PROCEEDINGS
AMERICAN ACADEMY
ARTS AND SCIENCES.
Vol.. XXXIV.
FR03I MAY, 1898, TO MAY, 1899.
BOSTON, MASS.:
JOHN WILSON AND SON.
Saiubcrsttg Press.
1899.
a ^'7 3
CONTENTS.
Page
I. On the Thermal Conductivities of certain Poor Conductors. — I.
By B. 0. Peirce and R. W. Willson 1
II. The Contact-Potential between Metals and Fused Salts, and the
Dissociation of Fused Salts. By Clarence McCheyne
Gordon 57
III. On Fluctuations in the Composition of Natural Gas. By Francis
C. Phillips 69
IV. Some Electrochemical and Thermochemical Relations of Zinc and
Cadmium Amalgams. By Theodore William Richards
and Gilbert Newton Lewis 85
V. Trinitrophenyhnalonic Ester. Second Paper. By C. Loring
Jackson and J. I. Piiinney 101
VI. On the Action of Sodic Ethylate on Trihromdinitrobenzol. By C.
Loring Jackson and Waldemar Koch 117
VII. On Certain Derivatives of Symmetrical Trichlorbenzol. By C. Lor-
ing Jackson and F. II. Gazzolo 137
VIII. Shoreline Topography. By F. P. Gulliver 149
IX. Japanese Collembola. By Justus Watson Folsom .... 259
X. Tlie Use of the Transition Temperatures of Complex Systems as
Fixed Points in Thermometry. By Theodore William
Richards and Jesse Briggs Churchill 275
XI. On the Thermal Conductivity of Cast Iron. By E. H. Hall and
C. H. Ayres 281
IV CONTENTS.
Page
XII. On the Optical Characters of the Vertical Zone of Amphiholes and
P0oxenes ; and on a new Method of determining the Extinc-
tion Angles of these Minerals by Means of Cleavage Pieces.
By R. a. Daly 309
XIII. A Revision of the Atomic Weight of Nickel. Second Paper. —
The Determination of the Nickel in Nickelons Bromide. By
Theodore William Richards and Allerton Seward
• CUSHMAN 323
XIV. A Revision of the Atomic Weight of Cobalt. Second Paper. —
The Determination of the Cobalt in Cohaltous Bromide. By
Theodore William Richards and Gregory Paul
Baxter 349
XV. A Comparative Study of Etch-Figures. The Amphiboles and Py-
roxenes. By R. a. Daly 371
XVI. On a neiv Variety of Hornblende. By R. A. Daly .... 431
XVII. The Orthojiteran Genus Schistocerca. By Samuel II. Scudder 439
XVIII. On Hardystonite, a new Calcium-zinc Silicate from Franklin
Furnace, New Jersey. By John E. Wolff .... 477
XIX. 1. Eleocharis ovata and its American Allies.
2. Scirpus Eriophorum and some related Forms. By 1\I. L.
Fernald 483
XX. 1. Revision of the Genera Montanoa, Perymeninm, and Zalu-
znnia. Bv B. L. Robinson and J. M. Greexman.
2. Synopsis of the Genus Verbesina, ivith an Analytic Key to the
Species. By B. L. Robinson and J. jNI. Greenman.
3. Some new Species, extended Ranges, and newly noted Identities
among the Mexican Phanerogams. By J. M. Greenman 505
XXI. Pctrographical Notes on some Rocks from the Fiji Islands. By
Arthur S. Eakle 579
XXn. Investigations on Periodicity in the Weather. By H. Helm
Clayton 597
CONTENTS. r
Page
Proceedings 621
A Table oj Atomic Weights. By Theodoue William Richards . 637
Report of the Council 639
Biographical Notices 641
Justin Winsor 641
Samuel Eliot 646
Jules Mai'cou 651
Theodore Lyman 656
List of the Fellows and Foreign Honorary jNIembers . . 668
Statutes and Standing Votes 673
Index 685
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. No. 1. — August, 1898.
ON THE THERMAL CONDUCTIVITIES OF CERTAIN
POOR CONDUCTORS. — I.
By B. O. Peirce and R. W. Willson.
Investigations on Light and Heat, made and published wholly or in part with Appropriation
FROM the KUMFORD FOND.
MARINE BIOLOGICAL LABORATORY.
Received %lJ^.,..I^.X^R'^..
Accession No. 1^. itzf-
Given by 0^rroy^Ji^L^:^C^.,0Ln^r^.^.9r:^:
Place,
ON THE THERMAL CONDUCTIVITIES OF CERTAIN
POOR CONDUCTORS. — L
By B. O. Peikck and R. W. Willson.
Presented April 13, 1898.
We have been engaged for several years in an attempt to measnre, by
the aid of the so called " Wall Method," the thermal conductivities of
certain relatively poor conductors,* and the variations of these conduc-
tivities witii the temperature. We have at length succeeded in over-
coming some' of the difficulties which we have encountered, and are now
ready to describe our apparatus and to give the results of a number of
observations made with it.
When one end of a regular right prism of 2 n sides made of homoge-
neous material is kept at a constant temperature, ]^„ and the other end at
a constant temperature, V^, while its other faces are kept as nearly as
possible at some constant temperature between F^ and FJ, the tempera-
tures on the axis of the prism in its final state depend very largely on the
ratio of the length of the axis of the prism to that of a diagonal of a cross
* Despretz, Ann. de Chimie et de Physique, 1827. Peclet, Ann. de Cliimie et
de Physique, 1841. Tyndall, Phil. Mag., 1853. Hopkins, Phil. Trans., 18.57.
Pfaff, Pogg. Ann., CXIII., 1801. J. D. Forbes, Proo. Edin. Soc, IV. AngstWlni,
Pogg. Ann., CXIV., 1801. Neumann, Ann. de Chimie et de Piiysique, 1862. G.
Forbes, Proc. Edin. Soc, VIII., 1873. Herschell, Lebour, and Dunn, Rep. Brit.
Assoc, 1873. V. Beetz, Pogg. Ann. Jubelband, 1874. Smith and Knott, Proc
Edin. Soc, 1875. Lodge, Phil. Mag. 1878. Less, Journ. de Phys., VII., 1878.
Ayrton and Perry, Phil. Mag., 1878. H. F. Weber, Vierteljahrschrift d. Ziiriclier
Naturf. Ges., 1879. Thoulet, Comptes Rendus, 1882. Lagarde, Comptes Rendus,
1882. V. Littrow, Wien. Per., LXXI. Stefan, Carl's Rep., XIII. Jannettaz,
Comptes Rendus, 1884. Tuchschmid, Beibliitter z. Wied. Ann., 1884. M. Ballo,
Dingler's Journ., 1886. H. Meyer, Wied. Ann., 1888. K. Jamagawa, Beibliitter
z. Wied. Ann., 1889. G. Stadler, Inaug. Diss., Berne, 1889. Venske, Gilttinger
Nachrichten, 1891. Grassi, Atti 1st. Napoli, 1892. Lees, Phil. Trans., 1892. R.
Weber, Bull. Soc. Science Nat. Ncuch., 1895. Lord Kelvin and Mr. Murray, Proc
Royal Soc, 1895. Peirce and Willson, American Journal of Science, 189-5. Lees
and Chorlton, Phil. Mag., 189G. Oddone, Rend. R. Ace d. Lincei, 1897. W. Voigt,
Wied. Ann., 1898. Lees, Proc. Royal Society, 1898.
4 PROCEEDINCJS OP THE AMERICAN ACADEMY.
section ; and, if this ratio be small enough, the temperature conditions to
which the sides are subjected are of slight importance. For instance,
the temperatures at points on the axis of a relatively thin disk, one face
of which is kept at 0" C and the other at 100° C, are not measur-
ably different, whether the curved surface is kept at 0° Cor 100° C,
from the temperatures at corresponding points on the axis of an infinite
disk of the same thickness, the faces of which are kept at 0° C. and
100° C. respectively.
On the other hand, if the temperature gradient on the side faces could
be made to follow the proper law, — or even if, for moderate values of
Fo — I'l, it could be kept constant, — the temperatures on the axis of the
prism would be much the same, whether the prism were slender or stout.
In view of the extreme difllculty of controlling, or even of measuring
with accuracy, the temperatures on the side faces of a prism, it seemed
to us desirable to determine beforehand, as accurately as we could from
theoretical considerations, under each of a number of different assump-
tions with respect to the side temperatures, how short a prism of given
cross section must be in order that the temperatures on its axis, in the
case mentioned above, might be sensibly the same as if its cross section
were infinite in area.
We shall find it convenient to write down at the beginning of our dis-
cussion some of the common equations* of the theory of heat conduction
in the forms which we shall need to use later on. If 6 represents
the temperature at -the time t at any point, J*, in an isotropic solid, the
rate of flow of heat at this time, at J\ in any direction, is usually assumed
to be the product of a scalar point function, /, and the negative of the
space derivative, taken at 7* in the given direction, of a certain function
of the temperature, /■ (6). If, therefore, u, v, and w are the components,
parallel to three mutually perpendicular co-ordinate axes, of the vector,
//, which represents the flow within the solid,
„_ . 9/(6) _ , 9 6
v = -/.f(6).l^^, w = -.'./' {6).^^^. (1)
* Fourier, Theorie Analytique de la Chaleur. Poisson, Theorie Matlie'niatique
de la Chaleur. Lame, Lei.ons sur la Theorie Analytique de la Chaleur. Kelvin,
Article " Heat " in the Encj'clopanlia Britanniea. Kelland, Brit. Assoc. TJcp , 1811.
Preston, Theory of Heat, llieniann, Partielle Differentialgleichungen.
PEIRCE AND WILLSON. — THERMAL CONDUCTIVITIES. 5
If t) v, t are aiuilytic point fuuctious which define a system of orthog-
onal curvilinear co-ordiniites, and h^, h,^, h^ are the gradients of these
functions, and if q^, 9,,, q^ are the coraponeuts of the heat flux taken at
every point normal to the surfaces of constant ^, 7/, ^ which pass ihrough
that point,
For a given material which would be homogeneous if it were at the
same temperature throughout, under given pressure conditions, k is as-
sumed to be constant, so tliut k' f (6) is a function of the temperature
only. This product is called the specific conductivity of the substance
under the given circumstances, and is denoted by F' (0) or by k. We
may write, therefore,
9 6 9F(0) 9F(6) 9F(0)
. ^ = 7^ — - , V = ^-^^ , w . —
dx d X d y d z
{-)
J 3F{6) . 9 Fid) , 9F{6) ^^,
If a closed analytic surface, aS", be drawn within the solid and if (4^, n),
(ij, n), (^, 11) represent the angles between the exterior normal to /S* at
any point on it and the directions at that point in which ^, »/, and l,
increase most rapidly, the flux of heat across ^from within outward may
be written
/'
{q^ cos (^, 11) + '7,, • COS {rj, 71) + q^ . cos {i, n) \d S. (5)
The surface integral, taken over S, of f^cos (^, «), where 6^^ is any
function which, with its space derivatives of the first order, is continuous
within and upon S, is equal to the volume integral, extended through the
U
9\ J J I
space enclosed by S, of /t^ . h^ . h^ . \''r, >'ij , so that the flux across
9$
S may be expressed by the integral
-^^^ ( 9^ ^~97, '^'^H )
If \j/ (0) is the specific heat per unit volume of the body under the
given pressure conditions, we may equate the expression just obtained to
6 PROCEEDINGS OP THE AMERICAN ACADEMY.
/ f / "A (^) • n^ • '^^ ■'■ J ^^^1 since this result is iudependeut of the form
of S and of the volume of the space euclosed by it, at every point within
the solid
90 h,.h,.h,\9(/]) 9(;'y) 9(,^)l .7.
5^ _ /<s . h, . kj I 9 f^ 9/(0) \
^^ 5^ _ //g . hr, . h^ f 9 fjtj_ 9F(0)
9t xpiO) y9^\hr,lH' 9
9_(k'k 9m\ 9(K^i 5m\\ (H)
9v\/is/ii- 9ri J^9i\hiK- 9^ J)' ^''
/Ji_
9t xp{b) \9^\Ihhi' 9$
LfA. l^M\+ 3 ( h^ 9_F(0)\'^
9v\hsf'i' 3-n )^9C\l'iK' H JI'
three different forms of the equation of continuity.
In Cartesian co-ordinates, this eciuation becomes
96 _ 1 (9 fK\9JW\,5 f>^m\ 9 f/^/(6)\^
^-^(e)\9x\ 9x )'^9y\ 9y ) ^ 9 z\ 9z Jj
^ v^i^C^). (10)
-^(6)
+-(^'(*)-ii)}- ^">
9
1
If the flow of heat within a solid is steady, -^ vanishes at every point,
q is a solenoidal vector, and the equation of continuity in terms of Cartesian
co-ordinates becomes
Or
PEIRCE AND WILLSON. THERMAL CONDUCTIVITIES. 7
It is usually assiiiued that i« continuous at the surface of separation
of two isotropic solids of different conductivities. If Hi and n.^ are nor-
mals at a [)oint of such a surface drawn into the first and second conduc-
tors respectively, aud if the flow of heat is steady,
9F,{6) 9 FAQ) ,, ^,. . 9A{e) , 9f,(6)
or Kj^^ [-K'iT^ — — 0. (14)
If the temperature differences within a body are comparatively slight,
we may often use Fourier's assumption and representy" (6^) approximately
by itself. As we shall need to compare the solutions of certain simple
problems in the steady flow of heat obtained on this hypothesis with the
corresponding solutions obtained ou the assumption that f (&) and 6 are
not identical, we may note certain facts in passing. It is easy to prove
by an elementary application of Green's Theorem that a function, F,
which is harmonic within a given closed surface S, and which upon two
given portions, Sy and ASg, of S has the constant values Ci and C2 respect-
ively, while at every other part of S its normal derivative is zero, is
determined by these conditions. If this function has been found, it is
easy to write down the unique function
F's^^5|-' F+^^i^^^-, (15)
which is harmonic within S, has the constant value C x on *S'i and the
constant value 0\ on S^^ and the normal derivative of which vanishes
at all points of S which do not belong to Sy oi' ^'i- T^'^ families of sur-
faces defined by the equations, F= constant, V = constant, are identi-
cal. If, therefore, two given portions of the surface of a solid isotropic
conductor in which there is a steady How of lieat be kept at constant
temperatures (Ci and G.) while there is no flow across the rest of its
surface, the function F, which on Fourier's hypothesis gives the tem-
peratures at all points within the solid, is connected with the function
K', which gives /(6') on the assumption that this is not identical with 6
itself, by means of the equation
Gy-C, ^ G,-G, ' ^ ^^
and the forms of the isothermal surfaces are inde23endent of the form of
the function f.
PROCEEDINGS OF THE AMERICAN ACADEMY.
Two harmonic functions can only have the same level surfaces when
one is a linear function of the other. If upon n given portions, Si, S^, S.^,
. . . S,„ of a given closed surface, S, Fhas the constant values Oi, (7i, C3,
. . . (7„, respectively, ami V the values, F {C^), F (C^), F (Gs), . . .
F(C„), while upon the remainder of S, if there is any, the normal deriva-
tives of V and V are zero, and if V and V are harmonic within S,
V caunot in general be expressed as a linear function of V, and, if « is
greater than 2, their level surfaces will not usually coincide. If n is 3,
the condition of coincidence is evidently
Ci F(00 1
C. F(C,) 1
C3 F(C,) 1
= 0.
(17)
If U has the constant values Ci, Co, C3 ; V the constant values
Ki, K2, A'a ; and W the constant values Zj, Z.j, L^ on Si, S.,, S^^, re-
spectively, if the normal derivatives of these functions are equal to zero
at every point of S not included in Si, S.,, or S3, and if all these func-
tions are harmonic within S, W can always be expressed uniciuely in the
form A(/+ UV-\- Z>, unless
Ci Ki 1
C\ K. 1
c. a; 1
= 0.
(18)
Before we were able to decide upon the forms and dimensions of our
apparatus and upon the manner in which it should be used, we found
it desirable to make some rather elaborate computations based on the
mathematical solutions of certain problems in heat conduction. In de-
scribing this work it will be convenient to state, first, some analytical
results to which we sliull afterwards give various piiysical interpretations.
We have purposely put these preliminary statements in purely mathe-
matical language lest they should seem to be narrower in their applica-
tions than they really are.
(1) ,The square bases of a rectangular parallelopiped of height /are 2a
long and 2 a broad. A function, V, harmonic within this parallelopiped,
has the constant value V\ at the lower base and the constant value F, at
the upper base. At every point of the other faces of the prism V satisfies
the equation
K^+/KF-F)=0, (19)
PEIRCE AND VVILLSON. — THERMAL CONDUCTIVITIES. 9
— 9 V
where Fis a constant, and ^ — represents the derivative of V taken in
d n
the direction of the exterior normal. If the origin of rectangular co-
ordinates be taken at the centre of the lower base while the axes of
X and y are parallel to the sides of this base, V is given by the equation
P ^ CO k =L 1X1
v_ '
p ^ CO a; =: CO
V +^(^j, • COS (n^ y) 2j'k • cos («^. x) O , (20)
where O represents the quantity
Here ni, ??2, ns, etc. are the successive roots of the equation
K n . tan {n d) ^ A,
and Ay,, h stands for the radical ^,j 2 _|_ ^j^/i, while Cj, c^, C3, etc. are the
coefficients of the successive terms in the development,
1 = d cos (?^l d) + Co, cos (wo 0) + Cg cos (??3 ^) +
so that t•^• = 4 sin (?^^^•a) -^ (2 nua + sin (2 rika)).
It is to be noticed that equation (20) would give, on Fourier's assump-
tions, the final temperatures within a homogeneous parallelopiped of spe-
cific internal conductivity k, and of external conductivity h, if the lower
base were kept at the constant temperature Vq and the upper base at the
constant temperature F„ while the sides were exposed to the atmosphere
at the temperature V. In this result the absolute dimensions of the
paralleIopi[)ed are inextricably involved with the value ot h / k.
(2) The square bases of a rectangular parallelopiped of height / are
2 a long and 2 a broad. A function, F, harmonic within this parallelo-
piped, has the constant value Vq at the lower square base, the constant
value Vi at the upper base, and the constant value F on the other faces
of the parallelopiped. If, then, the centre of the lower base be used as
origin of co-ordinates, with axes of x and y parallel to sides of the base,
Fis given by the equation
<'^^)-<^>*-(-)
10 PROCEEDINGS OF THE AMERICAN ACADEMY,
where $ represeuts the quantity
U V. - V) sinh {^/FT^') - ( n - V) sinh ( - <' - '>^^ ]
sioh ( 2^V/>2 + qA
and where /? and q are integers.
F is evidently the temperature on Fourier's hypothesis within the
parallelepiped, if its bases and sides are kept at the temperatures F^, F„
and V respectively, when the How is steady. In this case the specific
conductivity of the material of which the liomogeneous parallelepiped is
made does not aftect the temperatures within the solid, and the lelative,
not the absolute, dimensions of the parallelopiped are of importance. Tiie
interpretalion of the equation (21) when/(^) and 6 are assumed to be
different is obvious.
(3) A function V, which involves the time and the distance from the
9V 9V 9'V . .
co-ordmate plane z ■=. 0^ is contniuous, as are -7^ , -,^ , -?r-\ , in tne
at d z d z^
region It, bounded by the planes 2 = 0, z = I. Within R, V satisfies
9 V ,9- V
the equation -^^-~ = <r ^ ., • V vanishes when z ^ I, and has tlie
d t d z-
constant value J'^ when c = 0, whatever t is. If, when < = 0, F= Fo^(c)
for all points within R,
X[Vw + '-l]»i"^rfA]. (22)
If <^ (z) has the constant value c,
V= Fori-~^ + ^|(2c- l)[r'-sin'^+ vr'^'sin^ +
ir>sin^ + ...]-[U"-sin^+le--sinip + ...]M, (23)
where T=P /a'^TrK
Equation (23) would give, on Fourier's assumptions, the temperatures
at any time within a homogeneous infinite plane lamina of thickness I
initially at the uniform temperature c V^, if, from the time ^ = 0, one face
were kept at the constant temperature Fq and the other at the constant
temperature zero.
PEIRCE AND WILLSON. — THERMAL CONDUCTIVITIES. 11
(4) The radius of the hase of a riglit cylinder of revolution of length
Ms a. A function, V, harmonic within this cylinder, has the constant
value Fq on one (the lower) base, the constant value F, on the upper
base, and the conslant value Fon the convex surface. If, then, the axis
of the cylinder be used as axis of z with origin at the centre of the lower
base, Fis given by the equation
• x^ . /i {Xp) . sinh ( '^-~ \
(24)
where J^ and J^ represent Bessel's P'unctiou of the zeroth and first
order, respectively, and x^, is the jt»th root in order of magnitude of the
ecjuation J^, (./) = 0. The first ten values of x for which the Bessel's
Fuiictiuu of the zeroth order vanishes have been given by Meissel.*
We have computed the next thirty values of the xj?, by the aid of
Stokes's Formula,! and the values of the Bessel's Function of the first
order corresponding to these forty a;^'s either from the series which usu-
ally defines J^ (x) or from the semi-convergent series. This computation
was done by means of Vega's ten place table of logarithms,^ except in
the few cases where a greater number of places was necessary, and for
these we had recourse to Thoman's tables.§ All the values have been
checked by duplicate computation, and the first four values of Ji (x) by
comparison with Meissel's tables. The results of this work appear in
Table I. Table II. gives to seven places of decimals the values of the
Xj's from p = 4\ to /) = 65. The values of Kon the axis of the cylin-
der depend upon the corresponding values of the function
S
* Meissel, Matli. Abliandhiiigen der k. Akad. der Wissenscliaften zu Berlin,
1888.
t Stokes, Camb. Phil. Trans., IX. Lommel, Studien iiber die Bessel'schen
Functionen, Leipzig, 1868. Hayleigb, The Tiieory of Sound, London, 1878.
