) 13

194 PROCEEDINGS OF THE AMERICAN ACADEMY

an accurate remembrance of the nature of a beginner's difficulties,
and determined perseverance. The congratulations which this soci-
ety would so cordially have offered to the performer of this strikingly
useful task can now only be forwarded, with expressions of sympathy
and condolence, to his surviving relatives and friends ; with the hope
that astronomy, theoretical and practical, will flourish in the country
which has produced so remarkable a facilitation of the study of the
former, and so sound an example of the union of both."

In a note to the memoir of Dr. Bowditch by his son, it is stated
that in America two, and perhaps three persons, besides Dr. Bow-
ditch, were able to read the original work critically; but a compe-
tent judge has doubted whether the whole of it had been so read even
by one. It is also known, on the authority of his son, that Dr.
Bowditch was gratified whenever he received assurances that his
work had made Laplace accessible to young students. — more even
than he was by the approbation of his peers.

Benjamin Peirce, who was born in Salem, Mass., in 1809, and died
at Cambridge in 1880, and who was Professor of Mathematics or
Astronomy for forty-seven years, enjoyed the most intimate social
and scientific relations with Dr. Bowditch. How these relations
originated we are told by the writer of his obituary notice prepared
for the American Academy of Arts and Sciences.*

" In his early years he had the good fortune to come under the
influence of Dr. Nathaniel Bowditch. It is said that their first ac-
quaintance was made while Dr. Bowditch's son, Ingersoll, and young
Peirce were schoolmates. Ingersoll showed his comrade a solution
which his father had prepared of a problem that the boys had been
at work upon. Some error, real or conceived, was pointed out in the
work, which was reported by Ingersoll to his father. ' Bring me that
boy who corrects my mathematics ! ' was the invitation to an acquaint-
ance, the importance of which, in Professor Pierce's own estimation,
is told in the dedication, more than thirty years later, of his Analytic
Mechanics : ' To the cherished and revered memory of my Master in
Science, Nathaniel Bowditch, the father of American Geometry.'

"Peirce entered Harvard College in 1825. As Dr. Bowditch was
now in Boston, having removed from Salem in 1823, and was pre-
paring the first volume of his translation of Laplace's Mecanique
Celeste for the press, it followed almost as a matter of course that the
college student was more influenced in his studies by him than by

* Proceedings, vol. xvi. pp. 443, 444.

OF ARTS AND SCIENCES. 195

the college course. Dr. Bowditch's first volume was completed, and
the second entered for copyright, in 1829, the year of Peirce's gradu-
ation ; and the proof-sheets were regularly read hy him." *

That it should have been left to this country to open a way to
English- speaking nations for the study of the transcendental mathe-
matics of France and Germany may be explained by the following
lamentation of the younger Herschel, printed as late as 1845. f After
acknowledging the celerity with which everything valuable in British
journals is republished, examined, and criticised in those on the Conti-
nent, he writes : —

" This ought to encourage our men of science. They have a larger
audience, and a wider sympathy than they are, perhaps, aware of;
and however disheartening the general diffusion of smatterings of a
number of subjects, and the almost equally general indillerence to
profound knowledge in any, among their own countrymen, may be,
they may rest assured that not a fact they may discover, nor a good
experiment they may make, but is instantly repeated, verified, and
commented upon in Germany, and we may add too in Italy. We
wish the obligation were mutual. Here, whole branches of Conti-
nental discovery are unstudied, and indeed almost unknown even by
name. It is in vain to conceal the melancholy truth. We are fast
dropping behind. In mathematics we have long since drawn the rein,
and given over a hopeless race. In chemistry the case is not much
better."

At the present day Great Britain no longer lies under this sev
reproach. In pure mathematics there have been Boole, Hamilton,
Cayley, H. J. S. Smith, Clifford, and Sylvester; and in applied
mathematics, Adams, Stokes, Sir William Thomson, Lord Rayleigh,
Challis, and Powell. But the highest mathematical faculty will be
always the prerogative of the few. Not many can say with Lagrange,
" I found chemistry as easy as algebra ! "

The contrast is not less striking in this country between the pre*
condition of mathematical taste and acquirement and what they were
at the time of Bowditch's translation. This change has been lai

* In the memoir of Dr. Bo wd itch by his son it is said : " Whenever
hundred and twenty pages were printed, Dr. Bowditch had them bound in a
pamphlet form, and sent them to Professor Peine, who, in this manner,
the work for the first time. lie returned the pages with the list of errata,
which were then corrected with a pen or otherwise in every copy of the
whole edition."

t On Sound. Encyclo Metropolitan a, vol. iv. p. 810

196 PROCEEDINGS OF THE AMERICAN ACADEMY

due to the example and inspiration of Peirce. Though he was a poor
teacher for indifferent students, he was the best for superior intellects.
He simply led, and charmed them into following him as far and as fast
as they could. While his text-books were an enigma to those who
had little mathematical capacity, they fascinated minds of the highest
order in this direction. Hence, in every class in college, a very few
were found competent to read Bowditch's version of Laplace, and
obtain glimpses of the meaning of Peirce's ideal algebras. Observa-
tories, the United States Coast Survey, the Nautical Almanac, the
Smithsonian Institution, and schools of science exist, partly as cau!-e
and partly as effect of this new development. Bowditch's Practical
Navigator, first published in 1802, has carried his name, long ago,
over every ocean and into every port in the civilized world, and his
scientific reputation, as the earliest American expounder of the higher
geometry, is slowly following after.

The four volumes of Dr. Bowditch's Translation and Commentary
were published successively, in 1829, 1832, 1834, and 1839, at the
sacrifice of one quarter of hi whole property. The expense was
largely increased by the \ oluminous commentary. This was really
of the nature of an original work, and was rendered necessary by the
frequent ^aps which Laplace had left in his own publication. Mr. N.
I. Bowditch says, in his biography of his father, that Dr. Bowditch
himself was accustomed to remark, " Whenever I meet in Laplace
with the words, Thus it plainly appears, I am sure that hours, and
perhaps days, of hard study will alone enable me to discover how it
plainly appears."

The attention of Mr. James B. Francis of Lowell was attracted by
the following passage in a recent work of W. W. R. Ball of London :
" The Mecaniqne Celeste is by no means easy reading. Biot, who
assisted Laplace in revising it for the press, says that Laplace him-
self was frequently unable to recover the details in the chain of reason-
incr, ai d. if satisfied that the conclusions were correct, he was content
to insert the constantly recurring formula, 11 est aise a voir." *

Mr. Ball refers to the Journal des Savants, February, 1850. Mr.
Francis has found the volume in the Boston Public Library. The
title of the paper, translated, is, " An anecdote relating to Laplace :
read to the French Academy at its special session, on February 5,
1850, by Mr. J. B. Biot." I find that Biot reprinted the paper in
1858, in his Melanges Scientijiques et Litteraires. As Mr. Francis

* A Short Account of the History of Mathematics. London, 1388, p. 387.

OF ARTS AND SCIENCES. 197

has been at the pains of obtaining a translation of the paper by a lad v.
I shall adopt that with a few trifling changes.

" When a man of order prepares to depart on a long journey, he
arranges his affairs, and is careful to discharge all the debts he may
have contracted. This is why I am going to tell you how, some fifty
years ago, one of our scientists, the most distinguished, received and
encouraged a young debutant who had come to show him his first
attempts.

" This young debutant was myself, with your leave. Notice, to
excuse the epithet, that this goes back to the month Brumaire of the
year VIII. of the French republic, first edition.* Some months later,
they did me the signal honor of electing me an associate of the Na-
tional Institute ; but at that date, and especially at the little earlier
period when my story begins, I found myself completely unknown.
I was then quite a small professor of mathematics in the Central
School of Beauvais. Having recently left the Polytechnic School. I
had much zeal and less science. At that time, little was demanded
of young men but zeal. I was impassioned for geometry and many
other things. Fortune rather than reason saved me from yielding to
tastes too diverse. Bound from that time by the sweetest ties to the
interior of the family which had adopted me, happy in the present and
counting on the future, I only thought of following with delight the
bent of my mind towards all kinds of scientific studies, and of doing for
pleasure what the interest of my career would have prescribed to me
as a duty. I had, above all, an unbounded ambition to penetrate into
the high regions of mathematics, where men discover the laws of the
universe. But these grand theories, still scattered in the collections
of academies, were almost unapproachable, except to a small number
of superior men who had co-operated in establishing them; and to
launch into them without a guide over their footsteps, this would be
an enterprise in which one would have every chance of wandering for
a long time before rejoining them. I knew that Laplace had labor* A
in unifying this magnificent ensemble of discoveries in the work which
he has very justly called the Mecanique Celeste. The fust volume
was in press ; the others would follow, at long intervals, agreeably t ■<
my desires. A step which might appear very hazardous opened t.i
me a privileged entrance into this ?anctuary of genius. I ventured
10 write directly to the illustrious author, ami beg him to permit liis

* Brumaire began in 1790 on October 23: Biot was then twenty-five .years
ol 1.

198 PROCEEDINGS OP THE AMERICAN ACADEMY

publisher to send me the sheets of his book as they were printed.
Laplace replied to me with no less ceremony than if I had been a real
savant. However, as the end of the story, he refused my request ;
uot wishing, he said, that his work should be presented to the public
1 efore it was completed, that it might be judged as a whole. This
polite refusal was doubtless very obliging in its form ; but, at bottom,
it badly suited my business. I was not willing to accept it without
appeal. I immediately wrote again to Laplace to represent to him
that he did me more honor than I merited, and than I desired. I am
not, I said to him, of the public who judge, but of the public who
study. I added, that, wishing to follow and make again all the calcu-
lations throughout, for my own instruction, I might, if he yielded to
my request, discover and indicate the errors of press which may have
slipped into them. My respectful persistence disarmed his reserve.
He sent to me all the sheets already printed, adding thereto a charm-
ing letter ; this time not at all ceremonious, but filled with the most
lively and precious encouragements. I need not say with what ardor
I devoured this treasure. I could well apply to myself the maxim,
Violenti rapiunt Mud.

" After this, every time I went to Paris, I carried my work of typo-
graphical revision, and presented it in person to M. Laplace. He re-
ceived it always with kindness, examined it, discussed it ; and that
gave me an opportunity to submit to him the difficulties which too
frequently baffled my weakness. His condescension to remove them
was without bounds. But he himself could not always do it, with-
out giving attention to them ; sometimes pretty long. That happened
usually in places where, to save himself from the details of a too ex-
panded exposition, he had employed the expeditious formula, // est
aise de voir. The thing, in fact, had appeared at the moment very clear
to his eyes. But it was not so always, even for him, some time after-
wards. Then, if you asked him the explanation of it. he sought it
patiently, by different ways, on his own account as well as yours ; and
there was, without doubt, the most instructive of commentaries. Once
I saw him pass almost an hour trying to seize again the chain of rea-
soning which he had concealed under this mysterious symbol, 11 est
aise de voir. It must be said, for his acquittal, that if he had wished
to be completely explicit, his work would have required eight or ten
quarto volumes instead of five ; and perhaps he would not have lived
long enough to finish it."

These minute details relating to the intellectual life of Laplace, so
sacredly preserved by Biot, the eminent mathematician and physicist,

OF ARTS AND SCIENCES. 199

were not published until twenty-three years after the death of Laplace,
and twelve years after the death of Bowditch. Tliey must be taken
into account when the mathematical powers of the author of the Me-
canique Celeste and of its translator and commentator are compared.
There can be no doubt that a higher order of mind is demanded for
originating a profound analysis, than for understanding it when it is
once made. Dr. Bowditch said to one of his sons : " My order of
talent is very different from that of Laplace. Laplace originates
things which it would have been impossible for me to have originated.
Laplace was of the Newton class ; and there is the same difference
between Laplace and myself as between Archimedes and Euclid."
This confession, taken in connection with Bowditch's estimate of
Euclid, erred on the side of excessive modesty. Eor he was able
not only to detect and point out to Laplace accidental errors in his
work, but to provide an original commentary upon it more volumi-
nous than the work itself. If this required a great expenditure of
time and thought, it was no easy matter for Laplace himself. It was,
in fact, so hard, that Laplace was not willing to undertake it a second
time, after the subject had passed out of his mind ; and he left the
abyss open, with no stronger bridge to cross it than the simple for-
mula, 11 est aise de voir.

The circumstances under which Laplace and Bowditch did their
work are in striking contrast. Laplace was a favored child of for-
tune. He was able to give his long life of seventy-eight years to
his noble task, breathing an intensely intellectual atmosphere, and
surrounded by men of science only a little less distinguished than
himself. Bowditch's knowledge of affairs was as remarkable as his
science. He was weighted all his life with heavy cares and trusts,
and could spare only his hard-earned leisure for scientific work,
dying at the age of sixty-five. His intellect worked on a solitary
height, admired but not fully comprehended by those about him.
For example, instruction, and inspiration, he did as much for science

in America, and wherever the English language is read, as Lapli

did for that of France. It would be unjust to his memory, and to
his own great modesty, to make any other comparison.

No names more illustrious adorn the two original branches ol
membership in this Academy than those of Laplace and Bowditch.
Laplace was elected a Foreign Honorary Member in 1822, when he
had published only four volumes of the Mecanique I 'eleste. Bowditch
became a Resident Fellow in 1799, fifteen years before he had begun
his translation of Laplace's greatest work. He contributed twenty-

»
200 PROCEEDINGS OF THE AMERICAN ACADEMY

three valuable papers to the Memoirs, chiefly on Astronomy and the
Physics of the Globe. Having succeeded John Quincy Adams as
President of the Academy in 1829, he died in office, in 1838. An
elaborate eulogy was pronounced before the Academy, May 29, 1838,
by Mr. John Pickering, the Corresponding Secretary.

After the death of Laplace, his widow recognized the value of
Rowditch's labors, in extending the usefulness and enlarging the fame
of her husband's greatest work, by presenting to his translator and
commentator his marble bust. In the fourth volume of the last edi-
tion of Laplace's work, fifty-two corrections are specified as having
been made by Dr. Bowditch.

Dr. Bowditch's strength held out until he had corrected the proof-
sheet of the thousandth page of the fourth volume of his translation.
That was the last page his eye was to rest on. Eighteen more pa^es
remained. The incident has its parallel in the death of Copernicus.
He had spent twenty-three years in writing his book De Orbium
Ccelestium Revolutionibus, but delayed its publication for fear of a
popular outcry against his scientific doctrines. The book had been
printed at Nuremberg under the care of his disciple, Rheticus. A
few hours before he died, a copy arrived which was placed in his
hands. He touched it, and seemed to recognize it, and then relapsed
into unconsciousness.

No less suggestive was the death-bed of Lagrange. Though of
French extraction, he was born in Turin in 1736. His father lost
his property in speculation, — an event which the son did not regret.
" Had 1 been rich," he said, " I might never have known mathemat-
ics." At the age of twenty-six, Lagrange had reached the summit of
fame, though at the sacrifice of his health. In 1766, he was brought
to Berlin by Frederick the Great, " the greatest king in Europe
wishing to have the greatest mathematician at his court." After-
wards, various other courts competed for him. In 1787, he selected
Paris. He was patronized by Marie Antoinette, and lodged in the
Louvre. In the Revolution, when other scientific men were victims,
his name commanded respect. In reply to a petition in favor of
Lavoisier it was said, " The republic has no need of savants." Bona-
parte called Lagrange " La haute pyramide des sciences mathema-
tiques," and loaded him with offices and distinction. The first edition
of his Mecanique Analytique appeared in 178S. In 1811, he pub-
lished the first volume of another edition. The second volume, which
was on Dynamics, required more serious changes, and its preparation
was too much for his failing strength. On April 8, he had a conver-

OP ARTS AND SCIENCES. 201

sation with Monge and other friends, and looked forward to another
early interview, when he would complete some autobiographical de-
tails. But what he intended to say was never spoken. He died two
days afterwards, and was buried in the Pantheon.

Only one Fellow of this Academy is now living who was elected
before the death of Dr. Bowditch ; and there must be very few in its
present membership who knew him or ever saw him. But his work
and his reputation are their common inheritance. To keep alive his
memory, and do justice to his gigantic undertaking, is the loving ser-
vice of one who, in early life, was welcomed to the hospitality of his
thoughts and his books.

202 PROCEEDINGS OF THE AMERICAN ACADEMY

XVIII.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF

HARVARD COLLEGE.

ON A NEW METHOD OF DETERMINING
GAS DENSITIES.

By Josiah Parsons Cooke.

Presented June 12, 1889.

In the well known method employed by Regnault for determining
the density or specific gravity of air, oxygen, nitrogen, hydrogen, and
carbonic acid, we deal primarily with tares, of which the weights to
be determined are the differences. The glass balloon which holds the
gas is tared by a similar balloon of exactly the same volume and of
nearly equal weight, suspended from the opposite pan of the balance.
The small difference of weight required to establish perfect equilibrium
is alone measured with standard brass or platinum weights. What-
ever may be the form of the subsequent calculation, the primary ob-
ject is to obtain the tare of the empty balloon when absolutely vacuous.
This known, the differences between such tare and the tare of the
balloon filled with various aeriform substances, gives the weights of
equal volumes of these substances under the temperatures and pres-
sures at which the balloon was filled. The volume of the counter-
poise is exactly adjusted to that of the balloon by the aid of a small
subsidiary glass bulb (Plate II.) ; and by sealing up more or less mer-
cury in this bulb it is easy to make the difference of weight such that
the standard weights required to complete the equilibrium will measure
tlie differences of tare to be determined, and no more.

In the method of Regnault the tare of the empty balloon, or what
was equivalent to it, was found by exhausting the balloon with an
air-pump and weighing it after measuring the tension of the residual
gas while the glass was surrounded by ice. But it has been shown by
Agamennone * and Lord Rayleigh ,f that the results thus obtained are

* Atti (Rendiconti) d. R. Accad. dei Lincei, 1885.

i Proceedings of the Royal Society, vol. xliii. p. 362. 1888.

OF ARTS AND SCIENCES. 203

vitiated to a small extent by the circumstance that when the balloon
is exhausted the pressure of the atmosphere determines a slight shrink-
age of the external volume, which naturally disturbs the exactness of
the compensation between the buoyancy of the air on the balloon and
on its counterpoise. Although this shrinkage can be readily meas-
ured, as was done by Dr. T. W. Richards under my direction,* and still
more recently by Professor Crafts, t who experimented on the balloon
used by Reguault, which fortunately has been preserved, it seemed de-
sirable to develop a method by which this correction could be avoided.
For even if the new method should lead to no more accurate results
than before obtained, it might serve to confirm the validity of the cor-
rection iu question, and at least would gi\e additional data towards
establishing the value of important physical constants.

The new method we have devised for the purpose consists in first
taring the balloon when filled with carbonic acid gas, and then drawing
the gas through absorption tubes and determining its weight, as in the
well known method of organic analysis. This weight known, the tare
of the empty balloon is obviously the difference between the first tare
and the weight in question. The practical problem here presented is,
however, far more difficult than that of organic analysis. In the last,
we expect to determine the weight of only a few decigrams of car-
bonic acid within a few tenths of a milligram, while in the problem
now before us we must weigh at least nine or ten grams of carbonic
acid, not simply to a proportional, hut to an equal, degree of accuracy.
We only succeeded in securing such accuracy after many trials and a
careful study of all the conditions involved, and our primary object in
this paper will be to describe the precautions which are essential to
the success of the new experimental method. Incidentally it will appear
that our results confirm in a most striking manner the high value of
the specific gravity of hydrogen found by Lord Rayleigh,± and the low
value of the atomic weight of oxygen found by ourselves.

The Balance and Weights.

The balance and weights were the same as those used by us in our
previous work.§ The disposition of the apparatus whirl., after many
trials, we have found most suitable for accurate work, is shown in

* The.-e Proceedings, vol. xxiii. p. 177. 1888.

t Comptes Rend us, vol. cvi. p. 16G'2. 188b.

j Proceedings of the Royal Society, vol. xlv. p. 426. 1889

§ These Proceedings, vol. xxiii. p. 159. 1*

204 PROCEEDINGS OF THE AMERICAN ACADEMY

Plate II. The balance rests on the plank shelf of a wooden case,
whose glazed doors are shown open in the drawing ; and as this shelf
is firmly secured to a thick brick partition wall behind, great steadiness
is secured for the instrument. Around the iuside of the wooden case
are hung folds of drapery, which surround the balance case and protect
the beam from radiations, while a curtain rolling on a spring fixture
enables the experimenter to uncover the front as necessary. Fastened
by flanges and screws to the under side of the same shelf is a box,
made of tinned sheet iron, of somewhat larger dimensions than the
balance case. The front of this box can be almost wholly uncovered
by means of two large doors made of the same metal sheet, and se-
cured when closed by turning buttons, but shown open in the draw-
ing. In the metal top are two circular holes about half an inch in
diameter, corresponding to holes through the shelf and the base of
the balance case, and directly under the centres of the balance pans.
From hooks soldered to the under side of the pans the globe and its
counterpoise are suspended by means of brass wires, and the lengths
of the wires are so adjusted that the two glass vessels shall hang mid-
way in the metal box. The globe is hung from a wire stirrup, Plate I.,
which swings from an eye at the end of one wire; while on the neck
of the counterpoise is cemented a brass cap and hook, and both it and
the small subsidiary bulb are hung on an eye at the end of the other
wire. Of course these details might be varied. It is only important
that the whole system should swing freely, with as little friction as
possible, and that the balloon should be easily removed and replaced.

Through a third hole in the top of the metal box and shelf, placed
just in front of the balance case, passes a thermometer (not shown in
the drawiug) by which the temperature may be watched, and the in-
side of the box is painted with lamp-black so as to secure a uniform
temperature throughout the interior. In our previous paper we spoke
of the disturbance sometimes caused by currents of air within the box,
but we found on further experience that these currents could be pre-
vented by keeping the interior of the box as nearly as possible at the
temperature of the external air. For this purpose it is important that
the air of the room should circulate as freely as possible around the
box, and the required condition is more easily secured if the box is
made of thin metal sheet. It is also important that the temperature
of the room should not be constantly changing, and that all circum-
stances should be avoided which would cause a flow of heat either into
or out from the box. Sitting in front of the box in the process of
weighing immediately causes a disturbance, and hence the division of

OF ARTS AND SCIENCES. 205

the outer wooden case below the shelf is provided with wooden doors,
which should be shut in front of the metal box before approaching the
balance. When the thermometer hanging with its bulb inside the box
indicates the same temperature as one hanging from the balance case
outside within one tenth of a centigrade degree, the conditions are
most favorable for accurate weighing.

In order to maintain the atmosphere within the box in a constant
hygrometric condition, we placed in it four large open dishes of sul-
phuric acid, two on the bottum of the box and two on a shelf near the
top, as described in our previous paper. The acid was first boiled with
a fraction of a gram of amnionic sulphate, to remove the last traces
of nitrous fumes, and renewed as occasion required. ■

We have found that we can most accurately estimate slight differ-
ences of weight with a balance thus loaded by observing the amplitude
of the first swing of the pointer when the pans are relieved. With
the balances of Becker, after the beam has been set on its bearings by
the usual cam motion, the pans are still supported by two delicate arms,
which are held by a weighted lever against the under side of the pans,
but which can be pressed down by pushing against the button shown
in the drawing on the front of the balance case. Such a mechanical
arrangement is peculiarly adapted to this method of weighing, as it
relieves the pans suddenly and under the same conditions, so that the
amplitude of the subsequent swing is an accurate measure of the differ-
ence of weight acting at the two ends of the balance beam ; and in
this case the inertia of the heavy load renders the effects surprisingly
uniform. We give these details, not because they involve any new
principles, but because the results have been reached after a long se-
ries of experiments, and our experience may save others much loss of
time. With the apparatus arranged as we have described it, a dif-
ference of weight amounting only to one tenth of a milligram was
distinctly indicated, although the load exceeded half a kilogram on each
pan, and the compensation was made so perfect that the equilibrium
was not perceptibly altered by any ordinary changes of atmospheric
conditions.

In order to reach such compensation, a counterpoise is first select" <l
which on rough measurement appears to be nearly of the same volume
as the balloon, and at the same time a little lighter in weight. The
apparatus is then mounted with this counterpoise as described, the stop-
cocks of the balloon being closed. A series of observations are then
made of the variations of weight arising from changes of atmospheric
pressure and temperature, as carefully measured as possible. From

206 PROCEEDINGS OF THE AMERICAN ACADEMY

such observations we can readily calculate, by the method we have pre-
viously described in these Proceedings,* the volume of the small bulb
required to make the compensation perfect, and an experienced glass-
blower can readily make a bulb of the size indicated, which, before it
is sealed, may be adjusted as to weight by loading with mercury as
above described.

The weights used in connection with the balance had been adjusted
for our previous work with the greatest care, and not only compared
with each other, but also compared with a copy of the Washington
standard. Although we shall deal in this paper solely with relative
weights, yet the absolute values that may be deduced from our results
must be correct far within the limits of the experimental errors of our
own work. The sum of the weights required to measure the differ-
ences of tare never exceeded twelve grams, and of this small mass
any possible difference between our copy and the Paris standard kilo-
gram must be wholly inappreciable.

The Glass Balloon.

We have used in this investigation the same glass balloon that was
fitted up for a somewhat different purpose in our work on the atomic
weight of oxygen, but which has proved equally well adapted to deter-
minations of gas density. Regnault in his work on the same subject
used a balloon of ten liters' capacity, having its neck closed with a
brass cap and stop-cock, and filled it by alternately exhausting the in-
terior with an air-pump and refilling from a gasometer until the com-
plete purity of the enclosed gas was assured. After long experience
with this mode of experimenting, we have found it more convenient
to fill the balloon by displacement directly from the generator, and we
have found that we can reach a satisfactory result with least labor in
this way.

Our balloon is represented in Plate I., which shows all the details of
its construction. It was made by Emil Greiner of New York, and
all the joints are so finely ground that when properly lubricated they
will hold a vacuum indefinitely. The peculiar form of the stop-cocks
enables the experimenter to shut into the globe a definite volume of
gas and at the same time open a vent to the generator through the
side tube ; or to allow the gas to flow through the side tube until the
current is perfectly pure before it is admitted to the balloon ; and
again, when the stop-cocks are both closed and the balloon disconnected,

* Vol. xviii. p. 55.

OF ARTS AND SCIENCES. 207

all the nas left in the connecting tubes will soon be replaced by air.
At the freezing point the internal volume of the balloon, as nearly as
we have been able to measure it, is 4961.5 cubic centimeters, and the
weight of the glass 570.5 grams. After much experimenting, we have
concluded that as great accuracy in determinations of gas density ran
be obtained with a balloon of this size as with a larger one. The ad-
vantage to be derived from the greater mass of gas experimented on
is more than compensated by the greater liability to error which the
weighing of larger vessels and the longer duration of all the processes
involve.

The extent to which this balloon is actually compressed by the ten-
sion of the atmosphere when a vacuum is made in the interior has been
carefully measured,* and amounts to 1.66 cubic centimeters for a
difference of tension of 760 mm. between the interior and exterior of
the glass vessel ; and this corresponds to a loss of buoyancy of 1.96
milligrams in the air at 700 mm. and 22° C.

When the balloon was not hanging on the balance, it was always
protected by a metal case (see Plate II.), from which the neck ami
connecting tubes alone protruded, and in transferring the balloon to
and from the balance it was handled only by the neck.

In Regnault's experiments the balloon was filled with the gas und< r
examination at the temperature of 0° C. by surrounding the glass
with melting ice. This necessitates a careful cleaning of the globe at
least twice during each determination, and thus may arise slight varia-
tions of condition to impair the accuracy of the results. We have
found with Lord Rayleigh that it is preferable to fill the balloon at the
temperature of the laboratory, but we must then observe the tempera-
ture with very great accuracy. It is important to know the tempera-
ture to the one hundredth of a centigrade degree, for even so small a
difference of temperature as this causes a change in the amount of
carbonic acid gas which our balloon will hold under ordinary atmos-
pheric conditions corresponding to three tenths of a milligram.

We have found it possible to secure even such constancy of tem-
perature as this delicate measure implies by placing the balloon while
filling in a calorimeter case, and observing the temperature with a
calorimetric thermometer divided to the fiftieth of a centigrade degree.
The calorimeter case consists of a cylindrical vessel of sheet zinc wide
enough to receive the balloon with its tube and leave a play of lesa
than an inch on either side. This vessel is fastened by metal Btaya

* These Proceedings, vol. xxiii. \>. 182.

208. PROCEEDINGS OP THE AMERICAN ACADEMY

inside a second cylindrical vessel of the same material, leaving an an-
nular space about two inches wide and an equal space between the
bottoms of the two vessels, all of which is filled with water. In addi-
tion, the water vessel is further protected by a layer of felt one inch
thick kept in place by an outer covering of the same sheet metal.
When in use, the top is covered by a cap made also of thick felting,
and in the crown of the cap is a slit, through which a thermometer may
be inserted or the hand passed to turn the stop-cocks. The gas to be
used is conducted into the calorimeter through a very small lead tube,
which is connected with the lower tubulature of the stop-cock by a
rubber connection, and the balloon is so placed that the current shall
enter either at the top or at the bottom of the vessel, according as the
gas is lighter or heavier than the atmosphere. The calorimeter case
stands about four feet high, and the balloon rests in it on a down cush-
ion, with the connecting tube's fully six inches below the surface of the
water in the annular space above mentioned. "With this apparatus it
was possible to keep the temperature of the balloon constant within
the one hundredth of a centigrade degree for an hour at a time, unless
the temperature of the laboratory suddenly and greatly changed.

The general arrangement for filling the balloon was then as follows.
The gas delivered from the drying tubes of one of the generators here-
after to be described passed through the lead tube of which we have
spoken to one or the other, as the case might be, of the tubulatures of
the balloon standing in the calorimeter. Passing out by the opposite
tubulature, the current was conducted by a similar lead tube to a
second but much larger balloon serving as a gasometer, and from the
last vessel through a gas wash-bottle to the atmosphere. The bottom
of the wash-bottle was covered with strong sulphuric acid to dry any
returning air, but not deep enough for the inlet tube to dip under the
liquid. The current was continued until the proper test showed that
the gas issuing from the end of the apparatus was perfectly pure.
Then the inlet cock of the balloon was turned, and the temperature of
the calorimeter case watched until constant, and a sufficient time had
elapsed to establish a perfect equilibrium with the atmosphere through
the connecting tubes and vessels. This, as we found by experience,
often requires longer time than we were led to anticipate, and when
experimenting with carbonic acid gas it is not safe to allow less than
twenty or thirty minutes ; and now the large balloon and wash-bottle
come into play to prevent air diffusing into the balloon, or being
drawn back in consequence of changing pressure.

When all the conditions seemed to be satisfactory, the temperature

OF ARTS AND SCIENCES. 209

of the calorimeter and the height of the barometer were accurately
observed, the first to the hundredth of a centigrade degree and the
last to the twentieth of a millimeter, noting always the height of the
attached thermometer. The hand was then quickly thrust into the
calorimeter, aud the outlet cock at once closed. The balloon was now
removed, dusted with a large camel's-hair brush, and hung in its place
on the balance. When, however, the atmosphere is very dry, as it
often is in this climate during the winter months, it may happen that
the surface of the glass acquires a charge of electricity from the fric-
tion of the cushion in the calorimeter or of the brush just mentioued ;
and if there is auy suspicion of such an effect, the best way to dissipate
the charge is to stand the balloon for a few minutes in its metal case
(shown in Plate III.), after covering the bottom of the case with a thin
layer of water. If the balloon is hung on the balance with a charge of
electricity on its surface, however small, it will be unequally attracted
by the surrounding metallic walls, and most confusing anomalies ot
weight will be noticed. Moreover, in the very dry atmosphere of the
case the charge will last for days, or even weeks, as on the insulated
vanes of an electrometer.

Thermometers and Barometers, and their Correction.

The thermometers on whose indications we have relied in this
investigation are two of small range, but divided into fiftieths of a
centigrade degree and made by the house of the late Dr. H. Geissler
of Bonn. We were obliged to use the two, for our investigation was
continued into summer weather, which exceeded the range of the ther-
mometer first selected. As should always be the case with such in-
struments, the zero point is given in each case on a short subsidiary
scale, separated from the main scale by an enlargement of the tube.
Careful observation showed that the zero point of the instrument we
will designate as No. 1, was depressed 0.02 of a centigrade degree,
while the zero point of No. 2 was raised 0.28 of a degree. Making
the correction thus indicated, to be added to all temperatures observed
with No. 1, and subtracted from all observed with No. 2, the two in-
struments were found to agree exactly through all portions of the
scale common to both. Obviously, therefore, the differences of tem-
perature observed are perfectly trustworthy to the one hundredth of a
degree, but how far the absolute values — counting from the freezing
point of water — can be relied on when compared with the observa-
tions of other experimenters, we have no means of determining. Of
course, it will be understood that all temperatures hereafter given

VOL. XXIV. (n. S. XVI.) 14

210

PROCEEDINGS OF THE AMERICAN ACADEMY

have been corrected for the variation of the zero point. At the close
of the investigation the zero points were again tested and found to be
unchanged.

The barometer used was a large standard instrument of the familiar
mountain form, made by Greene of New York. Hanging at the side
of it was a smaller instrument of the same maker, whose indications
agreed with the first within the tenth of a millimeter. Here again, as
in the case of the thermometer, there can be no question as to the ac-
curacy of the differences of tension observed within moderate limits ;
but how far the absolute heights can be compared with those observed
by Regnault and others is also a problem that cannot be readily solved,
and one not within the means of a single experimenter to determine.
All heights of barometer given hereafter have been reduced to 0° by
the tables of Guyot.*

Tare op the Empty Balloon by Regnault's Method.

In the method of experimenting devised by Regnault the tare of the
empty glass globe used, although not explicitly given, was implicitly

determined. His arrangement for filling
the globe, shown by Figure 1, has already
been referred to, and his general method
of experimenting consisted in taking the
tare of the globe when filled with the same
gas at the temperature of melting ice under
as different tensions as possible. The ten-
sions were found by measuring with a
cathetometer the difference of height, a ft
(see Fig. 1), under each condition, and the
tares were found by means of the balance
by the system of compensation already de-
scribed. The difference of tare gives the
weight of gas which the globe contains at
0° at the tension H — h, and from this
may readily be deduced the weight at any
other tension ; as, for example, at the stand-
ard tension of 7 GO mm. The calculated
weight at either of the tensions H or h, deducted from the corre-
sponding tare, gives what we have called the tare of the empty globe ;
but obviously with Regnault's method this value is not needed in the
calculations of specific gravity or density.

Fig. 1.

Published by the Smithsonian Institution, Washington.

OF ARTS AND SCIENCES. 2 1 1

When the globe is filled by displacement it conduces to greater ac-
curacy to determine independently the tare in question ; and as soon
as the details of the new method had been worked out, we began our
final determinations by taking according to the old method the tare
of the balloon abovre described when empty, so as to obtain a definite
standard of comparison for further results. We however varied the
method of Regnault in so far that we always filled the globe with hy-
drogen before exhausting the interior. As with this exceedingly light,
material the weight of the residual gas seldom exceeded one half a
milligram, an error as great as one fifth in the capacity of the bal-
loon or in the tension of the residual gas would cause no appreciable
difference in the value of the small weight we sought to estimate.
The details of a single example will be sufficient to illustrate the
procedure.

Determination No. 1.

A current of hydrogen gas from the electrolytic generator described
in our previous paper * was run through the balloon from Saturday,
May 25th, at 4 o'clock p. m., to Monday, May 29th, at 10 o'clock A. M.,
1889, the balloon standing in calorimeter case connected with larger
balloon to receive overflow. At moment of closing, the height of
barometer and temperature of the case were observed, but these data
are not required for the present determination, and will be given
hereafter when used for calculating the specific gravity of hydrogen.
The balloon was exhausted with an excellent mechanical air-pump, f
and the tension of the residual gas measured by means of a siphon
manometer interposed between the balloon and the pump. The form
of manometer used is shown by Figure 2. It was made in the labo-
ratory, and in filling it with mercury the liquid metal, purified by
repeated distillation, was boiled in the tube with the greatest care.
For convenience, the difference of level of the mercury in the two
arms was measured with a cathetometer constructed by the Societe
Genevoise, but it might have been estimated with sufficient accuracy
by the millimeter scale of the instrument. Tension in balloon after
exhaustion 456.95 — 455.55 = 1.40 mm.

Assuming 0.4164 as the approximate value of the weight of hydro-
gen gas filling the balloon at 763.10 mm. and 21°.08 C. (p. 227), the
ht of the residual gas at 1.4 mm. and same temperature would be

weig

* These Proceedings, vol. xxiii. p. 1G8.
t Author's Chemical Physics, p- 881.

212

PROCEEDINGS OF THE AMERICAN ACADEMY

0.00076 gram or 0.76 m.g. No exact observation of the temperature
of the balloon was necessary ; for a difference of 4° C. would not alter
the above value more than T£o m.g., and we may therefore assume that
the temperature of the balloon was the same as that of the laboratory,

Fig. 2.

which at the time was 21° C. The correction for lessened buoyancy
(increased weight) on account of the contraction of the balloon by
atmospheric pressure amounts, as has been shown, to 1.98 m.g.,* and
this added to 0.76 makes the total correction 2.74 m.g. When the
balloon was hung on the balance, the weight required to complete the
tare was 2.5600 grams. Hence we have, —

Weight of balloon exhausted 2.5600 grams.

Correction for residual gas . . . 0.76 m.g.

" " lessened buoyancy . 1.98 " 27 "

Tare of empty balloon 2.5573 "

Two other determinations were made in the same way, and the
three results are united iu the following table.

* These Proceedings, vol. xxiii. p. 184.

OF ARTS AND SCIENCES. 213

Summary.

Tare of empty balloon, 1st value 2.5573 grams.

" " " 2d " 2.5572 "

« u 3d " 2.5574 "

Mean value, 2.5573 "

Take op the Empty Balloon by Chemical Method.

The general theory of this method has already been stated. The
balloon is filled with carbonic acid gas by displacement, and the tare
taken. The gas is then drawn through a series of absorption tubes,
as in the process of organic analysis, and the sum of the increased
weights of these tubes gives the weight of the total contents of the
balloon at the time the tare was taken. The weight of the balloon
filled with carbonic acid gas less the weight of the carbonic acid thus
determined is obviously the tare of the empty balloon required. In
attempting to perfect this apparently simple experimental method we
met with unexpected difficulties, arising from several circumstances.

In the first place, in order to wash out from the balloon the last
traces of carbonic acid, it was necessary to draw through the apparatus
a very large volume of air, and the system of purifiers and desiccators
needed absolutely to free the atmospheric air from the least admixture
of carbonic acid or aqueous vapor was found to be far more extensive
than we anticipated. In the second place, since the globe held, at
the ordinary temperature of the laboratory and at the average pres-
sure of the atmosphere, no less than nine grams of gas, the common
potash bulbs used in organic analysis were wholly inadequate for our
requirements, and we only succeeded after many trials in finding ;i
form of apparatus by which so great an amount of carbonic acid
could be determined with the necessary accuracy, that is, within a few
tenths of a milligram. The arrangement finally adopted is shown in
Plate III. The balloon is represented standing in its covered metallic
case, in which it is placed the moment it is taken from the balance.
On the right of the balloon and below it is a system of tubes tor
purifying and drying the atmospheric air which enters from outside
the laboratory by the flexible tube on the extreme right. On the
left of the globe is a system of tubes for absorbing the carbonic acid.
The flexible tube on the left is connected with a Bunsen pump through

214 PROCEEDINGS OF THE AMERICAN ACADEMY

a form of Mariotte's flask,* and by this means a current is maintained
through the whole system, under constant pressure.

Purifying and Drying Apparatus.

The air entering as we have described passed first through a wash-
bottle of familiar construction containing a solution of potassic hydrate,
Sp. Gr. = 1.44, and then through a similar wash-bottle containing a
concentrated solution of baric hydrate ; and if the rapidity of the cur-
rent was not greater than two bubbles a second, this double washing
was found sufficient to remove the last traces of carbonic acid. Indeed,
as soon as the baryta water showed the least indications of cloudiness,
the solutions were renewed. It was not so easy to remove the last
traces of moisture.

After the air had passed the washing flasks it entered a system of
desiccators, shown in the lower half of the plate, where in small bubbles
it travelled up first one tube, and then a second tube, both 5t feet long,
and filled with concentrated sulphuric acid previously boiled with a
small amount of ammonic sulphate to remove all nitrous fumes. Leav-
ing the second of these tubes it entered an elongated bulb four inches
long by two inches in diameter filled with phosphoric anhydride ; and
the fact that prolonged contact with sulphuric acid is not sufficient to
remove the last traces of moisture was shown by the circumstance that
after the current has passed for several days the dry white powder at
the opening of the bulb showed signs of deliquescence.

Our phosphoric anhydride was prepared by burning common yellow
phosphorus in a large sheet iron drum, through which a current of
dry air was drawn with sufficient rapidity to maintain the phosphorus
in rapid combustion, and the bottom of the drum was made tunnel-
shaped so that the anhydride could be shaken down into a self-sealing
fruit jar as fast as it formed. As thus prepared the anhydride has a
slight odor, which it loses after a prolonged current of dry air has
been drawn through the powder ; and after our drying tubes had
been filled their contents were submitted to a preliminary treatment of
this sort, and not until long after all perceptible odor had disappeared
were they used in our work. We noticed that after the anhydride
had been used for some time it appeared more granular and lost in
part its hygroscopic power. In order to make sure of this point, the
apparatus as far as described having been in use for several weeks, we
connected with the first phosphoric anhydride tube a weighed tube

* These Proceedings, vol. xxiii. p. 163.

OF ARTS AND SCIENCES. 215

containing fresh anhydride. After drawing air through the system
at the same rate and during the same time as in a carbonic acid
determination, we reweighed this second tube and found that it had
gained 5.5 m.g. We next added to our system the U tube filled with
phosphoric anhydride, which is shown in Plate III. immediately before
the balloon, and repeated the experiment. There was then no gain in
weight perceptible. The observations were as follows.

Before interposing the

U Tube.

■

ii

ipO

R

1st weight P205 tube 25.9212 grams

295.7

20.0

302.7

2d " " 25.9267 "

295.2

20.8

301.4

0.0055 " 1.4

After interposing the U Tube.

1st weight P2Os tube 25.9268 grams 299.3 19.6 306.7
2d " " 25.9268 " 299.2 21.6 304.6

2.1

In this table the column headed H contains the barometric pressure
at time of weighing in tenths of an inch ; that headed T° gives the
corresponding temperatures of the balance case in centigrade degrees.
By adding to the values H the quantities (27 — T)° we obtain the
values R, and these are the barometric heights which would give an
atmosphere of the same density, and therefore exerting the same buoy-
ancy, as the atmosphere at the time of weighing if the temperature
were uniformly 27° C. The correction that should be made for differ-
ences of temperature and pressure at successive weighings can at once
be calculated from these values. We found by experiment that for
the phosphoric anhydride tube used in these and in subsequent deter-
minations the correction amounted to 0.04 m.g. for each tenth of an
inch change of pressure, and hence in the above weighings the correc-
tion is too small to be appreciable. This simple method of correcting
weights for changes in the buoyancy of the atmosphere has been de-
scribed in a previous paper in these Proceedings.* In order to apply
it readily, the barometric heights, although usually read in millimeters,
were often also noted, as here, in tenths of inches, and the instrument
we used was provided with both scales.

* Vol. xviii p. 55.

216 PROCEEDINGS OF THE AMERICAN ACADEMY

It is obvious from the above experiments that the hygroscopic
power of phosphoric anhydride becomes impaired before the powder
actually deliquesces, and we have dwelt on the point because the ex-
perience was of great importance in our investigation by leading us
to add the second anhydride tube to our system, and also to watch
carefully the condition of its contents and to replace the tube as soon
as the least change was perceptible. We attribute the change to a
glazing of the amorphous grains due to the forming of a coating of
glacial phosphoric acid, but the effect is one that would not be noticed
except under such extreme conditions as existed in our work. We
now pass to the system of absorption tubes on the other side of the
balloon.

Absorption Apparatus.

The apparatus for the absorption of the carbonic acid gas must meet
very opposite requisitions. In the first place, it must offer a sufficiently
large mass of absorbent to the pure gas which first comes to prevent
overheating from the rapid chemical combination. In the next place,
it must be capable of removing the last traces of carbonic acid from
the air which is afterwards drawn through the tubes. Lastly, it must
not be so heavy or so complicated that it cannot be promptly and
sharply weighed, and the correction for changes of temperature and
pressure between successive weighings accurately estimated.

After trying in various ways to enlarge and modify the different
forms of potash bulbs and spiral tubes in general use, (and in these
attempts we had the aid of a skilful glass-blower,) we found that we
could reach a more accurate result with a system of separate tubes,
in spite of the number of weighings they involved. To absorb the
great mass of the carbonic acid we used the simplest form of gas
washer, shown in Plate III. at A. The bulb had a capacity of
200 c.c., and in this we placed 75 c.c. of a solution of potassic
hydrate having Sp. Gr. = 1.443. When empty the bulb weighed
only 57 grams, and when filled about 165 grams. Assuming that
the potash was fully neutralized, less than one half of this amount
would be sufficient to combine with nine grams of carbonic acid ; but
to insure prompt action and avoid excessive heating the amount
given is required.

The potash tube was followed by a small U tube, four inches long,
with ground caps filled with finely granulated soda lime, and weighing
when full about 40 grams. (See B, Plate III.) To. this was united a
similar U tube containing phosphoric anhydride to hold the aqueous
vapor, which is brought over by the gas current in no inconsiderable

OF ARTS AND SCIENCES. 217

amount from the potash solution and the soda lime, in consequence of
the heat set free by the absorption. The anhydride tube weighed
when full about 25 grams, and is shown at C. Next was a third l'
tube filled with glass beads and having a few drops of strong sulphuric
acid at the bend. This tube, D, was not weighed, but was necessary
in order to indicate any sucking back of the air and to prevent any
vapor from thus reaching the phosphoric anhydride. The double bulb
forming the end of the system, shown at E, was made by melting
together the usual inlet tubes of two gas washers like A. In this we
placed 50 c.c. of a standard solution of baric hydrate, and the peculiar
construction was necessary in order to provide against regurgitation
in consequence of back pressure. The solution is easily washed out
of the bulbs, and wdien retitrated at the end of the process gives the
amount of carbonic acid, if any, which escaped the other absorbents.
All of these tubes were united into one system by rubber connections
carefully lubricated both inside and out with vaseline, a soft parafline
which had previously been melted and heated to remove every trace
of moisture.

The manipulation of this apparatus was an interesting experiment.
When all had been set up as shown in Plate III., the globe being
full of carbonic acid, the Bunsen pump was first started, and after a con-
stant suction (regulated by the Mariotte's flask before described) had
been established, we cautiously opened the cock of the balloon, which
connected with the potash bulb, closely watching the flow of gas. As
soon as this flow began to slacken, we as cautiously opened the farther
cock, by which air was admitted to the balloon through the line of
purifiers, still closely watching and regulating the flow. We then
noticed a striking phenomenon. As pure carbonic acid was now
passing into the potash bulb, the air mixing but slowly with the
heavy gas at the surface of contact as the level sunk in the balloon,
the absorption at first was complete. If the air supply from behind
was not sufficient, a partial vacuum formed in A, and air was drawn
back through the tubes in front, as indicated at once by the bubbles
at the bend of D; but if the supply was carefully regulated, a perfect
equilibrium could be maintained in the tubes beyond, and we have
seen the carbonic acid flow into the potash bull) fur half an hour
without more than a few bubbles of air passing through the sulphuric
acid at the bend of I) in either direction.

After a while, however, when the great mass of the carbonic a. -id
has been absorbed, bubbles of air begin to pass through D, and also
bubble through the baryta water beyond, and soon a steady current

218 PROCEEDINGS OP THE AMERICAN ACADEMY

is established throughout the whole apparatus, which may then be left
to take care of itself. If the current does not exceed in rapidity
two bubbles a second, it may be allowed to flow for twenty-four hours
without risk, and in that time every trace of carbonic acid will have
been washed out of the 'balloon. The suction should now be stopped,
and the cocks of the balloon closed. The tubes A, B, and C may
then be disconnected, and after they have come to equilibrium in the
balance case reweighed. The double bulb E must also be discon-
nected, the baryta washed out into a suitable bottle, and the solution
retitrated with oxalic acid.

As the weights of the tubes A, B, and C must be corrected for
the buoyancy of the air, it is essential to note the temperature of the
balance case and the height of the barometer at the time of each
weighing. The corrections in question may be considered under
two heads ; first, that made necessary by a change of temperature or
pressure between successive weighings ; secondly, that required to
eliminate the effect of atmospheric buoyancy on the total weight of
carbonic acid absorbed.

The first correction is best made by the method already described,
and the application of this method was greatly facilitated in the
present investigation by the circumstance that the same tubes, wiih
the "same weights of contents, were used in all the carbonic acid
determinations made. We were able to determine once for all for
each tube the correction required for a variation of one tenth of an
inch in the barometric height reduced to 27° C.,* and then the cor-
rection in any case could be found by a simple multiplication. Thus
we estimated that the correction for one tenth of an inch in the
barometric height reduced to 27° C. was, —

For the potash bulb before absorption, . . . 0.306 milligram.

« " after " ... 0.321 "

" soda lime tube 0.067 "

" phosphoric anhydride tube .... 0.040 "

* That is, reduced to the height which would be required to give at the
temperature of 27°, an atmosphere of the same density and buoyancy as that
at the temperature of the balance case when the tube was weighed. These
Proceedings, vol. xix. p. 55.

OP ARTS AND SCIENCES. 219

Calculation of Correction for Potash Bulb.

Before After

Absorption. Absorption.

"Weight of glass 57.45 grams. 57.45 grams.

Volume of glass = 57.45 -r- 2.5 . . 23 c. c. 23 c. c.
« KOH solution .... 75 " 80

98 103

" brass weights, 164 -f- 8.2 20 "
" '• " 173 ~ 8.2 21 «

Uncompensated volume 78 " 82 "

Air displaced at 27° C.
and 300 tenths of an inch.

1.176 x 78 91.928 m.g.

1.176x82 96.332 m.g.

Variation for ^ 0.306 " 0.321 "

The correction of the total weight of carbonic acid for the buoyancy
of the atmosphere was not so simple a problem ; but by using the
same bulb and nearly the same volume of the same solution of potassic
hydrate in every case, it was sensibly the same for all the determina-
tions. The specific gravity of the solution of potassic hydrate before
the absorption was 1.443, and tested in two separate determinations
after absorption it was found to be 1.464 in each case. Using now
the arithmetical means of the weights of the potash bulb in the three
determinations under consideration, it was easy to find an average
value for the correction thus.

Average weight of potash bulb before absorption 165.02 gr.

Weight of glass 57.45 "

solution of KOII 107.57 "

Volume of solution of KOH = 107.57 -h 1.443 . 74.55 c. c.

Average weight of potash bulb after absorption . 173.57 gr.

Weight of glass 57.45 u

116.12 -

Volume of solution of KOIT = 1 16.12 -4- 1.464 . 79.32 c. <■.

Difference of volume = 79.32 -- 74.55 .... 4.77 -
Weight of C02 absorbed in potash bulb = 173.57 —

165.02 8.55 gr.

220 PROCEEDINGS OP THE AMERICAN ACADEMY

Volume of brass weights = 8.55 4- 8.2 1.04 c. c.

Uncompensated volume = 4.77 — 1.04 3.73 "

Weight of 3.73 c. c. of air at 300 tenths inch and

27° C. = 1.176 X 3.73 = corr. required . . 4.39 m.g.

In this calculation we have necessarily left out of the account both
the small amount of carbonic acid absorbed by the soda lime and the
water carried by the current from the potash bulb to the connecting
tubes ; for the increased weight of these tubes results from both these
causes, whose relative effects or whose influence on the correction we
canuot estimate! That they wuuld tend to raise slightly the value
of the correction is obvious ; and if we assume, as we have in the
following calculations, that the value is 4.5 milligrams we shall ap-
proximate as nearly to the precise amount as circumstances will per-
mit. For all these corrections the data are only known within certain
narrow limits ; but the total uncertainty thus arising does not exceed
two or three tenths of a milligram in the final result.

In weighing the potash bulb and U tubes it is of course essential
that they should be left in the balance case until a perfect equilibrium
of temperature has been reached, and then they must be weighed
when open so that a perfect equilibrium of pressure on the inside and
outside of the glass may be secured. And since we were here aiming
at very great accuracy, we felt it important to inquire further whether
any gain of weight from hygroscopic moisture was possible during
the time necessary to verify the weight while the tubes remained
open on the scale pan. We therefore carefully investigated this
point. The air in the balance case was kept as dry as possible by
an open dish of sulphuric acid, aud the experiments were made under
such conditions. It was found that the weight both of the potash
bulb and of the soda lime tube remained absolutely constant, hour
and hour together, except so far as they were influenced by changes
of temperature and pressure, but the weight of the phosphoric anhy-
dride tube slowly but uniformly increased at the rate of 0.025 m.g.
per hour. We followed the change on one occasion from I0h- 45m- a. m.
to 6h< 15m- p. m., and during these seven hours and a half the gain was
19 m.g., while the changes of temperature and pressure meanwhile
were not enough to cause a sensible difference in the atmospheric
buoyancy. Such experiments were continued for a sufficient length
of time to enable us to confirm the values of the corrections above
given, and they gave us great confidence in our weights. They
showed that the only precaution necessary for accuracy was to close

OF ARTS AND SCIENCES.

221

the tubes as soon as they were disconnected, and to leave each of them
closed in the balance case until it was time to take the weight. The
U-shaped tubes with perforated ground stoppers, by which connection
with the lateral tubes can be instantly opened or closed, are admirably
adapted to this work. As an example, we give next the details of
one determination of carbonic acid by the method we have described.

Details of a Determination of Carbonic Acid.

The gas was generated from statuary marble and dilute hydro-
chloric acid in an apparatus precisely similar to that described and
figured by us in a previous paper.* It was first washed by a solution
of potassic bicarbonate, and then, having passed in succession over

Fig. 3.

calcic chloride, sulphuric acid, and phosphoric anhydride, il was con-
ducted into the balloon and overflow in tlie manner already described.
The gas was tested by leading the current from the overflow into a
familiar form of nitrogen apparatus filled with a solution of p<>t

* These Proceedings, vol. xxiii. p. 160.

222 PROCEEDINGS OP THE AMERICAN ACADEMY

hydrate (Fig. 3), and not until no measurable residue was left by
a full liter of gas was its condition regarded as satisfactory. The
balloon stood in the calorimeter case, and when the cocks were closed
with the precautions already described the temperature of the case
and the height of the barometer were observed, and although these
values are not needed for determining the tare of the empty balloon,
they will be used for determining the density of carbonic acid here-
after. The balloon was then disconnected and hung on the balance,
and its tare found to be

W = 11.5907 grams.

The balloon was then transferred to its metal case, and, having been
mounted ;ts shown in Plate III., the determination of the weight of
its contents proceeded as just described.

Potash Bulb.

Grams.

H

TO

1st weight of potash bulb 165.3388

299.99

25.8

2d " " " 174.0132

299.83

27.0

8.6744

Correction* for change of H and T 4

8.6740

* Since the volume and density of the potash solution used were closely the
same in all the determinations, whether compared before or after the absorp-
tion, the value of the correction may be assumed to be the same in all cases,
and, as has been shown at page 218, amounts to 0.306 m.g. before absorption
and 0.321 m.g. after absorption for every change of one tenth of an inch pres-
sure or 1° centigrade temperature from the standard of 300 tenths inch and
27° C. or 300° absolute temperature. The correct mode of calculating the
correction in any case is to reduce both weights to what they would appear to
be at II = 300 and T0° = 300 = 27° C. before taking their difference, using
the value 0.30G for the smaller of the two, and 0.321 for the larger.

290.89 at 25° .8 corresponds in buoyancy to 301.09 at 27°
299.83 at 27° " " " 299.83 "

301.09—300 = 1.09 and 1.09 X 0.306 = 0.33 addative.
299.83 — 300 = —0.17 and —0.17 X 0.321 = —0.05 subtractive.

Then 165.3388 + 0.33 = 165.33913
174.0132 — 0.05 = 174 01315

8.67402 as before.

As regards the sign of the correction, notice that, when reduced to H = 300,
the buoyancy becomes less and the apparent weight greater if the previous
pressure was above 300, and vice versa.

Grams.

H

T°

42.8731

299.54

26.6

43.1857

299.87

27.4

OF ARTS AND SCIENCES. 223

Soda Lime Tube,
1st weight of soda lime tube

2(1 '< " it U

Correction insensible 0.3126

Phosphoric Anhydride Tube.

1st weight of anhydride tube 39.5196 299.50 27.0

2d " " " « 39.5493 300.00 27.2

Correction insensible 0.0297

Standard Solution of Baric Hydrate.

Before absorption, 50 c. c. required 46.60 c. c. solution of oxalic acid.
After " " " 43.65 « « "

2.95

One c. c. of decinormal solution of oxalic acid contains 0.0126 cram
of H2C204 . 2 H20, corresponding to 0.0044 gram of C02.

2.95 X 0.0044 = 0.0130 gram of CO,.

Total Weight of Carbonic Acid.

Grams.

From potash bulb 8.67 10

" soda lime tube 0.3126

" anhydride tube 0.0297

" titration 0.0130

9.0293
Reduction to vacuum 45

9.0338
Tare of balloon and gas 11.5907

Tare of empty balloon 2.5569

Value by Regnault's method 2.')-'n'-'>

The value of the tare of the empty balloon found by the new chemi-
cal method is directly comparable with that obtained by the Regnaull
method, since there was no change in the balloon meanwhile. But
after this determination was finished, the balloon, as it remained stand-
ing in its case, having been exposed to the direct sunshine of an un-
usually hot June day, some of the cerate lubricating the stop-cocks
melted and ran down into the tubulatures below, from which ii bad t"
be removed. Thus the tare was lessened by nearly a centigram, fl

224 PROCEEDINGS OF THE AMERICAN ACADEMY

was easy, however, to determine the exact loss, and as in a like inves-
tigation, even when using the greatest care, one must ever be liable to
accidents of this sort, it is important to dwell on this point.

In losing the tare we had apparently lost the thread of our inves-
tigation, but we were readily able to recover this thread without inter-
rupting the general course of the work. It must be borne in mind
that our immediate object was to obtain by the chemical method a
series of results for the tare of the empty globe which could be com-
pared with the mean value 2.5573 obtained by the method of Regnault.
Here had come a change of tare, and we could obviously find a new
standard of comparison by redetermining the new value by the old
method. But this was not necessary. In the first determination by
the chemical method, we had observed with great precision the pres-
sure and temperature at which the balloon was filled. We do the
same in the second determination, to which we next proceed, although
these values are not needed in this determination, by itself considered,
any more than they were in the first. If now from the tare of bal-
loon and gas in the first determination (11.5907) we deduct 2.5573,
we have the weight of carbonic acid which the balloon contains at a
known temperature and pressure as found by the method of Regnault.
If we deduct the same quantity from the tare found in the second de-
termination, we have the weight of gas which the balloon holds at
another temperature and pressure by the same method. If now there
has been no change of tare, these two weights ought to agree exactly
when one is reduced to what it would have been under the conditions
at which the other was taken ; and if there has been a change of tare,
the difference of the weights thus reduced will give the exact amount
of the loss, on the basis solely, let it be noticed, of results obtained
after the method of Regnault. Thus we have, —

Grams.

By No. 1, weight of C02 at 763.85 mm. and 25°.08 9.0273

" No. 2, " " 756.73 " " 22°.40 9.0334

Weight No. 1 reduced at " " " " 9.0241

Loss of tare 9.3 m. g.

Obviously the accuracy of this result depends upon the accuracy
with which the observations of temperature and pressure were made,
and upon the correctness of the data on which the reductions are based.
That the accuracy is extreme will appear from a comparison of the
values of the specific gravity of carbonic acid, given on page 229, which
were deduced from the same elements. And although in such work
as this there is an obvious liability to error from just such a change of

OF ARTS AND SCIENCES. 225

tare as we have been discussing, yet the loss or gain thus resulting is
usually of such a magnitude that it cannot be overlooked.

Besides the one of which the details have been given, we made by
the chemical method two other determinations of the tare of the empty
globe in precisely the same manner as before described, and obtained
the values 2.5481 and 2.5475 respectively. In order to bring these
results into comparison with that obtained by the first determination,
or those obtained by the method of Reguault, we must add 9.3 m.g.
to each value. We then have for

Tare of Empty Balloon.

Regnault's Method. Chemical Method.

No. 1 2.5573 2.5569

No. 2 2.5572 2.5574

No. 3 2.5574 2.55G8

Average 2.55730 2.55703

o

These results are essentially identical, and the two methods confirm
each other. But having reached this important conclusion, which was
the main object of our investigation, we freely concede the preference
to Regnault's method. The difference between the results of the
two methods, as here exhibited, is insignificant ; but so long as the
validity and value of the correction for the contraction of the balloon
on exhaustion has been thus established, Regnault's method will be
preferred because it is practically more simple, and hence the results
are more accordant than those obtained by the chemical method first
described in this paper. •

During the course of our investigation there were two marked
changes in the tare of our balloon. One of these has been de-
scribed. The other arose from the circumstance that the thin paraf-
fine (vaseline) which was most suitable for lubricating the stop-cocks
during the winter no longer kept the joints tight when the heat of
approaching summer rendered the material more liquid ; and it became
necessary to remove the stoppers and relubricate them, and it was the
excess of cerate then used which melted and ran into the tubulaturea
subsequently.

There are then three separate values of the tare of the empty bal-
loon to be used in our further calculations, all determined by or de-
pendent upon Regnault's method, and all known with equal accuracy.
These values we shall distinguish by the letters A, B, and C, and we
shall designate by the same letters the experimental data or calculated
results into which these values enter as elements.
VOL. xxiv. (n. s. xvi.) 15

226 PROCEEDINGS OP THE AMERICAN ACADEMY

Tare of Empty Balloon.

Values.

A 2.5098 grams.

B 2.5573 "

C 2.5480 "

Density of Air freed from Moisture and Carbonic Acid.

In these determinations the balloon was placed in the calorimeter
case and connected with the Bunsen pump through an overflow jar.
Air was then drawn into it through the purifiers we have described,
and the reflux current which followed for a moment when the inlet
cock was closed was supplied from the previous overflow. In other
respects the determinations were made as before described, and we
obtained the following data for the weight of air held in the balloon
at different temperatures and pressures.

Summary of Air Weights.

A No. 1, weight of air at 762.63 mm. and 18°.52 = 6.0347 grams.
A No. 2, " " 758.71 " " 19°.96 = 5.9745 "

B No. 1, " " 750.30 " " 23°.32 = 5.8411 "

Reducing now these weights to what they would have been at the
temperatures and pressures we have assumed as standards, we obtain
the following comparable values.

, Reduced Air Weights for

H == 761.99 mm. = 300 d. in* and

T = 27° C. = 300° absolute temperature.

A No. 1 5.8584 grams.

A No. 2 5.8588 "

B No. 1 5.8586 "

Average value 5.8586 grams, t

* We shall use the abbreviation d. in. for tenths of an inch, after the analogy
of d. m. for tenths of a meter.

t In the calculations of density and specific gravity, a still further, although
wholly unimportant, correction has been made for the buoyancy of the air on
the brass weights used to complete the tare in weighing the balloon, and in this
connection it must be borne in mind that the platinum or aluminum fractions of a
gram in a set of weights have always been adjusted to the larger brass weights
in the air, and are therefore the equivalent of brass weights of the same denomi-

OP ARTS AND SCIENCES. JlIT

Density of Hydrogen.

We have for the weight of hydrogen held hy the balloon the
following data.

B No. 1, weight of hydrogen at 763.10 mm. and 21°.08 = 0.41 G4 gr.
B No. 2, " " 764.92 " » 20°.52 = 0.4178 "

B No. 3, " " 758.96 « « 22°. 75 = 0.4123 "

nations. The volume of the counterpoise was adjusted when the balloon was
full of air; but although the compensation may be perfect through all ordinary
changes of temperature and pressure, it cannot be regarded as absolute. In
a laboratory, such changes seldom amount in the aggregate to more than the
equivalent of one twentieth of the normal atmospheric pressure ; and although
such a variation might not produce a sensible effect on the apparent weight of
the small amount of brass used in our weighings, twenty times this effect, which
we should have if the brass weights were wholly uncompensated, might be an
appreciable quantity. In weighing the balloon the largest amount of brass
weights used was about 11.5 grams, which at 300 d. in. and 27° C. displace
1.6 m.g. of air. One twentieth of this value would be barely, if at all, percep-
tible, but the whole quantity might cause a serious error in our results.

The tare of the empty balloon as first found by Regnault's method was
2.5573 grams ; and we must assume that in the system consisting of balloon
counterpoise and weights there may be outstanding a very small uncompen-
sated volume, so that the true tare would be represented by

WR = 2.5573 ± w.

We next find the weight of the balloon filled with hydrogen gas, and it is ob-
vious that under standard conditions the true weight, or

W = 2.5573 ± w + 0.4076 - w',

where w' represents the buoyancy of air on 0.4076 gram of brass. So the true
weight of the balloon filled with air must he

W" = 2.5573 ± w + 5.8594 - w",

where w" represents the buoyancy of air on 5 8594 grams of brass. In like
manner, the true weight of the balloon filled with carbonic acid gas is

W" = 2.5573 ± w + 8.9564 - w'"f

where w'" is the buoyancy of air on 8.9564 grams of brass.

Taking 1.176 m.g. as the weight of 1 c.c. of air at 300 d. in. and 27° C, the
standard conditions to which the weights have been reduced, we have

w' = 0.05 m.g., w" = 0.84 m.g., w'" = 1.28 m. g. ;

and further,

W — WR = 0 4076 - 0 00005 gram. True weight of hydrogen.

W" - Wi: = 5.8594 - 0.00084 " *' " air.

W'"-WK = 8.9564 -0.00128 " " " carbonic acid.

These corrections should be applied to the weights before calculating the
specific gravities, although they only alter the last values one or two unit- in

228 PROCEEDINGS OF THE AMERICAN ACADEMY

Reduced Weights.

H = 300 d.

in.

T = 27° C,

B No. 1

0.4076

B No. 2

0.4072

BNo. 3

0.4081

•age

0.4076

Specific Gravity of Hydrogen.

Air = 1.

The values were obtained by dividing each of the weights of hydro-
gen above given by the average weight of air (5.8586), after reduction
for the temperatures and pressures at which the hydrogen weights
were observed. The calculations were verified by dividing each of
the reduced weights of hydrogen by the average reduced weight of
air as above, and the same figures were obtained in both ways.

B No. 1 0.06957

B No. 2 0.06951

B No. 3 0.06966

Average 0.06958

Average using values of tare by chemical method 0.06962

Specific Gravity of Oxygen.

Hydrogen = 1.

In this investigation we have not made an additional determination
of the specific gravity of oxygen gas referred to air, for the causes of
error which are serious in the case of hydrogen are not appreciable
when working with the heavier gas, and we cannot expect to improvu

the last decimal place ; and in this paper, as is usual, the decimals are carried
out one figure beyond tlie limits of accuracy.

In the chemical method we also took the weight of the balloon filled with
carbonic acid gas, and, as before, we have

W" = 2.5573 ± w + 8.9564 - w'".
Then by chemical means we determined the absolute weight (tfi vacuo) of
the carbonic acid which the vessel held, that is, the quantity represented by
8 9564 — w"'. Subtracting this value from the total weight, we obtained the
tare of the empty balloon by the chemical method, and this obviously is

Wc = 2.5573 ± w.

Thus, the buoyancy of the air on the weights is eliminated, and "Wc is directly
comparable in all its relations with WR.

OF ARTS AND SCIENCES.

220

on the results of Regnault. Assuming, then, for the specific gravity
of oxygen gas referred to air the value 1.10562, as obtained by Re«-
nault and corrected by Crafts, and dividing this value by each of the
values for the specific gravity of hydrogen gas given above, we have
the following values for the specific gravity of oxygen referred to
hydrogen.

BNo.l
B No. 2
BiNo.3
Average

Average of chemical method

Value found by Lord Rayleigh

15.892
15.907
15.873
15.891

15.882

15.884

Density of Carbonic Acid.

For the weight of carbonic acid gas held in the balloon at different
temperatures and pressures we have the following data.

B No. 1, weight of C02 at 756.73 mm. and 22°.40 9.0334 gr.
C No. 1, " « 760.50 " " 25°.56 8.9821 «

C No. 2, " « 759.13 " " 26°.60 8.9348 "

Taking now the average value of the weight of air held in the bal-
loon at 300 d. in. = 761.99 mm. and 27° (viz. 5.8586 grams), and re-
ducing this weight to what it would have been under the condition at
which each of the four weights just given was observed, and then divid-
ing each of these weights (corrected as by previous note) by the
weight of air under the same conditions, we obtain the following
results.

Specific Gravity of Carbonic Acid. Air = 1.

Wt. of CO,.

Wt. of Air.

n

T

Sp. Gr.

B No. 1

9.0321

5.9088

756.73

22°.40

1.52858

CNo. 1

8.9808

5.8753

7C0.50

25°.56

1.52855

CNo. 2

8.9335

5.8445

759.13

26°.60

1.52855

Average value

L.52856

Aver

age value b\

' chemical r

irocess

1..-.2854

Taking lastly the average value of the weight of hydrogen held in
the balloon at 761.99 mm. and 27° (viz. 0.4076 gram), and dividing
the several weights of carbonic acid by this weight reduced to the
same conditions, we have, —

230 PROCEEDINGS OF THE AMERICAN ACADEMY

Specific Gravity of Carbonic Acid. Hydrogen = 1.

Wt. of C02. Wt. of H2. H T Sp. Gr.

B No. 1 9.0321 0.4111 756.73 22°.40 21.971
C No. 1 8.9808 0.4088 760.50 25°.56 21.971
C No. 2 8.9335 0.4066 759.13 26°.60 21.971

Average value 21.971

Average value by chemical process 21.957

The close agreement, indeed the essential identity, of these results
should be noticed, as this fully substantiates the extreme accuracy of
the method of recovering the lost tare of the empty balloon described
on page 224. It is true that, as the values of the specific gravity of
carbonic acid have been calculated with an average value of the sev-
eral reduced weights of hydrogen or air, an error in these data would
not affect the results ; but any error in the observed weights of car-
bonic acid would appear to its full extent.

Moreover, the striking fact should not be overlooked that the specific
gravity of carbonic acid gas referred to hydrogen approaches much
more nearly 22, the half molecular weight of carbonic acid gas as
generally assumed, than does that of oxygen gas when referred to the
same standard, the corresponding whole number 16, and the signifi-
cance of the circumstance is obvious. If we assume that the atomic
weight of oxygen as determined by Dr. Richards under my direc-
tion is 15.87,* then the corresponding atomic weight of carbon would
be 11.90, and the half molecular weight of carbonic acid 21.82.
Theoretically, this number ought also to define the specific gravity of
carbonic acid gas referred to hydrogen gas, if the two gases were com-
pared under the same conditions of high temperature and indefinite
expansion. But under the great pressure of our atmosphere the
molecular volume of carbonic acid gas is known to be condensed to
a measurably greater degree than the potentially equal molecular vol-
ume of hydrogen gas, and the result must be a proportionally increased
density ; and, moreover, the inequality in the condensation of the two
gases must be increased by the circumstance that at the ordinary tem-
perature of the air carbonic acid gas is below the critical point, while
hydrogen gas is very far above it. If, however, this is true in the case
of carbonic acid gas, our knowledge of the deviations from Mariotte's
law compels us to infer that the same must be true in some small
measure, although in a much less degree, in the case of oxygen gas ;

* These Proceedings, vol. xxiii. p. 185.

OF ARTS AND SCIENCES. 231

and it would lead us to expect that the specific gravity of oxygen gas
referred to hydrogen gas would be slightly greater, certainly not I
than the corresponding half molecular weight. Now we have from
the results published in this and our preceding paper, —

Half molecular weight of carbonic acid gas 21.82

Sp. Gr. of carbonic acid gas (referred to hydrogen gas) . 21.96

Half molecular or atomic weight of oxygen gas . . . 15.87
Sp. Gr. of oxygen gas (referred to hydrogen gas) . . . 15.88

It has been thought to discredit the low value of the atomic weight
of oxygen we have found by the very easy but wholly gratuitous as-
sumption that the hydrogen gas on which we experimented was im-
pure, and we had intended in this investigation to demonstrate the
identity of the gas from the several sources employed by a comparison
of their specific gravities. We are under great obligations to Lord
Kayleigh, who has relieved us from this necessity by anticipating the
work. He has experimented on hydrogen gas, not only from all the
sources we used, but also on gas which had been occluded by palla-
dium, and has obtained the same result in all cases ; namely, the value
1-3.S84,* essentially identical with that which we have found in this
investigation with the hydrogen gas from our electrolytic apparatus.!

This point, however, had not been overlooked in the previous paper.
It was there shown that the results obtained, in three distinct series of
experiments, with hydrogen gas prepared by three different processes
and with three different forms of apparatus, were essentially identical, t
On the doctrine of chances, such an agreement would have been prac-

*

tically impossible had there been an appreciable amount of accidental
impurities, in the gas from either of the sources. We say accidental
impurities, for there may be inherent impurities common to the
from all sources, of which we as yet know nothing; and, as we wrote
before, " The question still remains, Is the hydrogen gas thus prepared
the typical hydrogen element? But this is the same question which
must arise in regard to any one of the elementary substances ; and :ill
that we can say is, that the evidence in regard to the purity of the
hydrogen we have used is as good as any that can be adduced in regard
to any one of the elementary substances whose atomic weight has been
most accurately determined." §

* Proceedings of the Royal Society, vol. xlv. p. 426.
t These Proceedings, vol. xxiii. p. 108.
t See table, these Proceedings, vol. xxiii. p. 173.
§ These Proceedings, vol. xxiii. p. 171

232 PROCEEDINGS OF THE AMERICAN ACADEMY

It is not to be expected that our results, or those of our contempo-
raries, are final. The most that has beeu accomplished by recent in-
vestigation is to show that the ratio of the atomic weight of oxygen to
that of hydrogen, deduced from the elementary substances, as we know
them in their purest condition, is decidedly less than that of 10 to 1.
The evidence as to the exact value of this ratio is still conflicting, and
although after our experience we cannot see how greater accuracy
could be gained by any variation of our process, we are far from claim-
ing that our results have not been vitiated by unknown constant
errors. Fortunately, an exact knowledge of the ratio is at present of
no practical importance in chemical analysis. The only question on
which the actual small indefiniteness has a bearing is the unit of atomic
weights, a question that has been much discussed of late ; but here, as
it seems to us. one consideration should be conclusive.

Every one who has worked on the oxygen and hydrogen ratio
knows that the resources of experimental science have been taxed to
the utmost in these investigations. Greater accuracy is not to be i x-
pected with our present appliances, and yet there is an outstanding
uncertainty amounting to more than one half of one per cent of the
value. This corresponds to a variation of more than one unit in many
of the higher atomic weights ; and unless we are willing that these
chemical constants should fluctuate to this extent with every new de-
termination of the fundamental ratio, we must seek a more invariable
basis. The most natural and stable unit is the atomic weio-ht of
oxygen, not only on account of the wonderful combining power of this
element, but also because the ratio of the combining weights of most
of the elements to that of oxygen are known with great precision ;
and many cogent reasons could be urged for returning to the system
of Berzelius, which referred all the other weights to that of oxvsren,
assumed to be 100. But this system would not exhibit to advantage
the numerical ratios ou which modern chemical classification is based.
Hence for a provisional system we most warmly approve of that which
assumes 0 = 16 as its basis, and of the best known atomic weights
leaves only the value of H to vary with our changing knowledge.
This system has all the stability it is possible at present to secure, and
exhibits to advantage the relations which are important in classifica-
tion. Moreover, it is no small recommendation to this system that a
large number of the weights are whole numbers, within the limits of
error of ordinary analytical work, and for this reason can be easily
remembered.

In conclusion, I would express my obligations to my nephew, Dr.

PLATE I.

PLATE II.

OP ARTS AND SCIENCES. 233

Oliver W. Huntington, for the great aid he has given throughout this
investigation. With the help of his skill in glass-blowing I have been
able to secure all the connections of my gas generators with melted or
cement joints, except the direct connections with the globe, where india-
rubber could not be avoided. He has also constantly assisted my im-
paired eyesight in the observations with barometer and thermometer
on which the accuracy of the work greatly depended. I have likewise
been indebted to Dr. Arthur M. Comey for making the titrations
required, and to Mr. W. W. Bosworth for the drawings with which
this paper is illustrated.

234 PROCEEDINGS OF THE AMERICAN ACADEMY

XIX.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF

HARVARD COLLEGE.

ON THE ACTION OF SODIUM MALONIC ESTER ON
TRIBROMDINITROBENZOL.

SECOND PAPER.

By C. Loring Jackson and W. S. Robinson.

Presented May 28, 1889.

In a previous paper * on this subject we were obliged to publish sev-
eral results in an unfinished condition, because at the time we saw no
chance of being able to continue the work together. Since then, how-
ever, unforeseen circumstances have allowed us to go on with the re-
search, and in this paper we describe work which fills up most of the
gaps in our former publication.

The most important results contained in this paper may be stated
briefly as follows. The reactions by which the bromdinitrophenyl-
malonic ester is formed have been made out to be the following : —

I. C6HBr3(N02)2 + CHNa(COOC2H5)2 =

NaBr + CGHBr2(N02)2CH(COO"C2H5)2.

II. C,HBr2(N02)2CH(COOC2H5)2 + CHNa(COOC2H6)2 =

C6HBr2(N02)2CNa(COOC2H5)2 + CH2(COOC2Hs)2.

III. C6HBr2(N02)2CNa(COOC2H5)2 + CH2(COOC2H.)2 =

C6H2Br(NO,)2CNa(COOC2H5)2 + CHBr(COOC2H6)2.

IV. CHNa(COOC2H.)2 + H20 = NaOH + CH2(COOC2H,)2.

V. CHBr(COOC2H5)2 + NaOH = NaBr + CHOH(COOC2H.)2.

The proof of these reactions consisted in the isolation of tartronic acid
from the mixture obtained by the saponification of the oily secondary
product with hydrochloric acid.

The action of concentrated hydrochloric acid, or, better, sulphuric
acid of specific gravity 1.44, upon the bromdinitrophenylmalonic ester

* These Proceedings, xxiv. 1.

OP ARTS AND SCIENCES.

»

has been studied, and we are able to correct some of the preliminary
statements made on this subject in our first paper. The product is
the bromdiuitrophenylacetic acid

C6H2Br (N02) 2CH2COOH,

which melts at 177°, and is formed, when hydrochloric acid is used, by
the following reactions : —

C6H2Br(N02)2CH(COOC8H5)2 + 2 HC1 =

2 C.,H.C1 + C6H2Br(NOs)2CH(COOH)2 =
2 C2H.C1 + C02 + C6H2Br(N02)2CH2COOH.

The crystalline silver salt of this acid was analyzed, and proved to
have the formula

C6H2Br(N02)2CH2COOAg. H20.