Byerly, Treatise on Fourier's Series, etc., Boston, 1893. Gray and Matiiews, Be»-
sel Functions and their Applications to IMiysics.
X Vega, Thesaurus Logarithniorum Completus, Lipsiae, 1794.
§ Thoman, Tables de Logarithmes a 27 De'cimales pourles Calculs de Precision,
Paris, 1867.
12
PROCEEDINGS OF THE AMERICAN ACADEMY.
p-^^
^ ^
J^
I
c^
H-1
S
o
oo O' 'C ci t^ Ci ^^ c^i O ao
O l^ CO 1^ t^ C' CO ^ r- ' (M
O"— iCi ■^^^cO'— i-^^i^o
<M CC O CO O ''■O t^ t^ CO QO
• O-— iCOOiOCOCOCO'NGO
— I CO oo o r-, 1^ CO o 'X' lO
i>- >o ■<*' CO CO f^^ ">< "^ "" •— '
Cl fM O* CO --H GO t^ O iQ O
OC0OC0Ci-^l>.«0-^.-^
(^^--tlr-i^COOJlOOCSCO
CO CO '^ (M C.' l^ CO 1^ oc >o
l^ GO: O '^ GO ^f — CO O >0
CO ^ O QO '^r; 'O '-t^ oi r— O'
r-,_,-,0'C;'00000
O C5 Ci o
Ci Ci Ci C; Ci
Ci C: C^ C3 C: Ci O C C5 Ci
3.— lOcofMociio — 00 -*coocooco-*o
OC0C0G0-*C0G0>O<MCi Oi^OiO^fM — — 'GC■
■ OJ O^ O CO 'C ^^ — >0 o -^ C5 X' r- as — !>.
-^ -^ 'O 1-- Ol 1^ -^ '^ CM CO O CO (7^ -<!t C5 CC'
^- C^l O |-^ 00 — <M "^ 1^ ^- CO -^ l>. CO o o
1^ CO O GO' l^ CO lO '+ CO CO 04 fM ^ — O O
CMtMOM^^i— l^'—i'— I ^Hf— Ir-c^r-,— .1— II— I
■ O CO
03 CM
CO '—
o o
o o
+ I
oooooooo
+ I + I + I + I
C O O (
+ I +
o o
+ I
o o
+ I
+ I
OO^^OOCOfMt^COCOt>.
COCOCOCO^^COGO-^O'O
l^'TMCOTMOCJl^'^OO
i-O Ol O^ CO 'O t^ -— O l^ CO
CO 'O CO C: CO CO -^ CO Ci (M
CO -r O I- CO GO i^ -r 'M ^
O C '>} 'O O C3 'O >o •>! i»
^— ,1^ — -^^COCOCiCO
CO -+ CO 1^ i^ 'o 'M X CO a>
C0l>-C50'— iCMCOCO'*'^
d O O r-; .-I r^ r-J 1-; r^ r-i
CO
CO —1 OO CO
oD a f^i i>\
-p
'tl
GO X (M CO
1^ CO 't l^
o
>0 1^ 'O CO
~j CC -C X
1-
Ci
•O X CO — 1
• O ^M O l^
o
>o
1^ CO ^ i^
-ri -f "O CO
o
oj o X CO
>o X CO CO'
CO
CO
OM CO -t> c;
^ rt oo ^
1^
a.
l^ 'M >0 O
-* ^ CO o
fM
'M
CO O CO CO
Ci O-l "* |->.
<n
'O
iQ CO CO CO
CO i^ i^ i^
I-
t^ CO ~ ^H CO OS C3 X O -ti
1^ o CM c^ X 1^ ci c; ■» X
iQ _( _ CO ~ 'O Ol CO CO CO'
cr. -1" i^ ~ CO lo — ' -*
1^ CO CO — C3 CO
^ CO CO 1^ 1^ O
o c; CO -r -H CO
C' CO — — — — Ol CO -f
CO I- '— 'Q 05 CO
C^ .^^ ^^ CO -rf ^^
^ X —<■ -f i^" d
T-i .— I 1— I CM C^ (M CO
lO
.— 1
c;
-1"
'O
•r
1 ^
^
"M
1^
C^l
CO
X
1^
'O
'+
^'
CO
f— 1
o
Ol
IQ
n
-^
'O
CO
t^
CM
»o
X
.— 1
CO t^ CO CM !>. 'i*!
CO CO -* CO >— I ^>•
— o CO — . 1^ ci
c^i CO 1^ t^ CO X
O X 'C ^ X ~.
c^i CTi ■>! cv X c:
■X' O -f 1- -— I CO
'C 1^ X c:' ^ CM
i^ — >o cr. '-t* X
I- Ci O ^ CO '*"
CO* co' d CO CO Ci
CO CO ^ "^ ■^ "^
I— I O •— I C-1
r— lO CO O
'jf 'C c^i c;
X' I- 3-^ T— (
^^ O oo Oj
lO — X CO
'O C: ^
"* 'O CO' X
01 CO C '^^
CO i^ <r. c;
c^i 'O X' c^i
O lO iC CO
r-CMeO-^'OCOl^-XOsO ^CMCO-<*iOCOt>.XC-. o
PEIRCE AND WILLSON.
THERMAL CONDUCTIVITIES.
13
H
^
^ ^
-IS
ss
'^^
o
. Si,
—1 "O -*
CO
>0 'O "M CD fM
CO
CO CO 'O -M O ^ 'O t-^ CO CO
«
Ci) lO T
CO
l^ CD CO CO CO
,-H
(T^COCrjOO-Ht^-CDCDCiCO
^ — ^
>c CO t^
CO
t^ t^ O CO -f
(M
-:*< Ci 'O -M CO (>) eO O (^l Ci
^
^X) "-0 X)
'O
'O C7i t^ l^ o
CO
-*'!t<l^'7^1C^t>-I>-Ci'>»CO
-fH -ti -r
'O
CD t^ C5 — 1 ^
CD
C5 (M 'O Ci fM CD O '* C5 CO
4H
Ci -jo I^
'■Ci
'O -f CO CO ^
r— 1
O w CTj oo CO !» t^ CD 'O >0
en ci o
Ci
'^j C^ O C^ Ci
C5
C5 C5 00 00 CO 'CO 'X 00 'X) CO
be
o
GO yi 00
■:o
go' 00 CO GO oo"
GO
odGo'oooo'oocdcoGdcoGo
'O -H C5
1-H
O O CO ^ CO
CO
O>0OC5C0C0^OiCC0
'O o t^
'O
— 1 oo GO '>J Ci
'X)
t^ r-^ O GO ^ (M GO t-^ ^ (^
iTi •^ rjo
o
'O O CD -^ ^
I-H
GOO— icDcrioococoiocs
■^~>
(>j -n i^
l^
CO rt CO lo 'O
CO
l^CiCDXJOCO'CCD— iC5
^
00 'O CO
CO
-H l^ O 'f 'Cr^
'O
r-iCOCD'-tiCO<M<>J'MCOCO
'X> 'O -H
?>)
O 00 t^ 'O CO
(Tl
1— icriX)i^coiio-*co'>t'— 1
O O Cj
en
C5 X) CO 'GO 'X<
00
CO t^ i^ 1^ i^ t^ 1^ !>. t^ r^
^
o o o
o
o o o o o
o
O O O O O O O O O 'O
o o d
d
d d d d d
d
c:'i5':5c5cidi<5d> d d
+ 1 +
1
+ 1 + 1 +
■ 1
+ 1 + 1 + 1 + 1 + 1
•o -+I r^
CO CO CO ^ CO t^
'O
-f-*t>.-tl^C5(MC5'MCO
CO -:> O
CO
'^< O >0 -M 'O
l^
CI CO -f 'X' 00 — ' 00 t^ (M CO
-t< lO Oi
CO
>o o -H o r^
'O
O 'O O t^ CD 'M 00 'X '^ 'X
■•."■
1^ t^ oc
-rj
O -M CO O CO
o
G0OC0-*'0'^4'>lCDin'*
o o — .
CO
_t^ lO ^ o -t
CO
O Cl CD GO 'Tl 'O >— 1 O 'O O
to
CO Ol -f
T-J
CO CO '30 ^1 Ci
-t<
OCiCOfMCOCO^I^'TICi
o
— . O r-l
CO
t^ CI -+ -t< t^
CO
OXCO-*— i'*^0-f-+
h-;
-t< -r -+
^1
O 1^ -t* O 'O
■o
'O 'X rM >0 GO O ^1 -f 'C CD
^ CO 'O
l^
Ci O "M ^ >0
1^
C» Ci -^ (M CO 'O CD t^ GO C5
CO 00 CO
'3D
I-H
'7D C5 O Ci Ci
Ci
ci crs o <o o q o 'O 3 o
T^l C5 '^
r^
-t^ --^ Ci CO C<1
C5
o r^ o-i 'O CO >o t>- C5 o (M
O CI CO
CO
O ^ CO CO t^
^^
— - 'X -^ CD O O CO O GO CO
O CI o
-V
CO t^ 1^ -t" CO
00
'O 'O 'O ^ ^ GO 'O CO oo rt
00 CO <X)
CO
CD GO 1^ 'CO CD
l^
Ci 'X 'M C5 CO 'M CD CO O Ci
^^ en ^^
o
>0 'O O CTi 'M
x>
I^Cl^OCiO-MCOfMCO
CO CD CO
o
fM 'O oi C-l l^
t-H
CO — . t^ CO GO 'O — 1 t^ -* O
'VT
Oi -i< Ci
'O
O 'O O CO ^
t^
fM CO CO CI -t O CO -H t^ CO
O ^ 'M
-t<
CD t^ O O f>J
CO
>0 CD 00 O ^ CO -+' CO l^ Ci
CO CO t^
1-^
lO C5 CO CO 'M
CO
o -r CO 'T<i t^ r-H >o c:. CO t^
1-H CO -t<
o
t^ 'Xi O •— 1 CO
-)H
CO t^ GO q .-H CO -i^_ >o i^ CO
«6 oo' -^
-^
r-^ d -r t~>I d
CO
CO ci o-i CO ci T^i 'O GO ^ -TtH
CO CO i^
t^
t^ <X> CC CC Oi
Ci
CSCiOOO'— i"— ii— lO^S^I
^
1— 1 <N CO
'*
IJO CO t^ CO C5
o
^(MCO^'OCDI^COOO
■^
c^l f^ >>\
ITI
'M C-l C^l (M (>J
CO
CO CO CO CO CO CO CO CO CO '^
14
PROCEEDINGS OF THE AMERICAN ACADEMY.
TABLE II.
The /)th Root in Order of Magnitude of the Equation Jq (x)
is denoted by x^.
V
^P
Loga-^
I^ogV^
41
128.02087701
2.10728080
8.8482997
42
131.16244628
2.11780951
8.8430353
43
134.30401664
2.12808900
8.8378956
44
137.44558802
2.13823080
8.8328247
45
140.58716035
2.14794566
8.8279672
46
143.72873357
2.15754360
8.8231683
47
146.87030763
2.16693400
8.8184731
48
150.01188216
2.17617471
8.8138527
49
153.15345802
2.18512681
8.8093767
50
156.29503427
2.19394518
8.8019675
51
159.43601116
2.20258642
8.8006469
52
162.57818867
2.21106228
8.7961 0.S9
53
165.7i;»76675
2.21937431
8.7922529
54
168.86134537
2.22753025
8.7881749
55
172.00292150
2.23553583
8.7841721
56
175.14450412
2.24339651
8.7802418
57
178.28608520
2.25111745
8.7763813
58
181.42766171
2.25870351
8.7725883
59
184.56924564
2.26615934
8.7688604
60
187.71082696
2.27348932
8.7651954
61
190.85210865
2.28069765
8.7615912
62
193.99399070
2.28778828
8.7580459
63
197.13557308
2.29476500
8.7545576
64
200.27715580
2.30163142
8.7511244
65
203.41873881
2.30838096
8.7477496
and these latter we have computed for certain values of 2 / / and a \ I
by the help of Gudermaiin's tables.* The results appear in Table III.
To avoid possible errors arising from combining so many <pjaiitities, we
generally used seven places, although the time required for the com-
putation, which was done in duplicate, was thereby increased by some
weeks.
* Gnderniann, Tlieorie der Potenziiil oder Cykliseli-liypcrbolisclien Functinnen,
Berlin, 1833. Willson and Peirce, Bulletin of tlie American Mathematical Society,
1897.
PEIRCE AND WILLSON. — THERMAL CONDUCTIVITIES.
15
-s
TABLE III.
[ -'""(¥)
[^„/,(,r,)sinh(^)
z =
z = \l
. = 1/
z=-,l
z = l
n = \ I
O.OOOG
0.0065
0.0703
0.5
a = 1 I
0.0196
0.0697
0.2116
0.5
a^^l
0.0;358
0.1427
0.2908
0.5
a = I
0.0H57
0.1920
0.3320
0.5
n=%l
0.1144
0.2349
0.3642
0.5
a = 2 /
0.1224
0.2464
0.3724
0.5
<, = ?jI
0.1249
0.249S
0.3748
0.5
a = bl
0.1250
0.2500
0.3750
0.5
We shall wish to base an argument upon the values of S given in
the last line of Table III., and upon certain corresponding values of the
quantity
T= y -^^ — ^. (25)
We print, therefore, in Tables V. and VI., the numerical values of
the terms of the series which define these functions in the cases in
question.
It is evident that the three values of ^S* are in reality less than 0.125,
0.250, 0.375, respectively, though by quantities far too small to appear in
our results. Unavoidable errors introduced by adding together, in some
instances, hundreds of numbers determined by logarithms, make the last
figures given doubtful. Although our computations were made through-
out with the aid of seven place and ten place tables, we have contented
ourselves with four places in tabulating the values of T. It is interesting
16
PROCEEDINGS OP THE AMERICAN ACADEMY.
to notice the seemingly anomalous sequence of values in the terms of
the series for T. In fact, the relations between the successive terms
is, for some cases that we have studied, so complicated that the detec-
tion of accidental errors of computation becomes extremely difficult,
y — when r ^= a^ whatever s is, and equation (24) can be written in
the form
V= F(l - 2 n - 2 r,_,) + 2 F, n + 2 To r,_,.
TABLE IV.
all
'.//
2 S,
1-2 Si_,
i+^^_,^,_^
\-S-Si-.
\
1
4
.0012
.8595
.4303
.4291
\
1
2
.0130
.9870
.5000
.4870
1
3
.1405
.9988
.5695
.4291
2
1
4
.0395
.5768
.3080
.2688
jL
\
.1393
.8607
.5000
.3606
\
3
4
.4232
.9607
.6920
.2688
I
1
4
.1117
.4185
.2651
.1534
4
i
.2854
.7146
.5000
.2146
%
3
4
.5815
.8883
.7349
.1534
1
\
.1714
.3359
.2536
.0823
1
1
.3839
.6161
.5000
.1160
1
5
4
.6641
.8286
.7463
.0823
i
i
.2288
.2716
.2502
.0214
\
.4698
.5302
.5000
.0302
3
3
4
.7284
.77 1 2
.7498
.0214
2
JL
.2449
.2552
.2501
.0052
2
1
.4927
.5073
.5000
.0062
2
3
4
.7448
.7551
.7499
.0052
3
1
4
.2497
.2503
.2500
.0003
3
i
.4996
.5004
.5000
.0004
3
1
.7497
.7503
.7500
.0003
5
\
.2500
.2500
.2500
.0000
5
\
.5000
.5000
.5000
.0000
5
3
4
.7500
.7500
.7500
.0000
PEIUCE AND WILLSON, — THERMAL CONDUCTIVITIES.
IT
TABLE V.
s =
-'^ (^t)
Xp.Ji (xp) sinh
(^y
p
z-^yi
z = \l
z=ll
1
+0.1931944
+0.389186
+0.590810
2
-0.1108619
-0.230224
—0.367236
3
+ 0.0694976
+0.152211
+0.263868
4
-0.0434740
—0.102502
—0.198205
5
+0.0268426
+0.069353
+0.152345
6
-0.0163951
-0.047111
-0.118978
7
+0.0099432
+0.032159
+0.094069
8
-0.0060054
-0.022070
-0.075105
9
+0.0036196
+ 0.015227
+0.060435
10
—0.0021801
—0.010557
— 0.04.S940
11
+0.0013132
+ 0.007351
+ 0.039837
12
-0.0007914
-0.005139
—0.032567
13
+0.0004777
+0.003604
+0.026720
14
-0.0002886
-0.002536
-0.021990
15
+0.0001746
+0.001789
+0.018143
16
—0.0001057
—0.001264
-0.015007
17
+0.0000641
+0.000896
+0.012439
18
-0.0000389
—0.000635
—0.010325
19
+0.0000237
+0.000452
+0.008588
20
-0.0000144
-0.000321
-0.007150
21
+0.0000088
+0.000229
+0.005962
22
-0.0000053
—0.000163
—0.004975
23
+0.0000033
+0.000117
+ 0.004159
24
-0.0000020
-0.000083
—0.003479
25
+0.0000012
+0.000060
+0.002912
26
-0.0000007
—0.000043
-0.002441
27
+0.0000005
+0.000031
+0.002046
28
—0.0000002
—0.000022
-0.001717
29
+0.0000001
+0.000015
+0.001442
30
-0.0000001
—0.000011
—0.001211
31
• •
+0.000008
+0.001018
32
. .
—0.000006
—0.000856
33
.
+ 0.000004
+ 0.000721
34
. , .
—0.000003
—0.000607
35
.
+0.000002
+0.000511
36
• • •
—0.000002
-0.000430
VOL. XXXIV.
18
PROCEEDINGS OF THE AMERICAN ACADEMY.
TABLE V. — Conlinued.
P
\l
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
hi
+0.000001
-0.000001
0.1250000
3 I
4 *
0.250002
+0.000363
-0.000306
+0.000258
—0.000218
+0.000184
-0.000155
+0.000130
-0.000110
+0.000094
-0.000079
+0.000067
—0.000057
+0.000048
-0.000040
+0.000034
—0.000029
+ 0.000025
—0.000021
+0.000018
—0.000015
+0.000013
—0.000011
+0.000009
-0.000008
+0.000007
-0.000006
+0.000005
-0.000004
+0.000003
—0.000002
+0.000001
0.375004
PEIUCE AND WILLSON. THKIIMAL CONDUCTIVITIES.
19
TABLE VI.
7' =
^ Xp.Ji (a?p) sinh ^^ j
p
-^ = \l
z=^l
z=^,l
1
+0.182182
+0.367002
+0.557133
2
-0.079568
—0.165236
-0.263573
3
+0.026424
+0.057870
+0.100321
4
—0.001060
—0.002499
-0.004832
5
-0.006854
-0.017708
-0.038898
6
+0.006445
+0.018519
+ 0.046768
7
-0.003686
-0.011920
—0.034875
8
+0.001315
+0.004833
+0.016446
9
-0.000029
-0.000120
-0.000477
10
-0.000401
-0.001943
-0.009010
11
+0.000381
+0.002131
+ 0.011547
12
—0.000222
—0.001441
-0.009131
13
+0.000080
+0.000609
+0.004513
14
-0.000001
-0.000009
-0.000082
15
-0.000026
-0.000270
-0.002743
16
+0.000025
+0.000304
+0.003602
17
-0.000015
-0.000210
-0.002918
18
+0.000006
+0.000090
+0.001471
19
-0.000000
-0.000001
—0.000020
20
-0.000002
-0.000042
—0.000939
21
+0.000002
+0.000048
+0.001249
22
-0.000001
-0.000034
-0.001024
23
+0.000000
+0.000012
+0.000522
24
—0.000000
—0.000006
25
-0.000007
-0.000342
26
+0.000008
+ 0.000456
27
-0.000006
-0.000382
28
+0.000003
+0.000194
29
-0.000000
-0.000001
30
-0.000001
-0.000133
31
+0.000001
+0.000176
32
—0.000001
-0.000146
33
+0.000000
+0.000075
34
-0.000000
-0.000001
35
... .
-0.000051
36
+0.000069
20
PROCEEDINGS OF THE AMERICAN ACADEMY.
TABLE VI. — Continued.
p
z^ll
z=\l
z = %l
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
•
-0.000057
+0.000030
-0.000000
—0.000020
+0.000028
—0.000023
. +0.000012
—0.000000
—0.000008
+0.000011
—0.000009
+0.000005
-0.000000
—0.000003
+0.000005
—0.000004
+0.000002
—0.000000
—0.000001
+ 0.000002
-0.000002
+ 0.000001
-0.000000
-0.000000
+0.1250-
+0.2500-
+0.3749+
(5) The radius of the base of a right cylinder of revolution of height /
is a. The centre of the lower base is used as the origin of a system of
columnar co-ordinates (r, B, z), the axis of tiie cylinder being the axis of
2. A function F, which is continuous everywhere within the cylinder,
has the value zero on the curved surface and on the lower base, and the
constant value F; on the upper base. The planes 2 = /', z =^ I", divide
the cylinder into three portions (1), (2), and (3), in which V is repre-
sented analytically by three functions, Vi, V^, V3, respectively. If, when
z = l',
.^..
and when z = I", ko
9V.
h
9 Vo
, -^r—^ , ana wnen c = < , «<,
' d z a z o'
hi, k^, ^'3 are given constants, Vi, V.i, V3 are given by the equations
9V,
where
PEIRCE AND WILLSON. — THERMAL CONDUCTIVITIES.
/( = 00
+ B^ cosh { ^
a
where ^4^, jL, A3, B^, and ^3 are subject to the conditions
A, sinh f'^"J-\ = A, sinh ("^^ + i?2 cosh f"^) ,
21
(26)
(27)
(28)
(29)
k, A, cosh f^^ =: L Fj^ cosh ("^^ + i52 sinh (^')] ,
J^siuh (^^ + ^2C0sh C'^'^^ =^3 sinh (^') + i^3C0sh (^) ,
h r^ocosh (^) + ^2 sinh (^)l
J^cosh r^^ + ^3 sinh ('^)
^3 sinh ( — ) + i?3 cosh
k
xj.
2Vi
and where Xp is the p\\\ root in order of magnitude of the Bessel's Equa-
tion /u {x) = 0.
If, for brevity, we denote the quantities
sinh f '' " j , cosli ' '' '
-'), .„u(^-^'), ...(^.•/),
\ a ) \ « / x^.J^ {x^
by s, c, s', c', s", c", and O, respectively, these equations of condition
may be written
22 PROCEEDINGS OP THE AMERICAN ACADEMY.
Ai s' = A^s' + B^ c, Jc^ Ai c' — kz {A^ c' + B^ s'),
A^ s" + B., c" = As s" + Bs c", yC-2 A^c" + k^B^ s" ^ k^ A^ c" + t, B^ s",
AsS + BsC = il .
The determinant of the coefficients of the s's and c's may be reduced
to the form
c's'iki-k^) kic'^-k^s'^
c s" s c" — c s" c c"
k^ c c" ks {s s" — c c") kz c s"
(30)
and
A, = (31)
— ko Ag c n
c's'{ICi—k2) \ k-2CS"{sc"—CS")+k^Cc"(cc"S!i") j ■j-(kiC"^k<^'^) { Cs"/g(.ss"-Cf ")+^2Cc"(s"c-Ac") I
If in the special case where ^i and k^ are equal, we write ki = /x k^ = k.^ ,
we get
A, = (32)
-Mfl
c's'(fi. — l)\s"(sc" — cs") + fxc" {cc" — ss")\ + {fjLc''^ — s"^) \fj.s"(ss"—cc")+c"{s"c — sc")\
with corresponding values for the other coefficients.
We shall need at the outset only two or three applications of the fore-
going theory. We may ask first what must be the relative dimensions of a
homogeneous regular right prism, one end of which is kept at the uniform
temperature ^o ^I'J the other end at the uniform temperature 6„ while
its other faces are kept at some uniform temperature 6, between 6q and 61,
in order that the tem{)eratures on the axis of the prism in the final state
shall be sensibly the same whatever value has. If, for instance, the
difference between 60 and 61 is 100° C, what must be the ratio of the
radius a of the circumference inscribed in a right section of the prism to
the height I of the prism, in order that the temperature of every point on
the axis may be the same within less than 0°.01 C, whether 6 is ecjual
to 60 or to 61 ? Since we need merely to find a lower limit for a -^ /,
we shall do well to substitute for the prism the inscribed right cylinder
of revolution, and then apply the solution of Problem 4 given above.
We are to find a function of r and z, harmonic for values of r between
and a and values of z between and I, which (1) has the uniform
value F (Oo) when z — 0, whatever ?• is ; (2) has the uniform value F (0,)
PEIRCE AND WILLSON. THERMAL CONDUCTIVITIES. 23
when z ~ I, whatever r is ; and (3) has the uniform value F (0) when
r =^ a, whatever z is. The value of this function F (6) is evidently
F(e) . (1 - 2 T,_^- 2 T,) + 2F(eo) . r,_,+ 2 F(e,) . T;, (33)
or, for points on the axis,
F{e) {1 - 2 S,_^- 2 5J + 2 F(eo) . St_,+ 2F(e,) . S^; (34)
that is,
F{e)-F{e,) = 2 s,[F(e,) - f(0o)-] + [F(e)-F(Oo):\ (i - 2 5,_,- 2 *s;).
(35)
z I
In the case of an infinite lamina, where - = , - = ,
a a
F(6) = F (6,) + ~ (F (6,) - F {6,)). (36)
The difference between the values, at any point, of F {0) in the case of
the infinite lamina and in the case a = 5 I, is
[F{ei) - i^(6o)] [2 '^. - 7] + [^ W - ^ W] [1 - 2 S,_,- 2 S,-] .
It is easy to prove that for given values of? and a, 1 — 2 Si_^ — 2 S^
has its greatest value when z = ^ /, and if a -^ / is as great as 5, it is clear
from Table V. that neither 1 — 2 Si_^— 2 S^ nor [2 aS^ — j) can for any
point of the axis be nearly so great as 0.00001, so that whatever is, the
value of F (0) is surely equal, within less than one ten-thousandth part
of the greater of the quantities F (0,) — i^((9o), F(d) — F {6^), to the
value which it would have at the same point on the axis if the disk were
infinite. By exactly what amount the temperatures themselves would
differ in the two cases cannot be stated unless we know something of the
nature of the function F.
For certain substances, experiment seems to show that within wide
limits F {d) can be expressed as a linear function of 6, as Fourier as-
sumed. In the case of any one of these substances we may say, for
example, that the final temperature at a point on the axis of a disk the
radius of which is at least five times its thickness, if one face is kept at
100° C. and the other at 0° C, cannot be changed by nearly so much as
0°.01 C. by altering the temperature of the edge of the disk from 0° C.
24 PROCEEDINGS OF THE AMERICAN ACADEMY.
to 100° C. The effect of radiation or conduction from the edge is
therefore of no consequence.
Most experimenters have been able to reproduce mathematically the
results of their work on thermal conductivities by assuming that in every
case the conductivity, /c, is a linear function of 6, say k (1 + 2 hO), where
b is small (usually less than .OOo), so that F {B) — C + k 6 {} + h6).
On this assumption the temperatures within an infinite disk would be
given by the equation,
k' ^ (1 + ^» e) = ^ {0, (1 -\-he,)-eo{\ + b 6,)) + K e,{\ + b o,), (37)
or b (0- - 0,') + {B- Bo) = ^- {Oi -Bo+b {Bi" - 6„-)}.
Except in instances where near certain temperatures some great chem-
ical or physical changes take place in the materials concerned, experiment
appears to show that k always changes slowly with the temperature, and,
whether or not we know the exact nature of the connection between the
two, it is easy to get a superior limit for the effect on the final tempera-
tures at points on the axis of such a disk as has just been described, of
changes in the edge temperatures. Neither in our own experience nor
in any published reports that have come to our notice have we found any
substance in which the change of k with 6 is so rapid that in a disk,
where a > 5 I, made of it. with its faces kept at 0° C. and 100° C.
respectively, the final temperatures of points on the axis could be affected
by nearly so much as 0°.01 C. by changing the edge temperature fiom
0° C. to 100° C. We are here concerned merely with the magnitude of
a possible error, and in every case to which we need to apply our theory
we shall be well within bounds if we assume that the error is not greater
than twice the error which would be found if 6 and f(B) were identical,
as Fourier assumed them to be. We have, therefore, tabulated for a
numerical example the final temperatures computed on Fourier's hypoth-
esis at several points on the axis of a disk of radius a and length /, when
one face (z =^ 0) is kept at the uniform temperature 0° C. and the other
face (z = I) at the uniform temperature 100° C. on two or three different
assumptions with respect to the edge temperatures. If the face temper-
atures are B^ and Bf, and if the temperature has the same value, B, at all
points of the edge, the final axial temperatures are given by the equation
e = B(l -2S^-2S,_;) -\-2BoS,_,+ 2B,S„
PEIRCE AND WILLSON.
THERMAL CONDUCTIVITIES.
25
and from this expression, with the help of the numbers in the body of
Table IV., many special problems can be solved with very little labor.
The expression
A (1 _ 2 7; - 2 ?;_,) + 2 B Ti._,+ 2di.T^+ (0, - 5) (l - ^)
gives the final temperatures in a homogeneous disk of radius a and
height I, one face {z =. 0) of which is kept at the uniform temperature
^0, the other face (s = /) at the uniform temperature 0i, and the rim at
constant temperatures given by the law A -\- (0(i — B) il — ,) . From
this we may see, that, with a very rude approximation to a uniform
gradient in the temperatures of the edge of a disk of relatively large
thickness, the final temj^eratures on the axis are sensibly the same as for
an infinite disk of the same thickness.
100
Figure 1.
26
PROCEEDINGS OP THE AMERICAN ACADEMY.
Some of the results given iu the first column of Table VII., with some
others, are represented graphically iu Figures 1 and 2. In Figure 1 the
ordinates ai'e the final axial temperatures ; the abscissas, the distances from
the cold face of the slab. The straight line corresponds to an infinite
slab ; the other curves, in order, to disks where a = I, a =^ ^ I, a — ^ I,
and a = ll, respectively. In Figure 2, the ordinates of the three curves
u ^ , , . . . i I 3;
are the final temperatures on the axis at the points 2=-, z =^ - . s = — ,
respectively, and the abscissas are the values of a, each horizontal space
corresponding to a change in a of ^ I.
TABLE Vn.
Final Axial Temperatures in a homogeneous Disk of Radius a and
Thickness I, when one Face (c = 0) is kept at 100° C.,the other Face
(c = /) at 0° C, and the Edge at the uniform Temperature 0.
a/l
z/l
^ = 0°
6 = 100°
^=50°
I
i
14.05
99.88
56.95
i
h
1.30
98.70
50.00
i
f
0.12
85.95
43.03
h
\
42.32
96.07
69.20
1
'2
1
2
13.93
86.07
50.00
JL
.1
4
3.95
57.68
30.80
4
\
58.15
88.83
73.49
1
h
28.54
71.46
50.00
1
3
4
11.17
41..S5
26.51
1
1
4
GG.41
82.86
74.63
1
JL
38.39
61.61
50.00
1
1
17.14
33.59
25.36
i
I
72.84
77.12
74.98
1
i
46.98
53.02
50.00
i
22.88
27.16
25.02
2
1
5
74.48
75.51
74.99
2
h
49.27
50.73
50.00
2
■A
4
24.49
25.52
25.01
3
\
74.97
75.03
75.00
3
\
49.96
50.04
50.00
3
3
4
24.97
25.03
25.00
5
i
75.00
75.00
75.00
5
h
50.00
50.00
50.00
5
A
4
25.00
25.00
25.00
PEIRCE AND WILLSON. — THERMAL CONDUCTIVITIES. 27
100°
80°
60°
40^
20'
i
^
.^
/
/
/
/
/
,/"
"^^
/
/
/
/
,y^
,
,
'/
x*
u
/
o
2/
Figure 2.
3/
U
If one is to measure the quantity of heat that passes through a portion
of the disk, lying within a cylindrical surface of revolution of relatively
small radius co-axial with the disk, it is desirable to make the ratio of a
to I so large that possible changes in the edge temperatures shall not
sensibly affect the temperatures at any point within the portion in ques-
tion. It will be sufficient for our purpose to consider the temperatures
at a distance / from the axis in a homogeneous disk for which a = 5 /.
It is evident that the greatest effect of temperature changes on the edge
of the disk will apppear at those points on the inside cylindrical portion
nearest the edge, tliat is, farthest from the axis.
Taking the formula
e = ^ (1 - 2 n - 2 7;_,) + 2^0 • T'.-. + 2 ^, . t; ,
and using the values of T' given in Table VI., we see that, if Q^ = 100° C-
and $1 = 0° C, and, if the whole edtre is kept at the temperature 0° C.,
28 PROCKEDINGS OF THE AMERICAN ACADEMY.
the temperature at no point within the cylinder of radius I co-axial with
the disk differs by more than 0°.02 from the temperature at the corre-
sponding point in an infinite disk of the same tliickness and same face
temperatures. In practice there is always a gradual fall in edge temper-
atures as z increases from to /, and in such a case we may consider a
guard ring of width 4/ amply large- enough to make the final temperatures
within a right cylinder of radius / and thickness I sensibly ecjual to those
within an infinite slab of the same thickness and same face temperatures.
In our experimental work we have sometimes found it desirable to
introduce between two slabs of low conductivity a thin sheet of tinfoil of
comparatively very high conductivity. It is evident that under con-
ceivable conditions such a layer of metal might seriously affect the final
temperatures in the slabs near their common axis. To investigate the
disturbances that miglit arise from this cause, we may apply the solution
of Problem 5 given above to the extreme case where the uniform edge
temperature is equal to one of the face temperatures, and where ki — k^.
If we attempt to compute numerical values of the series
S-.-^y)- frO
p =
by using the expression for A^ given in equation (32), we shall find the
amount of labor involved enormous ; we will therefore change the form
of the expression so as to make the nature of its dependence upon the
dimensions of the cylinders and upon their conductivities more evident,
keeping in mind the fact that the ratio of ^o to ki is very large. If we
denote the denominator of the second member of (i32) by D, and write
«=sinh~^s', « -f 8 = sinh~ ^ s", and Mo^sinh^^s,
we have
Z> = S {(/. 5" 2 _ c" ') (fX C' "-s"-)-{l- /x)- C' S' C" S"}
+ c{\~fj.) {s" c" (/. c' 2 - ,' ■') - s' c' Ox c" ■' - /' ')}
= 1 s{(l _/x)2cosh 28 + (/x2 - 1 ) [cosh 2 (« + 8) - cosh 2 «] - (1 + /x)^}
— ^e(l— ;ut)-sinh28 + |c(l — /x2)[sinh 2(«-f 5)-sinli 2m]
= i {(1 - fxY smh («„ — 2 8) - (1 + fiY s\nh «„
— (1 — /x-)[sinh («o — 2« — 2 8) — sinh («o — 2(t)']},
PEIRCE AND WILLSON. — THERMAL CONDUCTIVITIES. 29
and Ai =
4jun
(I +/i)-sinliao— (1- ju)'"«>"'' ("o "'-^S) + (1 — ;U-)[sinh (ag -2a — 25) — sinli (ao — 2a)'
This expression, though still sufficiently complicated, shows that for
properly choseu cases, as good for our present purpose as any others, the
computation is comparatively simple.
If, for instance, the thickness of the lower slab is half that of the disk
formed of the two slabs and the intermediate sheet of metal, I' =z ^ I and
n(p = 2 u, so that
A ^ ^/^" ns^
' (l + yit)2sinh«o-(l-/x)2sinh(«o-2S)-(l-/i.'^)sinh2 8 '^ ^
O
. , ,, (1 — u)- . , ,_ (u— l)sinh'-S,, , ,, , . '.^ ^
sinhaf, 1 — — „ - sinh-5H — ; ^- 1 — cosh ao + i" (1 + cosh a,))]
Z/u 4jusinli a()
If we denote the denominator of this expression by sinh «o • (1 + ^)i ^^^
n
note that, if we make u e(\ua\ to unity, we shall have Ai = -^-, — , cor-
smh ixq
responding to the case of a homogeneous cylinder already treated in
Problem 4, we shall see that T'l in the case of the heterogeneous cylin-
der can be found by multi{)lying each term of the series for 7\ by the
quantity — ^— , and that in our problems the resulting series is usu-
ally more convergent than the original.
In order to exaggerate the magnitude of the disturbing effect of the
tinfoil, we have chosen for computation a value of /x much smaller and
a value of 8 much greater than the proper values of these quantities for
most of our experiments, assuming that a = 5^= 10Z' = 500 (I" — /'),
and that /x = 0.002, so that A is nearly equal to
^ Xj, tanh I «o - j^ ctnh ^ a, - ^^ , where «„ = -^
These values correspond in certain cases to large disturbances of temijcr-
ature on the axis of the slabs, as the results show. Consider, for instance,
the point z = /', r = 0, in a com|iound slab 2 cm. thick and 20 cm. in
diameter, built up of a slab of |)oorly conducting material 1 cm. thick, a
sheet of metal 0.2 mm. thi<'k, and a second slal) of the same material as
the first. Let the lower face be kept at the temperature 0° C, the other
face at the temperature 100'^ C, and let every point of the edge be kept
30 PROCEEDINGS OF THE AMERICAN ACADEMY.
at 0° C. The terms of the series which give the final temperature may
be found without much difficulty by the aid of the numbers in the third
column of Table V, Their values are
1. +61.3980
2. —19.6480
3. + 7.8876
4. - 3.7116
5. + 1.8474
6. - 1.0224
7. + 0.5938
8. — 0.3568
9. + 0.2198
10. - 0.1382
11. + 0.0882
12. - 0.0570
13. + 0.0372
14. - 0.0246
15. + 0.0162
16. — 0.0110
17. + 0.0074
18. — 0.0050
19. + 0.0034
20. - 0.0022
etc., so that the temperature required is 47°. 12+ C.
The terms of the series which give the final temperature at points for
which z = 1, r = 2, can be found in a similar way by the help of the
numbers in the third column of Table VI. Their values are
1.
+57.8982
2.
-14.1020
3.
+ 2.9988
4.
- 0.0904
5.
- 0.4716
6.
+ 0.4020
7.
— 0.2192
8.
+ 0.0782
9.
- 0.0018
10.
- 0.0254
11.
+ -0.0256
PEIRCE AND WILLSON. THERMAL CONDUCTIVITIES. 31
12.
—
0.0160
13.
+
0.0062
14.
—
0.0000
15.
—
0.0024
16.
+
0.0026
17.
—
0.0018
18.
+
0.0008
19.
—
0.0000
20.
—
0.0002
and the temperature is 46°.48.
If tlie radii of the slabs and the metal sheet had been infinite, the
temperature in these media would have been given by the expressions
Mz, M(fx c + 1 — fi), and M(z -\- ^\ (fi — 1)), respectively, where
31= 100/(1.98 + .02 /x). In all practical cases the temperatures of
points on the rim of the disk increase gradually from the cold face to the
warm face, and it would be easy to show that those portions of the
isothermal surfaces which we have used in computing the results of our
observations are sensibly plane.
The characteristic differential equation which gives the relation between
tlie temperature, the space co-ordinates, and the time in a body in which
there is an unsteady flow of heat, involves the specific heat of the body,
which is itself a function of tlie temperatui-e. Without attempting just
here to investigate the nearness of the approximation obtained in any
given case by assuming the specific heat to be constant, we will give for
future reference some numerical results obtained by using several differ-
ent values of z, t, and c in the solution of Problem 3.
An infinite homogeneous lamina of thickness I is originally at the
temperature c Vq throughout. From a given time, < = 0, one face is
kept at the constant temperature Fq, and the other face at the tempera-
ture 0°. The ratio of the conductivity of the slab to its specific beat is to
^be denoted by the constant d^, the ratio of /^ to a^ n"^ by T, and the
distance of any point in the lamina from the face which is kept at the
temperature Vq, by z.
The numbers in Table VITI. show the rate of flow across the cold face
of the lamina in fractional parts of the final rate for different values of
c and t, while the numbers in Table IX. give the rate of flow across
different planes parallel to the lamina faces at different times, for the
special case c = I.
32
PROCEEDINGS OP THE AMERICAN ACADEMY.
TABLE VIII.
t = \T
t=\T
t = \T
i= T
t=2T
^ = 47-
t=GT
c =
— 1
-5.013
-3.544
-2.434
-1.171
0.188
0.890
0.986
c =
-I
-2.506
-1.772
-1.199
-0.435
0.459
0.927
0.990
c =
1
4
-1.253
-0.886
-0.582
-0.067
0.595
0.945
0.993
c =
0.000
0.000
0.036
0.301
0.730
0.963
0.995
c =
1
4
1.253
0.886
0.654
0.669
0.865
0.982
0.998
c =
i
2.507
1.773
1.271
1.037
1.001
1 .000
1.000
c =
1
5.013
3.545
2.507
1.773
1.271
1.037
1.005
TABLE IX.
t = \T
t= \T
t= T
t = 2T
< = 47'
t = &T
z =
3.760
2.659
1.889
1.405
1.136
1.018
1.002
z = \l
2.762
2.279
1.756
1.366
1.125
1.017
1.002
^ = \l
1.095
1.438
1.420
1.260
1.096
1.013
1.002
Z=%1
0.235
0.682
1.030
1.115
1.051
1.007
1.001
z=.U
0.036
0.301
0.730
0.963
0.999
1.000
1 .000
z^ll
0.082
0.278
0.587
0.823
0.948
0.993
0.999
Z=%1
0.365
0.489
0.578
0.740
0.904
0.987
0.998
-r — Tl
- 8 '
0.922
0.761
0.627
0.686
0.876
0.983
0.998
Z= I
1.253
0.886
0.654
0.669
0.865
0.982
0.i)98
In Figure 3, the abscissas are the elapsed times (one division = \ T),
and the ordinate corresponding to any abscissa is the rate of flow of heat
at that instant across every unit of surfiice of the cold face of the lamina
PEIRCE AND WILLSON. — THERMAL CONDUCTIVITIES. 33
3A'
2K
-K
2K
\
\
\
\
\
\^
v^
^^
v
^ ^
-^"^
■
P
^
y^ y
y^
A
/,
/
/
/
/
V
' /
/
/
/
4 A'
2A'
2 A'
T 2 7' -iT \T
FlGUKlC 3
(J
VOL. XXXIV. 3
T 2 T
Figure 4.
3 T
34
PROCEEDINGS OF THE AMERICAN ACADEMY.