With sodic hydrate a deep Prussian green or with an excess a yellow-
ish brown solution is formed, from which acids throw down a pale
crimson precipitate, or a white one which becomes pale crimson when
moistened with alcohol ; none of these substances could be brought
into a state fit for analysis, but we found that bromine was removed
in their formation.

The bromdiuitrophenylacetic acid is broken up when boiled with
alcohol into bromdiuitrotoluol and carbonic dioxide ; boiling with
water produces the same effect, but much more slowly. A few drops
of sulphuric acid prevent the decomposition by boiling water. This
conversion into bromdiuitrotoluol gave us the means of determining
the constitution of the bromdinitrophenylmalonic ester and its deriva-
tives; as Messrs. W. B. Bentley and \V. II. Warren have, at our re-
quest, worked out the constitution of the substituted toluol, and found
that it is as follows :

CH8.Br.N02.N02.1.3. 4.G.

It follows, therefore, that all the substances mentioned in this and our
former paper must have a similar constitution, and thai in making the
bromdinitrophenylmalonic ester from bromdinitrobenzol the bromine
atom replaced by the malonic radical is one of those which is at the
same time para and ortho to the nitro groups, the one replaced by hy-
drogen is ortho to the two nitro groups, while the third atom of bro-
mine, which remains unaltered, occupies exactly the same position as
that replaced by the malonic ester radical, — certainly a curious result.

286 PROCEEDINGS OP THE AMERICAN ACADEMY

Preparation of Bromdinitrophenylmalonic Ester.

Our longer experience in the preparation of bromdinitrophenyl-
malonic ester has led us to introduce several improvements into the
process, which now we carry on as follows. A strong benzol solu-
tion of 20 grs. of tribromdinitrobenzol (benzol is to be preferred to
the ether used formerly, because the tribromdinitrobenzol is more
soluble in it) is mixed with 16 grs. of malonic ester previously con-
verted into sodium malonic ester by treatment with the sodic ethylate
from 2.3 grs. of sodium and about 100 to 125 c.c. of alcohol, and the
mixture allowed to stand over night at ordinary temperatures. The
red solution thus obtained is treated with water, which separates it
into two layers, a dark red aqueous solution containing the sodium
salt of bromdinitrophenylmalonic ester and sodic bromide, and a ben-
zol solution of the unaltered tribromdinitrobenzol with the oily product
of the reaction. The two layers are separated with a drop-funnel,
and the lower aqueous one acidified with dilute sulphuric acid ; this
throws down a yellowish white precipitate of the bromdinitrophenyl-
malonic ester, which is purified by crystallization from hot alcohol,
until it shows the constant melting point 70°.

In this way 20 grs. of tribromdinitrobenzol gave 7.9 grs. of the
bromdinitrophenylmalonic ester,* and 5.1 grs. of tribromdinitrobenzol
were recovered from the benzol solution. Subtracting this from the
20 grs. there are left 1-4.9 grs. of tribromdinitrobenzol, which entered
into the reaction, and should have given the same weight, 14.9 grs., of
the product, so that the yield was 53 per cent of the theory, or about
the same as that obtained by the process as given in our first paper,
which calculated in the same way becomes 50 per cent.

In the course of some other experiments the curious observation
was made that it is not necessary to use sodic ethylate to bring
about the action of malonic ester on the tribromdinitrobenzol, as a
little of the red salt of bromdinitrophenylmalonic ester was formed
when aqueous sodic hydrate was added to a mixture of the two organic

* Dr. G. D. Moore, in his work with me on the action of sodium acetacetic
ester on tribromdinitrobenzol, has found that a much better yield is obtained if
the mixture is boiled for an hour under a return condenser. This observation
was not made until the work described in this paper was finished and Mr.
Robinson had left Cambridge, so that it was not convenient to try the experi-
ment of boiling the solution of sodium malonic ester and tribromdinitrobenzol,
nor did it seem very desirable to do so, as when we first took up the subject
such an experiment was tried, and seemed to yield very unfavorable results.
I propose, however, to repeat the experiment next year, if I return to the study
of these compounds. C. L. J.

OF ARTS AND SCIENCES. 237

substances. This recalls the observation of Victor Meyer,* that hom-
ologues can be obtained from benzylcyauide by using the alkylhaloid
with powdered solid sodic hydrate instead of sodic ethylate, or from
desoxy benzoin by the use of an alcoholic solution of an alkaline
hydrate.

Study of the Reactions hj which the Bromdinitrophenylmalonic Ester

is formed.

One of the most important points, which want of time prevented us
from considering in our first paper, was the mechanism of the reac-
tions by which the bromdinitrophenylmalonic ester was formed from
the tribromdinitrobenzol, and accordingly this was one of the first
points to which we turned our attention in taking up the subject
again. The reaction consists essentially, as our aiwyses of the pro-
duct showed, in the replacement of one of the bromine atoms in the
tribromdinitrobenzol by the malouic ester radical C1I(C00C2II5)2,
and of a second by hydrogen forming CcH2Br(N02)2CH(COOC,H.)a,
and the obscure part of it is the manner in which this second atom of
bromine is replaced by hydrogen. The first step toward clearing up
this obscurity was obviously to make out whether this atom of bro-
mine was eliminated in the form of some organic compound, or as
sodic bromide, which was done by the following quantitative deter-
minations of the amount of sodic bromide formed in the reaction.

I. 20 grs. of tribromdinitrobenzol were treated with the sodium
malonic ester from 16 grs. of malonic ester in the way already de-
scribed, and, after precipitating the bromdinitrophenylmalonic ester
from the aqueous solution with dilute nitric acid, and filtering it out,
the amount of sodic bromide was determined by precipitation of argen-
tic bromide from an aliquot part of the filtrate, and calculating the
amount in the entire solution. In this way the whole solution was
found to yield 13.81 grs. of argentic bromide, corresponding to 5.88 grs.
of bromine. The benzol solution on evaporation yielded 5.1 gr>. of
tribromdinitrobenzol, which had not taken part in the reaction, leaving
14.9 grs. which had acted, and which would yield 5.89 grs. of bromine,
if two of the atoms of bromine had been removed from each molecule
as sodic bromide.

II. 20 grs. of tribromdinitrobenzol gave under the same conditions
13.8 grs. of argentic bromide corresponding to 5.87 grs. of bromine,
but only 4.9 grs. of tribromdinitrobenzol were recovered unaltered.

* Ber. d. cli. G. 1888, p. 1201.

238 PROCEEDINGS OF THE AMERICAN ACADEMY

The 15.1 grs. of tribromdinitrobenzol which entered into the reaction
should have lost 5.96 grs. of bromine.

Percentage of the theoretical amount of bromine removed from tri-
bromdinitrobenzol in the form of sodic bromide : —

I. ii.

99.86 98.49

These numbers agree as closely as could be expected, when the
necessary sources of error in the process are considered, and prove
that all the bromine removed by the reaction is converted into sodic
bromide.

The second point to be determined was from what source the atom
of hydrogen was obtained which replaced the atom of bromine. For
this purpose we investigated the oily product of the reaction, and,
after we had found that nothing definite could be obtained from it
by distillation under ordinary pressures, or with steam, and that sodic
hydrate or other alkaline saponifying agents gave most unpromising
results, we tried heating it with strong hydrochloric acid in a sealed
tube to 130° for 20 hours, as the bromdinitropheuylmalonic ester
which was dissolved in the oil would be converted by this treatment
into the corresponding phenylacetic acid, with the properties of which
we were familiar, and which therefore we thought could be removed
easily from the other products of the reaction. The tubes after heat-
ing contained an aqueous solution and a small amount of viscous
matter, which gradually changed into an amorphous solid. These
were separated by filtration, and the filtrate evaporated to dryness on
the water bath, leaving a crystalline residue, which was extracted with
cold water to leave behind the bromdinitrophenylacetic acid present.
The extract after being evaporated again on the water bath gave crys-
tals, which made up about one half the amount of the residue from
the first evaporation. They were freely soluble in water or alcohol,
nearly insoluble in ether, and could be sublimed at temperatures below
their melting point forming glassy white thick needles or prisms,
which melted at 185°. These properties prove that the crystals are
tartronic acid, as they are identical with those given by Conrad and
Bischoff for that substance.* We tried also to confirm this inference
by an analysis, but the small amount of the acid which after some
unsuccessful experiments remained at our disposal, little more than
0.1 gr., prevented us from purifying it thoroughly, as sublimation,

* Ann. Cliem., ccix. 222.

OP ARTS AND SCIENCES. 289

which had given the substance melting at 185°, turned out to be too
wasteful a process for working up large amounts ; we were not sur-
prised, therefore, that our analysis of the silver salt gave 62.82 per cent
of silver, instead of the 64. G9 per cent calculated for argentic tartio-
nate. This result, however, bad as it is, comes nearer to that for argen-
tic tartronate than to that for any other substance which could well be
formed, and therefore gives us the desired confirmation.* The small
amount of tartronic acid obtained, less than 0.5 gr. from about 10 grs.
of the oil, may be accounted for on the supposition that the larger
part of the tartronic acid was decomposed by the strong hydrochloric
acid used in the saponification of the oil, as Conrad and Bischoff f have
observed that tartronic acid is decomposed very easily by hydrochloric
acid, and we also were unable to obtain a trace of the acid from the
filtrate after argentic chloride had been precipitated from the silver
salt by a slight excess of hydrochloric acid. In fact, we should not
have thought it worth while to try to obtain tartronic acid by the ac-
tion of hydrochloric acid on our oil, if it had not been for the fact
that tartronic ester has been made from calcic tartronate by the action
of hydrochloric acid gas and alcohol, \ which would seem to show that
the tartronic acid is more stable in presence of strong than of dilute
hydrochloric acid, improbable as this sounds. This view is in harmony
with our observations detailed above, and Conrad and Bischoff state
that even very dilute acid decomposes it readily ; but even granting
this, a large proportion of the tartronic acid must have been decom-
posed during the treatment with hydrochloric acid, and this seems to
us sufficient to make it more than probable that the tartronic acid is
the product of the principal reaction, and not formed in a secondary
one taking place only to a limited extent. If this is admitted, the fol-
lowing reactions would give the most probable explanation of what
takes place.

I. C6HBr„(N02)2 + NaCII(COOC2H/)2 =

C6HBr2(N02)2CH(COOC2H5)2 + NaBr.
II. C6HBr2(N02)2CH(COOC2H5)2 + NaCH(COOC2H5)2 =
CcIIBr2(N02)2CNa(COOC2H.)2 + CH2(COOC2H5)2.

* No attempt was made to obtain a better analysis of the tartronic acid,
because the work upon, the corresponding trinitro compound described in an-
other paper furnished a complete confirmation of the theory of the reactions
given here.

t Ann. Chem., ccix. 223.

J Freund, Ber. d. eh. G., xvii. 786. Pinner, Ibid., xviii. 757.

240 PROCEEDINGS OP THE AMERICAN ACADEMY

III. C6HBr2(N02)2CNa(COOC2Hs)2 + CH2(COOCaH5)2 =

CeH2Br(N02)2CNa(COOC2H5)2 + CHBr(COOC2H5)2.

IV.. NaCH(COOC2H5)2 + H20 = NaOH + CH2(COOC2H5)2.
V. CHBr(COOC2H5)2 + NaOH = CHOH(COOC2H5)2 + NaBr.

Of these reactions, I., II., and III. take place before, IV. and V.
after, the addition of the water used in working up the product. That
Reaction III. takes place we infer from our discovery of tartronic i cid
among the products of saponification of the oil, as Conrad and Bischoff *
have proved that brommalonic ester is decomposed by sodic hydrate,
according to Reaction V.f

The experiments of one of us with G. D. Moore upon the action
of sodium malonic ester on tribromtrinitrobenzol have shown that
these reactions take place in a way analogous to that just worked out.
For the details we would refer to the paper on this subject.

The objection which might be raised against these reactions, that
the yield of bromdinitrophenylmalonic ester is at best only about 50
per cent of that required by theory, and therefore that some other
substance may be formed from the tribromdinitrobenzol in addition to
it, is disposed of by the fact that on making the dibromdinitrophenyl-
malonic ester from tetrabromdinitrobenzol by exactly analogous re-
actions, the yield rose to 80 per cent of the theoretical. t It follows,
therefore, that the 47 to 50 per cent unaccounted for in our preparations
remained dissolved in the oily secondary product ; but although we
have made many attempts to recover it, none of them have been
crowned with success.

The absence of malonic acid in the product of the action of hydro-
chloric acid on the oil seems to us extraordinary, as we purposely kept
the temperature of the saponification at 130°, in order to avoid decom-
posing the malonic acid, which we expected from the excess of malonic
ester undoubtedly present in the oil. We can account for this only
by supposing that malonic acid is decomposed in presence of strong
hydrochloric acid at a lower temperature than when heated alone.

Saponification of Bromdinitrophenylmalonic Ester-

In our previous paper we stated that strong hydrochloric acid de-
composed bromdinitrophenylmalonic ester, when the two substances

* Ann. Chetn., ccix. 222.

t A fuller discussion of these reactions will be found in the general paper at
the end of this series.

t These Proceedings, xxiv. 295.

OP ARTS AND SCIENCES. 241

were heated together to 140°-145° in a sealed tuhe, and described the
product formed, stating, however, that all our results up to that time
must be regarded as preliminary. Upon taking up the subject again,
we studied in the first place the gas evolved when the tube was
opened, and found that, in addition to the gas already mentioned burn-
ing with a green-bordered flame (which is undoubtedly ethylchloride),
carbonic dioxide was given off, as was proved by the precipitate formed
when the gas was passed through lime-water. It is to be observed,
however, that if the temperature to which the tubes were heated was
140°, or a little lower, very little, if any, carbonic dioxide was given
off, and the product insoluble in the hydrochloric acid was oily ; on
the other hand, when a crystalline product was obtained at a some-
what higher temperature, carbonic dioxide was given off freely. "VVe
next attempted to prepare the substance in open vessels by substitut-
ing for strong hydrochloric acid dilute sulphuric acid boiling at about
the temperature to which we heated the tubes, and found that not only
did we obtain the product in this way, but that it was much purer than
that made in the sealed tubes with hydrochloric acid. We have ar-
rived accordinslv at the following method as the most convenient for
saponifying the bromdinitrophenylmalonic ester. 2 grs. of the ester
are mixed with about 100 c.c. of dilute sulphuric acid, specific gravity
1.44 and boiling point 132°, and the mixture boiled in a flask with a
return condenser, until the oily drops of melted bromdinitrophenyl-
malonic ester have all gone into solution, which usually takes about
an hour and a half. After this the solution, as it cools, deposits long,
pale yellow needles of the new substance, which, if the preparation
has succeeded well, melt at 177° at once. If, however, the action has
not run so well, and the melting point is somewhat lower than this,
the substance can be easily purified by crystallization from boiling
water acidified with a few drops of dilute sulphuric acid, but in no
case should it be crystallized from pure water, or alcohol, as these
solvents decompose it in the curious way described later in this paper,
when we consider its properties.

Metabromdinitrophenylacetic Acid, C0II2Br(N02)oClI,COOII.

The substance prepared in the manner just described, and showing
the constant melting point 177°, was dried at 100°, and analyzed with
the following results : —

I. 0.2106 gr. of the substance gave on combustion 0.2426 gr. of

carbonic dioxide, and 0.0420 gr. of water.
vol. xxiv. (n. s. xvi.; 16

242 PROCEEDINGS OF THE AMERICAN ACADEMY

II. 0.1933 gr. of the substance gave 15.6 c.c. of nitrogen at a tem-
perature of 15°. 5, and a pressure of 759.2 mm.
III. 0.1821 gr. of the substance gave, according to the method of
Carius, 0.1152 gr. of argentic bromide.

Found.

in.

Carbon

Calculated for
C6H2Br( NO, )2CH2C02H.

3L47"

i.
31.43

Found.
II.

Hydrogen
Nitrogen

1.64
9.18

2.21

9.42

Bromine

26.23

26.93

These analyses supersede those already published,* which were
made with a substance prepared by crystallization from dilute alcohol,
before we had found that this solvent decomposed the compound ; it
was therefore decidedly impure, as indeed was shown by its melting
point, which was 170° instead of 177°.

The yield of the bromdinitrophenylacetic acid is essentially quanti-
tative, as 2 grs. of the bromdinitrophenylmalouic ester gave, after
treatment with dilute sulphuric acid, 1.4 grs. of bromdinitrophenylacetic
acid melting at 177°, instead of the calculated amount, 1.56 grs., that
is, 90 per cent of the theory; the missing 10 per cent without doubt
remained dissolved in the dilute sulphuric acid, in which the substi-
tuted acetic acid is not wholly insoluble, as the 1.4 grs. were obtained
by simple filtration, no attempt being made to recover the organic
substance from the filtrate.

In view of the observations described above, there can be no doubt
that the following reactions take place when bromdinitrophenylmalonic
•ester is saponified by heating it with hydrochloric acid at temperatures
in the neighborhood of 145° :

CcH2Br(N02),CH(COOC,H,)2 + 2 HC1 =

CGH2Br(X02)2CH(COOH)2 + 2 C2H.C1 =
CJI2Br(N02)2CH2COOH + C02+ 2 C2H5C1.

And there can be no question that a similar reaction takes place when
dilute sulphuric acid is substituted for strong hydrochloric acid. In
fact, this was so obvious that we have not thought it worth while to
try to isolate the sulphovinic acid which must be formed in this case
instead of the ethylchloride.

Properties. — The metabromdinitrophenylacetic acid crystallizes
from an aqueous solution containing a little sulphuric acid in yellowish

* These Proceedings, xxiv. 11.

OF ARTS AND SCIENCES. 243

white needles, often one to two centimeters long, which appear under
the microscope as tolerably thick prisms frequently tapering owing
to modification by the planes of a pyramid with, a very acute angle,
but almost always, even in addition to this tapering, terminated bluntly
by the planes of another very obtuse pyramid. Occasionally the ter-
mination is a single plane at a moderately acute angle to the sides, but
this looked as if it were due to cleavage. The crystals are generally
very well developed, but if they are small, a tendency to twin longi-
tudinally is observed. If the substance is crystallized from alcohol,
the forms like feathers, or half-feathers, described in our previous
paper, and groups of branching needles looking like certain feathery
seaweeds, appear ; but as the alcohol decomposes the substance, these
forms can hardly be ascribed to the substituted acetic acid itself. It
melts at 177° ; and is essentially insoluble in ligroine or carbonic
disulphide ; very slightly soluble in chloroform, slightly in benzol,
more soluble in both of these solvents when hot ; tolerably soluble in
ether, separating from the solution in a liquid form, which solidifies
on stirring ; easily soluble in acetone or glacial acetic acid ; soluble
in cold alcohol, more freely in hot, but the alcohol decomposes it, as
is indicated by the change in the appearance of the crystals, the long
yellowish prisms of unaltered bromdinitrophenylacetic acid becoming
roughly studded with fine needles of another substance, and by the
lowering of tke melting point, when the substance is crystallized from
alcohol ; evaporation to dryness of the alcoholic solution, if repeated
three times, reducing the melting point from 177° to 147° ; while after
two more evaporations, making five in all, the melting point had sunk
to 103°— 104°, at which point it remained constant after repeated
crystallization. The substance obtained in this way crystallized in
rectangular prisms, or plates, was not acted on by alkalies, and was
recognized as Grete's metabromdinitrotoluol.* For greater certainty
it was analyzed, with the following results : —

I. 0.1690 gr. of the substance gave by the method of Carius
0.1210 gr. of argentic bromide.
II. 0.2185 gr. of the substance gave 20.3 c.c. of nitrogen at a
temperature of 22°. 5 and a pressure of 773.3 mm.

Bromine
Nitrogen

Calculated for
C6H2Br(N02)2CH3.

30.65
10.73

Found.
I. II.

30.47

10.69

* Ann. Cliem., clxxvii. 258.

244 PROCEEDINGS OP THE AMERICAN ACADEMY

This conversion of the metabronidinitrophenylacetic acid into the
corresponding toluol by boiling with alcohol, when it is so stable in
presence of boiling dilute sulphuric acid, seemed to us so interesting
that we have examined it more carefully. Having in the first place
determined that the alcohol was neutral to test papers, to prove that
the reaction could not be due to a trace of alkali, or acid in it, a
quantity of the substance melting at 177° was boiled in a flask under
a return condenser with a little alcohol for three hours, and pure air
occasionally drawn through the apparatus to sweep out its contents
into a set of absorption bulbs filled with lime-water ; in this way a
heavy precipitate was obtained, which consisted in great part of
calcic hydrate precipitated by the alcohol vapor, but also contained
calcic carbonate, as was shown by effervescence when it was dissolved
in an acid, thus proving that carbonic dioxide had been given off.
The alcoholic solution, after it had been boiled for an hour and a half
longer, gave crystals, melting at 103° to 104°. It seems evident,
therefore, that the reaction runs as follows,

CeH2Br(N02)aCH2COOH = C6H2Br(N02)2CH3+C02,

and the alcohol seems to take no part in it. Methyl alcohol, in which
the substance is more soluble than in ethyl alcohol, brings about a
similar decomposition, but more slowly. In boiling water the meta-
bromdinitrophenylacetic acid is tolerably soluble, but nearly insoluble
in cold water. The decomposition produced by alcohol also takes
place when the aqueous solution is boiled, but very slowly, so that
long boiling is necessary to convert the substance completely into the
substituted toluol ; nevertheless, after distilling with steam for several
hours, the whole of the substance had passed over into the receiver
as metabromdinitrotoluol, since none of the substituted acetic acid,
which does not distil with steam, was left in the flask. The presence
of a very small amount of sulphuric acid (two or three drops of the
dilute acid to 100 c.c. of water) is sufficient to prevent this decomposi-
tion entirely, as a specimen of the substituted acetic acid showed no
signs of decomposition, even after having been distilled with steam for
three hours with only the amount of dilute sulphuric acid which
remained adhering to it after the excess of acid used in its preparation
had been removed by filtration. The best solvent for the bromdini-
trophenylacetic acid therefore is boiling water containing a few drops
of sulphuric acid. If the substance has become contaminated with
the substituted toluol, it can be purified conveniently by distilla-
tion with steam in presence of a few drops of sulphuric acid, when

OF ARTS AND SCIENCES. 2-15

the toluol passes over, and the acetic acid remains behind in the
flask.

Amnionic hydrate dissolves the metabromdinitrophenylacetic acid,
forming a colorless solution if the acid is in excess ; but if the amnio-
nic hydrate is added in slight excess, a green solution is obtained,
which turns dark brown if a large excess of ammonic hydrate is
added, but regains its green color on dilution with water. Both the
colorless and green solutions smell of ammonia, and this smell could
not be removed by gentle boiling, or by allowing the solution to stand
over sulphuric acid, from which we infer that the ammonium salt is
a decidedly unstable substance. When either solution is warmed on
the water bath for some time it is partially decomposed with formation
of a precipitate of metabromdinitrotoluol, and the green solution turns
yellow even when allowed to evaporate spontaneously. Hydrochloric
acid threw down from the solution a precipitate of metabromdini-
trophenylacetic acid. The ammoniacal solution, prepared with an
excess of the acid and freed as completely as possible from an excess
of ammonia, gave no precipitate with baric, strontic, or calcic chloride,
but with the salts of many of the heavy metals it gave precipitates,
of which the following were the most characteristic: —
Aluminic chloride, heavy, white.
Ferric chloride, heavy, very pale bluff.
Cupric sulphate, heavy, pale blue.
Mercuric nitrate, heavy, yellowish white.

Mercuric chloride, a very slight precipitate, probably white precipi-
tate from the slight excess of ammonic hydrate.
Mercvrous nitrate, heavy, white.
Plumbic acetate, heavy, white.
Argentic nitrate, heavy, white.

The behavior with argentic nitrate is especially characteristic ; if
the solutions are strong, the mixture becomes nearly solid from the
formation of a white flocculent precipitate, which frequently swells up,
forming a little heap or pyramid, and after standing for some hours
becomes converted into good-sized crystals. The analysis and proper-
ties of this salt are described more in detail below.

With sodic hydrate a solution of the acid turns a deep Prussian
green color, which passes in time into a yellowish brown, this latter
color appearing at once when an excess of sodic hydrate is added. If
the green solution is acidified, a precipitate is obtained, which, although
nearly colorless when dry, becomes purple when moistened with alco-
hol. This precipitate was oily at first, and, although it solidified later.

246 PROCEEDINGS OF THE AMERICAN ACADEMY

we did not succeed in finding any solvent from which it could be crys-
tallized. That the sodic hydrate had acted on the bromine was shown
by an actual experiment, which gave the following results.

0.67 gr. of the acid yielded by treatment with sodic hydrate so as
to form the green salt 0.1136 gr. of argentic bromide, which corre-
sponds to 0.0483 gr. of bromine instead of the 0.1757 gr. contained in
the amount of acid taken ; that is, 27.5 per cent of the whole had been
removed.

The yellowish brown solution obtained, when a large excess of sodic
hydrate was used, or the mixture of sodic hydrate and metabromdi-
nitrophenylacetic acid was boiled for some time, gave on addition of
an acid a viscous red substance, which was exceedingly unmanageable.
That bromine was removed in quantity in this way is shown by the
following determination.

0.2412 gr. of the metahromdinitrophenylacetic acid was boiled with
sodic hydrate for three quarters of an hour ; the product, after acidify-
ing with nitric acid, and filtering out the precipitate, gave 0.1180 gr.
of argentic bromide, corresponding to 0.0502 gr. of bromine instead of
the 0.0632 gr. contained in the weight of acid taken ; that is, 79.4 per
cent of the whole.

It is evident, therefore, that the original compound has undergone a
deep-seated alteration, and the new crimson substance is probably a
phenol. This view of its nature receives some confirmation from the
fact that it is decomposed with the greatest ease by even somewhat
dilute nitric acid. We have not yet succeeded in bringing either of
the products derived from sodic hydrate into a state fit for analysis,
although the slight solubility of the barium salt of the crimson acid
awakened hopes of success ; but these ended in disappointment, as the
salt proved quite as viscous as the free acid.

Potassic carbonate gives with the metahromdinitrophenylacetic
acid results similar to those obtained with sodic hydrate, but less
marked.

We hoped to be able to throw some light on this action of alkalies
on the metahromdinitrophenylacetic acid by the study of the correspond-
ing anilido compound, but have not succeeded as yet in preparing it,
as aniline converts the free acid direct into metanilidodinitrotoluol,*
C6H2CH8(C6H6NH)(N02)2, melting point 142°, and the saponification
of anilidodinitrophenylmalonic ester with sulphuric acid yielded only
uninviting viscous products.

* Hepp, Ann. Chem., ccxv. 371.

OF ARTS AND SCIENCES. 247

Radziszewski,* Gabriel and R. Meyer,f and Heckmann,^ have pre-
pared a dinitrophenylacetic acid which differs from that just described
only in containing no bromine. This substance also loses carbonic
dioxide easily, passing into diuitrotoluol, but none of the chemists
who have worked with it describe a decomposition by solvents similar
to that observed by us ; in fact, they crystallized it for analysis from
boiling water, but, as in all their methods of preparation acids were
used, it is possible that a sufficient amount of sulphuric acid was pres-
ent during the crystallization to prevent the decomposition. It ap-
pears that none of them tried to crystallize it from alcohol. The
statement is made, however, that the sodium or potassium salt of the
acid decomposes slowly on standing, instantly on boiling giving car-
bonate and diuitrotoluol ; from this it would seem probable that our
bromine acid was more stable than that containing no bromine, as our
ammonium salt (the sodium or potassium salt could not be obtained)
stood unaltered for many days, and even could be warmed on the
water bath for some minutes with only slight decomposition. The
curious change of color with sodic hydrate observed by us did not
occur with the dinitro body, confirming our conclusion that this action
was due to the removal of bromine by the sodic hydrate.

A rgentic Metabromdinitrophenylacetatey
C6H2Br(N02)2CH2COOAg . H20.

This substance was prepared by adding a solution of argentic nitrate
to a solution of the acid in amnionic hydrate, which had been freed
from the excess of ammonia as completely as possible, best by using
an excess of the acid in preparing the salt. The heavy flocculent
precipitate, which, if the solutions were strong, entirely filled the
liquid, and even piled up above its surface, became crystalline on
standing, and was purified by washing with cold water. Some of the
salt is also formed when argentic nitrate is added to an aqueous solution
of the free metabromdinitrophenylacetic acid. The analysis of the
salt gave the following results.

0.3530 gr. of the air-dried salt lost, when heated to 100°, 0.0149 gr.

Calculated for
C6H2Br(N02)2CH,C02Ag . H20. Found.

Water 4.19 4.22

* Ber. d. ch. G., ii. 210. J Ann. Chem., ccxx. 134

t Ibid., xiv. 823.

248 PROCEEDINGS OF THE AMERICAN ACADEMY

I. 0.1674 gr. of the salt dried at 100° gave, after being heated with
fuming nitric acid in a sealed tube * and precipitated with
potassic bromide, 0.0761 gr. of argentic bromide.
II. 0.1556 gr. of the salt treated in the same way gave 0.0707 gr.
of argentic bromide.

Calculated for Found.

C6H2Br(N02)2CH2CO,Ag. I. II.

Silver 26.21 26.11 26.10

Properties. — The argentic bromdinitrophenylacetate forms white
flat prisms terminated by a single plane at an oblique angle ; the ter-
minated end is somewhat broader than the other, so that the crystals
look like a flattened base-ball bat. They have also a tendency to
twin longitudinally. When put in the flame of a Bunsen lamp, they
burn with a sparkling flame, but when heated carefully on a piece of
porcelain, decompose quietly leaving a residue of silver. The sub-
stance is very slightly soluble in cold water, so that it can be washed
without too great loss ; more soluble in hot water, from which it crys-
tallizes apparently without decomposition ; insoluble in cold alcohol,
but slowly decomposed, if heated with it.

Constitution of Bromdinitrophenylmalonic Ester.

The formation of metabromdinitrotoluol from bromdinitrophenyl-
malonic ester by a series of reactions giving essentially a quantitative
yield and taking place at temperatures not above 150°, throws a great
deal of light on the nature of this substance. In the first place, its
conversion into a toluol derivative is a welcome confirmation of the
formula which we have given it, and which did not rest on the most
secure experimental foundations. In our first paper it was stated that
the analysis left the choice open between the following formulas,

I. C,H,Br(N02),CH(COOC,H5)2,

II. C6HBr(N02)2C(COOC2H5)2,

and we decided in favor of the first on account of the analyses of the
sodium salt, an amorphous substance for the purity of which we had
no guaranty, and also on account of the ease with which this salt was
formed and decomposed, the product of the decomposition being the
original ester. Against this formula stood only our inability at the

* We had hoped in this way to determine both the silver and bromine in one
operation, but found that the filtrate gave a precipitate with potassic bromide.
It would seem, therefore, that the presence of an excess of argentic nitrate is
necessary to retain all the bromine. *

OF ARTS AND SCIENCES.

249

time to explain the reactions by which it was formed, an objection
which is removed by this paper. The easy conversion of a substance
having Formula I. into a toluol compound is to be expected, whereas
with one having Formula II. it would be necessary to add two atoms
of hydrogen, an action which could hardly take place quantitatively
by boiling with dilute sulphuric acid, or distillation with steam, the
two processes employed in converting the bromdinitrophenylmalonic
ester into the bromdinitrotoluol. This formation of the substituted
toluol therefore establishes Formula I. beyond question.

In the second place, the way is opened to the determination of the
position of the substituting radicals upon the benzol ring in all these
compounds, as their position must be the same as in the bromdinitro-
toluol. Messrs. W. B. Bentley and W. H. Warren have accordingly,
at our request, determined the constitution of the bromdiuitrotoluol by
replacing its bromine successively by the amido group and by hydro-
gen, and have found in this way that its constitution is

CH3 . Br . NOa . N02 .1.3.4.6.

For the details of their work we would refer to their paper, which fol-
lows immediately after this. It follows from this that in the tribrom-
dinitrobenzol Br . H . Br . N0.2 . Br . N02 . 1 . 2 . 3 . 4 . 5 . 6, the bromine
numbered 1, which stands in the ortho position to one nitro group and
the hydrogen, in the para position to the other nitro group, has been
replaced by the malonic ester radical CH(COOC2H5)2 ; that the bro-
mine replaced by hydrogen is the one between the two nitro groups ;
and that the third bromine, which is in exactly the same relation to
the nitro groups as the one replaced by the malonic ester radical, is
left entirely unaltered. This last fact seems very strange to us. The
constitution of our bromdinitrophenylmalonic ester is therefore repre-
sented by the following graphical formula : —

CH(COOC2H5)2

250 PROCEEDINGS OF THE AMERICAN ACADEMY

XX.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF
HARVARD COLLEGE.

ON SOME NITRO DERIVATIVES OF
METABROMTOLUOL,

By W. B. Bentley and W. H. Warren.

Presented May 28, 1889.

Determination of the Constitution of Metabromdinitrotoluol,
C6II2CH3Br(N02)2.

Our attention was first called to this subject by Prof. C. Loring
Jackson, who, having with W. S. Robinson* obtained the metabrom-
dinitro toluol by the decomposition of their bromdinitrophenylmalonic
ester, asked us to determine its constitution.

This metabromdinitrotoluol, melting point 103°-104°, was discovered
by Grete,f who made it by the action of fuming nitric acid on meta-
bromtoluol, or metabrommononitro toluol, and assigned to it the con-
stitution CH3, N02, Br, N02, 1, 2, 3, 4, although he was very doubtful
about the position of the second nitro group, giving it the para posi-
tion without any experimental reason for doing so. His proof that
one of the nitro groups was in the ortho position consisted in estab-
lishing the identity of the bromtoluidine made by the reduction of his
bromnitrotoluol with that obtained by the action of bromine on ortho-
acettoluid ; % but this obviously leaves it doubtful whether this nitro
group stands in the position 2 or 6 to the methyl of the toluol, and in
fact later work has shown that Grete was wrong in ascribing to it
the position 2, as his mononitro compound really has the constitution
CH3, Br, N02, 1, 3, 6. The proof of this was given by Nevile and
Winther,§ who, by replacing the amido group in the metabromor-
thotoluidine (melting point 55°-56°) already mentioned by bromine,

* See preceding paper.

t Ann. Chem., clxxvii. 258.

X YVroblewsky, Ann. Chem., clxviii. 1G1.

§ Ber. d. eh. G., xiii. 962.

OF ARTS AND SCIENCES. £51

obtained a liquid dibrotntoluol, which on oxidation with nitric acid
yielded the dibrombenzoic acid melting at 151°-153° previously made
by Von Richter* from paradibrombenzol.

The only point therefore not settled in regard to the constitution
of Grete's metabromdinitrotoluol was the position of the second nitro
group, which might be either para or ortho to the methyl. To deter-
mine this the metabromdinitrotoluol was heated with alcoholic ammo-
nia f in sealed tubes to 100° for 12 hours, when it was found to be
converted into a yellow substance, which had partly separated in the
solid state, and partly remained in solution in the alcohol, from which
it was obtained by evaporation. The product was purified by wash-
ing with water to remove the amnionic bromide, and crystallization
from hot glacial acetic acid, until it showed a constant melting point
(193°-194°), when it was dried at 135°, and the following analysis
showed that a diaitrometatoluidine had been formed.

0.2658 gr. of the substance gave 51.2 c.c. of nitrogen at a temper-
ature of 24° and a pressure of 767.5 mm.

Calculated for

C6H2CH3NH,(N02)2.

Found.

21.31

21.71

Nitrogen

The melting point showed that this substance is identical with the
dinitrotoluidine made by Heppl by the action of alcoholic ammonia on
his y trinitrotoluol, but the constitution of this trinitrotoluol had not
been determined. Staedel § has also announced recently that one of
his students, Herr Adalbert Kolb, has prepared a dinitrotoluidine from
dinitrokresolether and determined its constitution as CH3, NH„ NO.,,
N02, 1, 3, 4, 6 ; but, as he neglected to give the melting point, we were
unable to tell whether it was identical with ours or not, and have
been forced to work out the constitution of our dinitrotoluidine by the
replacement of the amido group by hydrogen. In doing this some
difficulty was encountered because of the very slight solubility of the
dinitrotoluidine in alcohol, but we found on experiment that the Griess
reaction would take place satisfactorily in a mixture of acetone and
alcohol, and accordingly proceeded as follows. 2 grs. of the dinitro-

* Ber. d. ch. G., vii. 1146.

t When metabromdinitrotoluol was allowed to stand with alcoholic ammonia
in the cold, a dark blue solution was formed at first, but the color gradually
changed to reddish brown.

J Ann. Chem., ccxv. 371.

§ Ber. d. ch. G , xxii. 21-3.

252 PROCEEDINGS OF THE AMERICAN ACADEMY

toluidine were dissolved in about 12 c.c. of acetone, 8 c.c. of common
alcohol added, and the mixture acidified with strong sulphuric acid,
after which 1.5 grs. of sodic nitrite, about twice the calculated amount,
was added in small portions at a time, and the liquid warmed gently
until the evolution of nitrogen ceased, when the larger part of the
acetone and alcohol was distilled off at first on the water bath, finally
on the sand bath, and the residue distilled with steam ; the dinitrotoluol
passed over into the receiver in small white crystals, which were
removed by filtration, and crystallized from hot alcohol until they
showed the constant melting point 71°. It was not worth while to'
dilute the distillate of acetone and alcohol, as no precipitate was ob-
tained in that way. For greater certainty the product was analyzed,
with the following results : —

0.2456 gr. of the substance gave 33.1 c.c. of nitrogen at a tempera-
ture of 22° and a pressure of 764.2 mm.

Calculated for
CoILjCHaCNOj;);,. Found.

Nitrogen 21.31 21.71

There is no question therefore that the substance is the orthopara-
dinitrotoluol, CH3, NO.,, NO.„ 1, 2, 4, and the bromdinitrotoluol accord-
ingly has the following constitution: CH3, Br, N02, N0.2, 1, 3, 4, 6. It
follows also that our dinitrotoluidine is identical with that described
by Kolb, and, if Staedel had given its melting point, it would not have
been necessary for us to determine its constitution. As has been
already stated, this dinitrotoluidine melts at the same point as that
made by Hepp from alcoholic ammonia and his y trinitrotoluol. To
establish the relation between these substances more firmly, we made
the dinitrophenyltoluidine by treating our dinitrobromtoluol with ani-
line, and found that the product showed the same melting point, 142°,
as that of the compound made by Hepp from y trinitrotoluol and
aniline. The y trinitrotoluol of Hepp therefore has the following
constitution : CII.,, N0.2, NCX, N02, 1, 3, 4, 6 ; which is in harmony with
Laubenheimer's rule,* that a nitro group is removed by the action of
alcoholic ammonia only when it is in the ortho position to another
nitro group.

Metabromtrinitrotoluol, CcHCII3Br(N02) ,.

In preparing the metabromdinitrotoluol, we found, if a mixture of
fuming nitric acid and sulphuric acid was used, that a new substance

* Ber. d. oh. G., ix. 7G6, 1828.

OF ARTS AND SCIENCES. 253

melting above 104° was obtained, which turned out to be the as yet
undescribed metabromtriuitrotoluol. This compound is most easily
prepared from the metabromdinitrotoluol, when it is convenient to
proceed as follows. 5 to 10 grs. of metabromdinitrotoluol were placed
in a flask, and 10 to 20 c.c. of a mixture of two volumes of fuminf

9 c*

nitric acid and one of strong sulphuric acid added. The whole was
then boiled until the evolution of red fumes had nearly ceased, when,
after it was cool, it was poured in a fine stream into a beaker of cold
water, stirring the liquid vigorously during the addition of the acid
solution, as in this way the product is precipitated in a granular form
much more easy to manage than the large compact lumps obtained if
the stirring is neglected. The product was then washed with cold
water till free from acid, and purified by crystallization from alcohol,
until it showed the constant melting point 143°. The residue from
the mother liquors, consisting of a mixture of metabrom trinitrotoluol
and the corresponding dinitro compound, can be advantageously used
for preparing a fresh quantity of the trinitro body. The pure sub-
stance was dried at 120°, and analyzed with the following results: —

I. 0.3372 gr. of. the substance gave on combustion 0.3356 gr. of

carbonic dioxide and 0.0453 gr. of water.
II. 0.2651 gr. of the substance gave 32.9 c.c. of nitrogen at a tem-
perature of 21° and a pressure of 761.5 mm.
III. 0.1830 gr. of the substance gave, by the method of Carius,
0.1120 gr. of argentic bromide.

Found.

III.

Carbon

Calculated for
CGHCH3Br(X02)3.

27.46

i.
27.14

Found
II.

Hydrogen

1.31

1.49

Nitrogen

26.14

26.0'

Bromine

13.73

14.09

The yield was about 60 per cent of the theoretical.

Properties. — The metabromtrinitrotoluol crystallizes from alcohol
in small white needles, which melt at 143°. They are insoluble in
water, or ligroiue ; nearly insoluble in cold alcohol, only sparingly
soluble in hot ; slightly soluble in carbonic disulphide ; soluble in
ether, methyl alcohol, benzol, chloroform, glacial acetic acid, or acetone.
Boiling alcohol we found the best solvent for it, although it is so
slightly soluble in it. Aqueous sodic hydrate seemed to have no
action upon it, nor was it affected by the strong acids. The bromine
is removed easily, which we proved by the action of alcoholic ammonia
or aniline on it, as will be described later in this paper.

254 PROCEEDINGS OF THE AMERICAN ACADEMY

Constitution of Metabromtrinitrotoluol.

As the metabromtrinitrotoluol is made from the raetabromdinitro-
toluol, the only point to be determined is the position of the third nitro
group ; for this purpose we converted it into the corresponding trini-
trotoluidine by treatment with alcoholic ammonia in the cold. The
mixture was allowed to stand in a corked flask for about twelve hours ;
at first a dark blue color appeared in the liquid, but on longer stand-
ing this turned to a deep reddish brown, and a precipitate was de-
posited, which with the supernatant liquid at the end of the twelve
hours was poured into a dish, and the solvent allowed to evaporate
spontaneously. The residue, which was red and yellow, was washed
till free from ammonic bromide, and then purified by crystallization
from hot glacial acetic acid, until it showed the constant melting point
136°, when it was dried at 120°, and analyzed with the following
result : —

0.1144 gr. of the substance gave 24.2 c.c. of nitrogen at a tempera-
ture of 23° and a pressure of 754.7 mm.

Calculated for
C6HCH3NH,(N02)3. Found.

Nitrogen 23.14 23.51'

The melting point of this trinitrotoluidine shows that it is identical
with that prepared by Nolting and Salis* by the action of alcoholic
ammonia on trinitrometakresolethylether, to which they assign the
constitution CTI3, N02, NH2, N02, N02, 1, 2, 3, 4, 6, on the ground
that the alcoholic ammonia did not remove any of the nitro groups,
and therefore, according to Lauhenheimer,| no two of them could be
in the ortho position to each other, and this is the only possible ar-
rangement in which no two nitro groups are in the ortho position.
Our metabromtrinitrotoluol must consequently have the constitution
CH3, N02, Br, N02, N02, 1, 2, 3, 4, 6 ; and this conclusion is confirmed
by the fact recently discovered in this laboratory,! that the bromtrini-
trophenylmalonic ester is converted by boiling with sulphuric acid of
specific gravity 1.44 into this metabromtrinitrotoluol melting at 143°.
As the bromtrinitrophenylmalonic ester is made from the symmetrical
tribromtrinitrobenzol, Br, N02, Br, N02, Br, N02, 1, 2, 3, 4, 5, 6, it
can have only the constitution CH(COOC2tL)„, N02, Br, N02, N02,

* Ber d. ch. G., xv. 1864.

t Ber. d. cli. G., ix. 766, 1828.

% C. Loring Jackson and G. D. Moore. These Proceedings, xxiv. 268.

OF ARTS AND SCIENCES. 255

1, 2, 3, 4, 6, which leads to the constitution of our metabromtrinitro-
toluol given above.

Anilidotrimtrotoluol, C6HCH3(CGH3NH) (N02)3.

This substance was prepared by treating metabromtrinitrotoluol
with aniline in the proportion of two molecules of the base to one of
the nitro compound ; the action is violent, accompanied with consider-
able evolution of heat, and the product was easily purified by crystal-
lization from a mixture of alcohol and benzol, until it showed the
constant melting point 151°, when it was dried at 120°, and analyzed
with the following results : —

I. 0.2804 gr. of the substance gave on combustion 0.5031 gr. of

carbonic dioxide and 0.0936 gr. of water.
II. 0.2777 gr. of the substance gave 43.2 c.c. of nitrogen at a tem-
perature of 27°. 5 and a pressure of 772 mm.

Found.

II.

Carbon

Calculated for
C6HCH3(CcHflNH)(N02)3.

49.06

i.

49.03

Hydrogen

Nitrogen

3.14

17.61

3.71

17.41

Properties. — The metanilidotrinitrotoluol crystallizes from a mix-
ture of alcohol and benzol in well developed shining yellow plate's,
which melt at 151°. The substance is insoluble in water or ligroine ;
sparingly soluble in ethyl or methyl alcohol ; soluble in ether, chloro-
form, benzol, carbonic disulphide, glacial acetic acid, or acetone. The
most convenient solvent for it is a mixture of five parts of alcohol with
one of benzol. Aqueous sodic hydrate dissolves it with a red color,
and hydrochloric acid throws down from this solution the original sub-
stance apparently unaltered. Strong sulphuric or strong nitric acid
dissolves it, but strong hydrochloric acid does not.

The constitution of this substance is CH3, N02, C6HSNII, NO,,, N02,
1, 2, 3, 4, 6, as is shown by its preparation from the metabromtrinitro-
toluol, the constitution of which has been determined, as given earlier
in this paper.

256 PROCEEDINGS OF THE AMERICAN ACADEMY

XXI.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF

HARVARD COLLEGE.

ON THE ACTION OF SODIUM MALONIC ESTER UPON
TRIBROMTRINITROBENZOL.

By C. Loring Jackson and George Dunning Moore.

Presented May 28, 1889.

In a previous paper * one of us with J. F. Wing described the prepa-
ration of tribromtrinitrohenzol, and announced that its action with
sodium malonic ester would be studied. This work promised to be
of especial interest, because we hoped that each of the bromine atoms
would be replaced by the radical CH(COOC2EL)2, and that by the
reduction of the substance thus formed a compound might be obtained
consisting of three pyrrol molecules united to form a central benzol
ring, a sort of triple indol. Our first experiments, however, showed
that the reaction did not run in the way we had expected, but instead
of the removal of all three of the bromine atoms only two were re-
placed, one by the radical CH(COOC2H5)2, the other by hydrogen
giving a product with the following formula,

C6HBr(N02)3CH(COOC2H5)2,

that is, bromtrinitropheuylmalonic ester. After we had established
the composition of this substance, we decided that it was unwise to
undertake a complete study of this trinitro compound, which can be
obtained only with a very considerable outlay of time and material,
when the corresponding dinitro compound can be made much more
easily, and resembles it closely in most respects. "We have accord-
ingly confined our work principally to those properties of the trinitro
componnd in which we have observed marked differences from the
corresponding ones of the dinitro body, and for a fuller discussion of
those properties which the two substances have in common would

* These Proceeding, xxiii. 138.

OF ARTS AND SCIENCES. 257

refer the reader to a paper " On the Action of Sodium Malonic Ester
on Tribromdiuitrobenzol " by one of us and W. S. Robinson.*

The results described in this paper can be briefly stated as follows.
Sodium malonic ester forms in the cold with tribromtrinitrobenzol
bromtrinitrophenylmalonic ester,

C6HBr(N02)3CH(COOC2H.)2, melting point 104°-105°,

which has acid properties forming salts even with aqueous solutions
of the alkaline carbonates. Of these the red soluble sodium salt has
been studied, and its analysis led to the formula

C6HBr(N02)3CNa(COOC2H5)2.

The yellow insoluble copper salt, on the other hand, gave no con-
stant result on analysis, but on several occasions, in trying to make
it from a solution of cupric chloride in alcohol and the sodium salt, a
crystalline substance free from copper was obtained melting in the
neighborhood of 75°. Unfortunately the end of the term has pre-
vented us from studying this substance, which is the more interesting
because no similar compound has been obtained from the correspond-
ing dinitro body.

The reactions by which the bromtrinitrophenylmalonic ester is
derived from tribromtrinitrobenzol have been made out as follows: —

C6Br3(NO.,)3 + 3 CIINa(COOC2H3)2 =

NaBr + CGBr.,(N02)3CH(COOC2H5)2 + 2 CHNa(COOC2H.)2 =
CGBr2(N02)3CNa(c60C2H5)., + CH2(COOC2H5)2

' + CHNa(COOC2rL)2 + NaBr =
C6HBr(N02)3CNa(COOC2IL)2 + CIIBr(COOC2H5)2

" + CHNa(COOC2H5)2 + NaBr =
C6HBr(N02)3CNa(COOC2H5)2 + 2NaBr + C2H2(COOC2II,)4.

The acetylentetracarbonic ester formed according to the last re-
action was obtained from the oily secondary product of the action by
distillation under diminished pressure, and identified by its melting
point and analysis.

Perhaps the most striking difference between the dinitro and trini-
tro compounds consisted in the fact that the trinitro ester, or its salts,
when heated with an excess of common strong nitric acid, turned
bright blood-red, whereas no such action could be obtained from the
dinitro compound. The red product, on crystallization from alcohol,

* These Proceeding, xxiv. 1.
vol. xxiv. (s. s. xvi ) 17

258 PROCEEDINGS OP THE AMERICAN ACADEMY

was converted into colorless crystals melting at 125°, and as they
melted becoming blood-red and increasing very much in volume.
The study of this curious substance is still unfinished : we have only
established the fact that it is an ester. If the action of the nitric acid
is long continued, another body is formed melting at 156° to a color-
less liquid, and dissolving in aqueous sodic hydrate with a red color.

Sulphuric acid of specific gravity 1.44 converts the ester into the
metabromtrinitrotoluol melting at 143°-144°, recently discovered in
this laboratory by Bentley and Warren.

The trinitrophenylendimalonic ester (melting point 123°),

C6H(N02)3[CH(COOC2H5)2]2,

was also obtained by the further action of sodium malonic ester on
bromtrinitrophenylmalonic ester ; strangely enough, it has less marked
acid properties than the bromine compound, from which it is derived.

Preparation of Bromtrinitrophenylmalonic Ester.

The tribromtrinitrobenzol used for this purpose was prepared
according to the method already given by one of us and J. F. Wing;*
we have found, however, that if the proportion of fuming sulphuric
acid is increased, a better yield is obtained. The proportions finally
used were, 20 grs. of tribromdinitrobenzol, 500 c.c. of the nitric acid
of 1.52 specific gravity, and 200 c.c. of fuming sulphuric acid, instead
of one third the volume of the nitric acid as previously recommended.
The yield obtained from the new proportions was in the neighborhood
of 40 per cent of the theory, running in one case as high as 45 per
cent, whereas the proportions recommended by one of us and Wing
gave on the average from 15 to 20 per cent, and only in a single
instance ran as high as 40 per cent.

To convert the tribromtrinitrobenzol into bromtrinitrophenylmalonic
ester, one molecule of it must be treated with about three molecules
of sodium malonic ester. In practice we found it convenient to pro-
ceed as follows. 10 grs. of tribromtrinitrobenzol were dissolved in
about 200 c.c. of benzol with the aid of heat, mixed, while the solu-
tion was still moderately warm, with 10.6 grs. of malonic ester pre-
viously converted into the sodium compound by treatment with the
sodic ethylate from 1.7 grs. of sodium (a slight excess over the calcu-
lated amount) and about 15 c.c. of absolute alcohol, and the mixture

* These Proceedings, xxiii. 139.

OP ARTS AND SCIENCES. 259

allowed to stand 40 to 60 hours in a corked flask at ordinary tem-
peratures. As soon as the sodium malonic ester was added, the
liquid became dark blood-red, and on standing this color gradually
increased in intensity, while at the same time a precipitate of sodic
bromide was thrown down. The product of the reaction was mixed
with about three quarters of a litre of water, and acidified with dilute
sulphuric acid,* which decomposed the red salt, setting free the ester.
Ether was then added, and, after shaking thoroughly, the ethereal and
benzol solution separated from the aqueous liquid, which was ex-
tracted once more with ether. On distilling off the ether and benzol
from the extract, a dark oily residue was left, which was mixed with
a little alcohol, when, upon stirring, it solidified to a mass of prismatic
crystals. These were sucked out on the pump, washed with a little
cold alcohol to remove the adhering oil, and purified by crystallization
from hot alcohol, till they showed the constant melting point 104°-
105°. The oil which was sucked out from the crystals, or removed
from them by alcohol, upon standing, deposited an additional amount
of the substance, which was purified in the same way as the main
portion. The substance, after being dried in vacuo, was analyzed
with the following results : —

I. 0.2235 gr. of the substance gave on combustion 0.28G0 gr. of
carbonic dioxide, and 0.0629 gr. of water.
II. 0.1930 gr. of the substance gave 16.2 c.c. of nitrogen at a tem-
perature of 21°, and under a pressure of 775.9 mm.

III. 0.2548 gr. gave 20.1 c.c. of nitrogen at 18°.5, and 784 mm.

pressure.

IV. 0.2080 gr. gave by the method of Carius 0.0870 gr. of argentic

bromide.
V. 0.2512 gr. gave 0.1034 gr. of argentic bromide.

Calculated for Found.

CfiHBr(N02)3CH(C00C2H5)2. I. II. III. IV. V.

Carbon 34.67 34.89

Hydrogen 2.67 3.13

Nitrogen 9.33 9.75 9.38

Bromine 17.78 17.80 17.52

The yield was good, when compared to that obtained in similar
preparations from other substances ; the best result was as follows :

* Dilute nitric acid, which was used in some of the earlier preparations,
seemed to diminish the yield.

260 PROCEEDINGS OP THE AMERICAN ACADEMY

10 grs. of tribromtrinitrobenzol, treated as described above with 10.6
grs. of nialonic ester, gave 6.4 grs. of bromtrinitropbenyloialonic ester.
The amount required by theory, if all the tribromtriuitrobenzol had
been converted into bromtrinitrophenyl malonic ester, is 10 grs. ; there-
fore the yield is 64 per cent of the theoretical. The average yield
was between 50 and 60 per cent of the theoretical.

Properties. — The bromtrinitrophenylmalonic ester crystallizes in
white slender needles arranged in radiating groups, which, when ex-
amined with the microscope, are seen to be flat needles, or long plates
terminated by one plane, or, as commonly, by two at an obtuse angle
to each other. The crystals seem to belong to the monoclinic system.
The melting point is 104° to 105°, and if heated to 160° the substance
becomes dark red, and gives off gas. It is essentially insoluble in
cold water, very slightly soluble in hot ; insoluble in ligroine ; slightly
soluble in ether ; not very soluble in cold, freely in hot ethyl or
methyl alcohol ; freely in benzol, carbonic disulphide, or glacial acetic
acid ; very freely in chloroform or acetone. Hot alcohol is the best
solvent for it. Strong hydrochloric acid has no action upon it, even
when the substances are warmed together in open vessels ; it is prob-
able, however, that in sealed tubes the same decomposition would take
place as that observed with the corresponding dinitro body. Strong
sulphuric acid dissolves a little in the cold, more when hot, forming a
colorless solution. Strong nitric acid has little or no action in the
cold, but, if warmed with it, converts it into an intensely red viscous
substance swimming in the red acid liquid, which by further action of
strong nitric acid becomes solid and crystalline. A fuller discussion
of the action of strong nitric acid and that of dilute sulphuric acid will
be found later in this paper.

As was to be expected from the position of one of its hydrogen
atoms on a carbon surrounded by two carboxylester radicals and a
trinitrophenyl group, the substance possesses marked acid properties.
Sodic hydrate in excess gives only a pale red color with the solid
ester, owing to the very slight solubility of the sodium salt in sodic
hydrate, but upon adding water the coloration increases, and the addi-
tion of a few drops of alcohol produces a very dark red solution.
Potassic carbonate in aqueous solution gives a slight red color, on the
addition of a little alcohol a dark red solution ; acid sodic carbonate
acts in much the same way. but the color produced by the aqueous
solution is paler than that given with potassic carbonate. Amnionic
hydrate gives a red color at once, but this cannot be obtained free
from ammonia by evaporation on the water bath, or by using an excess

OF ARTS AND SCIENCES. 261

of the ester. The behavior of such a solution made by using an ex-
cess of the ester with amnionic hydrate was studied nevertheless, and
the following are the more characteristic precipitates which were
obtained : —

Magnesium salt, pale rust-colored.

Calcium salt, pale red flocks.

Strontium salt, like that obtained from calcium, but not so heavy a
precipitate.

Barium salt, an even less heavy precipitate than that obtained with
strontium, also less flocculent.

Manganese salt, yellowish brown.

Zinc salt, pale red.

Cobalt or Nickel salts, yellowish.

Ferric salt, rust-colored.

Capric salt, a rusty red to orange.

Mercuric chloride, yellowish brown.

Mercuric nitrate, rust-colored.

Mercurous salt, rusty precipitate mixed with the black product from
the excess of amnionic hydrate.

Cadmium salt, yellowish red.

Lead salt, brilliant rust-color.

Silver salt, vivid brown (" Bismarck brown").

The most characteristic point in its behavior with reagents is that
the magnesium and calcium salts are less soluble than the strontium
and barium salts, the order of solubility being magnesium and calcium
least soluble, strontium more soluble, barium the most soluble. A
similar observation has been made by Bischoff * in regard to the salts
of orthonitrobenzoylmalonic ester, and the bromdinitrophenylacetacetic
ester also exhibits the same peculiarity, as we mention in detail in the
next paper of this series.

Salts of Bromtrinitmphenylmalonic Ester.

"We had intended at first to make a ra;her thorough study of the
salts of the bromtrinitrophenylmalonic ester, but after a few experi-
ments became convinced that this would be a waste of time, as, owing
to the ease with which the atom of bromine can be removed, no very
satisfactory analytical results could be obtained ; and we were the
more ready to give up this part of the work, because our investigation
of the secondary oily product of the reaction, by which the bromtrini-

* Ann. Chem., ccli. 3G2.

262 PROCEEDINGS OF THE AMERICAN ACADEMY

trophenylmalonic ester is formed, had settled beyond a doubt the com-
position of the ester, which before this and the similar investigation
made by one of us and W. S. Robinson had rested principally on
the analyses of its salts (see these Proceedings, xxiv. 4). We have
confined ourselves therefore to a single analysis of the sodium salt,
and a preliminary study of the copper salt, the results of which are
given below.

Sodium Salt, C6HBr(N02)3CNa(COOC2H5)2. — This substance was
made in two ways. First, by digesting solid pure sodic carbonate
with a solution of bromtrinitrophenylmalonic ester in absolute alcohol.
The red solution was filtered from the excess of sodic carbonate,
evaporated to dryness on the water bath, and dried at 100°, when it
gave the following result on analysis : —

0.2373 gr. of the salt gave after being heated with sulphuric acid
0.0418 gr. of sodic sulphate.

Calculated for
C6HBr(N02)3CNa(C02C2H5)a. Found.

Sodium 4.87 5.71

The bad result is probably due to a slight excess of sodic carbonate,
which dissolved in the alcohol. The salt looked black and somewhat
decomposed.

The second and better method was that used for making the salt of
the corresponding dinitro compound, that is, by the action of an alco-
holic solution of sodic hydrate or ethylate on an alcoholic solution of
the ester, taking care that the ester is in excess. For the necessary
precautions see these Proceedings, xxiv. 7. No analysis was made
of this preparation, as from our experience with the acetacetic com-
pound we were sure that no accurate analytical results would be
obtained.

Properties. — The sodium salt forms an amorphous blackish red
mass, soluble in ethyl or methyl alcohol, water, or acetone ; tolerably
soluble in ether; slightly in chloroform ; insoluble in benzol or ligroine.
All the solutions have a deep blood-red color. When the salt is treated
with an excess of strong nitric acid (of specific gravity 1.36), it is at
first decolorized, but almost immediately turns vivid red owing to the
formation of the substance produced by warming the free ester with
nitric acid. This behavior is characteristic, as it appears with none
of the similar substances which we have studied.

Copper Salt. — We took up the study of this salt in the hope of
throwing light on the composition of the ester, but after analyzing

OF ARTS AND SCIENCES. 263

several samples were convinced that the precipitate had a varying
composition, and therefore describe our work only because of a curious
observation made in the course of it. The salt was made by adding a
solution of cupric sulphate to an alcoholic solution of the sodium salt,
which need not be free from sodic hydrate ; the copper salt was ex-
tracted with ether, and obtained on evaporating the ethereal eolutiou
as an orange mass, which when heated exploded with a blue flame.
It was insoluble in water, nearly insoluble in cold alcohol, soluble in
hot, and the yellow solution deposited the salt in rhombic crystals ; very
soluble in benzol or chloroform ; insoluble in ligroine.

AVe next substituted an alcoholic solution of cupric chloride for the
cupric sulphate, in hopes of getting a better result, when to our sur-
prise a product was obtained crystallizing from alcohol in long white
prisms, and melting in the crude state at 75°. This product was ob-
tained more than once, but as frequent crystallization was necessary
to purify it, we did not at first have enough to bring it into a state fit
for analysis ; and, on returning to the subject after some mouths, we
obtained under the same conditions nothing but the orange explosive
copper salt. .Unfortunately, we had postponed work on this subject
till so near the end of the term that we were unable to give it the
careful study it seems to deserve, but its investigation will be continued
in this laboratory next year.

Study of the Reactions by which Bromtrinitrophenylmalonic Ester

is formed.

The reactions by which the bromtrinitrophenylmalonic ester is
formed from the tribromtrinitrobenzol must consist in the replace-
ment of one of the atoms of bromine by the nialonic ester radical
CH(COOC2IL)2, and of another by hydrogen. The first of these pro-
cesses needs no explanation, but the mechanism of the second, the
replacement of the bromine by hydrogen, could be made out only by
experiment. Obviously, the first point to be settled was the form in
which the bromine was eliminated ; that is, whether as sodic bromide
alone, or partly as sodic bromide and partly as an organic com-
pound. For this purpose the following quantitative determinations
were made : —

I. 10 grs. of tribromtrinitrobenzol, treated with the sodium malonic
ester from 15 grs. of malonic ester, gave after standing two
davs and a half 8.83 grs. of argentic bromide, corresponding to
3.76 grs. of bromine.

264 PROCEEDINGS OF THE AMERICAN ACADEMY

II. The same weights under the same conditions gave 9.08 grs. of
argentic bromide, corresponding to o.SG grs. of bromine.

If two atoms of bromine had been removed, the amount of bromine
should have been 3.55 grs. Therefore the percentages of the theo-
retical bromine removed as sodic bromide were, —

I. ii.

105.9 108.8

It appears, therefore, that a little more than the amount calculated
for two atoms of bromine has been removed as sodic bromide, and this
is easily explained by the observation described later in this paper,
that sodium malonic ester can act on the bromtrinitrophenylmalonic
ester even in the cold to form the trinitrophenylendimalonic ester, a
small quantity of which was undoubtedly formed in these two experi-
ments by the large excess of sodium malonic ester present. At any
rate, there can be no question that all the bromine was removed in the
form of sodic bromide.

The next step consisted in determining the nature of the organic
secondary product. For this purpose, the oil, separated from the
bromtrinitrophenylmalonic ester by sucking out the crude product on
the pump and treatment with alcohol, was allowed to stand till it
ceased to deposit crystals of the ester, and then submitted to distilla-
tion under diminished pressure. The pressure varied from 22 to 25
mm., and a distillate began to appear when the thermometer inside
the flask stood at 98°, and was collected until the temperature had
reached 1G0°, when about one third of the total volume had passed
over. In this way a clear yellow liquid was obtained, which distilled
unaltered at ordinary pressure between 197° and 206°. We there-
fore inferred it was mostly malonic ester, boiling point 197°. 7, a view
of its nature which was confirmed by its smell. As the whole of the
distillate passed over below 206°, there could be no large amount ot
tartronic ester (boiling point 220°) present, which we had expected
after the work of one of us and W. S. Robinson on the corresponding
dinitro compound, and the secondary product of the reaction must be
looked for in the residue which had been left behind in the flask after
the distillation under diminished pressure. This was a thick blackish
brown oil, which on standing for about a week deposited crystals all
over its surface ; these were removed and allowed to stand on filter
paper until a large part of the oil had been sucked out, when they
were purified by washing with a small quantity of cold alcohol, and
then crystallizing from boiling alcohol, until they showed the constant

Carbon

Calculated for
C2H2( C02C2H6)4.

52.83

Hydrogen

6.92

OF ARTS AND SCIENCES. 265

melting point 76°. The substance formed long glittering colorless
needles or prisms, and contained no bromine ; we decided accordingly
that it was the acetylentetracarbonic ester (melting point 76°), and
this conclusion was confirmed by the following analysis : —

0.2270 gr. of the substance gave on combusion 0.4384 gr. of carbonic
dioxide, and 0.1506 gr. of water.

Found.

52.67
7.36

The amount of acetylentetracarbonic ester was so considerable, that
there can be no doubt it was a principal product of the reactions by
which the bromtrinkrophenylmalonic ester is formed, and these must
therefore be written thus : —

C6Br3(NO.v)3 + 3 CHNa(COOC2H5)2 =

C6Br,(XO,)3CH(COOC2H5)2 + NaBr + 2 CHNa(COOC2H5)2 =
C6Br2(N02)3CNa(COOC2Hs)2 + CH2(COOC2H.)2

+ CHNa(COOC,H5)9 + NaBr =
C6HBr(N02)3CNa(COOC2H5)2 + CHBr(COOC2H5)2

" + CHNa(COOC2H5)2 + NaBr =
C6HBr(N02)3CNa(COOC2H5)2 + C2H2(COOC2H5)4 + 2 NaBr.

The only objection which could be urged against this series of re-
actions is that the yield of bromtrinitrophenylmalonic ester is only
64 per cent of the theoretical, from which it might be argued that
nearly half of the tribromtrinitrobenzol may have undergone some
different transformation. This objection is disposed of, however, by
the fact that W. D. Bancroft and one of us * have succeeded in getting
a yield of 80 per cent of the dibromdinitrophenylmalonic ester from
tetrabromdinitrobenzol by a series of reactions exactly analogous to
those just given. The missing 36 per cent of the bromtrinitrophenyl-
malonic ester must therefore have remained dissolved in the oil, and
have been destroyed by the distillation even under the diminished
pressure used by us.

Action of Nitric Acid.

The intense red color produced by the action of nitric acid of spe-
cific gravity 1.36 on the sodium, or copper salt of the bromtrinitrophe-
nylmalonic ester in the cold, or on the ester itself at 100°, seemed to

* These Proceedings, xxiv. 295.

266 PROCEEDINGS OP THE AMERICAN ACADEMY

us of great interest, and we laid out what we thought would be time
enough for its investigation ; unfortunately, however, this was not the
case, since, after the work had been going on for some time, we found
that at least two substances were formed by this reaction, and conse-
quently the end of the term surprised us before we had reached any
definite results. We have decided, however, to publish here what
results we have obtained, as we are unable to go on with this work
together, but wish it to be understood that all these statements are
to be taken as preliminary.

If about 1 gr. of bromtrinitrophenylmalonic ester was mixed with
2-3 c.c. of strong nitric acid (specific gravity 1.36), no change took
place in the cold ; but if the mixture was warmed gently on the water
bath for less than five minutes, an intense vivid red color appeared
in both the acid liquid and the organic substance, which melted and
became converted into a viscous mass. If now the acid was poured
off, a fresh quantity added, and the gentle warming repeated, the vis-
cous drop became converted into a red crystalline mass, and by con-
tinuing* this treatment with successive portions of nitric acid the red
color could be removed partially, so that the product had a spotted red
and white appearance. This frequent treatment with nitric acid was,
however, unnecessary, as the red crystalline product after the second
warming with nitric acid gave, when crystallized from alcohol, well
formed white prisms, and a reddish mother liquor. The red nitric acid
poured off from the principal part of the product gave with water a
red precipitate, but a better mode of treatment seemed to be to evapo-
rate this red acid to dryness on the water bath. The residue, or pre-
cipitate obtained with water, was partly viscous, partly crystalline,
and by treating it again with warm strong nitric acid a new quantity
of the red crystalline substance was obtained, but the amount recov-
ered in this way was so small that it hardly paid for the trouble.
The white crystals, after purification by crystallization from alcohol,
showed the constant melting point 125°, and their behavior in melting
was very characteristic, as they turned from white to bright red, and
swelled to many times their original volume. They were not affected
by sodic hydrate in aqueous solution. Supposing that the substance
was homogeneous because of its constant melting point, we analyzed
it, but on studying its properties more carefully we began to doubt its
purity for the following reasons : — First. Although the microscopic
examination showed that the substance consisted principally of white,
short, rather thick monoclinic prisms, usually with both terminations
well developed and made up of two planes, there were mixed with

OP ARTS AND SCIENCES. 2G7

these longer prisms, and we could not decide whether these latter
were a different substance or merely a different habit of the same.
Second. When the crystals were treated with sodic hydrate and alco-
hol a little of a soluble red salt was formed around each crystal, but
we could not convert the whole of the crystals into this salt. This
seemed to point to the presence of an impurity, from which the salt
was formed. Third. We found almost at the very end of the term,
that by warming the bromtrinitrophenylmalonic ester with strong ni-
tric acid for three hours, instead of a few minutes, a red substance was
obtained, which crystallized from alcohol, became white, and melted
at 156° instead of 125°, and, what was as distinctive as the different
melting point, fused to a colorless liquid, and gave a red solution with
aqueous sodic hydrate. This substance was discovered so late that
we had no time to investigate it, but some of it seemed to be formed
even on shorter heating (15 minutes) with the nitric acid. For these
reasons, we have decided that it is wiser to postpone the publication
of our analyses of the substance melting at 125° until the work has
been repeated with samples in regard to the purity of which there
can be no doubt. We add such results of our work as are established
with certainty. The analyses made by us showed that there were
three atoms of nitrogen to one of bromine in the substance, and there-
fore the action of the nitric acid did not consist in the introduction
of another nitro, or nitroso group. The fact that it is insoluble in
aqueous sodic hydrate shows that it is not a free acid, and its action
with hydrochloric acid would indicate that it was an ester, as, when
heated to 135°-140° with this acid in a sealed tube for 36 hours, a
gas was given off burning with a green-bordered flame, and giving a
white precipitate with lime-water, which therefore must have contained
ethyl chloride and carbonic dioxide. The solid product of this action
was partly viscous and partly crystalline ; the latter melted in the
crude state above 180°. It has been stated already that the substance
melting at 125° turns red aud increases in volume when it melts; this
change, which takes place to a limited extent even when it is kept at
100° for some time, is accompanied by loss of weight, as a sample
kept at its melting point for some days lost at least 17 per cent, and
gave a residue consisting of two or more substances, one white and
crystalline, the other red and viscous.

268 PROCEEDINGS OP THE AMERICAN ACADEMY

Action of Sulphuric Acid. — Constitution of the Bromtrinitrophenyl-

malonic Ester.

The action of dilute sulphuric acid upon the bromtrinitrophenyl-
malonic ester was studied in the hope of obtaining the as yet un-
known bromtriuitrophenylacetic acid, as it had been found by one of
us and W, S. Robinson f that the corresponding dinitro compound was
decomposed in this way. For this purpose, about 2 grs. of the ester
were boiled in a flask under a return condenser with sulphuric acid of
specific gravity 1.44 and boiling point 132° until the ester had dis-
solved ; the liquid was then allowed to cool, when it deposited crystals,
which, after recrystallization from alcohol, were recognized by their
melting point 143°-144°, their appearance, and the absence of acid
properties, as the metabromtrinitrotoluol discovered in this laboratory
by Bentley and Warren.

It is evident therefore that the bromtrinitrophenylacetic acid is less
stable than the corresponding dinitro compound, as indeed was to be
expected, and was broken up as soon as formed into the substituted
toluol and carbonic dioxide according to the following reaction :

CGHBr(N02)3CH2COOH = C6HBr(N02)3CH3 + CO,.

The formation of this substance would settle the constitution of
the bromtrinitrophenylmalonic ester, if that were necessary, but
the preparation of the ester from symmetrical tribromtrinitrobenzol
leaves no doubt as to its constitution, which must be as follows,
CH(COOaH-).,, N02, Br, NO,, H, NO,, I, 2, 3, 4, 5, 6, and Bentley
and Warren have established the corresponding constitution for their
substituted toluol.

Trinitrophernjlendimalonic Ester, CcH(N02),[CH (COOC2H5),]2.

This substance was formed by the further action of sodium malonic
ester on bromtrinitrophenylmalonic ester. For this purpose, 1 gr.
of bromtrinitrophenylmalonic ester dissolved in ether was mixed with
1.5 grs. of malonic ester previously converted into sodium malonic
ester and dissolved in much absolute alcohol, and the mixture boiled
in a flask with a return condenser for an hour. The product, which
was very dark brown, almost black, was treated with water, acidified
with dilute sulphuric acid avoiding a large excess, the ether removed
with a drop funnel, and the aqueous liquid shaken out twice with
ether. The extract, after distilling off the ether and allowing it to

t These Proceedings, xxiv. 240.

OF ARTS AND SCIENCES. 269

get perfectly cool, was treated with a little alcohol, when crystals
separated on standing, which were purified by recrystallization from
alcohol, until they showed the constant melting point 123°. They
gave no test for bromine when heated on a copper wire, and, after
drying in vacuo, were analyzed with the following results : —

I. 0.3004 gr. of the substance gave on combustion 0.5043 gr. of

carbonic dioxide, and 0.1230 gr. of water.
II. 0.2392 gr. of the substance gave 1 6.8 c.c. of nitrogen at a tem-
perature of 25° and a pressure of 768 mm.

Calculated for Found.

CeHlNO^^CIKCOOC.IIs).,]^ I. II.

Garbon 45.37 45.78

Hydrogen 4.35 4.55

Nitrogen 7.94 7.92

A small quantity of this substance can also be formed in the cold,
and more than once some of it has been obtained in making the
bromtrinitrophenylmalonic ester by the process described earlier in
this paper (compare pages 258 and 264).

Properties. — The trinitrophenylendimalonic ester crystallizes from
alcohol in long colorless prisms, terminated by a single rhombic plane
at a tolerably sharp angle to the sides ; less frequently, the termination
consists of two planes at an obtuse angle to each other, so that the
general effect is as if the prisms had rounded ends. They are fre-
quently grouped, or twinned, parallel to their long axes, so that often
there are two or more terminations at one end of a group. The sub-
stance melts at 123°; and is very nearly, if not quite, insoluble in
water, whether cold or boiling; essentially insoluble in ligroine ; not
very soluble in cold ethyl or methyl alcohol, freely in either of these
solvents when hot ; very slightly soluble in carbonic disulphide ;
soluble in ether or glacial acetic acid ; and freely soluble in benzol,
chloroform, or acetone. Its acid properties are not so strongly de-
veloped as we expected. An aqueous solution of soilic hydrate
turns the solid pale red, but does not dissolve it to any extent ; if,
however, sodic hydrate is added to its alcoholic solution, it at once
takes on a dark brownish red color, much browner than any of the
similar salts which we have studied. — in fact it would be possible to
recognize the salt by this color. An aqueous solution of potassic car-
bonate gives a very faint red color with it, which is increased, but
not to a great extent, by adding alcohol to the aqueous solution. The
action was incomplete at best ; acid sodic carbonate had no action on

270 PROCEEDINGS OP THE AMERICAN ACADEMY

it in aqueous solution, but on the addition of alcohol a barely per-
ceptible red color appeared. Amnionic hydrate even in dilute aqueous
solution dissolved it easily with a brown color ; the solution turned
blacker when the attempt was made to drive off the excess of ammo-
nia on the water bath, and a white scum formed on the surface,
probably the original substance. The behavior of this solution was
studied with some of the commoner reagents ; but, as it showed such
evident signs of decomposition, we did not think it worth while to
extend this work to salts of all the basic radicals.

Ferric salt, cupric salt, silver salt, and lead salt all gave brown
ilocculent precipitates.

Calcium salt gave no precipitate.

Barium salt, a very slight dirty brown precipitate, but none if the
solution had not been warmed on the water bath in the vain attempt
to drive off the excess of ammonia.

Strong sulphuric acid dissolved the dimalonic ester, giving a color-
less solution. Strong hydrochloric acid had no action upon it, whether
hot or cold. Strong nitric acid dissolved it partially in the cold,
giving a yellowish solution, which when warmed became darker
yellow ; but if the warming on the water bath was continued for some
time, the acid liquid became red, and a red viscous substance was
also obtained, which, after washing with water and crystallization from
alcohol, was converted into yellow plates melting at 104°-105° in the
crude state. It is evident that the action here is similar to that of
bromtrinitrophenylmalonic ester with strong nitric acid, but the end
of the term has prevented us from continuing the study of this sub-
stance at present.

OF ARTS AND SCIENCES. 271

XXII.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF

HARVARD COLLEGE.

ON THE ACTION OF SODIUM ACETACETIC ESTER
UPON TRIBROMDINITROBENZOL.

By C. Loring Jackson asd George Dunning Moore.

Presented May 28, 1889.

After the action of sodium malonic ester ou tribromtrinitrobenzol
and on tribromdinitrobenzol had been studied, it seemed of interest to
determine whether sodium acetacetic ester acted in the same way, and
the following paper contains the results of experiments undertaken
with this intention upon the tribromdinitrobenzol, which was selected
because it can be prepared so much more easily than the correspond-
ing trinitro compound. These results can be summarized briefly as
follows. Tribromdinitrobenzol behaves with sodium acetacetic ester
in the same way that it does with sodium malonic ester ; that is, one
atom of its bromine is replaced by the acetacetic radical

CH3COCHCOOC2H5,

and a second by hydrogen, while the third remains unaltered ; so that
the product of the action is the bromdinitrophenylacetacetic ester,

C6H2Br(N02)3CH3COCIICOOC8Ha,

or to speak more accurately its sodium salt,

CTI2Br(NO,)2CH3COCNaCOOC2H5,

The bromdinitrophenylacetacetic ester melts at 96°, and possesses
marked acid properties, forming the sodium salt whose formula has
just been given, even with acid sodic carbonate in aqueous solution.
The salt has a red color, and is soluble in water ; in fact, it resem-
bles the sodium salt of the corresponding malonic compound most
closely.

272 PROCEEDINGS OP THE AMERICAN ACADEMY

The broradiuitrophenylacetacetic ester is saponified and decomposed
by heating with sulphuric acid of specific gravity 1.44, giving the
bromdinitrobenzylmethylketone,

C6H2Br(N02)2CH2COCH3,

but no trace of the corresponding acetic acid. This ketone melts at
112°— 1 13°, and also possesses acid properties, giving a purplish red
salt soluble in alcohol, but decomposed almost completely by water.
The acid properties of the ketone are less marked than those of the
acetacetic ester, as was to be expected, since the hydrogen, which is
replaced by basic radicals, is subject to the influence of an acetyl, a
carboxylester, and a dinitro phenyl group in the acetacetic compound,
only to those of an acetyl and a dinitro phenyl group in the ketone.

The bromine of the ketone can be replaced easily by the aniline
radical C6H.NH, forming anilidodinitrobenzylmethylketone,

C6H2(C6H5NH)(N02)2CH2COCH8,

which melts at 131°, and has not lost all acid properties, although
they have been much weakened by replacing the bromine atom by the
basic radical CCH.NH. It cannot form an ammonium salt, but the
sodium salt can be easily obtained in alcoholic solution, and on analysis
gave a number corresponding to the formula

CJIJ(CtiH5NH)(NO,),CHNaCOCH3.

It is completely decomposed by water, but dissolves in alcohol with a
brown color.

Both these ketones, therefore, show stronger acid properties than
desoxybenzoine C6H.COCH2C6H5, the metallic compounds of which,
according to Victor Meyer,* could not be isolated ; we are inclined to
ascribe this to the presence of the nitro groups in the phenyl, which
would heighten its acid-producing power, but it may also be due in part
to the fact that those ketones contain the acetyl group, which, as
Claisen and Ehrhardtf have pointed out, has a greater influence in
producing acidity than the benzoyl radical contained in desoxy-
benzoine.

The hydrazone of the anilirlodinitrobenzylmethylketone,

C6II.2(C6tI5NII)(N02)2CII2C(NNHC6H5)CH3,

was also prepared, and melted at 140°.

The full details of the work will be found in the remainder of the
paper.

* Ber. d. ch. G. 1888, p. 1291. t Ber. d. ch. G. 1889, p. 1019.

OF ARTS AND SCIENCES. 273

Preparation of Tribromdinitrobenzol.

The method used by us for preparing tribromdiuitrobenzol differs
from that formerly, in use only in certaiu details, but, as careful atten-
tion to these details insures a purer product at much less expense of
time and labor, we have thought it best to give a full account of our
mode of procedure.

To make the tribromaniliue, GO grs. of aniline were dissolved in
dilute hydrochloric acid, and, the solution having been made up to a
volume of about 3 litres, a rapid stream of air saturated with bromine
vapor was sucked through it by means of a Bunsen pump, until the
liquid assumed a distinct yellow color. The precipitate of tribrom-
aniliue was then removed by straining through cheese-cloth, and
washed with a stream of common water, until free from acid, when
the greater part of the water was squeezed out with a screw press, and
the product thoroughly dried on a steam radiator. In this way a
quantitative yield of the tribromaniliue was obtained, and it was free
from colored by-products.

In order to convert the tribromaniline into tribrombenzol, 50 grs. of
it, after being pulverized, were dissolved with the aid of heat in about
300 c.c. of common alcohol, and a concentrated aqueous solution of
21 grs. of sodic nitrite poured in slowly, but not in successive portions.
The hot mixture was then acidified with dilute sulphuric acid, and
allowed to stand over night, when it was filtered, and the precipitate
washed with hot water to remove the sodic sulphate and leave the
tribrombenzol. An additional quantity of this was obtained by con-
centrating the alcoholic filtrate, when it separated as an oil, that solidi-
fied on standing, and was then crystallized from alcohol. The yield
was nearly quantitative, as 45.50 grs. of tribrombenzol were obtained
instead of the 47.77 required by the theory, that is 95 per cent.
When made by this method, the tribrombenzol was usually pure
enough to be nitrired directly in spite of its brownish color.*

* The amount of sodic nitrite used in the process described above (two mole-
cules of nitrite to one of the base) is twice that required by the theory, but we
have found that this larg-e excess was necessary to bring all the tribromaniline
into the reaction. When a smaller amount of nitrite was used, the product was
much less pure, as shown by its lower melting point ; it was necessary in this
case to distil it from a little retort, and crystallize several times from alcohol, to
obtain pure tribrombenzol. The crystals obtained on evaporating the alcoholic
mother liquors, which made up about two thirds of the entire amount, were dis-
tilled with steam, pushing the distillation as rapidly as possible, when tribrom-
benzol passed over, and tribromaniline was left in the retort ; but all this tedious
purification can be avoided by using the excess of nitrite recommended above,
vor.. xxiv. (y. s. xvi.) 18

274 PROCEEDINGS OP THE AMERICAN ACADEMY

To convert the tribrombenzol into the dinitro compound, 25 grs. of
it were added to 100-120 grs. of nitric acid* of specific gravity 1.52,
(made from potassic nitrate and sulphuric acid in the laboratory.)
warmed gently until the solid had dissolved, and allowed to stand
about two hours, when tbe tribromdinitrobenzol crystallized out in
large white prisms. The yield was essentially quantitative.

Bromdinitroph enylacetacetic Ester,
C6H2Br(N02)2CH3COCHCOOC.2Hs.

Preparation. — This substance was made by the action of an alco-
holic solution of sodium acetacetic ester on a benzol solution of the
tribromdinitrobenzol in the proportion of about four molecules of the
former to one of the latter. For this purpose 15 grs. of tribromdini-
trobenzol were dissolved in about 200 c.c. of benzol, and mixed with
20 grs. of acetacetic ester previously treated with 3 grs. of sodium
dissolved in about 20 c.c. of absolute alcohol. At first there was very
little change of color, but on standing at ordinary temperatures the
liquid turned first yellow and then red (whereas with malonic ester
the red color appeared instantaneously). In order to complete the
reaction, the mixture was heated on the steam bath for about one
hour, at the end of which time it had become nearly black, and a con-
siderable precipitate of sodic bromide had formed. It was then mixed
with from one and a half to two litres of water, and the reddish benzol
solution, which was precipitated, removed from the dark red aqueous
liquid ; the latter was then acidified with dilute sulphuric acid,- taking
care to avoid a large excess ; this decomposed the red sodium salt of
the new substance, which was set free in the form of a yellowish oil,
and extracted by shaking the liquid with ether twice. On distilling
off the ether a dark reddish brown oil was left, which, when cold, was
treated with very little alcohol, and upon stirring became filled with
crystals, the quantity of which increased on standing for twelve or
more hours. When the quantity of the crystals did not increase
further, the pasty mass was sucked out on the pump, first adding a
little alcohol, if necessary, and by washing with a small quantity of
cold alcohol all the very soluble red oil was removed from the crys-
tals, which were nearly or quite insoluble in cold alcohol, and were
next purified by crystallization from hot alcohol until they showed the
constant melting point 96°. An additional quantity of the substance
can be obtained from the benzol solution by distilling off the benzol

* These Proceedings, xxii. 374.

OP ARTS AND SCIENCES. 275

and treating the residue with sodic hydrate and a little alcohol, which
convert the new substance into its sodium salt, and this can be re-
moved from the unaltered tribromdinitrobenzol by washing with
water ; when the wash-waters were acidified, and the precipitate puri-
fied in the same way as the maiu portion, the substance was dried in
vacuo, and analyzed with the following results : —

I. 0.2683 gr. of the substance gave on combustion 0.3788 gr. of
carbonic dioxide, and 0.0748 gr. of water.
II. 0.2512 gr. of the substance gave 16.7 c.c. of nitrogen at a tem-
perature of 22° and a pressure of 759.7 mm.

III. 0.2067 gr. of the substance gave, according to the method of

Carius, 0.1041 gr. of argentic bromide.

IV. 0.2141 gr. of the substance gave 0.1059 gr. of argentic bromide.

Calculated for Found.

C6H„Br(N02),CH3COOHCO,C,H5. I. II. III. IV.

Carbon 38.41 38.50

Hydrogen 2.93 3.09

Nitrogen 7.47 7.33

Bromine 21.33 21.43 21.05

The yield of bromdinitrophenylacetacetic ester, prepared by the
method described above, is satisfactory on the whole, the best result
being as follows : —

15 grs. of the tribromdinitrobenzol yielded 6.5 grs. of the ester, and
3.1 grs. of unaltered tribromdinitrobenzol were recovered, leaving 11.9
grs. which had entered into the reaction. 11.9 grs. of tribromdinitro-
benzol should yield 11.0 grs- of bromdinitrophenylacetacetic ester.
The actual yield, therefore, was 59 per cent of the theoretical. In
other preparations the yield varied from 54 to 34 per cent, the lowest
which was obtained from this process. The yield was not improved
by increasing or by diminishing the proportion of acetacetic ester to
the tribromdinitrobenzol. In our first preparations the reaction was
allowed to run in the cold, as an experiment, in which the mixture had
been heated less than one hour, seemed to give a very unmanageable
product; in this case, however, the yield was very much less than that
obtained by the process described above, in which the mixture was
heated one hour.

In order to throw light on the way in which the reaction runs, we
determined the amount of bromide of sodium formed in two prepara-
tions as follows : —

I. 12 grs. of tribromdinitrobenzol yielded 4.43 grs. of bromine in
the form of sodic bromide.

27G PROCEEDINGS OP THE AMERICAN ACADEMY

II. 12 grs. of tribromdinitrobenzol yielded 4.40 grs. of bromine in
tbe form of sodic bromide.
If two atoms of bromine are removed from each molecule by this re-
action, 12 grs. of tribromdinitrobenzol would lose 4.74 grs. of bromine.
Percentage of bromine removed as sodic bromide, —
i. ii.

93.5 93.0

From this it appears that all the bromine removed from the tribrom-
dinitrobenzol is finally in the state of sodic bromide, as the sligbt loss
of 7 to 6.5 per cent is not greater than wonld be expected when it is
considered that the aqueous solution containing the sodic bromide was
shaken out twice with ether. These determinations show also that in
these cases the whole of the tribromdinitrobenzol entered into the re-
action, and in fact none of it could be found unaltered in the products
of these two preparations ; but this was not always the case, as, for in-
stance, in the preparation described on page 275.

The full explanation of the reaction, however, was to be found in
the study of its secondary product, the red oil removed from the crys-
tals by washing with cold alcohol. This, upon long standing exposed
to the air, threw down a few more crystals of bromdinitrophenyl-
acetacetic ester, which would raise the yield a little above that given,
but the amount obtained in this way was very small. The oil, after
no more crystals could be obtained from it, was distilled under di-
minished pressure and yielded some acetacetic ester, leaving in the flask
a tarry mass which has shown no signs of giving crystals even after
long standing. This part of the work therefore was not carried fur-
ther, as from the formula of the ester and the removal of all the bro-
mine as sodic bromide there can be no doubt that the reaction with
acetacetic ester runs in the same way as that with malonic ester, and in
this latter case the reaction has been fully explained.* Reasoning on
this analogy, the following reactions may be taken as expressing what
takes place in the formation of the bromdinitrophenylacetacetic ester.

C6HBr3(N02)2 + 2 CH3COCHNaCOOC2H. =
' C6HBr2(Na),CH„COCHCOOC2H5 + NaBr

+ CH,COCHNaCOOC2H5 =
CsIIBr„(N02)2CII3COCNaCOOC2H5 + CH3COCII2COOC2H5

+ NaBr =
C6n2Br(N02)2CII!COCNaCOOC2II5 + CH,COCIIBrCOOC2H5

+ NaBr.

* These Proceedings, xxiv. 238, 265.

OF ARTS AND SCIENCES. ' 277

The broraacetacetic ester being afterward attacked either by a
molecule of sodium acetacetic ester or by sodic hydrate.

The only question with regard to these reactions, which might arise
would be due to the fact that only a little over 50 per cent (59 per
cent) of the ester was obtained, which might lead to the hypothesis
that nearly half of the tribromdinitrobeuzol went to form another
substance ; but this objection is removed by the fact that in another
analogous case as much as 80 per cent of a corresponding ester has
been found. We consider, therefore, that the missing 41 per cent of
the ester remained dissolved in the oily secondary product, but should
state, on the other hand, that upon treating the oil with sulphuric acid
of specific gravity 1.44 in hopes of decomposing the secondary products
and converting the bromdinitrophenylacetacetic ester into bromdinitro-
benzylmethylketone, we obtained nothing but black tarry substances,
from which no substance fit for analysis could be obtained.

Properties. — The bromdinitrophenylacetacetic ester crystallizes
well, usually in forms looking like a rhombohedron with a very acute
angle, less frequently in prismatic forms, which are twinned on a cen-
tral line parallel to their longer axis, the termination being in shape
either like a simple gable, or one with a notch in its point. The crys-
tals show a great tendency to twin, or group, often in forms like
those of frost, and are sometimes from 2 to 3 mm. long. The sub-
stance has a yellow color of about the shade of potassic chromate,
which is characteristic, as the corresponding malonic compound is
nearly white. Heckmann * has observed a similar deep yellow color
in the orthoparadinitrophenylacetacetic ester. It melts at 9G° ; is very
slightly soluble in water, rather more so in hot than in cold ; essen-
tially insoluble in ligroine ; not very soluble in cold alcohol, but freely
in hot ; if anything less soluble in methyl than in common alcohol both
cold and hot, but the solubility is much greater in the hot methyl alco-
hol than in the same solvent when cold ; tolerably soluble in carbonic
disulphide, or ether; soluble in glacial acetic acid; and freely in ben-
zol, chloroform, or acetone ; from ether, benzol, or chloroform it is
deposited as an oil. The best solvent for it is hot alcohol. Strong
sulphuric acid does not act upon it in the cold, but when warm dia
solves it, forming a slightly yellowish solution. Strong hydrochloric
acid has no action with it in open vessels, even if heated. The de-
composition of the substance by these two acids under proper condi-
tions is described later in this paper. Strong nitric acid has no action

* Ann < hem., ccxx. 133.

278 PROCEEDINGS OF THE AMERICAN ACADEMY

on it in the cold, but dissolves it when warm ; there is, however, no
marked change of color in which respect this substance shows a strik-
ing difference from the bromtrinitrophenylmalonic ester, but resembles
the broindinitrophenylmalonic ester.

The bromdinitrophenylacetacetic ester shows marked acid proper-
ties, as was to be expected from the position of one of its hydrogen
atoms upon a carbon adjacent to an acetyl, a enrboxylester. and a
phenyl group, the hater rendered still more efficient by the presence
of two oitro groups, and also from the acid properties of the corre-
sponding malonic compound. Aqueous sodic, or amnionic hydrate dis-
solves i: forming a red solution of the corresponding salt ; an aqueous
solution of potassie carbonate acts upou it in the same way giving,
however, a somewhat fainter color, but, if alcohol is added to the
solution, it turns at once dark red; acid sodic carbonate in aqueous
solution gives a very faint red color which is much intensified by
addition of alcohol to the solution. The red solution of the ester in
amnionic hydrate is decomposed by heating, or even by exposure to
the air at ordinary temperatures, anil we were uot able to obtain a
solution, which did not smell of ammonia, even when a large excess
of the ester was used, it is evident therefore that its ammonium salt
i- very unstable. We have, however, tried the action of a solution
prepared from an excess of the ester and amnionic hydrate with vari-
ous reagents, and found the following characteristic reactions.

Magnesium or Calcium suit, heavy flocculent precipitate of the
color of chrome yellow.

Strontium salt, a less heavy precipitate of a redder color.

Barium salt, a still smaller precipitate, also reddish.

Zinc salt, a pale yellow precipitate.

Cupric salt, a pale yellow precipitate.

Lead salt, a dark yellow precipitate.

Silver salt, a yellowish white precipitate.

The most striking thing about these salts is that the calcium salt is
i ^ soluble than that oi strontium, and this less soluble than the
barium salt. Bischoff* has observed a similar peculiarity in the salts
of orthonitrobenzoylmalonic ester, and we have found it in the salts
oi' bromtrinitrophenylmalonic ester.

With aniline the bromdinitrophenylacetacetic ester gave only a
waxy yellow mass, from which there seemed little chance of obtaining
a substance in a state tit for analysis. We have, therefore, abandoned

* Ann. Chem., ecli.

OF ARTS AND SCIENCES. 279

the further study of this reaction, in which undoubtedly the aniline
acted on the acetacetic ester radical as well as on the bromine in the
benzol ring. That this bromine had been removed we proved by de-
tecting aniline bromide among the products of the reaction.

Sodium Salt, C0H2Br(NO2)2CH3COCNaCOOC2H5.

In the earlier portions of our work on the bromdinitrophenylacet-
acetic ester and allied bodies tbe determination of the composition of
the salts was of the greatest importance, as this was the only ex-
perimental method which we had found for deciding between the
formulas,

I. C(;H,Br ( N02) 2CH,CO CH COOGH.,
II. C HI]] ND2)2CH8COCCOOC2H3.

This, at present, is no longer the case, as the discovery of acetylen-
tetracarbonic ester or tartronic acid as a secondary product in the for-
mation of the corresponding malonic compounds can be explained
only if a formula similar to I. is adopted, and the easy conversion of
bromdi- (or tri-) nitrophenylmalonic ester into the corresponding toluol
compound also could hardly be brought into harmony with a formula
like II. No similar proof has been applied to the acetacetic com-
pound, it is true: but when the close resemblance between this and
the malonic compounds is considered, there can be no doubt that they
are similarly constituted. The composition of the sodium salt there-
fore becomes of secondary importance, and this is fortunate, as we
have not succeeded in preparing it in a state of purity sufficient to
decide between the salts derived from formulas I. and II., although
our analyses leave no doubt as to its composition, if the first formula
i> adopted as correct.

The salt was prepared by adding to a solution of the bromdiui-
trophenylacetacetic ester in absolute alcohol a solution of sodic hy-
drate or ethylate also in absolute alcohol, taking care that the ester
was in decided excess. The deep red alcoholic solution thus obtained
was evaporated rapidly to dryness in a narrow beaker sunk throughout
its whole height in a water bath, some ether having been added pre-
viously in order that its vapor might protect the solution from the
carbonic dioxide of the air until the alcohol began to boil. The
excess of ester was washed out of the dry residue with benzol, and
the salt dried at 100°, and analyzed with the following results : —

I. 0.2102 gr. of the salt gave after treatment with sulphuric ;i
0.0361 gr. of sodic sulphate.

280 PROCEEDINGS OF THE AMERICAN ACADEMY

II. 0.2056 gr. of the salt gave 0.0407 gr. of sodic sulphate.
III. 0.2531 gr. of the salt gave 0.0390 gr. of sodic sulphate.

Calculated for Fouod.

C6H2Br(N02)2CH3COUNaC02C,H5. I. IJ. III.

Sodium 5.79 5.56 6.41 4.99

It is evident from these results that the method of preparation is
unsatisfactory, and we ascribe this to the action of the sodic hydrate
(or ethylate) on the bromine, and perhaps also on the acetacetic radi-
cal, as, when these two sources of error were removed by using the
anilidodinitrobenzylmethylketoue, an excellent result was obtaineii.
Bad as these analytical results are, they show that the salt contains
but one atom of sodium, and therefore must have the formula which
we have ascribed to it.

Properties. — The sodium saltof the bromdinitrophenylacetaceiiv-
ester forms a brick-red amorphous mass, easily soluble in water, al-
cohol, or ether, but insoluble in benzol. Much less freely soluble in
a solution of sodic hydrate than in water. Strong nitric acid decom-
poses it, giving apparently the unaltered ester.

Bromdinitrobenzyhnethylketone (Bromdim'trophenylacetone),
C6II2Br(N02)2CH2COCH3.

This substance can be made from the bromdinitrophenylacetacetic
ester by the action of dilute sulphuric acid in open vessels, or of strong
hydrochloric acid in sealed tubes. The method of preparation which
gave us the best results was as follows : 2 to 2.5 grs. of bromdinitro-
phenylacetacetic ester were boiled with about 200 c.c. of sulphuric
acid of specific gravity 1.44, boiling point 132°, in a flask with a
return condenser, until all the solid was dissolved, which took usually
from one hour to an hour and a half. The yellow solution was then
allowed to cool, when it deposited a heavy white flocculent precipitate,
which was increased in quantity by diluting largely with water; it
was filtered out, and after thorough washing with water was purified
by crystallization from alcohol until it showed the constant melting
point 112°-113°, when it was dried at 100° for analysis. If the ester
used in this process was not perfectly pure, a tarry impurity was
formed which could be removed only with great difficulty ; the best
plau in such a case was to wash the product with a small quantity
(20-30 c.c.) of benzol, which dissolved the ketone more readily than
its impurity, but even after this treatment tedious crystallization from
alcohol was necessary to obtain a pure substance.

OF ARTS AND SCIENCES. 281

The formation of this substance by heating the bromdinitrophenyl-
acetacetic ester with hydrochloric acid is not a good method of prepar-
ing it, because the process must be carried on in sealed tubes, and also
because the product is apt to be contaminated with the tarry impurity
just mentioned; but as it throws light on the reaction which takes
place, we will describe it briefly: 1 gr. to 1.5 grs. of the ester were
sealed in a tube with 20-30 c.c. of pure strong hydrochloric acid, and
heated from 130°-150° for two or three hours. Upon opening the
tube a gas was evolved, which burnt with a smoky green-bordered
flame (ethylchloride), aud also contained carbonic dioxide, as shown
by its giving a precipitate with lime-water. The contents of the
tubes consisting, in addition to the acid liquid, either of tufts of brown-
ish acicular crystals, or a brown oily semi-solid mass, were poured into
a large volume of cold water, and the insoluble portions purified as
already described.

The following analyses were made in part with substance prepared
by the hydrochloric acid process, and in part with that made with
sulphuric acid : —

L 0.2312 gr. of the substance gave on combustion 0.3005 gr. of

carbonic dioxide, and 0.0546 gr. of water.
II. 0.1638 gr. of the substance gave 14 c.c. of nitrogen at a tempera-
ture of 23° and a pressure of 755 mm.
III. 0.1862 gr. of the substance gave, according to the method of
Carius, 0.1173 gr. of argentic bromide.

Carbon '

Calculated for
CGH2Br(XO,),< H,UOCH3.

35.64

r.
35.44

Found.
II.

Hydrogen

2.31

2.62

Nitrogen

9.24

9.57

Bromine

26.40

III.

26.81

It is evident from the analyses and observations given above that
the reaction with hydrochloric acid runs as follows: —

CGH2Br(N02)2CH3COCHC02CJL + HC1 =

C2H3C1 + C02+ C6H2Br(N02)2CH3COCH8;

and that the reaction with sulphuric acid must be similar. We had
expected that bromdinitrophenylacetic acid would be formed also by
these processes, but after a most careful search for it not a trace could
be detected; and as its properties are so striking that we could not
have overlooked it, the conclusion is forced upon us that the reaction

282 PROCEEDINGS OF THE AMERICAN ACADEMY

consisted only in the formation of the ketone. This result is the more
remarkable because Ileckmaun * obtained from the orthoparadinitro-
phenylacetacetic ester by treatment with sulphuric acid of about 10
per cent exclusively the dinitrophenylacetic acid without a trace of the
corresponding ketone. As the presence of bromine in our compound
could hardly have caused such a great difference in the action of the
sulphuric acid, we infer that it must have been caused by the difference
iu strength of the sulphuric acid, Heckmann's containing about 10 per
cent of H2S04, while ours contained 54 per cent. This inference will
be tested by experiment in the ooming year.

Properties. — The bromdinitrobenzylmethylketone crystallizes from
hot alcohol by cooling in white rectangular plates, often with a right-
angled notch in one corner, sometimes also in plates with parallel sides
and a deep notch in each end, which makes them look like reels. If
crystallized by the evaporation of its alcoholic solution, it forms cylin-
drical tufts of needles looking like spires of moss, or much branched
forms resembling certain seaweeds. Both the plates and the branch-
ing needles commonly occur together. It melts at 112°-113° and is
essentially insoluble in ligroine ; nearly insoluble in cold water, more
soluble in hot, but still very sparingly ; slightly soluble in ether, car-
bonic disulphide, benzol, or methyl alcohol, its solubility in the last two
solvents is increased by heat ; tolerably soluble in ethyl alcohol iu the
cold, freely when hot ; ethyl alcohol dissolves it more freely than methyl
alcohol ; tolerably soluble in glacial acetic acid ; soluble in chloroform ;
aud freely soluble in acetone. Hot alcohol is the best solvent for it.
Strong sulphuric acid has no action on it in the cold, but when warmed
dissolves it, forming a colorless solution from which water precipitates
the ketone essentially unaltered, although it appears at first in spheri-
cal groups of thickly set radiating needles, a form in which it is also
obtained sometimes when prepared by the action of sulphuric acid on
bromdinitrophenylacetacetic ester, but these crystals are converted by
crystallization from alcohol into the rectangular plates described above.
Strong nitric acid acts like strong sulphuric acid.

The bromdinitrobenzylmethylketone has well marked acid proper-
ties, and in this respect far surpasses the desoxybenzoine from which
according to Victor Meyer f no sodium compound could be isolated.
In alcoholic solution the ketone gives a dark purplish red color with
sodic hydrate or ethylate. Aqueous amnionic hydrate gives only a
slight red color with it, which is increased by warming for a short

* Ann. Chem., ccxx 134. t Ber. d. ch. G. 1888, p. 1291.

OF ARTS AND SCIENCES. 283

time, and still more by the addition of alcohol, but the color disappears
if the heating is long continued. Aqueous potassic carbonate gave no
action, but on the addition of alcohol a very dark purplish red solution
of the salt. Acid sodic carbonate had no action in aqueous solution,
and only very slight on addition of alcohol. The dark red alcoholic
solution of the ammonium or sodium salt, if made with an excess of
the ketone, is decomposed by water, giving a white precipitate of the
ketone. The decomposition is not complete, however, as the liquid
retains a pale red color. In the presence of an excess of sodic hydrate,
the salt is much more stable. An attempt was made to study the
action of the pale red aqueous solution of the ammonium salt with
various reagents, but no characteristic precipitates were obtained,
probably on account of the small amount of salt left in solution.

We tried also to prepare and analyze the sodium salt, the method
being that adopted for the sodium salt of the bromdinitrophenylaceta-
cetic ester; but the results of the analyses came much too high, 9.73
and 8.64 per cent of sodium, instead of the 7.07 per cent required by
the formula. The reason for this difference is that the sodic hydrate
or ethylate removed a portion of the bromine from the ketone, and the
benzol, dissolving the organic product, left the sodic bromide formed
with the salt of the ketone, as was proved by dissolving the salt in
water, and acidifying with nitric acid, when, after removing the pre-
cipitate by filtration, argentic nitrate gave a heavy precipitate of argen-
tic bromide in the filtrate. We accordingly turned our attention to
the anilidoketone, our work on which is described later in this paper.
We add here the properties of the sodium salt of bromdinitrobenzyl-
metltylketone. It is a purplish black amorphous substance, very easily
soluble in alcohol forming a dark claret-red solution, the color of which
is so much more purple than that of the salt of the corresponding
acetaoetic ester, that the two substances can be easily distinguished in
this way ; it is decomposed almost, but not quite, completely by water,
and is insoluble in benzol.

The ketone is a decidedly reactive substance. When treated with
aniline, it <nves aniline bromide and the anilidodinitrobcnzvlmethyl-
ketone, which is described later. With phenylhydrazine it appears to
form a hydrazone, but at the same time the bromine was removed
from the benzol ring, so that the product seemed to be a pbenylhy-
drazidohydrazone. It was free from bromine, and exploded when
touched with a hot wire; but as its purification offered considerable
(litliculties, we did not try to investigate it thoroughly. The ketone
seems to react in the same way with hydio.xylamine, since in this •

284 PROCEEDINGS OF THE AMERICAN ACADEMY

also bromine seemed to be removed ; but as the product did not have
inviting properties, we have not attempted to isolate the oxime. If
the ketone is dissolved in chloroform, the solution mixed with bro-
mine, and allowed to stand at ordinary temperatures, hydrobromic
acid is given off in considerable quantity, and a new substance melting
at least 10° higher than the original is formed, which will be investi-
gated in this laboratory next term. It is a curious fact, that, if a solu-
tion of the ketone in carbonic disulphide instead of chloroform was
treated with bromine, no hydrobromic acid was given off, so far as we
could find.

Anilidodinitrobenzyhnethylketone, C6H2( CCH5NH) (N02)2CH2COCH3.

This substance is prepared easily by the action of aniline on the
bromdiuitrobenzylmethylketone. The substances were mixed in the
proportion of one molecule of the ketone to two of the base, and the
mixture, which had a bright red color, warmed for 15 to 20 minutes
on the water bath. The product was then freed from the aniline
bromide and any slight excess of aniline by washing with water to
which a little hydrochloric acid was added, the residue, which con-
tained no bromine, purified by crystallization from hot alcohol till it
showed the constant melting point 131°, dried at 100°, and analyzed
with the following result : —

0.1980 gr. of the substance gave 23.2 cc. of nitrogen at a tempera-
ture of 25° and a pressure of 771.2 mm.

Calculated for
CcH,CGU5NTII(N0,)2eH,C0CR"3. Found.

Nitrogen 13.33 13.26

Pro-perties. — The anilidodinitrobenzylmethylketone forms, when
crystallized from alcohol, bright yellow groups of curving needles,
which look at first like two heads of palm trees cut off at the point
where the leaves grow out of the trunks, and put with the stumps
together. As more crystals are formed, the groups develop into
irregular chestnut burs. It melts at 131°; is essentially insoluble in
ligroine ; nearly insoluble in cold water, slightly soluble in hot, form-
ing a pale yellow solution ; slightly soluble in ether; soluble in cold
ethyl or methyl alcohol, more freely iu either of these solvents when
hot ; soluble in benzol, carbonic disulphide, or glacial acetic acid ;
freely soluble in chloroform or acetone. Alcohol is the best solvent
for obtaining crystals. Strong sulphuric acid dissolves it, forming a
brown solution. Strong nitric or hydrochloric acid dissolves it with
a yellow color, but it is very slightly soluble in hydrochloric acid.

OF ARTS AND SCIENCES. 285

The acid properties of this substance are much less marked than
those of the corresponding bromine compound, as was to be expected
from the substitution of bromine by the basic radical C-HSNH.
Aqueous sodic hydrate has no action on it in the cold, and but slight
when warm, but if alcohol is added a brownish red solution of the
sodium salt is obtained. Amnionic hydrate, on the other hand, could
not be made to act on it, even by warming in presence of a large
excess of alcohol. Potassic carbonate in aqueous solution had no
action on it, either cold or hot, but gave a very slight action when a
large quantity of alcohol was added. The behavior of the solution of
a soluble salt with various reagents could not be studied, as the alco-
holic solution of the sodium salt was decomposed completely by dilut-
ing it with water. But in spite of these weaker acid properties we
selected the sodium salt of this substance for analysis, as there was no
danger of a decomposition of this ketone by the alkali used, such as
had prevented us from getting good results with the corresponding
bromine compound. For the same reason, the hydrazone for analysis
was prepared from this substance instead of from the bromdinitro-
benzylmethylketone.

Sodium Salt of Anilidodinitrobenzylmethylketone,
C6H2(CCH5NH) (N02)2CHNaCOCH3.

This substance was prepared by adding an alcoholic solution of
sodic ethylate to an excess of the ketone also dissolved in absolute
alcohol. The narrow beaker containing the mixture, after the addi-
tion of a little ether, was sunk to its rim in a steam bath. In this
way, at first the ether, and later the alcohol vapor, prevented the car-
bonic dioxide of the air from acting on the salt during the evapora-
tion. The dry residue was thoroughly washed with benzol to remove
the excess of the ketone, and the salt thus purified dried at 100°, and
analyzed with the following result: —

0.2380 <xr. of the salt gave after evaporation with sulphuric acid
0.0500 gr. of sodic sulphate.

Calculated for

CcII.,(C6H5NII)(NO:!):,C30H4Na. Found.

Sodium 6.82 6.80

Properties. — A brownish black mass, soluble in alcohol, giving a
much browner solution than any of the other salts described in this
paper, so that it could be recognized with ease by its color. It is
.decomposed at once by water, the brown alcoholic solution being

286 PROCEEDINGS OF THE AMERICAN ACADEMY

turned yellow and turbid. It seems to be nearly or quite insoluble
in ether, and is insoluble in benzol.

Anilidodinitrobenzyhnethylketonehydrazone,
C6H2(C6HfiNH)(N02)2CH3C(N2HC6H5)CH8.

Tbis substance was made by warming for 20-30 minutes on the
water bath in an open dish a mixture of anilidodinitrobenzylmethyl-
ketone with phenylhydrazine in the proportion of one molecule of the
former to about one and a half of the latter, so as to have a decided
excess of the hydrazine. At the end of this time the mixture had
changed from yellow to dark carmine red, and solidified to a tarry
mass, which was crystallized once or twice from alcohol to bring it
into a finely divided state, and washed with water containing a little
hydrochloric acid till the excess of phenylhydrazine had been re-
moved, after which it was boiled with alcohol on the water bath, and
while boiling enough benzol added cautiously to effect the solution ;
the mixture was then boiled for about a minute, when upon cooling
crystals were deposited, and this crystallization was continued until
the substance showed the constant melting point 140°, when it was
dried at 100°, and analyzed with the following result: —

0.2177 gr. of the substance gave 32.5 c.c. of nitrogen at a tempera-
ture of 22° and a pressure of 763 mm.

Calculated for
C6H2(C6H6NH)(N02)2CH2C(N2HC6H5)CH3. Found.

Nitrogen 17.28 16.97

o

Properties. — The hydrazone crystallizes well in reddish brown
scales, resembling strongly in general appearance the officinal ferric
citrate. The scales often reach a diameter of 3 to 4 millimeters.
Under the microscope plates were observed which seemed to belong
to the monoclinic system, but with these were very irregular forms,
often serrated on one or both sides, or irregularly diamond-shaped
with re-entering angles, usually either grouped in radiating masses, or
in crowded branching collections of thick needles. It melts at 140°,
but, as it seems to be slightly decomposed by crystallization, it was
difficult to determine the melting point with perfect exactness. It is
essentially insoluble in ligroine or cold water ; very slightly soluble in
boiling water or in ether; slightly soluble in cold alcohol, more solu-
ble, but still not freely, in hot; more soluble in methyl than in ethyl
alcohol ; slightly soluble in cold glacial acetic acid, freely in hot ; sol-
uble in cold benzol or carbonic disulphide, freely soluble in these sol-

OP ARTS AND SCIENCES. 287

vents when hot ; easily soluble in chloroform or acetone, even in
the cold. As the benzol solution deposits the substance in a vis-
cous state, it is best to crystallize from a mixture of alcohol and
benzol, as described above in speaking of its preparation. Strong
hydrochloric acid has no action upon it. Strong nitric acid seems to
decompose it, dissolving a little of the product. Strong sulphuric acid
dissolves it with a yellowish brown color.

Neither sodic or ammonic hydrate, nor potassic carbonate, nor acid
sodic carbonate, gives any action with it in aqueous solution, hot or
cold, or even on addition of alcohol ; but if it is warmed with sodic
hydrate, water, and alcohol, a little dissolves with a brown color, and
if a drop of sodic hydrate is added to an alcoholic solution, a dark
brownish red solution is formed, which is decomposed by water. It
may be, therefore, that the substance has not lost completely the power
of forming salts, but it is on the whole more probable that the salt is
derived from anilidodinitrobenzylmethylketone formed by the action of
the sodic hydrate on the hydrazone, as the action was certainly accom-
panied by decomposition, since there was a smell of isooyanphenyl,
and the hydrazone could not be recovered from the solution in sodic
hydrate.

The study of the bromdinitrobenzylmethylketone will be continued
in this laboratory.

2,88 PROCEEDINGS OF THE AMERICAN ACADEMY

XXIII.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF

HARVARD COLLEGE.

ON TETRABROMDINITROBENZOL,

By C. Lorixg Jackson and W. D. Bancroft.

Presented May 28, 1889.

After it had been determined that the tribromdinitrobenzol, melting
at 192°, reacted easily with a variety of substances, it seemed of in-
terest to try similar experiments with a tetrabrom compound, and we
selected for this purpose the tetrabromdinitrobenzol, melting at 228°,
and describe in this paper the results of our work, which may be briefly
summarized as follows. Tetrabromdinitrobenzol is not acted on by
alcoholic ammonia in open vessels, but, if heated with it in a sealed
tube, is converted into a yellow substance insoluble in all the common
solvents, which seems to be the bromtriamidodinitrobenzol. With ani-
line the bromtrianilidoclinitrobenzol, melting at 175°-176°, is formed.
With sodium malonic ester it gives in the cold dibromdinitrophenyl-
malouic ester, melting point 89°, which possesses acid properties form-
ing a red sodium salt, and is converted by aniline into the bromanili-
dodinitrophenylmalonic ester, melting at 127°. The action of these
reagents on the tetrabromdinitrobenzol is therefore similar to their
behavior with tribromdinitrobenzol, the fourth atom of bromine re-
maining unaffected in all these reactions. We were prevented by
want of time from trying to remove this fourth atom of bromine by
reactions carried on at higher temperatures. We have studied also
the reduction of the dibromdinitrophenylmalonic ester with tin and
hydrochloric acid, and have obtained in this way the bromamidooxindol,
C6H2BrNH2(CHaCONH), melting at about 212°, and its chloride,
C6H2BrNH2(CH2CONH)HClH20. We may add, that in preparing
the tetrabromdinitrobenzol from somewhat impure tetrabrombenzol we
obtained the as yet undescribed pentabromnitrobenzol, which melts at
248°, and on another occasion the hexabrombenzol.

OF ARTS AND SCIENCES. 289

Preparation of Tetrabromdinitrobenzol.

Our starting point in the manufacture of the tetrabrombenzol was
tribromaniline, which we prepared as follows. 60 grs. of aniline were
dissolved in dilute hydrochloric acid, and the solution, having been
made up to a volume of about three litres, a rapid stream of air satu-
rated with bromine vapor was sucked through it by means of a Bunsen
pump, until the liquid assumed a distinct yellow color. The precipi-
tate of tribromaniline was then removed by straining through cheese-
cloth, washed with a stream of common water until free from acid,
when the greater part of the water was squeezed out with a screw
press, and the product thoroughly dried on a steam radiator. In this
way a quantitative yield of tribromaniline free from colored by-
products was obtained at a much less expense of time and labor than
by the method formerly in use.

To convert the tribromaniline into tetrabrombenzol we used the
method of V. von Richter* slightly modified, which left nothing to be
desired so far as the yield was concerned, when the process was suc-
cessful ; this, however, was not always the case, as frequently the
product was a substance melting in the neighborhood of 87°, from
which tetrabrombenzol melting at 98° could be obtained only by re-
peated crystallization from a mixture of alcohol and benzol, and even
then not in large quantity.f In spite of many experiments, we have
not succeeded in determining with certainty the conditions under
which this mixture was formed, or the nature of the impurity which
it contained, and can give only the conditions which usually gave a
good result. 25 grs. of the tribromaniline were dissolved in a small
quantity of glacial acetic acid with the aid of heat, and, after the solu-
tion had cooled, a concentrated aqueous solution of hydrobromic acid t
was added, which threw down a precipitate of tribromaniline bromide ;
the mixture was stirred vigorously, and, disregarding the precipitate, a
concentrated aqueous solution of about the theoretical amount of sodic
nitrite added in small portions, keeping the solution cool by immersing
the beaker containing it in cold water ; the beaker was then warmed
gently on the water bath, and solid sodic nitrite added in small pieces

* Cer. d. ch. G , viii. 1428.

t Owing to this uncertainty in our modification of Von Richter's process, we
made two attempts to use Sandmeyer's method, but encountered such difficul-
ties in applying it to tribromaniline that we decided it was easier to prepare our
material by Von Richter's process than to overcome these difficulties.

J An attempt to substitute potassic bromide and sulphuric acid for hydro-
bromic acid gave a much poorer yield ; less than 50 per cent instead of 90.

vol. xxiv. (n. s. xvi.) 19

290 PROCEEDINGS OF THE AMERICAN ACADEMY

until the proportion of the total sodic nitrite to the tribromaniline was
three molecules to one (this very large excess of sodic nitrite being
found necessary to complete the reaction). The liquid on cooling
deposited a tarry mass, which was extracted repeatedly with small
quantities of hot alcohol, until it was converted into crystals (melting
at about 95°). An additional amount of these crystals was obtained
by pouring into water the red liquid from which the tar had been
deposited, but this amount was not large. The product, after purifica-
tion by one or two crystallizations from hot alcohol, showed a yield of
from 89 to 93 per cent of the theoretical.

To convert the tetrabrombenzol into tetrabromdiuitrobenzol, 10 grs.
of it were dissolved with the aid of heat in nitric acid of specific
gravity 1.52, prepared in the laboratory from nitre and sulphuric acid;
then the same volume of strong sulphuric acid was added, about 200 c.c.
of the mixture of the two acids being used, and, after standing in the
cold for ten minutes, the whole was boiled for about an hour in a flask
closed with a glass bulb. If the tetrabrombenzol was sufficiently pure
there was little foaming and no violent action, and on cooling a white
crystalline solid separated, which was nearly pure tetrabromdinitro-
benzol, and, after crystallization from benzol with a little alcohol, or
from chloroform, showed the right melting point 228°.*

If, on the other hand, a tetrabrombenzol was used containing some
of the impurity already mentioned, which, when present in quantity,
lowered the melting point to the neighborhood of 87°, the action was
violent, the boiling being accompanied by much foaming, the product
was often oily, and, after most of the solid had been precipitated by
water, the aqueous liquid contained decomposition-products, as was
shown by the blackish precipitate formed in it on the addition of sodic
carbonate. The main product in this case was a mixture of the tetra-
bromdiuitrobenzol and another substance, which was obtained by
treatment with aniline or sodium malonic ester, and subsequent sep-
aration of the derivative of the tetrabromdiuitrobenzol by crystalliza-

* One of us and J. F. Wing (These Proceedings, xxiii. 118) some years
ago obtained tetrabroradinitrobenzol as a secondary product in the manufacture
of tribromtrinitrobenzol, but we were unable to raise its melting point above
224°; this was ascribed at the time to the presence of a little tribromtri- (or di-)
nitrobenzol, an opinion which is confirmed by the fact mentioned above that
tetrabromdinitrobenzol prepared by us according to the usual method showed
the melting point 228° given by Von Richter (Ber. d. ch. G., 1875, p. 1427). The
difficulty in removing the tribrom compound can be accounted for by supposing
that it forms the addition-product observed by one of us and G. D. Moore (Ber.
d. ch. G. 1888, p. 1707), or an analogous one with the trinitro compound.

OF ARTS AND SCIENCES. 291

tion out of alcohol from the unaltered second product, which was then
purified by crystallization from boiling alcohol, or better chloroform,
until it showed the constant melting point 248°, when, after drying at
100°, it was analyzed with the following results : —

I. 0.1711 gr. of the substance gave on combustion 0.0855 gr. of
carbonic dioxide and 0.0151 gr. of water.
II. 0.2100 gr. of the substance* gave 0.1075 gr. of carbonic dioxide
and 0.0103 gr. of water.

III. 0.1846 gr. of the substance gave 3.5 c.c. of nitrogen at a tempera-

ture of 17° and a pressure of 773 mm.

IV. 0.1881 gr. of the substance gave, according to the method of

Carius, 0.3418 gr. of argentic bromide.

IV.

Calculated for

Found

C6Br5N02.

i.

n.

III.

Carbon

13.90

13.63

13.96

Hydrogen

0.00

0.98

0.54

Nitrogen

2.70

2.2

Bromine

77.23

77.34

There can be no doubt, therefore, that this is the as yet undescribed
pentabromnitrobenzol.

Properties. — The pentabromnitrobenzol crystallizes in very slender
little white needles, which melt at 248°. The substance is essentially
insoluble in water or ligroine ; nearly insoluble in cold ethyl or
methyl alcohol, more soluble when hot, but still sparingly ; the solu-
bility in glacial acetic acid or acetone is similar, except that it is much
more soluble in either of these solvents when hot than it is in alcohol ;
readily soluble in hot benzol or chloroform ; very soluble in ether or
carbonic disulphide.

Although pentabromnitrobenzol was the usual impurity, on one
occasion a different product was obtained, which consisted of small
needles melting, after crystallization from benzol, at about 312°, essen-
tially insoluble in alcohol or soclic hydrate, and containing no nitrogen.
We inferred therefore that this was the hexabrombenzol, which is said
to melt above 315°, and to be almost insoluble in boiling alcohol, and
this inference was proved to be correct by the following analysis : —

0.1303 gr. of the substance gave, by the method of Carius, 0.2666
gr. of argentic bromide.

Calculated for C6Br9. . Found.

Bromine 86.96 87.10

2?2 PROCEEDINGS OP THE AMERICAN ACADEMY

Preliminary Experiments.

After the tetrabromdinitrobenzol had been prepared, the following
experiments were tried to determine the ease with which it reacted
with various substances.

As has been stated in an earlier paper, alcoholic ammonia has no
action on tetrabromdinitrobenzol in the cold, or even if the substances
are heated together in open vessels. If however the mixture is heated
to 100° in a sealed tube for four or more hours, a reaction takes place
giving an orange precipitate and a red alcoholic solution in addition to
some unaltered tetrabromdinitrobenzol. The orange insoluble sub-
stance was undoubtedly the bromtriamidodinitrobenzol, but, as it was
insoluble in all the common solvents, we could only try to purify it by
washing, and even after repeated treatment with water, alcohol, chloro-
form, and benzol were unable to obtain a substance giving constant
results on analysis ; the bromine varied in different samples from
25.34 to 29.07 per cent, the amount required by the formula,

C6Br(N02)2(NH2)8,

being 27.39. Although it is probable that further work on this sub-
stance would have enabled us to find a method of purifying it, we did
not think it worth while to sacrifice the time, as these analyses show
that the reaction has run in the same way as that between tribrom-
dinitrobenzol and alcoholic ammonia, and nothing of interest was likely
to be found in its investigation. The substance is an orange-yellow
powder, insoluble in all the common solvents, and not melting even at
285°. The alcoholic filtrate contained a yellow substance melting at
185° in the crude state, but we did not obtain enough of it for investi-
gation. With aniline a more favorable result was obtained, which is
described fully in the next section of this paper.

When boiled with potassic sulphocyanate in alcoholic solution in a
flask with a return condenser, there was very little action, but, if amyl
alcohol was substituted for ethyl alcohol, so that the action would take
place at a higher temperature, a dark red substance was formed insolu-
ble in all the common solvents. In this respect the tetrabromdinitro-
benzol behaved exactly like the tribromdinitrobenzol, and, as it had
been found impossible to purify the corresponding product from the
tribrom compound, we did not think it worth while to continue work
in this direction.

With sodium malonic ester or sodium acetacetic ester it behaves
like the tribromdinitrobenzol. The action with sodium malonic ester
is described in detail later in this paper.

OF ARTS AND SCIENCES. 293

In all the derivatives of tetrabromdinitrobenzol which were analyzed
it was found that only three of the atoms of broniiue had been at-
tacked. We had hoped to try some experiments at higher temperatures
for the purpose of replacing the fourth atom of bromine, but the work
described in this paper has taken so much time that we have been
unable to take up this branch of the subject.

Bromdinitrotrianilidobenzol, C6Br(N02)2(Ct;H5NH)g.

To prepare this substance tetrabromdinitrobenzol was heated with
aniline in the proportion of one molecule of the former to a little more
than six of the base. Convenient amounts were 7 grs. of tetrabrom-
dinitrobenzol to 8.3 grs. of aniline. When the mixture was heated on
the water bath, the solid dissolved after some time, the solution being
accompanied by a change of color from yellow to bright red, and on
cooling the whole solidified to a mass of red needles; but in order to
get even a tolerable yield of the new substance the heating on the
water bath must be continued for at least four to six hours, and then
it is not by any means complete. A higher heat than the water bath
should not be used, as then the mixture shows a tendency to pass
into a purplish coloring matter, resembling impure rosaniline in ap-
pearance, and similar to the substance obtained in the same way from
tribromdinitrobenzol,* and like that undoubtedly produced by the
nitro groups taking part in the reaction. If the reaction had run
properly, the product consisted of a viscous or crystalline mass of the
color of red lead, from which in the first place the aniline bromide and
excess of aniline were removed by washing with water containing a
little hydrochloric acid, leaving a brick-red powder. The purification
was completed by crystallization from alcohol, and afterward from a
mixture of alcohol and chloroform, until it showed the constant melt-
ing point 175°-176°. If there was difficulty in obtaining crystals of
the proper melting point, it was found advisable to warm the substance
again with aniline for two or more hours, and then purify again in the
manner just described. In addition to unaltered tetrabromdinitroben-
zol a small amount of an impurity was found, which melted in the
crude state between 120° and 140°, but we did not succeed in isolat-
ing a substance fit for analysis from it. The main product, when
pure, was dried at 100°, and analyzed with the following results : —

I. 0.1667 gr. of the substance gave 20.4 c.c. of nitrogen at a tem-
perature of 25° and a pressure of 768.4 mm.

* These Proceedings, xxiii. 145.

291 PROCEEDINGS OF THE AMERICAN ACADEMY

II. 0.1624 gr. of the substance gave, by the method of Carius,
0.0578 gr. of argentic bromide.

Calculated for Found.

C6Br(N02)2(C6H6NH)3. I. II.

Nitrogen 13.46 13.81

Bromine 15.38 15.15

Properties. — Bromtrianilidodinitrobenzol forms a brilliant but rather
dark red crystalline powder, which, when examined with the micro-
scope, consists of crystals of two forms, rather short prisms terminated
by an obtuse angle, and thick groups shaped like an hour-glass and
made up of short prisms, which seem to have the same terminal angles
as those which are not in groups ; the groups are rather darker in
color than the free prisms, but this is probably due to their being
thicker. When crystallized from hot alcohol, instead of the usual
mixture of alcohol and chloroform, characteristic forms like sheaves
were obtained, also a small quantity of red nearly square plates. It
melts at 175°-176° ; is essentially insoluble in ligroine, or in water
either cold or boiling ; very slightly soluble in cold alcohol, more
soluble in hot ; more soluble in methyl than in ethyl alcohol ; slightly
soluble in cold ether, freely in hot ; moderately soluble in glacial acetic
acid ; freely soluble in benzol, chloroform, or carbonic disulphide ;
very freely in acetone. The best solvent for it is a mixture of alcohol
and chloroform, as it tends to separate in a viscous state from the
solution in chloroform alone. Strong sulphuric acid dissolves it
slightly, forming a pale yellow solution. Strong nitric acid acts in the
same way, but less energetically ; fuming nitric acid acts upon it vio-
lently even in the cold. Hydrochloric acid has no action either hot
or cold. The absence of basic properties is accounted for by the
presence of the two nitro groups. We tried also the action of sodic
hydrate, as certain nitroamido compounds possess weak acid proper-
ties,— for instance, the trinitrotoluidine of Nblting and Salis,* — but
found that it produced no effect.

Dibromdinitrophenylmalonic Ester, CcHBr2(N02)2CH(COOC2H5)2.

This substance was prepared by acting on one molecule of tetra-
bromdinitrobenzol with about four molecules of sodium malouic ester,
as follows : 15 grs. of tetrabromdinitrobenzol dissolved in about 20-
30 c.c. of benzol were mixed with 20 grs. of malonic ester previously
converted into sodium malonic ester by treatment with the sodic ethyl-

* Ber. d. ch. G., xv. 18G4.

OF ARTS AND SCIENCES. 295

ate made from 3 grs. of sodium dissolved in 100 to 150 c.c. of absolute
alcohol. The action began at once, as shown by the appearance of a
dark red color and a considerable evolution of heat, so that the flask
became too hot to hold in the hand with comfort ; in this respect it
differed from that with the tribromdinitrobenzol, as, although in that
case the red color appeared at once, no perceptible rise of temperature
was observed in any part of the reaction. To make certain that the
reaction had run as far as possible, the mixture was allowed to stand
in the cold for three or four days. At the end of this time the product
was diluted with about half a litre of water, and the benzol separated
from the red aqueous solution of the salt of the new substance. Upon
adding dilute sulphuric acid to this aqueous solution, the new substance
was precipitated, and after washing with water was purified by crys-
tallization from alcohol until it showed the constant melting point 89°,
when it was dried in vacuo for analysis. The benzol solution, which
separated on the addition of water, was evaporated to dryness, and the
residue, consisting of the new substance, a little unaltered tetrabrom-
dinitrobenzol, and an oil, after being freed from the oil on the pump,
was treated with alcohol, in which the tetrabrom compound is es-
sentially insoluble, and the small amount of the new substance thus
obtained was added to that from the aqueous solution.

I. 0.2183 gr. of the substance gave on combustion 0.2568 gr. of

carbonic dioxide and 0.0528 gr. of water.
II. 0.2144 gr. of the substance gave 11.4 c.c. of nitrogen at a tem-
perature of 23°. 1 and a pressure of 763.4 mm.
III'. 0.2372 gr. of the substance gave, by the method of Carius,
0.1833 gr. of argentic bromide.

Found.

III.

Carbon

Calculated for
C6HBr2(NO,)2CH( C02C2H5)2.

32.23

i.
32.07

Found,
II.

Hydrogen
Nitrogen

2.48
5.78

2.69

6.0i

Bromine

33.06

32.91

The yield of dibromdinitrophenylmalonic ester was very satisfactory.
On one occasion 5 grs. of tetrabromdiuitrobenzol gave 4 grs. of the
dibromdinitrophenylmalonic ester, instead of the 5 grs. required by the
theory, that is, 80 per cent of the theoretical yield. This large yield
proves conclusively that dibromdinitrophenylmalonic ester is the only
aromatic product of the reaction, a point which before this observation
seemed somewhat doubtful in the analogous preparations described in

296 PROCEEDINGS OP THE AMERICAN ACADEMY

other papers from this laboratory, as in them the yield rose but little
above 50 per cent of the theory.

The amount of bromine separated as bromide of sodium in the prep-
aration of the substance was determined in two cases as follows : —

I. 5 grs. of tetrabromdinitrobenzol yielded 3.707 grs. of argentic

bromide, corresponding to 1.578 grs. of bromine.
II. 5 grs. of tetrabromdinitrobenzol yielded 3.G31 grs. of argentic
bromide, corresponding to 1.545 grs. of bromiue.

The theoretical yield, if two of the atoms of bromine had been re-
moved, would be 1.653 grs.

Percentage of bromiue found as bromide of sodium, —

I. ii.

95.43 93.48

From these numbers it appears that essentially all the bromine is
removed as sodic bromide.

The oil which formed the secondary product of the reaction was
not investigated further, as from the composition of the ester, and the
fact that all the bromine appeared as sodic bromide, there could be no
question that the nature of this action is the same as that with the tri-
brom compounds ; and in those cases* the study of the oils has show a
conclusively the mode of formation of the substances. There can be
no doubt, therefore, that the reactions which take place in this case
should be written as follows : —

C6Br4(N02)2 + 2 CHNa(COOC2H5)2 =

NaBr+ C6Br3(N02)2CH(COOC2H5),-f CHNa(COOC2H.)2 =
C6Br3(N02)2CNa(c60C2H5)2 + CH2(COOC2H5)2 =
C6HBr2(N02)2CNa(COOC2H5)2 + CHBr(COOC,H5)2.

The brommalonic ester being converted afterward into tartronic or
acetylentetracarbonic ester.

Properties. — The dibromdinitrophenylmalonic ester crystallizes
from alcohol usually in slightly yellow to colorless needles, which unite
into groups like pompons if the solution is dilute; seen under the mi-
croscope they consist of rather long rhombic plates with a decidedly
acute angle, about six times as long as their breadth, or even slen-
derer, either free or in radiating groups, the crystals branching so
as to give the effect of being arranged in curved lines. When less
well developed, the groups are made up of fine needles very much

* These Proceedings, xxiv. 238, 265.

OF ARTS AND SCIENCES. 297

branched, like certain feathery seaweeds. From the oil which forms
the secondary product in its preparation, the substance crystallizes in
flattened prisms, often two centimeters in length, arranged in slightly
radiating groups. The melting point is 89°. It is essentially insolu-
ble in water or ligroine ; slightly soluble in cold, freely in hot methyl,
or ethyl alcohol ; freely soluble in carbonic disulphide, glacial acetic
acid, or acetone ; very freely in ether, benzol, or chloroform. Strong
sulphuric acid has little, if any, action on it in the cold, but dissolves
it slightly without change of color when hot. Strong nitric acid also
seems not to act upon it in the cold, but, when warmed with it, turns
yellow and imparts a yellow color to the drops of the melted sub-
stance which swim on its surface. Hydrochloric acid does not seem
to act on it in open vessels either cold or warm, although from analooy
with the corresponding monobrom compound it is fair to suppose that
it would decompose it if the two substances were heated to 150° in
a sealed tube.

The dibromdiuitrophenylmalonic ester has well marked acid prop-
erties, as was to be expected from the fact that one of its atoms of
hydrogen is attached to a carbon atom in direct contact with two car-
boxylester radicals and a dinitrophenyl. An excess of sodic hydrate
in aqueous solution has but little action on the solid, turning it pale
red ; if, however, a drop of alcohol is added, it turns a very dark red
at once, and the salt begins to dissolve. Potassic carbonate acts in
much the same way. Acid sodic carbonate in aqueous solution has
no effect, but if alcohol is added there is a slight action shown by
formation of a little of the red salt. Amnionic hydrate in aqueous
solution has but little action, but turns red if a little alcohol is added",
and upon warming this mixture a deep brownish red solution is ob-
tained, which smells strongly of ammonia, even if an excess of the
ester is used in making it. It seemed to be decomposed by heating,
or by exjiosure to the air, so that the ammonium salt must be a very
unstable substance. The solution of the ammonium salt made with
an excess of the ester (but still smelling of ammonia) was treated with
various reagents, and gave the following characteristic precipitates : —

With a salt of magnesium, calcium, strontium, or barium, very heavy
pale red flocks.

With a zinc salt, pale yellow flocks.

With a salt of manganese, cobalt, or nickel, pale red flocks.

With & ferric salt, pale brownish purple.

With a cupric salt, bright yellow.

With salts of mercury, pale red.

298 PROCEEDINGS OF THE AMERICAN ACADEMY

With a cadmium salt, reddish yellow.

With a lead salt, whitish red.

With a silver salt, pale orange.

Strong nitric acid gave with the salts a white precipitate, probably
the unaltered ester. In this respect it differs from bronitrinitro-
phenylmalonic ester.

Bromanilidodinitrophenylmalonic Ester,
C6HBr(C6H5NH)(N02)2CH(COOC2H5)2.

This substance was made by adding aniline to the dibromdinitro-
phenylmalonic ester in the proportion of a little more than two mole-
cules of the base to one of the ester. The reaction began in the
cold, but was brought to an end by warming the mixture for a short
time on the water bath. The product, after washing with water, to
which a little hydrochloric acid was added to remove aniline bromide
and the excess of aniline, was crystallized from hot alcohol, until it
showed the constant melting point 127°, when it was dried at 100°>
and analyzed with the following result : —

0.2003 gr. of the substance gave 14.85 c.c. of nitrogen at a tem-
perature of 20°. 5 and a pressure of 762 mm.

Calculated for
C6HBr(CGH5XII)i X02)2CH(C02C2H5)2. Found.

Nitrogen 8.47 8.49

Properties. — The bromanilidodinitrophenylmalonic ester crystal-
lizes from alcohol in bright red needles, which under the microscope are
seen to be slender prisms terminated usually by two planes, less com-
monly by one, and seeming to belong to the monoclinic system. The
substance melts at 127° ; and is essentially insoluble in cold water,
very sparingly soluble in boiling water, as shown by the faint yellow
color imparted to the liquid ; insoluble in ligroine ; very slightly solu-
ble in cold ethyl or methyl alcohol, more freely in either of these
solvents when hot; slightly soluble in ether; soluble in benzol or
glacial acetic%acid; freely soluble in chloroform, carbonic disulphide,
or acetone. Strong sulphuric acid dissolves it slightly, forming a
yellowish solution ; the solubility did not seem to be increased by
heat. Strong nitric acid dissolved it rather more freely than sul-
phuric acid, and the solubility was increased by heat. Strong hydro-
chloric acid had no action upon it, although it is probable that long
heating in a sealed tube with this acid would have decomposed it in
the way described * under the bromdinitropbenylmalonic ester.

* These Proceedings, xxiv. 240.

OF ARTS AND SCIENCES. 299

The acid properties of this substance have been much weakened by
the replacement of bromine by the basic aniline radical CtiH.NH, but
that they still exist is shown by the following observations. Sodic
hydrate in aqueous solution has no action on the solid, but on the
addition of a little alcohol a dark red solution of the salt is formed ;
sodic carbonate in aqueous solution produces no effect, but on the
addition of alcohol a very slight red coloration appears ; acid sodic
carbonate produces no effect even in presence of alcohol. Ammonic
hydrate in aqueous solution does not dissolve the solid substance, and,
t^en if alcohol is added, the action is very slight ; if, however, the mix-
ture of aqueous ammonic hydrate, alcohol, and the solid is warmed on
the water bath, a dark red solution is obtained, which smells of ammo-
nia, even if a large excess of the ester is used. The behavior of
such a solution with various reagents was studied with the following
results : —

Salts of magnesium, calcium, strontium, or barium gave rusty brown
precipitates.

The salts of the heavy metals gave yellow precipitates, except where
the color was modified by the excess of ammonia which could not be
removed from the solution of the salt.

Reduction of Dibromdinitrophenylmalonic Ester.

The investigation of this subject interested us especially, because it
seemed probable that the diamidophenylmalonic or acetic acid, which
would be the first product of the reaction, would lose water and be-
come converted into an amidooxindol, especially since Bischoff* ob-
tained from the reduction of his orthonitrobenzoylmalonic ester a-y-
diliyclroxychinoline. As with zinc and alcoholic hydrochloric acid, or
with zinc dust and acetic acid, he obtained more complex products,
some of which it was almost impossible to purify, we decided to try
first the action of tin and hydrochloric acid with alcohol upon our
dibromdinitrophenylmalonic ester, and for this purpose proceeded as
follows. 2 grs. of the dibromdinitrophenylmalonic ester were mixed
with alcohol, strong hydrochloric acid, and tin, a piece of platinum foil
being used to accelerate the action, and the mixture was kept upon a
steam radiator at a temperature of from 50°-70° until the whole of
the malonic compound had disappeared, which usually happened in an
hour and a half. The clear solution was poured off from the excess
of tin, and, after evaporation to dryness, dissolved again in water, and

* Ann. Chem., ccli. 364.

300 PROCEEDINGS OP THE AMERICAN ACADEMY

freed from tin with sulphuretted hydrogen ; upon concentrating the
filtrate, long needles separated, which varied in color from brown to
nearly white. By further concentration of the mother liquors a fresh
crop of crystals was obtained, and the precipitate of sulphide of tin
must be boiled out several times with water, as the reduction-product
is but slightly soluble. As the substance seemed to be decomposed to
a certain extent by our attempts to purify it by crystallization, we
analyzed some of it without further purification, while other samples
were crystallized once more from hot water, allowing the solution to
cool in vacuo to avoid oxidation by the air. It was dried in vacuo,
and analyzed with the following results : —

I. 0.0977 gr. of the substance gave on combustion 0.1224 gr. of car-
bonic dioxide and 0.0434 gr. of water.
II. 0.3108 gr. gave on combustion 0.3869 gr. of carbonic dioxide ami
0.1128 gr. of water.

III. 0.1113 gr. of the substance gave 9.8 c.c. of nitrogen at a temper-

ature of 25°. 5 and a pressure of 763 mm.

IV. 0.1010 gr. of the substance treated with argentic nitrate and the

precipitate washed with nitric acid and water gave 0.0498 gr.
of argentic chloride.
V. 0.4762 gr. gave 0.2461 gr. of argentic chloride.

Found.
I. II. III. IV. V.

Carbon 34.17 33.93

Hydrogen 4.93 4.03

"Nitrogen 9.84

Chlorine 12.19 12.77

The free base corresponding to this chloride was next prepared by
adding amnionic hydrate to a strong solution of it, when a white
precipitate swimming in a dark green liquid was obtained. It was
purified by washing with cold water, in wdiich it is as good as insolu-
ble, until the wash water gave no test for a chloride, then dried in
vacuo, and analyzed with the following results : —

I. 0.1690 gr. of the substance gave on combustion 0.2587 gr. of
carbonic dioxide and 0.0580 gr. of water.
II. 0.2021 gr. of the substance gave 23.7 c.c. of nitrogen at a tem-
perature of 24°. 5 and a pressure of 768.5 mm.

Found.
I. II.

Carbon 41.75

Hydrogen 3.81

Nitrogen 13.28

OF ARTS AND SCIENCES. 801

The following comparison shows that the numbers obtained from
the analyses of the chloride correspond to those calculated for the
formula CsHfJBrN.202HCl (with the exception of the hydrogen,* which
was undoubtedly brought too high by the passing over of a part of
the halogen into the sulphuric acid bulbs).

Calculated for

CgU9BrN202H01.

Found.

Carbon

34.12

34.17

33.93

Hydrogen

3.55

4.93

4.03

Nitrogen

9.95

9.84

Chlorine

12.61

12.19

12.77

On the other hand, the numbers given by the analyses of the free base
indicate a substance containing one molecule less of water.

Calculated for

Calculated for

C8H7BrN20.

Found.

C8H9BrN202.

Carbon

42.30

41.75

39.17 "

Hydrogen

3.08

3.81

3.67

Nitrogen

12.34

13.28

11.43

The want of agreement between the percentages of hydrogen f is ex-
plained in the same way as in the analyses of the chloride. Other-
wise the numbers come as near as could be expected, when it is
remembered that the free base was so unstable that we did not dare
to purify it except by washing, and that it gradually turned brown
even when dry.

In order to harmonize these results and determine the nature of
the substances we have found only three hypotheses.

First, and most obvious. The two substances belong to different
classes, i. e. one is the chloride of the bromdiamidophenylacetic, acid
C^HoBrNHgClNIIoCHjCOOH; the other is free bromamidooxindol,
CflH2BrNH2(CH2CONH).

If, on the other hand, the substances belong to the same class, —

Second. They are the bromamidooxindol and its chloride. In this
case the chloride must contain one molecule of water of crystallization.

Third. Tliey are bromdiamidophenylacetic acid and its chloride.
In this case our analyses of the free base are incorrect.

* The formula C8HnBrN.202HCl requires 4.23 per cent of hydrogen, and is
therefore distinctly too high for Analysis II., and the analysis of the " chloride
of the free base " given later. The small amount of substance used in Analysis
I. makes the per cent of hydrogen in it of no value.

t The formula C8II9I3rNoO requires 3.93 per cent of hydrogen.

302 PROCEEDINGS OF THE AMERICAN ACADEMY

If the first of these explanations is the true one, the chloride made
by adding hydrochloric acid to the free base would be the chloride
of the bromamidooxindol, and therefore different from the original
chloride of the bromdiamidophenylacetic acid. To settle this point,
we set free the base with amnionic hydrate from a quantity of the
original chloride, and, after washing until free from ammonic chloride,
dissolved it in dilute hydrochloric acid and crystallized it from the
slightly acid solution. The general habit of the crystals of the two
chlorides (the original one and that made from the free base) was the
same, but there were differences in the modifications on the ends of
the prisms which made the identity of the two somewhat doubtful ;
we accordingly analyzed the chloride made from the free base, with
the following results : —

0.1853 gr. of the substance gave on combustion 0.2328 gr. of car-
bonic dioxide and 0.0678 gr. of water.

Calculated for Calculated for Found. Analyses of

C8H713rN20IICl. C8H0BrN202HCl. Original Chloride.

Carbon 36.43 34.12 34.26 34.17 33.93

Hydrogen 3.04 3.55 4.06 4.93 4.03

There can be no doubt, therefore, of the identity of the two chlo-
rides, and the first explanation must be abandoned.

The second explanation requires that the chloride should contain
one molecule of water of crystallization. This point was tested by
heating the chloride with the following result : —

0.2517 gr. of the chloride heated for 6 hours at first at 110°, later
to 135°, lost 0.0024 gr.

Calculated for
C8H,BrN,OHClH20. Found.

Water 6.42 0.95

From this it appears that, if the chloride contains water of crystalli-
zation, it does uot lose it at 135°. The slight loss of only 2.4 mgrs.
could be sufficiently accounted for by the decomposition of the salt,
which had turned dark gray on the surface, and for this reason too the
heating could not be repeated at a higher temperature.

This result, unfavorable to the second explanation, necessitates the
discussion of the third, — that the free base is bromdiamidophenylacetic
acid, — which otherwise we should not have thought worthy of consid-
eration. As has been already stated, the analyses of the free base do
not agree with this explanation, and in order to adopt it we must
assume that the substance had undergone a decomposition sufficient to

OF ARTS AND SCIENCES. 303

raise the carbon 2.58 per cent, an assumption improbable in itself, and
not supported by the appearance of the preparation, which had only a
slight dirty pink color. This we think is enough to condemn this
third explanation, but we add the following considerations, all of
which tell strongly in favor of the second, and against the third ex-
planation. (1.) Gabriel and R. Meyer* by the reduction of diui-
trophenylacetic acid with tin and hydrochloric acid obtained the
paramidooxindol direct, and it does not seem possible that our brom-
diamidophenylacetic acid, which differs from theirs only in containing
an atom of bromine, should be so much more stable. (2.) Our free
base agrees fairly well in properties with the paramidooxindol of
Gabriel and R. Meyer. Both are easily oxidized, soluble in hot water
or in alcohol, slightly soluble in benzol or carbonic disulphide, the only
difference being that our base is nearly insoluble in ether, whereas
theirs is soluble in it. The melting points also stand about where we
should expect, amidooxindol about 200°, our bromamidooxindol about
212°; in both cases there was so much blackening that the melting
point could not be accurately determined. (3.) Our base (or its chlo-
ride) gives the indol reaction, turning a piece of pine wood red if
boiled with it and dilute sulphuric acid, or if the wood, after being satu-
rated with a solution of the base, is soaked in strong hydrochloric acid.
This seems to us conclusive. The color is rather dull, and appears only
after some time ; but the objection which might be urged against this
argument on this account, that the bromdiamidophenylacetic acid was
converted into the bromamidooxindol by treatment with the acids, has
no weight, as in the preparation of the original chloride it was most
thoroughly exposed to the action of strong hydrochloric acid, both hot
and cold ; and yet this chloride, as we have shown by the analyses,
contained two atoms of hydrogen and one of oxygen more than is re-
quired by the chloride of the bromamidooxindol.

These arguments seem to us to prove conclusively that the two
substances are bromamidooxindol CGH2BrNH2(CILCONH), and its
chloride CGH2BrNH,(CH2CONH)HCl.H20, in spite of the fact that
we could not drive off the water of crystallization from the chloride
even at 135°.

Properties of Bromamidooxindol, CGH2BrNH2(CH.,CONiI).

As precipitated from its chloride it forms a heavy flocculent white
precipitate, which, when examined with the microscope, is seen to be

* Ber. d. ch. G., xiv. 832.

304 PROCEEDINGS OF THE AMERICAN ACADEMY

a felt of small white needles, often arranged in fagots or groups like
an hour-glass ; it is not very stable, changing to a dirty pink on ex-
posure to the air, even when dry. The substance analyzed melted at
212° with a good deal of blackening, but we do not place much reli-
ance on this melting point, as the sample used was a good deal colored.
It is almost insoluble in cold, soluble in hot water ; soluble in hot
alcohol ; insoluble or nearly so in ether or chloroform ; not freely
soluble in benzol or carbonic disulphide, even when hot ; freely soluble
in hot glacial acetic acid. Sodic hydrate dissolves it, giving a pinkish
or pale magenta solution ; ammonic hydrate at first seems to have no
action, but on standing the liquid and solid turn dark bluish green ;
potassic carbonate in aqueous solution does not dissolve it. It seems
therefore to have the properties of a phenol. Neither picric acid in
henzol solution, nor ferric chloride with an alcoholic solution of the
base, produced any change of color. If some of the free base was
boiled with dilute sulphuric acid and a piece of pine, the wood was
turned a dull orange-red.

As to the constitution of this base, it has been determined by work*
done in this laboratory on the bromdinitrophenylmalonic ester that
the nitro group, from which the amido group is formed, is in the para
position to the carbon side chain. The bromine is probably in the
ortho position, since all our work with the tetrabromdinitrobenzol has
shown that only three of its atoms of bromine can be replaced easily ;
and, as the compounds made from this substance are so closely analo-
gous to those prepared from the tribromdinitrobenzol, it is fair to
infer that the three symmetrical (meta) atoms of bromine are those
which can be removed, and that the fourth more stable atom of
bromine is the one in the unsymmetrical (ortho) position ; the atom
of bromine, therefore, which reduction takes away from the dibrom-
dinitrophenylmalonic ester, would be that in the meta position, and
the base would be accordingly orthobromparamidooxindol.

Properties of the Chloride of Bromamidooxindol,
CGH2BrNH2(CH2CONH)HClH20.

When crystallized from water, the substance forms needles, or
prisms, sometimes as much as a centimeter long and a millimeter
thick, which seem to belong to the monoclinic system, and usually
have a yellowish color; we think, however, that this color is due to
partial oxidation, and that the substance is white when pure. When

* These Proceedings, xxiv. 249, 250.

OF ARTS AND SCIENCES. 30"

• >

examined with the microscope, the prisms are found to be arranged in
globular radiating groups, and are terminated usually by a single
rhombic plane at a not very oblique angle to the long sides of the
prism ; sometimes a number of other modifying planes are observed,
giving a sharp end to the prism. From alcohol it crystallizes in
branching ueedles set at a very acute angle to each other. It melts
only at a very high temperature, probably above the range of the
mercury thermometer. It is slightly soluble in cold water, freely in
hot; soluble in alcohol; slightly soluble in cold benzol, chloroform, or
glacial acetic acid, freely soluble in these solvents, when hot; almost
insoluble in ether; very soluble in carbonic disulphide. The best sol-
vent for the substance is boiling water. An excess of amnionic hy-
drate throws down from a concentrated solution a heavy precipitate
of the free base, which is white, but the liquid turns dark green ;
potassic carbonate gives a similar precipitate, not soluble in an excess ;
sodic hydrate in small quantity gives a similar precipitate, but dis-
solves it when added ho larger quantity, giving a reddish solution.
It gives a dull orange-red indol reaction with pine wood.

vol. xxiv. (n. s. xvi.) 20

306 PKOCEEDINGS OF THE AMERICAN ACADEMY

XXIV.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF

HARVARD COLLEGE.

GENERAL CONSIDERATIONS IN REGARD TO CER-
TAIN COMPOUNDS PREPARED FROM
BROMNITROBENZOLS.

By C. Loring Jackson.

Presented May 28, 1889.

Several compounds derived from various bromnitrobenzols have
been described recently by me in conjunction with other chemists in
a number of papers* from this laboratory, but in the course of the
work certain general observations were made in regard to them which
could not conveniently be introduced in the separate papers, and are
therefore collected here.

. The Action of Sodium Malonic {or Acetacetic) Ester on certain

Bromnitrobenzols.

The actions studied were the following : —

Sodium Malonic Ester on

Tribromtrinitrobenzol, C6Br3(N02)3, melting point 285°.

Tribromdinitrobenzol, C6HBr.,(N02)2, melting point 192°.

Tetrabromdinitrobenzol, CtiBr4(N02)2, melting point 227°.
Sodium Acetacetic Ester on

Tribromdinitrobenzol, C6HBr3(N02)2, melting point 192°.

In all these cases the action was essentially the same, and may be
represented by the following reactions for tribromtrinitrobenzol : —

I. C6Br8(N0.2)8+ CHNa(COOC2H5)2 =

NaBr+ CGBr2(N02)3CH(COOC2H5)2.
II. C6Br2(NO,),CH(COOC.,H.)2 + CHNa(COOC2EL),=

C6Br2(N02)3CNa(CObC2II5)2 + CH2(COOC2H5)2-
III. C0Br2(NO2)3CNa(COOC2H5)2 + CH2(COOC2H5)2 =

C,HBr(N02)3CNa(COOC2H5)2 + CHBr(COOC2H5)2.

* These Proceedings, xxiii. 138; xxiv. 1, 105, 234, 256, 271, 288.

OF ARTS AND SCIENCES. 307

The formation of brommalonic ester in III., the only point in these
reactions about which there was any doubt, was proved by the isola-
tion of acetylentetracarbonic ester* or of tartronic acid J from the
oily secondary product, as these substances could hardly have been
formed otherwise than by the following reactions, —

CHBr(COOC2IL), + CHNa(COOC2H5)2 =

CHCfI((JOOC2lL)4 + NaBr.
CHBr(UOOG,H5)2 + NaOM = CHOII(COOC2H.)2 + NaBr.

Reactions I.— III. are given because they show more clearly what
has taken place, but the discussion which follows does not depend
on them alone, since the analyses of the products are sufficient to
prove that in all the cases studied the action has consisted in the re-
placement of one atom of bromine by the malonic ester radical, and
of the second by hydrogen, while the third (and fourth) has remained
unaltered.! This difference in the behavior of the three atoms of
bromine is certainly very curious, and, so far as I can find, no case
analogous to it has been described as yet. The strangeness of the
replacement of bromine by hydrogen becomes especially evident when
it is remembered that in every case there was a large excess of sodium
malonic ester present, and that this second atom of bromine has there-
fore combined with the carbon of malonic ester (see Reaction III.) in
preference to the sodium of sodium malonic ester, so that the tendency
to introduce hydrogen rather than the malonic ester radical has been
strong enough in this case to overcome the attraction of the bromine
for the sodium, and to cause it to combine with carbon instead.

I have not succeeded in finding any explanation of these curious
observations which satisfies me, but hope to be led to one by further
experiment. It can be stated at present, however, that the difference
of behavior in the three bromine atoms is not due to differences in
their position on the benzol ring. This is shown easily in the case of
the tribromtrinitrobenzol, as here the three atoms of bromine occupy
exactly similar positions, each being ortho to two nitro groups and
para to the third. In the tribromdinitrobenzol.§ also, the first ami

* These Proceedings, xxiv. "2G5. t Ibid., xxiv. 238.

J Ih the trinitro compound the third atom of bromine is replaced by the ma-
lonic ester radical to a limited extent even in the cold, but it is acted on much
less easily than the first.

§ In the tetrabromdinitrobenzol the fourth atom of bromine seems to have
no influence on the reaction, and therefore all I say about the tribromdinitro-
benzol also applies to this substance.

808 PROCEEDINGS OF THE AMERICAN ACADEMY

third atoms of bromine are similarly placed (ortho to one, para to the
other nitro group), but the second (that replaced by hydrogen) is pe-
culiar in being ortho to both nitro groups. It is evident, then, that in
these reactions both similarly placed bromine atoms are not replaced by
the malonic ester radical ; but it is also to be observed that the bromine
which is replaced by hydrogen has in every case stood in the ortho
position to two nitro groups, and this suggested to me that perhaps an
atom of bromine in this position might be especially susceptible to the
action of free malonic ester. To test this hypothesis I mixed some
tribromdinitrobenzol dissolved in benzol with malonic ester, and, after
the mixture had stood for some time in the cold, warmed it for about
an hour on the water bath, but even after this treatment no action *
had taken place, as the tribromdinitrobenzol melted unaltered at 192°.
This result was confirmed by a similar experiment with acetacetic
ester, which stood with tribromdinitrobenzol for six mouths without
any action. It is evident, therefore, that this atom of bromine is not
attached in an especially loose way to the molecule, and that its re-
placement by hydrogen does not depend only on its position with
reference to the nitro groups, but is due also to the presence of the
malonic ester radical.

On the Relative Ease vrith which the Reactions take Place.

This subject will I hope prove susceptible of quantitative treatment,
and I propose next year to try some experiments of this sort ; I shall
therefore at present confine myself to two qualitative observations,
which were so marked that a quantitative confirmation of them seems
unnecessary.