(one vertical divisiou = ^^kVo^ I). Every curve correspoufls to a par-
ticular value of c, and the values represented are 1, i, \, 0, — i —A —1
respectively. All the curves have, of course, the common asymptote,
y = kVq^ I— K, where K is the final rate of flow.
If Vo is to be 100° C, and the slab is to be originally at room temper-
atures, we may put c = J . The ordiuates of the curves in Figure 4
represent the flow of heat, when c — J, across the hot face, the cold face,
and the plane midway between them, at the times indicated by the
abscissas. The horizontal unit is ] T, the vertical unit J — ^ .
3I\
2K
In Figure 5, the abscissas are values of z, the ordinates are rates of
flow. Each curve corresponds to a given epoch, and the epochs repre-
sented are | T, ] T, h T, T, 2 T.
Without waiting to discuss here certain theoretical questions which
will present themselves in the course of our work, we may briefly
describe some preliminary experiments.
We have used two different forms of apparatus in our work, the one
intended for measuring the absolute thermal conductivities at tempera-
PEIRCE AND WILLSON. — THERMAL CONDUCTIVITIES.
35
tures betvveeu 0° C. and 100° C. of relatively poor conductors like plates
of stone or glass ; the otiier designed merely for comparing the con-
ductivities of slabs which form a prism or " wall," through which there
is a steady flow of heat.
Of this second form of apparatus, which is much simpler than the other,
we have three of different sizes for plates G5 cm., 35 cm., and 20 cm.
in diameter respectively. Rough diagrams which show the essential
parts of two of these, without their elaborate stands and jackets, are
given in Figures 6 and 7. In each, the prism to be tested is enclosed
Figure 6.
between the horizontal planed plates of two castings, which are fastened
firmly together by bolts around their edges to insure close contact with
the body under experiment. Both castings are hollow ; one forms a
jacketed chamber through which steam or mercury vapor may be passed
for an indefinite period. The upper casting, which is provided with a
system of stirrers or scrapers operated by an electric motor, may be
kept at a low temperature by filling it with ice or by sending through
it a steady stream of water from a very large tank within the tower of
the laboratory.
In Figure G, A represents the hot chamber, weighing about two hun-
dred kilograms, which rests in a thick jacket on a heavy table or stand
made to hold it. A is connected directly with one (^B) of two stout-
walled copper boilers, B and i?', each of which holds about 40 litres of
water. A light cup-shaped weight, inverted and laid on a large tube
with squared end which projects above the top of the boiler, acts as a
sensitive safety valve and prevents any appreciable rise in temperature
within the boiler. B can be refilled when necessary with boiling water
36 PROCEEDINGS OF THE AMERICAN ACADEMY.
from B' without stopping the coustaut flow of steam through A, by means
of the siphou f, which is provided with a valve. The steam, after pass-
ing through the hot chamber, is led to the outer air by a jacketed pipe
h, descending* from the bottom of A.
The connections of the thermal elements are led out of the sides of the
prism shut in by A and Z>, and are held between slabs of wood, which act
as a sort of guard-ring jacket to the prism for about 40 centimeters before
they emerge. The platinoid or German silver leads of these thermal
junctions within the prism are soldered together, and to a copper wire
leading to the (copper) wire of a potentiometer, p. The copper ends of
the couples lead to a mercury switch by which any one of them, or any
pair pitted against each other, may be (|uickly connected with a second
copper wire leading to the potentiometer. On its way from the switch
to the cold junctions in C through the potentiometer wire, the current
encounters only copper. By means of a somewhat elaborate standard
potentiometer, not shown in the diagram, the resistance, A', in the poten-
tiometer circuit can be so adjusted that every millimeter on the poten-
tiometer wire corresponds to an}' desired small ])Oteutial difference, such
as one microvolt or one tenth of a microvolt. Katlier tlian make this
adjustment many times a day to conform to the varying temperature of
the copper wire, however, we lind it better to determine the slight correc-
tions necessary to reduce the readings to absolute measure, by noting at
frequent intervals the indications of a standard thermal couple, the elec-
tromotive force of which is well known. The potentiometer wire, which
is 0.25 mm. in diameter, can be changed in a few seconds for new wire,
if the old should become dented or stretched.
Into the vessel D about 100 kilograms of cracked ice can be put, and
this ice can be kept in constant motion over the smooth bottom by help
of the electric motor, M.
Figure 7 shows a similar but smaller ai)paratus without its elaborate
system of inch thick asbestos jackets. Z? is a closed iron drum contain-
ing a rotary stirrer and rubber scraper turned by a motor. Through D
a large volume of water can be sent at a steady rate. The hot chamber
is the iron box, B, planed on its upper surface and communicating at the
bottom with a retort chamber, C, in whi(;h about 20 kilograms of mercury
can be kept boiling. The outlet at fallows the vapor to escape to the
tube^f, connecting with a large wrought iron chamber where it condenses
* In the diagram, A is erroneously represented as ascendinfj, and as inserted in
the side of A.
PEIRCE AND WILLSON. — TIIKRMAL CONDUCTIVITIES.
37
and flows back into the retort through the trap h. This apparatus takes
slabs 35 centimeters square. Although we found it possible to maintain
with this arrangement a temperature above 350° C, for many hours at a
time, it was difficult to avoid superheating by conduction through the
massive iron of the hot box, and we intend to discard mercury in future
and use some less troublesome source of heat. If a substance of greater
heat of vaporization than mercury is employed, the retort can be removed
to such a distance that all danger of superheating is removed. We have
not yet been al)le to test au electrical stove which we hope may prove
to be a convenient and a sufficiently constant source of heat for many
purposes.
Figure 7.
The apparatus just described has been furnished with trunnions so that
the axis of the prism can be made horizontal or vertical at pleasure.
This renders it possible to use a layer of mercury on each side of the
slab to be tested, when this is desirable.
Our third apparatus of this kind is made entirely of brass. It is in-
tended only for small thin plates about 20 cm. in diameter, but is in
essentials like the apparatus just described.
Figure 8 represents the apparatus which we have used to determine
the absolute conductivities at temperatures between 0° C. and 100° C. of
various materials. The boilers and the hot chamber are those of the
apparatus shown in Figure 6; the ice box, which is the outcome of
PROCEEDINGS OP THE AMERICAN ACADEMY.
PEIRCE AND WILLSON. — THKRMAL CONDUCTIVITIES,
39
several years of experimeutatiou, is entirely different. An iron casting,
Z, seen in plan in Figure 11 and in elevation in Figure 9, accurately
planed below and turned true above, is the bottom of the box. Be-
tween this casting (which can be bolted to ^ as Z) is in Figure 6) and
Figure
A is held the prism to be experimented on. While Z was in the lathe
a small hole, i/, about 3 millimeters in diameter and 4 millimeters deep
was drilled exactly in the centre of its upper face. Subsequently a piece
of solid drawn brass tube 12.3 era. in outside diameter and 13.5 cm. high,
with carefully squared ends, was held centrally in Z, by means of a
40 PROCEEDINGS OP THE AMERICAN ACADEMY.
wooden disk turned to fit it, and a central pin inserted in H, and was
then soldered firmly to Z. This was accomplished, after many trials of
other materials, by the use of white pitch as a flux, and the result left
nothing to be desired. The walls of the pot tlius formed were jacketed
on the outside, except for a height of about 2 millimeters at the bottom,
by an inch thick casting of hard rubber made for the purpose in the
form of a cylindrical shell. This casting, which was cut off square at
the tO}:) of the pot, tapered to nothing near the bottom, but did not rest
upon the floor. (Figure 10.) Upon the top of this jacket was fastened
a hard rubber cover shaped somewhat like a cylindrical hat. This had
an opening at the top which could be closed by an accurately fitting
rubber plug. In the box P, thus made, is placed a thin-walled ice
holder, ^>, open at top and bottom, of the same outside diameter below
as the inside of the brass pot, but somewhat smaller above, so as to leave
an air space between it and the walls of the pot.
In order that the holder may be easily rotated, a pin soldered to a
thin diametral web, F^ which runs across the bottom of the holder, is
inserted in H, and a vertical brass rod soldered to a similar web, E, at
the top of the holder passes through a hole in the corner of the pot
which it fits closely. A hard rubber thimble fitting tightly on the rod
and turning with it permits the slow entrance of cold air into the pot
without allowing any water to leak in. The rod can be clamped at
pleasure to a brass yoke which is turned l)y the motor. In order to pre-
vent the introduction of heat into the pot by conduction down the rod,
the exposed ])ortion is buried in cracked ice held in a thin m(!tallic cup
carried by the yoke and resting on it. When the holder is filled with
ice and is turned by the motor, the web at the bottom compels the ice
to rub over the floor of the casting, since the holder itself has no bottom,
and as a result of this, the lower surface of the ice ipiickly acfjuires and
keeps a mirror-like surface. The drip from the pot comes out of the
edge of the casting Z through a straight hole about 26 cm. long and
0.6 cm. in diameter drilled in the plate afld ending just inside the pot.
The whole apparatus is very slightly tilted so as to insure the steady
outflow of the drip.
A large cylinder, K, 35 cm. high, made of rolled brass 4 mm. thick
and open at the top and bottom, is mounted on brass ball bearings placed
on the outside of the hard rubber jacket of the pot P, by means of six
vanes, one of which, X, is shown in Figure 9. K weighs about 20 kilo-
grams when empty, and rests upon 144 brass balls each about 12 mm. in
diameter. When set in motion by a slight push, it continues to rotate
PP:mCE AND WILLSON. THERMAL CONDUCTIVITIES. 41
r\
Figure 10.
42
PROCEEDINGS OF THE AJVIERICAN ACADEMY.
for about a minute before coming to rest. This, like most of our other
apparatus, was constructed by Mr. G. W. Thompson, the mechanician
of the Jefferson Laboratory, and we have been much indebted to his skill
and patience at every stage of our work. A' is so truly hung that the
outside can be used as a pulley and the whole can be rotated by the use
of the belt shown in Figure 8. The vanes reach to about 2 millimeters
o
^^^^
o
o
f
.J
^
o
o
x
o^
Ji
Ir
o
\
^^.^..^^J'^^^
w
o
o
o
o
Figure 11.
of the floor of the box, and when the whole is filled with cracked ice and
then rotated, the ice at the bottom which rubs on Z soon gets and holds
a very smooth surface. A hole in the bottom of Z carries away the drip
and prevents any accumulation of water on the floor of the ice box. We
were at first troubled by irregularities arising from honeycombing of the
ice in the ice box, and to remedy this a suitably loaded brass tripod is used
to pack the ice by light blows delivered at intervals of 21 seconds, by the
aid of th-. lever L. A train of four wheels is necessary to reduce the
PEIRCE AND VVILLSON. — THERMAL CONDUCTIVITIES. 43
speed of A' to one revolution in 20 seconds, though only two wheels are
shown in the drawings. The tripod slides in guides which revolve with
A', and a swivel at the top prevents the cord from twisting.
The rotation of A' and of the inside ice holder, Q, which is connected
with K by a thin yoke, are matters of much importance. The continual
rubbing of the ice over the flat surface of the casting seems to be neces-
sary if the latter is to be kept at a uniform constant temperature for
hours. The energy used in rotating Q is so little as to be quite negligible,
as we shall show further on. The ice in K is piled up so as to cover
P completely, and we have been unable to detect any difference be-
tween the temperatures within and without P by fine, properly protected
thermal junctions introduced for the purpose. If, while A' revolves, Q is
kept still, the amount of ice melted in Q becomes irregular, though
the whole amount of drip in two or three hours is not very different
from the amount of steady drip in an equal time when Q is rotating.
Only selected lumps of ice are put into Q. The ice to be used is first
broken up into pieces weighing something like 15 grams each, by means
of an ice-cracking machine, and these pieces are then put into ice
water so that their sharp edges may become slightly rounded. They
are then drained and dropped into Q. In this way a slight amount of
water attached to the ice is introduced into Q^ but the error due to this
cause appears to be of slight importance. In some of our experiments
the ice to be used was carefully dried in cold blotting paper, but this
precaution does not seem to be necessary, though the use of small bits
of ice with sharp edges is to be avoided, ^'s capacity is about 2,000
cubic centimeters. After Q has been freshly filled in the course of any
experiment while K is rotating, no record is kept for some time, perhaps
fifteen minutes, of the amount of drip. Before the expiration of this
interval the extra water introduced into Q with the ice has drained off",
and the indications have become steady. After this the apparatus is
allowed to run for about two hours until .jOO grams of ice or less has
been melted, and then Q is refilled. The drip tube always contains a
few drops of water, but this amount remains sensibly constant during the
progress of our experiment. The drip is collected in a graduated vessel,
and the approximate amount is noted from time to time to see whether
the flow is steady. The whole is then more accurately determined by
weighing, at longer intervals.
The regularity of drip is a far more sensitive test of the approximate
attainment of the final state of the body experimented on and its sur-
roundings than is a sensil)Iy constant tenqxM-ature gradient on the axis.
44 PROCEEDINGS OF THE AMERICAN ACADEMY.
In most of our experiments with the large apparatus just described, a
sufficiently steady state has been attained in about five hours from tlie
beginning of the heating. Sheets of blotting paper were generally in-
serted between the prism to be tested, and the hot and cold boxes, to
serve as elastic pads, and to prevent the possible wetting of the edge of
the prism by moisture condensed on the ice box. The presence of this
paper prolonged the time of waiting for the final state to be attained,
but did not influence the results of the measurement of the conductivity
of the prism. When filled with ice, Z and K weigh about 300 kilograms,
and the additional pressure due to the bolts is considerable, so that, when
the prism is made up of brittle material like glass, the blotting paper
or an equivalent must be used to prevent the prism from injury. We
have tried several different materials, and of these the blotting paper
is the most satisfactory. We may note in passing, however, tbat the
indications of thermal couples placed between soft jiads and the hard
prisms are often very anomalous, two thermal junctions placed side by
side sometimes differing very widely. In all the experiments that we
regard as trustworthy the slab to be tested with its attendant thermopiles
was placed between two other slabs of the same material^ in forming the
prism.
Most of our mercury thermometers were made by Alvergniat, or by
Richards & Co., but our final standard was Tonnelot No. 11,142, upon
which a very complete set of tests bas been made at the International
Bureau of Weights and Measures.
For temperatures higher than 100° C. we had two platinum ther-
mometers of the general form described by Messrs. Griffiths and Cal-
lendar. These served an excellent purpose, though the wire, about
0.2 mm. in diameter, seemed from the form of the calibration curve
not to be very pure. The resistance of one of them, as measured by
a Carey Foster Bridge was about 29.25, 3G.78, 42.85, 45.31, or 55.43
ohms, according as it was immersed in melting ice or the vapor, at 7G0° c.c.
pressure, of water, anilin, naphthalin, or mercury. We have another ther-
mometer made of pure platinum wire furnished by Messrs. Johnson and
Matthey, 0.005 inch in diameter. This we intend to make our standard.
All our thermal elements were made either of platinoid and copper, or
of German silver and copper; some were of wire, and some of narrow
ribbon carefully rolled for our use. Each specimen of platinoid or
German silver was " butt-jointed," generally by silver solder, to a piece
PEIliCE AND WILLSON. THERMAL CONDUCTIVITIES.
45
of the purest obtainable copper of equal cross section. Our finest wire
thermal elements, less than one tentli of a millimeter in diameter, were
so skilfully made by Mr. Sven Nelson, of Cambridge, that the joint was
hardly perceptible. Our German silver and copper ribbon thermal ele-
ments, about one eighth of a millimeter thick, were made by Mr. T. W.
Gleeson of Boston. These last were first soldered with the help of a
holder constructed for the purpose, and the joint was then rolled or
scraped until it was as nearly as might be of the same thickness as the
adjacent nietal.
For wire thermal elements we had large quantities of three kinds of
platinoid, approximately 0.74, 0.30, and 0.097 mm. in diameter. The
first two specimens were obtained about ten years ago from Messrs.
Elliott Brothers, and have been thoroughly seasoned. Each is thermo-
electrically pretty definite, though the two are quite different in their
properties. The electromotive force, in microvolts, of [)latinoid and
electrolytically deposited copper elements made of these wires may be
tabulated as follows for low temperatures.
Temperatures of the
Junctions.
Electromotive Force
of Platinoid No. 1
VH. Copper.
Electromotive Force
of Platinoid No. 2
VS. Copper.
0° and 10°
189
152
0° and 20°
379
306
0° and 30°
572
465
0° and 40°
769
628
0° and 50°
971
799
0° and G0°
1179
973
0° and 70°
1391
1159
0° and 80°
1609
1356
0° and 90°
1834
1569
0° and 100°
2063
1787
Besides platinoid we have used with copper for wire thermal elements
two kinds of German silver wire respectively about 0.1 mm. and about
46 PROCEEDINGS OP THE AMERICAN ACADEMY.
0.58 mm. iu diameter. The smaller German silver wire was connected
with the corresponding copper wire by a thin joint of electrolytically
deposited copper. These joints were very satisfactory, but extremely
tedious to make.
In some of our experiments we used fine wire thermal junctions in-
serted in shallow grooves accurately cut in the faces of the slabs to be
tested. These grooves were made in a Brown and Sharpe Universal
Milling Machine by extremely thin hard steel saws (No. 34 B. & S.
Gauge) held between flat disks of somewhat smaller diameters than the
saws to prevent buckling. The wire that we used fitted the grooves very
closely and we hoped that the indications of the thermal couples would
enable us to determine the mean temperature of the walls of the groove
when the grooved slab was placed against a flat one. We soon found,
however, that the results were most irregular, and, although we have
s[)eut some time in attempts to make observations obtained in this way
trustworthy, we have met with little success. Sometimes our results
have been good, and sometimes they have been considerably in error.
We do not yet know how to make them always good. It appears that
a thermal junction must be pressed firmly against a surface, the tempera-
ture of which it is to take approximately. Although we are not ready
to discuss this subject exhaustively, we mention our experiences to show
why we have abandoned for the present this very obvious manner of
inserting thermal junctions into a prism built up out of slabs, in favor
of what at first sight seems a less satisfactory device. After some pre-
liminary experiments with fine wire thermal junctions laid between the
slabs, with and without sheets of tinfoil at the sides of the wire, we
determined to use the thin ribbon thermal junctions, elsewhere described,
with varnished edges, so that sheets of tinfoil of the same thickness might
be laid at the sides of the ribbon, and in this way a sheet of metal be
introduced between the slabs.
It has been necessary for us to calibrate in the course of our work
a large number of thermal elements. Some of these when properly
protected we have heated with thermometers iu elaborately jacketed air
baths or in tanks of water or oil, and some in vapor baths. We have
had considerable quantities of nearly pure chloroform, benzol, a^thylen
bromide, bromoform, anilin, paratoluidin, naphthalin, chinolin, a naph-
thol, acetanilid, naphthylamin, diphenylarain, phenanthren, anthracen,
and a few other substances, the boiling points of which divide the ordi-
nary thermometric scale below the boiling point of sulphur into small in-
tervals. A good number of these, but not all, we have actually used.
PEIRCE AND WILLSON. — THERMAL CONDUCTIVITIES. 47
Some of our thermopiles have been cahbruted for us at temperatures
between 0" C. aud 100° C. by Mr. C. G. Tersous of the staff of the Jef-
fersou Laboratory, aud he has assisted us in much of our other work.
Thk Thermal Conductivity of Marble.
With the apparatus described in this paper we have made a large
uumber of experiments. As has beeu already intimated, we are not
entirely satisfied with the source of heat that we have used for temjjera-
tures higher than 300° C. because of the ditliculty of keeping these
temperatures constant for long intervals of time, while for tempera-
tures between 0° C. and 100° C, it has been easy to get closely agree-
ing results many times over. We have, nevertheless, made a good
many determinations at the higher temperatures, aud, while we are not
yet ready to state definitely the law of variation with the tempera-
ture of the thermal conductivities of materials in which we have found
such variations, we may say that, of the substances which we have ex-
amined, two, a special brand of glass of which we have a number of
large plates, and dry white marble,* show no appreciable change in
thermal conductivity within the limits of our measurements. We shall
therefore content ourselves in this preliminary paper with giving the
results of a number of determinations, made at different low tempera-
tures, of the conductivities of about twenty specimens of marble of
different kinds. Incidentally we shall need to describe very briefly
some experiments upon the glass plates just mentioned.
It will appear that the conductivity of a specimen of marble at
ordinary mean temperatures may de^jend to the amount of several per
cent, as Messrs. Herschell and Lebour have shown, upon the amount
of moisture which the specimen holds. For this reason we have aimed
at an accuracy of only 1 % in the determinations here recorded. A change
in conductivity much less than this was of course easily observable. The
difference of temperature between two thermopiles, one of which is only
a few degrees hotter than the other, can be measured with considerable
accuracy, but it will be sufficient here to state the results correct to tenths
of degrees.
* The conductivity of the specimen of marble upon wliich R. Weber has made
a set of extremely accurate measurements appears to change by only one two-
thousandth part of its own value between 0° C. and 100° C.
48 PROCEEDINGS OF THE AMERICAN ACADEMY.
While it takes a long day to make an accurate determiuation with
our large apparatus of the absolute couductivity of a slab, two determi-
nations may easily be made in the same time of the relative con-
ductivities of the slabs which go to form a prism, since the gradient
on the axis of the slab does not sensibly change after four hours
of heating, and it is then only necessary to note the readings of the
thermopiles. With our smallest apparatus and thin slabs two hours are
often sufficient for a measurement. Our experience seems to show that
this method of comparison is susceptible of great accuracy. We have
made a very large number of direct determinations of the conductivities
of different slabs of stone, but, in view of the fact mentioned above that
the amount of moisture in the stone affects the conductivity very appre-
ciably, even if the les.s tedious method of comparison were not ecpially
accurate, we should think it wise in future to determine with great care
the absolute conductivity of a standard substance unaffected by moisture,
and then compare with it the conductivity of the stone slabs. The ac-
curacy with wliich the comparison can be made is greater of course than
that of a single absolute determination.