The tribromtriuitrobenzol is distinctly more reactive than the tri-
bromdinitro- or tetrabromdinitrobenzol,as it acts on alcoholic ammonia
in the cold, whereas the other two must be heated to 100° in sealed
vessels to bring about this reaction.

The sodium acetacetic ester acts much less energetically than sodium
malonic ester upon tribromdinitrobenzol, since the red color of the
product appears instantaneously with the malonic ester, but only after
some time with acetacetic ester, and the yield of the acetacetic com-
pound is much smaller than that of the malonic derivative, if the
reactions are carried on under the same conditions.

* The addition of a little aqueous sodic hydrate was enough to cause action
at once, as shown by the appearance of the red color of the sodium salt.

OF ARTS AND SCIENCES.

809

On the Acidity of the substituted Malonic Esters, Acetacetic Ester,
and Ketones described in the preceding Papers.

In order to determine the relative acidity of these substance*, their
action with the following reagents was studied: acid sodic carbonate,
sodic carbonate, amnionic hydrate, sodic hydrate. Each reagent was
added in aqueous solution to the solid organic substance, and, after
observing what took place, alcohol was added, and any change in the
behavior of the substauce noted. All these experiments were tried at
the same time, great care being used to make the conditions as nearly
the same as possible. The results can be divided into those which
were perfectly definite, that is, where the differences consisted in the
fact that one compound formed a salt with the reagent, and another
did not ; and those depending to a certain extent upon my judgment,
in which the differences were only in the amount of salt formed, as
shown by the depth of color, or in the cpuantity of alcohol necessary
to produce the salt. The differences of the first class can be described
most clearly and succinctly by the following table, in which the sub-
stances are arranged in the order of their acidity, beginning with the
most acid. The columns correspond to the reagents, and the word
" Salt " indicates that a salt was formed, as shown by the color imparted
to the solution, or solid.

NaHC03

Na2C03

NH,OH

NaOH

0Q

=
O
0?

S

<
Salt

._ c

Salt

w

2
O

D
C
<

Salt

So

w

S
o

0)

s
<

"5
So

DO

s
o

3
<

"3

_ &

. - o

,1. CuH^BrtNO.JA^sOCHoCOa^Hs

Salt

Salt

Salt

Salt

Salt

II. C6HBr(NO.J3CH(C02C2H5)2

Salt

Salt

Salt

Salt

Salt

Salt

Salt

Salt

III. C6H„Br(N02)oCH(COoC.2H5)2

Salt

Salt

Salt

Sr.lt

Salt

Salt

Salt

IV. C6H2Br(XOo)oCII2COCri3

••

Salt

*

Salt

Salt

Salt

Salt

Salt

V. C,HBro(N02)2CH(C02C2H;;)2

••

Salt

Salt

Salt

Salt

Salt

Salt

Salt

VI. C6H2CBH,NH(N02)2CH(C02C.2H5)2

Salt

••

Salt

Salt

Salt

VII. (VnBrC,Hr,NH(NO,)2CH(C02C2H5),

::

Salt

Salt

Salt
Salt

VIII. C„H2C0H5NH(NO2)2CH2COCII3

* No salt was obtained from IV. with aqueous sodic carbonate, but I ascribe
this to a mistaken observation, probably due to the difficulty in moistening the
ketone with the aqueous solution.

•310 PROCEEDINGS OF THE AMERICAN ACADEMY

The observations recorded in this table divide these substances into
the following groups (indicated by the lines): — I. and II., most acid ;
III., IV., and V., less acid ; VI.. less acid ; VII., less acid ; VIII., least
acid. The relative acidity of I. and II. was determined by data of the
second sort mentioned above, since I. gave a much stronger coloration
than II. with aqueous sodic carbonate, amnionic hydrate, or sodic hy-
drate. The order of III., IV., and V. was determined by the following
observations. With acid sodic carbonate and dilute alcohol, III. and

IV. gave a stronger red color than V. With aqueous sodic carbonate,
III. gave a pale red, V. a much paler, in fact barely perceptible color.
With aqueous amnionic hydrate, III. gave a pale red, V. a very pale
barely perceptible red, and IV. stood between III. and V. in color.
With aqueous sodic hydrate, the color was confined to the crystals of
the organic substance, owing to the slight solubility of the salt in a
solution of sodic hydrate, and III. turned dark red, IV. pale red, while

V. showed a mixture of pale red particles of the salt and white ones
of the ester. Upon adding alcohol to the mixture of the substance
and either aqueous sodic carbonate or amnionic hydrate, a dark red
color was obtained from all three, but much more alcohol was neces-
sary to produce this color with V. than with III. or IV. My obser-
vations on the intensity of the color with different reagents confirm
the order of compounds VI., VII., and VIII., but I do not think it
worth while to give the details. I should add, also, to show that the
personal equation did not enter to auy great extent into these obser-
vations, that the substances during the experiments were arranged in
an order different from that of their acidity, and that the order of
acidity given above derived from these experiments was a surprise to
me in many particulars.

These experimental results justify the following inferences in re-
gard to the effect of differences in conijx>sition upon the acidity of
these substances.

(a.) The introduction of an additional nitro group (in the ortho
position) increases the acidity, since II. is more acid than III.

(b.) The replacement of hydrogen (in the ortho position) by bro-
mine diminishes the acidity, since V. is less acid than III.

(c.) The replacement of bromine (in the meta position) by the
aniline radical C'(,1LNTI reduces the acidity very considerably, since
III., IV., and V. are much more acid than VI., VIII., and VII.
respectively.

In all these cases the differences occur in the benzol ring, and the
effects are very marked (especially in c) when it is considered that

OF ARTS AND SCIENCES. 311

the hydrogen affected is attached to the side chain, and therefore more
removed from the influence of the acid radicals than in most of the
similar cases previously described, in which the element influenced was
attached to the benzol ring. These observations recall some made by
me several years ago, when 1 showed * that the ring bromine in the
three brombenzylbromides had a marked influence in diminishing the
ease with which the side-chain bromine was removed, varying with its
position on the benzol ring, the para bromine having less influence
than the meta, the meta than the ortho.

(d.) The nature of the side-chain also has a marked influence on
the acidity ; thus the substituted acetacetic ester I. is much more acid
than the corresponding malouic ester III., a result in harmony with
Claisen and Ehrhardt's f classification of the radicals according to their
acid-producing power, since they give the following list beginning with
that which has the least influence, "phenyl, carboxyl, benzoyl, acetyl,
formyl." An approximate measure of the relative acid-producing
power of carboxyl and acetyl is given by the comparison of III. and
IV., which possess nearly the same degree of acidity, and therefore it
follows that one acetyl has nearly, although not quite, the same effect
as two carboxyls.

(e.) The comparison of a dimalonic with the corresponding mono-
malonic compound in this respect is of interest. Unfortunately I had
none of the trinitrophenylendimalonic ester when the comparison de-
scribed above was made, so that this substance does not appear in the
table; but by comparing the observations made on it when its prop-
erties were studied with those on the bromtrinitrophenylmalonic ester,
I have found that the dimalonic compound is much less acid than the
brommonomalonic ester, since acid sodic carbonate in aqueous solution
has no action upon it, and even aqueous sodic carbonate gives only a
very faint red color.J This weaker acidity can be accounted for by
the hypothesis, that the influence of the three nitro groups being di-
vided between two malonic ester radicals, each will have the acidifying
influence of only one and a half nitro groups, instead of the three,
which act on the acid hydrogen in the monomalonic compound.

The stability of the salts of all these compounds is remarkable, the

* Those Proceedings, xvi. 241.

t Ber. d. ch. G. 1889, p. 1019.

i I should he inclined to place the trinitrophenylendimalonic ester between
III. and V., although from the absence of comparative work ita exact place cun-
not be determined with certainty ; this, however, is of little importance, as the
comparison of it with II. is the only one of especial interest.

312 PROCEEDINGS OF THE AMERICAN ACADEMY

sodium salt even of the anilidodinitrobenzylmethylketone (the least
acid of all these substances) having been prepared and analyzed with
no especial difficulty ; it shows therefore a marked difference in this
respect from desoxybenzoine (CgFLG'OCHgCytL), the sodium com-
pound of which, according to Victor Meyer,* has not been isolated.
In fact, he thinks it probable that it may not exist in the free state,
and leaves it doubtful whether the syntheses of homologues of desoxy-
benzoine are preceded by the formation of a sodium compound, as
with malonic or acetacetic ester, or whether the homologue is formed
direct by the removal of hydrogen and the halogen by sodic hydrate,
according to this reaction :

C6H.COCHCuH. + CH3 + NaOH = Nal + Ho0 + C6H.COCHCcH..

[H~ ~T\ CH3.

The reason for this striking difference in acidity between the sub-
stituted benzylmethyl ketones and desoxybenzoine can hardly be due
to the fact that the former contain the more active acetyl instead of
the benzoyl contained in the latter, although this may have some influ-
ence, but in my opinion is rather to be ascribed to the heightening of
the acid-producing power of the phenyl by the nitro groups which it
contains. The salts of the acetacetic or malonic esters are much more
stable than those of the ketones, as was to be expected, since in these
cases the hydrogen is exposed to the acidifying power of three radicals
(nitrophenyl and two carboxyls, or one carboxyl and an acetyl), and
their stability almost equals that of the salts of BischofF's f orthonitro-
beuzoylmalonic ester, in which the nitrophenyl of my compounds is
replaced by the even more acid nitrobenzoyl.

* Ber. (1. ch. G. 1888, p. 1291.
t Ann. Chem., ccli. 364.

OF ARTS AND SCIENCES. 313

XXV.

FEATURES OF CRYSTALLINE GROWTH.

By Oliver Whipple Huntington.

Presented June 12, 1889.

Variations in the Butcher Irons.

In a former paper * the author discussed in some detail the crystal-
line structure of iron meteorites, and in a second f described the
cleavage exhibited by various specimens of the Coahuila group of
irons, which are generally supposed to have come from the same
original mass. It was then shown that in this group of irons the
Butcher specimens differed so markedly from those of Saltillo (Santa
Rosa) that the cleavage seemed to offer a conclusive means for separat-
ing the two falls, though the appearance of the etched surfaces failed
to distinguish them.

As several very large individual masses of these irons are now in the
Harvard Cabinet, under the general name of " The Butcher Irons,"
we were interested to see whether any peculiarity of structure would
further distinguish them as independent meteorites, or whether on the
contrary they could be identified as having unquestionably come from
the same original body. A careful examination of the exterior of
these irons showed that they had begun to break up since becoming
the property of Harvard College, but this alteration had only taken
place in certain parts of the specimens. The largest specimen seemed
to have a zone running through the thickest part from which could be
detached with little difficulty octahedral fragments of iron made up of
distinct plates, showing a cleavage very similar to the well known Put-
nam County iron, while the rest of the mass was compact and un-
altered. Thus certain parts of the Butcher specimens appeared to
have an entirely different structure from what we are accustomed to
associate with these irons. They have heretofore been known as
typical " cubic irons," and, though it has been shown that all the twin

* Proceedings of the American Academy, vol. xxi. p. 478, May, 1886.
t Ibid., vol. xxiv. p. 30, October, 1888.

314 PROCEEDINGS OF THE AMERICAN ACADEMY

members of the regular octahedron cube and dodecahedron appear on
the surface of fracture, still the cube gives character to the iron, and,
together with the etched surface, serves to connect it closely with the
Brauuau (Hauptmannsdorf) and other so called cubic irons.

In seeking to explain this apparent variation in the structure of the
Butcher specimens, a number of large masses were polished and
etched. The most striking result was obtained on a specimen which
formed one end of a large mass originally cut up by Ward and Howell
for the late J. Lawrence Smith. This surface as it appeared after
etching is shown, one third the natural size, in Plate I., which was
made directly from a photographic negative. The figure, although so
much reduced, shows clearly a zone passing through the thickest part
of the mass characterized by a very different crystallization from the
rest of the piece. The greater portion of the etched surface is cov-
ered only with Neumann lines so characteristic of the Coahuila irons.
These lines, although not well shown in the photograph, are plainly
exhibited by the print from an etched plate previously published.*
This last feature is characteristic of the outside of the mass, where the
cooling must have been quicker and the crystallization more rapid and
disturbed, though following, nevertheless, the fundamental forms of
the regular system. In the central zone, however, corresponding to
the thickest part of the specimen, the iron appears to have crystallized
more slowly, and exhibits all the characters of Widmanstattian figures
resulting from well marked crystal plates ; and here the octahedron
appears to predominate, though the presence or absence of the cube
and dodecahedron cannot be established by the examination of only a
single etched surface cut at random.

The etched surface of another portion of the same mass seemed to
explain this contrast of coarse and fine grained crystallization. There
is in the Harvard collection the largest slab ever cut from the Coa-
huila specimens, being a full section of the large mass which we have
been discussing. The etched surface of this slab was fully as striking
as that shown in Plate L, but exhibited more uniformity in the. distri-
bution of the figures. Unfortunately, the slab was too large to allow
of its reproduction on paper within the limits of this publication, but
the figures can be best described by their resemblance to the mark-
ings of frost on a large window-pane. Around the edge the crystals
were so compact as to show no details, but these compact masses
joined the inner portion on a curved outline, looking like cloud masses

* Proceedings of the American Academy, vol. xxiii., Plate I.

OP ARTS AND SCIENCES. 315

which had floated in to the distance of from ten to one hundred
millimeters from the crust, and on the inner edge of these cloudy por-
tions the crystals became more distinct, and finally shot out in crystal-
line plates (which were seen of course only in section) some of them
from fifty to one hundred millimeters in extent. These again were
met and intersected by similar crystalline plates radiating from numer-
ous centres through the mass, these centres being spots of more com-
pact crystallization, like the cloudy masses around the edge, forming in
the interior a network of plates producing skeleton crystals.

Thus the entire surface presented exactly the appearance that one
might expect in a mass of iron which had slowly cooled. In such a
mass, the surface cooling first would become compact, and exhibit only
confused crystallization, while the interior, cooling more slowly, would
allow time for isolated crystal plates to form ; and at the same time
local causes would start a simultaneous crystallization from other
points in the interior, determined probably by the presence of nuclei
of foreign matter.

The possible variation in the character of the figures brought out by
etching different portions of the same iron, has been mentioned in a
former paper,* but at that time no such striking example had been
noticed as that now exhibited by the Coahuila irons. Specimens
might be cut from the iron shown in Plate I. so selected that one piece
when etched would give well marked Widmanstattian figures, with all
the characteristic features which have commonly been associated with
the so called typical octahedral irons, while another would show
equally typical Neumann lines, and still a third piece would appear
perfectly amorphous or made up of irregular grains.

Widmanstattian Figures on " Spiegel Eisex."

Hitherto Widmanstattian figures and Neumann lines have been
considered the strongest characteristic of meteoric irons ; but since it
is merely the evidence of slow crystallization attended by the exclu-
sion of foreign matter, one would expect that any impure iron, pro-
vided that it had cooled slowly from a state of fusion, would exhibit
similar characters. From depending too much upon these features the
Greenland iron was originally considered to be of meteoric origin, as
some of the masses show very good Widmanstattian figures. Since
specimens of cast iron have been frequently described as meteoric, and

* Proceedings of the American Academy, vol. xxi. p. 478, May, 1880.

316

PROCEEDINGS OF THE AMERICAN ACADEMY

Fig. 1.

are to be found in all the large museums of the world, I was inter-
ested in examining specimens of " Spiegel Eisen," which contains so
large an impurity of manganese, and is at the same time so strikingly
crystalline. In the first place, the fracture somewhat resembles that
of the Mexican irons which we have been discussing, as it shows cube
faces, although the octahedron is dominant. Moreover, thin plates
of " Spiegel Eisen " can in some cases be detached from the mass which
would make it appear still more like meteoric iron, were it not for
its great hardness. On a polished surface, however, the resemblance
is far more striking, since etching or tempering brings out well de-
fined Widmanstattian figures. These figures are shown of natural
size in Figure 1, which is an exact sketch of the etched surface on a
piece of this iron.

OF ARTS AND SCIENCES.

317

Crystalline Plates in Galena.

In this connection it may be of interest to describe some specimens
of galena which appeared among the minerals used for blowpipe
analysis by the class in mineralogy at Harvard College, but whose
origin is unknown.

Fig. 2.

There were found three crystals very much alike, one of which is
shown considerably enlarged in the accompanying sketch. On one side
it showed the perfect cubic cleavage of ordinary galena, but the mass
of the crystal consisted of a network of plates about 1 mm. thick, in-
tersecting each other for the most part at right angles, and thus form-
ing a skeleton cube. But that which makes the specimens of unusual
interest is the fact that there are certain additional plates, following
the direction of the faces of an octahedron, and thus running diago-
nally through the entire thickness of the crystals. Such a plate is
shown at A B in the sketch (Fig. 2).

Moreover, the cleavage of the galena appeared to be due to some
other cause than the form of the crystal, since the octahedral plates
were as well marked and nearly as abundant as the cubic ones, while
the cleavage was distinctly cubic. Thus it appears that when galena
crystallizes, just as when iron crystallizes, plates form through the
mass following the directions of both the cube and octahedron till

318

PROCEEDINGS OF THE AMERICAN ACADEMY

the mass becomes solid, but that subsequently either cleavage may
become so dominant as to wholly cover up all other crystalline
structure.

Crystalline Concretions of Mica.

Some time ago I received from Dr. D. F. Lincoln of Boston some
remarkable specimens of curved mica, one of which is shown of natu-
ral size in Figure 3. This crystal had its cleavage face very much

Fig. 3.

curved, while the successive layers tapered back to an apex at the
centre of curvature, the whole having the appearance of forming a
portion of a sphere. The occurrence of these micas is very well de-
scrihed in a letter from Dr. Lincoln : —

" The specimens of mica were procured by me in 1885, from Mt.
Apatite in Auburn, Maine. The rock of that locality resembles that

V

x

X

■^ ^>

*!=

V

-.

OP ARTS AND SCIENCES. 319

of Mt. Mica, being a granite, with all its components (quartz, ortho-
clase, clevelandite, muscovite, black tourmaline, etc.), in very lar^e
masses. At the eastern end of the ridge there are several places
where a very pure quartz is quarried in immense quantities. The
peculiar mica is found at one of these quarries at points on the border
of the quartz deposit. The coarser and less well formed specimens
are imbedded in quartz and feldspar ; the finer and well formed are
usually enclosed in feldspar alone. One vertical face of the quarry
presented a most interesting exposure. Here the mica stood out in
bosses, or huge breasts, looking like broken spheres developed from
numerous centres, and grouped very irregularly. The spherical sur-
faces, on analysis, consisted of groups of rhombic figures, each repre-
senting a crystal of mica, with curved cleavage and tapering to the
centre of the sphere.''

Thus it appears as if these crystals were the result of two distinct
tendencies acting together, one concretionary, exerted to make the
mass solidify from centres, and the other the crystallizing force, tend-
ing to form at the same time the rhombic prism of 60° and 120° so
characteristic of the mineral.

320 PROCEEDINGS OF THE AMERICAN ACADEMY

XXVI.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF
HARVARD COLLEGE.

I. ON CHLORPYROMUCIC ACIDS*
By Henry B. Hill and Louis L. Jackson.

Presented May 28, 1889.

Many years ago Malaguti f found that dry chlorine gas was quickly
absorbed by ethyl pyromucate, and that a thick viscous liquid was
formed, the composition of which showed that four atoms of chlorine
had been taken up. The oil was carbonized on distillation, and
yielded with potassic hydrate decomposition products which were not
further studied. Although Schmelz and Beilstein $ later found that
chlorine converted pyromucic acid in aqueous solution into muco-
chloric acid, and this reaction was afterwards further studied by Hill
and Bennett,§ no further experiments, as far as we could learn, had
ever been made with dry chlorine on pyromucic acid or its ethers.
In 1884, after it had been shown by investigations carried on in
this laboratory that substitution products could readily be formed
from pyromucic acid by the action of dry bromine, Mr. J. N. Garratt,
at that time an assistant in the laboratory, undertook the investigation
of the action of dry chlorine under similar conditions. Although he
succeeded in isolating a dichlorpyromucic acid melting at 1G7-1G80
in a state of purity, he found that the reaction differed in many re-
spects from the corresponding reaction with bromine, and that the
matter deserved a more careful study than he was then able to bestow
upon it. Mr. Garratt relinquished his investigation of this subject in
order to continue his studies in Zurich, and in the following winter

* A part of the work described in the following paper was pre«ented in the
form of a thesis to the Academic Council of Harvard University in May, 1888,
by Louis L. Jackson, then candidate for the degree of Doctor of Philosophy.

t Ann. Chim. Phys., Ixiv. 282; lxx. 371.

t Ann. Chem. u. Pharm., Suppl., iii. 276.

§ Berichte d. deutsch. chem. Gesellsch., xii. 655.

OP ARTS AND SCIENCES. 821

met with the fatal accident which so sadly ended his promising
career.

The observations which already had been made interested us so
much that in 1886 we undertook a more thorough study of the sub-
ject. We had already collected a great deal of material which we did
not think it necessary to publish in a preliminary form, when Denaro *
published in the Gazzetta Chimica a brief notice of a dichlorpyromucic
acid melting at 167-168°, which was evidently identical with that pre-
pared by Garratt. The paper contained a description of the acid and
two of its salts, but the analyses of the latter did not agree with those
which we had already obtained, and moreover contained such serious
misprints or arithmetical errors, or both, as to be wholly without
value. In consequence of the appearance of that paper, we thought
it best to publish a preliminary account f of our work, so far as that
one acid was concerned, more especially since our study of it was then
essentially completed.

In repeating the experiments of Garratt, we had no difficulty iD
isolating the dichlorpyromucic acid which he had prepared, but we
found that the reaction was still more complex than we had supposed.
At ordinary temperatures chlorine was rapidly absorbed, but the pyro-
mucic acid was soon so coated with the viscous addition product that a
homogeneous product could not be obtained. On heating, the reaction
could easily be pushed to the end, and the gain in weight closely cor-
responded with the addition of four atoms of chlorine. At the same
time hydrochloric acid and carbonic dioxide were given off, and the
resulting product contained a percentage of chlorine much larger than
that demanded by any simple addition product. Under these circum-
stances it seemed advisable to return to the ethyl pyromucic tetrachlo-
ride of Malaguti, which, from his description and analyses, appeared to
be homogeneous ; and to take up the study of the more complex ac-
tion of chlorine upon the acid itself after this investigation had made
us better acquainted with some of the products likely to be formed.
Although we were unable to prepare a perfectly homogeneous product
by following the directions of Malaguti, we found no difficulty in so
modifying the conditions that the resulting compound should give us
pure fiy dichlorpyromucic acid melting at 168-169°, S chlorpyro-
mucic acid melting at 176-177°, or trichlorpyromucic acid melting at
172-173°. The (38 dichlorpyromucic acid we were unable to prepare
by direct substitution, or by saponification of ethyl pyromucic tetra-

* Gazz. Chim., xvi. 333. t Berichte d. deutsch. chem. Gesellsch., xx. 252.
vol. xxiv. (n. s. xvi.) 21

322 PROCEEDINGS OF THE AMERICAN ACADEMY

chloride ; but by distillation of ethyl pyromucic tetrachloride, of the
tetrachloride of pyromucyl chloride, or of ethyl 8 chlorpyroniucic tetra-
chloride, the ySS dichlorpyromucic acid melting at 155-156° may be
obtained, and in the first two cases a third dichlorpyromucic acid melt-
ing at 197-198° is also formed. By careful reduction of the j38 or
the /3y dichlorpyromucic acids, the /3 chlorpyromucic acid melting at
145-146° is formed.

Ethyl Pyromucic Tetrachloride.

The ethyl pyromucate which we have used in the course of our
investigations we have made by warming a solution of pyromucic acid
(3 parts) in absolute alcohol (5 parts) with concentrated sulphuric
acid (3 parts acid, Sp. Gr. 1.84). After heating for four hours on the
water bath, the mixture was allowed to cool, precipitated with water,
and the ether washed with a dilute solution of sodic carbonate. The
crude ether was then dried by exposure to the air and distilled. The
yield of pure distilled ether which is thus obtained amounts to about
85 per cent of the weight of the acid taken. This is somewhat less
than that which may be obtained by means of hydrochloric acid, but
the method is much more expeditious and convenient.

When dry chlorine is passed over ethyl pyromucate we found that
the ether was rapidly liquefied with the evolution of heat, and that the
formation of the addition product proceeds as described by Malaguti
until the ether has increased in weight very nearly the amount re-
quired by the addition of four atoms of chlorine. We soon found,
however, that, no matter how carefully the chlorine was dried, hydro-
chloric acid escaped, showing the formation of a substitution product,
and that the amount of hydrochloric acid formed depended largely
upon the rapidity of the stream of chlorine and the consequent eleva-
tion of temperature. Even when the ether was carefully cooled to 0°
during the whole of the treatment with chlorine, the formation of
hydrochloric acid could not wholly be avoided. In this respect our
experience agrees with that of Hill and Sanger,* who were unable to
form the ethyl pyromucic tetrabromide without the simultaneous for-
mation of substitution products. While the statements of Malaguti
seemed to leave no room for doubt that the ethyl pyromucic tetrachlo-
ride was completely carbonized by heat, experiments made with bro-
mine in this laboratory had shown that substitution products could
readily be formed from the ethyl pyromucic tetrabromide by heat,

* These Proceedings, xxi. 155.

OF ARTS AND SCIENCES. 323

and it seemed to us hardly conceivable that complete carbonization
should ensue. A preliminary trial showed so much less charring
under ordinary pressure than we had anticipated, that we were en-
couraged to hope that the ethyl pyromucic tetrachloride itself might
be distilled under diminished pressure without essential decomposition.
Ethyl pyromucate was therefore carefully treated with chlorine at 0°
until it ceased to gain in weight. Hydrochloric acid was given off in
small quantity, and the total gain in weight was 94 per cent of the
weight of ethyl pyromucate taken, instead of the theoretical gain of
101 per cent. The product was then fractionally distilled under a
pressure of 15 mm. and showed its complex nature by the wide range
of its boiling point, 104-160°. Above 160° there remained a residue
which even under 15 mm. pressure could not be distilled without
decomposition.

None of the fractions showed any tendency to solidify in a freezing
mixture. After several distillations, it became evident that the greater
portion boiled without essential decomposition between 150-160°,
and that a smaller portion boiled below 110°.

In order to determine the nature of the several products, the follow-
ing analyses were made. Each fraction was collected under 15 mm.
pressure.

A. Boiling point 104-110° ; weight = 2.6 grm.

0.2235 grm. substance gave 0.2139 grm. AgCl.

Calculated for

C6H2C103C2US. Found.

CI 20.34 23.63

B. Boiling point 145-150° ; weight = 3.2 grm.

0.2334 grm. substance gave 0.4418 grm. Ag. CI.

C Boiling point 152-153° ; weight = 7.7 grm.

I. 0.1678 grm. substance gave 0.3431 grm. AgCl.
II. 0.1415 grm. substance gave 0.2889 grm. AgCl.

D. Boiling point 153-157° ; weight = 12.0 grm.
0.1502 grm. substance gave 0.3094 grm. AgCl.

Found.
Calculated for B. C. D.

C5H30,C2II5C14. I. n.

CI 50.36 46.77 50.55 50.49 50.97

From these analyses it was evident that the fraction C consisted of
pure ethyl pyromucic tetrachloride ; the fraction D contained a slight

324 PROCEEDINGS OF THE AMERICAN ACADEMY

admixture of a product containing a higher percentage of chlorine ;
the fraction A approached in composition the ethyl chlorpyromucate ;
while the fraction B was a mixture. In order to be quite sure of the
correctness of our conclusions, we treated these successive fractions
with an alcoholic solution of sodic hydrate, and obtained acids in each
case confirming the results of our analyses. From fraction A we
obtained an acid crystallizing from hot water in irregular leafy plates
which melted at 176-177°. The physical properties and melting
point were sufficient to identify this acid as the 8 chlorpyromucic acid
later described, and to prove that the lower boiling fraction consisted
chiefly of ethyl 8 chlorpyromucate. Fraction C yielded us without
difficulty, and in nearly theoretical quantity, an acid which when re-
crystallized twice from benzol melted sharply at 168-1G9°, and in other
respects proved to be identical with the /3y dichlorpyromucic acid
which we shall hereafter fully describe. The formation of this acid in
nearly theoretical quantity and without recognizable admixture proves
that the fraction C consisted of pure ethyl pyromucic tetrachloride.
Fraction D yielded us a somewhat less pure j3y dichlorpyromucic
acid, as did also fraction B. In the latter case, however, we were
able to isolate a small quantity of a sparingly soluble acid, crystal-
lizing from water in minute needles which closely resembled trichlor-
pyromucic acid. It was therefore evident that the main product formed
by the action of chlorine at 0° upon ethyl pyromucate was the tetra-
chloride and this might be distilled under diminished pressure without
essential decomposition. At the same time, it was shown that substi-
tution was effected even in the cold, and that the product contained
ethyl 8 chlorpyromucate and very possibly its tetrachloride.

8 Chlorpyromucic Acid.

Since substitution had taken place at low temperatures, it seemed
not unlikely that at 100° the substitution might be so rapid as to offer
a convenient mode for the preparation of chlorpyromucic acids. Ethyl
pyromucate when treated with bromine at 100° yields 8 brompyro-
mucic and y3S brompyromucic acids still more readily than pyromucic
acid itself. We accordingly passed dry chlorine through ethyl pyromu-
cate heated to 100°, but found that a higher temperature was necessary
in order that substitution might promptly be effected. At 145° the
action was sufficiently rapid, and the chlorine was passed through the
melted ether at this temperature until a gain in weight was noted
which corresponded to the substitution of one hydrogen atom by chlo-
rine. The viscous liquid which was thus obtained was then slowly

OF ARTS AND SCIENCES. 325

added to an excess of a concentrated alcoholic solution of sodic hydrate,
care being taken to avoid any great elevation of temperature. "When
the action was completed the sparingly soluble sodium salt was re-
moved by filtration and dried. The alcoholic filtrate contained small
quantities of this sodium salt in solution, and contained further a small
amount of some volatile furfuran compouud. We converted the
excess of sodic hydrate into carbonate with carbonic dioxide, removed
the sodic carbonate by filtration, and distilled the alcoholic filtrate.
The distillate when mixed with water threw down a minute quantity
of a colorless oil of peculiar aromatic odor, which we have as yet
been unable to identify through lack of material. The small amount
of sodium salt obtained from the alcoholic mother liquors was then
added to the main portion. The sodium salt when dissolved in water
and acidified with hydrochloric acid then yielded the crude 8 chlorpyro-
mucic acid, which not infrequently needed no other purification than
recrystallization from benzol. If, however, the treatment with chlorine
had been too long continued, or the action had taken place at too low
a temperature, the product contained dichlorpyromucic and trichlorpy-
romucic acids which could not be removed by simple recrystallization.
Whenever the crude acid failed on trial to crystallize from hot water
in shining irregular plates, it was necessary to resort to chemical
means for its purification. The acid was suspended in twenty times
its weight of cold water, ammonia added in slight excess, and then
baric chloride in quantity sufficient to precipitate the sparingly soluble
barium salts of the admixed acids. After the separation of these salts
is complete, the filtered solution gives on acidification an acid which
crystallizes from water in leafy plates, and which may further be puri-
fied by recrystallization from .benzol. The yield of chlorpyromucic
acid thus obtained amounts to about 40 per cent of the weight of ether
taken, or 38 per cent of the theoretical amount.

I. 0.2761 grm. substance gave 0.4145 grm. CO., and 0.0494 grm. II20.
II. 0.2G67 grm. substance gave 0.2642 grm. AgCl.
III. 0.2990 'grm. substance gave 0.2950 grm. AgCl.

Calculated for Found.

cysjCio,. i. ii. in.

C 40.96 40.94

H 2.05 1.99

CI 24.23 24.49 24.39

The 8 chlorpyromucic acid is readdy soluble in hot water, sparingly
in cold water, and crystallizes in large irregular leafy plates, which

326 PROCEEDINGS OF THE AMERICAN ACADEMY

closely resemble the corresponding bromine derivative. It dissolves
readily in alcohol, ether, or hot benzol, but sparingly in cold benzol.
The acid recrystallized from benzol melted at 176-177°, and a prepa-
ration made from the pure etbyl ether melted at the same point. Its
solubility in cold water was determined according to the method of
V. Meyer. A weighed portion of the solution saturated at 19°. 5
was boiled with baric carbonate, and the barium dissolved determined
by precipitation with sulphuric acid.

I. 37.8458 grm. solution saturated at 19°.5 gave 0.0833 grm. BaS04.
II. 33.5216 grm. solution saturated at 19°.5 gave 0.0742 grm. BaS04.

According to these determinations, the solution saturated at 19°. 5
contained the following percentages of acid.

i. n.

0.28 0.28

Baric S Chlorpyromucate, Ba(C5H2C10„)2 . H20. — The barium salt
was prepared by boiling the acid with an excess of baric carbonate.
The salt is quite readily soluble even in cold water, still more readily
soluble in hot water, and separates from a hot concentrated solution
in thin leafy plates, usually aggregated in globular form. The air-
dried salt contains one molecule of water, which it loses rapidly at 100°.

I. 1.2193 grm. air-dried salt lost at 100° 0.0491 grm. H20.
II. 0.5932 grm. air-dried salt gave 0.3094 grm. BaS04.

Calculated for Found.

Ba(C5H2C103)2 . H20. I. II

H20 4.04 4.04

Ba 30.72 30.67

0.6909 grm. salt dried at 100° gave 0.3763 grm. BaS04.

Calculated for

Ba(C5II2C103)2. Found.

Ba 32.01 32.02

The solubility of the salt in cold water was determined according
to the method of V. Meyer.

I. 9.3523 grm. solution saturated at 19°.5 gave 0.2932 grm. BaS04.
II. 8.3395 grm. solution saturated at 19°.5 gave 0.2534 grm. BaS04.

The solution saturated at 19.5° therefore contained the following
percentages of anhydrous salt.

I. n.

5.76 5.58

OF ARTS AND SCIENCES. 327

Calcic o Chlorpyro?nucatei Ca(C5H2Cl(X)2 . 3 H20. — The calcium
salt was prepared by boiling the acid with an excess of calcic car-
bonate. It is readily soluble in hot water, rather sparingly soluble in
cold water, and separates from a hot concentrated solution in clustered
prisms with rectangular terminations. The air-dried salt contains
three molecules of water. It effloresces slowly over sulphuric acid,
and loses all its water readily at 100°.

I. 1.0814 grm. air-dried salt lost at 100° 0.1480 grm. H,0.
II. 1.2247 grm. air-dried salt lost at 100° 0.1689 grm. H20.

Calculated for Found.

Ca(C5H2C103)2 . 3 H20. I. II.

H20 14.03 13.70 13.79

I. 0.6236 grm. salt dried at 100° gave 0.2564 grm. CaS04.
II. 0.5509 grm. salt dried at 100° gave 0.2244 grm. CaS04.

Calculated for

Found.

Ca(C6II2C103)2.

i.

II.

12.09

12.09

11.98

Ca

The solubility of the salt in cold water was determined according
to the method of V. Meyer. The calcium was precipitated as oxalate,
and the oxalate ignited with sulphuric acid.

I. 21.6494 grm. solution saturated at 19°. 5 gave 0.1002 grm. CaS04.
II. 17.7467 grm. solution saturated at 19°.5 gave 0.0809 grm. CaS04.

The solution saturated at 19°. 5 contained therefore the following
percentages of anhydrous salt.

i. n.

1.13 1.11

Potasstc 8 Chlorpyromucate, KC5H2C103. — The potassium salt is
readily soluble even in cold water, and separates from a hot concen-
trated solution on rapid cooling in fine branching needles ; on slower
cooling, in small thin oblique plates. The salt lost nothing in weight
when dried at 120°, and analysis showed it to be anhydrous.

I. 0.6543 grm. salt dried at 120° gave 0.3100 grm. K2S04.

II. 0.6390 grm. salt dried at 120° gave 0.3018 grm. K2S04.

Calculated for

Found.

KC5H,C103.

I.

II.

21.18

21.27

21.20

K

Argentic S Chlorpyromucate, AgC5H2C103. — The silver salt may
best be prepared by precipitating a solution of the barium salt with
argentic nitrate.

328 PROCEEDINGS OF THE AMERICAN ACADEMY

For analysis the precipitated salt was well washed, recrystallized
from hot water, and dried in vacuo over sulphuric acid. It is spar-
ingly soluble even in hot water, and separates from a hot concentrated
solution in irregularly branching flattened needles.

I. 0.4843 grm. substance gave 0.2730 grm. AgCl.
II. 0.3948 grm. substance gave 0.2220 grm. AgCl.

Calculated for Found.

AgC3H2C103. I. II.

Ag 42.60 42.44 42.34

Ethyl 8 C/tlorpyromucate, C5H2C103C2H5. — The ethyl ether was pre-
pared by heating at 100° a solution of 5 parts of 8 chlorpyromucic
acid in 10 parts of absolute alcohol, with the addition of 5 parts of
concentrated sulphuric acid (Sp. Gr. 1.84).

After heating for three hours, the ether was precipitated by the
addition of water, washed first with dilute sodic carbonate, then with
water, and finally dried with calcic chloride. The ether is a colorless
heavy oil boiling at 216-218° (column completely in vapor) under a
pressure of 76.8 mm., solidifying at 1-2°, and melting at the same
point.

I. 0.2740 grm. substance gave 0.2219 grm. AgCl.
II. 0.2918 grm. substance gave 0.2402 grm. AgCl.

Calculated for Found.

C6H2C103C2H5. I. II.

CI 20.34 20.02 20.35

8 Chlorpyromucamide, C5H2C10.,NH2. — 8 chlorpyromucic ether is
but slowly attacked by concentrated ammonia in the cold, and the
amide may more readily be prepared from the acid chloride by means
of solid ammonic carbonate. The amide is readily soluble in hot water,
sparingly soluble in cold water and separates from a hot concentrated
solution in fine branching needles, which melt at 154-155°.

For analysis, the amide was dried over sulphuric acid.

I. 0.1696 grm. substance gave 0.1677 grm. AgCl.
II. 0.2933 grm. substance gave 25.7 c.c. moist N at 25° and under
a pressure of 764 mm.

Found.

II.

9.82

Preparation I. was made from the ether, and II. from the acid
chloride.

CI

Calculated for
CEU2C102NH2.

24.40

I.

24.44

N

9.62

OF ARTS AND SCIENCES. 329

There seemed no room for doubt that the chlorpyromucic acid thus
formed by the direct replacement of hydrogen by chlorine, like the
8 brompyromucic acid obtained in a similar way, contained its chlorine
in the 8 position. We considered it necessary, however, to establish
this point by direct experiment.

Action of Bromine and Water.

Although the action of bromine on 8 brompyromucic acid suspended
in water varied greatly with the conditions, Hill and Sanger* found
that, when the oxidation was carefully conducted, fumaric acid alone
was formed. We therefore suspended the 8 chlorpyromucic acid in
about twenty-five times its weight of water, and passed in bromine
vapor slowly with a current of air. The acid gradually dissolved, and
but little more than two molecules of bromine were needed to com-
plete the oxidation. After standing for some time in the cold, the
solution was evaporated to a small volume, and the crystalline acid
which appeared as the solution cooled recrystallized from hot water.
In this way we obtained a white crystalline acid which contained
neither chlorine nor bromine, which was sparingly soluble in cold
water, more readily in hot water, and which remained unchanged
when heated to 200°. Although the acid was thus sufficiently char-
acterized as fumaric acid, we further analyzed its silver salt.

0.4943 grm. substance dried at 120° gave 0.5591 grm. AgBr.

Calculated for
Ag,C4H204. Found.

Ag 65.46 64.97

In the oxidation of the 8 chlorpyromucic acid by bromine water,
fumaric acid, had therefore been formed according to the equation :

C5H3C103 + 2 Br2 + 3 H20 = C4H404 + COa + 4 II Br + IIC1.

Action of Nitric Acid.

8 chlorpyromucic acid is not as readily attacked by nitric acid as the
8 brompyromucic acid, and for its complete oxidation we found it neces-
sary to heat one part of the acid with three parts of strong nitric acid
(Sp. Gr. 1.42) diluted with twice its weight of water. After heat-
ing for three hours the oxidation was completed, and on evaporation
fumaric acid was obtained, which was recognized by its sufficiently
characteristic physical properties.

* These Proceedings, xxi. 144.

330 PROCEEDINGS OP THE AMERICAN ACADEMY

(3 Chlorpyromucic Acid.

Hill and Sanger* found no difficulty in preparing ft brompyro-
mucic acid by the reduction of either of the two dibrornpyroniucic
acids described by them. The corresponding dichlorpyromucic acida
we found to be much more refractory, and we succeeded in preparing
the j3 chlorpyromucic acid only after many unsuccessful attempts.
The fSy dicblorpyromucic acid, which was the more accessible, proved
to be the more difficult of reduction, so that almost the whole of the
material for this investigation was made from the y8S dichlorpyromucic
acid. When zinc dust is added to a cold ammoniacal solution of this
acid, no perceptible reaction takes place, and even after long standing
no essential change can be detected. Even when the ammoniacal
solution is boiled with a large excess of zinc dust the reduction is but
slowly effected, and long continued heating is necessary to complete
the reaction. We have fouud it advantageous to dissolve the acid in
about twenty times its weight of dilute amnionic hydrate, to add an
equal weight of zinc dust, and to boil for eight or ten hours, taking
care to keep the solution strongly ammoniacal. The filtered solution
is then cooled and acidified with dilute sulphuric acid, the acid which
separates removed by filtration, the filtrate extracted with ether, and
the residue left upon the evaporation of the ether added to the main
product. The crude acid thus obtained melted at about 130°, and
evidently contained unaltered /38 dicblorpyromucic acid. We there-
fore dissolved it in dilute amnionic hydrate, and added to the ammoni-
acal solution calcic chloride as long as a sparingly soluble calcium salt
was promptly precipitated. The filtered solution then gave, when
acidified with hydrochloric acid, a crystalline acid, which after recrys-
tallization from hot water melted at 145-146°, and proved upon
analysis to be a chlorpyromucic acid. From the sparingly soluble
calcium salt, which had been removed by filtration, we obtained only
unaltered fi& dichlorpyromucic acid, which in its turn was treated with
zinc dust. In this way we found it possible to obtain from the /?S di-
chlorpyromucic acid at least 80 per cent of the theoretical amount of
pure /3 chlorpyromucic acid. Zinc dust, even after long boiling, seems
to have but little action upon /3y dichlorpyromucic acid in ammoniacal
solution, but sodium amalgam slowly reduces it at 100°. In order to
effect complete reduction, it proved to be necessary to use a large ex-
cess of sodium in the form of a one per cent amalgam, and the reaction
was then completed after heating for three hours on the water bath.

* These Proceedings, xxi. 147.

OF ARTS AND SCIENCES. 831

The acid obtained by acidification melted at 145-146°, contained the
percentage of chlorine required by a chlorpyromucic acid, and appeared
to be in all respects identical with the acid obtained from the /38 dichlor-
pyromucic acid. While the yield of pure acid was in this case but
50 per cent of the theoretical amount, it is possible that it might be
somewhat increased by further experiments. From the third iso-
meric dichlorpyromucic acid, subsequently described, we have as yet
been able to obtain no definite reduction product.

The {3 chlorpyromucic acid dried over sulphuric acid gave on analy-
sis the following results : —

I. 0.2710 grm. substance gave 0.4073 grm. C02 and 0.0513 grm. H20.
II. 0.1385 grm. substance gave 0.1355 grm. AgCl.
III. 0.1736 grm. substance gave 0.1692 grm. AgCl.

i.

in.

c

Calculated for
C6HaC103.

40.96

i.
40.99

Found.
II.

H

2.05

2.10

CI

24.23

24.18

24.10

Analysis III. was made with material prepared from the fiy dichlor-
pyromucic acid.

/3 chlorpyromucic acid is readily soluble in hot water, sparingly sol-
uble in cold water, and crystallizes from hot aqueous solution in thin
irregular striated plates or flattened prisms. It is readily soluble in
alcohol, ether, hot benzol or hot chloroform, and is but sparingly sol-
uble in cold benzol or cold chloroform. "When repeatedly crystallized
from hot water it melted at 145-146°, and recrystallization from other
solvents failed to raise this melting point.

In order to determine the solubility of the acid in cold water, a
weighed quantity of a solution of the acid saturated at 19°.8 was boiled
with baric carbonate, and the barium taken into solution precipitated
as sulphate.

I. 31.6300 grm. solution saturated at 19°.8 gave 0.201 9 grm. BaS04.
II. 34.0328 grm. solution saturated at 19.°8 gave 0.2176 grm. BaS04.

The solution saturated at 19°.8 therefore contained the following

percentages of the acid : —

i. n.

0.80 0.80

For the further characterization of the acid we prepared certain of
its salts.

332 PROCEEDINGS OP THE AMERICAN ACADEMY

Baric ft Chlorjvjromucate, Ba(C5H2C103)2 . H20. — The barium salt
was prepared by boiling a solution of the acid with an excess of baric
carbonate. The filtered solution was evaporated, and the salt which
separated on cooling recrystallized from hot water. It proved to be
readily soluble in hot water, more sparingly soluble in cold water, and
crystallized in long obliquely terminated prisms, which contained one
molecule of water. The salt is permanent in the air or over sulphuric
acid, but loses its water readily at 100°.

I. 0.6092 grm. of the air-dried salt gave 0.3180 grm. BaS04.
II. 1.3540 grm. of the air-dried salt lost at 100° 0.0576 grm. H20.

Calculated for Found.

Ba(C5H2C103)2.H20. I. II.

Ba 30.72 30.69

H20 4.04 4.25

0.4749 grm. of the salt dried at 100° gave 0.2587 grm. BaS04.

Calculated for
Ba(C5H2C103).,. Found.

Ba 32.01 32.03

The solubility of the salt in cold water was determined in the usual
way.

I. 10.2683 grm. solution saturated at 1 9°.l gave 0.1 1 49 grm. BaS04.
II. 6.8737 grm. solution saturated at 19°. 1 gave 0.0747 grm. BaSQ..
According to these determinations, the aqueous solution saturated at
19°. 1 contained the following percentages of the anhydrous salt: —

i. II.

2.06 2.00

Calcic /? Chlorpyromucate, Ca(C5H2C10„)2 . 3 H20. — This salt was
prepared by neutralizing with calcic carbonate a boiling solution of the
acid. If the solution thus obtained is evaporated upon the water bath
clustered, pointed prisms of an anhydrous salt are formed when the
solution becomes highly concentrated. By evaporation in vacuo over
sulphuric acid at ordinary temperatures, tufts of fine branching prisms
are obtained which contain three molecules of water. The latter salt
is permanent in the air, effloresces over sulphuric acid, and loses its
water readily at 105°.

I. 0.4226 grm. of the air-dried salt gave 0.1490 grm. CaS04.
II. 1.3627 grm. of the air-dried salt lost at 105° 0.1902 grm. H20.

Calculated for Found.

Ca(C6lI2C10.,),.3H20. I. II.

Ca 10.39 10.37

H20 14.03 13.95

OP ARTS AND SCIENCES. 333

0.5160 grm. of the salt dried at 105° gave 0.2111 grm. CaS04.

Calculated for
Ca(C5H2C103)3. Found.

Ca 12.09 12.03

The pointed prisms obtained by evaporating the solution at 100°
gave the following results.

0.4433 grm. of the air-dried salt gave 0.1796 grm. CaSOr

Calculated for

Cu(,C5II,C103)?. Found.

Ca 12.09 11.92

The solubility of the calcium salt in cold water was determined in
the usual manner. The calcium was precipitated as oxalate, and con-
verted into the sulphate before weighing.

I. 5.0960 grm. solution saturated at 19°. 5 gave 0.0653 grm. CaS04.
II. 5.8981 grm. solution saturated at 19°. 5 gave 0.0752 grm. CaS04.

According to these determinations, the aqueous solution saturated at
19°.5 contained the following percentages of the anhydrous salt: —

i. ii.

3.12 3.10

Ethyl (3 Chlorpyromucate, C.H2C103. C2H5. — The ethyl ether was
prepared by heating at 100° for four hours a mixture of 2 parts of the
acid. 3 parts of absolute alcohol, and 2 parts of sulphuric acid (Sp.Gr.
1.84). On dilution with water the ether separated as an oil, which
was thoroughly washed and dried over calcic chloride. It then dis-
tilled without essential decomposition at 217° (mercury column com-
pletely in vapor) under a pressure of 764 mm. The liquid distillate
solidified on cooling, with the formation of concentrically grouped
prisms, which, after repeated recrystallization by cooling the melted
ether, showed the constant melting point of 29-30°.

0.1864 grm. of substance gave 0.1523 grm. AgCl.

Calculated for
C6H2C103 . C2H6. Found.

CI 20.34 20.19

Action of Nitric Acid.

/3 chlorpyromucic acid is readily attacked by dilute nitric acid, but
the oxidation is not smoothly effected, and the yield of chlorfumaric
acid is comparatively small. The best results were obtained when the
acid was boiled with 2 parts of nitric acid (Sp. Gr. 1.42) diluted with

334 PROCEEDINGS OF THE AMERICAN ACADEMY

5 parts of water. After the lapse of two hours, the action appeared
to be completed, although a few drops of oil were still suspended in
the clear solution. The products of the reaction were then extracted
from the diluted solution with ether, the residue obtained by the
evaporation of the ether pressed thoroughly with filter-paper and
dried at 100°. The dry residue was then repeatedly washed with
benzol in which chlorfumaric acid is but sparingly soluble. The
product thus obtained was readily soluble in water, almost insoluble
in benzol, melted at 188°, and contained the percentage of chlorine
required by chlorfumaric acid.*

0.1089 grm. substance dried over H2S04 gave 0.1027 grm. AgCl.

Calculated for

C4H3C104. Found.

CI 23.59 23.31

The reaction may therefore be represented, in part at least, by the
equation

C5H3C103 + 30 = C4TI3C104 + C02.

Action of Bromine and Water.

Since j3 brompyromucic acid in aqueous solution is readily con-
verted by an excess of bromine into mucobromic acid,f it seemed to us
of interest to study the behavior of the (3 chlorpyromucic acid under
the same conditions, since a product containing bromine and chlorine
might then be formed. The acid was therefore suspended in five
times its weight of water, and an excess of bromine at once added.
After heating for a short time a clear nearly colorless solution was
obtained, which was concentrated by evaporation on the water bath.
On cooling, the solution deposited a crystalline acid which was readily
soluble in hot water or hot benzol, and but sparingly soluble in these
solvents in the cold. The acid recrystallized from water formed thin
rhombic plates which melted at 120-121°, but this melting point could
easily be raised to 121-122° by recrystallization from benzol.

An analysis of the substance dried over sulphuric acid showed that
it was the mucochlorbromic acid whose formation we had been led to
expect.

0.2871 grm. substance gave 0.4486 grm. AgCl + AgBr.

Calculated for
C4U2BrC103. Found.

Br+Cl 54.11 54.45

* Kauder (Journ. f. prakt. Chem., [2], xxxi. 28) gives the' melting point of
chlorfumaric acid as 191°.

t These Proceedings, xxi. 152.

OP ARTS AND SCIENCES. 335

The reaction was then precisely analogous to that by means of
which mucobromic acid was formed from (3 brompyromucic acid :

C5HX103 + 3 Br2 + 2 H20 = C4H2BrC103 + Co2 + 5 HBr.

With the material which for the moment was at our disposal, we were
unable to study the decomposition of the mucochlorbromic acid by
alkalies. We shall hope in the future to prepare in this way a chlor-
bromacrylic acid, and compare it with the acid of the same composi-
tion already described by Mabery and Loyd.*

(3y DlCHLORPYROMUCIC ACID.

Hill and Sanger f had shown that in the decomposition of pyromu-
cic tetrabromide or of ethyl pyromucic tetrabromide by alcoholic sodic
hydrate, two isomeric dibrompyromucic acids are formed in not widely
unequal quantities. There was, therefore, every reason to expect that
two isomeric dichlorpyromucic acids could be found in the product
formed in a similar way from ethyl pyromucic tetrachloride. Since
we had found that a low temperature was essential to the preparation
of a pure product, we allowed the ethyl pyromucate to absorb chlorine
at 0° until a constant weight was reached, expelled the excess of chlo-
rine by a current of dry air, and decomposed the product at once with
an excess of an alcoholic solution of sodic hydrate. The best results
were obtained when the tetrachloride was slowly added to a concen-
trated sodic hydrate solution, taking care to keep the mixture cold.

The sodium salts formed are sparingly soluble in alcohol, and after
a short time can be removed by filtration. The alcoholic solution was
freed from the excess of sodic hydrate by means of carbonic dioxide,
and, after removing the sodic carbonate by filtration, distilled. The
sodium salts left on distillation appeared to be identical with those
already obtained and were therefore added to the main portion. The
alcoholic distillate grew turbid when mixed with water, and gradually
deposited a small quantity of a colorless oil which had a peculiar aro-
matic odor. The quantity of this oil was so small that no investiga-
tion of it has as yet been made.

The sodium salt which was obtained from the tetrachloride was
dried, dissolved in hot water, and acidified with hydrochloric acid.
In this way an acid was obtained which crystallized in finely felted
needles which usually melted at 155°. After two recrystallizations
from benzol the acid melted at 108-169°, and further recrystalliza-

* These Proceedings, xvi. 238.
t These Proceedings, xxi. 150.

336 PROCEEDINGS OP THE AMERICAN ACADEMY

tions failed to raise this melting point. Analysis showed this acid to
be a dichlorpyromucic acid.

I. 0.4235 grm. substance gave 0.51 35 grm. C02 and 0.0455 grm. H20.
II. 0.2332 grm. substance gave 0.3695 grm. AgCl.
III. 0.2195 grm. substance gave 0.3490 grm. AgCl.

Calculated for Found.

CSH2C1,03. I. II. III.

C 33.15 33.07

H 1.10 1.19

CI 39.22 39.16 39.31

The acid, which proved on subsequent investigation to be the
(3y dichlorpyromucic acid, is sparingly soluble in cold water, readily
in hot, and crystallizes as the solution cools in finely felted needles.
It is readily soluble in alcohol, ether, or in hot benzol ; in cold benzol
it is sparingly soluble and crystallizes from it in short prisms. In hot
chloroform it is also readily soluble, sparingly soluble in the cold.
The ready purification of this acid by recrystallization rendered it
improbable that any sensible amount of an isomeric acid was formed
with it. The most patient search has failed to show the formation of
such a product in appreciable quantity, and from pure ethyl pyromucic
tetrachloride almost pure /3y dichlorpyromucic acid is obtained at once.
The yield of pure acid ordinarily obtained from pyromucic ether is not
wholly satisfactory, since it amounts to about 50 per cent of the weight
of ether taken, or about 39 per cent of the theoretical amount.

The solubility of the acid in cold water we determined as usual.
A weighed quantity of a solution of the acid saturated at 19.5° was
boiled with baric carbonate and the barium dissolved precipitated by
sulphuric acid.

I. 36.2505 grm. solution saturated at 19°.5 gave 0.0619 grm. BaS04.
II. 35.6546 grm. solution saturated at 19°.5 gave 0.0615 grm. BaS04.

According to these determinations, the solution saturated at 19°. 5
contains the following percentages of acid : —

i. ii.

■ 0.27 0.27

Baric fiy Dichlorpyromiicate, Ba(C5HCl203) . 3 H20.* — The barium
salt may be most conveniently prepared by precipitating a solution of the

* We have already referred to the brief description of the &y dichlorpyro-
mucic acid published by Denaro in the Gazzetta Chimica (xvi. 333), and have
asserted that the analyses of the two salts which he describes are not worthy of
confidence. In support of our assertion we append his results in full, together

OF ARTS AND SCIENCES. 337

ammonium salt with baric chloride. It is rather sparingly soluble even
in hot water, and still less soluble in cold water. It separates from a
hot concentrated solution in fine clustered needles which contain three
molecules of water. The crystals are permanent in the air, but lose
their water readily at 100°.

I. 1.2048 grm. air-dried salt lost at 100° 0.1170 grm. H20.
II. 1.2118 grm. air-dried salt lost at 100° 0.1132 grm. H20.

III. 0.5694 grm. air-dried salt gave 0.2393 grm. BaS04.

IV. 0.3783 grm. air-dried salt gave 0.1593 grm. BaS04.

Calculated for Found.

Ba(C5HCl203)2 . 3 H20. I. II. IH. IV.

H20 9.80 9.71 9.34

Ba 24.86 24.71 24.76

I. 0.4828 grm. salt dried at 100° gave 0.2250 grm. BaSOv
II. 0.5540 grm. salt dried at 100° gave 0.2579 grm. BaS04.

Calculated for Found.

Ba(C5HCl2Os)2. I. II.

Ba 27.56 ' 27.40 27.36

The solubility of the salt in water at 19°.5 was determined accord-
ing to the method of V. Meyer.

with the analytical data upon which they depend. The errors in the calculated
percentages are corrected in parenthesis.

Barium salt :
0.4480 grm. of the salt lost at 110° 0.00204 grm. H20.

Calculated for
Ba(C6HCl203)2 . 3 H20. Found.

H.20 9.80 8.30 (0.445)

0.1939 grm. of the dry salt gave 0.0769 grm. BaS04.

Calculated for

Ba(C5HCl,03)2. Found

Ba 23.54 (27.56) 23.32

Calcium salt:
0.6942 grm. of the salt lost at 110° 0.0300 grm. H20.

Calculated for
Ca(C5HClo03), . 3£ H20. Found.

H20 13.60 14.45 (4.32)

0.3024 grm. of the salt gave 0.1072 grm. CaS04.

Calculated for

Ca(C5HCl20,)2. Found.

Ca 10.00 10-42

These results can hardly be explained unless it is assumed that they con-
tain both typographical arid arithmetical errors. The two remaining analyses
contained in Denaro's paper, two chlorine determinations in the free acid are
correctly calculated, and agree well with the theory.
vol. xxiv. (n. s. xvi. ) 22

338 PROCEEDINGS OP THE AMERICAN ACADEMY

I. 23.0222 grm. solution saturated at 19°.5 gave 0.0503 grm. BaS04.
II. 22.4890 grm. solution saturated at 19°.5 gave 0.0487 grm. BaS04.

The solution saturated at 19.5° therefore contained the following per-
centages of anhydrous salt.

i. n.

0.46 0.46

Calcic /3y Dichlorpyromucate, Ca(C5HCl203)2 . 4 H20. — The calcium
salt was prepared by boiling a solution of the acid with an excess of
calcic carbonate. It is readily soluble in hot water, less soluble in
cold, and crystallizes from a hot concentrated solution in long clustered
needles which contain four molecules of water.

The crystallized salt is permanent in the air, effloresces slowly over
sulphuric acid, and loses all its water readily at 110°.

I. 2.7640 grm. air-dried salt lost at 11"0° 0.4160 grm. H20.
II. 3.1275 grm. air-dried salt lost at 113° 0.4707 grm. H20.

Calculated for Found.

Ca(C6HCK03)2 . 4 H20. I. II.

H20 15.26 15.05 15.05

I. 0.6595 grm. salt dried at 110° gave 0.2227 grm. CaS04.
II. 0.7004 grm. salt dried at 113° gave 0.2366 grm. CaS04.

Calculated for Found.

Ca(C6HCl.,03)2. I. II.

Ca 10.00 9.93 9.94

The solubility of the salt in water at 19.5° was determined as usual.
The calcium was precipitated as oxalate and the oxalate ignited with
sulphuric acid.

I. 22.8797 grm. solution saturated at 19°. 5 gave 0.0944 grm. CaS04.
II. 21.5915 grm. solution saturated at 19°. 5 gave 0.0891 grm. CaS04.

The solution saturated at 19°. 5 therefore contained the following
percentages of anhydrous salt:

i. n.

1.21 1.21

Potassic fiy Dichlorpyromucate, KC5HC1208. — The potassium salt
is readily soluble in hot water, but rather sparingly soluble in cold
water, and separates from a hot concentrated solution in small prisms
which are anhydrous.

I. 0.6130 grm. substance dried at 120° gave 0.2435 grm. K2S04.
II. 0.8200 grm. substance dried at 120° gave 0.3265 grm. K2S04.

Calculated for Found.

KCr>HCUOs. I. II.

K 17.84 17.83 17.87

OP ARTS AND SCIENCES. 339

Argentic fty Dichlorpyromucate, AgC5HCl203. — If argentic nitrate
is added to a cold aqueous solution of the acid, the silver salt is pre-
cipitated in the form of fine needles. For analysis 'we prepared the
salt by precipitating a dilute neutral solution of the ammonium salt
with argentic nitrate, and recrystallizing the precipitated salt from
hot water. It is sparingly soluble even in hot water, and separates on
cooling the hot saturated solution in fine needles.

I. 0.3230 grm. substance dried over H2S04 gave 0.1612 grm. Ao-Cl.
II. 0.2580 grm. substance dried over H2S04 gave 0.1282 grm. AgCl.

Calculated for Found.

AgC5HCl2Os. I. II.

Ag 37.50 37.57 37.41

Ethyl /3y Dichlorpyromucate, C5IIC1203C2H5. — The ethyl ether was
prepared by heating at 100° for three hours a mixture of 5 parts of
acid, 8 parts of absolute alcohol, and 5 parts of concentrated sulphuric
acid (Sp. Gr. 1.84). It was precipitated with water, washed with
dilute sodic carbonate, and then with water. It is readily soluble in
hot alcohol, more sparingly in cold, and crystallizes in large needles
which melt at 63-64°.

I. 0.2645 grm. substance dried over H2S04gave 0.3640 grm. AgCl.
II. 0.1933 grm. substance dried over H2S04 gave 0.2673 grm. AgCl.

Calculated for Found.

CfiHCljO^Hs. I. II.

CI 33.98" 34.02 34.18

fiy Dichlorpyrmucamide, C5HC1202NII2. — The ethyl ether of /?y
dichlorpyromucic acid is readily attacked by strong aqueous ammonia
even in the cold, and in a short time is converted into the amide.
The amide is sparingly soluble even in hot water, and crystallizes in
long slender needles which melt at 176-177°.

0.4015 grm. substance dried over II2S04 gave 28.5 c.c. moist N
at 23° under a pressure of 767 mm.

Calculated for

C3nCU02NII2. Fcund.

N 778 8.03

The formation of but one dichlorpyromucic acid in the decomposi-
tion of the ethyl pyromucic tetrachloride by alkalies, while two iso-
meric acids are always obtained under the same circumstances from
the corresponding bromine compound, made it impossible to predict its
constitution from the method of formation. Still, its comparatively
high melting point, its physical properties, and the ready solubility of

340 PROCEEDINGS OF THE AMERICAN ACADEMY

its calcium salt, in which it closely resembled the /?y dibrompyromucic
acid, gave fair ground for the conjecture that it had a similar struc-
ture. A study of its oxidation products showed that such was the
case, since we easily obtained from it mucochloric and dichlormaleic
acids.

Action of Bromine and Water.

If bromine is added to f3y dichlorpyromucic acid suspended in
six or eight times its weight of cold water, oxidation rapidly ensues,
with the escape of carbonic dioxide. A slight excess of bromine was
added, and the reaction completed by the aid of heat. On cooling, the
solution solidified with the separation of colorless crystals, which when
recrystallized from water and finally from benzol melted at 124—125°,
and had the characteristic form of mucochloric acid. Analysis also
gave the proper percentage of chlorine.

I. 0.2015 grm. substance dried over H2S04 gave 0.3408 grm. AgCl.
II. 0.2320 grm. substance dried over H2S04 gave 0.3922 grm. AgCl.

Calculated for Found.

C4H2C1203. I. II.

CI 42.01 41.81 41.79

In this case the reaction may be represented by the equation

C6H2C1203 + 2Br2 + 2 H20 = C4H2C1203 + C02 + 4 HBr.

Action of Nitric Acid.

Nitric acid acts but slowly upon the fty dichlorpyromucic acid, and
for its complete oxidation we have found it necessary to take for one
part of the acid 3 parts of concentrated nitric acid (Sp. Gr. 1.42)
diluted with twice its weight of water. After boiling for five hours
the action appeared to be complete, and as the clear solution deposited
nothing on cooling we extracted it with ether. The ether left upon
evaporation a colorless crystalline mass, which proved to be a mixture
of two substances, one readily soluble, the other but sparingly soluble
in cold water. The sparingly soluble substance when recrystallized
from hot water was recognized by its crystalline form and by its
melting point, 124-125°, as mucochloric acid. The acid, which was
readily soluble in cold water, was neutralized with baric carbonate,
the barium salt precipitated from its aqueous solution by alcohol, and
the acid liberated from this purified barium salt again extracted with
ether. The crystalline acid left by the evaporation of the ether was
then dissolved in a small amount of water, and on standing well

OF ARTS AND SCIENCES. 341

formed rhombic plates separated which were dried over sulphuric
acid for analysis.

0.2233 grm. substance gave 0.3433 grm. AgCl.

Calculated for
C4U2C1204. FouDd.

CI 38.38 38.01

The identity of this acid with dichlormaleic acid was further de-
termined by the melting point of the anhydride prepared by sublima-
tion, which we found to be 119-120°, in agreement with the statement
of Ciamician and Silber.*

The oxidation therefore took place in accordance with the following
reactions,

C5H2C1208 +20= C4H2CI208 -f C02,
C5H2C1203 +30 = C4H2C1204 + C02,

and the dichlorpyromucic acid in question is thus conclusively shown
to have its chlorine atoms in the j3 and y positions.

Preparation of Isomeric Dichlorpyromucic Acids.

Although we had been unable to find the (38 dichlorpyromucic acid
among the products formed by the action of alkalies upon ethyl
pyromucic tetrachloride, it seemed to us probable that it might be
formed under the proper conditions by the direct action of chlorine
upon ethyl pyromucate, or by the decomposition of its tetrachloride by
heat alone.

Tunnies f had already shown that 8 brompyromucic acid could be
formed by heating pyromucic tetrabromide, and Hill and Sanger $ had
further shown that 8 brompyromucic acid and /38 dibrompyromucic
acid could conveniently be made by the action of bromine upon pyro-
mucic acid at high temperature. As subsequent experiments in this
laboratory had shown that these two acids could more advantageously
be made by substituting the ethyl ether for the acid, and our own
experiments had shown that the S chlorpyromucic acid could readily
be made in this way, we proceeded to study the action of chlorine
upon ethyl pyromucate at high temperature. It seems hardly neces-
sary to describe in detail the numerous experiments which we made
under widely varying conditions, each one of which yielded us purely
negative results. Ethyl pyromucate was treated with chlorine at

* Berichte d. deutsch. chem. Gesellscb., xvi. 2390.

t Ibid., xi. 1088.

| These Proceedings, xxi. 130, 100.

342 PROCEEDINGS OP THE AMERICAN ACADEMY

temperatures ranging from 145° to its boiling point, both by itself and
after the addition of iodine or aluminic chloride, but in no case was
the desired product obtained. Pyromucic acid and 8 chlorpyroniucic
acid also failed to give such a product when treated at high tempera-
tures with chlorine, and we were equally unsuccessful when we used
the chloranhydrides of these acids either with or without an excess of
phosphoric pentachloride. We then studied the action of heat upon
the tetrachloride of pyromucic acid itself, of its chloranhydride and
ethyl ether, and of the ethyl ether of S chlorpyromucic acid, and found
in each case that small quantities of a dichlorpyromucic acid were
formed which closely resembled the (38 dibrompyromucic. A more
careful investigation further proved that in the decomposition of the
ethyl pyromucic tetrachloride a second new dichlorpyromucic acid was
formed concerning whose constitution we are not yet able to speak
definitely. This acid we propose to call the ^ dichlorpyromucic acid
until its structure is established. The amount of the dichlorpyromucic
acids which we have been able to obtain is but small, and we have
made many unsuccessful attempts to increase the yield by varying the
temperature or the mode of heating, or by adding iodine, or aluminic
or ferric chloride, before heating.

We first obtained the (38 dichlorpyromucic acid by distilling under
ordinary pressure the product formed by treating pyromucic acid with
chlorine at 100°. Hydrochloric acid is given off in quantity, and,
although a large carbonaceous residue is left in the retort, a liquid
distillate is obtained which, after repeated distillation through a
Hempel's column, amounts to about 85 per cent of the weight of
the pyromucic acid taken and then distils leaving but an insignificant
carbonaceous residue. This distillate is extremely complex in its
nature and we have as yet made no thorough study of its constituents.
We found, however, that the portions which boiled between 196° and
220° gave considerable quantities of (38 dichlorpyromucic acid when
treated with cold water. They therefore contained the corresponding
chloranhydride of the acid. The yield thus obtained amounted to
but about 4 per cent of the pyromucic acid taken. From ethyl
pyromucic tetrachloride we succeeded in obtaining a somewhat better
yield. We found it advantageous to purify the tetrachloride by one
distillation in vacuo, and to distil the product thus obtained under
ordinary pressure. More or less carbonization ensued, and on frac-
tional distillation in vacuo through a Hempel's column * the distillate

* For fractional distillation under diminished pressure we used the extremely
convenient apparatus of Anschiitz. We found that the ease of separation could

OF ARTS AND SCIENCES. 343

was found to contain considerable unaltered tetrachloride. That
portion which boiled above 140° under 16 ram. pressure was there-
fore redistilled under ordinary pressure and the distillate again frac-
tioned in vacuo. After repeated distillations we found that the portion
which boiled between 118° and 123° under 16 mm. pressure partially
solidified on cooling, and that a few crystals were also formed in the next
lower fraction. These fractions were therefore strongly cooled and the
crystalline solid removed by filtration and the liquid portions further
distilled. The solid thus obtained proved to be the ethyl ether of the
new x dichlorpyromucic acid melting at 197-198°, which we shall
presently describe. When no more of this crystalline ether could be
obtained by cooling, the liquid fractions were saponified by alcoholic
sodic hydrate, the acids liberated by hydrochloric acid and separated
through their calcium and barium salts. From the fraction boiling
below 110° (16 mm.) we obtained chiefly 8 chlorpyromucic acid
melting at 176-177°, although it yielded also a small quantity of
(38 dichlorpyromucic acid, which was readily isolated by means of its
sparingly soluble barium salt. The fractions 110-118° and 118-123°
apparently consisted chiefly of the ethyl ether of j3d dichlorpyromucic
acid, but the latter necessarily contained also a small amount of the
crystalline ethyl % dichlorpyromucate held in solution. The two
dichlorpyromucic acids could readily be separated through the differ-
ent solubilities of their calcium salts, the calcic (38 dichlorpyromucate
like the calcium salt of the corresponding bromine derivative being
very sparingly soluble in water. From 123° to 130° but an insignifi-
cant fraction was collected and the fraction 130°-153° consisted in
part at least of unaltered ethyl pyromucic tetrachloride, from which
j3y dichlorpyromucic acid was obtained.

100 grin, of ethyl pyromucate yielded us 197 grm. of the tetrachlo-
ride and from this we obtained the following weights of pure products :

be materially increased by filling a few inches of the stem of the distilling flask
with glass beads, and thus combining the Hempel's column with the vacuum
distillation. The beads were supported upon a perforated disk of platinum foil
slipped over the capillary air tube and held in place by a slight enlargement of
the tube. The thermometer was then raised to the proper line by a short bit
of small glass tubing dropped into the air tube. Ilantzsch (Ann. Chem. u.
Pharm., ccxlix. 57) has used with advantage for distillation under ordinary
pressure a long necked boiling flask whose stem is partially filled with beads
supported upon a platinum foil forced into the neck. We have for a long time
used such a Hempel's column, but have supported the beads conveniently upon
a glass bulb which nearly fills the stem and whose sealed neck is long enough
to rest upon the bottom of the flask.

344 PROCEEDINGS OP THE AMERICAN ACADEMY

5.0 grm. 8 chlorpyromucic acid.
12.2 grm. /38 dichlorpyromucic acid.

2.7 grm. x dichlorpyromucic acid.
16.1 grm. ethyl x dichlorpyromucate.

If pyromucyl chloride is treated with chlorine at 0°, the chlorine is
absorbed very slowly, but the gain in weight finally approximately cor-
responds with that required for the formation of a tetrachloride. If
the tetrachloride is distilled under ordinary pressure, only an inconsid-
erable carbonaceous residue is left in the retort, and after repeated
distillations through a Hempel's columu under ordinary pressure
a product is obtained which, when treated with water, yields the
/J8 dichlorpyromucic acid and the x dichlorpyromucic acid. While
the yield of the /SS dichlorpyromucic acid thus obtained is somewhat
greater than that obtained from the ethyl pyromucic tetrachloride, the
yield of the x dichlorpyromucic acid is much smaller.

By treating ethyl 8 chlorpyromucate in the cold with chlorine, and
distilling the addition product thus formed under ordinary pressure, we
also obtained the (38 dichlorpyromucic acid. While the yield was
somewhat larger than that obtained from the ethyl pyromucic tetra-
chloride, it did not repay us for the loss of time and material involved
in making the 8 chlorpyromucic acid. Moreover, to our surprise, we
could obtain in this way none of the crystalline ethyl x dichlorpyro-
mucate.

/38 Dichlorpyromucic Acid.

The acid whose preparation has just been described can most readily
be purified by repeated precipitation from ammoniacal solution with
calcic chloride, and recrystallization from chloroform. The acid thus
purified gave on analysis the following results :

I. 0.2577 grm. substance gave 0.3129 grm. C0.2 and 0.0259 grm. H20.
II. 0.1933 grm. substance gave 0.3053 grm. AgCl.

Calculated for Found.

C5U2C103. I. II.

C 33.15 33.11

H 1.10 1.12

CI 39.22 39.04

/3S dichlorpyromucic acid is readily soluble in ether or alcohol. It
dissolves readily in hot water, benzol, or chloroform, and the greater
part of the acid is in each case deposited on cooling in oblique prisms,
which are frequently twinned in forms which can hardly be distin-

OF ARTS AND SCIENCES. 345

guished from those of the (38 dibrompyromucic acid. The acid melts
sharply at 155-156°, and sublimes unaltered at a higher temperature.
The solubility of the acid in water at 19°. 5 was determined in the
usual manner. A weighed quantity of a solution of the acid saturated
at 19°. 5 was boiled with baric carbonate, and the barium dissolved
determined by precipitation with sulphuric acid :

I. 42.8007 grm. solution saturated at 19°.5 gave 0.0724 grm. BaS04.
II. 51.4109 grm. solution saturated at 19°.5 gave 0.0907 grm. BaS04.

According to these determinations, the solution saturated at 19°. 5
contained the following percentages of acid :

I. II.

0.26 0.27

Baric (38 Dichlorpyromucate, Ba(C5HCl.203)2. 4 H20. — The barium
salt was prepared by precipitating a dilute solution of the ammonium
salt with baric chloride, and recrystallizing the product thus obtained
from water. The salt is sparingly soluble in hot water, still less solu-
ble in cold water, and crystallizes from a hot concentrated solution in
long irregular flat prisms which contain four molecules of water.

The salt is permanent in the air, effloresces over sulphuric acid, and
loses its water readily at 100°.

I. 1.4513 grm. air-dried salt lost at 100° 0.1795 grm. H20.

II. 0.4908 grm. air-dried salt gave 0.1981 grm. BaS04.

Calculated for Found.

Ba(C5HCl203)2 .4H20. I. II.

H20 12.66 12.37

Ba 24.07 23.73

0.4751 grm. salt dried at 100° gave 0.2205 grm. BaS04.

Calculated for

Ba(C5HCl203)j. Found.

Ba 27.56 27.29

The solubility of the salt in water at 19.5° was determined in the
usual manner.

I. 28.4671 grm. solution saturated at 19°.5 gave 0.0561 grm. BaS04.
II. 28.6844 grm. solution saturated at 19°.5 gave 0.0573 grm.BaS04.

The solution saturated at 19°.5 therefore contained the following
percentages of anhydrous salt :

I. ii.

0.42 0.43

346 PROCEEDINGS OF THE AMERICAN ACADEMY

Calcic /38 Dichlorpyromucate, Ca(C5HCl.203)2 . 3 H20. — Tlie cal-
cium salt was prepared by precipitating a dilute solution of the ammo-
nium salt with calcic chloride, and recrystallizing the product from hot
water. It is but sparingly soluble in hot water, still less soluble in
cold water, and separates from a hot concentrated solution in flattened
prisms with rectangular terminations. The crystallized salt is perma-
nent in the air, effloresces slowly over sulphuric acid and loses all its
water at 125°.

I. 1.5490 grm. air-dried salt lost at 125° 0.1820 grm. H20.
II. 0.4067 grm. air-dried salt gave 0.1219 grm. CaS04.

Calculated for Found.

Ca(C5HCl,03)2 . 3 H20. I. II.

H,0 11.90 11.75

Ca 8.81 8.81

0.6899 grm. salt dried at 125° gave 0.2332 grm. CaS04.

Calculated for

Ca(C6HCl203)2. Found.

Ca 10.00 9.94

The solubility of the salt in water at 19°. 5 was determined in the
usual manner. The calcium was precipitated as oxalate and ignited
with sulphuric acid.

I. 32.1733 grm. solution saturated at 19°. 5 gave 0.0244 grin. CaS04.
II. 36.7555 grm. solution saturated at 19°.5 gave 0.0289 grm. CaS04.

The solution saturated at 19°. 5 therefore contained the following
percentages of anhydrous salt :

I. IT.

0.22 0.23

Ethyl /3S Dichlorpyromucate, C5HC1203C2H5. — The ethyl ether was
prepared by the action of ethyl iodide on the silver salt, and is a heavy
liquid of pleasant odor. It melts at 2-3° and distils at 116-118°
under a pressure of 16 mm., the oil bath being at 165-175°. An
analysis of the distilled ether gave the following result.

0.2157 grm. substance gave 0.2957 grm. AgCl.

Calculated for Found.

C6HC1203C2H5.

CI 33.98 33.89

/3S Dichlorpyromucamide, CSHC1202NH2. — The ethyl ether is but
slowly attacked by concentrated ammonia in the cold, but at 100° it is
readily converted to the amide. The amide is sparingly soluble in

OF ARTS AND SCIENCES. 347

cold water, more readily in hot, and separates from a hot concentrated
solution in long needles which on standing are converted into oblique
prisms which melt at 153-154°.

0.2781 grm. substance dried over H2S04 gave 20.0 c.c. moist N at
25.5° under a pressure of 766 m.m.

Calculated for

C5HC1302NH2. Found.

N 7.78 8.06

Although the melting point, the crystalline form, and the insolubility
of its calcium salt showed that this acid closely resembled the (38
dibrompyromucic acid, it was evidently necessary to establish the posi-
tion of the chlorine atoms by means of its oxidation products.

Action of Bromine and Water.

Hill and Sanger* found that the (38 dibrompyromucic acid was
readily. attacked by aqueous bromine in the cold with the formation
of monobrommaleyl bromide. The (38 dichlorpyromucic acid is but
slowly attacked by aqueous bromine in the cold, and we therefore sus-
pended it in five times its weight of water, added at once somewhat
more than four atoms of bromine, heated until the oil which was first
formed had almost all disappeared, and evaporated the solution at a
gentle heat. The crystalline mass thus obtained was readily soluble
even in cold water, and very sparingly soluble even in hot chloroform
or benzol. The product was dissolved in cold water, filtered, and
evaporated nearly to dryness. The crystals which separated were
pressed dry with filter paper, washed carefully with hot benzol, and
recrystallized from a little hot water. The acid thus prepared crys-
tallized in microscopic crystals, which melted at 189-1 90°,t and proved
on analysis to contain the percentage of chlorine required by mono-
chlorfumaric acid.

0.1288 grm. substance dried over H2S04 gave 0.1227 grm. AgCl.

Calculated for

C4H,C104. Found.

CI 23.59 23.55

The oxidation with bromine and water therefore takes place in
accordance with the following equation :
CLBLCLO, + 2Br2 + 3H.20 = C4II,C104 + C02 + HC1 + 4 HBr.

* These Proceedings, xxi. 165.

t Kauder (Journal fur prakt. Cheinie, [2], xxxi. 28) gives the melting point

as lfJl°.

348 PROCEEDINGS OP THE AMERICAN ACADEMY

Since the (38 dichlorpyromucic is but slowly attacked even by
concentrated nitric acid, it was not thought worth while to study the
reaction in detail.

X Dichlorpyromucic Acid.

The formation of the ethyl ether of this acid by the decomposition
of ethyl pyromucic tetrachloride has already been described. The
ether was recrystallized from hot alcohol, and saponified by alcoholic
sodic hydrate. On the addition of hydrochloric acid to the aqueous
solution of the sodium salt, a sparingly soluble acid separates, which
may easily be purified by recrystallization from hot water.

The acid dried over sulphuric acid gave on analysis the following
results : —

I. 0.2196 grm.substance gave 0.2672 grm. C02 and 0.0237 grm. H20.
II. 0.2212 grm. substance gave 0.3492 grm. AgCl.

Found.

II.

c

Calculated for
C5II2C1203.

33.15

I.
33.18

H

1.10

1.20

CI

39.22

39.03

X dichlorpyromucic acid is readily soluble in ether or alcohol, and
but sparingly soluble in cold water. In hot water it is freely soluble,
and crystallizes as the solution cools in long needles which melt at
197-198°. The acid readily sublimes unchanged below its melting
point. Sodium amalgam slowly reduces it to pyromucic acid melting
at 129-130°.

The solubility of the acid in water at 19°. 5 was determined in
the usual manner. A solution of the acid saturated at 19°. 5 was
boiled with baric carbonate, and the barium dissolved determined by
precipitation with sulphuric acid.

I. 38.2670 grm. solution saturated at 19°.5 gave 0.0322 grm. BaS04.
II. 37.5125 grm. solution saturated at 19°. 5 gave 0.0301 grm. BaS04.

The solution saturated at 19°. 5 therefore contained the following
percentages of acid : —

i. n.

0.13 0.12

Baric x Dichlorpyromucate, Ba(C5HCl203)2 . 4 H20. — The barium
salt may most readily be prepared by boiling the acid with baric car-
bonate. The salt is quite readily soluble in cold water, more soluble
in hot water, and crystallizes in columnar aggregations of prisms which

OF ARTS AND SCIENCES. 349

contain four molecules of water. The crystallized salt is stable in the
air, effloresces over sulphuric acid, and loses all its water at 100°.

I. 1.0471 grm. air-dried salt lost at 100° 0.1285 grra. H20.
II. 0.5464 grm. air-dried salt gave 0.2235 grm. BaS04.

Calculated for Found.

Ba(C5HCl203)2 . 4 H20. I. II.

H20 12.66 12.27

Ba 24.07 24.05

0.4495 grm. salt dried at 100° gave 0.2091 grm. BaS04.

Calculated for

Ba(C5IICl203)2. Found.

Ba 27.56 27.35

The solubility of the salt in water at 19°.5 was determined in the
usual manner.

I. 11.1910 grm. solution saturated at 19°.5 gave 0.0820 grm. BaS04.
II. 8.6544 grm. solution saturated at 19°. 5 gave 0.0644 grm. BaS04.

According to these determinations, the solution saturated at 20°
contained the following percentages of anhydrous salt: —
I. ii.

1.56 1.59

Calcic x Dichlorpyromucate, Ca(C5HCl203)2. 4 H20. — The calcium
salt was prepared by boiling the acid with calcic carbonate. The fil-
tered solution was then concentrated on the water bath to a small vol-
ume, and the salt which separated was filtered out and washed with a
little water. Since the salt appeared to be about as soluble in cold
water as in hot, it was dissolved in water and the solution concen-
trated in vacuo over sulphuric acid. The salt which separated was
then pressed dry with filter paper. It is quite readily soluble in water,
and crystallizes in prisms vhich contain four molecules of water. The
crystallized salt is permanent in the air, effloresces slowly over sul-
phuric acid, and loses all its water at 117°.

I. 0.9075 grm. air-dried salt lost at 117° 0.1380 grm. H20.
II. 0.6080 grm. air-dried salt gave 0.1746 grm. CaSOv

Calculated for Found.

Ca(C5HClo03), . 4 HjO. I. n.

H20 15.25 15.21

Ca 8.48 8.44

0.3377 grm. salt dried at 117° gave 0.1152 grm. CaSOv

Calculated for
Ca (C5HCU03),. Found.

Ca 10.00 10.02

350 PROCEEDINGS OF THE AMERICAN ACADEMY

The solubility of the salt in water at 19°. 5 was determined in the
usual manuer. The calcium was precipitated as oxalate and iguited
with sulphuric acid.

I. 6.5743 grm. solution saturated at 19°.5 gave 0.1522 grm. CaS04.
II. 7.2115 grm. solution saturated at 19*°. 5 gave 0.1689 grm. CaS04.

The solution saturated at 19°. 5 therefore contained the following
percentages of anhydrous salt:

i. ii.

6.81 6.89

Ethyl x Dichlorpyromucate, C5HC1203C2H.. — The process by which
the ethyl ether was obtained has already been described. It is readily
soluble in hot alcohol, sparingly soluble in cold alcohol, and crystallizes
in clustered prisms with rectangular truncations which melt at 72-73°.
A sample of the ether melting at 72-73° was distilled under a pressure
of 16 mm. and boiled constant at 12 2°. 5, temperature of bath 170-175°.
For analysis the ether was dried over sulphuric acid.

I. 0.2034 grm. substance gave 0.2790 grm. AgCl.

II. 0.2302 grm. substance gave 0.3160 grm. AgCl.

Calculated for Found.

C6HC1203C2H6. I. n.

CI 33.98 33.91 33.94

In order to determine the constitution of this dichlorpyromucic acid
we studied its oxidation products with aqueous bromine.

Action of Bromine and Water.

As it was desirable to compare the oxidation product obtained from
this acid with that obtained from the fih dichlorpyromucic acid under
the same conditions, the acid was suspended in five times its weight of
water, somewhat more than four atoms of bromine added as rapidly as
possible, and the solution boiled until the oil which was at first formed
had nearly disappeared. The filtered solution was evaporated to dry-
ness at a gentle heat, the crystalline residue dissolved in a little cold
water, filtered, and again evaporated nearly to dryness. The acid
which separated was readily soluble even in cold water, almost insolu-
ble in hot chloroform or benzol. It was therefore pressed dry with
filter paper and carefully extracted with hot benzol. The acid as thus
prepared crystallized in microscopic crystals which melted at 188-189°,
and moreover gave on analysis the percentage of chlorine required by
monochlorfumaric acid.*

* KauJer, loc. cit.

OF ARTS AND SCIENCES. 351

0.1441 grm. substance dried over H2S04 gave 0.1377 grm. AgCl.

Calculated for

C4H3C104. Found.

CI 23.59 23.62

The x dichlorpyromucic acid, like the (38 dichlorpyromucic acid,
therefore gives with aqueous bromine as tbe chief product chlorfumaric
acid according to the equation

C5H2C1203 + 2 Br2 + 3 H20 = C4H3C104 + C02 + HC1 + 4 HBr.

Since the oxidation with bromine had shown with sufficient precision
that one of the chlorine atoms of the x dichlorpyromucic acid occupied
the 8 position, we thought it unnecessary for our purposes to study
also the action of nitric acid.

Action of Fuming Sulphuric Acid.

The two isomeric dibrompyromucic acids had shown marked differ-
ences in their behavior toward fuming sulphuric acid.* While the
/Sy acid had rapidly been converted into the corresponding sulphonic
acid, brommaleic acid in nearly theoretical quantity had been formed
from the (38 dibrompyromucic acid, and no trace of a sulphonic acid
could be detected. Preliminary experiments proved that the /?y and
/8S dichlorpyromucic acids showed a similar difference in their behavior
toward fuming sulphuric acid. While the formation of chlormaleic
acid in the latter case could not definitely be pi'oved, certainly no
sulphonic had been formed, and it seemed to us of interest to study
also the behavior of the third isomeric dichlorpyromucic acid under
the same conditions, x dichlorpyromucic acid dissolves readily in
fuming sulphuric acid, and if care be taken to prevent any marked
elevation of temperature no very essential decomposition ensues,
although a slight effervescence is noticeable. After the solution of
the acid in four times its weight of fuming sulphuric acid had stood for
thirty-six hours, it was poured into a large amount of cold water, the
solution cooled, and thoroughly extracted with ether. The ethereal
extract left on evaporation a small quantity of a colorless oil which
reduced silver oxide, and whose vapor vigorously attacked the eyes
and nose. The amount of oil thus obtained was wholly insufficient
for further study. On standing, it deposited a few clustered needles,
which probably were unaltered acid, although they may possibly hive
been the decomposition product subsequently described. From the
aqueous solution the barium salt was prepared in the usual way, and,

* These Proceedings, xxiii. 218.

352 PROCEEDINGS OF THE AMERICAN ACADEMY

since the properties of the neutral salt were unfavorable to purifica-
tion, it was converted into the acid salt, and this recrystallized from
hot water. The acid salt was readily soluble in hot water, more spar-
ingly soluble in cold water and crystallized in triclinic (?) prisms which
effloresced on exposure to the air. It gave an excellent qualitative
reaction for sulphur, and when dried over sulphuric acid* gave the
percentage of barium required by the formula Ba(C.HCl2SOc)2.

0.3886 grm. of the salt dried over sulphuric acid gave 0.1338 grm.
BaS04.

Calculated for
Ba(C5HCl2S06)2. Found.

Ba 20.84 20.24

Baric x Dichlorsulphopyromucate, BaC5Cl2S06 . 2 H20. — From the
acid barium salt we prepared the neutral salt by neutralizing its aque-
ous solution with baric carbonate. Since the hot saturated solution
deposited little or nothing on cooling, it was evaporated in vacuo over
sulphuric acid. The salt then crystallized in sheaves of prisms which
appeared to be triclinic. It was permanent in the air, effloresced over
6ulphuric acid, and lost its water completely at 160°.

1.4159 grm. of the air-dried salt lost at 160° 0.1277 grm. H20.

Calculated for
BaC5Cl2S06 . 2 H20. Found.

H20 8.33 9.02

0.5296 grm. of the salt dried at 160° gave 0.3126 grm. BaS04.

Calculated for
BaC5Cl2S06. Found.

Ba 34.59 34.69

The formation of a dichlorsulphopyromucic acid by the action of
fuming sulphuric acid upon the x dichlorpyromucic acid is thus suffi-
ciently established. The bromsulphop3Tomucic acids are so readily
reduced in alkaline solution that we hoped to be able to prepare from
this dichlorsulphopyromucic acid the corresponding sulphopyromucic
acid, and thus establish the position of the two chlorine atoms. We
soon found, however, that the chlorine was held with unusual persist-
ence, and with the material at our disposal we have as yet been unable
to reach decisive results.

* The single determination of the water of crystallization was unfortunately
defective. It gave 13.34 per cent of water in the salt dried by short exposure
to the air, while a salt crystallizing with 2£ molecules of water should contain
13.70 per cent.

OF ARTS AND SCIENCES. 353

Decomposition by Hydrochloric Acid.

If the x dichlorpyromucic acid is heated with water in a sealed tube
to 170°, no change is effected, but a reaction which we had in no way
anticipated takes place if it is heated upon the water bath in an open
flask with concentrated hydrochloric acid. Carbonic acid is evolved,
and in a short time the acid is completely decomposed with the forma-
tion of a neutral body which volatilizes when the solution is boiled,
and which can be extracted, although with difficulty, from the distil-
late, or from the original solution with ether. The ethereal solution
left on evaporation a white crystalline solid which was sparingly soluble
in water, readily soluble in cold chloroform or benzol, and but sparingly
soluble in ligroin. When recrystallized from ligroin it formed long
slender lustrous prisms, which melted at 52-53°, and sublimed rapidly
at ordinary temperatures. It reduced argentic oxide on warming, and
dissolved in aqueous alkalies, forming a yellow solution. The physical
properties and the behavior of this substance at once recalled to our
minds the crystalline body melting at 77° which Hill and Sanger*
had obtained in small quantity from the by-products of the decompo-
sition of pyromucic tetrabromide by alcoholic sodic hydrate. The
formula of this body had been shown to be C4H3Br02 and an analysis
of the new substance left no doubt of its similar composition.

0.1045 grm. substance gave 0.1261 grm. AgCl.

Calculated for

C4H3C102. Found.

CI 29.95 29.84

This interesting body is probably formed according to the reaction

C5H2C1203 + H20 = C4H3C102 + C02 + HC1.

And since we found little difficulty in obtaining 40 per cent of the
yield which this equation demands, it will be possible to study it more
in detail. Unfortunately, it was discovered so late in our work as to
make it impossible to present the results of such a study in this paper.
While it would be easy for us to venture a conjecture as to its struc-
ture, we prefer to await the results of a future investigation.

Trichlorpyromucic Acid.

For the preparation of the trichlorpyromucic acid it was evidently
most convenient to decompose with alkalies the tetrachloride of the
& chlorpyromucic acid. It did not seem necessary, however, to pre-

* These Troceerfings, xxi. 158.
vol. xxiv. (n. s. xvi.) 23

354 PROCEEDINGS OF THE AMERICAN ACADEMY

pare this tetrachloride in a pure condition, and indeed we first obtained
the trichlorpyromucic acid from one of the earlier preparations of the
ethyl pyromucic tetrachloride in which an unusually great spontaneous
elevation of temperature had taken place through the rapid absorption
of chlorine. For its preparation we heated the ethyl pyroinucate to
145°, and passed in chlorine at this temperature until the gain in
weight showed that one atom of hydrogen had been replaced by
chlorine. We then allowed the temperature to fall to about 120°.
and continued the chlorination to saturation. The total train iu
weight then corresponded approximately to that required by the
formation of the tetrachloride of the ethyl chlorpyromucate. On
decomposing this product as usual with a cold concentrated alcoholic
solution of sodic hydrate, the alcoholic solution filtered from the
insoluble sodium salts contained, as in the previous cases, small quan-
tities of liquid furfuran derivatives, but the amount was so minute that
no separate study of them was made. The sodium salts dissolved iu
hot water gave with hydrochloric acid an impure trichlorpyromucic
acid as a more or less colored oil, which solidified as the solution
cooled. For the purification of the acid we have found it convenient
to take advantage of the slight solubility of the ammonium salt in cold
water. The crude acid was suspended in about thirty times its weight
of water, ammonic hydrate added in excess, and the hot solution treated
with bone-black. The filtered solution deposits on cooling the greater
part of the trichlorpyromucic acid as the ammonium salt from which
the pure acid can readily be obtained. The small amount of trichlorpy-
romucic acid remaining in the ammoniacal solution may be recovered,
although at the expense of considerable trouble, by precipitation with
calcic chloride and repeated recrystallization from water and dilute
alcohol of the acid obtained from the insoluble calcium salt.

The yield of pure trichlorpyromucic acid was far from satisfactory,
as we could obtain only 30 per cent of the weight of the ethyl pyromu-
cate taken, but 15 per cent of the theoretical amount.

For analysis the acid was dried over sulphuric acid.

I. 0.3214 grm. substance gave 0.3248 grm.C02 and 0.021 1 grm. H20.
II. 0.2107 grm. substance gave 0.4192 grm. AgCl.
III. 0.2135 grm. substance gave 0.4255 grm. AgCl.

i

in.

49.27

c

Calculated for
CfiIICl303.

27.85

i.
27.56

Found
II.

II

0.46

0.73

CI

49.43

49.U

OP ARTS AND SCIENCES. 355

Trichlorpyromucic acid is readily soluble in alcohol or ether, quite
readily soluble in boiling benzol, and but sparingly soluble in cold
benzol. Hot water dissolves it but sparingly, and as the solution cools
most of the acid is deposited in microscopic needles which melt at
172-173°. This melting point was so much below that which we had
been led to expect from analogy to the known acid containing bromine
that we felt some doubt of its correctness, more especially since the
crude acid was so far from pure. A sample of the acid melting at
172-173° was therefore recrystallized twice from water and then three
times from benzol without perceptibly changing the melting point.
The acid was then converted into the calcium salt, and this separated
by crystallization into three successive fractions. The acid from
these three fractions melted simultaneously and sharply at 172-173°.
Finally, the ethyl ether was made, and the acid prepared from the
repeatedly recrystallized pure ether melted at the same point.

The solubility of the acid in water at 19°. 5 was determined by boil-
ing with baric carbonate a weighed quantity of a solution of the acid
saturated at that temperature, and determining the barium dissolved as
sulphate.

I. 49.1742 grm. solution saturated at 19°.5 gave 0.0344 grm. BaS04.
II. 47.0228 grm. solution saturated at 19°.5 gave 0.0331 grm. BaSo4.

The solution saturated at 19.5° therefore contained the following
percentages of acid :

i. ii.

0.13 0.13

Baric Trichlorpyromucate, Ba(C5Cl.,03)2 . 4 H20. — The barium salt
was prepared by precipitating a dilute solution of the ammonium salt
with baric chloride, and recrystallizing the sparingly soluble salt thus
thrown down from hot water. The salt proved to be but sparingly
soluble even in hot water, and still less soluble in cold water. It sep-
arates from a hot concentrated solution in needles which apparently
contain four molecules of water. The salt is permanent in the air,
but loses three molecules of its crystal water over sulphuric acid.
When dried at 120°, it still retains a half-molecule of water which
cannot be expelled without essential decomposition.

I. 0.6847 grm. air-dried salt gave 0.2511 grm. BaS04.

II. 0.6205 grm. air-dried salt gave 0.2273 grm. BaS04.

III. 0.5672 grm. air-dried salt gave 0.2076 grm. BaS< \.

IV. 1.3625 grm. air-dried salt lost over II2S04 0.1148 grm. H.,0 and

at 120° 0.1338 grm. II20.

Found.

I.

II.

iii.

23.86

23.72

23.84

356 PROCEEDINGS OP THE AMERICAN ACADEMY

V. 1.1645 grm. air-dried salt lost over H2S04 0.1003 grm. H20 and
at 100° 0.1124 grm. H20.
VI. 1.3000 grm. air-dried salt lost over H2S04 0.1124 grm. H20 and
at 100° 0.1256 grm. H20.

Calculated for Found.

BaJCjCLjOa).; . 4 H,0. I. II. III. IV. V. VI.

Ba 21.47 21.56 21.54 21.52

3H20 8.46 8.43 8.61 8.65

3£H20 9.87 9.83 9.65 9.66

I. 0.7484 grm. substance dried at 120° gave 0.3038 grm. BaS04.

IT. 0.6247 grm. substance dried at 100° gave 0.2521 grm. BaS04.

III. 0.6497 grm. substance dried at 100° gave 0.2634 grm. BaS04.

Calculated for i

Ba(C5CI303)2.^n20.

Ba 23.83

The solubility of the salt in water at 19°. 5 was determined in the
usual way.

I. 41.4499 grm. solution saturated at 19°.5 gave 0.0470 grm. BaS04.
II. 44.9822 grm. solution saturated at 19°.5 gave 0.0500 grm. BaS04.

The solution saturated at 19°. 5 therefore contained the following
percentages of anhydrous salt : —

i. ii.

0.27 0.27

Calcic Trichlorpyromucate, Ca(C5Cl303)2 . 4 H.,0. — The calcium
salt was prepared by precipitating a dilute solution of the ammonium
salt with calcic chloride. The precipitated salt was then crystallized
from hot water, and proved to be sparingly soluble in hot water, still
less soluble in cold water, and separated from a hot solution in irreg-
ular aggregations of small leafy plates.

The salt is permanent in the air, effloresces over sulphuric acid, and
loses all its water at 110°.

I. 1.3008 grm. air-dried salt lost at 110° 0.1697 grm. H20.

II. 1.8389 grm. air-dried salt lost at 110° 0.2417 grm. H20.

Calculated for Found.

(C6Cl303),Ca . 4 11,0. I. II.

H20 13.30 13.05 13.14

I. 0.6399 grm. salt dried at 110° gave 0.1855 grm. CaS04.
II. 0.6974 grm. salt dried at 110° gave 0.2019 grm. CaS04.

Calculated for Found.

Ca(CBCl303)2. I. n.

Ca 8.53 8.53 8.51

OF ARTS AND SCIENCES. S57

The solubility of the salt in water at 19°. 5 was determined in the
usual manner. The calcium was precipitated as oxalate, and the
oxalate ignited with sulphuric acid.

I. 18.1457 grm. solution saturated at 19°.5 gave 0.0342 grm. CaS04.
II. 14.6806 grm. solution saturated at 19°. 5 gave 0.0265 grm. CaS04.

The solution saturated at 19°. 5 therefore contained the following
percentages of anhydrous salt : —

i. ii.

0.65 0.62

Potassic Trichlorpyromucate, KC5C1303. — The potassium salt was
prepared by boiling the acid with a slight excess of potassic carbonate.
The solution was then concentrated until the salt began to separate,
and on cooling it solidified to a mass of crystals, which were filtered off
on the pump and pressed dry with filter paper. The salt is readily
soluble even in cold water, and separates from a hot concentrated
solution in sheaves of fine needles, which lose nothing in weight when
heated to 120°. Analyses of the salt dried at this temperature proved
it to be anhydrous.

I. 0.5338 grm. salt dried at 120° gave 0.1851 grm. K2S04.
II. 4425 grm. salt dried at 120° gave 0.1532 grm. K2S04.

Calculated for Found.

KC5C1303. I. II.

K 15.42 15.57 15.54

Argentic Trichlorpyromucate, AgC5Cl,03. — The silver salt may
best be prepared by precipitating a hot solution of the calcium salt
with argentic nitrate. For analysis the precipitated salt was well
washed, and then recrystallized from hot water. It is sparingly solu-
ble even in hot water, less soluble in cold water, and crystallizes from
a hot concentrated solution in moss-like aggregations of fine needles.

The aii-dried salt gave on analysis much too low a percentage of
silver, and although the sample analyzed was free from calcium salt
and free acid, it failed to give perfectly satisfactory results even when
dried in vacuo over sulphuric acid or at 100°.

I. 0.5142 grm. salt dried in vacuo over H2S04 gave 0.2254 grm.

AgCl.

II. 0.4545 grm. salt dried at 100° gave 0.1998 grm. AgCl.

Calculated for Found.

C5Cl,0,Ag. i. a

Ag 33.48 33.00 33.10

358 PROCEEDINGS OP THE AMERICAN ACADEMY

Ethyl Trichlorpyromucate, C5C1803C2H5. — The ethyl ether was
prepared in the usual manner, by warming an alcoholic solution of the
acid with concentrated sulphuric acid. 5 parts of the acid, 10 parts of
absolute alcohol, and 5 parts of concentrated sulphuric acid (Sp. Gr.
1.84) were heated together for three hours at 100°. The ether was
then precipitated with water, washed with warm dilute sodic car-
bonate, and finally with water. The ether is readily soluble in hot
alcohol, more sparingly in cold alcohol, and crystallizes in flat prisms
which melt at 62-63°.

I. 0.2142 grm. substance dried over H2S04 gave 0.3793 grm. AgCl.
II. 0.2536 grm. substance dried over H2S04 gave 0.4476 grm. AgCl.

Calculated for

Found.

C6C1303C2H0.

I.

ii.

43.74

43.77

43.64

CI

Trichlorpyromucamide, C5Cla02NH2. — Concentrated aqueous am-
monia attacks the trichlorpyromucic ether but slightly at ordinary tem-
peratures. Even after three hours' heating in a sealed tube at 100°
most of the ether was found to be unaltered, and we were obliged to
prepare the amide from the acid chloride by the action of solid am-
nionic carbonate. The amide is but sparingly soluble even in hot
water and crystallizes in long slender needles, which melt at 160-161°.

0.3804 grm. substance dried over H2S04 gave 23.2 c.c. moist N at 21°
under a pressure of 765 mm.

Calculated for
C6C1302NH2. Found.

N 6.53 6.95

Although there could be no doubt as to the constitution of the
trichlorpyromucic acid it seemed better for the sake of completeness to
study its behavior with oxidizing agents.

Action of Bromine and Water.

Trichlorpyromucic acid was suspended in ten times its weight of
cold water, and a little more than one molecule of bromine was added.
Carbonic dioxide was evolved and the color of the bromine rapidly
disappeared. When the reaction had been completed, a white insolu-
ble substance remained, which was removed by filtration and washed
with a dilute solution of sodic carbonate. The alkaline solution gave
on acidification a copious precipitate of unaltered trichlorpyromucic
acid, whose identity was established by the melting point 172-173°.
The substance which remained undissolved by the alkaline solution

OP ARTS AND SCIENCES. 359

had the characteristic odor of the tetrabromfurfuran, and crystallized
from hot alcohol in irregular plates, which melted at 75-76°. Analy-
sis showed this substance to be a trichlorbromfurfuran.

I. 0.1749 grm. substance dried over H2S04 gave 0.4269 grm. AgCl
and AorBr.
II. 0.1411 grm. substance dried over H2S04 gave 0.3447 grm. AgCl

and AgBr.

Calculated for

Found.

C4Cl3Br0.

I.

ii.

Cl3 + Br

74.45

73.60

73.67

The strongly acid filtrate from the trichlorbromfurfuran and unal-
tered trichlorpyromucic acid yielded on extraction with ether a crys-
talline acid which was readily soluble in water. The amount of this
acid was too small for analysis, but it was easily identified as dichlor-
maleic acid by the melting point of its anhydride, 119-120°.*

The reaction had therefore taken place in accordance with the
following equations :

C5HC1303 4- Br2 = C4CLBrO + C02 + HBr.
C5HC1,03 + 2 Br2 + 3 H20 = C4H2C1204 + C02 + HC1 + 4 HBr.

Action of Nitric Acid.

Trichlorpyromucic acid is but slowly attacked by nitric acid, and
prolonged heating with moderately strong nitric acid is needed for
complete oxidation. Even after heating the acid for six hours with
six times its weight of nitric acid (Sp. Gr. 1.42) diluted with an equal
weight of water, a portion of the trichlorpyromucic acid escaped oxi-
dation. Carbonic dioxide was slowly given off, and a small quantity
of an insoluble oil was formed, which had a peculiar penetrating odor
not unlike that of substituted furfuran derivatives. Its quantity was
too small to admit of its identification. In solution we were able to
find nothing but dichlormaleic acid, which we purified through the
barium salt and identified by the melting point of its anhydride and
by analysis.

0.1997 grm. substance dried over H2S04 gave 0.3086 grm. AgCl.

Calculated for
C4H2C1204- Found.

CI 38.38 38.20

The reaction had therefore taken place in accordance with the fol-
lowing equation :

C5IIC130; + 2 O + H20 = C4H2CL04 + CO, + IIC1.

* Ciamiciau and Silber, he. cit.

360 PROCEEDINGS OP THE AMERICAN ACADEMY

We have also prepared several other trisubstituted pyromucic acids,
which may conveniently be described here.

/3y-DlCHLOR-S-BROJIPYROMDCIC ACID.

Hill and Sanger * found that fiy dibrompyromucic acid was easily
converted into tribrompyromucic acid by the action of bromine at
ordinary temperatures.

If /3y dichlorpyromucic acid is exposed to the vapors of bromine at
ordinary temperatures, bromine is rapidly absorbed, hydrobromic acid
is evolved, and the gain in weight approaches that required by the dis-
placement of hydrogen by bromine. The product was treated with
small quantities of boiling water to remove unaltered dichlorpyromucic
acid, and then recrystallized from dilute alcohol, and finally from water.
The acid is readily soluble in alcohol or ether, dissolves freely in boil-
ing benzol, more sparingly in cold benzol. Even in boiling water it is
very sparingly soluble. It crystallizes in short clustered prisms, which
melt at 185-186°.

0.3189 grm. substance dried over II.2S04 gave 0.5817 grm. AgCl
and AgBr.

Calculated for
C0HCl2Br03. Found.

Cl2 + Br 58.08 57.99

/?y-DlBROM-S-CHLORPYROMUCIC ACID.

This acid we made by treating the ethyl fiy dibrompyromucate
with chlorine and decomposing the product with alcoholic sodic hy-
drate. On acidifying the aqueous solution of the sodium salts thus
obtained, a crystalline acid was precipitated, which after one recrystal-
lization from dilute alcohol melted at 192-193°. After recrystalliza-
tion from benzol the melting point rose to 193-194°. The acid was
readily soluble in alcohol, ether, or hot benzol ; sparingly soluble in
cold benzol, or even in boiling water.

0.2315 grm. substance dried over H2S04 gave 0.39G0 grm. AgCl
and AgBr.

Calculated for
C5UClBr,03. Found.

CI + Br2 64.19 64.37

We have attempted to prepare these acids also by the decomposition
with alcoholic sodic hydrate of products formed by the addition of

* These Proceedings, xxi. 172.

OF ARTS AND SCIENCES. 861

bromine to 8 chlorpyromucic acid and of chlorine to ethyl 8 brompyro-
mucate. Although we readily prepared in this way trisubstituted
acids, the products in neither case were homogeneous, tribrom- or
trichlorpyromucic acids being formed together with the acid containing
both halogens.

/3y-DiCHLOR-8-NiTROPYito:\iucic Acid.

Hill and Palmer* have shown that /3y dibrompyromucic acid is
easily converted into a sulphonic acid by means of sulphuric acid, and
that the sulpho group in this acid may readily be replaced by the nitro
group. /3y dichlorpyromucic acid was dissolved in fuming sulphuric
acid, and the barium salt of the sulphonic acid isolated in the usual
way. Since a more complete study of the salt did not fall within the
plan of our work, we precipitated the barium exactly with sulphuric
acid, and evaporated the dilute solution of the acid first at 100° and
afterwards in vacuo over sulphuric acid. The crystalline somewhat
deliquescent acid thus obtained was dissolved in several times its
weight of fuming nitric acid, to which half its volume of concentrated
sulphuric acid had been added. After the action was finished, the
nitric acid was partially expelled, the residue diluted with water, and
extracted with ether. The ether, was then shaken with a dilute solu-
tion of soclic carbonate, and the acid precipitated from this alkaline
solution by the addition of hydrochloric acid. The ethereal solution
proved to contain no substance which invited further investigation.

The /3y-dichlor-S-nitropyromucic acid is readily soluble in alcohol,
ether, or hot benzol. Hot water dissolves it freely, and on cooling the
greater part of the acid is deposited in flattened leafy prisms which
melt at 189-190°.

0.2168 grm. substance dried over H2S04 gave 0.2756 grm. Ag. CI.

Calculated for
C6HC12N06. Found.

CI 31.42 31.42

"We have as yet made no further study of this acid.

Theoretical Considerations.

The acids melting at 168-169° and 155-156°, which in the pre-
ceding pages we have called the (3y and /3S dichlorpyromucic acids, are
without doubt identical in structure with the two dibrompyromucic
acids described by Hill and Sanger. | Their physical properties and

* Tliese Proceedings, xxiii. 201, 205.
t These Proceedings, xxi. 137.

362 PROCEEDINGS OP THE AMERICAN ACADEMY

the solubility of their salts show their close relationship to these bro-
mine derivatives, while the products formed from them by oxidation
conclusively jjrove that they are structurally isomeric, and that the
chlorine atoms occupy the same relative position that the bromine
atoms hold in the dibrompyromucic acids. On reduction these two
acids both yield the same chlorpyromucic acid, which in its turn closely
resembles the j3 brompyromucic acid in its jmysical properties, although
its melting point, 145-146°, is comparatively high. It is to be noted,
however, that the f3y dichlorpyromucic acid alone is formed by the
action of alkalies upon the ethyl pyromucic tetrachloride, while the two
isomeric dibrompyromucic acids are simultaneously formed under sim-
ilar conditions. This simultaneous formation of the two structurally
isomeric acids from pyromucic tetrabromide to which the formula

Br Br
H--C-C- COOH
\

0
/

HC - C - H
Br Br

may be assigned led Hill and Sanger * to the conclusion that these
acids must have their bromine atoms in the /?y and ft8 positions
respectively, and that pyromucic acid itself probably had the formula

HC = C - COOH

O
/
HC = CH.

The structure of the fSy and /?S dichlorpyromucic acids seems to us
to be thus sufficiently established.

As to the third isomeric form, which for the present we have called
the x dichlorpyromucic acid, since it gives chlorfumaric acid by oxida-
tion, it is evident that it must either have its chlorine atoms in the
y and 8 position, or that it must be a geometrically isomeric form of
the (38 acid. Unfortunately, we have been able to obtain as yet no
evidence which conclusively proves either view to be correct. A
dichlorpyromucic acid having its chlorine atoms in the y and 8 posi-
tions could not well be formed from ethyl pyromucic tetrachloride by
the simple loss of hydrochloric acid. Tonniesf had shown that 8 brom-

* These Proceedings, xxi. 182.
\ t Berichte de deutsch. chem. Gesellsch.

OP ARTS AND SCIENCES. 863

pyroinucic acid could be formed by tbe action of heat upon pyromucic
tetrabrornide, when bromine must be eliminated as well as hydrobromic
acid. Since we had noticed that hydrochloric acid invariably was
evolved in the preparation of ethyl pyromucic tetrachloride, it seemed
to us not improbable that the ^ dichlorpyromucic acid was in fact the
yd acid which had been formed in a similar way from the tetrachloride
of the 8 chlorpyromucic ether necessarily contained in the crude
product.

CI CI

HC - C - COOH

\

0

/

Cl(

: cci

I

I Cl

We found, however, that no x dichlorpyromucic acid was formed
on heating ethyl 8 chlorpyromucic tetrachloride, and that our product
in this case contained the /?8 acid alone, so that the molecule of
chlorine, which had been eliminated in the reaction, had taken its
chlorine atoms from the y and 8 positions. We hoped also to get
further evidence as to the structure of the ^ dichlorpyromucic acid by
reducing it to a chlorpyromucic acid, or by substituting the chlorine of
the x dichlorsulphopyromucic acid by hydrogen; but in neither case
have we yet succeeded in obtaining well characterized products.

While it seems to us probable that the % dichlor pyromucic acid is
geometrically isomeric with the /38 acid, we have as yet been unable
to convert one acid into the other. Both acids volatilize unchanged
when heated, and the ordinary reagents which usually effect conver-
sion in such cases have failed to bring abont any perceptible isomeri-
zation. While lack of material has to a great degree limited our work
with the x acid, we have studied the behavior of the fi8 acid under a
variety of conditions with purely negative results. If tlie two acids
are in fact geometrically isomeric, the simultaneous appearance of both
forms in the distillation of ethyl pyromucic tetrachloride with the
escape of large quantities of hydrochloric acid shows that the conver-
sion of one acid into the other is effected with difficulty, while the
formation of the (3d acid alone from the ethyl 8 chlorpyromucic tetra-
chloride under similar conditions is especially worthy of note. Since
pyromucic acid apparently contained two pairs of carbon atoms doubly

364 • PROCEEDINGS OF THE AMERICAN ACADEMY

tied, the existence of geometrically isomeric forms of its derivatives
seemed at the outset more thau prohable, and during the progress of
the investigations which for many years have been carried on in this
laboratory careful search had been made for such bodies, but hitherto
only the two isomeric dibromfurfuran tetrabromides of Hill and Harts-
horn * had been discovered. Since the discovery of the ^ dichlorpyro-
mucic acid, which we made toward the close of our work, we have
attempted to prepare the corresponding bromine derivative, but as yet
without success. The isolation of the body C4H3Br02, which Hill
and Sanger f found among the products of the decomposition of
pyromucic tetrabromide by alkalies, certainly is an indication of the
previous existence of such an acid, although it is by no means clear in
what stage of the process the decomposition of the acid was effected,
nor how the decomposition product itself escaped further alteration.
Any detailed consideration of the geometrical structure of pyromucic
acid we' feel obliged to postpone until more facts bearing upon the
question are at our disposal.

II. — ON CERTAIN DERIVATIVES OF FURFUR-
ACRYLIC ACID.

By H. B. Gibson and C. F. Kahnweiler. <

Presented May 28, 1889.

The preparation of furfuracrylic acid was described by v. Baeyer t
1877. At that time, however, the investigation of the acid was
pushed in but one direction, and the successive steps by which it
could be converted into a pimelic acid were alone described in detail.
It seemed that a further study of the acid in other directions also
could hardly fail to yield interesting results.

The ready decomposition of furfuracrylic acid by mineral acids
compelled us to relinquish the line of work we originally had pro-
posed to follow, and we consequently began an investigation into the
nature of this decomposition. The study of the action of hydrochloric
acid upon an alcoholic solution of furfuracrylic acid was just beginning

* Berichte der deutsch. chem. Gesellsch., xviii. 450.

t These Proceedings, xxi. 158.

t Ber. d. deutsch. chem. Gesellsch., x. 355.

OF ARTS AND SCIENCES. 365

to yield us results when the paper o'f Markwald * upon the same sub-
ject reached us, and we were therefore obliged to abandon our work
in this direction, and turned our attention to the products formed by
the action of bromine, which promised to repay investigation. We
felt ourselves quite at liberty to continue our study of furfuracrylic
acid in this direction, since Markwald had declared himself unable
to obtaiu well defined crystalline products in this way.

Furfuracrylic Acid.

In the preparation of furfuracrylic acid we have found it to our
advantage to modify slightly the proportions given by v. Baeyer, and
in this respect our experience agrees more closely with that of Mark-
wald. We have heated 2 parts of furfurol with 3 parts of fused sodic
acetate and 3 parts of acetic anhydride with reverse cooler at 160-170°
for about ten hours, extracted the acid with a dilute solution of sodic
carbonate, precipitated with hydrochloric acid, and recrystallized the
product thus obtained from boiling water. Although the yield varied
somewhat in successive preparations, we frequently obtained an amount
equal to the weight of furfurol employed. The melting point of the pure
acid we found to be 140°, in agreement with the observations of Jaffe
and Cohn,f instead of 136° as given by v. Baeyer. From the acid we
made the methyl ether and the amide, which we may describe a little
more in detail, since no description of them has yet been published.

Methyl Furfuracrylate, C.H.CX . CH3. — By precipitating with ar-
gentic nitrate a solution of ammonic furfuracrylate the silver salt was
obtained as a heavy curdy precipitate. This was carefully dried and
treated with an excess of methyl iodide. When the decomposition
was complete the product of the reaction was extracted with ether,
and purified by distillation under diminished pressure. Under a
pressure of 15 mm. the methyl furfuracrylate distils unchanged at
112°, and the distillate solidifies in beautiful rhombic crystals which
melt at 27°. Under a pressure of 774 mm. we found that the ether
distilled without apparent decomposition at 227-228°. It possessed
an agreeable characteristic odor, and dissolved readily in alcohol, ether,
ligroin, or benzol. An analysis gave the following results : —

0.2224 grm. substance gave 0.5157 grm. C02 and 0.1084 grm. H20.

Calculated for

C8H9Oa. Found.

C 63.17 63.23

H 5.26 5.42

* Ber. d. deutsch. chem. Gesellsch., xx. 2811. t Ibid., xx. 2315.

366 PROCEEDINGS OP THE AMERICAN ACADEMY

Furfur acrylamide, C-H.02. NH2. — Concentrated aqueous amnionic
hydrate acted but slowly upon methyl furfuracrylate in the cold, but
at 100° in sealed tube the decomposition was readily effected. The
amide was sparingly soluble in cold water, more readily in hot water,
and crystallized in pearly scales which melted at 168-1G90.

I. 0.4076 ffrm. substance gave 35.5 c.c. moist nitrogen at 21°. 4 and
under a pressure of 766 mm.
II. 0.4590 grm. substance gave 41.9 c.c. moist nitrogen at 21°. 7
and under a pressure of 767 mm.

Calculated for Found.

C7II7X0,. I. II.

N 10.22 9.98 10.45

Action of Bromine.

It had already been observed by v. Baeyer that furfuracrylic acid was
readily attacked by aqueous bromine, but he had made no careful study
of the products formed. Since the product formed in this way was un-
inviting, and the reaction undoubtedly complicated, it seemed advisable
to study instead the action of dry bromine, which might yield well char-
acterized addition products. We soon found, however, that with dry
bromine even in the cold substitution was readily effected, so that the
simplest well defined product contained three atoms of bromine. Sub-
sequent study showed that this product was a saturated compound iu
which one atom of bromine had taken its place in the furfuran ring.

Bromfurfurdibrompropionic Acid, C7H.Br303. — If furfuracrylic
acid is suspended in ten times its weight of carbonic disulphide and two
molecules of dry bromine added, a clear deep red solution is at once
obtained, from which hydrobromic acid is soon evolved in quantity.
Although the reaction progresses steadily at ordinary temperatures, it
may be greatly hastened by heat, and we usually have heated the
mixture upon the water bath soon after the addition of the bromine.
When the evolution of the hydrobromic acid slackens, and a crystalline
substance has separated in quantity, the solution is well cooled, filtered,
and the crystalline solid well washed with cold carbonic disulphide.
The substance then gives on analysis percentages which closely agree
with those required by the formula C7H5Br303 ; but it may still further
be recrystallized from hot benzol or carbonic disulphide.

I. 0.2080 grm. substance gave 0.3117 grm. AgBr.
II. 0.3798 grm. of substance recrystallized from carbonic disulphide
gave 0.3099 grm. C02 and 0.0502 grm. H20.

OP ARTS AND SCIENCES. 867

III. 0.2005 grm. substance recrystallized from benzol gave 0.2998 grm.

AgBr.

IV. 0.2003 grm. substance recrystallized from benzol gave 0.2993 grm.

AgBr.

Calculated for

C7H5Br303.

c

22.28

H

1.33

Br

63.66

Found

II. III.

IV.

2.26

1.47

63.65

63.60

63.78

Bromfurfurdibrompropionic acid crystallizes in small flat oblique
prisms, which are sparingly soluble iu cold benzol or carbonic disul-
phide, more readily in hot. It is readily soluble in alcohol or ether,
insoluble in water, although slowly decomposed by it. On heating,
the acid is gradually decomposed, with the evolution of hydrobromic
acid and slight carbonization. As might be expected, we found it im-
possible to prepare salts, or even ethers, of the acid.

The action of cold water upon the acid was so well marked, that we
proceeded to determine the products formed. When bromfurfurdi-
brompropionic acid is suspended in water, carbonic dioxide soon
begins to escape, and after long standing the crystalline solid is com-
pletely converted into a colorless oil. The aqueous solution then
contains hydrobromic acid in abundance. As the decomposition was
greatly facilitated by gentle heat, we usually have allowed the reaction
to proceed at about 40°. When the decomposition was complete, we
distilled with steam, and dried the heavy colorless oil thus obtained
with calcic chloride. Ou distilllation under diminished pressure, we
found that this oil began to boil at 108° under a pressure of 14 mm.,
and that the thermometer gradually rose to 112° the pressure remain-
ing constant. The distillate was at first nearly colorless, and of
high refractive power, but on standing it quite rapidly became dark-
colored. Analyses of the freshly distilled substance showed it to be a
bromfurfurbromethylen.

I. 0.3173 grm. substance gave 0 3350grm. C02 and 0.0476 grm. ILO.
II. 0.1928 grm. substance gave 0 2883 grm. AgBr.
III. 0.2045 grm. substance gave 0.3056 grm. AgBr.

and.

III.

63.60

Calculated for

Found

CcII4Br20.

i.

II.

c

28.57

28.79

H

1.59

1.67

Br

63.49

63.65

368 PROCEEDINGS OF THE AMERICAN ACADEMY

The reaction is then strictly analogous to that by which /? brom-
styrol may be made from phenyldibrompropionic acid :

C7H5Br303 = C6H4Br20 + C02 + HBr.

The aqueous solution left in the retort after distillation with steam
deposited on long standing a small quantity of a crystalline acid. The
quantity of this acid was so insignificant that we were able to iden-
tify it only after uniting the products obtained in many successive
preparations. Two determinations of the percentage of bromine
showed the acid to be the bromfurfuracrylic acid subsequently de-
scribed. Lack of material made it impossible to prove with precision
the identity of the two acids, and we can therefore only assert their
close resemblance.

I. 0.1262 grm. substance gave 0.1091 grin. AgBr.
II. 0.1956 grm. substance gave 0.1700 grm. AgBr.

Calculated for Found.

C7II,Br03. I. II.

Br 36.86 36.79 36.99

We have hitherto been unable to find other definite products of
the reaction.

By the action of bromine upon bromfurfurbromethylen we could
obtain no products which invited further study. On the other hand,
alcoholic potassic hydrate gave us the corresponding acetylen deriva-
tive, although we have not yet isolated it in a pure condition. The
oil obtained as the direct product of the reaction contained much too
large a percentage of bromine, and prolonged action at a higher
temperature gave no more satisfactory results. That the oil con-
tained the acetylen derivative was readily shown by its behavior with
an ammoniacal solution of silver or copper. The silver compound
was white, the copper compound a greenish yellow. The latter was
made in larger quantity, carefully washed by decantation, and dried
over sulphuric acid. It then exploded violently by heat or on contact
with fuming nitric acid, and contained the required percentage of
copper.

I. 0.6479 grm. substance gave by electrolysis 0.1748 grm. Cu.
II. 0.6178 grm. substance gave 0.1681 grm. Cu.

Calculated for Found.

[CgHjBrOJjCuj. I. II.

Cu 27.16 26.98 27.20

From the copper compound we attempted to prepare the bromfurfur-
acetylen in a pure condition. By decomposing with acid, extracting

OP ARTS AND SCIEiNCES. 369

with ether, and distilling the well dried extract under diminished
pressure, we obtained a colorless oil which boiled at 65-68° under a
pressure of 19 mm. This oil contained a percentage of bromine, which
left no doubt as to its identity, but showed at the same time that it
was still impure. We have as yet made no further attempt at its
complete purification.

It seemed to us of interest to prepare by oxidation the diacetylen
derivative also, analogous to the diacetenylphenyl of Glaser,* which
was afterward more fully studied by v. Baeyer and Landsberg.| In
its preparation we followed closely the method of v. Baeyer. To the
copper compound was added one molecule of potassic ferricyanide in
aqueous solution. When the oxidation was complete, the precipitate
formed was collected on a filter, dried, and extracted with hot alcohol.
On cooling the alcohol deposited the new compound in small iridescent
plates which melted at 126°. It was readily soluble in hot alcohol,
sparingly in cold alcohol, and dissolved freely in ether, chloroform,
benzol, or ligroin. Analysis showed the substance to be dibromdifur-
furdiacetylen (diacetenylbromfurfuryl) :

C4H2BrO -C=C-C=C- C4H2BrO.

I. 0.2476 grm. substance gave 0.3888 grm C02 and 0.0383 grm. H20.
II. 0.1538 grm. substance gave 0.1696 grm. AgBr.
III. 0.1548 grm. substance gave 0.1702 grm. AgBr.

d.

in.

c

Calculated for
CjoH^Br-jC^.

42.36

i.

42.82

Found,
II.

H

1.18

1.72

Br

47.06

46.93

46.79

Bromfurfuracrylic Acid.

The behavior of the acid containing three atoms of bromine which
was obtained by the direct action of bromine upon furfuracrylic acid
left little doubt that it was a bromfurfurdibrompropionic acid. If
this view were correct, it seemed not improbable that the bromine
attached to the the side-chain might be removed without disturbing the
bromine in the furfuran ring, and that a bromfurfuracrylic acid would
thus be formed. After a kw preliminary experiments it was found
that zinc dust added to an alcoholic solution of the bronifnrftirdibrom-
propionic acid brought about the desired result. The reaction is vig-

* Ann. Chem. u. Pharm., cliv. 159.
t Ber. d. deutsch. chem. Gesellsch., xv. 57.
vol. xxiv. (n. s. xvi.) 24

870 PROCEEDINGS OP THE AMERICAN ACADEMY

orous, and the boiling point of the alcohol soon reached. When the
reaction is over, water is added to the filtered solution, and the acid
which is thus precipitated is purified by conversion iuto its sodium
salt, reprecipitation with hydrochloric acid, and recrystallization from
dilute alcohol or boiling water. The yield of the new acid thus ob-
tained was between 80 and 90 per cent of the theoretical amount, and
analysis showed that it had the desired composition.

I. 0.2645 grm. substance gave 0.3768 grm. C02 and 0.0566 grm. H20.
II. 0.2511 grm. substance gave 0.2161 grm. AgBr.

III. 0.2223 grm. substance gave 0.1920 grm. AgBr.

IV. 0.1490 grm. substance gave 0.1288 grm. AgBr.

Found.
II. III. IV.

c

Calculated for
C7H0Br03.

38.71

i.

38.86

H

2.30

2.38

Br

36.86

36.63 36.76 36.80

Bromfurfuracrylic acid crystallizes in long slender prisms, which are
sparingly soluble in cold water, more readily in hot water. The acid
dissolves readily in alcohol, ether, hot benzol, or hot chloroform, in cold
benzol or cold chloroform it is but sparingly soluble, and it is nearly
insoluble in ligroin. It melts sharply at 176-177°, but since it is
somewhat decomposed and blackened by long continued heating above
150°, the bath must first be heated to within a few degrees of the
melting point. Small quantities of the acid may readily be sublimed
between watch-glasses. The sublimed product also melts at 176—177°,
and darkens upon long heating. With bromine the acid unites readily,
and forms the original substituted propionic acid. We have made
many attempts to convert the acid into the corresponding substituted
pyromucic acid, or to obtain from it by oxidation a product which
should enable us to fix the position of the bromine atom, but hitherto
without success. We can, therefore, at present only conjecture that
the bromine in this case, as with the pyromucic acid, enters first the
8 position. For the further characterization of the acid we have
prepared a few of its salts.

Baric Br omfurfaracrylate, Ba(C7.II4Br03)2 . H20. — The barium salt
is sparingly soluble in cold water, somewhat more readily in hot
water, and crystallizes in clustered needles. When air-dried it loses
nothing over sulphuric acid, gives up one molecule of water slowly
at 100°, and is decomposed at a somewhat higher temperature.

I. 1.1830 grm. of the air-dried salt lost at 100° 0.0380 grm. II20.

Calculated for

Ba(C,Il4Br03)2 . H20.

H20

3.07

Ba

23.34

OF ARTS AND SCIENCES. 371

II. 0.7064 grm. of the air-dried salt lost at 100° 0.0227 grm. H20.
III. 0.4898 grm. of the air-dried salt gave 0.1951 grm. BaS04.

Found.
I. 11. III.

3.21 3.21

23.42

I. 0.5258 grm. of the salt dried at 100° gave 0.2149 grm. BaS04.
II. 0.4787 grm. of the salt dried at 100° gave 0.1955 grm. BaS04.

Calculated for Found.

Ba(C7lI4Br03)2. I. II.

Ba 24.07 24.02 24.01

Calcic Bromfurfuracrylate, Ca(C-H4Br03)2. 3 H20. — This salt crys-
tallizes in columnar aggregations of plates, and is sparingly soluble in
cold water, more readily in hot water. The air-dried salt effloresces
over sulphuric acid, and is gradually decomposed at 100°, but may be
dried without difficulty at 80-85°.

I. 1.1639 grm. of the air-dried salt lost at 82° 0.1184 grm. H20.
II. 0.7362 grm. of the air-dried salt lost at 83-85° 0.0740 grm. H20.
III. 0.6331 grm. of the air-dried salt gave 0.1648 grm. CaS04.
IV. 0.5287 grm. of the air-dried salt gave 0.1375 grm. CaS04.

Calculated for Found.

Ca(C7H4Br03)2 . 3 H20. I. II. III. IV.

H20 10.07 10.17 10.05

Ca 7.61 7.65 7.65

0.6123 grm. of the salt dried at 82° gave 0.1774 grm. CaS04.

Calculated for

Ca(C7H4Br03)2. Found.

Ca 8.47 8.52

Sodic Bromfurfuracrylate, NaC7H4Br03. — This salt is readily soluble
even in cold water, and crystallizes in anhydrous nodular aggregations.

I. 0.7612 grm. of the salt gave 0.2228 grm. Na2S04.
II. 0.7589 grm. of the salt gave 0.2206 grm. Na2S04.

Calculated for Found.

NaC7H4Br03. I- H.

Na 9.62 9.48 9.42

Argentic Bromfurfuracrylate, AgCrII4BrOr — The silver salt, formed
by the addition of argentic nitrate to a neutral solution of the ammo-
nium salt, proved to be nearly insoluble in water, and was apparently
amorphous.

I. 0.3919 grm. of the salt gave 0.2261 grm. AgP»r.
II. 0.4065 grm. of the salt gave 0.2347 grm. AgBr.

372 PROCEEDINGS OF THE AMERICAN ACADEMY

Calculated for Found.

AgC7H4Br03. I. rr.

Ag 33.34 33.13 33.16

Ethyl Bromfurfuracrylate, C7H4Br03 . C2H5. — To a solution of the
acid (5 parts) in somewhat more than its own weight of absolute alco-
hol (6.5 parts) was added concentrated .sulphuric acid (3 parts), and
the mixture heated upon the water bath for two or three hours. "We
found that too long heating should be avoided since decomposition
then ensued, and a product was formed which undoubtedly was analo-
gous to that obtained by Markwald.* We have, however, made no
careful study of it. As soon as the etherification appeared to be com-
plete, the ether was precipitated with water, dried with calcic chloride,
and distilled uuder diminished pressure. The greater part of the
product distilled at 151-152° under a pressure of 14 mm., and on
cooling this distillate completely solidified. From ligroin the ether
could be obtained in large flat prisms, which melted at 42°. It was
readily soluble in alcohol, ether, chloroform, benzol, or ligroin and
had a faint aromatic odor.

I. 0.1968 grm. substance gave 0.31 68 grm. C02 and 0.0673 grm. H20.
II. 0.2251 grm. substance gave 0.1722 grm. AgBr.
III. 0.2545 grm. substance gave 0.1943 grm. AgBr.

m.

c

Calculated for
C9H9Br03.

44.08

i.

43.88

Found.
II.

H

3.67

3.80

Br

32.65

32.55

32.49

We further attempted to make the amide, but found that the ether
was hardly attacked by concentrated ammonic hydrate at 100°, and
that at a much higher temperature (215°) most of the ether employed
also remained unchanged.

Bromfurfurbromacrylic Acid.

It already has been mentioned that bromfurfurdibrompropionic acid
is decomposed by heat with the evolution of hydrobromic acid, and the
simultaneous formation of a disubstituted furfuracrylic acid might be
inferred. We have made many attempts to bring about this loss of
one molecule of hydrobromic acid by more convenient and more
economical methods, but without success. In alcoholic solution potas-
sic hydrate gave chiefly the ethylen derivative and potassic carbonate.

* Loc. cit.

OP ARTS AND SCIENCES. 373

no matter how the conditions of the reaction were varied. A number
of high boiling neutral solvents gave no better results than those ob-
tained by heating the acid by itself. The bromfurfurdibrompropionic
acid was therefore carefully heated in shallow crystallizing dishes im-
mersed in a sulphuric acid bath. The hydrobroinic acid thus escaped
without blackening the product to any great extent. The tempera-
ture was slowly raised to 130°, and when at this point the formation
of hydrobromic acid was no longer noticeable the residue was boiled
with dilute ammonic hydrate filtered, heated with bone-black, and the
filtered solution concentrated by evaporation. In this way sparingly
soluble finely felted needles of ammonic bromfurfurbromacrylate were
obtained, which yielded the acid in a state of purity.

I. 0.2246 grm. substance gave 0.2328 grm. C02 and 0.031 7 grm. H20.
II. 0.2290 grm. substance gave 0.2373 grm. C02 and 0.03H grm. H20.

III. 0.2059 grm. substance gave 0.2G15 grm. AgBr.

IV. 0.2025 grm. substance gave 0.2571 grm. AgBr.

Calculated for Found.

C7H4Br20s. I. II. HI. IV.

C 28.38 28.27 28.26

H 1.35 1.57 1.53

Br 54.05 54.04 54.03

Bromfurfurbromacrylic acid is almost insoluble even in hot water.
It dissolves freely in ether or alcohol, and when precipitated by water
from a hot alcoholic solution is obtained in fine clustered needles. At
ordinary temperatures it is very sparingly soluble in chloroform or
carbonic disulphide, but dissolves somewhat more freely on heating.
The acid melts at 178-179° and small quantities of it may easily be
sublimed without essential decomposition.

In spite of many attempts, we have been unable to obtain experi-
mental evidence as to the position of the bromine atoms. We have
also been unable to prepare definite products by the addition of bro-
mine to the acid, or to form from it bromfurfurpropiolic by the elimi-
nation of hydrobromic acid. The following salts may serve for the
better characterization of the acid.

Baric Bromfurfnrbromacrylate, Ba(C7H3Br20..)2 . 2 ILO. — This
salt was prepared by adding baric chloride to a solution of the ammo-
nium salt. It is very sparingly soluble even in boiling water, and
crystallizes in lustrous scales. The air-dried salt contain-, two mole-
cules of water, which it does not lose over sulphuric acid.

I. 1.4911 grm. of the air-dried salt lost at 120° 0.0674 grm. 11,0.
II. 0.8824 grm. of the air-dried salt lost at 120° 0.0399 gnu. 11./).

374 PROCEEDINGS OF THE AMERICAN ACADEMY

Calculated for

Found.

Ba(C7H3Br203)2.2Ha0.

I.

II.

4.72

4.52

4.52

H20

I. 0.2130 grin, of the anhydrous salt gave 0.0684 grm. BaS04.

II. 0.2229 grm. of the anhydrous salt gave 0.0712 grm. BaS04.

III. 0.4376 grm. of the anhydrous salt gave 0.1398 grm. BaS04.

Calculated for Found.

Ba(C7H3Br203)2. I. II. III.

Ba 18.85 18.88 18.78 18.78

Argentic Bromfurfurbromacrylate, AgC7H3Br203. — The silver salt
made by precipitation is very sparingly soluble even in boiling water,
and is apparently amorphous.

I. 0.2052 grm. of the salt dried over sulphuric acid gave 0.0951
grm. AgBr.
II. 0.2101 grm. of the salt dried over sulphuric acid gave 0.0977
grm. AgBr.

Calculated for Found.

AgC,H3Br20j. I. II.

Ag 26.80 26.62 26.71

Potassic Bromfurfurhromacrylate, KC7H3Br203. — The potassium salt
is sparingly soluble in cold water, more readily soluble iu hot water,
and crystallizes in slender clustered needles which are anhydrous.

I. 0.2275 grm. of the salt gave 0.0589 grm. K2S04.
II. 0.5386 grm. of the salt gave 0.1407 grm. K2S04.

Calculated for Found.

KC7HsBr20,. I. n.

K 11.70 11.62 11.72

Ethyl Bromfurfurbromacrylate, C7H3Br203 . C2Hj. — The ethyl ether
was made by heating for three hours on the water bath a solution of
the acid (4 parts) in absolute alcohol (40 parts), with the addition of
a small amount (3 parts) of concentrated sulphuric acid. The ether
was then precipitated with water, washed with a dilute solution of
sodic carbonate, and recrystallized from ligroin. It crystallized easily
in radiated needles which melted at 55-56° and dissolved readily in
alcohol, ether, chloroform, benzol, ligroin, or carbonic disulphide.

I. 0.3774 grm. of substance gave 0.4592 grm. C02 and 0.0849
grm. H20.
II. 0.2125 grm. substance gave 0.2464 grm. AgBr.
III. 0.2179 grm. substance gave 0.2525 grm. AgBr.

OF ARTS AND SCIENCES.

375

Calculated for

Found

C9H815r203.

i.

II.

c

33.34

33.19

H

2.47

2.50

Br

49.39

49.3,

III.

49.32

While we have been unable to establish with precision the constitu-
tion of the foregoing derivatives of furfuracrylic acid, the analogies
offered by the derivatives of pyromucic and cinnamic acids naturally
suggest the following formulas for the substances in question : —

HC = C - CHBr - CHBr - COOH
\

O
/

IIC = C
\
Br

Bromfurfurdibrompropicmic Acid.

HC = C - CH = CHBr
\

O
/

HC = C
\
Br

Bromfurfurethylen.

HC = C - CH = CBr - COOH

\
O

/
HC = C

\
Br

HC = C - CH = CH - COOH

\
O

/
HC = C

\
Br

Bromfurfurbromacrylic Acid.

Bromfurfuracrylic Acid.

376 PROCEEDINGS OF THE AMERICAN ACADEMY

III. ON THE SO CALLED DIOXYMALEIC ACID.

By W. S. Hendrixson.

Presented May 28, 1889.

Several years ago Bourgoin * announced the discovery of two new
acids which he had obtained by replacing the bromine of brommaleic
and dibrommaleic acid by hydroxyl, and which he therefore called oxy-
maleic and dioxymaleic acid. Although the experimental evidence as
to the existence of these acids was very slight, and their assumed struc-
ture wholly anomalous, they soon found place in many text-books.

In 1886, at the suggestion of Fittig, the work of Bourgoin upon
oxymaleic acid was repeated by Scherks,f who found that his state-
ments were without foundation, and that brommaleic acid in aqueous
solution was not in the least attacked by argentic oxide, even at 100°.
Scherks further concluded that Bourgoin's statements concerning his
dioxymaleic acid must also be incorrect, because the dibrommaleic acid
which he used he claimed to have made from his tribromsuccinic acid,
an acid which could not be formed under the given conditions, as shown
by Fittig and Petri,:? or if formed would at once be decomposed into
dibromacrylic acid and carbonic dioxide. This conclusion of Scherks
seems hardly justified by Bourgoin's statements. The dibrommaleic
acid which he used undoubtedly was made by the action of aqueous
bromine upon succinic acid, and, while his material may have been far
from pure, the question of its identity is in no way affected by his erro-
neous assumption that it had been formed by the decomposition of
tribromsuccinic acid. In any case the subject seemed to demand a
more careful investigation. At the suggestion of Prof. H. B. Hill,
I have, therefore, repeated Bourgoin's work upon dioxymaleic acid,
and find that decomposition is in this case readily effected, but that
the product formed is not dioxymaleic acid.

The dibrommaleic acid needed for this investigation was made by
the oxidation of mucobromic acid with cold fuming nitric acid, a
method which had already been found in this laboratory to be much
more advantageous than the methods previously described. Muco-

* Bull, de la Soc. Chim.,xix. 482; xxii. 443.

t Ann. d. Chem. u. Pharm., ccvii. 223. } Ibid., cxcv. 70.

OF ARTS AND SCIENCES. 377

bromic acid was dissolved in cold fuming nitric acid, and the nearly
saturated solution allowed to stand for several days at ordinary tem-
peratures. When the mixture had completely solidified, the nitric
acid was expelled by gectle heat, and the dry residue distilled from a
small retort. In order to obtain a perfectly pure product, the distil-
late was dissolved in water, neutralized with baric carbonate, and the
baric dibrommaleate precipitated by the addition of alcohol. The
barium salt was then converted into the sodium salt, and from this the
silver salt was made by precipitation with argentic nitrate.

According to Bourgoin, dioxymaleic acid is formed by heating ar-
gentic dibrommaleate with water to 150°. On opening the tubes after
heating for five hours he found that carbonic dioxide escaped, that
argentic bromide had been formed, and that the liquid in the tubes
was distinctly acid. On the addition of argentic nitrate to this acid
solution carefully neutralized with ammouic hydrate he obtained a
white insoluble silver salt, which gave on ignition a residue closely
agreeing with the weight of metallic silver required by the formula
Ag2C4H206. He found tbe acid to be a white crystalline solid, which
was readily soluble in water and alcohol, and barely soluble in ether.
With tbe alkalies and alkaline earths it formed readily soluble salts,
and showed itself to be non-saturated in that it was capable of fixing hy-
drogen aud bromine. Bourgoin gives, however, no analytical results
whatsoever except the one already mentioned.

On following closely the directions of Bourgoin, I found that the
argentic dibrommaleate had been almost completely decomposed, and
that at least 95 per cent of the theoretical amount of argentic bromide
had been formed. The liquid in the tubes was strongly acid, but the
carbonic dioxide which escaped was by no means insignificant in quan-
tity. Although the aqueous solution gave a crystalline precipitate with
argentic nitrate after careful neutralization with amnionic hydrate, it
was thought advisable to distil the liquid with steam before preparing
salts for analysis, since it was found that the acid volatilized readily
under these conditions. The acid distillate was neutralized with calcic
carbonate, the silver salt precipitated by the addition of argentic
nitrate to the concentrated solution, and recrystallized from hot water.
The silver salt thus obtained closely resembled argentic acetate, and
its identity was established by analysis.

I. 0.4677 grm. of the air-dried salt gave 0.5249 grm. A^Br.
II. 0.2263 grm. of the air-dried salt gave 0.1465 grm. Ag.
III. 0.2056 grm. of the salt dried over sulphuric acid gave 0.1082 grm.
C02, 0.0330 grm. H20, and 0.1327 grm. Ag.

Calculated for
AgC2H302.

64.67

i.
64.46

Found.
II.

64.73

m.
64.54

14.37

14.35

1.80

1.78

378 PROCEEDINGS OF THE AMERICAN ACADEMY

H

c

H

The barium salt was also made by neutralizing the acid distillate
with baric carbonate. The salt obtained on evaporation was either
recrystallized from water, or precipitated from a concentrated aqueous
solution by alcohol. I failed to find any definite statements as to the
composition of the salt thus precipitated by alcohol, but experiments
of my own showed that it contained one molecule of water like the salt
crystallized from water.

I. 1.1456 grm. of the air-dried salt lost at 150° 0.0780 grm. H20.

II. 0.9187 grm. of the air-dried salt lost at 125° 0.0609 grm. H20.

III. 0.4561 grm. of the air-dried salt gave 0.3899 grm. BaS04.

IV. 0.6868 grm. of the air-dried salt gave 0.5841 grm. BaS04.

Calculated for Found.

Ba(C2H302)2 . H20. I. II. IH. IV.

H20 6.59 6.80 6.62

Ba 50.18 50.25 50.02

0.8554 grm. of the salt dried at 125° gave 0.7811 grm. BaS04.

Calculated for
Ba^HA)* Found.

Ba 53.73 53.69

It thus appears that acetic acid is formed in the decomposition of
argentic dibrommaleate by water at 150° and the reaction may be
expressed by the equation :

Ag2C4Br204 + 2 H20 = 2 AgBr + C02 + C2H402.

The weight of baric acetate which was obtained from the distilled
liquid amounted to about 60 per cent of the theoretical yield required
by this equation, while 85 per cent of the theoretical amount of free
acid, calculated as acetic acid, was found by direct titration of the con-
tents of the tubes. The higher result in the latter case may in part
have been due to incomplete expulsion of the carbonic dioxide, but the
most careful search failed to show that any well defined organic acid
except acetic acid had been formed in the reaction. Dibrommaleic
acid therefore yields, under the conditions prescribed by Bourgoin for
its conversion into dioxymaleic acid, carbonic dioxide and acetic acid.
The facts observed give no certain information as to the mechanism
of the reaction. If it is assumed that the body

OF ARTS AND SCIENCES.

379

C

H

OH

is first formed, it would pass at once into the aldehyde alcohol

H

OH

from which acetic acid would then be formed, precisely in the same
way that tartronic acid is formed from dibrompyruric acid,* or from
dioxytartaric acid.f It is more probable, however, that the ketone acid

COOH
I
CO

Ih

?OH

COOH

is the first product, and that this passes into acetic acid in a similar
way through the loss of carbonic dioxide.

* Griniaux, Bull, de la Soc. China., xxvii. 440.
t Kekule, Ann. Chem. u. Pharm., ccxxi. 230.

880 PROCEEDINGS OP THE AMERICAN ACADEMY

XXVII.

AN ADDRESS

DELIVERED AT THE MEETING OF APRIL 10, WHEN THE RUM-
FORD MEDALS WERE PRESENTED TO PROFESSOR A. A.
MICHELSON.

By Joseph Lovering.

For many generations it was assumed that no sensible time was
taken by light in moving over the largest distances. The velocity
of sound was found by noting the time which elapsed between seeing
the flash and hearing the report of an explosion. It was only in the
vast spaces of astronomy that distances existed large enough to un-
mask the finite velocity of light, and, in extreme cases, to make it
seem even to loiter on its way.

The satellites of Jupiter were discovered by Galileo in 1610; and
the eclipses of these satellites by the shadow of Jupiter became an
interesting subject of observation. It was soon noticed that the in-
terval between successive eclipses of the same satellite was shorter
when the earth was approaching Jupiter, and longer when the earth
was receding from Jupiter. The change of pitch in the whistle of a
locomotive, under similar motions, would suggest to the modern mind
an easy explanation. A Danish astronomer, Romer, without the help
of this analogy, deciphered the problem in astronomy. The eclipse
was telegraphed to the observer by a ray of light, and the news was
hastened or delayed in proportion to the distance from which it came.
In this way it was discovered that light took about 18 minutes to run
over the diameter of the earth's orbit. This discovery was published
by Romer in the Memoirs of the French Academy in 1675. The
mathematical astronomer Delambre, from a discussion of one thou-
sand of these eclipses observed between 1662 and 1802, found for the
velocity of light 193,350 miles a second.

Meanwhile Rbmer's method, after fifty years of waiting, had been
substantially confirmed in an unexpected quarter. Dr. Bradley of
the Greenwich Observatory, the greatest astronomical observer of his
day, was perplexed by certain periodical fluctuations, of small amount,

OF ARTS AND SCIENCES. 381

in the position of the stars. Suddenly the explanation was flashed
upon him by something he observed while yachting on the River
Thames. He noticed that, whenever the boat turned about, the
direction of the vane altered. He asked the sailors, Why ? All they
could say was, that it always did. Reflecting upon the matter, Brad-
ley concluded that the motion of the boat was compounded with the
velocity of the wind, and that the vane represented the resultant direc-
tion. He was not slow in seeing the application of this homely illus-
tration of the parallelogram of motion to his astronomical puzzle. The
velocity of light was compounded with the velocity of the earth in its
orbit, so that its apparent direction differed by a small angle from its
true direction, and the difference was called aberration. In spearing
a fish or shooting a bird, the sportsman does not aim at them, but
ahead of them. This inclination from the true direction is similar, in
angular measure, to what the astronomer calls aberration. Struve's
measurement of aberration combined with the velocity of the earth in
its orbit gave for the velocity of light 191,513 miles a second. Both
of the two methods described for obtaining the velocity of light depend
for their accuracy upon the assumed distance of the earth from the
sun. The distance adopted was the one found by the transits of Venus
in 1761 and 1769, viz. 95,360,000 miles.

During the last forty years, the opinion has been gaining ground
amon<r astronomers that the distance of the sun, as deduced from
the transits of Venus in 1761 and 1769, was too large by three per
cent. Expeditions have been sent to remote parts of the earth for
observing the planet Mars in opposition. The ablest mathematical
astronomers, as Laplace, Pontecoulant, Leverrier, Hansen, Lubhock,
Airy, and Delaunay, have applied profound mathematical analysis to
the numerous perturbations in the planetary motions, and proved that
the sun's distance must be diminished about 2,000,000 miles in order
to reconcile observations with the law of gravitation. Airy reduced
the distance of the sun by more than 2,000,000 miles, to satisfy the
observations on the transit of Venus in 1874. Glasenapp derived
from observed eclipses of Jupiter's satellites a distance for the sun of
only 92,500,000 miles. From these and similar data, Delaunay con-
cluded that the velocity of light is about 186,420 miles a second.

These triumphs of astronomical theory recall the witty remark of
Fontenelle, that Newton, without getting out of his armchair, calcu-
lated the figure of the earth more accurately than others had done by
travelling: and measuring to the ends of it. And Laplace, in contem-
plation of similar mathematical achievements, says : " It is wonderful

382 PROCEEDINGS OP THE AMERICAN ACADEMY

that an astronomer, without going out of his observatory, should be
able to determine exactly the size and figure of the earth, and its
distance from the sun and moon, simply by comparing his observa-
tions with analysis ; the knowledge of which formerly demanded long
and laborious voyages into both hemispheres."

The ancients supposed that light came instantaneously from the
stars ; a consolation for those who believed that the heavens revolved
around the earth in twenty-four hours. Galileo and the academicians
of Florence obtained even negative results.

While the number of physical sciences has received numerous addi-
tions during the last half-century, new affiliations and a more inti-
mate correlation have been manifested. In this mutual helpfulness
light has played an important part. The optical method of studying
sound, and the many varieties of flame apparatus, have made acoustics
as intelligible through the eye as through the ear.

Velocity being expressed by space divided by time, it is evident
that, in measuring an immense velocity,1 we must have at our com-
mand an enormous distance, such as we find only in astronomy, or
else possess the means of measuring fractions of time as small as one
millionth of a second. The first successful attempt to measure such
a velocity was made by Wheatstone in 1834. Discharges from a
Leyden jar were sent through a wire, having two breaks in it, £ of a
mile apart. The wire was in the form of a loop, so as to bring the
breaks into the same vertical line. The sparks, seen at these breaks,
were reflected by a mirror, at the distance of 10 feet, and revolving
800 times per second. The images of the two sparks were relatively
displaced in a horizontal direction. As the displacement did not ex-
ceed one half of an inch, the time taken by electricity to go from one
break to the other was less than a millionth of a second. Since the
distance was one quarter of a mile, the electricity travelled, in that
case, at the rate of 288.000 miles a second. If this experiment is
interpreted to mean that electricity would go over 288,000 miles of
similar wire in one second, as it probably often was at that time, the
conclusion is fallacious. The velocity of electricity, unlike that of
sound or light, diminishes when the length of wire increases.

In 1838, Wheatstone suggested a method for measuring the velocity
of light, which he thought was adequate for giving, not only the abso-
lute velocity, but the difference of velocity in different media.

In that year Arago communicated to the French Academy the
details of an experiment which he thought would give the velocity of
light in air or a vacuum. As his own health was broken down (he

OF ARTS AND SCIENCES. 383

died in 1853) he appealed to two young French physicists to under-
take the experiment. On July 23, 1849, Fizeau, by a method wholly
his own, made a successful experiment. A disk, cut at its circumfer-
ence into 720 teeth and intervals, and made by Breguet, was rapidly
rotated by a train of wheels and weights. A concentrated beam of
light was sent out through one of the intervals between two teeth of
the disk, which was mounted in a house in Suresne, near Paris, and
was sent back by a mirror placed on Montmartre at a distance of
about five miles. The light, on its return, was cut off from the eye or
entered it, according as it encountered a tooth or an interval of the
disk. If the disk turned 12.6 times in a second, the light encountered
the tooth adjacent to the interval through which the light went out.
With twice as many rotations in the disk, the light could enter the eye
through the adjacent interval. With three times the original velocity,
it was cut off by the next tooth but one, and so on. From the num-
ber of teeth and the number of rotations in a second, the time taken
by the light in going and returning was easily calculated. In this
way the velocity of light was found to be 195,741 miles per second.
In 1856 the Institute of France awarded to Fizeau the Imperial Prize
of 30,000 francs in recognition of this capital experiment.

In 1862, Foucault succeeded in measuring the velocity of light by a
wholly different method, all parts of the apparatus for it being em-
braced within the limits of his laboratory. The light emanated from
a fine reticule, ruled on glass and strongly illuminated by the sun. It
then fell upon a plane mirror revolving 400 times a second, by which
it was reflected successively to five other mirrors, the last of which
was plane, and returned it back, by the same path, to the revolving
mirror and reticule. The total distance travelled was only about 66
feet. As the revolving mirror had moved while the light was making
this short journey, the image of the reticule was displaced in reference
to the reticule itself ; and this displacement was the subject of measure-
ment. Although the time involved was only about one 15,000,000th
of a second, this brief interval was translated by the method of the
experiment into a measurable space, and gave 185,177 miles per
second for the velocity of light, differing from the best results of
astronomical methods by only 1,243 miles. Foucault was prompted
to this experiment by Leverrier, Director of the Observatory. Arago
was the first to propose the experiment. To obtain greater accuracy,
he placed the moving mirror in a vacuum, but without any advantage :
he said, " Le mieux est l'ennemi du bien." His modest claim was that
he had suggested to Foucault the problem, and indicated certain means

384 PROCEEDINGS OF THE AMERICAN ACADEMY

of resolving it. Babinet thought that the experiment admitted of ten
times greater accuracy. With three times only it might correct
Struve's value of aberration.

In 1873, Cornu, another French physicist, repeated the experi-
ments of Fizeau with a toothed wheel, the work extending over three
years. The observer was stationed at the Ecole Polytechnique.
The reflecting mirror and collimating telescope were placed on Mont
Valerian, at a distance of about 33,816 feet. Three different wheels
were tried, having 104, 116, and 140 teeth respectively, and rotating
between 700 and 800 times a second, the velocity being registered by
electricity. Cornu used at times all the eclipses from the first to the
seventh order. Calcium and petroleum light were tried, as well as
sunlight. A chronograph with three pens recorded automatically
seconds, the rotations of the toothed wheel, and the time of the eclipse.
More than a thousand experiments were made, 600 of which were
reduced. The velocity of light as published by Cornu in 1873 was
185,425.6 miles per second. The probable error was one per cent.
In 1874, Cornu gave the result of a new set of experiments made
by him in conjunction with Fizeau, over a distance of more than
14 miles between the Observatory and Montlhery. The experiments
were repeated more than five hundred times, mostly at night with
the lime light. The light was sent through a 12 inch telescope and
returned through a 7 inch telescope. The toothed wheel which pro-
duced the eclipse was capable of rotating 1,600 times a second. From
these experiments the velocity of light was placed at 186,618 miles.
The probable error did not exceed 187 miles. The time was re-
corded accurately within a thousandth of a second.

I come now to that which most interests us to-night, viz. the part
taken in this country for the measurement of these great velocities.
About 1854, Dr. Bache, chief of the U. S. Coast Survey, appropri-
ated $1,000 for the construction of apparatus to be used in repeating
Wheatstone's experiment on the velocity of electricity. But those
who were expected to take part in the investigation were called to
other duties, and the money was never drawn.

In 1867, Professor Newcomb recommended a repetition of Foucault's
experiment, in the interest of astronomy, to confirm or correct the re-
ceived value of the solar parallax. In August, 1879, Mr. Albert A.
Michelson, then a Master in the United States Navy, presented a
paper to the meeting of the American Association for the Advance-
ment of Science, on the measurement of the velocity of light. This
paper attracted great attention. Mr. Michelson adopted Foucault's

OF ARTS AND SCIENCES. 385

method with important modifications. In Foucault's experiment, the
deflection of the light produced by the revolving mirror was too
small for the most accurate measurement. Mr. Michelson placed the
revolving mirror 500 feet from the slit (which was 10 times the dis-
tance in Foucault's experiment) and obtained a deflection 20 times
as great, although the mirror made only 128 turns in a second. With
apparatus comparatively crude, he obtained for the velocity of light
186.500, with a probable error of 300 miles. This preliminary ex-
periment, made in the laboratory of the Naval Academy in May,
1878, indicated the directions in which improvements must be made
in order to insure greater accuracy. The distance from the slit to
the revolving mirror must be increased, the mirror must revolve at
least 250 times a second, and the lens for economizing the light must
have a large surface and a focal length of about 150 feet. With the
aid of $2,000 from a private source, Mr. Michelson was able to carry
out his ideas on a liberal scale.