The particular kind of glass which we have found useful as a compari-
son substance was selected some years ago from the stock of the Boston
Plate Glass Company. The faces of each plate are very nearly plane,
but the planes are not in every specimen quite parallel. The conduc-
tivities of different plates are somewhat different, but the conductivity of
each plate remains sensibly constant within large ranges of temperature.
Cut from this glass we have a number of slabs 60 centimeters square, a
number of slabs 30 centimeters square and some disks about 20 centi-
meters in diameter.
We shall wish to discuss the properties of this kind of glass at higher
temperatures more particularly on another occasion. For our present
purposes, it is worth while to measure the temperatures to tenths of
degrees only and the thickness of a slab to the nearest twentieth of a
millimeter, and an account of a few experiments to this degree of ac-
curacy, chosen almost at random from the large number of which we have
records, will suffice.
Slabs A, B, C, and D are cut from one particular large homogeneous
piece of this glass, the conductivity of which, according to our determi-
nations, is to that of Plate III. mentioned below as 187 to 175. We
shall assume the conductivities of these slabs to be 0.00277 at all ordinary
temperatures. We have not been able to detect any differences in their
conductivities.
PEIRCE ANP VVILLSON. — THERMAL CONDUCTIVITIES. 49
Experiment (a). A compound slab, made up of slabs Z? and A with
their thermal elements, was placed between two other glass plates to form
a prism. The thickness of B is 0.950 cm. and of A 0.935 cm. In the
final state of the prism, the thermal elements on the warmer face of B,
between B and A, and on the colder side of A, indicated 88°. 1, 63°. 4,
and 38°. 9 respectively. A fall of 14°. 7 in 0.950 cm. is very nearly equal
to a fall of 14°.5 in 0.935 cm.
Experiment (b). In the final state of a prism made up of slabs A and
B shut in between two other glass plates, the thermal elements on the
warmer face of A, between A and i?, and on the colder face of B,
indicated 85°.0, 62°.2, and 39°. 1 respectively. A fall of 22°. 8 in
0.935 cm. is very nearly equal to a fall of 23°. 1 in 0.950 cm.
Experiment (c) . Three slabs A, C, and E of the standard glass with
three other glass plates, which we may denote by F, Q, and H, were
built up into a prism PA Q C E R with thermal elements between P and
A, A and Q, Q and (7, E and R. In the final state the temperatures
of the thermal elements were very nearly 88° .2, 74°. 2, 58°.8 and 30°.0,
respectively, so that the gradient in the slab A of thickness 0.935 cm. is
almost exactly the same as in the double slab G E oi thickness 1.93 cm.
There seemed to be, therefore, no appreciable contact resistance (Ueber-
gangswiderstand) between the two slabs.
Experiment (d). After experiment (c) had been finished, a narrow
ring of blotting paper, the inside diameter of which was only slightly less
than the diameter of the disks, was inserted between C and E so as to
have a dead air space between them 0.7 mm. thick, when the prism was
under pressure. In the final state the temperatures were now 89°. 9,
78°. 3, 66°. 5, and 25°. 9, so that in this particular case the dead air space
was nearly equivalent to a glass plate 4.8 mm. thick.
Experiment (e). In this experiment Plate III., of 0.875 cm. thick-
ness, was a part of a prism heated in the larger apparatus intended for
the determination of absolute conductivities. The temperatures of the
thermal elements on the faces of the plates in the final state were 69°. 7
and 58°.8 respectively. In 9060 seconds 464.5 grams of ice were melted.
Assuming the area of the bottom of the ice pot to be 126.7 square centi-
meters and the latent heat of melting of ice to be 79.25, this corresponds
to a conductivity of 0.00258. It is obvious, however, that the last
figure of this number is not quite definitely determined.
VOL. XXXIV. — 4
50 PROCEEDINGS OP THIO AMERICAN ACADEMY.
Experiment (/). lu the final state of a prism similar to the one used
in the last experiment, 311.9 grams of ice were melted in 5340 seconds
when the temperatures of the thermal elements on the faces of Plate III.
were 66^.4 and 54°. 1. This corresponds to a conductivity of 0.00260.
Here again the last figure is in doubt.
We had occasion to measure the absolute conductivity of only one
other of the 60 cm. s(juare plates bought at the same time as Plate III.
This was Plate I. The results of two experiments made on it were
0.002 G2 and 0.00259. The crown glass used b}^ Oddone had a con-
ductivity of 0.00245, that of Lees* a conductivity of 0.00243.
We will next cite a single experiment to show how much the con-
ductivity of the particular kind of statuary marble that we used could be
changed by moistening the stone.
Experiment (g). A prism was made up of three plates of glass, A, P,
and Q, and three dry slabs of statuary marble, C, D, and E, arranged in
the order P A Q ED C with thermal junctions between P and A, A and
Q, E and D, D and G. The temperatures indicated by the thermal
junctions when the prism had sensibly reached its final state were 84°. 6,
67°. 7, 38°. G, and 27°. 7. D was then well moistened with water, and
the experiment was then repeated. The temperatures were then 85°. 3,
70°. 5, 46°.0, and 38°. 1, so that the conductivity oi D had been increased
in the ratio of 1.21 to 1.
Experiment (Ji). In order to form an idea of the amount of change
with the state of the weather of the conductivity of a piece of our Carrara
statuaiy marble, we made three comparisons on three different occasions
of the relative conductivities of a slab of it (0) 1.08 centimeters thick,
and a plate {A) of standard glass. Between the experiments, was left
in a room the windows of which were much of the time open. The results
were as follows : —
Temperature of the warm side of the glass.
Temperature of the cool side of the glass,
Temperature of the warm side of the marble,
Temperature of the cool side of the marble,
Ratio of the conductivities of the marble and the glass, 1 .84
Average conductivity of the slab O, 0.00509
Another specimen of Carrara marble had a conductivity of 0.00501.
* We have not yet seen the paper by Mr. Lees mentioned in the March, 1898,
number of the Beibliitter zii den Annalen der Physik und Cheniie.
85°.5
84°.6
84°.8
08°. 1
67°.l
67°.4
43°.0
42°.l
40°.3
32°.0
31°.l
29°.4
1.84
1.84
1.83
PEIliCE AND WILLSON. THERMAL CONDUCTIVITIES.
51
Before we state the results of our own observations upon other speci-
mens, we will give for purposes of com[)arison some determinations of the
thermal conductivities of marble made by other observers.
o
o
o
o
o
d
>.
1
1
!fc
1—1
3
o
1-H
1—1
(M
xt<
'^
^ CO
t^
t^
-tH
lO
•o
lO
'Ci
lO
t^ t^
l^
c:.
o
O
o
o
o
o
o o
o
O
O
o
o
o
q
o
o o
o
o
d
d
d
d
d
d
d d
d
d
fl
a
s
Q
-J
^
^
n3
C
c3
u
S
o
^
^
^
S
-Q
~*
^
**
O
"3
03
• 1
03
S
-
*
^
03
^
^
03
^
t/3
t>
0.
a
Tj
'o
»
a>
a
<A
-0)
-0)
(^
w
e4
h-^
kH
Oh
Ph
•
^
•
*
3
'
^H
03
^
tH
.*
'^
f^
QD
^
;3
3
03
:;
^
1
03
■73
^
^03
-rs
bC
C3
'rt
03
■*
•*
3
03
03
f^
'^
a
03
03
^
fl
?j
s
03
!3
03
a
3
03
1
^
03
03
m
r-4
es
o
03
g
2
2
03
03
03
m
O
03
3
03
03
1
Oj
o
03
3
c3
a
'5
be
03
t4
13
CO
33
rH
la
bo
a
P^
M
^
13
^
(73
fe
52 PROCEEDINGS OF THE AMERICAN ACADEMY.
Takiug a certain piece of " Pyrenees Marble " as a standard, Dr. Less
found the conductivities of specimens of " Carrara Marble " and " Italian
Marble " to be 0.769 and 0.763 respectively.
In determining the thermal conductivities of the specimens of marble
mentioned below, the prism clamped between the hot and the cold box
of our apparatus was made up of six slabs in series, a plate of standard
glass 0.935 cm. thick between two thin plates of glass, and the shib to be
tested between two thin slabs of marble. A ribbon thermal element
and tinfoil wings were placed on each side of the standard glass, and on
each side of the marble to be experimented on, so that there were four of
these thermal elements in all. When the prism had sensibly reached its
final state, the temperatures of the thermal junctions were determined and
the ratio of the conductivities of the glass and the marble was assumed
to be etjual to the reciprocal of the ratio of the gradients in the two
slabs. By introducing an extra plate or a sheet or two of blotting paper
into the prism, the two gradients could be altered at pleasure but not
their ratio. So far as we could see, it was immaterial in the case of
these substances whether the marble base of the prism or the glass base
was placed uppermost, but we generally placed the marble on top, so that
the mean temperature of each specimen might be about 30° C. In stating
the results of some of these determinations, we shall give the tem])era-
tures of the four thermal junctions in order, then the ratio of the con-
ductivities of the marble to be tested and the standard glass, and finally
the absolute conductivity of the marble on the assumption that that of the
glass is 0.00277. We shall give the absolute conductivity of the marble
to three significant figures, but it is evident that the last of these is not
determined. All the specimens were artificially dried for some time in
the hot air space over the boilers which furnish steam for heating the
Jefi'erson Laboratory, and were then allowed to stand for some weeks
at ordinary room temperatures so that their conditions might be normal.
The artificial heating drove off the excess of moisture acc^uired by the
marble while being cut under water at the mill.
Most of our stone was obtained from Messrs. Bowker and Torrey of
Boston, who kindly collected for us representative specimens of such
materials as are commonly used for decorative and monumental purposes.
We have given to the slabs the names used by stone workers and have
called them all "marbles," though one or two might more properly be
called " limestones." The " Mexican Onyx " is really travertine. Our
thanks are due to Prof. J. E. Wolff for help in identifying our specimens.
PEIRCE AND WILLSON. — THERMAL CONDUCTIVITIES. 53
Possiliferous Tennessee Marble.
(Red with numerous white fossils.)
Thickness in centimeters, 2.40
Temperatures of the faces of the glass plate, 82°. 3 and 63°. 2
Temperatures of the faces of the marble slab, 43°.3 and 24°.4
Ratio of the couductivities of the marble and the glass, 2.73
Absolute conductivity of the marble, 0.00756
American White Marble.
(Cream white.)
Thickness in centimeters, 2.68
Temperatures of the faces of the glass plate, 83°. 6 and 64°. 6
Temperatures of the faces of the marble .slab, 45°. 4 and 20°. 3
Ratio of the couductivities of the marble and the glass, 2.15
Absolute conductivity of the marble, 0.00596
Vermont Statuary Marble.
(Snow white with coarse but uniform grain.)
Thickness in centimeters, 2.40
Temperatures of the faces of the glass plate, 82°. 9 and 64°. 2
Temperatures of the faces of the marble slab, 44° .7 and 21°. 7
Ratio of the couductivities of the marble and the glass, 2.09
Absolute conductivity of the marble, 0.00578
Lisbon Marble.
(Light terra-cotta with darker veins.)
Thickness in centimeters, 2.30
Temperatures of the faces of the glass plate, 80°. 9 and 60°. 8
Temperatures of the faces of the marble slab, 39°. 6 and 19°. 6
Ratio of the conductivities of the marble and the glass, 2.47
Absolute conductivity of the marble 0.00685
St. Baume Marble.
(Yellow, red, and yellowish white brecciated.)
Thickness in centimeters, 2.36
Temperatures of the faces of the glass plate, 80°. 9 arid 61°. 2
Temperatures of the faces of the marble slab, 40°. 3 and 22°. 1
Ratio of the couductivities of the marble and the glass, 2.75
Absolute conductivity of the marble, 0.00761
54 PROCEEDINGS OP THE AMERICAN ACADEMY.
Rose Ivory Marble.
(From Djebel-er-Roos, Algiers. White with very sUght piukish tinge. Very fine
in grain.)
Thickness in centimeters, 2.64
Temperatures of the faces of the glass plate, 80°. 3 and 60° .2
Temperatures of the faces of the marble slab, 39°. 8 and 19°.0
Ratio of the conductivities of the marble and the glass, 2.7C
Absolute conductivity of the marble, 0.00756
Italian Egyptian Marble.
(Breccia. Slate colored witli oclire-yellow and white veins.)
Thickness in centimeters, 2.55
Temperatures of the faces of the glass plate, 83°. and 63°. 3
Temperatures of the faces of the marble slab, 43°. 1 and 19°. 2
Ratio of the conductivities of the marble and the glass, 2.25
Absolute conductivity of the marble, 0.00623
Mexican Onyx.
(Alabaster white, translucent.)
Thickness in centimetres, 2.29
Temperatures of the faces of the glass plate, 82°.9 and 63°. 8
Temperatures of the faces of the onyx slab, 43°.l and 19°. 8
Ratio of the conductivities of the onyx and the glass, 2.01
Absolute conductivity of the oynx, 0.00556
Vermont Dove Colored Marble.
(Dove colored witli light and dark striaj.)
Thickness in centimeters, 2.19
Temperatures of the faces of the glass plate, 80°.5 and 59°. 3
Temperatures of the faces of the marble slab, 39°. 1 and 18°.9
Ratio of the conductivities of the marble and the glass, 2.47
Absolute conductivity of the marble, 0.00684
Bardiglio Marble.
(From the Seravazza quarries. Cloudy white, with network of distinct dark lines.)
Thickness in centimeters, 2.44
Temperatures of the faces of the glass plate, 81°. 8 and 61°. 3
Temperatures of the faces of the marble plate, 41°. 1 and 19°. 3
Ratio of the conductivities of the marble and the glass, 2.45
Absolute conductivity of the marble, 0.00680
PEIRCE AND WILLSON. — THERMAL CONDUCTIVITn]S. 55
sienna Marble.
(Yellowish white with blue veins.)
Thickness in centimeters, 2.48
Temperatures of the faces of the glass plate, 81 ".5 and 60°. 9
Temperatures of the faces of the marble plate, 40°. 9 and 18°. 6
Ratio of the conductivities of the marble and the glass, 2.44
Absolute conductivity of the marble, 0.00676
St. Anne Marble.
(Brown black with white patches.)
Thickness in centimeters, 2.34
Temperatures of the faces of the glass plate, 80°. 9 and 60°. 1
Temperatures of the faces of the marble plate, 38°. 8 and 19°. 7
Ratio of the conductivities of the marble and the glass, 2.73
Absolute conductivity of the marble, 0.00755
American Black Marble.
(Dark slate.)
Thickness in centimeters, 2.43
Temperatures of the faces of the glass plate, 81°. and 61°.l
Temperatures of the faces of the marble slab, 40°. 1 and 19°. 2
Ratio of the conductivities of the marble and the glass, 2.47
Absolute conductivity of the marble, 0.00685
Vermont Cloudy Marble.
(Cloudy white with darker patches.)
Thickness in centimeters, 2.55
Temperatures of the faces of the glass plate, 82°. 3 and 62°. 1
Temperatures of the faces of the marble slab, 41°. 8 and 19°. 4
Ratio of the conductivities of the marble and the glass, 2.46
Absolute conductivity of the marble, 0.00681
Knoxville Marble
(Pink with occasional dark serrated veins.)
Thickness in centimeters, 2 37
Temperatures of the faces of the glass plate, 81 °.6 and 61°.
Temperatures of the faces of the marl>le .slab, 38°.9 and 20°.l
Ratio of the conductivities of the marble and the glass, 2.62
Absolute conductivity of the marble, 0.00757
56 PROCEEDINGS OP THE AMERICAN ACADEMY.
Arranging the results in the order of the couductivities of the speci-
mens, we get the subjoined table. We call attention to the two groups
of fine-grained marbles, which have conductivities of about 0.0068 and
0.0076 respectively, at about 30° C.
Variety of Marble. Conductivity.
« Carrara Statuary " 0.00501
0.00509
« Mexican Onyx " . 0.00556
" Vermont Statuary " 0.00578
" American White " 0.00596
"Egyptian" 0.00623
"Sienna" 0.00676
" Bardiglio " 0.00680
"Vermont Cloudy White" .... 0.00681
" Vermont Dove Colored " . . . . 0.00684
"Lisbon" 0.00685
" American Black " 0.00685
"Belgian" 0.00755
" African Rose Ivory " 0.00756
"Tennessee Fossiliferous" .... 0.00756
"Knoxville Pink" 0.00757
" St. Baume " 0.00761
We reserve for a second paper the results of observations made upon
other materials.
Our acknowledgments are due to the American Academy of Arts and
Sciences, which has made an appropriation from the Rumford Fund in
aid of our work.
Jefferson Physical Laboratory,
Cambridge, Mass.
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. Xo. 2. — Xovember, 1898.
CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.
THE CONTACT-POTENTIAL BETWEEN METALS AND
FUSED SALTS, AND THE DISSOCIATION
OF FUSED SALTS.
By Clarence McCheyxe Gordon.
CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF
llAKVAKD COLLEGE.
THE CONTACT-POTENTIAL BET^yEEN METALS AND
FUSED SALTS, AND THE DISSOCIATION OF FUSED
SALTS.
By Clarence McCheyne GoRDOjir.
Received October 10, 1808.
Presented by Theodore W. Richards, October 12, 1898.
The potential difference between fused salts and metals immersed in
them is a quantity of great importance because of its relation to the theory
of the origin of contact-potentials, on account of the light it throws upon
the decree of dissociation of salts in their fused state, and in view of its
bearing upon the electrolytic separation of the metals. Notwith^taud-
insj; these important relations, the subject has received practically no
attention at the hands of scientific investigators. In a few cases cells
containing fused salts as the electrolyte have been measured, but always
with some other end in view than the study of the single potential differ-
ence between metal and salt.*
The practical difficulties in the way of carrying out an exhaustive in-
vestigation of this subject are many. Among the more important of
these are the fact that many salts decompose below or slightly above
their melting point, the disturbing effect of side reactions which at ordi-
nary temperatures would be so slow as to cause little or no inconvenience,
and the impossibility of using glass vessels at temperatures as high as the
melting points of most inorganic salts. Such difficulties as these have
prevented me from extending the investigation rapidly, but the results
* Lash Miller (Zeit. fiir pliys. Chemie, X. 459, 1892) used fused salts in experi-
ments to prove that there was no change in the contact-potential as the metal
chantred from liquid to solid state.
Poincare (Ann. cliim.et phys., [6.], XXI. 289, 1890) measured reversible cells con-
taining melted zinc and tin salts with the heat of reaction as the end in view. He
investigated also some polarization phenomena with silver electrodes.
60 PROCEEDINGS OF THE AMERICAN ACADEMY.
from the cells so far measured seem of such value that I give them in this
preliminary jjaper, hoping to extend the investigation to other salts and
metals in the near future.
For dilute water solutions the variation of the potential hetween a
metal and a solution of a salt, whose kation is the metal of the electrode,
is given by the well known Nernst formula,*
^^ - ^^ = V ^°s« ^'
where E^, E^, are the respective potentials for the concentrations C^, C, ;
R, the gas constant ; T, the absolute temperature; and m, the valency of
the metal. In the derivation of this formula the gas law, jo v= 11 T, is
applied to the dissolved salt, and complete dissociation is assumed. In
case the dissociation is not complete, Ci, Co, denote the concentration of
the dissociated part.
For fused salts, in case the solutions are sufficiently dilute, we might
expect the same or an analogous formula to hold. Any experimental
evidence, however, for the applicability of the gas law to fused salt solu-
tions, or as to the amount of dissociation, has been lacking.
Accordingly, the first question to be determined was whether or not,
for any measureable range of concentrations, the potential difference
would vary according to the Nernst formula.
So far all my observations have been upon silver electrodes immersed
in fused mixtures containing varying amounts of silver salts. Most of
the cells measured were simple silver nitrate concentration cells of the
type,
Ag
TiAgNOs
in
a:oAgN03
y,KN03 + y^NaNOs
Aff.
The mixture of potassium and sodium nitrates, rather than either alone,
was used as solvent in order to obtain a lower melting point.
Experimental.
General. — In a few preliminary measurements I tried keeping the
temperature constant with an air bath, but was unable to attain much
constancy, and therefore substituted a vapor bath. The practical ar-
rangement of the cell in its constant temperature vapor bath is shown in
the opposite figure.
* See Nernst, Zeit. fiir phjs. Cliemie, IV. 129, 1889.
GORDON. — DISSOCIATION OF FUSED SALTS.
61
The jacket Z>, containing the vapor, is a thin-walled glass cylinder
about 5 cm. in diameter, joined below by means of a Wood's metal or
mercury joint, W, to a cast-iron cup, /, on which the heating flame plays.
This vapor bath arrangement is such as is in
common use for V. Meyer vapor density deter-
minations at moderately high temperatures, ex-
cept that in the present case the jacket tube
was shortened. Inside of this vapor jacket is
the large test tube C, extending about 25 cm.
below the top of the outside jacket. So much
of the upper opening of the cylinder D as tiie
tube C does not fill is closed with the asbestos
sheet Bi. This covering is of course not air
tight, but makes it possible for the vapor to
ascend near the top without any rapid escape
into the room. In the tube (7 is the cell to be
measured. It consists of the silver electrodes
JlJi, £!.,, in the tubes Ai, A.^^ which contain the
two differently concentrated solutions of silver
nitrate, and the connecting solution *S^. The
capillary ends of the tubes A^, Ao, were lowered
beneath the surface of the connecting licjuid
only long enough for the measurements to be
taken. In order to prevent air circulation, the
top of the tube C was covered with an asbestos
sheet having holes for the tubes A and the ther-
mometer T.
The heat was sup[ilied by a three-tube Bunsen
burner, fed from a gas supply containing a good
pressure regulator. By reason of this regu-
lator, when the gas was once adjusted, no care
was required to keep the heiglit of the vapor
constant, except as the amount of boiling sub-
stance made slow escape into the room. The
top of the condensing vapor column could be jilainly seen, and was kept
constant about 18 cm. above the lower end of the tube C. With this
arrangement any difference in tempei-atnre between the upper and lower
jvart of the liquid S was unnoticeable.
Temperature. — The cells were measured at two temperatures, that of
boiling diphenylamine, and of boiling chinoline. The former was taken
One Fourth Natural Size.
62 PROCEEDINGS OF THE AMERICAN ACADEMY.
from the laboratory stock, and the latter was made for me by Mr. J. B.
Churchill, whom I wish to thank for j^roviding me with a very consider-
able amount. Since the temperatures were read on a mercury ther-
mometer in the cell itself, no attempt was made to attain absolute iiurity
in the boiling substances.
The most of the cells measured were liable to variations due to other
causes than change of temperature of about 0.001 volt. This was the
accuracy of measurement aimed at. As it was found to require a change
of about 5° ill temperature to cause a change of 0.001 volt, the tempera-
tuies were read on a thermometer graduated to degrees, and considered
constant as long as tiiey did not vary more than 1°.
In order to standardize the thermometer used, it was tested with pure
naphthalene. According to Crafts,* uai)hthalene boils under 756 mm.
pressure at 21 7°. 9. The thermometer used for these experiments regis-
tered 218" as the boiling point under tlie same pressure. Tlie actual
readings of the thermometer were therefore taken as correct.f
Change in Concentration. — In order to render the danger of change
ill concentration in the neighborhood of the electrodes as small as pos-
sible, the tubes A were given the shape shown in the figure. In the
preliminary measurements, however, there was still trouble on this ac-
count. This was due to the fact that the salts were put into the tubes in
the unmelted state. Air was unavoidably retained when the salt was
fused, which must necessarily be blown out after the tubes were lowered
under the level of the connecting solution, and the latter was thus
brought into the tubes. As the potentials were measured by means of a
capillary electrometer, even when the air filled the whole cross-section of
the capillary, the measurement could be made; the value was, however,
alwavs different from the true one. In order to avoid this trouble due
to air bubbles, the filling of the cell tubes with the pulverized salts was
* Amer. Chem. Journ., V. 307, 1883-84.
t Graebe (Annalen [Liebig's], CCXXXVIII. 362) finds tlio boiling point of
diphenylamine to be 302°. The earlier determinations of Hofmann, Girard and
Wilson, and Kreis, etc., are evidently incorrect, as Graebe standardized his ther-
mometers with benzophenone, according to the figures of Crafts. Whether the
difference between Graebe's 302° and the reading 298° of my thermometer in the
cell surrounded by the diphenylamine was due to impurities in the latter, to a
change in the value of the marked degree between 218° (the point found to be
correct), or to the fact that the thermometer in the cell was at a little lower tem-
perature than the surrounding vapor, I have not attempted to determine, since a
change of even 4^ in temperature would cause a change in voltage in the cells
measured of less than 0.001 volt.
GORDON. — DISSOCIATION OF FUSED SALTS. 63
abandoned. Instead, the salts were first melted in a test tube; and when
they were about the temperature of the coustant temperature bath, the
cell tubes were filled by immersion, and then transferred to the tube C.
After this method of filling was adopted, no change of concentration was
noticed.
The Connecting Solution. — The solution .S", which served to connect
the two differently concentrated silver nitrate solutions, consisted of equal
parts potassium and sodium nitrates, with some silver nitrate. For the
cells of small concentration it contained the same proportion of silver
nitrate as the less concentrated of the solutions in the tubes A. Since
it was found experimentally that the amount of silver nitrate in this
connecting solution had no effect on the value of the potential, for the
concentrated cells it was not accurately determined, but was about
10% AgNO..
Preparation of Solutions. — The potassium and sodium nitrates used
were purified by recrystallization. Crystallized argentic nitrate is easily
obtained in a condition pure enough for the end in view.
The concentrations of the silver nitrate ranged from 0.001 of total
weight to pure silver nitrate. 50 7c. and 10% solutions were weighed
out and fused in rather large quantities, and the other solutions made by
dilution of these.
21ie Molecular Concentrations. — While the solutions were naturally
made up by weight, it is the molecular volume concentrations, which are
to be used in the calculation. In order to obtain these it was necessary to
measure the specific gravities of the several mixtures. This was done at
the temperature of the chiaoline bath, 236°. Since the expansion coeffi-
cients for these salts, as found by Poincare,* are small, we can for our
purpose consider the concentrations at 298° to be the same. In order
to obtain the specific gravities, tubes of about 0.5 cm. internal diameter
were narrowed still more a few centimeters from the lower closed end,
and filled to a marked point on the narrow portion of the tubes while in
the chinoline bath. After cooling, the tubes were broken oflT at the
marked point, weighed, and the volumes determined by weighing with
water, applying the correction for expansion of the glass. The specific
gravities measured, and the molecular concentration calculated therefrom
are given in the following table.
* Loc. cit.
6i
PROCEEDINGS OF THE ABIERICAN ACADEMY.
TABLE I.
Per Cent bv weight
AgNO's. .
Specific Gravity.
g-niol. AgNOj
iu liter.
0.1
1.
10.
50.
100.
1.84
2.02
2.61
3.82
0.0108
0.11 (estimated).
1.16
7.68
22.5
Tlie Electrodes. — Pure silver wires served as electrodes. It was found
experimeutally that no special care need be taken as to the character of
their surface. In view of the irregularities shown by solid metal elec-
trodes in water solutions, this independence of the character of the surface
and previous treatment of the electrodes is quite remarkable. As a
matter of precaution the wires were scraped with a clean knife, washed,
and dried with filter paper before using. Tests were made on several
electrodes so treated by putting them into the same solution. They did
not give a difference of potential of more than 0.0001 volt.
Measurement of the Potential Difference. — Above the cell tubes the
silver wires were soldered to copper ones, and the potential difference
measured by the Poggendorf method, using a Leclanche cell, an Ostwald
potential box, and a capillary electrometer. A previously standai-ilized
one-volt Helmholtz cell was taken as the standard of electromotive force.
The electrometer was read with a microscope, and would readily show a
difference of 0.0004 volt.
No noticeable error could result from the possible difference in temper-
ature of the two copper-silver junctions.
The Conductivity of Glass. — Although the temperatures used for these
nitrate cells were far below the softening point of glass, its conductivity
is, even at 200°, very considerable, being of about the same order of mag-
nitude as a j^iotfcT normal KCl solution. Measurements could indeed
be made with the capillary electrometer without having any liquid con-
nection between the two AgNOs solutions. The potentials so measured
(with the two cell tubes closed at the lower end) differed from the true
ones generally by about 0.02 volt. The conductivity, however, of tlie
fused salts is so great as to preclude any disturbance from conductivity of
the glass when the liquid connection is made.
GORDON. — DISSOCIATION OF FUSED SALTS.
65
Results of the Experiments.
The potentials of four such silver nitrate cells are given in the follow-
ing table. The calculated values are obtained by substitution of the
absolute temperatures and the molecular concentrations, as given in Table
I., in the Nernst formula,
a
Ex- E^ = RT loc
C{
TABLE II.
Silver Xituate Cells.
Xo.
Per Cent AgNO.,.
Chinoline bath.
Diphenylamine bath.
Sol. 1.
Sol. 2.
T.
E. M. F.
t.
E. M. F.
1
2
3
4
1.
10.
50.
100.
0.1
1.
10.
50.
°C.
232
236
Calc
0.101
0.102
0.082
0.045
Observed.
0.100
0.100
0.071
0.039
298
Calc.
0.114
0.115
0.003
0.051
Observed.
O.IIO
0.112
0.080
0.045
III general I can say that the measurements were all repeated several
times, and only in a few cases were there single measurements varying as
much as 0.00 1 of a volt from the mean as here given. The behavior
of the several cells with regard to their constancy requires individual
mention.
No. 1, at 298°, decreased rapidly in potential after first getting up, but
when the 0.1% solution was replaced by that of the outside connecting
solution (of same concentration), the value 0.110 volt was always ob-
tained. Five minutes later it had always decreased, in one case to as
low as 0.100 volt. This rapid decrease in value may account for the
fact that the observed value is 0.004 volt less than the calculated. Had
the measurement been taken immediately after the fresh solution came in
contact with the electrode, the observed value might have been larger.
This value was obtained repeatedly, however, the measurement being
taken about one minute after the introduction of the fresh solution into
the cell tube.
In the chinoline bath cell No. 1 remained constant for from ten to
fifteen minutes, and then decreased. It could be brought back to the
VOL. XXXIV. — 5
66 PROCEEDINGS OF THE AMERICAN ACADEMY.
original value by replacing the weaker of the two solutions witli that of
the connecting solution (of the same concentration).
Cell No. 2, at 298°, remained constant for at least twenty minutes
after the 1% solution in the neighborhood of the electrode was replenished.
Of four cells measured, one decreased in value, and three increased, on
standing.
This cell in the chinoline bath remained constant for several hours.
Cells 3 and 4 were extremely constant for several hours at a time, and
the same solutions iu the same tubes could be lieated up several days in
succession, and would always give the same value.
The inconstancy of the cells containing very dilute solutions would
seem to indicate that in these there is a slow reaction going on between
the solution and the electrode. The fact that the replenishing of the solu-
tion with that which had not been in contact with tlie electrode always
brought the cell back to the same value, shows that the reaction did not
take place throughout the solution as a whole, but only in the neighbor-
hood of the electrode. The more rapid variation at the higher tempera-
ture shows that, whatever the reaction may be, its speed increases with
the temperature. Obviously a slow dissolving of the silver electrode in
the fused salt would account for the observations, the weak solutions be-
ing much more affected by this possible irregularity than the strong ones;
but this hypothesis is not advanced as a certainty.
The observed values for the cells No. 1 and No. 2 agree remarkably
well with the calculated. We must conclude that the osmotic theory
of electromotive forces can be extended to the case of fused salts, and
that even in a 10% solution the silver nitrate is almost wholly dissociated.
In considering the observed potential to be the difference between the
two electrode potentials, we assume that the potential difference between
the two differently concentrated solutions is so small as to be negligible.
While the good agreement between the values found and calculated is
the strongest argument in favor of this supposition, it is to be inferred
also from what we know of aqueous solutions. Such potential differ-
ences depend on the difference in transference numbers of the two ions.
Nernst and Loeb,* and more especially W. Bein,t have shown that the
transference numbers all tend toward the value 0.5 as the temperature
increases. For AgNO^ especially, they are never far from 0.5, and ap-
proach very near it at the highest temperature measured. W. Beiu
* Zeit. fur phys. Chemie, II. 962, 1888.
t Wied. Ann.,' XLVI. 69, 1892.
GORDON. — DISSOCIATION OF FUSED SALTS.
67
found for the transference number of NOg iu AgNOg, 0,470 at 20°, and
0.490 at 90°.
Dissociation. — In view of the agreement with the formula in case of
dilute solutions, the cousideralion of the cells containing the more con-
centrated solution becomes of greatest interest. It seems probable that
the deviations from the calculated values are here due entirely to in-
complete dissociation. "We thus have a means of calculating the degree
of dissociation for the 50% AgNOg solution, and for the pure salt. The
results of this reckoning are given in Table III.
TABLE III.
Dissociation of Silver Nitrate.
Temperature 2'ZQ°. Dissociation 0.1% sol. assumed complete.
Per Cent
AgNOj.
Logp
Degree of Dissociation.
By Measurement.
Calc. from E M. F.
50.
100.
2.8508
3.3172
2.6912
3.0784
0.69
0.58
This is, so far as I know, the first determination of the degree of dis-
sociation of a fused salt. To say that pure fused silver nitrate is 58%
dissociated seems at first thought somewhat incredible ; especially when
we think of the almost infinitesimal dissociation of water and other
liquids at ordinary temperatures. These are not the only measurements,
however, that go to show that the degree of dissociation is large. Their
large conductivities, and the small range of the same when different
salts are considered, are in favor of it. The careful measurements of
Poincare * on the conductivity of fused salt mixtures show a behavior
very different from that of electrolytic solutions at ordinary temperatures.
In the latter case the conductivity is generally greatly changed, or first
made possible, by the mixing, while for fused salts the conductivity is
almost an additive property of the separate salts. The large amount of
dissociation gives at least a qualitative explanation of this apparently
anomalous behavior.
* Loc. cit.
68 PROCEEDINGS OF THE AMERICAN ACADEMY.
Before the above cells were measured, cells of the following types
were studied : —
Zn - ZnCl, - AgCl in ZnCIo - Ag,
Zn — ZnBi'a — AgBr in ZuBro - Ag.
The observed values corresponded roughly with those calculated, but
the readings were not constant enough to be worthy of detailed men-
tion. In the hope of securing greater constancy silver was substituted
for zinc, and two different strengths of argentic halide dissolved in
zincic halide were used around the electrodes, as in the experiments
with the nitrates. These also were not constant, and some interesting
2)henomena concerning their inconstancy are worthy of further investiga-
tion. After the partial failure of these two attempts, the nitrates were
resorted to with the satisfactory and interesting results described above.
These measurements were carried on in the Laboratory of Physical
Chemistry in Boylston Hall, Harvard University. I wish to thank
Professor Theodore W. Richards for his kindly interest and aid in the
investigation, and especially for his many helpful suggestions when ex-
perimental difficulties were encountered.
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. Xo. 3. — November, 1898.
ON FLUCTUATIONS IN THE COMPOSITION OF
NATURAL GAS.
By Francis C. Phillips.
Aid in the Work described in this Paper was given bv the Acade.mt from the C. M. Warren
FcND FOR Chemical Research.
ON FLUCTUATIONS IN THE COMPOSITION OF
NATURAL GAS*
By Francis C. Phillips.
Presented by the C. M. Warren Committee, October 12, 18S8.
Received October 12, 1898.
At the mills and factories of the Pittsburgh region the opinion is
often expressed by men in charge of steam boilers where natural gas
is the fuel used, that the gas fluctuates in its heating power, and that at
certain times more gas must be used than at others to accomplish the
same work. Changes of pressure in the mains, owing to varying
demands upon the supply, requiring that the valve controlling the
admission of gas to a boiler fire should occasionally be opened more
widely, might readily lead to the supposition that the gas at sucli times
possesses less heating power and consequently a different composition.
No data as regards results in practice have been obtainable, but in
analyses of gas from various wells in the Pittsburgh region reasons have
been found for supposing that slight fluctuations actually occur in its
composition. With a view to a more complete study of the question, the
experiments described in this paper were carried out.
As regards the character and number of its chief constituents natural
gas differs widely from coal gas, and from gas manufactured at high
temperatures in the various forms of producers. While in artificial gas
unsaturated compounds are present in great variety, natural gas is com-
posed mainly of hydrocarbons of the paraflin series, associated with very
small quantities of nitrogen, carbon dioxide, and water vapor. The
defines, represented mainly by ethylene, are found sometimes in extremely
minute proportion, so minute in fact that quantitative determinations are
a matter of difficulty, although they are qualitatively recognizable when
large volumes of gas are employed. Traces of organic sulphur com-
* Acknowledgment is liere made of aid received from the C. M. Warren Fund
of the American Academy of Arts and Sciences, in conducting the experiments
described in this paper.
72 PROCEEDINGS OF THE AMERICAN ACADEMY.
pounds are also present. Free hydrogen, carbon monoxide, hydrogen
sulphide, and oxygen do not occur.
It is a difficult matter to single out any of the constituents of natural
gas as specially suited for a series of determinations having for their
purpose to ascertain whether fluctuations actually occur in its composi-
tion. One of its constituents — nitrogen — seems to be less prone than
the others to enter into chemical changes locally in the rocks, and there-
fore less likely to undergo diminution in the original gas as it is stored
in the interstices of the sandstones and limestones of the Devonian forma-
tion. It might be proper to begin such a study with nitrogen.
Under the term nitrogen is here understood the incombustible residue
obtained when natural gas is burnt in such manner as to prevent the
admixture of air or oxygen with the products. That the element nitro-
gen is actually contained in natural gas has been frequently shown by
passing the gas over heated magnesium. The magnesium, on being
afterwai'ds moistened, yielded ammonia, recognizable by its odor and
reactions, and indicating the presence of nitrogen in the original gas.
The method employed for the determination of nitrogen consisted in
burning a measured volume of natural gas by passing it over heated
copper oxide. The resulting steam was condensed and the carbon
dioxide absorbed by potassium hydroxide solution, leaving the residual
nitrogen to be measured over mercury. This adaptation of the Dumas
ifcethod has been proposed by Stockmann for the determination of nitro-
gen in coal gas.* Arth has described an apparatus for a similar purpose,
the gas measurements being made over water. f
Description of the Process.
The gas sample was collected in a glass cylinder of from 150 to
400 c. c. capacity, and having stopcocks at both ends. In the drawing
of the apparatus A represents the gas sample vessel with its stopcocks
and capillary endings; Vessels of 150 c. c. proved sufficiently large for
the determination, although in the later work, where it was desirable
to collect a larger volume of the residual incombustible gas, vessels of
350 to 400 c. c. were used. These vessels were calibrated by weighing
the mercury required to fill them. They were filled with gas under
slight excess of pressure (about two ounces). With a view to deter-
mining by calculation the volume at 0° and 760 mm. pressure of the
* Zeitschrift fiir Analyt. Chemie, 1875, p. 46.
t Bull. Soc. Chim., 1897, p. 30.
PHILLIPS.
COMPOSITION OF NATURAL GAS.
73
74 PROCEEDINGS OF THE AMERICAN ACADEMY.
contained gas, the vessels after return to the laboratory were placed
vertically in a large box having glass sides, and in such a position that
the lower stopcock projected through a thick rubber disk set in the
bottom of the box and having an opening just large enough for the stop-
cock to pass through it. The temperature in the box was indicated by
a thermometer, which could at all times be read through the glass panes.
When this temperature had remained constant for at least one hour, the
lower stopcock was opened for an instant allowing the surplus gas to
escape. A glass tube, connected by rubber joint with the lower end of
the stopcock, and just touching, but not dipping under the surface of
water, afforded a simple and safe means of preventing the possibility
of the entrance of air during the equalizing of the pressure. After
opening and shutting the stopcock the pressure of the contained gas was
assumed to be that of the air outside, as indicated by the barometer.
A drop of water was always introduced into the sample vessel before it
was filled, so tliat the gas could be regarded as saturated with moisture
when used for analysis.
The gas vessel was connected with a mercury bottle B, the capillary
tubes at the ends of the cylinder having been filled with mercury befcre
the connections were made. By raising the bottle B and opening the
stopcocks the gas was slowly driven over from A through the glass tube
WD into the porcelain combustion tube HH, which contained a layer
of heated copper oxide forty centimeters long. The porcelain tube was
heated in a combustion furnace (partly shown in the sketch). The pro-
ducts of combustion of the gas passed by way of the glass bulb / and
the stopcocks W and X to the absorpion vessel Q, in which the carbon
dioxide was absorbed by solution of potassium hydroxide. This solu-
tion was delivered over when needed from the bottle 0, by pouring
mercury into the tap funnel U. The residual gas, after the absorption
of the carbon dioxide in Q was then caused to return by the same route
to the cylinder A, by adjusting the levels of the mercury reservoirs B
and R. A third passage of the gas over from ^ to ^ rendered it certain
that all hydrocarbons were completely burnt. The bulb J, and the
combination of bulb with three way tube Z)FF served to condense and
hold as water any steam due to burning of hydrogen of the gas in the
combustion tube, and prevent its return into the heated porcelain tube.