His new experiments were made in the summer of 1879. The
revolving mirror, made by Alvan Clark and Sons, was moved by a
turbine wheel. Its rapidity of revolution was measured by optical
comparison with an electric fork which made about 128 vibrations
a second, the precise value being accurately measured by reference
to one of Kbnig's standard forks. The velocity generally given to
the mirror was about 256 turns a second. The distance between the
revolving and the fixed mirror was 1,986.26 feet. The light from
the moving mirror was concentrated on the fixed mirror by a lens
8 inches in diameter, with a focal length of 150 feet. These im-
provements on Foucault's arrangement were so advantageous that
Mr. Michelson obtained, even with a smaller speed in the revolving
mirror, an angle of separation between the outgoing and returning
rays of light so great that the inclined plate of glass in front of the
micrometer was not necessary ; the head of the observer not shutting
off the light. The mean result of 100 observations taken on 18
different days made the velocity of light 186,313 miles per second;
with a probable error of 30 miles.

In 1882, at the request of Professor Newcomb, Mr. Michelson made
a redetermination of the velocity of light at the Case Institute, in ( !leve-
land, Ohio, by the method already described, with some modifications.
The space traversed by the light in going and returning between the
two mirrors was 4,099 feet. Two slight errors in the reduction of his
former work were corrected in this. The velocity deduced from 563
new observations was 186,278 miles, with a probable error of 37 miles.
vol. xxiv. (n. s. xvi.) 25

386 PROCEEDINGS OP THE AMERICAN ACADEMY

In March, 1879, Congress had voted an appropriation of $5,000 for
experiments on the velocity of light, to he made under the direction
of Professor Newcomb. All the delicacy of instrumental construction,
all the skill of scientific observation, and all the resources of mathe-
matical discussion, were enlisted in this service. The method adopted
was that of the revolving mirror. The movable mirror was mounted
at Fort Myer. Two different locations were selected for the fixed
mirror, viz. the Naval Observatory, and the Washington Monument.
In one case the distance was 2,550.95 meters, or about 8,367.12 feet;
in the second case, 3,721.21 meters, or about 12,205.57 feet. Mr.
Michelson assisted in the observations until his removal to Cleveland,
in the autumn of 1880. The observations began in the summer of
1880, and were continued into the autumn of 1882, the most favorable
days in spring, summer, and autumn being selected. In all 504 sets
of measurements were made; viz. 276 by Professor Newcomb, 140
by Professor Michelson, and 88 by Mr. Holcombe. After a full
discussion of all the observations and the possible sources of error,
Professor Newcomb decided to rest the final result on the 132 sets
of observations made iu 1882 over the long distance between Fort
Myer and the Washington Monument. The velocity then obtained
was 186,282 miles. The velocity deduced from the three sets of
observations was 186,251 miles. The probable error of the first
result was about 19 miles.

For some future attack upon this problem Professor Newcomb
suggested a prism for the reflector with a pentagonal section, and
placed at such a distance that it could revolve through an arc of 36°
while the light was going and returning : 500 turns a second and a
distance of 19 miles would fulfil this condition. In the Rocky
Mountains, or the Sierra Nevada, stations from 20 to 30 miles
distant could be found, and with no greater loss of light from absorp-
tion than is produced by 2 or 3 miles of common air.

The first experiments made in Great Britain for the measurement
of the velocity of light were published by James Young and Professor
G. Forbes in the Philosophical Transactions of 1882. They adopted
the method of Fizeau. In 1878, between 600 and 700 observations
were made ; but the number of teeth in tlie rotating wheel was insuffi-
cient. New experiments were made in 1880-81 across the river
Clyde. Two reflectors were used at unequal distances, and the time
was noted when an electric light after the two reflections was at its
maximum. The corrected distances for the two mirrors were 18,212.2
and 16,835 feet. After an elaborate mathematical discussion of the

OF ARTS AND SCIENCES. 387

theory of this method, the velocity of light was placed at 187,221
miles. This value exceeded those obtained by Cornu or Michelson ;
but this might be explained by the color of the light used in the differ-
ent experiments. Mr. Young and Professor Forbes made some ex-
periments with lights of different colors, in confirmation of this view.
But Professor Michelson compared his 318 observations with sunlight
and 267 observations with electric light, and found that the difference
was in the opposite direction ; and in a differential experiment, when
half the slit was covered with red glass, he found no displacement.
Young and Forbes were attracted to their experiments on the velocity
of light by Maxwell's speculations on the electromagnetic theory of
light, and also as promising the most accurate method of obtaining the
parallax and distance of the sun. Their velocity of light combined
with Struve's constant of aberration made the sun's parallax 20".44o,
and its distance 93,223,000 miles.

When Arago, in 1838, suggested to the French Academy an exper-
iment on the velocity of light, and explained his method of making it,
which was essentially the one afterwards adopted by Foucault, he had
in view the settlement of the long controversy between the advocates
of the corpuscular and undulatory theories. Almost all of the differ-
ent classes of phenomena in geometrical optics can be explained by
either one of these theories, though even here the undulatory has the
advantage of greater simplicity. But in one respect the two theories
are antagonistic. According to the corpuscular theory, light should
move faster in glass or water than in air, for example. The undula-
tory theory reversed this proposition. Here was an experiment nm
cruets. In 1850, Fizeau and Foucault made the experiment, each in
his own way, and in both experiments the result was in favor of the
theory of undulations. It has been shown that in the case of air
alone lengths of many thousand feet are practicable. But the absorb-
ing power of water prevents the use of greater lengths than about
10 feet. Light would pass through 10 feet of air in less time than
T_i.T.?T of a second ; and the difference of time for air and water would
be only a fraction of that small fraction. Hence the exceeding deli-
cacy of the experiment.

In 1883, Mr. Michelson, at the request of Professor Newcomh, re-
heated Foucault's experiments for finding the difference of the velocity
of light in air and water. Foucault did not aspire to quantitative pre
cision in his results. The experiments of Michelson proved that the
ratio of the velocities was inversely as the indices of refraction. The
velocity with sunlight was a little greater than with the electric light;

388 PROCEEDINGS OF THE AMERICAN ACADEMY

which opposes the conclusion of Young and Forbes. When Mr.
Michelson covered half of the slit with red glass, the two halves of the
image were exactly in line. Experiments were also made ou the
velocity of light in carbon disulphide, which led to the inference that
its index of refraction was 1.77, and that orange-red light travelled
from one to two per cent faster than greenish blue light. Mr. Mich-
elson was enabled to make this investigation by a grant from the
trustees of the Bache Fund.

Various other methods of measuring the velocity of light have been
proposed. About 1850, Laborde suggested, in a letter to Arago, a
mechanical method of measuring the velocity of light. He supposes
two disks, with many holes at the outside, connected by a very long
axis and rotating rapidly. The light which was sent out through a
hole in one wheel would be transmitted or arrested by the second
wheel, behind which an observer was stationed. The distance between
the wheels, the time of rotation, and the order of the eclipse, would be
sufficient for calculating the velocity of light. Laborde imagined an
enormous axis more than 200,000 miles long. Moigno recommended
the substitution of a mirror for the observer and the second wheel,
which would double the distance travelled by the light. A distance of
1,640 feet, a disk 25 feet in radius, with 1,000 holes, and turning 360
times a second, would be more than sufficient to surprise the reflected
ray and stop it.

In 1874, Burgue suggested a new way of finding the velocity of light
by experiment. If a white disk, with a black radius, is rotated rapidly,
and at each turn is illuminated by an instantaneous flash, this radius
will appear immovable. If this flash is reflected on the disk from a
distant mirror, the black radius will be displaced. No details of the
arrangement of apparatus and no experiments were published.

In 1885, Wolf proposed the following arrangement. Two mirrors
were placed five meters apart and facing each other. The radius of
curvature of each mirror was five meters. The first mirror was 0.20
of a meter in diameter : the other, 0.05 m., revolved rapidly (200
turns a second). A slit was made in the centre of the large mirror
through which light was sent to the small mirror, forming an image on
the surface of the large mirror : this image became an object for the
small mirror, forming another image on the large mirror, at a distance
from the first mirror depending on the velocity of rotation. These
images could be sent out laterally by an inclined plate of thin glass,
and their distances measured by a micrometer. Wolf expected
advantages from the proximity of the two mirrors which would

OF ARTS AND SCIENCES. 389

more than balance those of the long distances used by Foueault and
Michelson.

The greatest difficulty which the undulatory theory of light has
encountered is found in the attempted reconciliation between the
requirements of the refraction of light and the aberration of light.
To explain refraction, the density of the lumiuiferous ether must be
greater when the index of refraction is greater. If a body moves, it
must carry its enclosed ether with it, as its refractive power does not
change. On the other hand, to explain the aberration of light, it must
be supposed that the ether in the telescope does not move with the
telescope ; that the ether sifts through the telescope, the ether in front
taking the place of the ether left behind ; or, as Young expressed it,
that the ether flows through the air and solid earth as easily as the
wind blows through the trees of a forest.

The difficulty can be eluded by supposing that a refracting body
carries along with it as much of the ether as it possesses in excess of
what would exist in a vacuum of the same bulk. This, added to what
is always sifting through it, would maintain its ether at a constant
density. What this fraction is which must travel with the body was
calculated by Fresnel. But while the refracting power has been pro-
tected, how is it with aberration ? That would be increased to a small
extent. But as the aberration is very small, only about 20 J" at its
maximum, the required change in its value might be masked by ordi-
nary errors of observation. Boscovich suggested to Lalande, in 1766,
that a telescope filled with water instead of air would test the theory ;
but he made no experiment. Wilson of Glasgow also proposed a
water telescope in 1782. In the course of time it appeared that not
only was the effect of the earth's motiou on refraction and aberration
under trial, but also the solar parallax, the motion of the solar system,
and that of other stars.

The case is clearly stated by Lodge in this way. Sound travels
quicker with the wind than against it. Is it the same with light?
Does light travel quicker with the wind? Well, that depends alto-
gether on whether the ether is blowing along as well as the air. If it
is, then its motion must help the light on a little ; but if the ether is
at rest, no motion of the air, or of matter of any kind, can make any
difference. According to Fresnel, the free is at rest, the bound is in
motion. Therefore the speed of light will be changed by the motion
of the medium ; but only by a fraction, depending on its index of re-
fraction, — infinitesimal for air, but sensible for water.

At an early day Arago investigated the effect which a change in the

390 PROCEEDINGS OF THE AMERICAN ACADEMY

velocity of light would produce on aberration and refraction. He saw
that a change of five per cent in the velocity of light would alter the
aberration by only one second, whereas the refraction in a prism of
45° would be affected to the extent of two minutes. He observed the
zenith distances of stars with and without the prism ; and also the
deviation of stars which passed the meridian at 6 a. m. and 6 p. m.
The observations were made with a mural circle and a repeating cir-
cle. Arago expected to find a difference of 10 or 15 seconds, but
found none. He thought that a difference no greater than one ten-
thousandth would have been manifested by his observations had it
existed. Arago attempted to explain his negative results by assump-
tions based upon the corpuscular theory of light. But Lloyd thought
that the change in the length of the wave would balance the chauge in
the direction of the ray. Arago's observations were communicated to
the Institute on December 10, 1816, and excited great interest. They
were quoted by Laplace and Biot. But the manuscript was mislaid
and not found until 1853, when it was published. Mascart thinks that
this experiment of Arago owes its reputation to Fresnel's explanation
of it by his fraction.

In regard to the wave-motion involved in the transmission of light,
Maxwell says : " It may be a displacement, or a rotation, or an elec-
trical disturbance, or indeed any physical quantity which is capable of
assuming negative as well as positive values. But the ether is loosely
connected with the particles of gross matter: otherwise they would
reflect more light." Then he asks the question, " Does the ether
pass through bodies as water through the meshes of a net which is
towed by a boat?" It is difficult to obtain the relative motion of the
earth and ether by experiment, as the light must move forward and
then back again. One way is to compare the velocities of light ob-
tained from the eclipses of Jupiter's satellites when Jupiter is in oppo-
site points of the ecliptic. Cornu referred, in 1883, to the difficulty
of observing these eclipses, especially when Jupiter is in conjunction
with the sun. On account of this difficulty observations have been
neglected for the last fifty years. Observations must be made near
quadratures. Cornu suggests a proper arrangement for this purpose.

At various times between 1864 and 1868 Maxwell repeated Arago's
experiment in a more perfect form. A spectroscope was used, having
three prisms of 60° each. A plane mirror was substituted for the
slit of the collimator. The cross-wires of the observing telescope
were illuminated by light reflected by a plate of thin glass placed at
an angle of 45° . Light went to the mirror and was sent back to the

OF ARTS AND SCIENCES. 891

wires from which it started after passing through six prisms. The
experiment was tried when the light started in the direction of
the earth's motion, and when in the opposite ; also, at different sea-
sous of the year. In all cases the image of the wires coalesced with
the wires.

Lodge states the case clearly thus : " If all the ether were free,
there would have been a displacement of the image of the wires. If
all the ether were bound to the glass, there would have been a differ-
ence on the other side. But, according to Fresnel's hypothesis, there
should be no difference either way. According to his hypothesis, the
free ether, which is the portion in relative motion, has nothing to do
with the refraction. It is the addition of the bound ether which
causes the refraction, and this part is stationary relatively to the
glass, and is not streaming through it at all. Hence the refraction is
the same whether the prism be at rest or in motion through space."
Maxwell is more guarded in his own statement of the case. He says:
•• We cannot conclude certainly that the ether moves with the earth.
For Stokes has shown from Fresnel's hypothesis that the relative
velocities of the ether in the prism and that outside are inversely as
the square of the index of refraction, and the deviation in this case
would not be sensibly altered ; the velocity of the earth being only
one ten-thousandth of the velocity of light."

In 1879, Maxwell wrote to Professor D. P. Todd, then at the Nau-
tical Almanac Office in Washington, asking him if he had observed an
apparent retardation of the eclipses of Jupiter's satellites depending on
the geocentric position of the planet. Such observations, he thought,
would furnish the only method he knew of finding the direction and
velocity of the sun's motion through the surrounding medium. Iu
terrestrial methods of measuring the velocity of light, it returns on its
path, and the velocity of the earth in relation to the ether would alter
the whole time of passage by a quantity depending on the square of
the ratio of the velocities of the earth and light, and this is quite too
small to be observed.

In 1839, Babinet made a very delicate experiment on the relation
of the luminiferous ether to the motion of the earth. He found that
when two pieces of glass of equal thickness were placed across two
beams of light which interfered so as to produce fringes, one of them
moving in the direction of the earth's motion and the other contrary
to it, the fringes were not displaced. The experiment was made
three times by Babinet, with new apparatus each time. He con-
cludes that here is a new condition to be fulfilled by all theories is

392 PROCEEDINGS OF THE AMERICAN ACADEMY

regard to the propagation of light in refracting media. According to
all the theories admitted or proposed, the displacement of the fringes
should have heen equal to many lengths of a fringe, that is, many
millimeters, while by observation it was nothing. Stokes has calcu-
lated the result according to Fresnel's theory, or his own modification
of it, and found that the retardation expressed in time was the same
as if the earth were at rest. Fizeau has pointed out a compensation
in the effect of Babinet's experiment. He says : " When two rays
have a certain difference of march, this difference is altered by the
reflexion from the turning mirror." By calculating the two effects
in Babinet's experiment, Fizeau finds that they have sensibly equal
values, and of opposite sign.

In 1860, Angstrom communicated to the Royal Society of Upsala
a method of determining the motion of the solar system by obser-
vations on the bands of interference produced by a glass grating. In
18G3 he published the results which he had obtained. After allow-
ing for Babinet's correction on account of the motion of the grating,
Angstrom finds that a difference in the direction of the observing
telescope with reference to the earth's motion might produce a dis-
placement of the fringes amounting to 49".8. Selecting the line D
in the fourth spectrum, he thought that the influence of the earth's
annual motion was verified, but that of the motion of the solar system
was less decided. The observations were more consistent with the
assumption that the solar system -moved with a velocity equal to one
third of that in its orbit, than with an equal velocity, or none at all.
In 1862-63, Babinet presented to the Academy of Paris a paper on
the influence of the motion of the earth on the phenomena produced
by gratings, which depend not on reflection, refraction, or diffraction,
but on interference. His principal object was a study of the motion
of the solar system. He calculated the effects to be expected, but
published no observations. In 1867, Van der Willigen measured the
length of waves of light by means of a grating. When a slit was
used, no effect was produced by the motion of the earth, the slit par-
taking of that motion. With a star, a movement of the earth in the
direction of the light had an effect. This is the theoretical result, and
agrees with Babinet's experiment, but is not applicable to solar light
when reflected by a mirror. That behaves as light from a terrestrial
source. In 1873, he rejects the proposition that the refraction of light
is modified by the motion of its source or of the prism. In 1874, he
seems to doubt the reality of the effect produced on diffraction.

In 1867, Klinkerfues used a transit instrument having a focal

OF ARTS AND SCIENCES. 393

length of eighteen inches. In the tube was a column of water eight
inches long, and a prism. He observed transits of the sun and of
certain stars whose north polar distance was equal to the sun's, and
which passed the meridian at midnight. The difference of right as-
cension is affected by double the coeilicient of aberration. He com-
puted that the column of water and the prism would increase the
aberration by 8". The amount observed was 7".l. In 1868-G9,
Hoek of Amsterdam discussed the influence of the earth's motion on
aberration. Delambre had calculated from the eclipses of Jupiter's
satellites that light must take 493". 2 in coming from the sun.
Hence the aberration must be 20".25o. Struve's observed aberration
made the time 497".8. Hoek decided in favor of Struve ; but he
thought that it was desirable that a new set of observations should be
made on the eclipses. Klinkerfues espoused the side of Delambre.
Hoek said that, if the earth's motion was taken into account, according
to Fresnel's fraction, different results would be harmonized. In 1868
he made experiments on a divided beam of light, the two parts going
in opposite directions through tubes filled with water. There was no
interference attributable to the effect of the earth's motion. As to
any influence to be expected from the motion of the solar system, he
thinks that motion must be insignificant compared with the initial mo-
tion of the comets, and with the cometary orbits, which are parabolas
with few hyperbolas.

In 1872, and on several previous occasions, one of the grand prizes
of the Academy of Paris was offered for an investigation of the effect
produced by the motion of the luminary or of the observer. This
prize, consisting of a gold medal or three thousand francs, was awarded
in 1874 to Mascart. He maintained that in Arago's experiment the
change in refraction produced by the fraction of the earth's motion
was compensated by the displacement of the observing telescope.
Mascart repeated Babinet's experiment with gratings, where the effects
of the motion of the telescope and of the grating would be addi-
tive, and found the sum small compared with Babinet's calculation.
He thinks that the change in the length of the wave caused by the
motion is compensated by the displacement of the measuring apparatus.
He concludes that reflection, diffraction, double refraction, and circular
polarization are powerless to show the motion of the earth, either with
solar light or that from a terrestrial source.

In 1871, Airy used a vertical telescope, and measured the merid-
ional zenith distance of y Draconis, the star by which Bradley dis-
covered aberration. It is about 100" north of the zenith. The tube

894 PROCEEDINGS OF THE AMERICAN ACADEMY

♦
of the telescope, which was 35.3 inches long, was filled with water.
The days of observation included the seasons of the equinoxes, when
the star is most affected iu opposite directions by aberration. The
observations were repeated in the spring and autumn of 1872. No
increase was produced in the aberration by the water in the telescope.

In 1873, Ketteler, in the Preface to the " Laws of the Aberration
of Light," enumerates thirty-nine persons who have investigated the
effect of motion on the phenomena of sound and light. From his own
analysis he concludes: 1. that a motion of the prism and telescope
perpendicular to the direction of a star produces no effect on the re-
fraction ; 2. that when the motion is in the direction of the star, the
velocity of the light is changed according to Fresners fraction of that
motion ; and 3. that for any intermediate direction it is changed
to the extent of that fractional part of the motion multiplied by
the cosine of the angle between the direction of the motion and the
direction of the star.

In 1859, Fizeau proposed an experiment for ascertaining if the azi-
muth of the plane of polarization of a refracted ray is influenced by
the motion of the refracting medium. When a ray of polarized light
passes through an inclined plate of glass, the plane of polarization is
changed, according to certain laws investigated by Malus, Biot, and
Brewster. The degree of change depends upon the inclination of the
ray, the azimuth of the plane of primitive polarization, and the index
of refraction of the glass. The incidence and azimuth being constant,
this rotation of the plane of polarization increases with the index of
refraction. This index being inversely as the velocity of light, the ro-
tation is smaller the greater this velocity. Fizeau used two bundles
of glass, four plates iu each, and slightly prismatic, inclined to one
another. One bundle was made of common glass ; the other of flint
glass. The angle of incidence for the ray was 58° 49'. When the
azimuth of the primitive plane of polarization was 20°, the rotation of
the plane of polarization was 18° 40' and 24° 58' for the two bundles.
By Fresnel's hypothesis the change in the velocity of light from the

motion of the medium is ± f - — ^ — ) v. The greatest available velocity

for the medium is that of the earth in its orbit, viz. 101,708 feet per
second (31,000 meters). At the time of the solstices this motion is
horizontal, and from east to west at noon. If the incident light comes
from the west, the velocity of light is diminished by Fresnel's fraction
of the velocity of the earth. If the light comes from the east, its
velocity is increased by the same amount. The change in the index of

OP ARTS AND SCIENCES. 395

refraction (or — j is equal to - (ju.2 — 1) ; this for an index of l.olo

amounts to TTtio* Measurements show that in glass, the index in-
creasing by a certain fraction, the rotation increases by a fraction
4^ times greater, and the consequent change in the plane of polari-
zation would be s-W. The total change on reversing the direction
from which the light came would be y^Vo* ^ tne incidence is 70°,
and allowance is made for the change of direction inside of the glass,
the fraction becomes j^^. When a ray of light falls on a single
plate of glass at an angle of 70°, if its plane of primitive polarization
makes an angle of 20° with the plane of refraction, this plane is
changed by 6° 40'. This multiplied by y^Vt? gives 16 seconds for
the probable effect of the earth's motion. With 40 such plates the
effect would be increased to lOf minutes. Two mirrors were used,
one to the east and the other to the west, and light could be sent by a
heliostat upon either one. The apparatus was easily turned through
180° so as to receive successively the light which travelled with or
against the earth's motion.

With a single pile of plates, highly inclined, and a second pile, less
inclined, of more highly tempered glass and in the opposite azimuth, a
rotation of 50° could be produced while the tendencies to elliptical
polarization were exactly balanced. The motion of the earth could
modify this result to the extent of only 2 minutes ; which is too small
for accurate observation. Fizeau then resorted to a device already
indicated by Botzenhart for amplifying this effect. A small variation
in the primitive .plane of polarization produces a greater effect the
smaller the azimuth of this plane. If the original azimuth is only 5°,
a small change in the azimuth trebles the value of the rotation. A
large rotation is first produced on a ray whose azimuth is large, and
then this rotation is largely changed by another pile so placed that the
ray enters it under a small azimutli. More than 2,000 measurements
were made under various conditions. For noon observations at the
time of solstice the rotation was always greater when the light came
from the west, and was less at other times of day. The excess in
the value of the rotation when the light came from the west varied
between 30' and loo', according to the different ways in which the piles
of plates were combined. The difference in the values of the rotation
according as the lisht came from the west or east was consistent with
a change in the index of refraction corresponding to Fresnel'a hypoth
esis. Fizeau indicated his intention of renewing the research with
improved apparatus, but no further publication on the subject by him
can be found.

396 PROCEEDINGS OP THE AMERICAN ACADEMY

Faye has criticised this investigation of Fizeau, on the ground that
he has taken no account of the motion of the solar system towards the
constellation Hercules. This motion, recognized by astronomers on
substantial evidence, amounts to 25,889 feet per second (7,894 meters)
at its maximum. Its influence is almost zero at noon of the solstices.
But it increases after noonday. Faye examines Fizeau's observations
at 4 p. M., and finds discrepances of 12' or 15' between the results of
theory and observation. By neglecting the term which corresponds to
the motion of the solar system, Fizeau's observations accord better at
all hours of the day. Must the inference be, Faye asks, that the solar
system does not move ? Tessan, in reply to Faye, says that the sun,
from which Fizeau derived the light used in his experiments, moves
with the rest of the solar system ; and that therefore Fizeau was justi-
fied in neglecting the term which expresses this motion, as of no effect
on his calculations. Fizeau's theory depends only on the relative
velocity between the source of light and the body which receives it;
that is, the velocity of revolution and rotation of the earth.

In 1881, Professor Michelson published the results of his investiga-
tion on this delicate problem. He first calculates the probable differ-
ence of time taken by the light in going and returning o^jr a given
distance, according as that distance lies in the direction of the earth's
motion or at right angles to it. If the distance were 1200 millimeters,
the difference of time translated into space would be equal to 5*5 of a
wave-length of yellow light. The apparatus was ingeniously devised,
so as to bring about fringes of interference between the two rays
which have travelled on rectangular paths. The whole apparatus
was then turned round bodily through 90°, so as to exchange the
conditions of the two interfering rays. Special apparatus was made
for this experiment by Schmidt and Haensch of Berlin, and was
mounted on a stone pier at the Physical Institute of Berlin. It was
so sensitive to accidental vibrations that it could not be used in the
daytime, nor indeed earlier than midnight. To secure greater stability
the apparatus was moved to the Astrophysikalisches Observatorium
in Potsdam, in charge of Professor Vogel. But even here the stone
piers did not give sufficient protection against vibration. The appara-
tus was then placed in the cellar, the walls of which formed the foun-
dation for an equatorial. But stamping with the feet, though at a
distance of 100 meters, made the fringes disappear.

The experiments were made in April, 1881. At this time of the
year, the earth's motion in its orbit coincides roughly with the motion
of the solar system, viz. towards the constellation Hercules. This

OP ARTS AND SCIENCES. 397

direction is inclined about 26° to the plane of the earth's equator, and
a tangent to the earth's motion in its orbit makes an angle of 20?,°
with the plane of the equator. The resultant would be within 25°
from the equator. The nearer the components are in magnitude, the
more nearly would the resultant coincide with the equator. If the
apparatus is placed so that the arms point north and east at noon, the
eastern arm would coincide with the resultant motion of the earth,
and the northern arm would be at a right angle to it. The displace-
ment produced by revolving the whole apparatus through 90° should
amount to ^V °f the interval between two fringes. If the proper
motion of the solar system is small compared with the velocity of the
earth in its orbit, the displacement would be less. Mr. Michelson
drew from these experiments the conclusion that there was not a suffi-
cient displacement of the fringes to support the theory of aberration,
which supposes the ether to move with a certain fraction of the earth's
velocity. The displacement, however, was so small that it easily
might have been masked by errors of experiment. Mr. A. Graham
Bell supplied Mr. Michelson with the money required for this inves-
tigation.

In 1886, Mr. Michelson and Mr. Morley published a paper on the
influence of the motion of the medium traversed by the light on its
velocity. Fizeau had made similar experiments. In both cases the
interfering rays were changed in velocity in opposite ways by flowing
air or water through which they were transmitted. With air having
a velocity of about 82 feet (25 meters) a second, the effect was so
small that it might easily be covered up by errors of experiment; but
with water it was measurable, and the result corresponded with the
assumption of Fresnel, that the ether in a moving body is stationary,
except, the portions which are condensed around its particles. In this
sense, it may be said that the ether is not affected by the motion of
the medium which it permeates. For this investigation, which was
made possible by a grant from the Bache Fund of the National
Academy, Mr. Michelson and Mr. Morley devised a new instrument,
called the Refractometer. Cornu writes of Michelson's experiments on
moving media : " Leur travail cone.ii dans l'esprit le plus eleve Exe-
cute* avec ces puissant moyens d'action que les savants dea Ktats-l nis
aimant a deployer dans les grandes questions scientifiques fait Le plus
grand honneur a leurs auteurs."

In 1887, Professor Michelson published another investigation of
the question whether the motion of the earth in its orbit carried its
ether with it. In his previous experiment his apparatus was sensitive

398 PROCEEDINGS OP THE AMERICAN ACADEMY

to the smallest jars, and it was difficult to revolve it without pro-
ducing distortion of the fringes, and an effect amounting to only ^V of
the distance between the fringes might easily be hidden by accidental
errors of experiment. In the new experiment the apparatus was
placed on a massive rock, which rested on a wooden base, which
floated upon mercury. The stone was 1.5 meters square and 0.3 of a
meter thick. At each corner four mirrors were placed, by reflection
from which the length of path traversed by the light was increased
to ten times its former value. The width of the fringes of interfer-
ence, which were the subject of observation, measured from 40 to
60 divisions of the observing micrometer. The light came from an
Argand burner sent through a lens. To prevent jars from stopping
and starting, the float was kept constantly in slow circulation, revolv-
ing once in six minutes. Sixteen equidistant marks were made on
the stationary framework within which the float moved. Observa-
tions were taken on the fringes whenever any one of these marks
came in the ranee of the micrometer. The observations were made
near noon, and at 6 p. m. The noon and evening observations were
plotted on separate curves. One division of the micrometer measured
5o of a wave-length. Mr. Michelson was confident that there was
no displacement of the fringes exceeding T^ of a wave-length. It
should have been from twenty to forty times greater than this. Mr.
Michelson concludes that this result is in opposition to Fresnel's
theory of aberration.

As late as 1872, Leverrier thought that a new measurement of the
velocity of light by Fizeau very important in the interest of astron-
omy; and in 1871 Cornu wrote that the parallax of the sun, and hence
the size of the earth's orbit, were not yet known with the desirable
precision. In 1875, Villarceau made a communication to the Paris
Academy on the theory of aberration. He says that the parallax of
the sun by astronomical measurement is 8". 86. Foucault's velocity
of light combined with Struve's aberration makes the sun's parallax
8".86. Cornu's velocity of light gives the same result only when it
is combined with Bradley's aberration, which differs from that of
Struve by O'^O. Villarceau thinks that there is an uncertainty
about the value of aberration on account of the motion of the solar
system. In 1883, M. O. Struve discussed seven series of observa-
tions made by his father, Nyren, and others, with various instruments
and by different methods, at the Observatory of Pulkowa. He was
certain that the mean result for the value of aberration was 20". 492,
with a probable error of less than TJj of a second. This aberration,

OF ARTS AND SCIENCES. 399

combined with the velocity of light as deduced from the experiments
of Cornu and Michelson, made the parallax of the sun 8". 78-4 ; dif-
fering from the most exact results of the geometric method hy only a
few hundredths of a second. Villarceau proposed to get the solar
motion by aberration ; selecting two places on the earth in latitude
35° 16' north and south, and, after the example of Struve, observing
the zenith distances of stars near the zenith. The tangents of these

latitudes are ± — - ; so that they contain the best stations for obtaining

the constant of aberration, and the three components of the motion of
translation of the solar system. In 1887, Ubaghs, a Belgian astron-
omer, published his results on the determination of the direction and
velocity of the movement of the solar system through space. For
finding the direction, he used the method of Folie. For calculating
the velocity, he combined the observations on three groups of stars,
the brightest belonging probably to the solar nebula. The resulting
velocity was only about ten million miles a year. Homann, working
on the spectroscopic observations at Greenwich, had obtained a ve-
locity of 527 millions of miles. As late as 1887, Fizeau studied the
nature of the phenomena when light was reflected from a mirror
moving with a great velocity, and inferred that aberration was the
same in this case as when the light was taken directly from a star.

The solar parallax, calculated from Cornu's last experiment on the
velocity of light and Delambre's equation of light (493".2 beinu
the time for passing over the radius of the earth's orbit) = 8".878
From Struve's observed aberration, 8".797

From Bradley's observed aberration, 8". 881

From Foucault's velocity with Struve's aberration, 8".8G0

From Leverrier's latitudes of Venus by transits, 8". 853

From meridian observations of Venus during 106 years, 8".859

From occupations of ^ Aquarius in 1672, 8".866

Glasenapp calculated the time taken by the light in travelling the
mean distance of the earth's orbit as equal to 500".85 ± 1.02. This
time combined with Michelson's velocity of light makes the solar par-
allax 8".76. Struve's constant of aberration with Michelson's velocity

gives a parallax of 8".81. From Gill's mean of the nine best i [era

determinations of aberration (= 20".496) the parallax comes out equal
to 8".78. If we regard the solar parallax as known, the eclipses u i \ - ■
nearly the same velocity as aberration, though the former is a group-
velocity and the latter a wave-velocity. Gill's parallax from ob

400 PROCEEDINGS OF THE AMERICAN ACADEMY

vations of Mars (8". 78) agrees with Michelson's velocity of light and
the mean constant of aberration.

In 1877-78, Lord Rayleigh, in his profound treatise on the Theory
of Sound, discussed the distinction between wave-velocity and group-
velocity. In 1881, he recognized the same difference in the case of
luminous waves. In the experiments of Young and Forbes, the wave-
velocity might be nearly three per cent less than the group-velocity.
With toothed wheels and the revolving mirror, group-velocity was the
subject of observation. Aberration gave wave-velocity ; Jupiter's sat-
ellites, group-velocity ; experiment, however, showed but little differ-
ence. Lord Rayleigh found formulae for the relation between these
two kinds of velocity, which involved the wave-length and the index
of refraction, and J. Willard Gibbs has compared them, and other for-
mulae proposed by Schuster and Gouy, with the experimental veloci-
ties of light. Michelson's experiment on the index of refraction of
carbon disulphide agrees with the assumption that he was dealing with
the group-velocity.

Although there is not a complete accordance between the results of
different methods of investigation, astronomers and physicists will be
slow to abandon the theory of undulations, and take up again the cor-
puscular theory of light. The latter theory has received fatal blows
from which it cannot recover. The undulatory theory, which started
with Huyghens more than two hundred years ago, and was elaborated
by Fresnel sixty years ago has survived many crises in its history,
and is supported by a wonderful array of experiments. Some of the
experiments of Mr. Michelson may require a modification in Fresnel's
interpretation. Stokes and Challis have worked for many years upon
it, and established it on mathematical principles differing from Fres-
nel's and from each other. Ketteler in his TJieoretische Optlh, pub-
lished in 1885, builds upon the Sellmeier hypothesis, that ponderable
particles are excited by the ethereal vibrations and then react upon
them. There remains Maxwell's electromagnetic theory of light,
which has been elaborated by Glazebrook and Fitzgerald, and is sup-
ported, to say the least of it, by remarkable numerical coincidences.

Discrepancies between theory and experiment are always to be
welcomed, as they contain the germs of future discoveries. We have
learned in astronomy not to be alarmed by them. More than once
the law of gravitation has been put again on trial, resulting in a new
discovery or in improved mathematical analysis. We may not expect
in light such a brilliant discovery as that of the planet Neptune. The
luminiferous ether is a mysterious substance, enough of a fluid for the

OP ARTS AND SCIENCES. 401

planets to pass easily through it, but at the same time enough of a
solid to admit of transverse vibrations. Stokes suggests water with a
litile glue dissolved in it as a coarse representation of what is required
of the ether.

Mr. G. A. Hirn has written recently on the constitution of celestial
space. He decides against the existence of an all-pervading medium.
He thinks that matter exists in space only in the condition of distinct
bodies, such as stars, planets, satellites, and meteorites. In nebulae it
is in a state of extreme diffusion ; but elsewhere space is empty. But
how would it be after the correction is applied for the equation of
light ? Humboldt said that the light of distant stars reaches us as a
voice from the past. The astronomer is not seeing for the most part
contemporaneous events. He is reading history ; and often ancient
history, and of very different dates. Stellar photography reveals mil-
lions of stars which cannot be seen in the largest telescopes, and new
harvests of these blossoms of heaven (as they have been called) spring
up like the grass in the night. Numbers fail to express their probable
distances and the time taken by their light in coming to the earth.
In the theogony of Hesiod, the brazen anvil took only nine days in
falling from heaven to earth. On the other hand, the reduction of
the sun's distance by three per cent not only affects its mass and
heat, but it changes the unit of measure for the universe. Such are
the remote results of any change in the estimated velocity of light.

We may thank Professor Michelson not only for what he has estab-
lished, but also for what he has unsettled. In his various researches,
which I have hastily sketched, but which require diagrams or models
to be clearly understood, he has displayed high intelligence, great
experimental skill and ingenuity, and unflagging perseverance. With
a high appreciation of his work, the Rumford Committee recommends 1,
and the Academy voted, that the Rumford Premium be awarded to
him.

vol. xxiv. (n. s. xvi.) 26

PROCEEDINGS.

Eight hundred and thirteenth Meeting.

May 29, 1888. — Annual Meeting.
The President in the chair.

In the absence of the Recording Secretary, Mr. Henry W.
Haynes was appointed Recording Secretary pro tempore.

The report of the Council was read by the Corresponding
Secretary.

The Treasurer's report was read by Mr. Augustus Lowell,
accepted, and'placed on file.

The Librarian's report was read by Mr. Henry W. Haynes,
accepted, and placed on file.

The report of the Rumford Committee was read by the
President as follows.

The Rumford Committee present the following report for
the year ending May 29, 1888 : —

An appropriation of $250 was recommended by the Com-
mittee, and voted by the Academy, to assist Professor
Trowbridge in his investigations on metallic spectra. The
committee have approved of bills amounting to 8219.72 for
printing in the Memoirs or Proceedings of the Academy
papers on the subjects of Light or Heat.

After mature deliberation, the Committee have voir.]
unanimously to recommend to the Academy the award of the
Rumford Medals to Professor Albert A. Michelson for his
determination of the velocity of light, for his researches upon
the motion of the luminiferous ether, and for his work on the

404 PROCEEDINGS OF THE AMERICAN ACADEMY

absolute determination of the wave-lengths of light ; and the
annexed vote is offered for the consummation of this award.

For the Committee,

Joseph Lovering, Chairman.

The report was accepted, and it was

Voted, That the Rumford Premium be awarded to Pro-
fessor Albert A. Michelson, as recommended by the Rumford
Committee, and that the Treasurer be authorized to pay from
the income of the Rumford Fund the amount required for the
preparation of the Rumford Medals.

On the motion of the Corresponding Secretary it was

Voted, To meet on adjournment on the second Wednesday
in June.

On the motion of Professor Jackson, it was

Voted, To amend No. 9 of the standing votes by chan-
ging the hour for the Annual Meeting from half-past three
o'clock, P. M., to half-past seven o'clock, P. M.

Dr. William Everett read a biographical notice of the late
Hugh A. J. Munro.

Professor Arthur Searle read a notice of the late Alvan
Clark.

The annual election resulted in the choice of the following
officers : —

Joseph Lovering, President.
Andrew P. Peabody, Vice-President.
JosiAH P. Cooke, Corresponding Secretary.
William Watson, Recording Secretary.
Eliot C. Clarke, Treasurer.
Henry W. Haynes, Librarian.

Council.

C. Loring Jackson,

Charles R. Cross, \ of Class I.

Amos E. Dolbear,

OF ARTS AND SCIENCES. 405

Henry W. Williams,

William G. Farlow, J of Class II.

Alexander Agassiz,

John C. Ropes,

Frederick W. Putnam, J> of Class I [I.

William Everett,

Rumford Committee.

Wolcott Gebbs, Jostah P. Cooke,

Edward C. Pickering, Joseph Lovering,

John Trowbridge, George B. Clark,

Erasmus D. Leavitt.

Member of the Committee of Finance.
Thomas T. Bouve.

The President appointed the following standing commit-
tees : —

Committee of Publication.

Josiah P. Cooke, Alexander Agassiz,

John C. Ropes.

Committee on the Library.

Henry P. Bowditch, Amos E. Dolbear,
Edward J. Lowell.

Auditing Committee.
Henry G. Denny, Thomas T. Bouye\

The President reported that he had requested Mr. W. W.
Story, of Rome, Associate Fellow, to represent the Academy
at the celebration of the eighth hundredth anniversary of
the University of Bologna, and this action was ratified by the
Academv.

406 PROCEEDINGS OP THE AMERICAN ACADEMY

Eight hundred and fourteenth Meeting.

June 13, 1888. — Adjourned Annual Meeting.

The President in the chair.

The President announced the deaths of Charles S. Brad-
ley and John Dean, Resident Fellows ; of Spencer F. Baird,
Associate Fellow ; and of Matthew Arnold, Foreign Honorary
Member.

The meeting was devoted to a commemoration of the late
Asa Gray. The President opened the proceedings in the fol-
lowing words : —

The death of Dr. Asa Gray has already been announced to the
Academy in the usual simple manner, and the Council has discharged
its duty in providing for an appropriate notice of his life and scientific
work. Before calling for the reading of this paper, I desire to say a few
words, and present some resolutions.

Nowhere else, except in his home, will Dr. Gray be so much missed
as in this hall, and from these meetings. He was elected into the
Academy on November 10, 1841, — a year before he took up his resi-
dence in Cambridge as Professor of Natural History in Harvard Uni-
versity. From the first, he was devoted to the scientific interests of the
Academy, and active in its administration.

He was the Corresponding Secretary for seventeen years, viz. from
1844 to 1850, and again from 1852 to 1863. He was highly qualified
for this office, as he had a large personal acquaintance with the scientific
men of Europe. He was Chairman of the Committee of Publication
for four years, viz. from 1846 to 1850, and in this capacity inaugurated
the publication of the Proceedings of the Academy in octavo form, to
supplement the ponderous volumes of Memoirs. He was President for
ten years, viz. from 1863 to 1873. As Corresponding Secretary and
President, he was an ex officio member of the Council, and for six years,
at other times, by election, — in all for seventeen years. But all these
official duties are only the means to an end : this end is the advance-
ment and diffusion of science. The richness of Dr. Gray's contribu-
tion to the Memoirs and Proceedings of the Academy admits of no
comparison ; though it was only the overflow from his abundant
learning, which was circulating at the same time in numerous other
channels.

In view of Dr. Gray's many and varied services to the Academy, of

OF ARTS AND SCIENCES. 407

his devotion to his chosen science, his exalted character, and his inspir-
ing example, I propose to the Academy to put on record the following
resolutions : —

Resolved, That, as Fellows of the Academy, we are deeply sensible of
the loss it has suffered by the death of one who has been associated with it
for forty-seven years, who has served it zealously in many relatious, and
who has done much to maintain its usefulness and the honor of its name at
home, and to make it respected throughout the world of science.

Resolved, That, as members of the community, we realize that it also
sustains a bereavement in the death of our lamented associate which is
not fully measured by his scientific work and reputation, great as these
were. By his unselfish devotion to his favorite studies, by his wide sym-
pathies, which condescended to the youngest and least knowing lover of
nature, by his joyous spirit and his simple Christian life, he has shown
that complete devotion to science can be reconciled with the highest inter-
ests of humanity.

Resolved, That a copy of these resolutions be sent to Mrs. Gray, his
helpful companion in his life's work, with the respectful sympathies of the
Fellows of the Academy.

The resolutions offered by the President were seconded by
Mr. Augustus Lowell.

Mr. President, — I have been asked to second the resolutions
presented, and to offer a few thoughts suggested by the affectionate
reverence with which we dwell upon the memory of our dear friend,
Dr. Gray.

In recalling him, we naturally think first of the affection he inspired.
"We remember the charming smile, the ready grasp, the strong expres-
sions of interest, with which he greeted us. We remember his readiness
to help, his patience, his sympathy. None ever sought his aid without
being impressed, first, by the certainty of his knowledge, and then by
the great kindness with which he gave it. The affection of all who
have ever benefited by his teaching is the highest tribute that can be
paid to that side of his character.

Dr. Gray was a man of singular simplicity. Thatjconsciousness of
power he had which is essential to high achievement, but so easily and
almost from the start did he assume the place that lie was to fill in
science, that he bore about him none of the signs of conflict, but held it
as a birthright. Ranking with the highest botanists of the ai:'-, 1"- did
more than perhaps any American to raise the appreciation of our schol-
arship abroad, while the same simple, lovable qualities which endeared
him to us charmed and fascinated his co-laborers there.

408 PROCEEDINGS OP THE AMERICAN ACADEMY

But perhaps the most remarkable trait in Dr. Gray was the stead-
fastness of his religious convictions under a strain to which so many
other minds succumbed. Early captivated with the theory of Darwin,
which ended by making its author an Agnostic, and has driven into
the ranks of unbelief nearly every one of its most earnest disciples,
Dr. Gray preserved his religious faith, and strove to reconcile the truths
of nature to what he knew to be the truth of God. Unlike Darwin, he
could not be dazzled by his own processes of thought, nor could he suf-
fer the ingenuity of any human theory to undermine his convictions of
fundamental truth. We all know with what earnestness he argued to
reconcile the two, and even those of us who cannot quite follow him in
both directions bear witness to the nobleness of his aim, and to the
essential service he performed for each.

Others better qualified than I will speak of his professional triumphs.
Mine is a simple tribute to the character of one whom I honored, rever-
enced, and loved.

I ask leave, Mr. President, to second the resolutions you have
offered.

After Mr. Lowell, President Eliot of Harvard University
addressed the Academy.

The life of Asa Gray always seemed to me a singularly happy one.
His disposition was eminently cheerful, and his circumstances and
occupations gave fortunate play to his natural disposition and capacity
for enjoyment. From opening manhood he studied with keenest in-
terest in a department of natural history which abounds in beauty,
fragrance, and exquisite adaptation of means to ends, and opens in-
exhaustible opportunities for original observing, experimenting, and
philosophizing. For sixty years he enjoyed to the full this elevating
and rewarding pursuit. These years fell at a most fortunate period ;
for the continent was just being thoroughly explored, and its botani-
cal treasures brought to light. Dr. Gray's labors therefore cover
the principal period of discovery and of accurate classification in
American botany. Merely to have one's intellectual life-work
make part of a structure so fair and lasting, is in itself a substantial
happiness.

His pursuit was one which took him out of doors, and made him
intimate with Nature in all her moods. It required him to travel
often, and so enabled him to see with delight different lands, skies,
and peoples ; it gave him intellectual contact with many scholars of

OF ARTS AND SCIENCES. 409

various nationalities, whose pursuits were akin to his own. Intel-
lectual sympathy and co-operation led to strong friendships founded
securely upon common tastes and mutual services. All these are
elements of happiness, — love of nature, acquaintance with the
wide earth, congenial intercourse with superior minds, and abiding
friendships.

Although Dr. Gray had no children, his domestic experience was
one of rare felicity. His life illustrated a remark of his friend Dar-
win, that with natural history and the domestic affections a man can
be perfectly happy. His way of living was that most agreeable to
a philosopher ; for it was independent, comfortable, and free alike
from the restrictions of poverty and the incumbrances of luxury.
With simplicity and regularity of life went health and a remarkable
capacity for labor.

All appropriate honors came in due course to Dr. Gray from acade-
mies, scientific associations, and universities, at home and abroad.
The stream began to flow as early as 1844, and continued to the
end of his life. With these honors came the respect and affection
of hundreds of persons who were devoted to the pursuit in which
he was a leader. His reputation was larger than that of a specialist ;
he was recognized as a clear thinker on philosophical and religious
themes, a just and sagacious critic, and a skilful and vigorous writer.
It is the greatest of human rewards to be thus enfolded, as years ad-
vance, in an atmosphere of honor, gratitude, and love.

Finally, Dr. Gray enjoyed the conscious satisfaction of having ren-
dered during his long and industrious life a great and lasting service
to his kind. For many years past he could not but know that he had
made the largest and most durable contribution to American botanical
science which had ever been made, and that he had done more than
any other man to diffuse among his countrymen a knowledge of botany
and a love for it. He knew, moreover, that by his own work, and
by the interest which his labors inspired in others, he had placed
on a firm foundation the botanical department of the University which
he served for forty-six years, and that the collections he had created
there would have for generations a great historical importance. To
have rendered such services was solid foundation indeed for heartfelt
content.

Professor G. L. Goodale, the successor of Dr. Gray as
Fisher Professor of Natural History in Harvard University.
next spoke.

410 PROCEEDINGS OF THE AMERICAN ACADEMY

Mr. President, — You have touchingly and truly said that this
was one of the scientific homes of him whom we commemorate. Here,
almost as distinctly as in the Herbarium which he founded, and which
was the principal seat of his ceaseless activities, we can feel his pres-
ence. Under the influence of that presence, those who knew the sim-
plicity and modesty of Asa Gray, as the members of this Academy were
permitted to know him, realize how completely out of place and taste
would be any formal eulogium. Therefore these tributes of affection
which are brought to-night do not partake of the nature of formal eu-
logies ; they are merely the presentation of certain phases of what has
been so often called a beautiful, fruitful, and well-rounded life.

Not long since, it fell to my lot, as an officer of another scientific
organization, to present a sketch of some of the leading events in Dr.
Gray's career. During the preparation of the outlines, two things kept
pressing themselves constantly upon my attention : first, the singleness
of purpose with which Professor Gray devoted himself to the establish-
ment of an Herbarium for the elucidation of the North American Flora ;
secondly, the zeal with which he engaged in the consideration of every
question touching the educational interests of Botany.

This evening you are to hear from one of his successors a biographi-
cal notice, in which these two features must be presented as parts of the
whole ; from another successor, to whom has been wisely intrusted the
care of the Herbarium and the completion of the monumental work,
you will learn of the devotedness and self-sacrifice by which Professor
Gray placed his collection among herbaria of the very first rank.
There would therefore appear to be occasion for a special, though very
brief, consideration of the second point, namely, some of the relations of
Professor Gray to botanical education. On this subject, the Correspond-
ing Secretary has invited me to speak.

Concerning Dr. Gray as a college instructor, I am permitted to quote
a few words from a letter by a leading jurist. He says : —

"I was in college when Dr. Gray was appointed to his professorship at
Harvard, and was, I think, the first, or one of the first, to whom he lectured.
I remember his lectures well, they were so full of knowledge and of enthusi-
asm, and so calculated to impress the young mind. I suppose he had
not lectured much of late years, and in his many other successes his
powers as a lecturer may have been overlooked by those who have written
of him."

Even early in his connection with the College, Dr. Gray was accus-
tomed to open his library and herbarium to those who expressed any
inclination to carry their botanical studies farther than the very narrow

OF ARTS AND SCIENCES. Ill

limits of the prescribed course. This generosity was accepted by many,
and most highly appreciated by them. It may be said that Dr. Gray
always maintained a deep interest in those who had availed themselves
of these special opportunities.

It was, however, by his educational works that Dr. Gray was able to
exert a wider influence on botanical education than if he had confined
himself to the formal college instruction. The works of his educational
series are of three grades : 1st, for very young children ; 2d, for our
grammar schools ; 3d, for high schools and colleges. All of these
treatises are characterized by the perspicuous style with which you
are all familiar in his general writings. The style is never involved :
it is essentially French in its clearness. This is as true of his first
work, published when he was barely twenty-six, as of his last.

Careful examination of all these many works shows that Dr. Gray's
constant endeavor was to lead the student to examine the plant, and not
the book, and thus to become self-reliant.

It is worthy of note, that in his books prepared for children Dr. Gray
never writes down to them ; there is no air of condescension. The sub-
ject is developed under his charmed hand as naturally as foliage and
flowers unfold from buds. The same may be said of his books for the
higher grades. There is never any straining after a certain pedagogic
effect. Hence his works have an attraction even for those who have
little or no interest in Botany itself. Even Mr. Ruskin, in that pleas-
ing, and yet withal provoking, sketch entitled " Proserpina," has a kind
word to say for Asa Gray's Text-Book.

One of the most remarkable things connected with Dr. Gray's series
is the fact that his Text-Book, now in its plan expanded into a four-
volume work, is laid down on substantially the same lines as his earliest
treatise, the " Elements." His earliest work is on the whole, however,
rather more dogmatic than the last.

Perhaps no single subject treated of in these earliest and latest
works is of greater general interest than his presentation of the doc-
trine of species. One realizes, as he compares the first hook with the
last, how difficult it must have been to detect relationship, when one
felt at heart that there was no more real relationship between different
species than exists between the different sorts of trinkets in a toy-shop;
those of the same kind beinu reckoned of one kind because of their
shapes, size, substance, and the fact that they have been perhaps
turned out by the same machine. In his last treatise, relationship
among plants is recognized as true kinship ; allied plants are of com-
mon blood. And yet the transition from one view to the other is

412 PROCEEDINGS OF THE AMERICAN ACADEMY

made so naturally, that no shock is felt at the profound change iu
statement.

Although Dr. Gray welcomed new methods of botanical instruction,
there was one which never met his approval ; namely, that which at-
tempts to introduce the student first to the lowest and microscopic
organisms, and passes thence to the higher plants. He held it to be
wrong in theory and vicious in practice to lead a student from the
unseen to the seen, — to investigate the unknown before taking up
the known. He naturally preferred to have examinations of flower-
ing plants precede a study of microscopic forms. It is of interest to
observe that the writers against whom some of his strictures were
directed have seen fit in a recent treatise to reverse their earlier
sequence.

Professor Gray manifested a sincere interest in public and private
schools, especially those designed to prepare boys for college. He
watched with attention the attempt to introduce elementary botany
as one of the requirements for admission to our College, and
he was one of the first to realize that, under existing conditions,
the requisition was likely to do more harm than good to botanical
education.

In 1872, after almost thirty years of college service, he relinquished
it to other hands. But he by no means gave up his care over the de-
partment. He remained to the last its judicious counsellor, patiently
answering questions and aiding in every way the happy solution of dif-
ficult problems of policy. Mr. President, I dare not trust myself to
dwell upon this phase of the subject which has been assigned me. The
memories of those hours of conference in which the learned teacher
gave advice to his younger colleague are yet too fresh to permit of any
other than the most general allusion. To the affection which every
member of this Academy felt for him was added, in the case of his col-
leagues, associated in the work of teaching, and brought into daily con-
tact with him, a feeling almost of reverence for a patience which never
overstepped its bounds. Hence you will pardon me if reference is made
only to the general subjects which appeared to command his interest in
the curriculum of his University.

Professor Gray viewed with pleasure the enlargement of facilities for
laboratory instruction in Botany. Twice in the fifteen years after he laid
down the charge of the department, he had the gratification of seeing
the natural development of his well-matured plans demand larger and
larger rooms for work. He watched with keen interest attempts to
secure accommodations for the accumulating illustrations of Economic

OF ARTS AND SCIENCES. 413

Botany, — collections which he had done everything to divert into the
channels marked out by himself. He was particularly attracted to the
important adjunct for instruction and research, and for popular education,
the Arnold Arboretum. He devised plan after plan to bring back into
the course of medical study its traditions of Botany, and he was never
wholly reconciled to give this science up to the schools of Pharmacy.
He urged his colleagues to overcome their prejudice against having at
least one large general elective in Botany, which might serve as a pos-
sible attraction to those intending to study the profession of medicine,
and he was fond of noting the success attending a smaller elective termed
Biology, specially designed for them. Believing that the study of
Botany might be made of great use as a discipline in a far different
field, he suggested establishing at the Bussey Institution an exceedingly
general elective, open to college students, under the title of " Rural
Affairs." This elective, comprising a modicum of Botany, was to be
arranged for the increasing number of those on whom sooner or later
might devolve the care of estates. In short, every phase of botanical
study was carefully examined by him with reference to the needs of his
University, and of sound learning in general.

His visits to St. Louis, to consult with the venerable founder of the
Shaw School of Botany in regard to the establishment, attest the fact
that his interests in education were not local in their character. The
wisdom with which that great enterprise was planned shows how far-
reaching was his good judgment. It may with truth be said, that the
Shaw School of Botany is at once a memorial of Henry Shaw,
and of his friends George Engelmann and Asa Gray. It may well
be numbered among Professor Gray's most enduring distinctions, that
he assisted in shaping the future of two working schools of Botany.

Besides his educational works, his critical reviews have exerted a
profound influence on the direction and scope of botanical education in
this country. These reviews were judicial and impartial, — discriminat-
ing rather than laudatory. His censure was given in a kindly spirit,
frequently with a light touch of humor, which was never suggestive of
cynicism. He could find faults, but not as a fault-finder; bis aim
was always to secure improvement. When, as in one memorable
instance, the errors pointed out by him were perpetuated in a second
edition, he could bring to bear a severity of criticism which v.
promptly remedial. His command over written speech, by which
sometimes words long unused by others were brought out of tlie
armory for a single thrust, make many of his reviews instructive
studies iu style.

414 PROCEEDINGS OP THE AMERICAN ACADEMY

If the present speaker were to refer to Dr. Gray's contributions to
education in general, instead of his influence on botanical education in
particular, it would be necessary to bring to your notice also his treat-
ment of Darwinism, and his popular expositions of scientific subjects.
Of the effect which these have had in shaping educational methods in
this country, in stimulating interest in natural history, and in promoting
research, others will perhaps speak.

And others must allude to the incalculable influence of his di^ni-
fied simplicity and noble character. But you will permit me, in closing
these remarks, to emphasize that Professor Gray was directly and indi-
rectly a teacher. Whoever examines the subject carefully will find
that Harvard University and this Academy account among the highest
services of the three naturalists whom we associate together their
direct and indirect influence as teachers, — Louis Agassiz, Jeffries
Wyman, and Asa Gray.

Professor Goodale was followed by Mr. Sereno Watson,
the Curator of the Herbarium which Dr. Gray founded.

It is an interesting coincidence that the life of Asa Gray was closely
parallel in this century to that of Linnasus in the last, — the one born
in 1707, and dying in January, 1788; the other born in 1810, and dy-
ing in January, 1888. Linnseus was easily the master of all that was
known in his day through the entire round of the natural sciences.
Now, the ablest man finds ample scope for all his energies in any single
branch of science. It is no longer derogatory to a man's fame that
the field of his work has been circumscribed ; it is rather a necessary
condition to success.

The life-work of Asa Gray was thus in' a great measure limited; and
it is as a systematic botanist that he stands forth as especially eminent.
The qualities of his mind fitted him remarkably for success in this de-
partment, — his accuracy of observation, not only seeing but knowing
what it was that he saw, his power of recognizing the resemblances and
affinities of plants as well as their differences, his clear judgment of the
weight to be given to characters in any given case, and not least his
wonderful memory, by virtue of which a plant once examined was never
forgotten, and anything once read or heard was ever at command. His
keen eye, his sound, quick, and critical judgment, his own confidence in
that judgment, and his power of expressing his opinions in clear and
concise language, gave him that weight of authority which has been ac-
knowledged by all contemporary botanists.

OP ARTS AND SCIENCES. 41")

Mr. Bentham, than whom there could be no one more competent to
express a just opinion, said of him in 1866 : —

"No botanists in America are more actively at work than Asa Gray,
whose well known accuracy of detail, connected with great correctness of
general views, a thorough knowledge of general botany in all its branches,
and a philosophical mind, has given him so high a rank among the votaries
of the science."

And again in 1873 : —

" Of all the modern contributions to the study of Composite none are*
more important for the accuracy of observation and the due appreciation
of characters and affinities than those of Asa Gray. His views may always
be implicitly followed without any danger of being led into error, although
sometimes a difference of opinion may exist upon such minor points as the
generic or subgeneric value to be given to a group."

It was the study of the flora of North America to which he devoted
these abilities, — the critical mastery of its orders, genera, and species in
their relations to each other as forming one great flora, and as related
more or less closely to the floras of other regions and other times.
Everything that was done by him had a more or less direct reference
to this main object, or was subordinate to it, whether it was giving
instruction to his classes, preparing text-books and manuals, or writ-
ing articles and notices for the journals. It may even be said that it
was only incidentally that he became a participant in the discussions
on Darwinism, for which his previous study of species had eminently
fitted him.

His early botanical work was done under difficulties. The prime
requisites to the work of any systematic botanist are an herbarium and
a library. Of these there were then none worthy the name in America.
The best to be found were those belonging to the Academy of Natural
Sciences at Philadelphia, and what Dr. Torrey had gathered about him
in New York, and it was here that Dr. Gray made all his early re-
searches. But type specimens of species for comparison were almost
wholly wanting, as none of the collections which contained them were
in this country. The collections of Clayton, Kalm, and Cat. shy; of
Walter, Bartram, and Michaux ; of Bradbury and Pursh, and even of
our own Government explorers, Lewis and Clark ; of Menziea ai '1
Douglas, and of all the arctic explorers, — upon which collections a
very large part of the species then known had been founded, — were
scattered through different European herbaria. Many of these were
consulted by Dr. Gray on his first visit to Europe, and this fact

416 PROCEEDINGS OF THE AMERICAN ACADEMY

gave to the published volumes of Torrey and Gray's Flora their great
value.

With his settlement in Cambridge began in earnest an endeavor to
gather together the herbarium and library that were absolutely necessary
to the continuation of his work. He had already plants which he had
himself collected in New York, New Jersey, and the Alleghanies, speci-
mens received from many correspondents throughout the country, and
some very valuable contributions from European botanists with whom
he was now in intimate and most friendly relationship. In the pre-
face to the early Flora acknowledgments are made of such gifts of
plants, illustrative of the American flora, from Sir William Hooker,
Dr. Richardson, Bentham, Lindley, Arnott, Spach, the elder De Can-
dolle, Schlechtendal, Trinius, Bongard, and Lehmann. By a contin-
uance of similar liberality on the part of both foreign and home friends,
and by purchase from his own limited means (it is impossible now
to say in what relative proportions), both herbarium and library gradu-
ally grew year by year, until after twenty-two years the house which
he occupied became too strait to hold the accumulations. The whole,
books and plants together, were then presented to Harvard University,
in the building which had been erected for them by Nathaniel Thayer,
and a small fund was raised by private subscription affording an income
barely sufficing for the most necessary expenses of their care.

Twenty-four years more have passed, and herbarium and library
have more than doubled in extent. Cases for plants have been added
once and again, and a large room has been built for the library, in which
the shelf-room is already too small. These later accretions have also
been for the greater part the gift of Dr. Gray. Through his member-
ship in numerous foreign societies, his position as botanical editor of the
Journal of Science, and especially his eminence as the great botanist of
America, whose acceptance and approval of a work were an honor,
many more additions have been made to the library than by purchase
with the small amount of money that could be spared for that purpose.
And thus the library has grown to be by far the best working botanical
library in the United States, though not indeed complete, nor what it
needs to be. In like manner the herbarium has from many sources
become very general in its character, and there is scarcely an accessible
region on the globe whose flora is not to some extent, and often very
largely, represented in it, and the number of known genera of phenoga-
mous plants of which it does not possess specimens is remarkably
small. This has been due mainly to Dr. Gray's wide acquaintance with
the botanists of the world, and especially to the close intimacy that has

OF ARTS AND SCIENCES. 417

always existed between him and the directors of the Royal Gardens at
Kew, Sir William J. Hooker, Sir Joseph D. Hooker, and the present
director, Mr. W. T. Thistleton Dyer, and to their generous sharing of the
wealth of specimens continually coming in to them from all parts of the
British Empire and from elsewhere, and to his like friendly relations
with Russian botanists, from whom have come large collections from
Central and Eastern Asia. But it is in the richness of its American
collections that the great value of the Gray Herbarium to the Ameri-
can botanist lies. In the possession of typical specimens it stands
unrivalled, including as it does the fullest sets of the collections of
Fendler, Lindheimer, Wright, Hall, Parry, Bolander, Brewer, Palmer,
Pringle, and others, the basis of many of Dr. Gray's classical papers,
and procured often by a costly outlay of time and labor in their de-
termination. It has also large sets of the collections made upon the
Government Pacific Railroad Surveys, the Mexican Boundary Survey,
and Geological Surveys, and very many specimens from collectors in
British America and the Canadian Government Surveys, and from
Greenland, Arctic America, and Alaska from various sources. It is
enriched, moreover, with his determinations and notes, and with his
memoranda upon the types of early species and the specimens of early
collectors which have been examined by himself at various times in
the herbaria at Kew, London, Paris, Madrid, Geneva, Berlin, Munich,
Vienna, and elsewhere. In short, the Herbarium has become what Dr.
Gray endeavored to make it, a mass of material fit to form the basis for
a Flora of North America.

That Flora he labored upon for years, but left incomplete. Finis
coronat opus. The crown of a finished life-work it was not permitted
him to wear. But his monument is the Gray Herbarium, that makes'
completion still possible. The care of this Herbarium and its Li
brary Harvard University has assumed ; and Harvard University may
be trusted, in honor of the man who has been its high honor, to main-
tain their integrity and pre-eminence, and to perpetuate, so far as
possible, that high grade of botanical work and research of which they
have hitherto been the main centre in America, and for which alone
they have been created.

The resolutions were adopted by a unanimous vote.

Professor William G. Fallow then read a biographical notice
of Dr. Gray, which had. been prepared at the request n\' the
Council, and is printed with the Report of the Council in Vol.
XXIII. of these Proceedings.

VOL. XXIV. (s. S. XVI.) 27

418 PROCEEDINGS OF THE AMERICAN ACADEMY

The following papers were presented by title : —

On Sodic Zincates. By C. Loring Jackson and Arthur M.
Comey.

Action of Sodium Malonic Ester on Tribromdinitrobenzol.
By C. Loring' Jackson and W. S. Robinson.

Contributions from the Physical Laboratory of Harvard
University : 1. Limits of the Ultra Violet Spectra of Metals.
By John Trowbridge and W. C. Sabine. 2. Photography of
the Infra Red Spectra. By John Trowbridge and J. C. B.
Burbank. 3. On a new Method of Measuring the Efficiency
of Electrical Transformers. By John Trowbridge. 4. Ex-
periments on the Electromagnetic Theory of Light. By
John Trowbridge.

Eight hundred and fifteenth Meeting.

October 10, 1888. — Stated Meeting.

The President in the chair.

The President announced the death of Rudolph Julius
Emmanuel Clausius, Foreign Honorary Member, and of
Samuel Kneeland, Resident Fellow.

The Corresponding Secretary read a letter from the Secre-
tary of the Franklin Institute, announcing that the Institute
is empowered to award certain medals for meritorious dis-
coveries and inventions which tend to the progress of the
arts and sciences.

The following gentlemen were elected members of the
Academy : — ■

John William Strutt, Lord Rayleigb, of Witham, to be a
Foreign Honorary Member in Class I., Section 3, in place of
the late Balfour Stewart.

Friherre Adolf Erik Nordenskiold, of Stockholm, to be a
Foreign Honorary Member, in Class II., Section 1, in place
of the late Bernhard Studer.

Carl Johann Maximowicz, of St. Petersburg, to be a For-
eign Honorary Member in Class II., Section 2, in place of the
late August Wilhelm Eichler.

OP ARTS AND SCIENCES. 419

Henry Sidgwick, of Cambridge, to be a Foreign Honorary
Member in Class III., Section 1, in place of the late Sir
Henry James Sumner Maine.

John Evans, of Hemel Hempstead, to be a Foreign Hon-
orary Member in Class III., Section 2, in place of the late
Hugh Andrew Johnstone Munro.

Johann Wilhelm Adolf Kirchhoff, of Berlin, to be a For-
eign Honorary Member in Class III., Section 2, in place of
the late Georg Curtius.

Charles Jacques Victor Albert, Due de Broglie, of Paris,
to be a Foreign Honorary Member in Class III., Section 3,
in place of the late Leopold von Ranke.

Leslie Stephen, of London, to be a Foreign Honorary
Member in Class III., Section 4, in place of the late Mat-
thew Arnold.

George Brown Goode, of Washington, to be an Associate
Fellow in Class II., Section 3, in place of the late Spencer
Fullerton Baird.

On the recommendation of the Rumford Committee, it was

Voted, That an appropriation of five hundred dollars
($500) be made from the income of the Rumford Fund
to assist Professor Trowbridge in his work on the metallic
spectra.

Voted, That an appropriation of five hundred dollars
($500) be made from the income of the Rumford Fund
to enable Mr. W. H. Pickering to observe the solar corona
during the eclipse of January 1, 1889, which is visible in
California.

On the motion of the Corresponding Secretary, it was

Voted, To meet, on adjournment, on the second Wednes-
day in November.

The following papers were read : —

Atmospheric Economy of Solar Radiation. By Arthur
Searle.

Further Study of the Structure of Iron Meteorites. By
Oliver W. Huntington.

Some Defects in Physical Terminology. By Amos E.
Dolbear.

420 PROCEEDINGS OP THE AMERICAN ACADEMY

The following papers were presented by title : —
Studies from the Newport Marine Laboratory, communi-
cated by Alexander Agassiz. A Preliminary Account of the
Development and Histology of the Eyes in the Lobster. By
G. H. Parker.

List of Plants collected by Dr. Edward Palmer, in 1887,
at Guaymas, Sonora, and at Los Angeles Bay and adjacent
Localities in Lower California. By Sereno Watson.

Eight hundred and sixteenth Meeting.

November 14, 1888. — Adjourned Stated Meeting.

The President in the chair.

Letters were read from John Evans and Henry Sidgwick,
acknowledging their election as Foreign Honorary Members ;
also, a letter from George Brown Goode, acknowledging his
election as Associate Fellow.

The following papers were presented : —

A New Galvanic Cell. Bv Amos E. Dolbear.

The Strength of the Microphone Current as influenced by
' Variations in Normal Pressure and Mass of the Electrodes.
By Annie W. Sabine.

Researches on Microphone Currents. By Charles R. Cross
and Annie W. Sabine.

The Strength of the Current of the Magneto Telephone as
influenced by Variations in the Strength of the Magnet. By
Charles R. Cross and Arthur S. Williams.

Eight hundred and seventeenth Meeting.

December 12, 1888. — Monthly Meeting.

The President in the chair.

Dr. Charles S. Minot presented a paper entitled "Researches
on the Problem of Growth and Death."

OP ARTS AND SCIENCES. 421

Eight hundred and eighteenth Meeting.

January 9, 1889. — 'Stated Meeting.

In the absence of the President, Dr. Henry W. Williams
took the chair.

The Corresponding Secretary read letters from the Due de
Broglie, A. Kirchhotf, A. E. Nordenskibld, and Lord Rayleigh,
acknowledging their election as Foreign Honorary Members;
from the Royal Bohemian Society of Sciences, announcing
the death of its president, Josef Jirecek ; and from Mr. W.
W. Story, the Delegate of the Academ}' appointed to attend
the celebration of the eight hundredth anniversary of the
University of Bologna.

Professor Joseph Lovering : —

Dear Sir, — I have to apologize for having so long delayed to
report to you, as President of the Academy of Arts and Sciences, the
proceedings and ceremonies which took place at Bologna on the cele-
bration of the eight hundredth anniversary of its famous University;
but I rely on your kindness to find an excuse for me, without my going
into the reasons for my delay.

In conformity with the mandate of the Academy, I proceeded to
Bologna on the 10th of June, and there was received most kindly by
the Rector and the Syndic, who furnished me with a most pleasant
apartment in a private house, where your representative had no possi-
ble cause of complaint, but, on the contrary, everything to be grateful
for personally. I make this statement because I have heard that
much complaint was made by others on the manner of their accommo-
dation, and this naturallv grew out of the fact of the numerous eon-
course of representatives from every quarter of the world, and of the
confusion naturally incident to such a state of things.

On the 11th, at nine in the morning, there was a formal presenta-
tion of all the delegates from the various universities, academies, and
scientific institutes to the Syndic. At noon the inauguration of the
equestrian monument to Victor Emanuel took place, at which all the
various delegates were present. At half-past eight in the evening
there was an artistic torchlight procession in honor of the King and
Queen, which was eminently imposing and picturesque.

On the 12th, all the professors and delegates met at the University
at nine o'clock, and there formed in procession, procc Win - through the

422 PROCEEDINGS OF THE AMERICAN ACADEMY

main street of the city to the building of the Archiginnasio. Nothing
could be more picturesque and interesting than this procession. By
special request of the authorities,' the greater portion of the delegates
were dressed in their academical and official robes, the effect of which
was most striking. Every variety of costume was there to be seen,
from all the quarters of the world. Mr. James Russell Lowell
and myself appeared in Oxford gowns as Doctors of Law. The
streets through which we passed were thronged with a dense crowd.
Every window and balcony was filled, and as the procession passed
along each delegation was saluted with hearty cheers, while from the
windows and balconies showers of roses and bay and laurel leaves
were rained upon it. The great Cortile of the Archiginnasio, which is
a noble old structure, was covered with a striped awning of blue and
red. The upper porticos were thronged by invited guests, the inter-
mediary columns and arches being decorated with wreaths and flowers,
while the floor was occupied by the students and professors of the
University and the delegates. In the central arch, facing the entrance,
a lofty canopy was erected, with a dais on which were seated the
King, Queen, and Prince Royal, attended by the officials of the court,
all in uniform and court dress. When the company had assembled, a
Cantica, the words of which were composed by Professor Panzacchi
and set to music by Baron Alberto Franchetti, was sung by a grand
chorus, and after this Professor Giosue Carducci, the distinguished
writer and poet, delivered an interesting discourse, the subject of
which was " Lo Studio Bolognese."

On the same day, at six p.m., all the representatives of the various
foreign universities and academies were invited to a great banquet
given by the government of His Majesty, the King of Italy, which
was largely attended, all present being in full dress, with their decora-
tions, by particular request.

On the 13th, all the delegates were again brought together in the
Cortile of the Archiginnasio, all in their academic or official robes and
with their decorations. The scene was similar to that of the previous
day, and equally imposing. On this occasion degrees of honor were
conferred by the Rector, one being given to each delegation. The
degree conferred on the American delegation was given to Mr.
Lowell, in accordance with the general wish of all the representatives
from America.

As each degree was given by the Rector, the whole delegation of
which the recipient of the honor was a member was called, and accom-
panied him to the foot of the canopy under which the royal family was

OF ARTS AND SCIENCES. 423

seated. The Rector then made a short address, placed on his finger
the large ring of the University, and then, withdrawing it, presented
him with his diploma. To this the Laureate replied in an address,
either in Latin or Italian, to the King and Queen, as well as to the
Rector and Professors.

In accordance with the request of the American Legation, the duty
of making this address in their behalf and that of the Laureate was
performed by me in Italian, and I am happy to say that it was well
received by the royal family, as well as the Rector and Professors and
the audience in general, so that I hope I did not do discredit to the
distinguished Academy which I had the honor to represent.

All the delegates were furnished with tickets to the Exhibition,
which was extremely interesting, and to the various concerts and cer-
emonies and entertainments given by the authorities. Among these
I have omitted to speak of the interesting discourse by Professor
Eurico Panzacchi, to which we were all invited by a committee of
the students.

A grand reception was also given by the King and Queen at the
Palace of the Prefettura, at which all the delegates were personally
presented to their Majesties and received with the greatest courtesy
and kindness.

The ceremonies and receptions having come to an end on the even-
ing'of the 13th, I left Bologna, and have nothing further to report.
With high regard, I have the honor to be

Yours most faithfully,

W. W. STORY.
London, July 10, 1888.

On the motion of the Corresponding Secretary it was

Voted, To meet, on adjournment, on the second Wednesday
in February.

The following paper was read : —

Comparison of Right Ascensions of Polar Stars in Recent
Catalogues. By Truman H. Safford.

The following papers were presented by title : —

Determination of Chromium in Chrome Iron Ore. By
Leonard P. Kinnicutt and George W. Patterson.

Measurements of Specific Inductive Capacity. By William
W. Jacques.

424 PROCEEDINGS OP THE AMERICAN ACADEMY

Eight hundred and nineteenth Meeting.

February 13, 1889. — Adjourned Stated Meeting.

The President in the chair.

The President announced the death of John Call Dalton,
of New York, Associate Fellow.

The Corresponding Secretary read letters from the New
York Academy of Sciences, requesting aid in erecting a
monument to Audubon ; from the United States Commis-
sioner to the Paris Exhibition, requesting information in
reference to the library of the Academy ; and from the Royal
Academy of Sciences of Turin, announcing the conditions of
competition for the seventh Bressa Prize.

The following papers were presented : —

The Application of the Stroboscopic Method to the Study
of the Excursion of the Electrode of a Microphone Trans-
mitter. By Charles R. Cross.

On the Rotation resulting from the Vibration of a Chladni
Plate. By Amos E. Dolbear.

On Pentamidobenzol. By C. Loring Jackson and A. W.
Palmer.

On Reversible Chemical Reactions. By P. C. Freer.

Eight hundred and twentieth Meeting.

March 13, 1889. — Stated Meeting.

The President in the chair.

The following gentlemen were elected members of the
Academy : —

Cecil Hobart Peabody, of Boston, to be a Resident Fellow
in Class I., Section 4.

Peter Schwamb, of Boston, to be a Resident Fellow in
Class I., Section 4.

Henry Willey, of New Bedford, to be a Resident Fellow
in Class II., Section 2.

OF ARTS AND SCIENCES. 425

Harold Clarence Ernst, of Boston, to be a Resident Fellow
in Class II., Section 3.

Samuel Henshaw, of Cambridge, to be a Resident Fellow
in Class II., Section 3.

Edward Hickling Bradford, of Boston, to be a Resident
Fellow in Class II., Section 4.

Arthur Tracy Cabot, of Boston, to be a Resident Fellow in
Class II., Section 4.

David Williams Cheever, of Boston, to be a Resident Fel-
low in Class II., Section 4.

Reginald Heber Fitz, of Boston, to be a Resident Fellow
in Class II., Section 4.

John Homans, of Boston, to be a Resident Fellow in Class
II., Section 4.

Frederick Irving Knight, of Boston, to be a Resident Fel-
low in Class II., Section 4.

George Hinckley Lyman, of Boston, to be a Resident Fel-
low in Class II., Section 4.

Henry Pickering Walcott, of Cambridge, to be a Resident
Fellow in Class II., Section 4.

Franklin Carter, of Williamstown, to be a Resident Fellow
in Class III., Section 2.

Franklin Bache Stephenson, of Boston to be a Resident
Fellow in Class III., Section 2.

Frank William Taussig, of Cambridge, to be a Resident
Fellow in Class III., Section 3.

Barrett Wendell, of Boston, to be a Resident Fellow in
Class III., Section 4.

John Nelson Stockwell, of Cleveland, Ohio, to be an Asso-
ciate Fellow in Class I., Section 1, in place of the late Ezekiel
Brown Elliott.

Alexander Johnston, of Princeton, New Jersey, to be an
Associate Fellow in Class III., Section 3, in place of the late
Samuel Oilman Brown.

Dmitri Ivanowitsh Mendeleeff, of St. Petersburg, to be ;i
Foreign Honorary Member in Class I., Section •">.

Wilhelm Edward Weber, of Oottingen, to be a Foreign
Honorary Member in Class I., Section 3, in place <»!' the
late Gustav Robert Kirchhoff.

426 PROCEEDINGS OF THE AMERICAN ACADEMY

Anatole Francois Hiie, Marquis de Caligny, of Versailles,
to be a Foreign Honorary Member in Class I., Section 4, in
place of the late Rudolph Julius Emmanuel Clausius.

The Rumford Committee, through its chairman, Professor
Lovering, submitted the following report : —

The Rumford Committee recommend to the Academy an
appropriation of $100 to Professor E. Hall to enable him to
make certain investigations as to the fluctuations of tempera-
ture which occur at the inner surface of the cylinder of a
steam-engine in operation.

They also recommend an appropriation of $250 from the
same fund to Professor C. C. Hutchins, of Bowdoin College,
for the purchase of a large parabolic mirror of glass and its
mounting to be used for the completion of his investigation
on Lunar Radiation.

The report of the committee was adopted, and the appro-
priations voted.

The following papers were read : —

Experimental Proof of Newton's Law of Cooling as ap-
plied to Thermometers and to Rectangular Bars of Metal.
By William A. Rogers.

Experimental Proof that the Relative Coefficient of Expan-
sion between Brass and Steel is Constant between the Limits
of —5° Fahr. and +100° Fahr. By William A. Rogers. .

Professor Charles R. Cross read a paper entitled " On the
Relation of the Candle-Power of Incandescent Electric Lamps
to Current, Potential Difference, and Energy consumed in
them." By L. A. Ferguson and D. A. Center.

Professor Benjamin O. Peirce presented the following
papers by title : —

On Lunar Radiation. By C. C. Hutchins.
• On the Charging of Condensers by Galvanic Batteries.
By B. O. Peirce and R. W. Wilson.

OF ARTS AND SCIENCES. 427

Eight hundred and twenty-first Meeting.

April 10, 1889. — Monthly Meeting.

The President in the chair.

Letters were read from Messrs. Bradford, Cheever, Ernst,
Fitz, Knight, Lyman, Peabody, Schwamb, Stephenson, and
Wendell, accepting Fellowship in the Academy ; also, from
Messrs. Johnston and Stockwell, acknowledging their elec-
tion as Associate Fellows.

The President, having stated that the special business of
the meeting was the presentation of the Rumford Medals,
made the address printed among the papers of this volume,
on the numerous investigations which had been made on the
velocity of light, and their relation to the theories of optics
and delicate questions in astronomy. He then addressed
Professor Michelson in these words : —

Professor Michelson, — It is now my duty and privilege to pre-
sent to you, in the name of the* Academy and in the presence of these
members, here assembled to administer Count Rumford's Trust, the
gold and silver medals which together constitute the Rumford Pre-
mium. Be pleased to receive with these medals my congratulations,
and those of all the other members of the Academy, for the intelli-
gence, ingenuity, and perseverance with which you have labored to-
wards the solution of problems of the greatest concern to more than
one of the physical sciences.

Professor Michelson, having received the medals, made the
following reply: —

Mr. President and Gentlemen of the Academy, — I wish
to thank you sincerely, and to express my appreciation of the high
honor you have just conferred upon me. These medals I shall always
regard as a token of your encouragement and approval ; hut in addi-
tion they will, I trust, be a strong incentive for the future to do the
best that is in me to justify this expression of your kind appreciation.

The following paper was presented by title : —
Classification of the Atomic Weights in two Ascending

Series, corresponding to the Groups of Artiadsand Perissads.

By W. R. Livermore.

428 PROCEEDINGS OF THE AMERICAN ACADEMY.

Eight hundred and twenty-second Meeting.

May 8, 1889. — Monthly Meeting.

The President in the chair.

The President announced the death of Jonathan Ingersoll
Bowditch, senior Resident Fellow, elected on November 12,
1834, who died on February 19, 1889. He served as Treasurer
from 1842 to 1852, and was a member of the Council from
1854 to 1856, and from 1858 to 1863.

WRITERS OF NOTICES IN REPORT OF THE

COUNCIL.

George Rumford Baldwin, by Rev. George E. Ellis.
Jonathan Ingersoll Bowditch, by Rev. A. P. Peabody.
Samuel Kneeland, by Prof. J. D. Runkle.

E. B. Elliott, by Gen. F. A. Walker.

F. A. P. Barnard, by Prof. J. Lovering.
John C. Dalton, by Dr. H. P. Walcott.
M. E. Chevreul, by Prof. F. H. Storer.

R. J. E. Cladsius, by Prof. J. Willard Gibbs.
F. C. Donders, by Dr. H. W. Williams.

REPORT OF THE COUNCIL.

MAY 28, 1889.

Since the last Annual Meeting, on May 23, 1888, the
Academy has lost by death nine of its members ; — viz. three
Resident Fellows, George Rum ford Baldwin, Jonathan In-
gersoll Bowditch, and Samuel Kneeland ; three Associate
Fellows, Frederick A. P. Barnard, John C. Dalton, and
Rowland G. Hazard ; * and three Foreign Honorary Mem-
bers, Michel Eugene Chevreul, Rudolph Julius Emanuel
Clausius, and Franciscus Cornelius Donders.

RESIDENT FELLOWS.

GEORGE RUMFORD BALDWIN.

George Rumford Baldwin was elected a Fellow of the Acad-
emy on December 12, 1871, in Class I., Section 4, Technology and
Engineering. He was born on January 26, 1798, in the Baldwin
mansion, North Woburn, Mass.. and died there on October 11, 1888,
having devoted his lengthened life, with the full possession of his
faculties till its close, to the pursuits of practical science, as a sur-
veyor, a civil engineer, and a constructor.

The "Baldwin mansion," as an edifice, the estate connected with
it, and the generations and individuals successively owning and occu-
pying it in the performance of many valuable and distinguished pub-
lic services, justify a special notice here. The first of the family line
of the Baldwins in this Colony was Henry Baldwin, an emigrant from
Devonshire, England, coming with the earliest settlers. He was one
of the first founders of the town of Woburn, in 1640, took up lands
there, still held by his descendants, now including between five and
six hundred acres of farm and woodland, and was one of the select-
men of the town and deacon of the church.

* A notice of Rowland (i. Hazard could not be prepared for this volume;
but the notice of Ezekiel Brown Elliott, necessarily omitted last year, ia now
given.

430 GEORGE RUMFORD BALDWIN.

The old mansion, dating, as marked on one of its massive oak tim-
bers, from 16G2, with successive additions and improvements, is one
of the noteworthy survivals of our earliest times, in size, arrangement,
adornment, and in its well-preserved relics. A history and descrip-
tion of it might engage the zeal of the relic-hunters of the day. In
it are to be found implements, household utensils, paintings, orna-
ments, and sundry furnishings, with luxurious appliances, gathered
by the generations which have occupied it from birth to death. Piles
of trunks and boxes contain their private papers and settlements of
estates. Most interesting among its contents is a large, select, and
valuable library of many thousand volumes, collected principally by
the father and brothers of our associate, and by himself, giving evi-
dence of their scientific and literary tastes. Learned tomes in many
languages, costly illustrated works, series of scientific publications on
construction and engineering, and sumptuous editions of the best
writers in various departments of literature, are among its treasures.
It may be that, sooner or later, the collection will find its deposit in
the noble Public Library of the city of Woburn, as a memorial of one
of the oldest and most distinguished families of its citizens.

Of the fourth generation from the original colonist was that dis-
tinguished military officer in our Revolutionary war, honored with
the high esteem of Washington, and afterwards eminent as the earli-
est civil engineer in this State, Colonel Loammi Baldwin. He was
born in 1744. and died in 1807. Of his relations with his fellow
townsman, playfellow, schoolmate, and life-long friend, the farmer's
son, Benjamin Thompson, afterward Count Rumford, mention will
be made further on. Baldwin manifested from his earliest years an
ardent passion for acquiring knowledge, with skill in practical works,
and inventive ingenuity. He sought and obtained permission to at-
tend the lectures of Professor Winthrop at Harvard, walking to Cam-
bridge for that purpose with young Thompson, whose tastes and
genius were similar. After his patriotic service in the war, Colonel
Baldwin devoted himself to public offices in the town and in the
Legislature, and was the first High Sheriff of the County of Middle-
sex, never intermitting his scientific pursuits. On the projection of
the first of our public enterprises for more extended internal commu-
nication, the connection of the waters of the Merrimac with those of
the harbor by the Middlesex Canal, chartered in 1793, he was one of
its leading promoters. Its course lay through his own estate. It was
completed in 1803. Of this then signal enterprise, Colonel Baldwin,
as surveyor, engineer, and constructor, was the efficient aid of Mr.

GEORGE RUMFORD BALDWIN. 4ol

Weston, the English surveyor from Philadelphia. The canal served
its uses till superseded by the Lowell Railroad. Colonel Baldwin
was a Fellow of the Academy, and received the honorary degree of
A.M. from Harvard College in 1785.

The fourth child of the Colonel was the equally eminent Loammi
Baldwin, Esq., born in 1780, graduated at Harvard in L800. and a
Fellow of the Academy. The father, this son, and his two brothers,
yet to be mentioned, by their inheritance of scientific proclivities and
skill, furnish a marked illustration of transmission by heredity. The
second Loammi was employed in many public works by the govern-
ment, and was the constructor of the Naval Dry Docks in Charles-
town and Norfolk. His brother, James Fowler Baldwin, was
eminent as a Boston merchant and a State Senator, and, as one of
the Commissioners for the introduction of the Cochituate water into
Boston, found opportunity for the exercise of the family ingenuity,
and for engaging the professional services in his chosen work of his
younger brother, the subject of this sketch.

Our late associate was a son of Colonel Loammi Baldwin by a
second wife. His middle name of Rumford recalls the relations just
referred to, which existed between his father and the distinguished
Count. The farmer boy Thompson early developed the ingenuity,
the inventive and philosophical genius, and the skill in practical
science, especially in utilitarian directions, which won bim advance-
ment and distinction. As the close companion of young Baldwin,
thev eno-ao-ed together in the tentative trial of their experiments, and
walked to the lectures in Cambridge. After Thompson left his coun-
try in early manhood, under a cloud which obscured his patriotism,
the intercourse of the friends was suspended.* When Thompson had
attained rank and title at Munich, a correspondence began between
the easily reconciled friends, which is of great personal and historical
interest. In a letter following the birth of our present Bubject, the
Colonel writes to the Count, " I have had a son born to me to
whom I have given your name." The father wished this hoy. as he
trrew up, to succeed his brother Loammi at College; but the Bon was
disinclined to the pursuit of general scholarship in thai institution.
From his earliest years his bent was for mathematical and scientific
studies pursued by himself, and for practical out of door work in
waterwavs. surveying and engineering, in the examination ol mills
and water-power, dams and raceways. A- has been noticed, he bad
marked facilities for practice of this sort — after preliminary training

* See Memoir of Count Rumford, by the present wriier, published by the

Academy.

432 GEORGE RUMFORD BALDWIN.

in a school kept by Dr. Stearns, in Medford — by accompanying his
father and brother in field and office work. I find among his papers
some " Sketches " of proposed works made by him when employed
by his brother Loammi on the fortifications of Boston harbor in the
war of 1812, he being then in his fourteenth year.

I have in my hands the series of his diaries for more than fifty
years, with the daily entries upon them of his employments and
occupations. They contain most miscellaneous and comprehensive
notes of a life of marvellous industry, of wide travel and of useful
service. Only the most conspicuous and important of the enterprises
and undertakings in which he had the chief part and responsibility in
planning and supervising mechanical scientific and industrial schemes
for public improvements, can be noted here. We find him called upon
in all directions, on a large variety of occasions, as expert, witness, ref-
eree, or examiner of wharves and docks, in making surveys, drawing
maps and plans, exploring state and city archives, tracing estates, esti-
mating damages, assessing costs, consulting lawyers, instructing legis-
lative committees, and alone or with associates disposing of a vast
number of trivial or serious interests at a period when the develop-
ment of our railroad and manufacturing enterprises made a demand
for talent and skill not then easily obtained. He was instrumental in
forming the first associated company of engineers. It should be noted
here, that the varied and constant demand for his presence and ser-
vices, especially when called upon for any utterance in public before
many persons, was a serious strain upon one of his peculiar tempera-
ment. He was naturally shy, modest, diffident, and reticent, of most
retiring and undemonstrative ways. His social intercourse was very
limited, but his domestie ties and habits drew out from him very en-
gaging and tender qualities. Under no stress of circumstances could
he have made in public a speech of advocacy or argument. His deli-
cacy and refinement made him personally attractive to his intimates.

His brother Loammi, living in Charlestown in a spacious house
with garden and open grounds, had married the widow of Mr. Beck-
ford, who had been the partner of^ Mr. Joshua Bates, afterward the
London banker, when the firm had done business in Boston. Mrs.
Beckford had two daughters by her first marriage, but no child by
the second. One of these daughters, Catherine Richardson, became
the wife of George Rumford Baldwin, thus bringing the brothers into
a peculiar relationship. The younger brother had his first home in
Boston; but in 1842, after the death of Loammi in 1838, he moved
with his family to the Beckford house. Here, with occasional sojourn

GEORGE RUMFORD BALDWIN. 433

at "Woburn, he made his home, till his important work at Quebec —
soon to be referrc d to — occupied him there so many years, that be
became a householder and resident there. It was on an autumn visit
to the family mansion at Woburn that his life of ninety years closed
where it began.

On a fragment among his papers is the following brief mention of
some of his early engagements: " 1821. Built P. C. Brooks' Stoue
Bridge [over the Mystic, in Medford]. 1822, Sept. 16, to 1823,
Jan. 13, in Pennsylvania, with Loammi Baldwin. 1823, May l,to
1825, June 25, at Factories in Lowell. 1826. Surveyed Navy Yard,
Charlestown. Executed Marine Railway, Boston. 1831, Nov. 2.").
to 1833, Jan. 5, in England. 1833, Feb. or March, to 1834, Nov. 18,
on Lowell Eailroad. 1831, Nov., to 1836, April, in Nova Scotia.
1837, in Georgia on Brunswick Canal."

On November 6, 1845, he left Boston on an appointment as chief
engineer to make a survey for the route of the Buffalo and Mississippi
Railroad. His report, with specifications and estimates, was rendered
in 1847. In 1846 he was employed in the examination of water power
in Augusta, Georgia, and by the national government on the Board of
Commissioners for Dry Docks, in Washington and Brooklyn.

It was in July, 1847, that he was summoned to Quebec by a letter
from the Mayor, engaging him on a professional task which was to
occupy him diligently and laboriously till he completed it, in 1856,
though many years afterwards he was employed in further perfecting
some details and improvements in it. The enterprise resulted in the
introduction of water into the city from the Indian village of Lorette.
From August to October he was engaged in his survey. His field-
books and journals give full evidence of the general and some special
difficulties of the undertaking, in a region exposed to extreme changes
of temperature, to ice and freshets, with marshes, streams, and hard
rocky cliffs. He brought home with him his Profiles and Estimates,
on which he worked, presenting his report in print in the following
June. A contract for pipes was to be made in Scotland. He was
put in full superintendence of the work, under the Mayor and a
Water Board, in November, 1850, and in February, 1851, he had
made a survey of the city. After a short visit home, he sailed with
his family for Europe, in November, to superintend at a foundry in
Glasgow the casting of pipes, gates, etc., and to make arrangemi
for their shipment. The water-jet flowed in the city in 1852, and at
the end of 1853 he returned temporarily to Boston. Bis resignation
as Chief Engineer took effect in 1856.
vol. xxiv. (n. s. xvi.) 28

434 GEORGE RUMFORD BALDWIN.

From April, 1857, to July, 1858, he spent with his family in Eu-
rope, principally in Paris and London, with many excursions. He
was a most intelligent and inquisitive traveller. With accomplished
skill in draughting and etching, his pencil was ever busy in sketching
all the objects of special interest to him. His descriptions of such
objects in his journals are illustrated by a mass of drawings, more or
less perfected, preserved in his numerous collections. On October
25, 1848, he witnessed the introduction of the Cochituate water into
Boston. His papers show that his services in this undertaking had
been frequently engaged by his brother James, already mentioned as
one of the Water Commissioners. He mentions that on this occasion
he had in charge Sarah Countess Rumford. She had been permitted
by the Elector of Bavaria to take this title after the death of her father
in France. She spent her closing years in her fine family home in
Concord, N. H., the children of her father's friend having the over-
sight of her interests.

Mr. Baldwin was Assistant Consulting Engineer of the Mystic
Water Works in Charlestown. He notes that his photograph as
such an official was deposited in the copper box under the engine-
house in 1860. In February, 1861, he was appointed by the Legis-
lature as Engineer for the proposed Ship Canal at Cape Cod. His
report is among the State archives. Under date of Boston, June 6,
1864, he made his report on the survey for the Boston, Hartford,
and Erie Railroad.

In 1867 he was commissioned by Governor Bullock as State
Engineer of Improvements in Boston Harbor.

In this brief sketch, as already intimated, only the more important
of the great undertakings of a professional character which occupied
Mr. Baldwin's long life have been mentioned. His journals show
how fully every interval between these public works, and every mo-
ment of leisure, was employed in a vast variety of occasions in which
he was called upon as an expert. But he always found time to indulge
himself as an amateur in many congenial pursuits. He was skilled
in all farming, horticultural, and agricultural labors. He had his
work-bench, with the best of tools and scientific instruments. His
pen was ever busy in his own affairs, or for the service of friends.

It may be that his voluminous and carefully arranged and filed
papers may yet need to be consulted for the facts and information
contained in them.

Mr. Baldwin had but one child, a daughter, now married and
residing mainly in Quebec.

JONATHAN INGERSOLL BOWDITCIJ. 4.;."

JONATHAN INGERSOLL BOYVDITCH.

Jonathan Ingersoll Bowditch, the second son of Nathaniel
and Mary (Ingersoll) Bowditch, was born in Salem, October 15, 1806.
He inherited more amply than any other member of the singularly
gifted family his father's love and aptness for mathematical science,
aud had opportunity served he would have attained distinguished emi-
nence, as he won no little reputation, in that department. But after
his school days were over he entered upon a mercantile career, which
for many years left him scanty leisure for scientific pursuits, except
when at sea, and then, of course, with limited access to books and
none to teachers. He commenced business in Boston as a clerk in the
office of Messrs. Ropes and Ward, East India merchants, and in their
service made several voyages as supercargo. At that time a pas
to or from the East Indies round the Cape of Good Hope occupied
seldom less than four, and often five months. During these voyages
Mr. Bowditch made diligent use of his father's " Practical Navigator,"
taking observations, and keeping the ship's reckoning. At the same
time, by well chosen books and well directed courses of reading he
supplied in no small measure what in his earlier culture fell short of a
liberal education. In 1836 he became President of the American
Insurance Company in Boston, and held that office till 1864, retaining
his place on the Board of Directors till 1884. As a business man he
was distinguished not only by integrity of the most rigid type, but
equally by promptness, energy, efficiency, and a practical wisdom
closely akin to intuition. For these qualities his services were
sought as a director in institutions of all kinds, financial, industrial,
and charitable, and, numerous as were the trusts thus devolved upon
him, he never suffered one of them to be a sinecure, though very
many of them involved the gratuitous bestowal of large amounts of
time and labor. Indeed, he sustained for many years two characters
not often united in a superlative degree ; in business circles being re-
garded as of exceptionally sound and safe judgment and superior exec-
utive capacity, while the outside world looked upon him as a public
benefactor.

Mr. Bowditch shared, and no one who knew him can doubt that he
shared with his whole heart and soul, the honor that rests upon all
his father's children in determining unselfishly the question whether
he should devote a full third part of what would be their patrimony
to the publication of his great astronomical work, or Bhould accept
the subsidies that might be offered by the American Academy and the

436 JONATHAN INGERSOLL BOWDITCH.

friends of science. After his father's death in 1838, Mr. Bowditch
assumed the editorship of frequent successive editions of " The Prac-
tical Navigator," making such corrections and new calculations as
were needed, until the copyright was purchased by the United States
government, and so became public property.

Mr. Bowditch erected a private astronomical observatory in connec-
tion with his summer residence at Canton, and for many years rnadn
there observations of celestial phenomena, while he kept at the same
time a full meteorological register. He early interested himself in the
Observatory of Harvard College. He was the fellow townsman and
the early and lifelong friend of Benjamin Peirce, the first Professor of
Astronomy in the University, and was in intimate association with both
the Bonds and with Professor Winlock. Professor Pickering is of his
near kindred ; but Mr. Bowditch, though fully aware of his scientific
talents and attainments, hesitated to recommend him for his present
office, on the ground that he bad never made a specialty of astronomy,
and withdrew his objections only when persuaded that the training of
an accomplished physicist implies an ability iu the invention, improve-
ment, and direction of instruments and methods of observation which
the mere scientific knowledge of astronomy cannot furnish, — a propo-
sition which, if it needs proof, is abundantly demonstrated in the
recent history and present condition of the Harvard Observatory.
Mr. Bowditch was for many years a member of the committee ap-
pointed by the Overseers of the College for visiting the Observatory,
and rendered important aid to its administration, especially in the in-
vestment and increase of its funds, to which he was himself a generous
contributor, and for which he secured in many instances donations and
annual subscriptions. He took an equally active and beneficent inter-
est in various other departments and enterprises of the University, —
in the erection of Memorial Hall, in the establishment of the Scientific
Museum, and in the various improvements that have been made in the
Medical and in the Divinity School. He was among the efficient pro-
moters of the plan of furnishing cheap board for poor students, which
resulted in the erection and furnishing of a large and commodious
dining hall by the munificence of the late Nathaniel Thayer, and was
superseded only by the ample provision of board at cost for several
hundred students in Memorial Hall. The University recognized at
once Mr. Bowditch's long and varied services, and his claim to high
regard for scholarly and scientific attainments, by conferring on him
the degree of Master of Arts in 1849, and that of Doctor of Laws on
its two hundred and fiftieth anniversary, in 1887.

JONATHAN INGEBSOLL BOWDITCH. 437

Mr. Bowditch was also a benefactor of the Massachusetts Institute
of Technology, and a member of its Corporation and of its Financial
Committee.

In addition to all these public interests Mr. Bowditch for many
years had the management of large and important private trusts, es-
pecially of estates of widows and orphans, — a charge in which equal
reliance was placed, and never misplaced, in his painstaking fidelity,
his far-seeing prudence, and his financial skill. It is impossible to
overestimate the benefit conferred on dependent families by the ad-
ministration of trusts of this class in the hands of men who, like Mr.
Bowditch, employ the treasured experience and wisdom of a lifetime
for the security and well-being of those who cannot take due thought
for themselves.

It would be difficult to name any public charity, or any enterprise
for the welfare of the community, which has not had aid and further-
ance from Mr. Bowditch. He was a liberal giver, and in a good
cause he knew how to elicit gifts, even from those whose sympathies
are not easily moved. He could, almost in a literal sense, command
resources where others might plead in vain. His private character
gave weight to his influence. Truth, honor, purity, and benignity,
while they were manifested in his relations with society and the out-
side world, made him unspeakably precious in the more intimate circle
of home, kindred, and friends. Impulsive, but with only generous
impulses; free-spoken, but with the freedom of one who has nothing to
hide ; with rpuick indignation, but only for meanness and depravity, —
he has left the memory of a truly noble life, and a void place in the
community, larger and more diversified than it is often the lot of any
one man to fill. He retained much of his working power and active
usefulness till he had passed his eightieth year, though the last few
months of his life were a season of decline and infirmity.

Mr. Bowditch was happy in his domestic life and relations. He
married, in 1836, Lucy Orne Nichols, daughter of Benjamin R. and
Mary (Pickering) Nichols, and granddaughter of Timothy Picker-
ing. She died several years ago. Of their eight children, six. three
sons and three daughters, reached years of maturity, and are all still
living.

438 SAMUEL KNEELAND.

SAMUEL KNEELAND.

Dr. Samuel Kneeland was born in Boston, August 1st, 1821.
He was a thoughtful boy, early developed a taste for reading and
study, and entered the Boston Latin School in 1831.

A classmate for five years at the Latin School, and afterwards at
Harvard College, says : —

"My acquaintance with Kneeland dates from August 31, 1831.
On the morning of that day we met in front of the old Latin School-
house in Scbool Street. Here a friendship sprang up between us
which lasted till the day of his death.

" At school he was industrious and studious, stood well with his
teachers, took a high rank in his class, and at the end of the five
years' coui'se was awarded a Franklin medal. He left the Latin
School a fine classical scholar, but says himself that he knew next to
nothing of mathematics. He was, however, admitted to college with-
out conditions even in mathematics, and sustained a creditable rank in
that study. Without being especially popular, he kept up friendly
relations with all of the class. He was an inveterate reader of mis-
cellaneous books, but did not allow this habit to interfere with his
studies. He took a respectable rank at graduation, and was a member
of the Phi Beta Kappa Society."

Another classmate says : " Kneeland and I were excellent friends
all through our college course, without a break; and yet it was hard
to get at his inner life on account of a certain reserve which kept even
near friends somewhat at a distance. He had no liking for mathe-
matics, but developed an interest in and fondness for natural history
and the practical sciences."

Still another classmate says : —

" My recollections of Dr. Kneeland are very agreeable. I recall
him as one eminently self-poised, one who had come to feel

' My mind to me a kingdom is.'

Hence he was restrained from mingling in any boisterous carousals,
and kept also from courting the companionships of his classmates.
We could have pardoned his being a little more demonstrative, or his
taking a little more pains to hunt us up. When, however, we were
thrown into company with him, I, for one, can testify to his inspiring
by his pure life and manly bearing, and by his genial greeting, a
hearty respect.

" It was always a pleasure to meet him. He was one of the few

SAMUEL KNEELAND. t39

persons who have a decidedly handsomer and more impressive aspect
at sixty than at twenty. As I met him in recent years, I never
thought, ' How 1 wish I could make you young again ! ' He had
matured without losing the vivacity or the energy of youth. Indeed,
from his love of nature and his researches into things of lofty import
there had come a charm and an elevation into his face which led me
to rejoice that there was such a classmate so near at hand, and caused
me. to deplore his sudden death."

I have given these extracts from the letters so kindly furnished me
because they show us Dr. Kueeland during his early manhood, and
foreshadow so plainly what those who knew him well in his later years
found him to be. After graduating from Harvard College, in 1840,
he studied medicine under Doctors D. H. Storer, O. W. Holmes, Jacob
Bigelow, and Edward Reynolds, and at the Medical School, then
situated in Mason Street, from which he graduated in 1^43. From
1843 to 1845 he studied medicine and surgery in Paris. Returning
home after fourteen years of continuous study since his entrance into
the Latin School, we may suppose him particularly well prepared to
enter upon the practice of his profession, which he continued in Boston
till failing health in the early part of 1848 led him to make a voyage
to Brazil. After an absence of about six months, he returned in
good health, was married, and resumed the practice of his profession.
But a love of the study of natural history and a desire to travel led
him to relinquish the practice of his profession. He soon found conge-
nial employment as the Secretary of the Boston Society of Natural
History, a post which he filled with dignity and ability for five years.
In 18o6 he went to Portage Lake, the copper district of Lake Su-
perior, as physician and surgeon to several copper mining compan
where he remained one year. On his return he accepted the position
of contributor to Appleton's New American Cyclopaedia, then in
course of publication. The medical and zoological articles were as-
signed to him, of which he prepared over one thousand, beginning
with Volume III.

For two years he served as Secretary of the American Academy of
Arts and Stiences. He took an active part in those measures which
finally led to the grant by the State of the land on which the build-
ings of the Society of Natural History and Institute of Technology
now stand.

In 18G1 he entered the service of the United States government
surgeon, and served until February, I860, when lie w&a mustered out
with the rank of Brevet Lieutenant-Colonel of Volunteers.

440 SAMUEL KNEELAND.

In August, 1866, he was chosen Secretary of the Massachusetts
Institute of Technology, and here my personal acquaintance with
Dr. Kneeland begins. From this time till his resignation, Decem-
ber 31, 1878, twelve years, our official and personal relations were
constant and intimate. He was the secretary of the Faculty and of
the Society of Arts, as well as of the Corporation, and also for some
years the Professor of Zoology and Physiology. He performed the
duties of these various positions with ability aud with signal fidelity.

But few ever really knew Dr. Kneeland. Outside of his immediate
family he had but few intimate friends who were permitted to see and
know his inner self. His manner was cold, and to casual acquaintance
seemed somewhat forbidding, if not morose. He had little or no desire
or power to conciliate, and to those who were indifferent to him, or as
he thought disliked him, he was literally a sealed book. But for those
who had gained his entire confidence, and with whom he was in full
sympathy, no man was ever more transparent. He had a large and
loving heart, aud when he gave his friendship it was in full measure.
He hated deceit and all forms of indirection, and would never counte-
nance such practice either in word or manner. But he had a large
power of self-control, seldom lost or even exhibited temper, and
usually when he could not speak favorably and kindly was absolutely
silent.

He was not in the usual sense a popular man, but at the Institute
teachers aud students alike soon became accustomed to his manner,
and held him in high regard. The qualities of his character which
most deeply impressed me were strength, depth, and purity, and
during the more than twenty years of my intimate acquaintance with
him he lived a life of almost Spartan simplicity. His long vacations
at the Institute, and the period between his resignation, at the end of
1878, and his death, on September 28, 1888, were spent in travel and
literary work, but the results in books, miscellaneous papers, and in
Lowell Institute lectures, it is not needful here to enumerate.

In the latter part of March, 1888, he sailed from New York for
Hamburg, on what proved to be his last voyage across the Atlantic, to
meet a friend who was on his way home from the Philippine Islands,
and also to visit old friends, residents of that city. He had also the
pleasure during the summer of visiting several of the most interesting
cities and places in Germany, accompanied by his daughter and her
husband, to whom he bade a last good-by at Cologne.

He died very suddenly at Hamburg on the morning of September 28,
1888. The previous evening he spent with his friends, expecting to

FREDERICK AUGUSTUS PORTER BARNARD. 441

leave the next day for London, where he intended to remain a few
days, and then go on to Liverpool, where his passage for the home-
ward voyage had already beeu engaged.

Soon after an early breakfast taken in his room, he was heard to
fall, and was found with his overcoat on prepared to go out. By his
side lay a letter addressed to his son-in-law, evidently just written,
and stamped but not sealed, which he was on the point of mailing.
It was written in excellent spirits, detailing his preparations for the
homeward voyage, and the pleasure he should experience in again
seeing his daughter and her family. Such a termination of his life
had been anticipated for some years by Dr. Kneeland. He knew and
had informed some intimate friends of a chronic trouble of the heart,
which would in all probability suddenly end his days.

His remains were tenderly cared for by his friends, and laid by
loving hands in the cemetery at Ohlsdorf, near Hamburg. They have
also placed over his grave, for themselves and for his family and
friends on this side of the Atlantic, a monumental stone in affectionate
remembrance.

ASSOCIATE FELLOWS.

FREDERICK AUGUSTUS PORTER BARNARD.

Frederick Augustus Porter Barnard was born in Sheffield,
Massachusetts, on May 5, 1809. He died in New York on April 27,
1889. He was descended on his father's side from Francis Barnard
of Coventry, Warwickshire, England; and on his mother's side from
John Porter of Warwickshire, who emigrated to Massachusetts in
1628. In 1886 he furnished to a magazine an interesting article
" How I was educated." After receiving some elementary instruction
from his mother, he was sent, with a sister two years older, to the
village school in Sheffield. At six years of age, he had read from
Shakespeare, Cowper, Burns, Addison, etc., and began the study of
Latin. When nine years old, he lived for a time at Saratoga Springs
with his grandfather, General B. P. Porter, afterward. Secretary of
War under John Quincy Adams. At his leisure, be learned the art
of printing so effectually that he might have supported himself by it
had circumstances compelled him. From Saratoga Springs he went
to Stockbridge, Mass., and became interested in scientific stud
He entered Yale College in 1824, and graduated honorably in L828.

442 FREDERICK AUGUSTUS PORTER BARNARD.

Thirty years after his graduation he delivered the annual oration
before the society of the Yale Alumni. After some experience in
teaching at the Hartford Grammar School, he served as Tutor at Yale
in 1830. In those days young graduates of the College were broken
in for tutors at Hartford. Having an apprehension of deafness, which
was in the family and which had begun to reveal itself in his own
case, he connected himself, in 1831, as instructor with the Asylum for
Deaf and Dumb at Hartford, and a year later with a similar institu-
tion in New York. He originated a system still used in institutions
for the deaf and dumb, and published an analytic grammar with sym-
bolic illustrations. His leisure he devoted to the study of theology,
and to writing for the Hartford Review, which the poet Whittier was
editing. The friendship which he then formed with Whittier lasted
for sixty years, and was recognized by the poet in 1870 by dedicating
to liim his "Miriam." It was a graceful and generous compliment
which the great poet paid to the President of Columbia College when
he wrote that in those old days of their early acquaintance Dr. Bar-
nard had a greater chance of distinction even in literature and poetry
than himself.

In 1837 Dr. Barnard accepted an appointment as Professor of
Mathematics and Natural Philosophy in the University of Alabama;
in 1848 he became Professor of Chemistry. Meanwhile resuming his
theological studies, he took orders in the Episcopal Church. In 1854
he went to Oxford University in Mississippi as Professor of Mathe-
matics and Astronomy. In 1856 he was elected President; a title
changed in 1858 to Chancellor. While in the South, he urged the
claims of science to a higher place in a liberal education. In the
agitation between Northern and Southern views in regard to Califor-
nia, and other burning questions of the day, he spoke boldly for the
Union. On July 4, 1851, he delivered an address at Tuscaloosa, and
wrote a patriotic ode. At the outbreak of the civil war he found
himself in a delicate position ; he resigned his place in the University
in 1861, and sought to return North, but was refused a passport by
Jefferson Davis. For a time he was not permitted to leave the Con-
federate States, but finally was allowed to go to Europe. On his
journey, by the way of Fortress Monroe, he was stopped at Norfolk,
where he remained until the city came into the hands of the Union
army. Going then to Washington, he found friends and employment.
In 1862 he was engaged on the reduction of Gilliss's observations on
southern stars, under the Director of the Naval Observatory. In 1863
he was placed in charge of chart-printing and lithography for war maps

FREDERICK AUGUSTUS PORTER BARNARD. 440

in the office of the U. S. Coast Survey. In 1864, Mr. McCulloch, Pro-
fessor of Physics in Columbia College, N. Y., resigned, and linked his
fortunes with the Southern Confederacy. Professor Barnard applied
for the vacant office ; but the Trustees made him President.

President Barnard for many years had been interested in the im-
provement of education of every grade, and had claimed for science
its proper place in the curriculum. At his accession to the presi-
dency, the managers of Columbia College were induced by him to
enlarge the sphere of collegiate education by the establishment of the
School of Mines with an independent faculty. The friendly rivalry
between this department and the classical has invigorated both, and
brought greater numbers and more funds to the College. The cen-
tennial celebration of the revival aud confirmation by the Legislature
of the State of New York of the Royal Charter granted in 1754,
which occurred on April 13, 1887, took place under circumstances
of great present prosperity, and of rich promises for the future.

The labors of Dr. Barnard were never limited by his professional
duties. When the American Astronomical Expedition was organized
by the Superintendent of the U. S. Coast Survey, under the authority
of Congress, for the purpose of observing the solar eclipse of July 17,
1860, Professor Stephen Alexander was chief, and President Barnard
one of a large corps of able assistants. Mr. Alexander made the
report on the observations taken on Cape Chadleigh in Labrador, and
Mr. Barnard furnished an account of them for the American Journal
of Science. In 1867 Dr. Barnard was appointed one of the U. S.
Commissioners to the Universal Exposition in Paris, and his report
was printed by Congress. He was also Assistant Commissioner
General to the Exposition of 1878. He received the Cross of Officer
of Legion of Honor from the French Ministry, and a gold medal for
his work as editor in chief of Johnson's Cyclopaedia.

Dr. Barnard joined the American Association for the Advancement
of Science at the seventh, meeting, held at New York in 1846. He
was elected President of the fifteenth meeting, which was expected to
convene at Nashville, Tenn., on April 15, 1861. But that meeting
was prevented by the impending civil war. In L866 Dr. Barnard
thought that the time had come for renewing the meetings, and, in
co-operation with Joseph Lovering, the Permanent Secretary, he
arranged for a meeting in Buffalo, beginning on August 15, over
which he presided. Absence in Europe prevented him from giving
his valedictory address as retiring President at the next meeting, m
Burlington, Vt., but it was delivered at the meeting in Chicago in

444 FREDERICK AUGUSTUS PORTER BARNARD.

1868. Iu this address, after a comprehensive survey of recent ad-
vances in the science of all the physical forces, in the widest meaning
of that phrase, culminating in the doctrine of the conservation and
correlation of force, he strenuously opposed the philosophy which
confuses mental or moral power with physical force. His last words
were : " In conclusion, gentlemen, thanking you for the kind attention
with which you have listened to me, permit me to congratulate you
on the cheering auspices under which you are once more assembled.
You are here in a strength which recalls the happy days when your
Association was in the zenith of its prosperity and its usefulness, and
which justifies the hope that a fresh career of still more fruitful labors
and of higher services to humanity is before it." When, at a later
day, the American Metrical Society was instituted, Dr. Barnard was
its President.

Although Dr. Barnard made no great discovery in science, his
acquaintance with it was accurate and of wide range. Besides many
articles in the Cyclopaedia, he published eleven papers in the American
Journal of Science, and four in the Proceedings of the American
Association for the Advancement of Science. In 1859, as chairman
of a committee of twenty, he made an elaborate report to this As-
sociation on the History, Methods, and Results of the U. S. Coast
Survey. In 1860, he gave an interesting course of lectures in the
Smithsonian Institution, which were published in the Report of 1862.
In 1869, he published " Recent Progress in Science," and in 1871, a
volume on the Metric System. Dr. Barnard was no specialist ; his
love of knowledge included all the sciences, mathematical, physical,
and social ; and with it was blended a love of poetry and art. But
above all and through all he was an educator. For sixty years edu-
cation was the great subject of his thoughts and his professional
activity. In early life he published text-books on arithmetic and
grammar for schools and colleges. At a later period, he wrote a book
upon Art-Culture. He was a frequent contributor to Dr. Henry
Barnard's Journal of Commerce and Education. His essays and
reports on collegiate and university education, and his letters on col-
lege government, attracted attention throughout the country. Books
and the periodical and daily press equally served him for the dissemi-
nation of his independent views. His pamphlet on University Educa-
tion, addressed to the Trustees of the University of Mississippi, was a
plea for building up a university in this country; an idea afterwards
realized in more than one college, though at the time a voice crying
in the wilderness.

JOHN CALL DALTON. 445

The value of Dr. Barnard's many services to science and education
were recognized by academies and seats of learning. He was elected
an Associate Fellow of the American Academy of Arts and Sciences
in 1860. He was included among those named in the act of incor-
poration of the National Academy of Science by Congress, in 18G3,
was chairman of an important committee on weights, measures, and
coinage, and Foreign Secretary for a time. In 1871 he was chosen a
member of the Philosophical Society of Philadelphia. He was also
a corresponding member of the Royal Society of Liege. He received
the degree of LL. D. from Jefferson College in 1855, and from Yale
in 1859 ; and the degree of S. T. D. from the University of Missis-
sippi in 1861. The Regents of the University of New York made
him a Doctor of Letters in 1872.

It is understood that Dr. Barnard, devoted to the end to the cause
of education, and faithful to the College which had adopted and hon-
ored him, and to which he had given twenty years of his ripened
thought, has left his whole property (about 80,000 dollars) to Colum-
bia College, to be available after the death of Mrs. Barnard.

JOHN CALL DALTON.

Johx Call Daltox was born in Chelmsford, Mass., on February
2, 1825. He was the son of John Call Dalton,a practitioner of medi-
cine of unusual attainments and great success in his profession.

The subject of this notice was a graduate of Harvard College, of
the class of 1844, and received his degree in medicine from the same
institution in 1847. He at once devoted himself to the study of
experimental physiology and comparative pathology. After a short
residence in Boston, he was appointed to the Professorship of Physi-
ology in the medical department of the University of Buffalo. II«'
subsequently occupied the same position in the Vermont Medical Col-
lege, and in the Long Island College Hospital. In 1861 he became
assistant surgeon of the New York Seventh Regiment, and, later.
Brigade Surgeon U. S. Volunteers, serving in many important medical
duties till February 14, 186 I.

In 1864 he began his active work in New York and was until hia
death, on February 12, 1889, connected with the College of Phy-
sicians and Surgeons, at first as Professor of Physiology, and from
1883 as its President.

Dr. Dal ton was distinctly a teacher. To a thorough knowledge of
his subject he added the skill of a successful experimenter, and a com-

446 JOHN CALL DALTON.

mand of simple, exact, and attractive language, and in consequence
never failed to attract and retain the attention of listener and reader.
His great merit was this, that he made a place here for the study of
experimental physiology. He had been preceded by Magendie and
Claude Bernard in Europe, but in this country he was incontestably
the leader.

His first published work of importance was an essay on the " Cor-
pus Luteum of Pregnancy." This received a prize offered by the
American Medical Association, and was at once recognized as the
work of an independent and careful investigator.

In 1859 appeared the first edition of the "Treatise on Human
Physiology." This work was soon found to be better adapted to
the uses of students of medicine than any other then in existence,
and it has been the text-book of the majority of the physicians of this
country. It was one of the few text-books in which a teacher of phy-
siology had ventured to discard the well-worn commonplaces, which
seem to have a meaning, but are, in fact, nothing but abstract ideas,
and scarcely recognize a connection between the laws that govern the
universe and those that affect the human body. Vital force was a
phrase much used ; and so long as the student was taught that this
mysterious something was the only explanation of the functional
activities, investigation was necessarily at a standstill.

Dalton brilliantly asserted the claims of original investigation upon
the living subject to the chief place in the study of physiology. Both
the merits and the defects of this book were eminently characteris-
tic of the man. The subjects not sufficiently elaborated in this
work — and they are few in number — were those which had not
been experimentally studied by himself. That this text-book after
the lapse of thirty years is still in general use, is conclusive evidence
of the careful revisions and additions which the author was constantly
making.

His pronounced position brought upon him the reproaches and
threats of those who oppose vivisection under all circumstances.
Before legislative bodies, as well as at the bar of public opinion, he
maintained most convincingly the position he had early assumed, —
that vivisection, as he practised and taught it, was essentially humane.

In 1885 he published his "Topographical Anatomy of the Brain."
In this, his last contribution to science, as in his earliest essay, the
same good qualities are manifest. lie sought to display the structure
of the brain, — so far at least as the unaided eye can study it. For-
tunately he brought his great work to a conclusion, and has built by

EZEKIEL BROWN ELLIOTT. , 447

it a permanent monument to his skilful, well-trained eye, powers of
patient investigation and exact observation, and that felicity of state-
ment in which he never failed.

Upon this and upon the Treatise on Physiology his reputation here-
after will rest ; but they represent by no means all his useful activi-
ties. In all the higher relations of medical teaching he was .constantly
busied. A large number of instructive, original, and attractive ad-
dresses are proofs of the eagerness with which men listened to him.

He became an Associate Fellow of the Academy in 1855, and was
an early member of the National Academy of Sciences.

EZEKIEL BROWN ELLIOTT.

Ezekiel Brown Elliott was born in Sweden, near Brockport,
Monroe County, New York, on July 16, 1823. His father, John B.
Elliott, was a doctor of medicine. In 1845 the family moved to
Waterloo, Seneca County, New York. Ezekiel Elliott attended the
High School in that place, and later, the Academy at Geneva, New
York, where he was prepared for admission to Hamilton College
which he entered in 1840. He graduated in 1844, having been a
diligent student, displaying marked capacity for mathematics, as-
tronomy, and physics.

Subsequent to graduation from college, Mr. Elliott pursued the vo-
cation of teacher, until 1849, having charge of schools in Grand Rapids,
Michigan, Lyons and Macedon, New York, and Eastport, Maine. In
the last named year Mr. Elliott went to Boston and opened an office
as " Actuary and Electrician." It was to work in the latter capacity
that, during the first succeeding years, he devoted the greater part of
his time. I find, in Mr. Elliott's handwriting, the following account
of his connection with the extension of the telegraph service in the
Eastern States: "Just before the latter half of the year 1849, I
aided in opening the House Printing Telegraph line between Boston
and New York, taking charge of the office in Boston, having previ-
ously spent a few weeks in Providence, R. I., making myself familiar
with the operations necessary. Subsequently I became, for a short
time, joint proprietor of the line with Mr. W. O. Lewis, of Hartford,
Connecticut ; and still subsequently Mr. Lewis and myself were joint
superintendents. Later, I accepted the superin tendency of the Bos-
ton, Troy, and Albany Printing [House] line of telegraph. I relin-
quished my relatipn with the telegraph, I think, in 1854." Mr.
Elliott, however, industriously kept up his work in physics. I find

448 EZEKIEL BROWN ELLIOTT.

that in 1853 he was awarded a medal by the Massachusetts Chari-
table Mechanics' Association for " white flint telegraphic insulators."
Among his subsequent inventions were a portable dynamo ; an elec-
tric motor; a "storer" for electricity; and an "auroral telephone or
phono-flash."

From 1855 onward, Mr. Elliott was increasingly engaged in actu-
arial work. In that and the following year he prepared for the New
England Mutual Life Insurance Company tables of Two-Life Sur-
vivorships, comprising about 18,000 logarithmic values, computed on
the basis of the London Actuary's life table, at four per cent interest.
He had, in 1851, been employed by Mr. Amasa Walker, Secretary of
State, in actuarial work upon the pension lists of Massachusetts ; and
in 18G0 he prepared, at the instance of Mr. Oliver Warner, Secretary
of State, a pamphlet, to be issued to city and town officials, contain-
ing Instructions concerning the Registration of Births, Marriages and
Deaths in Massachusetts (pp. 56, octavo). In all his actuarial work,
whether done for private employers or for the Commonwealth, he
won a high reputation for accuracy of computation, ingenuity in
method, and wide range of knowledge.

In 1856 Mr. Elliott read before the Association for the Advance-
ment of Science, at its Buffalo meeting, the following papers : —

A. Tables of Prussian Mortality, interpolated for annual intervals
of a ore ; accompanied with formula and process for construction.

B. Discussion of certain methods for converting ratios of deaths to
population, within given intervals of age, into the logarithm of the
probability that one living at the earlier of two ages will attain the
later ; with illustrations from English and Prussian data.

C. Process for deducing accurate average duration of life, present
value of life annuities, and other useful tables involving life-contin-
gencie0, from returns of population and deaths, without intervention
of general interpolation.

At the meeting of the Association in Montreal, in 1857, Mr. Elliott
presented a paper on the Law of Mortality in Massachusetts, with
practical tables.

At the outbreak of the War of Secession, Mr. Elliott was called
to join in the work of the United States Sanitary Commission, into
which he entered with intense zeal and unflagging industry. In 1862
he made a Preliminary Report on the Mortality and Sickness of the
Volunteer Forces of the United States Government during the Pres-
ent War. This report led to his being elected a Fellow of the
American Academy of Arts and Sciences, and to his appointment as

EZEKIEL BROWN ELLIOTT. 449

Actuary to the Sanitary Commission. In the same year he made a
Statistical Report to the United States Sanitary Commission on the
Mortality and Sickness of the United States Volunteers. He also
contributed to the Republic Magazine of July, 1863, a paper on
Mutual Relations, as to Price, of Gold, Greenbacks, Silver Bullion,
and Silver Coin.

In 1863 he was sent as a delegate from the American Statistical
Association to the International Statistical Congress at Berlin, Sep-
tember 6th to 12th, where he presented a paper on the Military Statis-
tics of the United States of America, which was published, in Eng-
lish, by the Royal Printer of Prussia (44 pages, 4to), with appended
charts.

In 1864 Mr. Elliott was sent by the Sanitary Commission to
inspect the hospital and ambulance services of the armies engaged
in the Danish war. In the discharge of this duty, Mr. Elliott visited
the hospitals of the contending forces on both sides. The results of
his observations were submitted to the Sanitary Commission, but
have not, so far as I am advised, been published. During his tour
Mr. Elliott prepared a report on Prussian Mortality, which was pub-
lished in the Zeitschrift of the Royal Statistical Bureau of Prussia.

At the close of the War of Secession, Mr. Elliott was appointed
Secretary of the United States Revenue Commission, under the chair-
manship of Mr. David A. Wells, to which had been assigned the al-
most hopeless task of bringing order out of the weltering chaos of
customs duties and internal revenue taxes which had been imposed,
in defiance of all recognized laws of finance, by an uninstructed Con-
gress urged on by the continually recurring exigencies of a war of
unexampled proportions. Mr. Elliott's services in this capacity were
of incalculable value to the country, although they did not appear in
a form distinct from the general work of the Commission. After the
Commission was dissolved, and Mr. Wells became sole Special Com-
missioner of the Revenue, Mr. Elliott continued his work upon the rev-
enue system.* On the discontinuance of Mr. Wells's ollice, in 1869,f
Mr. Elliott was assigned to duty in the Bureau of Statistics, Treasury

* At the Chicago meeting of the American Association, in lSi'.s, Mr. Elliott
presented a paper on "Redemption Periods of Monetary Values involving
Life-Contingencies." Also, a paper on the " Metrical Unification of Interna-
tional Coinage," suggested by the International Monetary Conference of that
year in Paris.

t In this year Mr. Elliott prepared the extended section (00 pnges, 4to) on the
Moneys, Weights, and Measures of different Countries, embraced in Webster's
Counting-House Dictionary.

vol. xxiv. (n. s. xvi.) 29

450 EZEKIEL BROWN ELLIOTT.

Department, under the present writer. On the accession of Mr. Ed-
ward Young to the charge of the Bureau, Mr. Elliott became its
chief clerk, although his time continued to be devoted chiefly to the
preparation of reports and tables for the use of the Secretary of the
Treasury, and to answering special calls for information coming from
Congress, where financial questions were resuming the importance
they had lost during the periods of war and reconstruction.

In 1879 Mr. Elliott was appointed as the representative of the
Treasury Department upon the commission, under the chairmanship
of Mr. George William Curtis, constituted by President Grant to
frame rules and regulations to govern the admission of persons into
the civil service of the United States. He was chosen Secretary, and
occupied that position until the discontinuance of the work of the
Commission, which was, in fact, never legally dissolved, but ceased
to act through the failure of appropriation on the part of Congress,
owing to the hostility of the advocates of the Spoils System.

During this period Mr. Elliott continued to carry on his actuarial
duties in the Treasury Department. On December 1, 1871, he pub-
lished a letter on the Credit of the United States Government, with
comparative tables, addressed to the Secretary of the Treasury. In
1872, at the request of the Superintendent of the Census, he prepared
two papers, which are contained in the volume on Vital Statistics of
the Ninth Census, as follows : —

A. On the Table of Births, — correcting manifest incongruities in
the distribution, as to age, of the population under five years (pp.
517-531, inclusive).

B. Upon the Statistics of Mortality, — their reduction to the practi-
cal form of life tables (pp. x-xv. inclusive).

In 1879 he was joined with Mr. Thomas L. James, afterwards
Postmaster General, in a commission to report upon trials of the
electric light in the post-office building of New York City, and to
compare the relative economy of illuminating gas and of electricity
in lighting. The report of the commission is on file in the Treasury
Department.

In January, 1880, he prepared a series of Tables on the Credit of
the United States Government, giving the prices of the several classes
of the securities of the United States, together with the corresponding
rates of interest realized to investors therein. These tables were
published in the letter of the Secretary of the Treasury to the Chair-
man of the Finance Committee of the Senate, bearing date January
30, 1880.

EZEKIEL BROWN ELLIOTT. 451

Meanwhile he had continued his contributions to the meetings of
the American Association. In 1874, he presented, at the Ilartfunl
meeting, a paper on the Future Population of the United States. In
1875, he presented to the Detroit meeting a paper on the Subsidiary
Principle applied to Coinage and Money of Account.* At the
Buffalo meeting of 1876, he presented a paper on the Relative
Market Prices of Gold and Silver, and their influence on the Metallic
Monetary Standard of the United States, accompanied by a diagram
showing for thirty-four years the relative values of gold and silver ;
also the mint ratios adopted by the government with regard to the
gold and silver coinage, and their successive changes. At the Nash-
ville meeting of 1877, he presented two papers, one on the Monetary
Standard, the other on Standard Time. The last-named subject had
strongly attracted Mr. Elliott's attention, f and he continued to work
in this line until the adoption of the principle throughout the United
States.

In 1881, Congress passed an act creating the office of Government
Actuary ; and on the 1st of July Mr. Elliott was appointed to that
offk'e, which he held during the remainder of his life. The reports
prepared by him, in this capacity, are too numerous and various to
be individually noticed. Among the questions presented to him for
investigation and discussion were the apportionment of representa-
tion in Congress under the Tenth Census ; the condition of the United
States Sinking Fund ; the probable population of the country, at suc-
cessive dates ; the interest borne by various species of United States
Government securities ; and the effect of innumerable schemes for re-
funding the debt of the United States.}: The calls upon Mr. Elliott
as Government Actuary came from all the executive departments,
from the Smithsonian Institution, from hoth Houses of Congress, and
from many individual members and committees of Congress.

From the date of his appointment as Government Actuary, his

* In this year lie wrote the article on Coinage in Johnson's Cyclopaedia.

t In 1881, lie joined with Professor Cleveland Abbe in forming a Universal
Time Table, and in a report on Standard Time, made to the Metrological
Society, of which he was an active member.

| Senator Ilawley relates an amusing incident concerning a scheme which
had passed the House, and was being very indulgently considered by the
Finance Committee of the Senate, when Mr. Elliott appeared before the Cum
mittee and in a few minutes showed conclusively that the measure, if adopted,
would cost the government in excess of thirty-three millions of dollars. The
bill was thereupon incontinently dropped, and no one could afterwards be
found who had ever been in favor of it.

452 MICHEL EUGENE CHEVREUL.

official duties, and perhaps also his advancing years, restricted some-
what his contributions to the American Association. In 1881 he
presented papers to the Association at its Cincinnati meeting ; in
1882 he was President of Section I., and delivered an address upon
Economic Science and Statistics ; and in 1886 he made the last of
his many contributions, at the Buffalo meeting, in a paper dealing
with the rate of interest realized to investors in government securi-
ties, and offering formulas for determining the United States gold
value of silver bullion, when the London price per ounce of standard
silver and the price of Sterling Exchange between New York and
London are known.

His last scientific work was in connection with the Metrological
Society, in 1887. The second volume of the Proceedings of that
Society contains several communications from bis pen.

Mr. Elliott reached the close of his busy and useful life, at Washing-
ton, on May 24, 1888, in his sixty-third year. He was never married.

This brief review of his numerous and varied works fully justifies
the conclusion that he was among the most meritorious of American
statisticians and actuaries. Personally, Mr. Elliott was one of the
most amiable of men, with a childlike simplicity of character and
utterly incapable of offence or of guile.

FOREIGN HONORARY MEMBERS.

MICHEL EUGENE CHEVREUL.

Michel Eugene Chevreul died in Paris, April 9, 1889. He
was born at Angers, on a branch of the Loire, August 31, 1786.
The mere fact that any one prominent in the world of science should
have lived to the extraordinary age of almost one hundred and three
years is in itself so surprising and unprecedented that our thoughts
tend to dwell upon it unduly. A moment's reflection is needed to
bring into clear light the deeds of merit which made the man
illustrious. It is to be remembered that Chevreul was the contem-
porary of chemists who died when most of us here present were
children or not yet born. The friends of his youth were Amprre
and Gay-Lussac, D'Arcet, Davy, Oersted, Wollaston, and Berzelius.
Proust was his townsman, Vauquelin was his teacher, and it is evi-
dent that Fourcroy also had no little influence upon his scientific life.

MICHEL EUGENE CHEVREUL. 453

With Cuvier and with Ampere, as he tells us, he was accustomed
to hold unending friendly debates upon plans of classification and
upon the aims and methods of scientific research. Years after Chev-
reul had been left the sole survivor of the chemists of his time, he
was accustomed, in his lectures at the Gobelins, to make fond allu-
sion to the works of these men, almost as if they had been still alive.
Chenevix, also, as he assured us, "was a very conscientious man, with
immense encyclopedic knowledge."

It is said that Chevreul's early education was carefully looked to
by his father, who was a distinguished physician. Between the ages
of eleven and seventeen the boy studied at the Hcole Centrale of
Angers. He then went to Paris, where he immediately attracted
the attention of his teachers. AVhen twenty years old he was
assistant in charge of Vauquelin's laboratory, and teacher in a
scientific school which Fourcroy had founded. In this same year,
1806, he began to publish scientific papers, and before ten more
years had passed it was recognized by every one that he had gained
a place in the foremost rank of chemists. In 1810 he was appointed
assistant naturalist at the Garden of Plants, and he subsequently
succeeded to the Chair of Chemistry at the Garden, which had been
occupied, successively, by Geoffroy, the Brothers Rouelle, Lemery,
Macquer, Fourcroy, and Vauquelin. It is noticeable that no small
part of Chevreul's work related to subjects in the domain of natural
history, and that he published many papers in the Annales and
Memoires of the Museum. In 1816 he was appointed Professor of
Chemistry at the Gobelins, and director of its dye-houses. In 1826
he became a member of the French Academy of Sciences and a
foreign honorary member of the Royal Society of London.

As a chemist, Chevreul investigated a great variety of substances,
ran<dncr from the sujjar of diabetes and the oxidation products of
cork, on the one hand, to compounds of tungsten, uranium, and
zirconium, on the other. He edited several scientific journals also,
and contributed articles to encyclopaedias and dictionaries of science
and of technology. As an example of work of tliis kind may be
cited a review written by him, in 1822, of a French translation of Sir
Humphry Davy's Lectures on Agricultural Chemistry. This review
is well worth reading to-day, and it has a certain historic interest,
since it marks better even than the Lectures themselves the very
considerable progress which had been made at thai early period by
chemists who sought to classify and explain the phenomena of v<
table growth by referring them to scientific principles.

454 MICHEL EUGENE CHEVREUL.

At the very first, Chevreul devoted much attention to the investi-
gation of the coloring matters in dye-woods, not because the subject
was at that time in the line of his duty, but, apparently, because of
the facility with which the materials could be procured. It has long
been a maxim of the French chemists, that it is well for beginners in
research to work upon some technical product, such as can readily be
obtained at a manufactory, since by so doing much of the trouble and
cost of procuring materials fit for investigation may be avoided. The
results of several of these early investigations were valuable additions
to scientific knowledge. It was Chevreul who discovered brazil in,
white-indigo, and hematoxylin, — also creatin. But by far the most
noteworthy of his works at this period was the classical research on
the composition of fats, which was published in book form in 1823.
He had been actively engaged in studying this subject during the
years 1811 to 1820. Of this research chemists have always held one
opinion, namely, that it was masterly. Hermann Kopp says of it,
in his Geschichte der Cheraie, " Chevreul's work on the fats is a
model of a complete and exhaustive research in organic chemistry."
It may well be said of it, that it gave to the world a new source of
power, akin to those gained by the investigation of the properties of
steam and of electricity. A great flood of light was thrown upon a
matter of extreme darkness. By mere force of its clearness, the
publication of this work led directly to numberless improvements in
the arts and in household economy. At a time when methods of
investigating organic substances were little known, this research
brought out, as by a touch, complete, accurate, and enduring knowl-
edge of the composition and properties of a large and important class
of chemical substances, the study of which had previously been sup-
posed to be peculiarly difficult. The research itself was thoroughly
and purely scientific. As Kopp has set forth, it disclosed the proxi-
mate and ultimate composition of the fats, both in general and in
particular, qualitatively and quantitatively, as well as the composition
of the products of the decomposition of fats ; it developed a rational
theory as to the chemical structure of the fats, and displayed the
relations which the fats bear to other chemical substances. Mean-
while, the processes of organic analysis which Chevreul used were
decided improvements on those of his predecessors.

It is because of the purity and fulness of this particular fountain,
so to say, of scientific knowledge, that so many useful technical
applications have flowed from it. Even up to the present time it has
exerted an enormous influence upon all that relates to the mauufac-

MICHEL EUGENE CHEVREUL. 455

ture and use of soaps, oils, glycerine, and candles. Such familiar
words, as cholesterin, stearin, margarin, oleic acid, butyric acid, and
the like, have come down to us from Chevreul's work. It is safe to
say that, if the man had died iu his fortieth year, his name and fame
would still have been familiar to us.

Subsequently to his work upon the fats, Chevreul occupied himself
with a great variety of chemical studies, and gave special attention to
matters relating to dyeing and painting. It was at this epoch that he
began his investigations of the laws of the h irmony of colors. His
book upon this subject, published in 1839, has been widely read, with
no little interest, by several generations of men.

In studying the question of the harmony of colors, Chevreul was
impressed with the importance of having a definite standard with
which all colors could be compared, and by means of which all colors
could be accurately named. He argued that in the technical applica-
tion of colors, either for painting or dyeing, it is impossible to lay
dowu definite rules for obtaiuing a desired result unless each of the
colors to be employed can be accurately described. In order to do
this, he devised his so called chromatic circle, which was manifestly
a great improvement both upon the common practice of comparing
colors with natural objects, and upon the ideas of several scientific
men who have urged that definite standards of comparison might be
chosen among colored minerals, or birds, or flowers. This chromatic
circle well illustrates one peculiarity of Chevreul which was not a
little conspicuous ; namely, his leaning towards the methods of natu-
ralists. He sought always to classify ideas and substances into families,
genera, and species. He insisted on carefully distinguishing the chemi-
cal properties of bodies from their physical and organoleptic proper-
ties. To him the characteristic peculiarities of the " immediate
principles" (proximate constituents) in a compound were of deeper
interest than the question of their ultimate composition.

For the fundamental ponts of comparison in his circle of colors he
chose from the solar spectrum definite rays of red, yellow, and blue,
the position of each of which was marked by that of certain stated
lines of Fraunhofer. These three colors were placed at equidistant
points on the circle, and between each pair of them were intercalated
twenty-three mixtures, so that the complete circle comprised seventy-
two colors in all. For example, half-way between yellow and blue
came green, and half-way between yellow and green came "yellow-
green." But between yellow and yellow-green were live different
tints, named yellow with green 1, or 2, or 3, or 4, or .0, respectively ;

456 MICHEL EUGENE CHEVREUL.

and so between blue and red, and between red and yellow. In addi-
tion to this "normal" circle, he had eight others, the colors of, which
were toned down, respectively, with one tenth, two tenths, three
tenths, and so on, of black. In order to compare shades of colors, he
prepared standard rows, or columns, by mixing his normal colors with
one tenth, two tenths, and so on, of black in one direction, and with
one tenth, etc. of white in the other. Chevreul's idea was that stan-
dard circles of color should be kept in the custody of public officers,
just as standards of weights and measures are kept. But in seeking
to carry out this scheme a grave difficulty is encountered in that most
colors speedily fade. Moreover, while it is comparatively easy to
compare colored threads with the colors of a circle made in silk or
woollen, it would not be easy to compare colored woollen stuffs with
the colors of a circle made of porcelain ; that is to say, the varying
lustres of colors, according as they have been put upon one or another
kind of material, hinders very decidedly the accurate comparing of
them. It is in fact difficult to compare the colors of metals or alloys
with those of dyes fixed upon wool, or of pigments printed on paper.
In the course of his lectures, Chevreul was accustomed to construct
witli skeins of worsted upon the floor of the great exhibition hall at
the Gobelins a mammoth circle, which was an object of great beauty.
It carried conviction to his audience that the device would be one of
great practical utility, if it were but possible to obtain permanent
colors with which to construct the circle.

It was in 1832 that Chevreul printed his first paper on the har-
mony of colors, and his book upon the subject was published seven
years later, when he was fifty-three years old. He continued to
write about the contrast of colors as long as he lived ; but not with
his old-time sagacity. It must be said, indeed, that, in spite of his very
long life, the term of Chevreul's useful working years was practically
no longer than that of many another scientific man whose life was
shorter than his by a quarter of a century. Even in Chevreul's seven-
tieth year it was very evident that he was already "a savant somewhat
past his faculties."

During his declining years he had much to say, both by way of
reclamation and of explanation, in favor of the views and labors of ear-
lier investigators, and there can be no question that he held in mind
many matters of historic interest ; but, sad to say, in his enfeebled
condition the task of putting these things in order was beyond his
powers. Worst of all, perhaps, the ways of thinking of the new gen-
erations had become so unlike his ways that hardly any one could find

MTCHEL EUGENE CHEVREUL. 457

time to give heed to what he had to say. Indeed, it is by no means
easy, even if it be possible, to apprebend justly the import of some of
his later views and statements. It may well be questioned whetber
the somewhat painful impression which must have been made upon
his grandchildren and great-grandchildren by many of the utterances
of ChevreuFs last years does not explain the singular request made
by them to the French Academy, that no eulogy should be pronounced
upon their ancestor.

In many notices which have been printed since ChevreuFs death
stress has been laid, justly enough, upon the great influence which be
exerted in the field of industrial chemistry. All this is true. Few
men have bad a wider influence than he for good in this department.
On the approach of his hundredth birthday, it was suggested that there
should be displayed to the public at the Garden of Plants specimens
of the various products of industry which had been made better by
means of his thoughts and studies. But this idea had to be given up,
because not even the large new hall of the Zoological Museum could
heve held all the materials which were offered for exhibition. In
point of fact, on his hundredth birthday there came marching with
banners into the great hall of the Museum, to do him honor, two thou-
sand and more delegates from the learned societies, the schools, the
museums, and the workshops in whose behoof he had labored so faith-
fully during many useful years. It should be borne in mind, none the
less, that Chevreul was not an industrial chemist, in the modern ac-
ceptation of the term. It was not for money prizes that he lived and
labored. But the brilliant technical results which followed the publi-
cation of his researches showed to all men what stores of wealth may
be gained by the judicious application of scientific knowledge for the
correcting of rules of art. In recent days there has been no lack
of workers in this alluring field.

It is a matter of history that the influences of the Revolution, and of
the succeeding long struggle between France and the rest of Europe,
gave more or less of a practical or economic bias to the minds
of most French teachers and scientific men of the period in which
ChevreuFs youth was passed. With his fellows, he felt the influence
of the wars of the Revolution, and of the public feeling which found
expression in the Fcole Normale and the Kcole Poly technique. In
this sense, it is true that he worked by preference upon subjects inti-
mately connected with the ordinary affairs of life; but it should be
remembered of him that he was not oidy an accomplished chemist,
but always, from first to last, a thoroughly scientific man of an admi-
rable type.

458 RUDOLF JULIUS EMANUEL CLAUSIUS.

RUDOLF JULIUS EMANUEL CLAUSIUS.

Rudolf Julius Emanuel Clausius was born at Coslin in
Pomerania, January 2, 1822. His studies, after 1840, were pursued
at Berlin, where he became Privat-docent in the University, and
Instructor in Physics in the School of Artillery. He was Professor
of Physics at Zurich iu the Polytechnicuui (1855-67) and in the
University (1857-67), at Wurzburg (1867-69), and finally at Bonn
(1869-88), where he died on the 24th of August, 1888.

His literary activity commenced in 1847, with the publication of a
memoir in Crelle's Journal, " Ueber die Lichtzerstreuung in der At-
mosphiire, und iiber die Intensitat des durch die Atmosphare reflec-
tirten Sonnenlichts." * This was immediately followed by other
writings relating to the same subject, two of which were subsequently
translated from PoggendorfFs Annalen f for Taylor's Scientific Me-
moirs. A treatise entitled " Die Lichterscheinungen der Atmos-
phare " formed part of Grunert's " Beitrage zur meteorologischeu
Optik."

An entirely different subject, the elasticity of solids, was discussed
in his paper, (1849,) " Ueber die Veranderungen, welche in den bisher
gebriiuchlichen Formeln fur das Gleichgewioht und die Bewegung
fester Korper durch neuere Beobachtuugen nothwendig geworden
sind." t

But it was with questions of quite another order of magnitude that
his name was destined to be associated. The fundamental questions
concerning the relation of heat to mechanical effect, which had been
raised by Rumford, Carnot, and others, to meet with little response,
were now everywhere pressing to the front.

"For more than twelve years," said Regnault in 1853, "I have
been engaged in collecting the materials for the solution of this ques-
tion : — Given a certain quantity of heat, what is, theoretically, the
amount of mechanical effect which can be obtained by applying the
heat to evaporation, or the expansion of elastic fluids, in the various
circumstances which can be realized in practice ? " § The twenty-first
volume of the Memoirs of the Academy of Paris, describing the first
part of the magnificent series of researches which the liberality of the
French government enabled him to carry out for the solution of this

* Vol. xxxiv. p. 122, and vol. xxxvi. p. 185.
t Vol. lxxvi. pp. 161 and 188.
\ Pogg. Ann., vol. lxxvi. p. 4G (1849).
§ Comptes Kendus, vol. xxxvi. p. 676.

RUDOLF JULIUS EMANUEL CLAUSIUS. 459

question, was published in 1847. In the same year appeared Ilelra-
holtz's celebrated memoir, " Ueber die Erhaltung der Kraft." For
some years Joule had been making tbose experiments which were to
associate his name with one of the fundamental laws of thermody-
namics and one of the principal constants of nature. In 18-19 he
made that determination of tbe mechanical equivalent of heat by the
stirring of water which for nearly thirty years remained the unques-
tioned standard. In 1848 and 1849 Sir William Thomson was en-
gaged in developing the consequences of Carnot's theory of the motive
power of heat, while Professor James Thomson in demonstrating the
effect of pressure on the freezing point of water by a Carnot's cycle,
showed the flexibility and the fruitfulness of a mode of demonstration
which was to become canonical in thermodynamics. Meantime Ran-
kine was attacking the problem in his own way, with one of those
marvellous creations of the imagination of which it is so difficult to
estimate the precise value.

Such was the state of the question when Clausius published his
first memoir on thermodynamics : '• Ueber die bewegende Kraft der
Warme, und die Gesetze, welche sich daraus fiir die Warmelehre
selbst ableiten lassen." *

This memoir marks an epoch in the history of physics. If we say,
in the words used by Maxwell some years ago, that thermodynamics
is "a science with secure foundations, clear definitions, and distinct
boundaries," f and ask when those foundations were laid, those defini-
tions fixed, and those boundaries traced, there can be but one answer.
Certainly not before the publication of that memoir. The materials
indeed existed for such a science, as Clausius showed by constructing
it from such materials, substantially, as had for years been the com-
mon property of physicists. But truth and error were in a confusing
state of mixture. Neither in France, nor in Germany, nor in Great
Britain, can we find the answer to the question quoted from Regnault.
The case was worse than this, for wrong answers were confidently
urged by the highest authorities. That question was completely
answered, on its theoretical side, in the memoir of Clausius, and the
science of thermodynamics came into existence. And as Maxwell
said in 1878, so it might have been said at any time since the publica-
tion of that memoir, that the foundations of the science were secure,
its definitions clear, and its boundaries distinct.

* Read in the Berlin Academy, February 18, I860, and published in the
March and April numbers of PoggendurlT'.s Annalen.
t Nature, vol. xvii. p. '1'ii.

4G0 RUDOLF JULIUS EMANUEL CLAUSIUS.

The constructive power thus exhibited, this ability to bring order
out of confusion, this breadth of view which could apprehend one
truth without losing sight of another, this uice discrimination to sep-
arate truth from error, — these are qualities which place the possessor
in the first rank of scientific men.

In the development of the various consequences of the funda-
mental propositions of thermodynamics, as applied to all kinds of
physical phenomena, Clausius was rivalled, perhaps surpassed, in
activity and versatility by Sir William Thomson. His attention,
indeed, seems to have been less directed toward the development of
the subject in extension, than toward the nature of the molecular
phenomena of which the laws of thermodynamics are the sensible
expression. He seems to have very early felt the conviction, that
behind the second law of thermodynamics, which relates to the heat
absorbed or given out by a body, and therefore capable of direct
measurement, there was another law of similar form, but relating to
the quantities of heat (i. e. molecular vis viva) absorbed in the per-
formance of work, external or internal.

This may be made more definite, if we express the second law in
a mathematical form, as may be done by saying that in any reversible
cyclic process which a body may undergo

f

= 0,

where dQ is an elementary portion of the heat imparted to the body,
and t the absolute temperature of the body, or the portion of it which
receives the heat. Or, without limitation to cyclic processes, we
may say that for any reversible infinitesimal change,

dQ = t dSy

where S denotes a certain function of the state of the body, called by
Clausius the entropy. The element of heat may evidently be divided
into two parts, of which one represents the increase of molecular
vis viva in the body, and the other the work done against forces,
either external or internal. If we call these parts dH and dQw, we
have

dQ = dff+dQw.

Now the proposition of which Clausius felt so strong a conviction
was that for reversible cyclic processes

f-

jf = °>

RUDOLF JULIUS EMANUEL CLAUSIUS. 4G1

and that for any reversible infinitesimal change

d Qw = t d Z,

where Z is another function of the state of the body, which he
called the disgregation, and regarded as determined by the positions m
of the elementary parts of the body without reference to their veloci-
ties. In this respect it differed from the entropy. An immediate
consequence of these relations is that for any reversible cyclic process

-*dH

r-

= 0,

and therefore that 77, the molecular vis viva of the body, must be a
function of the temperature alone. This important result was ex-
pressed by Clausius in the following words : " Die Menge der in
einem Korper wirklich vorhandenen Warme ist nur von seiner
Temperatur und nicht von der Anordnung seiner Bestandtheile
abhiingig."

To return to the equation

dQw = tdZ. ■

This expresses that heat tends to increase the disgregation, and that
the intensity of this tendency is proportional to the absolute tempera-
ture. In the words of Clausius : " Die mechanische Arbeit, welche
die Warme bei irgend einer Anordnungsanderung eines Kbrpers thun
kann, ist proportional der absoluten Temperatur, bei welcher die Aen-
derung geschieht."

Such in brief and in part were the views advanced by Clausius in
18G2, in his memoir, '• Ueber die Anwendung des Satzes von der
Aequivalenz der Verwandlungen auf die innere Arbeit."* Although
they were advanced rather as a hypothesis than as anything for
which he could give a formal proof, he seems to have little doubt of
their correctness, and his confidence seems to have increased with the
course of time.

The substantial correctness of these views cannot now be called in
question. The researches especially of Maxwell and Boltzmann have
shown that the molecular vis viva is proportional to the absolute
temperature, and Boltzmann has even been able to determine tin-
precise nature of the functions which Clausius called entropy and
disgregation .f But the anticipation, to a certain extent, at so early a

* Pogg. Ann., vol. cxvi. p. 73. See also vol. exxvii. p. 477 (1800).
t Sitzungsberichte Wien. Akad., vol. lxiii. p. 728 (1871).

462 RUDOLF JULIUS EMANUEL CLAUSIUS.

period in the history of the subject, of the ultimate form which the
theory was to take, shows a remarkable insight, which is by no
means to be lightly esteemed on account of the acknowledged want of
a rigorous demonstration. The propositions, indeed, as relating to
'quantities which escape direct measurement, belong to molecular
science, and seem to require for their complete and satisfactory
demonstration a considerable development of that science. This de-
velopment naturally commenced with the simplest case involving the
characteristic problems of the subject, — the case, namely, of gases.

The origin of the kinetic theory of gases is lost in remote antiquity,
and its completion the most sanguine cannot hope to see. But a
single generation has seen it advance from the stage of vague sur-
mises to an extensive and well established body of doctrine. This is
mainly the work of three men, Clausius, Maxwell, and Boltzmann, of
which Clausius was the earliest in the field, and has been called by
Maxwell the principal founder of the science.* We may regard his
paper, (1857,) " Ueber die Art der Bewegung, welche wir Wiirrae
nennen," | as marking his definite entrance into this field, although
many points were incidentally discussed in earlier papers.

This was soon followed by his papers, " Ueber die mittlere Liinge
der Wege, welche bei der Molecularbewegung gasformiger Korper
von den einzelnen Moleciilen zuriic.kgelegt werden, " $ and " Ueber die
Warrneleitung gasformiger Korper." §

A very valuable contribution to molecular science is the conception
of the virial, defined in his paper, (1870.) " Ueber einen auf die
Warme anwendbaren Satz," | where he shows that in any case of
stationary motion the mean vis viva of the system is equal to its virial.

In the mean time, Maxwell and Boltzmann had entered the field.
Maxwell's first paper, " On the Motions and Collisions of perfectly
elastic Spheres, " ^f was characterized by a new manner of proposing
the problems of molecular science. Clausius was concerned with the
mean values of various quantities which vary enormously in the
smallest time or space which we can appreciate. Maxwell occupied
himself with the relative frequency of the various values which these
quantities have. In this he was followed by Boltzmann. In reading

* Nature, vol. xvii. p. 278.
t Ibid., vol. c. p. 353 (1857).

J Ibid., vol. cv. p. 239 (1858). See also Wied. Ann., vol. x. p. 92.
§ Ibid., vol. cxv. p. 1 (1862).

|| Ibid., vol. clxi. p. 124. See also Jubelband, p. 411.
1 Phil. Mag., vol. xix. p. 19 (1860).

RUDOLF JULIUS EMANUEL CLAUSIUS. 463

Clausius, we seem to be reading mechanics; in reading Maxwell, and
in much of Boltzmann's most valuable work, we seem rather to be
reading in the theory of probabilities. There is no doubt that the
larger manner in which Maxwell and Boltzmann proposed the prob-
lems of molecular science enabled them in some cases to get a more
satisfactory and complete answer, even for those questions which do
not at first sight seem to require so broad a treatment.

Boltzmann's first work, however, (1866,) " Ueber die mechanische
Bedeutung des zweiten llauptsatzes der Wiirmetheorie," * was in a
line in which no one had preceded him, although he was followed by
some of the most distinguished names among his contemporaries.
Somewhat later (1870) Clausius, whose attention had not been called
to Boltzmann's work, wrote his paper, " Ueber die Zuriickfuhrung des
zweiten Hauptsatzes der mechanischen "Wiirmetheorie auf allgemeine
mechanische Principien." f

The point of departure of these investigations, and others to which
they gave rise, is the consideration of the mean values of the force-
function and of the vis viva of a system in which the motions are
periodic, and of the variations of these mean values when the exter-
nal influences are changed. The theorems developed belong to the
same general category as the principle of least action, and the prin-
ciple or principles known as Hamilton's, which have to do, explicitly
or implicitly, with the variations of these mean values.

Among other papers of Clausius on this subject, we may mention
the two following: "Ueber einen neuen mechanischen Satz in Bezug
auf stationare Bewegung," t (1873,) and "Ueber den Satz vom mit-
tlcren Ergal und seine Anwendung auf die Molecularbewegungen der
Gase"§ (1874).

The first problem of molecular science is to derive from the ob-
served properties of bodies as accurate a notion as 2>ossible of their
molecular constitution. The knowledge we may gain of their molecu-
lar constitution may then be utilized in the search for formulas to
represent their observable properties. A most notable achievement
in this direction is that of van der Waals, in his celebrated memoir
"On the Continuity of the Gaseous and Liquid States." To this part
of the subject belong the following papers of Clausius: "Ueber .las
Verhalten der Kohlensaure in Bezug auf Druck, Volumen und Tetn-

* Ritzungsberichte Wien. Akad., vol. liii. p. 195.

t Fogg. Ann., vol. cxlii. p. 433.

j Ibid., vol. cl. p. 106.

§ Ibid., Erganzung«band vii. p. 215.