These bulbs proved very necessary to prevent breakage. The stopcock
on the bulb J served to discharge this condensed water. As it was
necessary to rinse the porcelain tube and its glass connections after the
passage of the gas from A to Q, or from Q to A, this was accomplished
PHILLIPS. — COMPOSITION OF NATURAL GAS. 75
by means of a slow stream of carbon dioxide generated by tbe action
of concentrated hydrochloric acid upon calcite in the large jar M. This
jar contained about five pounds of calcite at starting. The escaping
carbon dioxide passed through fragments of calcite in the tube K, and
through water in the bottle G. From this wash-bottle the iias stream
could be deflected to the right hand end of the combustion tube, as seen
in the sketch, entering this tube by the stopcock Z in rinsing gas into
the absorption vessel Q, or to the left, by way of the stopcock ZZ in
rinsing gas back into the vessel A. In this manner the rinsing out of
the products of combustion into either A or Q could be made very com-
plete. The tube F, descending into water in the open cylinder FF,
served as a safety valve to permit the escape of surplus carbon dioxide
from the generator, and thus any unsafe pressure in the apparatus was
avoided. The carbon dioxide could, therefore, be generated freely and
utilized only in so far as it was needed. It is to be observed that the
long glass tube connecting the stopcocks Z and ZZ lies in the, same
horizontal plane with the combustion tube HH, although in the sketch
it appears lower, in order that the arrangement of parts may be rendered
clear. The carbon dioxide having been absorbed in Q, the residual gas
was ready for measurement which was accomplished in the eudiometer
P over mercury. This eudiometer was of 100 c. c. capacity, being made
very short as shown (about forty centimeters long). It was graduated
only in its uppermost and lowermost portions. This is a convenient form
of eudiometer where mercury is used, as it avoids the great pressure of
a high mercury column, and consequent danger of leakage through the
stopcocks, so common where a high column of mercury is used. It is
easily seen that by lowering sufficiently the bottle R any volume of gas,
up to 100 c. c. may be readily measured. The combustion furnace stood
somewhat higher than the upper end of the eudiometer. It was sur-
rounded closely by sheet iron sides, which served to carry upward the-
waste heat. No difficulty was experienced in maintaining a constant
temperature about the eudiometer, as indicated by a thermometer read-
ing to 0.05°, placed in contact with its sides. By reason of the strong
upward draught produced by a sheet iron box placed around and in close
contact with the sides and ends of an ordinary combustion furnace, much
may be done towards diminishing the discomfort of the experimenter,
while the temperature of the interior of the furnace is increased, and
the ends of the tube are more readily kept cool. The tube tt, dipping
into water in the beaker T, serves to discharge the absorption vessel
after a determination. In a similar manner the eudiometer may be
76 PROCEEDINGS OF THE AMERICAN ACADEMY.
discharged of its contents by the tube ss dipping into the beaker S.
The tube B serves to discharge carbon dioxide through a Liebig's bulb
having a water seal, when the apparatus is being cleared of air prepara-
tory to a determination. The various forms of stopcocks used are suffi-
ciently indicated by the sketch. The time required for a nitrogen
determination was from one and a half to two hours when 150 c. c. of gas
were used. Two or three days were sometimes occupied in expelling
the last traces of air from the apparatus by the slowly passing carbon
dioxide stream preparatory to a series of determinations.
The potassium hydroxide solution was of 1.258 specific gravity, as
recommended by Kreusler, * and the measurement of the nitrogen was
always made with a little of the fresh solution resting upon the mercury
in the eudiometer. The mercury reservoirs, H, M, were attached to sup-
ports sliding vertically in wooden frames (not shown), and the reservoir
connected with the eudiometer could be adjusted by screw movement so
as to bring the mercury in the reservoir and that in the eudiometer to the
same level. Readings were made by an accurate cathetometer. The
pressures of mercury columns were all calculated at 0°.
A very high temperature was found to be necessary for the complete
combustion of the hydrocarbons of natural gas, under the conditions of
the method, the excess of carbon dioxide produced causing retardation.
No difficulty was experienced, however, as repeated tests demonstrated
that the residual gas did not contain carbon monoxide or free hydrogen.
Moreover, it was repeatedly found that on passage of the residual gas for
a fourth and fifth time through the copper oxide, and absorption of the
carbon dioxide, no further reduction of volume was produced. The con-
stantly increasing amount of reduced copper in the porcelain tube, as
combustion goes on, serves to prevent the escape undecomposed of any
oxides of nitrogen, should such compounds tend to form during the
process.
In order to procure pure carbon dioxide for the Dumas method of
nitrogen determination in organic bodies, it has been recommended that
the marble to be used be first pulverized and then boiled in water before
its carbon dioxide is liberated by the action of an acid. Bernthsen f
frees the pores of the marble from air by exhaustion with an air pump.
In experiments tried with a view to producing pure carbon dioxide these
methods have not always proved satisfactory. The carbon dioxide stored
* Zeitschrift fiir Analyt. Cliemie, 1885, p. 445.
t Zeitschrift fur Analyt. Chemie, 1882, p. 63.
PHILLIPS. — COMPOSITION OF NATURAL GAS. 77
in liquid form in steel cylinders was tried. It was found, however, that the
gas leaves a considerable volume of unabsorbed residue when treated witii
the solution of a caustic alkali. Sodium carbonate was tried instead of
marble. A hot saturated solution of the salt was allowed to crystallize
in the generator and the mother liquid poured off. The carbon dioxide
produced by action of an acid is purer than that from marble, but the
evolution of the gas is tumultuous and uncontrollable. Experiments
were tried with marble from various localities, but with little success. A
marble from Tate, Georgia, was found to yield very pure carbon dioxide
after it had been coarsely pulverised and well boiled in water. Another
sample of apparently the same rock, from the same locality, yielded
after similar treatment a small residue of gas unabsorbed by jwtassium
liydroxide solution. A calcite from Lampasas, Texas, in translucent
cleavable crystals, was found to yield satisfactory results. The mineral
was crushed coarsely, boiled for six hours in water, and then transferred
with a portion of the boiled water to the generator. The carbon dioxide
evolved proved to be very pure. Experience has shown that no reliance
can be placed upon marble or calcite as a source of carbon dioxide
because a specimen of apparently the same mineral known to come from
the same locality has proved satisfactory. Every batch must be sepa-
rately tested as regards the purity of the carbon dioxide which it
evolves.
To overcome the danger of the action of strong alkali upon stopcocks
in such work, it was necessary to use an unsaponifiable lubricant. After
numerous trials it was found that a mixture, consisting of 70 parts
melted rubber and 30 parts unbleached beeswax, softened with a little
vaseline, served the purpose quite well, protecting the stopcocks com-
pletely. More than 150 determinations have been made in the same
eudiometer without accident to stopcocks.
Some difficulty was experienced in expelling air from the copper oxide
when the apparatus was being prepared for work. In beginning a series
of determinations several days were often required for the purpose. The
porcelain tube was strongly heated, while a slow stream of carbon dioxide
was maintained, and the copper oxide was not considered to be in proper
condition for use until the escaping carbon dioxide was found to be ab-
sorbed without residue by potassium hydroxide solution. About 300 c. c.
of the escaping gas were used for the trial. When the copper oxide was
freed from air, determinations of nitrogen in natural gas could follow
each other until from experience it was shown to be necessary to reoxi-
dize the partially reduced copper oxide. About ten determinations
(8 PROCEEDINGS OF THE AMERICAN ACADEMY.
could be made before this reoxidation was needed. Regeneration of the
copper oxide was effected by drawing air through the apparatus while
hot. It was found that the copper oxide when once ioapreguated with
carbon dioxide while strongly heated could be reoxidized by an air current
with very little tendency to occlusion of air. After the passage of air
for a few hours, a stream of carbon dioxide readily expelled the remain-
ing air and the apparatus was again ready for further determinations.
But if the copper oxide was allowed to cool in contact with air, much
time was lost in removing the air by the current of carbon dioxide, even
wlien strong heat was applied during the process. It appears that little
or no occlusion of air takes j^lace when the copper oxide is first impreg-
nated with carbon dioxide^ while the same substance exposed to air in
the cold occludes the air and hoUls it with much persistence.
Selection of Samples of Gas.
Many'of the wells are drilled through several different gas producing
sand rocks, separated by deep layers of impervious shales and other strata.
The gas from these different sands mingles, and tlie product flowing from
a single well is often a mixture of gas from formations many hundred
feet apart in the vertical scale. It was attempted as far as possible in
the present work to secure samples from wells yielding gas from a single
sand rock. It was desirable that the samples be taken as far as possible
from wells situated at no great distance from the laboratory, in order
that as short a time as possible should elapse between the collection of
the sample and the commencement of the analysis.
Results of Determinations of Nitrogen.
1. Gas well at Shields, 14 miles west of Pittsburgh. This well was
drilled in 1892, and yields gas exclusively from the Fourth Sand, which
was reached in drilling at a depth of 1,7G0 feet.
Date of Collection of Samples. Percentage of Nitrogen found.
August 0, 1896 (1) 1.25
(2) 1.26
February 5, 1897 (1) 2.70
(2) 2.67
(3) 2.68
April 6, 1897 (1) 1.79
(2) 1.80
April 20, 1897 (1) 1.85
(2) 1.85
Junel, 1898 (1) 1.10
(2) 1.10
PHILLIPS. — COMPOSITION OF NATURAL GAS. 79
2. Well on the Auderson farm at Sewickley, 12^ miles west from
Pittsburgh. This well was drilled in 1894, and yields gas exclusively
from the Third Sand, which was found at a depth of 1,850 feet.
Date of Collection of Samples. Percentage of Nitrogen found.
July 7, 1896 (1) 2.48
(2) 2.50
August 14, 1896 (1) 1.72
(2) 1.71
March 22, 1897 (1) 2.11
(2) 2.10
3. Well on the Mliller farm at Glenfield, Pa., 9^ miles west from
Pittsburgh. Drilled in 1887. The gas is produced mainly from the
Fourth Sand, although a little gas finds access to this well from the
upper sands. The Fourth Sand was reached in this well at a depth of
1,800 feet.
Date of Collection of Samples. Percentage of Nitrogen found.
July 27, 1896 (1) 1.52
August 17, 1896 (1) 1.69
(2) 1.69
April 2, 1897 (1) 3.24
(2) 3.25
April 16, 1897 (1) 3.23
(2) 3.21
April 28, 1897 (1) 3.23
(2) 3.20
May 4, 1897 . (1) 3.20
(2) 3.22
April 8, 1898 (1) 2.10
(2) 2.10
April 21, 1898 (1) 2.12
(2) Lost. :
June 9, 1898 (1) 1.27
(2) 1.30
June 13, 1898 ....... (1) 1.23
(2) 1.22
4. Well on the Bayley farm, Neville Island in the Ohio River, 6^
miles west from Pittsburgh. The well was drilled in 1892. Gas is
80 PROCEEDINGS OF THE AMERICAN ACADEMY.
derived from tlie " thirty-foot " sand alone, which rock was found at
a depth of 1,510 feet.
Date of Collec tion of Samples. Percentage of Nitrogen found.
August 1, 1896 (1) 1.46
(2) 1.48
April 12, 1897 (1) 2.10
(2) 2.11
April 24, 1897 (1) 2.10
(2) 2.09
May 27, 1898 (1) 1.49
(2) 1.48
5. Well on the Hamilton farm, Neville Island in the Ohio River,
7 miles west from Pittsburgh. The gas was derived from the Third
Sand, which was reached at a depth of 1,580 feet, although there was a
slight flow of gas from the Fifth Sand, which was reached at 1,729 feet.
The well was drilled during the spring of 1898. The first sample of
gas below mentioned was taken about 24 hours after the gas had been
turned into the mains.
Date of Collection of Samples. Percentage of Nitrogen found.
May 16, 1898 (1) 1.78
(2) 1.76
May 19, 1898 (1) 1.74
(2) 1.74
6. Well on the King farm at IMurraysville, 18 miles east from Pitts-
burgh. The gas is derived from the 3Iurraysville Sand at a depth of
1,336 feet. The well was drilled in 1887. Analyses of samples taken
on a single date are presented, but this gas is of interest in the present
connection, as the Murraysville gas field was among the earliest explored.
The pressure in the rock has fallen from about 500 pounds to so low a
point that at certain seasons of the year, when the demand is greatest,
the gas is' regularly pumped from the wells into the mains. Pumping
was not in progress when the samples were taken.
Determinations of nitrogen in samples collected June 6, 1898, gave: —
(1) 1.29.
(2) 1.28.
7. Another well on the same farm at Murraysville yields gas exclu-
sively from a sand about 100 feet deeper. Gas collected on the same
PHILLIPS. — COMPOSITION- OF NATURAL GAS. 81
date from this well was found to contain the following percentages of
nitrogen : —
(1) 1.38.
(2) 1.40.
8. "Well on the Souder farm, 12 miles east of Buffalo. This well
was drilled in 1892, and yields gas from the Trenton Limestone exclu-
sively. Mr. E. Coste, Engineer for the Provincial Natural Gas Com-
pany of Buffalo, is authority for the statement that the drill passed
through the base of the Trenton Limestone 30 feet below the point at
which the gas was obtained. Hence the horizon from which this gas
comes is extremely low in the geological scale as compared with the
productive formations of the Pennsylvania gas fields.
Samples were collected at this well on September 3, 1896. The
results of the analyses are as follows : —
(1) 4.57.
(2) 4.55.
9. Well on the Reinhart farm, near Sherkston, 10 miles east from
Buffalo. In this well the gas is derived solely from the Clinton Lime-
stone, which was reached at a depth of 590 feet. Examinations were
made of samples collected on September 1, 1896. Percentage of nitro-
gen found : —
(1) 3.64.
(2) 3.G1.
10. "Well No. 12 of the Provincial Natural Gas Company at Sherk-
ston, Canada. Gas is derived solely from the Medina Sandstone, which
was reached at a depth of 850 feet.
The samples were collected on September 1, 1896. Percentage of
nitrogen found : —
(1) 5.17.
(2) 5.10.
In the case of all these Canadian wells the gas samples were shipped
at once to the laboratory, and the determinations made without delay.
All the samples so far mentioned were taken directly at the wells.
Several determinations of nitrogen have been made in the case of natural
gas from the mains supplying Allegheny. This gas was derived from
various wells scattered through a region of considerable area.
VOL. XXXIV. — 6
82 PROCEEDINGS OF THE AMERICAN ACADEMY.
Date of Collection of Samples. Percentage of Nitrogen found.
August 25, 1896 (1) 1.48
(2) 1.50
August 26, 1896 (1) 1.52
(2) 1.52
August 28, 1896 (1) 1.56
(2) 1.54
August 28, 1896 (1) 1.52
(2) 1.52
February 28, 1898 (1) 1.93
(2) 1.99
March 9, 1898 (1) 1.30
(2) 1.31
Careful tests for oxygen were made in the case of the gas from the
mains, as there was a possibility of access of air to the gas. The method
of testing consisted in causing the gas to bubble through a solution of
manganous sulphate, to which a little sodium hydroxide had been added.
A change in color of the manganous hydroxide which was precipitated
would have indicated oxygen. Tests made in this way were continued
frequently for an entire day while the nitrogen determinations were in
progress, but no oxygen was found, and iience no air could have gained
access to the gas.
There seems to be reason for the assertion that fluctuations occur in
the composition of natural gas, but until the study of the subject is
carried further, and more complete data obtained, no attempt can be
made to connect such fluctuations with any known facts as to the geology
of gas.
There is some little evidence for supposing that gas from the deeper
horizons is richer in nitro<j:en, and that the older productive wells yield
gas containing a little less nitrogen, but such may i^rove not to be the
case when more data are at hand.
It seemed to be of interest to subject the incombustible gas residue,
obtained in the preceding work, to further study. Portions of this
residue were mixed with oxygen and subjected to the action of electric
sparks. The experiments are still in hand, and their results will be \jve-
sented in a later paper. It may be mentioned here that the gas subjected
to this treatment yields oxides of nitrogen, and in presence of caustic
alkali undergoes a considerable shrinkage in volume.
If natural gas occurs in liquefied form in the rocks owing to the
PHILLIPS. — COMPOSITION OF NATURAL GAS. 83
pressure to which it is there subjected, it is probable that when a drill
taps the gas-beariug rock, causing relief of pressure, the more volatile
among the constituents of the liquefied gas would escape in relatively
laro-er proportion at the outset. The process occurring would be of the
nature of fractional distillation, and would tend to the production of a
gas especially rich in the most volatile constituent ; but the most volatile
constituent of natural gas is nitrogen, since all of the hydrocarbons and
carbon dioxide would be more readily liquefied than nitrogen. The first
yield of a gas well should therefore contain a higher proportion, relatively,
of nitrogen, and this nitrogen should gradually diminish as the liquefied
gas continued to evaporate. There would result after a time a gaseous
mixture containing less and less of nitrogen, and when the reduction of
pressure had progressed so far as to permit of the conversion of the
least volatile of the constituents into gas, the proportion of nitrogen in
the escaping gas would become constant, because the process would then
be one of outflow of a gas mixture, and not of volatilization of an
extremely low boiling liquid whose constituents have different boiling
points.
If natural gas occurs liquefied in the rocks, we should expect to find that
the newly drilled wells yield at first a gas relatively richer in nitrogen.
The rock pressure in the Pennsylvania natural gas fields has in rare
instances attained to 1,000 pounds per square inch. Some cases are
reported in Northern New York State where rock pressures of 1,500 and
2,000 pounds have been measured. Such pressures are probably the
highest ever observed in any natural gas field, but such pressures would
be insufficient to liquefy natural gas. Should further determinations of
nitrogen furnish evidence that a gradual diminution of the percentage
of nitrogen is in progress, support would be given to the view that
natural gas occurs in liquefied form in the rocks. The great productive-
ness of many single wells is, upon this supposition, more readily ex-
plained, for in the interstices of the rock might be stored a much larger
quantity of gas if in liquefied form. The study of the composition of
natural gas will have much to do with determining this point.
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. No. 4. — December, 1S9S.
CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.
SOME ELECTROCHEMICAL AND THERMO CHEMICAL
RELATIONS OF ZINC AND CADMIUM AMALGAiMS.
By Theodore William Richards and Gilbert Newton Lewis.
CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF
HARVARD COLLEGE.
SOME ELECTROCHEMICAL AND THERMOCHEMICAL
RELATIONS OF ZINC AND CADMIUM AMALGAMS.
By Theodore William Richards and Gilbert Newton Lewis.
Presented October 12, 1898. Received October 17, 1898.
Introduction.
The nature of amalgams, although a matter of much interest, especially
in the light of the modern theory of solutions, is still obscure. The
subject has been studied from the standpoint of electrochemistry by
several investigators, notably by Meyer, * who observed the electromo-
tive forces of cells of the following type: — Dilute zinc amalgam of
concentration x ; Solution of zinc salt ; Dilute zinc amalgam of con-
centration y. Since the total change in a cell of this kind consists in
the transfer of metal between amalgams of different concentrations, the
electrical energy obtained is equal to the maximum osmotic work obtain-
able by the process. On the assumption that the zinc in the dilute
amalgam obeys the laws of dilute solution, Meyer derived the following
expression : —
^=L908|.riog-|; (1)
in which E represents electromotive force ; ^, the electrochemical equiv-
alent, in grams, of the metal carried in one second from one amalgam
to the other ; J/, the molecular weight of the metal in the amalgam ;
T^ the absolute temperature ; log — , the common logarithm of the ratio
of concentration of the amalgams. By comparing with this formula the
experimental results, Meyer showed with sufHcient exactness that in
dilute amalgams the molecules of the metals studied by him are mon-
atomic. If, on the other hand, the atomic weight is substituted for M
in the formula, the agreement between the calculated and his observed
electromotive forces is not close enough to show whether the dilute
* Zeitschr. phys. Chera., VIL 477 (1891),
88 PROCEEDINGS OP THE AMERICAN ACADEMY.
amalgams follow rigidly the laws of dilute solutions, oa account of wide
deviations in the observed values.
It has been the object of the present research to determine the electro-
motive forces of cells of the above type at varying temperatures, and
with amalgams of all degrees of concentration, and also of similar cells
in which one amalgam is replaced by the pure metal. Zinc and cad-
mium as metals, and normal solutions of their sulphates as electrolytes,
were chosen as best adapted to the purpose. The experimental results
were studied in relation to the two following equations: —
£= — Tin — = .0000^0 T log -^. (2)
dE _E Q
Equation (2) is the simpler form that (1) assumes when the atomic
weight is substituted for M. E is the observed electromotive force ; R
is the gas constant ; n, the valence of the metal in question (« = 2 in the
case of zinc and cadmium) ; e^ is the quantity of electricity in coulombs
which is carried by one gram-equivalent; In — is the natural logarithm
of the concentration ratio. A comparison of this formula with the
experimental results shows the extent of applicability of the laws of
dilute solutions to amalgams (assuming the molecule of the metal to be
monatomic when amalgamated).
Equation (3) is the Ilelmholtz equation for the temperature coefficient
of a cell, where Q is the heat given off by the cell during a transfer of
n gram-equivalents. In the cells under consideration the only change
produced by the current is the transfer of metal from the solid electrode
to the amalgam, or from one amalgam to another more dilute. Q then
represents either the heat of amalgamation of w gram-ecjuivalents (one
gram-atom) of the metal, or the heat of dilution of an amalgam contain-
ing one gram-atom. The use of this equation permits the calculation of
these quantities from the temperature coefficient of the cell.*
Since the heat capacity of the amalgam is approximately the sum
of the heat capacities of its constituents, the heat of amalgamation is
Q
practically constant. If we place a constant, h, in place of — , equation
(3) becomes
* Compare the interesting paper by Bugarszky, Zeit. anorg. Chem., XIV. 145.
RICHARDS AND LEWIS. — ZINC AND CADMIUM AMALGAMS. 89
dE_E-h dE _dT
integrating, In {E-K) = lnT-\- 0= hiKT,
where C is the integration constant and C = InK, hence
E = KT -{- h. (4)
Upon comparing this result with equation (3), which can be written
E — T -\ one sees that the outcome of this reasoning is simply
d T ne^
dE
the proof that if Q is constant -7^, (the temperature coefficient of the
electromotive force) is also a constant. Thus the electromotive force of
any cell of this type should be a linear function of the temperature.
Materials and Apparatus.
The materials used in this research were of known purity. The
mercury had been twice distilled in vacuo. The zinc and cadmium were
prepared by electrolysis from chemicall}^ pure salts. The zinc sulphate
had been prepared in this laboratory for atomic weight investigation.