464 RUDOLF JULIUS EMANUEL CLAUSIUS.

peratur," * and " Ueber die theoretische Bestimmung des Dampf-
druckes und der Volumina des Dampfes und der Fliissigkeit " (two
papers), f

Another matter in which Clausius showed his originality and power
was the vexed subject of electrodynamics, as treated in his memoir,
" Ueber die Ableitung eines neuen electrodynamischen Grundge-
setzes/'t Various points in the theory of electricity in which the
principles of thermodynamics or of molecular science were involved,
had previously been treated in different papers, of which the earliest
appeared in 1852, § while the doctrine of the potential (electrical and
gravitational) was treated in a separate book, which appeared in 1859,
with the title, " Die Potentialfunction und das Potential, ein Beitrag
zur mathematischen Physik." This subsequently went through several
editions, in which it was revised and enlarged. All these subjects, with
others, were brought together in a single volume, " Die mechauische
Behandlung der Electricitat," which appeared in 1879, forming the
second volume of his " Mechanische Warmetheorie." Later papers
on electricity related to the principles of electrodynamics,! electrical
and magnetic units,** and dynamo-electric machines.ft

The Royal Society's catalogue of scientific papers, and the excellent
indices to the Annalen der Physik und Chemie, in which Clausius's
work usually appeared, render it unnecessary to enumerate in detail
his scientific papers. The list, indeed, would be a long one. The

* Wied. Ann., vol. ix. p. 337 (1880).

t Ibid., vol. xiv. p. 279 and p. 692 (1881).

} Crelle's Journal, vol. lxxxii. p. 85 (1877).

§ " Ueber das mechanische Aequivalent einer electrischen Entladung und die
dabei stattfindende Erwarmung des Leitungsdrahtes." Pogg. Ann., vol. lxxxvi.
p. 337. " Ueber die bei einem stationaren electrischen Strome in dem Leiter
getbane Arbeit und erzeugte Warme." Pogg. Ann., vol. lxxxvii. p. 415 (1852).
"Ueber die Anwendung der mechanischen Warmetheorie auf die thermoelec-
trischen Erscheinungen." Pogg. Ann., vol. xc. p. 513 (1853). "Ueber die
Electrieitatsleitung in Electrolyten." Pogg. Ann., vol. ci. p. 338 (1857).

|| The first volume of this work appeared in 1876, and contained the general
theory with the more immediate consequences of the two fundamental laws.
The third volume has not yet appeared, but it is expected very soon, edited by
Professor Planck and Dr. Pulfrich. In a certain sense this work may be re-
garded as a second edition of an earlier one (18G4 and 1867), which consisted
of a reprint of papers and had the title " Abhandlungen iiber die mechanische
Warmetheorie."

IT Wied. Ann., vol. x. p. 608; vol. xi. p. G04.

** Ibid., vol. xvi. p. 529 ; vol. xvii. p. 713.

tt Ibid., vol. xx. p. 353 ; vol. xxi. p. 385.

FRANCISCUS CORNELIUS DONDERS. 465

Royal Society's catalogue gives seventy-seven titles for the years
1847-1873. Subsequently tweuty-five papers have appeared iu the
Annalen alone, and about half as many others elsewhere.

But such work as that of Clausius is not measured by counting
titles or pages. His true monument lies not on the shelves of libra-
ries, but in the thoughts of men, and in the history of more than one
science.

FRANCISCUS CORNELIUS DONDERS.

In assuming the honorable duty of preparing *a biographical notice
of Professor Donders, the writer must acknowledge indebtedness, for
many details, to accounts of his life and work by his coadjutors and
friends, Moleschott, Snellen, Nuel, and Landolt, and to his own speech
at the festival in his honor on his seventieth birthday.

Franciscus Cornelius Donders was born at Tilbury, in Holland, on
the 17th of May, 1818 ; the tenth child and only son of parents in
very moderate circumstances. His father's death occurred about a
year later. From his seventh to his thirteenth year he was at
school in Duizel, where he learned to write his own language, some-
thing of French, mathematics, and music. From eleven to thirteen
years he served as submaster at the school, thus defraying his ex-
penses. His studies were continued at Liege, in pursuance of his
mother's wish that he should become a priest; but this plan being
defeated by the occurrence of the Belgian revolution, he returned to
the French school at Tilbury. Becoming fond of study, he was sent
to Boxmeer, where he learned to write and speak Latin with great
fluency, and a little Greek, but where mathematics, for which he
seems to have had special aptitude, was neglected. At the age of
seventeen he entered the School of Military Medicine at Utrecht,
where he became enthusiastic in the experimental study of chem-
istry, the natural sciences, and physiology. After four years of
study, desiring to obtain his diploma, he applied for examination ; but
a technical objection being made by the Faculty, he was advised to
present himself at the University of Leyden. Here, through the
merits of his dissertation and an address in Latin in which he con-
vinced the Faculty that he was worthy to receive the degree of
Doctor, he made such an impression of his capacity that in three days
he returned to Utrecht as M. D. Of his " Dissertatio Inaugurate
sistens Observationes Anatomico-pathological" Professor Moleschotl
says, "In this trial test I could well discern the future master."
vol. xxiv. (n. s. XVI.) 30

466 FRANCLSCUS CORNELIUS DONDERS.

Before he was twenty-two years of age, Donders was sent as
junior military surgeon to the garrison at Vlissingen. Here, and
especially at The Hague, the seat of government, where he was sta-
tioned a year later, he had the advantage of polite society, and of
contact with distinguished men, — with opportunities for culture in
art and literature, as well as in science and general and professional
knowledge. He amply profited by these advantages, and perfected
his acquaintance with French, English, and German, so that he
wrote and spoke these as if they had been his native tongue ; acquir-
ing also a grace and urbanity of manner for which he was distin-
guished in all his social and professional life. The attention of his
superiors being attracted by some of his published scientific papers, he
was sent again to Utrecht, in 1842, to reorganize the Military Medical
School ; and was appointed Professor of Anatomy and of Physiology
at the institution where so recently he had been a pupil.

Here began Donders's real scientific life. Convinced that book
knowledge, especially in the natural sciences, has little value in con-
tributing to further advancement, unless completed by careful personal
experimental investigation, he accepted the offered position without a
moment's hesitation, although he thus gave up the pleasures and
advantages of the capital, and accepted a smaller income : " For I
felt," he says, " that to teach was my vocation."

Thus he established himself for life, as it proved, in a small city
of Holland. Quickened into still greater activity by the labors he
assumed, and animated by the example of Professor Mulder, — whose
laboratory became, as Donders expressed it, the cradle of physiologi-
cal chemistry, — and in co-operation with his venerated teacher, Pro-
fessor Schroeder, he devoted himself with ardor to explorations in
every part of anatomy and physiology, verifying each observation with
his own eyes, and accepting nothing as proved which his own experi-
ence had not confirmed; but showing marvellous lucidity in directing
researches, in forming conclusions, and in appreciating the values of
results gained and the means of utilizing them. " There is no domain
in the vast science of Physiology," says Landolt, " in which Donders
did not leave traces of his labors. The vitality of tissues, the circula-
tion of the blood, digestion, secretions, movements, language, the
organs of sense, the secrets of the nervous system, have all in turn
been investigated by this indefatigable explorer." He founded, with
Ellermann and Jansen, the " Nederlandsch Lancet," that they might
have an organ for the announcement of new discoveries ; and he
largely augmented his own labors by frequent contributions to it*

FRANCISCUS CORNELIUS DONDERS. 467

pages. Rousiug emulation in others, he disarmed jealousy hy his
candid and cordial recognition of their meritorious work. Hoarding
none of his own acquisitions, he added to the joys of discovery the
pleasures of disclosure. As a teacher, he was radiant; he seemed
superb in the lucidity, conciseness, elegance, and adaptiveness of
his style of explanations, — which he often made, in several ver-
naculars, where he saw that he was not understood by an intelligent
disciple. Donders himself says : " To teach is as great a joy as to
learn. Acquired knowledge is as a hidden treasure, which slumbers
useless until it is disclosed in teaching." His instruction was most
suggestive, for himself as well as for his hearers, — opening as it
were new horizons of thought. Kuel says of his manner of teaching,
" Few have equalled, none surpassed him." And Moleschott, in his
address of greeting on his seventieth birthday, says of him, " Donders
has remained at Utrecht, but all the world has come to him."

We find him, at the age of twenty-four, giving, as Professor, eigh-
teen lectures weekly for forty-six weeks of the year, on anatomy,
histology, and physiology, and yet finding time for a vast amount of
original investigations. In his modest address on the occasion of his
jubilee, last year, he thus expresses his appreciation of the favorable
circumstances by which he had been surrounded : " No other period
has been comparable with this for great discoveries in so many fields
of biological science. Von Baer had discovered the ovule of the
mammif'ers ; Bischoff had demonstrated that the embryo is built up
exclusively by means of segmentation of cells; Schwann found in
the cell the origin of all the fundamental forms of life ; Ilenle had
created, in his k'Anatomie Generale," an organon of histology. At the
same epoch appeared the Physiology of Jean Muller. It was in sucli
a world I had the good fortune to have place; every circumstance
seemed to be adapted to render my life and my work prolific."

In 1847 Donders hecarue Professor Extraordinarius at the Univer-
sity of Utrecht, giving courses on Legal Medicine, Hygiene, Anthro-
pology, and Ophthalmology. " After a time,'" he says, "m\ teaching
of Ophthalmology gave a new direction to my life." Two distin-
guished men from Edinburgh were one day among his auditors, and
urged him to visit the great eye hospitals of England al the time of
the International Exhibition in 1851. At London he saw Bowman,
equally renowned as physiologist and ophthalmologist, and also \ on
Jaeger of Vienna and Yon Grade of Berlin. These friends became
to him " t.he most precious treasure of my whole career." He could
announce to them the recent discovery, by his compatriot Cramer

468 FRANCISCUS CORNELIUS DONDERS.

of Haarlem, of the modus operandi by which was effected the accom-
modation in the human eye, hitherto unexplained ; and he learned from
them of the invention of the Ophthalmoscope by Helmholtz; by ineaus
of which all the secrets of the before unexplored interior of the eye
could be revealed to the wondering observer. Before returning to
Holland he went to Paris in company with Von Graefe ; there visiting
the large Cliniques for diseases of the eye, and comparing their meth-
ods of diagnosis and practice with those seen at London. Thenceforth,
without abandoning general physiology, Donders worked especially
in physiological optics as applied to eye affections, in so doing largely
enhancing his already great renown. Impatient to await the arrival
of an ophthalmoscope which Helmholtz was to send to him, he con-
structed one for himself; not of superimposed glass plates, as contrived
by Helmholtz, but of a silvered mirror with a central perforation, as
now generally used ; — and was enthusiastic in his instant perception
of the value of its disclosures.

In 1864 appeared Donders's monumental work on the Refraction
and Accommodation of the Eye, published by the Sydenham Society
at London, and soon translated into many other languages. It came to
the world of Ophthalmology as a revelation, — a complete and finished
creation, involving infinite labor and research, — from which nothing
could be retrenched without loss, and to which nothing could be added
without superfluity. It created scientific Ophthalmology. His dis-
covery of Hypermetropia, his explanation of Astigmatism, his indi-
cation of the relations between different forms of Strabismus and the
hypermetropic or myopic conditions of refraction of the eye, were
and must remain masterpieces of absolute demonstration.

It would be almost impossible to give even a catalogue of Donders's
published works. His friend Nuel, in the " Annales d'Oculistique,"
cites two hundred and eight of these, and adds, " This list is by no
means complete." Already, in 1846. he had translated into Dutch a
German work on Ophthalmology by Ruete, which seems unquestiona-
bly to have inclined his spirit of investigation towards the eye. He
says, " Thanks to the progress of the histology and physiology of the
eye, its diseases are those which best admit of a physiological explana-
tion." To elucidate some questions which had suggested themselves
in this translation, Donders published within about a year, in the
Nederlandsch Lancet, three notable papers on Physiological Optics, —
among these a monograph on " The Relations between the Conver-
gence and the State of Accommodation of the Eyes."

After his return from England, in 1851, Donders obtained endow-

FRANCISCUS CORNELIUS DONDERS. 469

ments for the Netherlands Hospital for Diseases of the Eyes, which,

with the co-operation of Snellen, he established and conducted at
Utrecht; — thus entering a field where the large opportunities for
direct and skilled personal service towards his suffering fellow men
afforded a new satisfaction to his sympathetic nature. The mean-
thus offered for continuing scientific observation wee also most
advantageous, and afforded ample fruition in the hands of such a
man, — who loved accurate research, not only for its own sake, but
for the delight of imparting and applying the knowledge thus gained
for the relief of humanity.

So great a man cannot be wholly unconscious of his worth; hut
Donders's modesty was sometimes almost diifidence. We rejoice that
in his later years he was allowed the gratification of knowing that he
had been eminently useful, not only by his own labors, which had
advanced Ophthalmology a century, but hugely also by rendering
applicable the results of the labors of others. The invention of Von
Helmholtz, which indeed opened a new world to Ophthalmology,
might have rested in the laboratory of the renowned physicist who
devised it, but perhaps little suspected its practical value, as a mere
bit of apparatus for experiment, because he lacked opportunities for
its clinical use. Moreover, as an apostle of Ophthalmology, Don-
ders became the exemplar and teacher of disciples from near and far
countries.

When, in 1862, upon the decease of Schroeder, the Professorship
of Physiology at the University was tendered to Bonders, he once
more obtained subscriptions of large sums for the building and equip-
ment of a new physiological laboratory, so supplied with all modern
means of research that it should be a model institution. But his
enormous labors in so many departments made it imperative that he
should divide his responsibilities with worthy assistants, — Snellen
assuming the management of the Ophthalmic Hospital, and Engle
mann the charge of the laboratory and of the courses in Microscopic
Anatomy and Physiological Chemistry, — Donders reserving to him-
self the instruction in General Physiology and Physiological Optics,
and the time for original investigations. At tin- F< itsitzung der
Ophthalmologischen Gesellschaft, in 1880, in tin- greal ball of the
University of Heidelberg, for the presentation to Von Helmholtz of
the Orach- Medal, which, according to the statutes of its founder,
should he given once in ten years to him of any nationality who had
done most for the promotion of Ophthalmology, Donders, a- the mosl
distinguished member of the society, was selected as the orator. After

470 FRANCISCUS CORNELIUS DONDERS.

a feeling eulogy of the founder, Von Graefe, he spoke of Von Helm-
holtz, who, as the inventor of the Ophthalmoscope, had marked the
dawn of a new era in our knowledge of diseases of the eye ; and of
the gratitude due to him for having thus endowed the profession and
humanity, taking away the reproach of our former utter ignorance of
the pathology of the deep-seated parts of the eye, and well deserving
the hi^h award then for the first time made.

In May, 1888, Donders attained his seventieth birthday; at which
time, according to the laws of the University of Utrecht, its Professors
retire from office. This event was made the occasion of a festal
homage, rendered by the King of Holland, the most distinguished of
his fellow countrymen, his fellow citizens, his coadjutors, and by
representatives of his disciples and of Ophthalmology from every part
of the world. It was an apotheosis. There he gave a most simple
and unostentatious narration of his life and work, with grateful ac-
knowledgment of the friendships by which he had been encouraged,
and of the circumstances by which he had been favored.

Once again, in August, 1888, he assumed the place of honor at the
Heidelberg annual convocation of Ophthalmologists. Now, alas ! we
shall see him no more. Since March the 24th, Ophthalmology mourns
its chief, — the world of science one of its brightest ornaments.

Since the last Report, the Academy has received an acces-
sion of twenty-four members, — fifteen Resident Fellows,
two Associate Fellows, and seven Foreign Honorary Mem-
bers. One Resident Fellow has resigned. The list of the
Academy, corrected to November 30, 1889, includes 190
Resident Fellows, 89 Associate Fellows, and 71 Foreign
Honorary Members.

To this volume are added the Statutes and Standing Votes
of the Academy, which have been out of print for several
years.

LIST

OF THE FELLOWS AND FOREIGN HONORARY MEMBERS.

(Corrected to November 30, 1889.)

RESIDENT FELLOWS. — 190.

Class I.

(Number limited to two hundred.)
Mathematical and Physical Sciences. — 77.

Section I. — 7.
Mathematics.

Gustavus Hay,
Benjamin O. Peirce,
James M. Peirce,
John D. Runkle,
T. H. Safford,
Edwin P. Seaver,
William E. Story,

Boston.
Cambridge.

Cambridge.
Brookline.
Williamstown.
Newton.
Worcester.

Section II. — 11.
Practical Astronomy and Geodesy.

Seth C Chandler, Cambridge.

A Ivan G. Clark, Cambridgeport.
George B. Clark, Cambridgeport.
J. Rayner Edmands, Cambridge.
Henry Mitchell, Boston.

Edward C. Pickering, Cambridge.
John Ritchie. Jr., Boston.
William A. Rogers, Waterville, Me.
Edwin F. Sawyer, Brighton.
Arthur Searle, Cambridge.

O. C. Wendell, Cambridge.

Section III. — 43.

Physics and

A. Graham Bell,
Clarence J. Blake,
Francis Blake,
John II. Blake,
Josiah P. Cooke,
James M. Crafts,
Charles R. Cross,

Chemistry.

Washington.

Boston.

Weston.

Boston.

Cambridge.

Boston.

Boston.

William P. Dexter,
Amos E. Dolbear,
Thos. M. Drown,
Charles W. Eliot,
Moses G. Farmer,
Thomas Gaffield,
Wolcott Gibbs,
Frank A. Gooch,
Edwin II. Hall,
Henry B. Hill,
N. D. C. Hodges,
Silas W. Holman,
William L. Hooper,
Eben N. Horsford,
T. Sterry Hunt,
Charles L. Jackson,
William W. Jacques,
Alonzo S. Kimball,
Leonard P. Kinnicutt,
Joseph Lovering,
Charles F. Mabery,
Arthur Michael,
A. A. Michelson,
Lewis M. Norton,
William II. Pickering,
Robert II. Richards,
Edward S. Ritchie,
A. Lawrence Rotch,
Stephen P. Sharpies,
Francis II. Storer,
Klihu Thomson,
John Trowbridge,
Cyrus M. Warren,
Harold Whiting,
Charles II. Wing,

Edward S. Wood,

Roxbury.

Somerville.

Boston.

Cambridge.

Eliot, Me.

Boston.

Newport, R. I.

New Haven.

Cambridge.

Boston.

New York.

Boston.

Somerville.

Cambridge.

New York.

Cambridge.

Newton.

Worcester.

Worcester.

Cambridge.

Cleveland.

Worcester.

Worcester.

Newton.

Cambridge.

Boston.

Brookline.

Boston.

Cambridge.

Boston.

Lynn.

( ambridge.

Brookline.

Cambridge.

Boston.
Cambridge.

472

RESIDENT FELLOWS.

Section IV

■16.

Technology and Engineering.

John M. Batchelder, Cambridge

Chas. O.Boutelle,

Winfield S. Chaplin,

Eliot C. Clarke,

James B. Francis,

Gaetano Lanza,

Washington.

Cambridge.

Boston.

Lowell.

Boston.

E. D. Leavitt, Jr., Cambridgeport.

William R. Lee,
Hiram F. Mils,
Cecil H. Peabody,
Alfred P. Rockwell,
Peter Schwamb,
Charles S. Storrow,
George F. Swain,
William Watson,
Morrill Wyman,

Roxbury.

Lawrence.

Boston.

Boston.

Boston.

Boston.

Boston.

Boston.

Cambridge.

Class II. — Natural and Physiological Sciences. — 58.

Sectiox I. — 8.

Geology, Mineralogy, and Physics of
the Globe.

Thomas T. Bouve,
Algernon Coolidge,
William O. Crosby,
William M. Davis,
O. W. Huntington,
Jules Marcou,
William H. Niles,
Nathaniel S. Shaler,

Boston.

Boston.

Boston.

Cambridge.

Cambridge.

Cambridge.

Cambridge.

Cambridge.

Section II. — 7.
Botany.

William G. Farlow,
George L. Goodale,
H. H. Hunnewell,
Charles S. Sargent,
Charles J. Sprague,
Sereno Watson,
Henry Willey,

Cambridge.

Cambridge.

Wellesley.

Brookline.

Boston.

Cambridge.

New Bedford.

Section III. — 20.

Zoology and Physiology.

Alex. E. R. Agassiz, Cambridge.

Robert Amory, Boston.

James M. Barnard, Milton.

Henry P. Bowditch, Boston.

Edward Burgess, Boston.

Harold C. Ernst, Boston.

J. Walter Fewkes,
Hermann A. Hagen,
Samuel Henshaw,
Alpheus Hyatt,
Theodore Lyman,
Edward L. Mark,
Charles S. Minot,
Edward S. Morse,
James J. Putnam,
Samuel H. Scudder,
William T. Sedgwick,
D. Humphreys Storer,
Henry Wheatland,
James C. White,

Boston.

Cambridge.

Cambridge.

Cambridge.

Brookline.

Cambridge.

Boston.

Salem.

Boston.

Cambridge.

Boston.

Boston.

Salem.

Boston.

Section IV. — 23.

Medicine and Surgery.

Samuel L. Abbot, Boston.

Henry J. Bigelow, Boston.

Henry I. Bowditcb, Boston.

Edward H. Bradford, Boston.

Arthur T. Cabot, Boston.

David W. Cheever, Boston.

Benjamin E. Cotting, Roxbury.

Frank W. Draper, Boston.

Thomas Dwight, Boston.

Reginald H. Fitz, Boston.

Charles F. Folsom, Boston.

Richard M. Hodges, Boston.

Oliver W. Holmes, Boston.

John Homans, Boston.

RESIDENT FELLOWS.

473

Alfred Hosmer,
Frederick I. Knight,
George H. Lyman,
Francis Minot,
Wrn. L. Richardson,

Watertown.

Boston.

Boston.

Boston.

Boston.

George C. Shattuck,
Henry I'. Walcott,
John C. Warren.
Henry W. Williams,

Boston.
Cambridge.
Boston.
Boston.

Class III. — Moral and Political Sciences. — 55.

Section I. — '9.

Philosophy and Jurisprudence.

James B. Ames,
Phillips Brooks,
Charles C. Everett,
Horace Gray,
John C. Gray,
Nathaniel Holmes,
John Lowell,
Henry W. Paine,
James B. Thayer,

Cambridge.

Boston.

Cambridge.

Boston.

Boston.

Cambridge.

Newton.

Cambridge.

Cambridge.

Section II. — 19.

Philology and Archaeology.

William S. Appleton, Boston.
William P. Atkinson, Boston.

Boston.
Williamstown.
Boston.
Boston.
Boston.
Cambridge.
Quincy.

Lucien Carr,
Franklin Carter,
Joseph T. Clarke,
William C. Collar,
Henry G. Denny,
Epes S. Dixwell,
William Everett,
William W. Goodwin, Cambridge.
II Miry W. Haynes, Boston.
Bennett H. Nash, Boston.
Frederick W. Putnam, Cambridge.
F. B. Stephenson, Boston.
Joseph II. Thayer, Cambridge.
Crawford II. Toy, Cambridge.
John W. White, Cambridge.

Justin Winsor, Cambridge.

Edward J. Young, Waltham.

Section III. — 17.
Political Economy and History.

Chas. F. Adams,
Edward Atkinson,
John Cummings,
Charles F. Dunbar,
Samuel Eliot,
George E. Ellis,
Henry C. Lodge,
Augustus Lowell,
Edward J. Lowell,
Francis Parkman ,
Andrew P. Peabody,
John C. Ropes,
Denman W. Ross,
F. W. Taussig,
Henry W. Torrey,
Francis A. Walker,
Robert C. Winthrop,

Quincy.
Boston.

Woliurn.

Cambridge.

Boston.

Boston.

Boston.

Boston.

Boston.

Boston.

Cambridge.

Boston.

Cambridge.

Cambridge.

Cambridge.

Boston.

Boston.

Section- IV. — 10.
Literature an<! th Firu Arts.

George S. Boutwell,
Martin Brimmer,
J. Elliot Cabot,
Francis J. Child.
Charles ('< Loring,
James Russell Lowell,
Charles Eliot Norton,
Horace E. Scad I r.
Barrett Wendell,
John (',. Whittier,

Groton.

Boston.

Br " dv line.

Cambridge.

Boston.

Cambri

Cambridge.

Cambridge.

on.
Anieshury.

474

ASSOCIATE FELLOWS.

ASSOCIATE FELLOWS.

89.

(Number limited to one hundred.)

Class I. — Mathematical and Physical Sciences. — 34.

Section I. — 6.

Mathematics.

William Ferrel, Kansas City, Mo.
Thomas Hill, Portland, Me.
Simon Newcomb, Washington.
H. A. Newton, New Haven.
James E. Oliver, Ithaca, N.Y.
J. N. Stockwell, Cleveland.

Section II. — 12.
Practical Astronomy and Geodesy.

W.H.C.Bartlett,
J. H. C. Coffin,
Geo. Davidson,
Wm. H. Emory,
Asaph Hall,
J. E. Hilgard,
George W. Hill,
E. S. Holden,
Sam. P. Langley,
C. H. F. Peters,
George M. Searle,
Chas. A. Young,

Yonkers, N.Y.
Washington.
San Francisco.
Washington.
Washington.
Washington.
Washington.
San Jose, Cal.
Washington.
Clinton, N.Y.
New York.
Princeton, N.J.

Section in. — 10.
Physics and Chemistry.
J. Willard Gibbs, New Haven.

S. W. Johnson,
M. C. Lea,
John Le Conte,
J. W. Mallet,
A. M. Mayer,
Ira Remsen,
Ogden N. Rood,
H. A. Rowland,
L.M. Rutherfurd

New Haven.
Philadelphia.
Berkeley, Cal.
Charlottesville, Ya.
Hoboken, N. J.
Baltimore.
New York.
Baltimore.
, New York.

Section IV. — 6.

Technology and Engineering.

Henry L. Abbot,
Geo. W. Cullum,
Geo. S. Morison,
John Newton,
William Sellers,
W. P. Trowbridge,

New York.
New York.
New York.
New York.
Philadelphia.
New Haven.

Class II. — Natural and Physiological Sciences. — 30.

Section I. — 15.
Geology, Mineralogy, and Physics of

the Globe.
Cleveland Abbe, Washington.

George J. Brush,
James D. Dana,
Sir J.W. Dawson,
-I. C. Fremont,
F. A. Genth,

New Haven.
New Haven.
Montreal.
New York.
Philadelphia.

James Hall,
F. S. Holmes,
Clarence King,
Joseph Le Conte,
J. Peter Lesley,
J. S. Newberry,
J. W. Powell,
R. Pumpelly,
Geo. C. Swallow,

Albany, N.Y.
Charleston, S. C-
New York.
Berkeley, Cal.
Philadelphia.
New York.
Washington.
Newport, R.I.
Columbia, Mo.

ASSOCIATE FELLOWS.

it:>

Section II. — 2.

Botany.

A. W. Chapman, Apalachicola, Fla.
D. C Eatou, New Haven.

Section III. — 7.

Zoology and Physiology.

Joel A. Allen, New York.

G. B. Goode, Washington.

Joseph Leidy, Philadelphia.

0. C. Marsh, New Haven.

S. Weir Mitchell, Philadelphia.
A. S. Packard, Providence.
A. E. Verrill, New Haven.

Section IV. — 6.

Medicine and Surgery.

Fordyce Barker, New York.
John S. Billings, Washington.
Jacob M. Ua Costa, Philadelphia.
W.A.Hammond, New York.
Alfred Stille, Philadelphia.

H. C. Wood, Philadelphia.

Class III. — Moral and Political Sciences. — 25.

Section I. — 7.

Philosophy and Jurisprudence.

D. R. Goodwin,
A. G. Haygood,
James McCosh,
Charles S. Peirce,
Noah Porter,

E. G. Robinson,
Jeremiah Smith,

Philadelphia.
Oxford, Ga.
Princeton, N.J.
New York.
New Haven.
Providence.
Dover, N.H.

Section II. — 6.
Philology and Archaeology.

A. N. Arnold, Pawtuxet, R.I.

D. C. Gilman, Baltimore.

A. C. Kendrick, Rochester, N.Y.

E. E. Salisbury, New Haven.

A. D. White,
W. D. Whitney,

Ithaca, N.Y.
New Haven.

Section IH. — 6.
Political Economy and History.

Henry Adams, Washington.

George Bancroft, Washington.

M. F. Force, Cincinnati.

Henry C. Lea, Philadelphia.

W. G. Sumner, New Haven.

J. II. Trumbull, Hartford.

Section IV. — 6.
Literature and the Fine Arts.
James B. Angell, Ann Arbor, Mich
L. P. di Cesnola, New York.
F. E. Church, New York.

R. S. Greenough, Florence.
William W. Story, Rom.'
Win. It. Ware, New York.

476

FOREIGN HONORARY MEMBERS.

FOREIGN HONORARY MEMBERS. — 71.

(Elected as vacancies occur.)

Class I. — Mathematical and Physical Sciences. — 24.

Section I. — 6.
Mathematics.

John C. Adams,
Sir George B. Airy,
Francesco Brioschi,
Arthur Cayley,
Charles Herniite,
J. J. Sylvester,

Cambridge.

London.

Milan.

Cambridge.

Paris.

Oxford.

Section II. — 5.

Practical Astronomy and Geodesy.

Arthur Auwers,
J. H. W. Dollen,
H. A. E. A, Faye,
Eduard Schoni'eld,
Otto Struve,

Berlin.

Pulkowa.

Paris.

Bonn.

Pulkowa.

Section III. — 10.

Physics and Chemistry.
Adolf Baeyer, Munich.

Marcellin Berthelot,
R. Bunsen,
H. L. F. Helmholtz,
A. W. Hofmann,
Mendeleeff,
Lord Rayleigh,
G. G. Stokes,
Julius Thomsen,
W. E. Weber,

Paris.

Heidelberg.
Berlin.
Berlin.
St. Petersburg.
Witham.
Cambridge.
Copenhagen.
Gbttingen.

Section IV. — 3.

Technology and Engineering.
Marquis de Caligny, Versailles.

F. M. de Lesseps,
Sir Wm. Thomson,

Paris.
Glasgow.

Class II. — Natural and Physiological Sciences. — 26.

Section I. — 6.

Geology, Mineralogy, and Physics of
the Globe.

H. Ernst Beyrich, Berlin.
Alfred Des Cloizeaux, Paris.
A. E. Nordcnskicild, Stockholm.
C. F. Rammelsberg, Berlin.
Sir* A. C. Ramsay, Beaumaris.
Ikinrich Wild, St. Petersburg.

Section II. — 7.
Botany.

J. G. Agardh, Lund.

Alphonse de Can lolle, Geneva.
Sir Joseph D. Hooker, London.
C. J. Maximowicz, St. Petersburg.
Carl Niigeli, Munich.

Julius Sachs, Wiirzburg.

Marquis de Saporta, Aix.

FOREIGN HONORARY MEMBERS.

477

Section III. — 10.
Zoology and Physiology.

P. J. Van Beneden, Louvain.

Du Bois-lleymoud, Berlin.

Thomas H. Huxley, London.

Albrecht Kolliker, Wiirzburg.

Lacaze-Duthiers, Paris.

Rudolph Leuckart, Leipsic.

C. F. \V. Ludwig, Leipsic.

Sir Richard Owen, London.

Louis Pasteur, Paris.

J. J. S. Steenstrup, Copenhagen.

Section IV. — 3.

.I/- dicine and Surgery.

C. E. Brown-Sequard, Paris.
Sir James Paget, London.

Rudolph Virchow, Berlin.

Class III. — Moral and Political Sciences. — '2 1 .

Section I. — 3.

Philosophy and Jurisprudence.

James Martineau, London.

Henry Sidgwick, Cambridge.

Sir James F. Stephen, London.

Section II. — 7.

Philology and Archaeology.

John Evans, Hemel Hempstead.
Pascual de Gayangos, Madrid.
Benjamin Jowett,
J. W. A. Kirchhoff,
G. C. C. Maspero,
Max Miiller,

Oxford.
Berlin.
Paris ?
Oxford.

Sir H. C. Rawlinson, London.

Section III. — 7.

Political Economy and History.

Due de Broglie, Paris.

Ernst Curtius, Berlin.

W. Ewart Gladstone, London.

Charles Merivale, Ely.

Theodor Mommsen, Berlin.

Jules Simon, Paris.

William Stubbs, Chester.

Section IV. — 4.

Literature and the Fine Arts.
Jean Leon Gerome, Paris.
John Ruskin, Coniston.

Leslie Stephen, London.

Lord Tennyson, Isle of Wight.

STATUTES AND STANDING VOTES

OF THE

AMERICAN ACADEMY OF ARTS AND SCIENCES.

(Adopted May 30, 1S54 : amended September 8, 1857, November 12,

1862, May 24, 1864, November 9, 1870, May 27, 1S73,

January 26, 1876, and June 16, 18S6.)

CHAPTER I.
Of Fellows and Foreign Honorary Members.

1. The Academy consists of Fellows and Foreign Honorary J/em-
bers. They are arranged in three Classes, according to the Arts and
Sciences in which they are severally proficient, viz. : Class I. The
Mathematical and Physical Sciences ; — Class II. The Natural and
Physiological Sciences; — Class III. The Moral and Political Sci-
ences. Each Class is divided into four Sections, viz. : Class L, Sec-
tion 1. Mathematics; — Section 2. Practical Astronomy and Geodesy;
— Section 3. Physics and Chemistry; — Section 4. Technology and
Engineering. Class II., Section 1. Geology, Mineralogy, and Physics
of the Globe; — Section 2. Botany ; — Section 3. Zoology and Physi-
ology;— Section 4. Medicine and Surgery. Class III., Section 1.
Philosophy and Jurisprudence; — Section 2. Philology and Archae-
ology;— Section 3. Political Economy and History; — Section 1.
Literature and the Fine Arts.

2. Fellows resident in the State of Massachusetts can alone vote at
the meetings of the Academy.* They shall each pay to the Treasurer
the sum of ten dollars on admission, and an annual assessment of ten
dollars, with such additional sum, not exeeediny; five dollars, as the
Academy shall, by a standing vote, from time to time determine.

3. Fellows residing out of the State of Massachusetts shall be
known and distinguished as Associate Fellows. They shall nol be
liable to the payment of any fees or annual dues, bnl on removing

* The number of Resident Fellows is limited by the Charter 10

480 STATUTES OF THE AMERICAN ACADEMY

within the State shall be admitted to the privileges,* and be subject
to the obligations, of Resident Fellows. The number of Associate
Fellows shall not exceed one hundred, of whom there shall not be
more than Jorty in either of the three Classes of the Academy.

4. The number of Foreign Honorary Members shall not exceed
seventy-Jive ; and they shall be chosen from among persons most emi-
nent in foreign countries for their discoveries and attainments in either
of the three departments of knowledge above enumerated. And there
shall not be more than thirty Foreign Members in either of these
departments.

CHAPTER II.
Of Officers.

1. There shall be a President, a Vice-President, a Corresponding
Secretary, a Recording Secretary, a Treasurer, and a Librarian, which
officers shall be annually elected, by written votes, at the Annual
Meeting, on the day next preceding the last Wednesday in May.

2. At the same time and in the same manner, nine Councillors
shall be elected, three from each Class of the Academy, but the same
Fellows shall not be eligible on more than three successive years.
These nine Councillors, with the President, Vice-President, the two
Secretaries, the Treasurer, and the Librarian, shall constitute the
Council. It shall be the duty of this Council to exercise a discreet
supervision over all nominations and elections. With the consent of
the Fellow interested, they shall have power to make transfers
between the several Sections of the same Class, reporting their action
to the Academy.

3. If any office shall become vacant during the year, the vacancy
shall be filled by a new election, and at the next stated meeting.

CHAPTER III.

Of the President.

1. It shall be the duty of the President, and. in his absence, of the
Vice-President, or next officer in order as above enumerated, to pre-
side at the meetings of the Academy ; to summon extraordinary

* Associate Fellows may attend, but cannot vote, at meetings of the Acad-
emy. See Chapter I., 2.

OF ARTS AND SCIENCES. 481

meetings, upon any urgent occasion ; and to execute or see to the
execution of the Statutes of the Academy.

2. The President, or, in his absence, the next officer as above
enumerated, is empowered to draw upon the Treasurer for such sums
of money as the Academy shall direct. Bills presented on account of
the Library, or the Publications of the Academy, must be previously
approved by the resjDective committees on these departments.

3. The President, or, in his absence, the next officer as above
enumerated, shall nominate members to serve on the different com-
mittees of the Academy which are not chosen by ballot.

4. Any deed or writing to which the common seal is to be affixed,
shall be signed and sealed by the President, when thereto author-
ized by the Academy.

CHAPTER IV.

Of Standing Committees.

1. At the Annual Meeting there shall be chosen the following
Standing Committees, to serve for the year ensuing, viz.: —

2. The Committee of Finance, to consist of the President, Treas-
urer, and one Fellow chosen by ballot, who shall have charge of the
investment and management of the funds and trusts of the Academy.
The general appropriations for the expenditures of the Academy shall
be moved by this Committee at the Annual Meeting, and all special
appropriations from the general and publication funds shall be
referred to or proposed by this Committee.

3. The Rumford Committee, of seven Fellows, to be chosen by
ballot, who shall consider and report on all applications and claims for
the Rumford Premium, also on all appropriations from the income of
the Rumford Fund, and generally see to the due and proper execu-
tion of this trust.

1. The Committee of Publication, of three Fellows, to whom all
memoirs submitted to the Academy shall be referred, and t<> whom
the printing of memoirs accepted for publication shall be intrusted.

5. The Committee on the Library, of three Fellows, who shall
examine the Library, and make an annual report on its condition and
management.

G. An Auditing Committee, of two Fellows, for auditing the
accounts of the Treasurer.

vol. xxiv. (n. s. xvi.) 81

482 STATUTES OF THE AMERICAN ACADEMY

CHAPTER V.

Of the Secretaries.

1. The Corresponding Secretary shall conduct the correspondence
of the Academy, recording or making an entry of all letters written in
its name, and preserving on file all letters which are received ; and at
each meeting he shall present the letters which have been addressed
to the Academy since the last meeting. With the advice and consent
of the President, he may effect exchanges with other scientific associ-
ations, and also distribute copies of the publications of the Academy
among the Associate Fellows and Foreign Honorary Members, as
shall be deemed expedient ; making a report of his proceedings at the
Annual Meeting. Under the direction of the Council for Nomination,
he shall keep a list of the Fellows, Associate Fellows, and Foreign
Honorary Members, arranged in their Classes and in Sections in
respect to the special sciences in which they are severally proficient ;
and he shall act as secretary to the Council.

2. The Recording Secretary shall have charge of the Charter and
Statute-book, journals, and all literary papers belonging to the Acad-
emy. He shall record the proceedings of the Academy at its meet-
ings; and after each meeting is duly opened, he shall read the record
of the preceding meeting. He shall notify the meetings of the Acad-
emy, and apprise committees of their appointment. He shall post up
in the Hall a list of the persons nominated for election into the
Academy ; and when any individual is chosen, he shall insert in the
record the names of the Fellows by whom he was nominated.

3. The two Secretaries, with the Chairman of the Committee of
Publication, shall have authority to publish such of the proceedings
of the Academy as may seem to them calculated to promote the in-
terests of science.

CHAPTER VI.
Of the Treasurer.

1. The Treasurer shall give such security for the trust reposed in
him as the Academy shall require.

2. He shall receive officially all moneys due or payable, and all
bequests or donations made to the Academy, and by order of the

OF ARTS AND SCIENCES. 483

President or presiding officer shall pay such sums as the Academy
may direct. He shall keep an account of all receipts and expendi-
tures ; shall submit his accounts to the Auditing Committee; and
shall report the same at the expiration of his term of office.

3. The Treasurer shall keep a separate account of the income and
appropriation of the Rumfbrd Fund, and report the same annually.

4. All moneys which there shall not be present occasion to expend
shall be invested by the Treasurer, under the direction of the Finance
Committee, on such securities as the Academy shall direct.

CHAPTER VII.

Of the Librarian and Library.

1. It shall be the duty of the Librarian to take charge of the books,
to keep a correct catalogue of the same, and to provide for the delivery
of books from the Library. He shall also have the custody of the
publications of the Academy.

2. The Librarian, in conjunction with the Committee on the Library,
shall have authority to expend, as they may deem expedient, such sums
as maybe appropriated, either from the Rumford or the General Fund
of the Academy, for the purchase of books, and for defraying other
necessary expenses connected with the Library. They shall have au-
thority to propose rules and regulations concerning the circulation,
return, and safe-keeping of books; and to appoint such agents for
these purposes as they may think necessary.

3. To all books in the Library procured from the income of the
Rumford Fund, the Librarian shall cause a stamp or label to be
affixed, expressing the fact that they were so procured.

4. Every person who takes a book from the Library shall give a
receipt for the same to the Librarian or his assistant.

5. Every book shall be returned in good order, regard being had
to the necessary wear of the book with good usage. And if any
book shall be lost or injured, the person to whom it stands charged
shall replace it by a new volume or set, if it belongs to a set, or pay
the current price of the volume or set to the Librarian ; and there-
upon the remainder of the set, if the volume belonged to a set, shall
be delivered to the person so paying for the same

6. All books shall be returned to the Library for examination, at
least one week before the Annual Meeting.

484 STATUTES OF THE AMERICAN ACADEMY

CHAPTER VIII.
Of Meetings.

1. There shall be annually four stated meetings of the Academy ;
namely, on the clay next preceding the last Wednesday in May (the
Annual Meeting), on the second Wednesday in October, on the second
Wednesday in January, and on the second Wednesday in March ; to
be held in the Hall of the Academy, in Boston. At these meetings
only, or at meetings adjourned from these and regularly notified, shall
appropriations of money be made, or alterations of the statutes or
standing votes of the Academy be effected.

2. Fifteen Fellows shall constitute a quorum for the transaction of
business at a stated meeting. Seven Fellows shall be sufficient to
constitute a meeting for scientific communications aud discussions.

3. The Recording Secretary shall notify the meetings of the Acad-
emy to each Fellow residing in Boston and the vicinity ; and he may
cause the meetings to be advertised, whenever he deems such further
notice to be needful.

CHAPTER IX.

Of the Election of Fellows and Honorary Members.

1 . Elections shall be made by ballot, and only at stated meetings.

2. Candidates for election as Resident Fellows must be proposed
by two or more Resident Fellows, in a recommendation signed by
them, specifying the Section to which the nomination is made, which
recommendation shall be transmitted to the Corresponding Secretary,
and by him referred to the Council for Nomination. No person rec-
ommended shall be reported by the Council as a candidate for elec-
tion, unless he shall have received a written approval, signed at a
meeting of the Council by at least eight of its members. All nomi-
nations thus approved shall be read to the Academy at a stated meet-
in"-, and shall then stand on the nomination list during the interval
between two stated meetings, and until the balloting. No person
shall be elected a Resident Fellow, unless he shall have been resident
in this Commonwealth one year next preceding his election ; and any
Resident Fellow who shall remove his domicile from the Common-
wealth, shall be deemed to have abandoned his Fellowship. If any

OF ARTS AND SCIEN*

person elected a Resident Fellow shall neglect for one year to pay his
admission fee, his election shall be void ; and, if any Resident Fellow
shall neglect to pay his annual assessments for two years, provided
that his attention shall have been called to this article, he shall be
deemed to have abandoned his Fellowship ; but it shall be in the
power of the Treasurer, with the consent of the Council, to dispense
(sub silentiu) with the payment both of the admission fee and of the
assessments, whenever in any special instance he shall think it advisa-
ble so to do.

3. The nomination of Associate Fellows shall take place in the
manner prescribed in reference to Resident Fellows ; and after such
nomination shall have been publicly read at a stated meeting previous
to that when the balloting takes place, it shall be referred to a Council
for Nomination ; and a written approval, authorized and s'gned at a
meeting of said Council by at least seven of its members, shall be
requisite to entitle the candidate to be balloted for. The Council may
in like manner originate nominations of Associate Fellows, which
must be read at a stated meeting previous to the election, and be
exposed on the nomination list during the interval.

4. Foreign Honorary Members shall be chosen only afcer a nomi-
nation made at a meeting of the Council, signed at the time by at
least seven of its members, and read at a stated meeting previou
that on which the balloting takes place.

5. Three fourths of the ballots cast must be affirmative, and the
number of affirmative ballots must amount to eleven to effect an elec-
tion of Fellows or Foreign Honorary Members.

6. Each section of the Academy is empowered to present lists of
persons deemed best qualified to fill vacancies occurring in the num-
ber of Foreign Honorary Members or Associate Fellows allotted to
it; and such lists, after being read at a stated meeting, shall he
referred to the Council for Nomination.

7. If, in the opinion of a majority of the entire Council, any Fellow
— Resident or Associate — shall have rendered himself unworthy of
a place in the Academy, the Council shall recommend to the Academy
the termination of his Fellowship; and, provided that a majority <>i
two thirds of the Fellows at a stated meeting, consisting of not less
than fifty Fellows, shall adopt this recommendation, his name shall be
stricken off the roll of Fellows.

486 STATUTES OP THE AMERICAN ACADEMY

CHAPTER X.

Of Amendments of the Statutes.

1. All proposed alterations of the Statutes, or additions to them,
shall be referred to a committee, and, on their report at a subsequent
meeting, shall require for enactment a majority of two thirds of the
members present, and at least eighteen affirmative votes.

2. Standing Votes may be passed, amended, or rescinded, at any
stated meeting, by a majority of two thirds of the members present.
They may be suspended by a unanimous vote.

CHAPTER XI.

Of Literary Performances.

1. The Academy will not express its judgment on literary or
scientific memoirs or performances submitted to it, or included in its
publications.

OF ARTS AND SCIENCES. 487

STANDING VOTES.

1. Communications of which notice has been given to the Secre-
tary shall take precedence of those not so notified.

2. Resident Fellows who have paid all fees and dues chargeable to
them are entitled to receive one copy of each volume or article printed
by the Academy, on application to the Librarian personally or by
written order, within two years from the date of publication. And
the current issues of the Proceedings shall be supplied, when ready
for publication, free of charge to all the Fellows and Members of the
Academy who desire to receive them.

3. The Committee of Publication shall fix from time to time the
price at which the publications of the Academy may be sold. But
members may be supplied at half this price with volumes which they
are not entitled to receive free, and which are needed to complete
their sets.

4. Two hundred extra copies of each paper accepted for publica-
tion in the Memoirs or Proceedings of the Academy shall be placed
at the disposal of the author, free of charge.

5. Resident Fellows may borrow and have out from the Library
six volumes at any one time, and may retain the same for three
months, and no longer.

6. Upon special application, and for adequate reasons assigned, the
Librarian may permit a larger number of volumes, not exceeding
twelve, to be drawn from the Library for a limited period.

7. Works published in numbers, when unbound, shall not be
taken from the Hall of the Academy, except by special leave of the
Librarian.

8. Books, publications, or apparatus shall be procured from the
income of the Rumford Fund only on the certificate of the Rumford
Committee, that they, in their opinion, will best facilitate and encour-
age the making of discoveries and improvements which may merit the
Rumford Premium.

.9. The annual meeting and the other stated meetings shall be
holden at half-past seven o'clock, P. M.

10. A meeting for receiving and discussing scientific communica-
tions shall be held on the second Wednesday of each month not ap-
pointed for stated meetings, excepting July, August, and September.

488 RUMFORD PREMIUM.

RUMFORD PREMIUM.

In conformity with the terms of the gift of Benjamin, Count Rum-
ford, granting a certain fund to the American Academy of Arts and
Sciences, and with a decree of the Supreme Judicial Court for carry-
ing into effect the general charitable intent and purpose of Count
Rumford, as expressed in his letter of gift, the Academy is empowered
to make from the income of said fund, as it now exists, at any annual
meeting, an award of a gold and silver medal, being together of the
intrinsic value of three hundred dollars, as a premium to the author of
any important discovery or useful improvement in light or in heat,
which shall have been made and published by printing, or in any way
made known to the public, in any part of the continent of America, or
any of the American islands ; preference being always given to such
discoveries as shall, in the opinion of the Academy, tend most to pro-
mote the good of mankind ; and to add to such medals, as a further
premium for such discovery and improvement, if the Academy see fit
so to do, a sum of money not exceeding three hundred dollars.

.3 I •'.—

INDEX.

A.

Abronia umbellata, Lam , 70.
Abutilon aurantiacum, Watson, 41.
crispum, Don, 40.
incanum, Don, 40.
Palmeri, Gray, 40.
scabrum, 41.
Acacia cochliacantha, Humb. &
Bonpl., 49.
Farnesiana, Willd., 49.
filicina, Willd., 48.
flexicaulis, Benth., 48.
Wrightii, Benth., 48.
Acalypha Pringlei, Watson, var., 77.
Achyronychia Cooperi, Torr. &

Gray, 71.
Agrostis verticillata, Vill., 81.
Air, density of, freed from moisture

and carbonic acid, 226.
Allionia incarnata, Linn., 68.
Allium hyalinum, Curran, 87.
Amarantus fimbriatus, Benth., 71.
Palmeri, Watson, 71.
Palmeri, Watson, var. (?), 71.
venulosus, Watson, 71.
Amoreuxia palmatifida, DC, 40.
Acetacetic ester, on the acidity of

the substituted, 309.
Acidity of the substituted malonic
esters, acetacetic ester, and
ketones, 309.
Anilidodinitrobenzvlmethylketone,
284.
properties, 281, 285.
sodium salt of, 285.
Anilidodinitrobenzylmethylketone-
hydrazone, 286.
properties, 286.
Anilidotrinitrotoluol, 255.
Anoda pentaschista, Gray, 40.
Antigonon leptopus, Hook. & Arm,
73.

Antirrhinum cyathiferum, Benth.,
66.

Kingii, Watson, var., 66.
Aplopappus spinulosus, DC, 54.
Apodanthera (?) Palmeri, 50.
Argemone albiflora, Horuem., 38.

Mexicana, var. (?), 38.
Argentic S chlorpyromucate, 327.
Aristida bromoides, HBK., 80.

dispersa, Trim, 80.

fugitiva, Vasey, 80.

Schiedeana, Trin. & Rupr., 80.
Aristolochia brevipes, Benth., 73.

var. acuminata, AVatson, 73.
Argythamnia Neo-Mexicana, Mull.,
77.

Palmeri, 77.

sericophylla, Gray, 77.
Asclepias albicans, 59.

subulata, Decaisne, 59.
Aster frutescens, 55.
.Ast-icralus Nuttallianus, Gray, 46.
A^-jiisquea emarginata, Miers,

39.
Atmospheric economy of solar radi-
ation, 26.
Atriplex Barclayi, Dietr., 72.

elegans, Dietr , 72.

linearis, 72.
Ayenia Berlandieri, Watson, 42.

filiformis, 42.

B.

Baccharis sarothroides, Gray, 55.

Bseria Parishii, 88.

Bahia Palmerii, 83.

Baric 8 chlorpyromucate, 326.

/3 chlorpyromucate, 332.

/3-y dichlorpyromucate, 336.

$8 dichlorpyromucate, 345.

X dichlorpyromucate, 318.

490

INDEX.

Baric x dichlorsulphopyromucate,

oo'j.

Bebbia juncea, Greene, 58.
Beloperone Californica, Benth., 67.
Beta vulgaris, Linn., 72.
Bigelovia diffusa, Gray, 54.

veneta, Gray, var. (?), 55.
Bouchea dissecta, 68.
Boerhaavia alata, 69.

erecta, Linn., 69.

erecta, Linn., var., 69.

Palmeri, 69.

paniculata, Rich., 68.

scandens, Linn., 70.

spicata, Choisy, 70.

var. (?) Palmeri, 70.

triquetra, 69.

Wrightii, Gray, 69.

Xanti, 69.
Bouteloua arenosa, Vasey, 81.

aristidoides, Thurb., 81.

bromoides, Lag., 81.

polystachya, Torr., 81.

var. major, 81.

Rothrockii, Vasey, 81.
Bourreria Sonorse, 62.
Brickellia Coulteri, Gray, 51.

floribunda, Gray, 54.
Brodiasa Palmeri, 78,
Bromanilidodinitrophenylmalonic
ester, 298.

properties, 298.
Bromamidooxindol, properties of,
303.

chloride of, properties of, 304.
Bromdinitrotrianilidobenzol, 293.

properties, 294.
Bromdinitrobenzylmethylketone,
280.

properties, 282.
Bromdinitrophenylmalonic ester,

preparation of, 236.

reaction by which it is formed,

saponification of, 240.

constitution of, 248.
Bromdinitrephenylacetic ester,

preparation of, 274.

properties, 274.

reactions, 278.
Bromfurfuracrylate, argentic, 371.

baric, 370.

calcic, 371.

ethyl, 372.

sodic, 371.
Bromfurfuracrylic acid, 369, 372.

Bromfurfurbromacrylate, argentic,
374.
baric, 373.
ethyl, 374.
potassic, 374.
Bromfurfurdibrompropijiiic acid,

366.
Bromnitrobenzols, general consid-
erations in regard to certain
1 compounds prepared from,

306.
reactions, 307.
bromtrinitrophenylmalonic ester,
preparation of, 258.
constitution of. 268.
precipitates, 261.
properties, 260.
salts of, 261.
sodium salt, 262.
study of the reaction by -which
it is formed. 263.
Bunchosia parvifolia, 42.
Bursera 41indsiana, Benth. & Hook.,
44.
laxiflora, 44.
micropbylla, Gray, 43.
pubescens, 44.

Cacalia tussilaginoides, HBK., 84.
Csesalpinia gracilis, Benth., 47.

Palmeri, 47.
Calathea crotalifera, 86.
Calcic 8 chlorpyromucate, 327.

/3 chlorpyromucate, 332.

/3y dichlorpyromucate, 338.

X dichlorpyromucate, 349.
Californian Plants, descriptions of
new species of, with miscel-
laneous notes, 82.
Calliandra Coulteri, Watson, 49.

eriophylla, Benth., 49.
Capsicum annuum, Linn., 65.

baccatum, Linn., 65.

cordiforme, Mill., 65.

var. globosum, Dun. (?), 65.
Carbonic acid, determination of,
221.

density of, 228.

specific gravity of, 230.
Cardamine angelorum, 39.

Palmeri, 38.
Cardiospermum halicacabum,Linn.,
45.

INDEX.

401

Cassia Covesii, Gray, 47.

nictitans, Linn., 47.
Cathesticum erectum, Vasev &

Hack., 80.
Celtis pallida. Torr., 77.
Cenchrus etehinatus, Linn., 80.

myosuroides, II BK., 80.

Palmeri, Vasey, 80.

tribuloides, Linn., 80.
Cereus Pringlei, Watson, 52.
Cheilanthes myriophylla, Desv., 81.

Pringlei, Davenport, 81.
Chenopodium album, Linn., 72.

ambrosioides, Linn., 72.
Cliloris elegans, Kunth, 81.
Chlorpyromucainide. 8, 328.
Chlorpyromucic acids, 320.

acid, 0, 330.
Chrome iron ore, the determination

of chromium in, SS.
Chromium, the determination of,

in chrome iron ore, 88.
Citharexylum flabellifolium, 67.
Cladothrix lanuginosa, Nutt., 71.
Cleome tenuis, 39.
Coahuila irons, crystalline struc-
ture of the, 30.

examination of irons, —
Allen Co., 33, 34.
Butcher (Coahuila), 30-32.
Chattooga Co., 31.
Maverick Co., 34.
Saltillo, 32.
Santa Rosa, 34.
figures, 30, 31, 32.
Cocculus diversifolius, DC, 38.
Coldenia angelica, 62.

brevicalyx, 62.

Palmeri, Gray, 62.
Collinsia Wrightii, 84.
Colubrina glabra, 41.
Communications, —

W. I). Bancroft, 288.

W. B. Bentley, 250.

Arthur M. Comey, 14.

Josiah Parsons Cooke, 202.

Charles R. Cross, 94, 113.

W. S. Hendrixson, 376.

Henry B. Hill, 320.

Oliver W. Huntington, 30, 313.

C. C. Ilutchins, 125.

C. Loring Jackson, 105, 234,
256, 271, 288, 306,320.

Leonard P. Kinnicutt, >8.

W. R. Livermore, 164.

Joseph Lovering, 185, 380, 441.

George Dunning Moore, 256.

271.
Daniel Edward Owen, 1:25. ■

A. W. Palmer, 105.
G. H. Parker, 24.
George W. Patterson, 88.

B. O. Peirce, 146.

\V. S. Robinson, 1, 234.

Annie W. Sabine. 90, 94.

Arthur Searle/26.

Samuel Sheldon, 176, 181.

John Trowbridge, 176, 181.

W. \V. Warren, 250.

Sereno Wat -on, 36.

Arthur S. Williams, 113.

R W. Willson, 148.
Conobea intermedia, Gray, 66.
Cordia (.Ireggii, Torr., 61.

var. (?) Palmeri, 61.

Palmeri, 62.
Cottea pappophoroides, Kunth, 81.
Cracca Edwardsii, Gray, 46.
Crescentia alata. HBK., 66.
Cressa Cretica, Linn., 64.
Croton Pringlei, Watson, 77.
Cryptocarpus (?) capitatus, 71.
Crystalline concretions of mica, 318.
Crystalline growth, features of, 313.

" The Butcher Irons," 313.

Widmanstattian figures in Spie-
gel Eisen, 315.
Crystalline plates in galena, 317.
Cucurbita cordata, 50.
Cuscuta Americana, Linn., 64.

Palmeri, 64.

umbellata, HBK., 64.
Cyperus aristatus, Rottb., 79.

articulatus, Linn., 79.

esculentus, Linn., 79.

ferax, Rich., 79.

laevigatas, Linn., 79.

speciosus, Vahl, 79.

D.

Dalea Emoryi, Gray, 46.

megacarpa, Watson, 45.

mollis, Benth., 16.

Parryi, Gray, 46.

Pringlei, Gray, 15.
Desmanthus Jamesii, Torr.& Gray,
var., 48

virgatus, Benth., 48.
Desmodium Bcopalorum, 47.
Dianthera SononB, 67.

492

INDEX.

Dibrom-8-chlorpyromucic acid, /3y,

360.
Dibromdinitrophenylinalonic ester,
291.
properties, 296.
precipitates, 297.
reduction of, 299.
Dichlor-8-brompyromucic acid, /3y,

360.
Dichlor-S-nitropyromucic acid, /3y,

361.
Dichlorpvromucaniide, /3y, 339.

/3S, 346.
Dichlorpyromucic acid, j3y, 335.
/3S, 344.
x, 348.
Dichlorpyromucic acids, preparation

of isomeric, 341.
Dioxymaleic acid, on the so called,

376.
Diplachne dubia, Benth., 81.
imbricata, Vasey, 81.
Tracyi, Vasey, SI.
viscida, Scribner, 81.
Diphysa sennoides, Benth , 46.
Distichlis maritima, Ral, 81.
Dodonsea viscosa, Linn., 45.
Drymaria crassifolia, Benth., 40.
Dysodia porophylloides, Gray, 58.

E.

Echinopepon insularis, 51.

Palmeri, 52.
Echinopterys Lappula, Juss., 43.
Eclipta alba, Hassk., 56.
Electrodes, the strength of the mi-
crophone current as influ-
enced by variations in normal
pressure and mass of the. 90.
Ellis J a chrysanthemifolia, Benth ,

61.
Eleocharis capitata, R. Br., 79.
Eleusine iEgyptiaca, Pers., 81.

Tndica, Gaertn., 81.
Elytraria tridentata, Vahl, 66.
Encelia farinosa, Gray, 56.
Eragrostis major, Host, 81.
Purshii, Schrad., 81.
var. diffusa, 81.
Erigeron sanctarum, 83.
Eriochloa punctata, Desv., 80.
Eriogonum fasciculatum, Benth.,
73.
Esmeraldense, 85.

Eriogonum inflatum, Torr., 73.

insigne, Watson, var., 73.

gracilipes, 85.
Erodium Texanum, Gray, 43.
Eschscholtzia csespitosa, Benth., 38.
Ethyl 5 chlorpyromucate, 328.

j3 chlorpyromucate, 333.

(3y dichlorpyromucate, 339.

/33 dichlorpyromucate, 346.

X dichlorpyromucate, 350.

pyromucic tetrachloride, 322.
Eucnide cordata, Kell., 50.
Eupatorium sagittatum, Gray, 54.
Euphorbia albomarghiata, Torr. &
Gray, 74.

Brasiliensis, Lam., 74.

Californica, Benth. (?), 76.

capitellata, Engelm., 74.

var. laxiflora, 74.

eriantha, Benth., 76.

florida, Englm., 75.

glyptosperma, Engelm., 75.

intermixta, 74.

maculata, Linn., 75.

Magdalense, Benth. (?), 74.

misera, Benth., 76.

pediculifera, Engelm., 75, 76.

var. linearifolia, 76.

petrina, 75.

polycarpa, Benth., 75.

var. hirtella, Boiss., 75.

portulana, 75.

serpyllifolia, Pers., 75.

setiloba, Engelm., 75.

tomentulosa, Watson, 74.

trachysperma, Engelm., 74.
Evolvulus linifolius, Linn., 63.

F.

Fagonia Californica, Benth., 43.
Ficus fasciculata, 78.

Palmeri, 77.

Sonorse, 78.
Fimbristylis laxa, 79.
Fellows, Associate, deceased, —

Spencer F. Baird, 406.

Frederick A. P. Barnard, 429,
441.

John Call Dalton, 424, 415.

Rowland G. Hazard, 429.
Fellows, Associate, elected, —

George Brown Goode, 419.

Alexander Johnston, 425.

John Nelson Stockwell, 425.

INDEX.

493

Fellows, Associate, list of, 474.
Fellows, Resident, deceased, —
George Ruinford Baldwin, 429.
Jonathan Ingersoll Bowditch,

423, 435.
Charles S. Bradley, 406.
John Dean, 406.
Samuel Kneeland, 418, 438.
Fellows, Resident, elected, —

Edward Hickling Bradford,

4 25.
Arthur Tracy Cabot, 425.
Franklin Carter, 425.
David Williams Cheever, 425.
Harold Clarence Ernst, 425.
Reginald Heber Fitz. 425.
Samuel Henshaw, 425.
John Homans, 425.
Frederick Irving Knight, 425.
George Hinckley Lyman, 425.
Cecil Hobart Peabody, 424.
Peter Schwamb, 424.
Franklin Bache Stephenson,

425.
Frank William Taussig, 425.
Henry Pickering Walcott, 425.
Barrett Wendell, 425.
Henry Willey, 424.
Fellows, Resident, list of, 471.
Foreign Honorary Members de-
ceased, —
Matthew Arnold, 406.
Michel Eugene Chevreul, 429,

452.
Franciscus Cornelius Donders,

129, 465.
Rudolf Julius Emanuel Clau-

sius, 418, 458.
Foreign Honorary Members elect-
ed, —
Charles Jaques Victor Albert,

Due de Broglie, 419.
John Evans, 419.
Anatole Francois Hue, 426.
John William Adolf Kirchhoff,

419.
Carl Johann Maximowicz, 418.
Dmitri Ivauowitsh Mendeleeff,

425.
Friherre Adolf Erik Norden-

skiold, 418.
Henry Sidgwick, 119.
Leslie Stephen, 419.
John William Strutt, Lord

Rayleigh, 418.
Wilhelm Edward Weber, 425.

Foreign Honorary Members, list of,

476.
Frankenia Palmeri, Watson, 4o.
Frauseria ambrosioides, Caw, 55.

dumosa, Nutt, 56.

ilicifolia, Gray, 55.

teuuifolia, Gray, 56.
Frcelichia alata, Watson, 71.
Furfuracrylamide, 366.
Furfuracrylic acid, on certain de-
rivatives of, 365.

G.

Galium stellatum, Kell., 53.
Galphimia angustifolia, Benth.,
42.
var. oblongifolia, Gray, 42.
Gas densities, a new method of de-
termining, 202.
absorption apparatus, 216.
balance and weights, 202.
carbonic acid, determination

of, 221.
correction for potash bulb,

219.
density of air freed from moist-
ure and carbonic acid, 226.
density of hydrogen. 227.
purifying and drying appara-
tus, 214.
specific gravity of oxygen, .

228.
thermometers and barometers
and their corrections, 209.
Gaura parviflora, Dough, 49.
Genipa echinocarpa, Gray, 53.
Gilia Palmeri, 61.
Gomphrena Sonora?, Torr., 71.
Gossypium Davidsoni, Kell., 41.

herbaceum, Linn.. 1 1 .
Gray, Asa, meeting in commemo-
ration of, 406.
Guaiacum C'oul t <-ri. Gray, !•">.
Guaymas, Mexico, collection of
plants made at, by Dr. K
Palmer, in 1887, 30.
Gulf of California, collection of
plants made by Dr. E.
Palmer, in 1887, on the
island of San Pedro Martin
in the, 36.
Gutierrezia EuthamisB, Torr. &
Gray, 54.

494

INDEX.

H.

Hsematoxylon boreale, Watson, 47.
Haplophytou Cimicidum, A. DC,

59.
Helianthus annuus, 56.
Heliotropium Curassavicum, Linn.,
63.

phyllostachy'um, Torr., 63.
Hermannia pauciflora, Watson, 42.
Heteropogon contortus, R. & S., 80.
Hibiscus Coulteri, Gray, 41.

denudatus, Benth., 41.
Hilaria cenchroides, HBK., 80.

var. longifolia, 80.
Himantostennna Pringlei, Gray, 61.
Hirtea macroptera, DC., 43.
Hoffmanseggia microphylla, Torr.,
47.

var. glabra, 47.
Hofmeisteria crassifolia, 53. •

pubescens, 54.
Horsfordia Palmeri, 40.

Newberryi, Gray, 40.

rotundifolia, 40.
Hosackia rigida, Benth., 45.

strigosa, Nutt., 45.
Hydrogen, density of, 227.

specific gravity of, 228.
Hyptis Emoryi, Torr., var., 68.

Palineri, 68.

I.

Indigofera Anil, Linn., 46.

inucronata, Spreng., 46.
Ionidium polyalsefolium, Vent., 40.
Ipomcea Bona-nox, Linn., 63.

coccinea, Linn., 63.

hederacea, Jacq., 63.

leptotorna, Torr., 63.

Palmeri, 63.

triloba, Linn., var., 63.
Iresine alternifolia, 72.
Irons, Coahuila, crystalline struc-
ture of the, 30.

J.

Jacobinia ovata. Gray, 67.

var. subglabra, 67.
Jacquinia pungens, Gray, 59.
Jacquernoutia Palmeri, 63.

Jacquemontia Pringlei, Gray, 63.

var. glabrescens, Gray, 63.
Janusia Calitornica, Benth., 4^5.
Jatropha canescens, Midi., 76.

Palmeri, 76.

spathulata, Miill., 76.

var. sessiliflora, Miill., 76.
Juncus robustus, Watson, 79.
Jussisea octonervis, Lam., 49.

K.

Ketones, on the acidity of the sub-
stituted, 309.
Kosteletzkya Coulteri, Gray, 41.
Krameria canescens, Gray, var., 40.

parvifolia, Benth., 40.
Krynitzkia angusti folia, 63.

ramosissima, Greene, 63.

L.

Lagascea decipiens, Hemsl , 55.
Lantana Camara, Linn., 67.
Laphamia ( ?), 57.
Lepidium Palmeri, 39.
Leptochloa mucronata, Kunth, 81.
Leptosyne partlienioides, 56.

var. dissecta, 56.
Lippia Palmeri, 67.
Lobelia splendens, Willd., 59.
Lobster, a preliminary account of
the development and histol-
ogy of the eyes in the, 24.
Loliuin perenne, Linn., 81.
Loranthus Sonorse, 73.

spirostylis, DC, 73.
Los Angeles Bay, collection of
. plants made at, by Dr. E.
Palmer, in 1887, 36.
Louteridium Donnell-Smithii, Wat-
son, 85.
Lovering, Joseph, an address on
the presentation of Rumford
Medals to Prof. A. A. Mich-
elson, 380.
Lupinus, 45.

Arizonicus, Watson, 45.
Lyciuin, 65.

Andersoni, Gray, 65.
var. pubescens, 65.
barbinodum, Miers, 65.
carinatum, 65.
Richii, Torr., 65.

INDEX.

495

Lympetaleia rupestris, Gray, 50.
Lyrocarpa Coulteri, Hook. & Arn.,

39.
Lysilouia microphylla, Benth., 49.]

M.

Magnet, the strength of the induced
current with a magneto tele-
phone transmitter, as influ-
enced by the .strength of the,
113.
Magneto telephone transmitter^ the
strength of the induced cur-
rent with a, as influenced by
the strength of the magnet,
113.
apparatus, 113.

diaphragm, magnetizatiou of
the, 117.
polarizing the, 123.
results witli the thick or thin,
123.
explanation of results, 110.
figures, 114, 118.
tables, 115-117, 119-122.
Malonic esters, on the acidity of

the substituted, 309.
Malperia, 54.

tenuis, 54.
Manihot angustiloba, Mull , 77.
Marsdenia edulis, 61.
Martynia althesefolia, Benth., 60.
fragrans, Lindl., 66.
Palmeri. 00.
Maximowiczia Sonorse, 51.
Maytenus phyllanthoides, Benth.,

44.
Microphone current, the strength of
the, as influenced by varia-
tions in normal pressure and
mass of the electrodes, 90.
apparatus, 90.
figures, 92,
method, 90.
table, 91.
Microphone currents, researches
on. Ill
actual strength of working cur-
rents, in;!.
figures, 90, 97, 100, I'll,
loss of current in long distance

telephony, 101.
tables, 91, 95, 98, 09, 102, 103.
Microseris anomala, Watson 84.

Mimosa laxiflora, Benth., 48.
Mhnulus deflexus, Si
Mimulus moschatus, Dough, 00.
Mirabilis California, Gray, 08.

tenuiloba, Watson, 08.
Mohavea viscida, Grajr, 05.
Mollugo Cerviana, Ser., 52.

verticillata, Linn., 52.
Melilotus parviflorus, Desv., 45.
Melochia speciosa, 12.

tomentosa, Linn., 42.
Mentzelia adherens, Benth., 50.

inultiflora, Nutt.. 50.
Metabronidinitrophenylacetic acid,
241.

ammonincal solution of, 245.

precipitates of same, 215.
Metabronulinitrophenylacetate, ar-
gentic, 1217.
Metabromtoluol, on some nitro de-
rivatives of, 250.
Metabromtrinitrotoluol, 252.

properties of, 25:5.

constitution of, 251.
Metastelma albiflora, 60.

Pringlei, Gray, var. (?), 60
Methyl furfuracrylate, 365.
Mexico, collection of plants made
by Dr. E Palmer, in 1687,
about Guaymas, 36.
Muhlenbergia debilis, Trin., 80.

spicifonnis, Trin., 80,

tenella, Trin., 80.
Muleje, collection of plants made
at, by Dr. E. Palmer, iu
1867, 30.

N.

Naias major, All., 79.
Nasturtium (V) laxum, 39.
Nemastylis Dugesii, 80.

Pringlei, 85.
Nicolletia Edwardsii, Gray, 58
Nicotiana Clevelandi, Gray, 65.

trigonophylla, 1 inn., 05.
Nissolia Schottii, < Iray, 10.
Notholsena cretacea, Liebm., 81.

Lemmoni, Eaton, 81.

o.

(Enothera angelorum, 19.
cajspitosa, Nutt., 19.

496

INDEX.

CEnothera cardiophylla, Torr., 49.
Oligomeris glaucescens, Camb., 39.
Opuntia, 52.
Oxygen, s|)ecific gravity of, 228.

Palafoxia linearis, Lag., 57.
Panicum capillare, Linn., var., 80.

colonum, Linn., 80.

dissitiflorum, Vasey, 80.

fasciculatum, Sw., 80.

var. niajus, 80.

Hallii, Vasey, 80.

lachnanthum, Torr., 80.

paspaloides, Pers., SO.

sanguinale, Linn., 80.

var. ciliare, 80.
Pappophorurn aperturn, Munro, 81.

Wrightii, Watson, 81.
Parallax, solar, 399.
Paspalum distichum, Linn., 79.

publiflorum, Ruprecht, 79.
Passiflora foetida, Linn., 50.
Pattalias, 60.

Palmeri, 60.
Paulliuia Sonorensis, 45.
Pectis angustifolia, Torr., 58.

Conlteri, Gray, 58.

Palmeri, 58.

prostrata, Cav., 58.

punctata, Jacq., 58.
Pedilanthus macrocarpus, Benth.,

74.
Pellsea Seemanni, Hook., 82.

Wrightiana, Hook., 82.
Pelucha, 55.

trifida, 55.
Pentaniidobenzol, on, 105-

trichloride of, 108.
Perezia Palmeri, 58.
Perityle Calif or nica, Benth., 57.

deltoidea, 57.

Palmeri, 57.
Petalonyx linearis, Greene, 50.
l'eucephyllum Schottii, Gray, 58.
Phacelia crenulata, Torr., 61.

pauciflora, 61.
Phaseolus atropurpurens, DC, 47.

var. sericeus, Gray, 47.

filiform is, Benth., 47.
Phaulothamnus spinescens, Gray,

73
Philibertia linearis, Gray, 59.

var. heterophylla, Gray, 59.

Philibertia Pavoni, Hemsl., 59
Phoradendron Californicum, Nutt.,
73.
flavescens, Nutt., 73.
Phragmites communis, Trin., 81.
Physalis angulata, Linn., 64.
var. Linkiana, Gray, 64.
pubescens, Linn., 64.
Pithecolobium Sonorae, 49.
Plantago Patagonica, Jacq., 6S.
Plants, descriptions of some new
species of, chiefly Califor-
nian, with miscellaneous
- notes, 82.
Plants, upon a collection of, made
by Dr. E. Palmer, in 1887,
about Guaymas, Mexico, at
Muleje and Los Angeles Bay
in Lower California, and on
the Island of San Pedro Mar-
tin in the Gulf of California,
36.
indeterminable species, 82.
Polygonum Persicaria, Linn., 73.
Porophyllum crassifolium, 57.
gracile, Benth., 57.
Seemanni, Schultz Bip., 57.
Portlandia pterosperma, 52.
Portulaca oleracea, Linn., 40.
Potamogeton pectinatus, Linn., 79.
Potassic 8 chlorpyromucate, 327.
Prosopis articulata, 48.

heterophylla, Benth., 48.
Palmeri, 48.
Psilactis Coulteri, Gray, 55.

R.

Randia obcordata, 53.

Thurberi, 53.
Rhizophora Mangle, Linn., 49.
Rhynehosia phaseoloides, DC, 47.
Riddellia Cooperi, Gray 57.
Rothrockia cordifolia, Gray, 61.
Ruellia tuberosa, 66.
Rumford Medals, presentation of,
to Prof. Albert A. Michel-
son, 380, 403, 427.
Rumford Premium, 488.

Salvia privoides, Benth., 68.

San Pedro Martin, island of, collec-

INDEX.

497

tion of plants made on, by
Dr. E. Palmer, in 18S7, 36.
Scirpus Olneyi, Gray, 79.
Sebastiania (?) bilocularis, Watson,

77.
Senecio Lemmoni, Gray, 58.
Serjania Palmeri, 45.
Sesbania macrocarpa, Muhl., 46.

var. picta, 46.
Setaria caudata, R. & S., 80.
var. pauciflora, 80.
composita, II BK., 80.
Sida carpinifolia, Linn., 40.
Sideroxylon leucophyllum, 59.
Silene Bernardina, S2.
Simmondsia Californica, Xutt., 76.
Sisymbrium canescens, Xutt., 39.
Sisyriuchium anceps, Cav., 86.
Sodic zincate, fusible, analysis of
the, 17.
atomic ratio of zinc to so-
dium in, 18.
properties of the, 19.
infusible, analyses and proper-
ties of the, 20.
atomic ratio of zinc to so-
dium in the, 21.
properties, 21.
Sodic zincates, on, 14.

experiments with magnesic ox-
ide, 22.
preparation of the, 16.
research by Saux, Fremy, and

others, 14, 15.
study of other zincates, 22.
Sodium malonic ester, on the action
of, on tribromdinitrobenzol,
234.
Solatium Hindsianum, Benth., 64.
nigrum, Linn.. 64.
var. nodiflorum, Gray, 64.
Solar parallax, 'i'XK
Solar radiation, atmospheric econ-
omy of, 26.
Sorghum Halepense, Pers., 80.
Sphseralcea, sp., 41.
ambigua, Gray, 41.
axillaris, 41.
Spirostachys occidentalis, Watson,

72.
Sporobolus cryptandrus, Gray, 80.
Domingensis, Kunth, 81.
humifusus, Kunth, 80.
Virginicus, Kunth, 81.
Stachys cocci nea, Jacq., 68.
Statutes and Standing Votes, 479.

VOL. XXIV. (n\ 9. XVI. J

Stegnospermahalimifolium, Benth.,
73.

Stemodia durantifolia, Sw., 66.

Stipa Californica, Vasey, 80.

Story, W. W., letter on the cele-
bration of the eight hun-
dredth anniversary of the
University of Bologna. 121.

Suseda Torreyana, Watson, 72.

Tamarindus Indica, Linn., 47.
Telephone transmitter, magneto, the
strength of the induced cur-
rent with a, as influenced by
the strength of the magnet,
113
Tephrosia constricta, 46.
Palmeri, 46.
tenella, Gray, 46
Tetrabromdinitrobenzol, on, 288.
preparation of, 289.
properties, 291.
Tragia nepetsefolia, Cav., 77.

var. amblyodonta, Mull., 77.
Triamidodinitrobenzol, 106.

properties, 107.
Trianilidodinitrobenzol, 111.

properties, 112.
Trianthema monogyna, Linn., 52.
Tribromdinitrobenzol, on the ac-
tion of sodium acetacetic
ester upon, 271.
preparation of, 273.
action of sodium malonic ester
on, 234, 256.
Tribulus Californieus, Benth., 43
grandiflorus, Benth. & Hook.,

43.
maximus, Linn., var. ?, 43
Trichloride of pentamidobenzol,
108.
properties, 109.
Trichlorpyromucate, argentic, 357.
calcic, 356.
ethyl, 358.
potassic, 357.
Trichlorpyromucamide, 858.
Ti iihloipvromucic acid, 353.
Trichoptilium incisum, Gray, 58.
Trinitrophenylendimalonic ester,
268.
properties. 209.
Triodia pulchella, IIBK., 81.

32

498

INDEX.

Triphasia trifoliata, DC, 43.
Trixis angustifolia, DC, 59.
var. latiuscula, Gray, 59.

V.

Vallesia dichotoma, Ruiz & Pavon,

59.
Verbesina Palmeri, 56.
Viguiera laciuiata, Gray, 56

Parishii, Greene, 56.
Viscainoa geniculata, Greene, 43.
Vitex mollis, HBK., 68.
Votes, Standing, and Statutes, 479.

W.

Washingtonia Sonorse, 79.
Waltheria detonsa, Gray, 42.
Wislizenia Palmeri, Gray, 39.

z.

Zincates, sodic, on, 14.
Zizyphus lycioides, Gray, 44.

var. canescens, Gray, 44.

Sonorensis, 44.

^, "7 V - - -

^^ ./>

rS O^

'&!£

' -2S5js X

^1

V^7 /%

7? A

:-j ■-

■" ^* - ~

£-K>.?^ *

2Jz

~?V

KsS:

;S Hfe>

*>?*■

vt*

New York Botanical Garden Librar

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIHIIIIMI

3 5185 00257 9132

— w

l*Jk

^

7^

> '■

^
-^

:t> ; j?

^

T-^s

3*2

->"> ?.

:?■' .

<>£«

x .^x

?<? ■»:>»

'^x^'\

^

- >

. ^-4 »

■7V7)

m

■iv;

"}*■">