The cadmium sulphate was prepared by dissolving the electrolyzed metal
in pure sulphuric acid, and crystallizing twice in order to free it from the
excess of acid. The amalgams were prepared by mixing weighed
amounts of mercury and the metal used and diluting as desired. The
amalgams wore weighed and kept under solutions of their sulphates to
prevent oxidation.
Both cadmium and zinc, when electrolyzed from ammoniacal solutions
of their sulphates, separate in tree-like forms, varying according to the
conditions of electrolysis from large and distinct crystals to finely divided
spongy masses. The latter form is produced by the stronger currents
and greater dilutions.
The metal thus made, after suitable washing, was sometimes dried
with alcohol and ether and converted into amalgam, or sometimes used
at once in the jDure state in the trial cells. The method of using this
spongy material as an electrode is described later, on page 90.
The vessels in which the measurements were made consisted of large
H tubes, with their lower extremities drawn out and turned upwards.
Platinum wires run in through these fine tubes established the connection
with the galvanometer. It was occasionally convenient to use a double
H tube, in which one electrode could be connected through the same
electrolyte with either of two electrodes.
90
PROCEEDINGS OF THE AMERICAN ACADEMY.
Diagram of Apparatus.
The glass part at the left of the diagram is drawn about one third of the actual
size. The remainder of tlie apparatus is not drawn to scale, being annexed
merely to show the connections. For explanation, see page 91.
The cell was immersed in a thermostat which could be maintained
constant at any desired temperature within one tenth of a degree.
The electromotive forces were measured by means of an astatic gal-
vanometer according to the Poggendorff method. An Ostwald * " poten-
tial box," the resistances of which were carefully verified, was connected
* Ostwald, Hand- und Hilfsbuch, p. 252.
RICHARDS AND LEWIS. — ZINC AND CADMIUM AMALGAMS. 91
with a Leclanche cell through such external resistance as to maintain a
total fall of potential of 0.1 volt. This was adjusted before each series
of measurements by comparison with a standardized Helmholtz cell which
remained constant in potential throughout the course of the experiments.
The direct measurement of potential as far as three figures when de-
sirable was made possible by a device, the principle of which is shown
in the accompanying diagram, ^i'^ represents a potential box containing
three sets of resistances: AB, 9 times 100 ohms; CD, 11 times 10
ohms; JSF, 10 times 2 ohms. These resistances are fitted with pegs
for connection as in the ordinary potential box. ABCD forms a simple
circuit between the terminals A and D. EF'xi, a shunt whose terminals
can be connected with movable caps to the pegs in the row CD. If
they are connected to two pegs, G and H, for example, between which
there is 20 ohms resistance, the current between the pegs is evenly
divided by the shunt. Thus arranged the total resistance of (7— Z^ is
100 ohms, that of A—B is 900 ohms, making 1,000 ohms for the whole
box. A definite potential is maintained between A and D. \i for
example this is 0.1 volt, then the fall between two adjacent pegs is in
^5 0.01 volt; in CD 0.001 volt (except between G and H, where
the total fall is 0.001 volt); in EF 0.0001 volt. The cell, P, whose
potential is to be measured, is connected through the galvanometer to
the box by means of the caps /and K. Its potential is compensated, and
there is no current through the galvanometer when it is equal to the
total fall from B to I, from G to H, and from E to K. This value can
be read directly from the box ; thus the arrangement in the figure indi-
cates a potential of 0.0374 volts. Of course even the next decimal place
may be estimated from the deflections of the galvanometer when two
adjacent pegs in EF on each side of the true point are connected in
succession, or might be measured directly by yet another shunt similar to
E F, with a total resistance of four ohms. In this case E F mxxsi be
provided with an extra peg.
The external resistance of 13,000 ohms was calibrated during the
course of the experiments. Owing to an error found in this box, it was
necessary to correct the direct readings of potential of some of the early
cadmium amalgam cells by multiplying with a constant factor. The
corrected readings are given below without comment.
Results.
Four classes of cells were studied: — (1) electrodes of cadmium
amalgam of different concentration ; (2) electrodes of zinc amalf^am of
92
PROCEEDINGS OP THE AMERICAN ACADEMY.
different concentration ; (3) solid cadmium electrode opposed to cadmium
amalgam ; (4) solid zinc electrode opposed to zinc amalgam.
Class 1.
Table I. gives the values found for the electromotive forces of cells
with electrodes of cadmium amalgams of different concentration. The
lower extremity of each arm of the H tube was filled with amalgam to
a depth of two or three centimeters, so that a fairly large volume might
be present. Measurements were made at such temperatures that the
amalgams were wholly liquid, for their partial solidification is an insidious
cause of error capable of producing serious results, especially in this
case of cadmium. The three per cent amalgam begins to freeze at about
0°, becoming wholly solid at a slightly lower temperature (about — 3°) ;
while the nine per cent amalgam is not wholly liquid until 45° is reached.
For this reason measurements 10 in Table I. had to be made at high
temperatures. Normal cadmic sulphate was the electrolyte.
TABLE T. Cadmidm Amalgams.
fi
<-j
t
E (Obs.)
.E(Calc)
E
T
1
.V
^T
3
30
.01470
.0143
.0000485
2
i
i7
3
30
.01470
.0143
.0000485
3
1
3
h
3
30
.01460
.0143
.0000482
.01320
.0129
.0000483
4
i
1
a
3
30
.01445
.0143
.0000477
.01303
.0129
.0000477
5
1
1
3
3
30
.01460
.0143
.0000482
.01313
.0129
.0000481
6
1
i
3
30
.01450
.0143
.0000479
.01313
.0129
.0000481
7
1
1
3
8
30
.014.50
.0143
.0000479
.01306
.0129
.0000478
8
1
1
3
30
.014.50
.0143
.0000479
.01306
.0129
.0000478
9
3
1
3
30
.01470
.0143
.0000485
10
9
3
3
72
.01993
.0163
.0000578
60
.01915
.0157
.0000575
RICHARDS AND LEWIS. — ZINC AND CADMIUM AMALGAMS.
93
In the table, Cj and Cg represent the percentage by weight of cadmium
in the amalgam, which in the case of dilute amalgams is proportional
to the concentration by volume ; t is the Centigrade temperature. The
columns under E give the observed electromotive forces and those calcu-
late,d from equation (2).
It was noticed by Meyer that the electromotive force of a cell of this
kind increases rapidly upon standing. A similar eflfect was noticed by
Jaeger* with solid cadmium amalgams. Although no explanation can
be given of this phenomenon, one may prevent it by using as electrolyte
a solution which has remained standing in contact with cadmium amal-
gam for several weeks before being used. The constancy thus reached
permits much greater accuracy than could otherwise be obtained.
A study of the data of Table I. shows that the behavior of the last cell
containing nine per cent amalgam differs materially from the rest. Re-
garding the electromotive forces of the other cells, it is to be noticed that,
(1 ) at the same temperature they are all equal within limits of .0003 volt,
therefore the potential depends on the ratio of c^ to Cj, and not on their
absolute values; (2) they are proportional to the absolute temperature ;
(3) they are uniformly higher than the values calculated from the
formula, ranging from one percent to three per cent too high. In the first
two respects the amalgams obey rigidly the laws of dilute solutions. The
small apparent deviation from these laws, indicated by the difference
between the observed and calculated values, is therefore probably not a
real deviation, but the effect of some slight side reaction in the cell.
The wide departure in the case of the last cell, on the other hand,
shows that in amalgams of concentration as great as nine per cent there
is a considerable deviation from the laws of dilute solution.
Theoretically, the nature of the anion of the electrolyte or the con-
centration of the kation should be without effect upon the rcfult. This
TABLE n. Varying Electrolytes.
f\
«-2
/
E
CdS04('^) . . .
1
\
30
.02925
CdS04(«J . . .
1
1
30
.02915
Cdl. ('{) ....
1
i
30
.02920
* Wied. Ann., LXV. 106, 1898.
94
PROCEEDINGS OF THE AMERICAN ACADEMY.
prediction was verified by experiment. Tliree similar cells were set up
with different electrolytes, (1) normal cadmium sulphate, (2) tenth nor-
mal cadmium sulphate, (3) normal cadmium iodide. The results are
shown in Table II. Theory demands 0.0286 instead of 0.0292.
Cl(xss 2.
The cells with electrodes of zinc amalgams were less constant in
potential than those with cadmium, and the measurements were less
trustworthy. Table III. gives the results with four cells of this tyjie.
The remarks on the preceding case apply here.
TABLE III. Zinc Amalgams.
«i
^2
t
E (Obs. )
J?(Calc.)
E
T
1
1
\
9
30
.02890
.0286
.0000954
.02010
.0258
.0000956
2
\
1
5?
9
30
.029'J0
.0280
.0000064
.02025
.0258
.0000965
3
\
\
3
30
.0]-12o
.0143
.0000470
.01280
.0120
.0000469
4
i
h
3
SO
.01515
.0143
.0000500
.01305
.0129
.0000500
Class 3.
The measurement of the contact-potential of solid electrodes has
always been subject to considerable uncertainty, due to accidents of
crystallization, condition of surface, polarization, and other unknown
causes. It seemed possible that by sufficiently increasing the extent and
diversity of the surface an electrode might be obtained whose potential
would be the mean of a large number of different values, and therefore
constant. An electrode consisting of a quantity of finely divided metal
perhaps a centimeter in depth, packed loosely around a sealed-in platinum
wire seemed likely to satisfy the necessary conditions most nearly, and
experiments made with zinc and cadmium electrodes of this sort yielded
remarkably satisfactory results.
RICHARDS AND LEWIS. — ZINC AND CADMIUM AMALGAMS.
95
In order to justify the use of this solid electrode one must show that
it always gives the same potential, and that this is equal to the true
potential existing between metal and solution. The first experiments
were made with a cell whose electrolyte was cadmium sulphate and
whose electrodes consisted of electrolyzed cadmium, of medium fineness,
which had been washed successively in dilute suli^huric acid, distilled
water, and absolute alcohol, and then dried. The two electrodes being
exactly similar, the electromotive force of the cell should be zero. Tlie
actual electromotive force, therefore, indicates the amount of deviation
in potential of electrodes of this kind. Several such cells were measured.
The largest electromotive force found was .0004 volt, the majority being
about .0001 to .0002 volt. Better results were obtained with metal
which, instead of being dried, was washed in the electrolyte. The differ-
ence of potential under these circumstances never exceeded .0001 volt
when proper care was used in preparation. Moreover no greater difference
was found when two electrodes made of entirely different samples of
electrolyzed cadmium were used. Since, therefore, the same potential
is obtained from such electrodes, whether the metal be in a finely divided
spongy state, or consist of a coarser network of crystals, this may safely
be considered the true potential of metallic cadmium. Similar experi-
ments were made with zinc, with equally satisfactory results. The data
given below illustrate the constancy of these electrodes. The electrodes
were usually not tested for any considerable length of time, the cells
being prepared anew for each series of observations, which only lasted
a few hours. Table IV. gives the electromotive forces of cells where
finely divided metallic cadmium was thus pitted against a dilute cad-
mium amalgam containing one per cent of cadmium. '
TABLE IV. Cadmium versus Amalgam.
E at 0"^
E at 24.45°
Cell No. 1 . . .
Cell No. 2 . . .
Cell No. 3 . . .
Cell No. 4 . . .
0.068.3-5
O.OfiSSO
0.06845
0.06835
0.07345
0.07345
0.073G0
0.07350
Average . . .
0.0683G
0.07350
96 PROCEEDINGS OP THE AMERICAN ACADEMY.
On referring to equation (4),
E = KT -V h,
it is obvious that we have here data for calcuhiting h ; for
dE 0.07350 — 0.06836
dT 24.45
= 0.0002102.
Upon this basis, h = 0.01096. But according to our original definition
h = — , or Q = hneQ. In this case « = 2, and Bq = 23040, expressed
in such units that if E is in volts, CqE will be in gram calories. Hence
we have
Q = +505 small calories.
This result represents the small quantity of heat given off when one
gram atom (or 112.3 grams) of cadmium is dissolved in 11,100 grams of
mercury. The constancy of the quotient in the last column of Table I.
shows that further dilution does not increase this heat evolution.
Class 4.
A further application of equation (4) is presented in Tables V. and
VI., which give the electromotive forces of cells whose electrodes were of
zinc and one per cent zinc amalgam. The cells were kept at each tem-
perature for ten minutes after constancy was reached.
Between the observations (4) and (5) in Table V. the thermostat
was raised to higher temperatures, but above 37° hydrogen bubbles weie
evolved at the solid electrodes and the electromotive force became incon-
stant. The thermostat was then cooled to 34.5° and the electrodes were
stirred and shaken to drive off the accumulated hydrogen. The readings
were then resumed.
In Table VI. between observations (7) and (8) forty-eight hours inter-
vened ; so that the later results are not so trustworthy as the earlier
ones. These observations are not used in the calculation, but are given
merely in order to show that even a long immersion in the electrolyte
does not seem to affect very greatly the condition of the spongy zinc.
The equation, based upon the starred observations in Table V., is as
follows : —
E = 0.0002000 T — 0.04895.
Everywhere between 0° and 36° the values found agree almost exactly
with the values calculated from this formula.
Here, as before, Q = hneQ ; yi = 2 ; e^ — 23040 ; but in this case
RICHARDS AND LEWIS.
ZINC AND CADMIUM AMALGAMS.
97
h = —.04895. Therefore Q = —2255. That is, a gram atom of zinc
(65.4 grams) takes up 2255 gram calories iu dissolving iu 6500 grams
of mercury, or to greater dilution.
This method of determining thermal quantities is evidently one of
great accuracy and convenience in cases where it is applicable. The
data concerning the heats of amalgamation of zinc and cadmium which
have previously been obtained are meagre. With these, however, the
present results are in agreement. Thus Obach* found a cooling effect
when zinc was amalgamated, a warming when cadmium was amalgamated.
P'uvre t found for the heat of solution of amalgamated zinc 39.43 Kg-
cal., for that of pure zinc 37.34 Kg-cal. The difference between the
two represents the heat of amalgamation of zinc or — 2100g-cal. The
agreement of our result — 2255 with this is striking.
The difference iu potential between the solid metal and its saturated
amalgam should be emphasized. It is often stated that the potential of
TABLE V. Zinc veksds Amalgam.
t
E (Obs.)
E (Calc.)
1
30.0
(a)
.01175
(b)
.01160
.01165
2
0.0
.005701
.00560t
.005651
3
30.0
.01 not
.OllGOt
.011651
4
36.2
.01285
.01289
5
34.5
.01270
.01200
.01255
6
32.6
.01230
.01215
.01217
7
30.0
.01170
.01160
.01165
8
28.0
.01125
.01120
.01125
9
26.6
.01095
.01085
.01097
10
26.7
.01100
.01085
.01099
11
23.8
.01045
.01030
.01011
12
15.7
.00885
.00870
.00879
13
16.0
.00890
.00875
.00885
* Jahn, Grundriss der Elektrocheniie, p. 8.
J These values were taken as the basis of the formula.
VOL XXXIV. — 7
t Ibid.
98
PROCEEDINGS OP THE AMERICAN ACADEMY.
TABLE VI. Zinc versus Amalgam.
t
E (.Obs.)
E (Calc.)
1
2G.0
(0)
.01075
.01075
.01085
2
25.7
.01070
.01080
.01079
3
0.0
.00580
.00595
.00565
4
25.3
.01060
.01085
.01071
5
25.4
.01065
.01085
.01073
6
21.5
.00985
.01010
.00995
7
19.4
.00960
.00980
.00953
8
24.2
.01080
.01080
.01049
9
24.4
.01080
.01080
.01053
10
0.0
.00600
.00600
.00565
11
24.5
.OlOSO
.01085
.01055
12
0.0
.00595
.00605
.00565
13
24.8
.01080
.01100
.01061
a saturated amalgam may be considered the potential of the pure metal.
In the case of zinc this is true within a few thousandths of a volt ; in the
case of cadmium the difference between the solid metal and the saturated
amalgam is 0.45 volt at 30°, and .054 volt at 0°.
Summary.
The main points of the present paper may be summarized as fol-
lows : —
(1) A convenient method of measuring electromotive force directly to
any desired number of decimal places is described.
(2) Cadmium amalgams as far as concentrations of three per cent and
zinc amalgams to concentrations of at least one per cent obey closely the
laws of dilute solution.
(3) The use of the Helmholtz equation for the temperature coefficient
of a cell offers in these cases an accurate method of determining thermal
quantities.
EICHARDS AND LEWIS. ZINC AND CADMIUM AMALGAMS. 99
(4) The heat of amalgamation of cadmium is thus found to be +505
gram calories.
(5) The heat of amalgamation of zinc as thus found is — 2255 gram
calories.
(6) A solid electrode composed of finely divided electrolyzed metal
gives a very constant and reliable potential.
(7) In the case of cadmium the contact-potential given by the satu-
rated amalgam in reversible relation to an electrolyte differs by the
twentieth of a volt from that given by the metal.
(8) In the case of zinc this difference is very slight.
Cambridge, Mass., October, 1898.
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. No. 5. — Decembp:r, 1898.
CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.
TRINITR0PHEN7LMAL0NI0 ESTER: SECOND PAPER.
By C. Loring Jackson and J. I, Phixney.
CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF
HARVARD COLLEGE.
TRINITROPHENYLMALONIC ESTER.
SECOND PAPER.
By C. Loring Jackson and J. I. Phinney.
Presented October 12, 1898. Received October 20, 1898.
The work described in this paper was undertaken with the intention of
preparing some derivatives of the trinitrophenylmalonic ester (picrylraa-
louic ester), C6Ho(N02);5CH(COOaH5)2, discovered by C. A. Soch and
one of us.* It has led to the discovery of a second and more stable
form of the trinitrophenylmalonic ester, which melts at 64° instead of 08°,
the melting point of the form at first obtained ; and we have also pre-
pared the nitrite of this ester,
CeH2(NO,)sCN02(COOC2H5)2,
melting at 109°, the corresponding trinitrophenyltartronic ester,
C6H2(N02)3COH(COOC2H5)2,
melting at 117°, its acetyl derivative,
CcH2(NOo)3COCOCH3(COOC2H5)2,
which melts at 125°, and the trinitropheuylacetic acid,
C6H2(N02)3CHoCOOH,
melting at 161°.
The Two Modifications of Trinitrophenylmalonic Ester.
The trinitrophenylmalonic ester, as prepared by Soch and one of U8,t
crystallized from alcohol in white long rather slender rectangular plates,
or when better developed in thick prisms with blunt ends, often as much
as two centimeters long. It melted at 58°. J When we first took up the
* These Proceedings, XXX. 40L
t Ibid.
X In the previous paper, this melting point is given as 59°, but it must be changed
to 58°, as it was found after that paper was published that the zero point of the
thermometer used had changed.
104 PEOCEEDINGS OF THE AMERICAN ACADEMY.
work again, we also obtained the trinitrophenylmalonic ester exclusively
in this form, and confirmed repeatedly the observations given above on
its crystalline form and melting point. Somewhat later we converted a
sample of the triuitrophenylmalonic ester into its very characteristic am-
monium salt, and upon acidifying this and recrystallizing the ester thus
set free we obtained instead of the rectangular plates four- (or six-) sided
plates in which two of the opposite angles were unlike, — one being
obtuse and the other acute, — so that the crystals were shaped somewhat
like a kite ; and not only did the form of these crystals differ in such a
marked way from the rectangular plates previously obtained, but they
also melted at G4° instead of at 58°. During this experiment a solution
of the ester from one of our preparations of it was evaporating spon-
taneously on the desk ; the next morning these kite-shaped crystals were
deposited from it, although it had previously yielded only the rectangular
plates melting at 58°, and since then we have obtained from all our
preparations only the form melting at 64°, whereas before exactly sim-
ilar pre[)arations had given us exclusively the form melting at 58°, as
Las been already stated. Varying the conditions of the preparations,
such as carrying them on in cooled or warmed solutions, or with longer
or shorter standing, did not modify the result, so that we have not suc-
ceeded in adding to the stock of the modification melting at 58°, which
we had on hand, when we encountered the other form.
The following experiment seems to us to suggest a possible explanation
of these results : a solution of the form melting at 58° was evaporated
until it began to deposit the rectangular crystals, and then inoculated
with a speck of the form melting at 64°, when at once the kite-shaped
crystals of this form began to appear, and no more rectangular crystals
were produced. This experiment was tried several times early in our
work ; it could not be repeated later, as then recrystallization alone of
the form melting at 58° was enough to convert it partially or even com-
pletely into the form melting at 64°. As then a small amount of the
form melting at 64° is enough to convert a large quantity of the other
form into this, it seems probable that the small amount of the more
stable form floating in the air of the laboratory as dust was enough to
bring about this change, and to give us this more stable modification as
the sole product of our preparations, and even of recrystallizations of the
less stable form. The view that the inoculation of the solutions pro-
ceeded from the dust in the air is supported by the following observa-
tions. A preparation made with new apparatus and fresh material,
but in the laboratory where these experiments had been tried, yielded
JACKSON AND PHINNEY. PICRYLMALONIC ESTER. 105
only the stable form melting at 64°. Crystallizations of the form
melting at 5S° gave, after the work had been in progress some time,
the form melting at 64°. We then crystallized a specimen melting
at 58° in a fresh room, which had not been used for these experi-
ments, and obtained in this way the rectangular plates melting at 58° ;
but upon recrystallizing the specimen, the form melting at 64° appeared,
and after this the first crystallization of a fresh specimen melting at 58°
in this room gave crystals of the form melting at 64°. A probable ex-
planation of these latter observations is that there was a little dust of
the more stable modification in this room coming from the clothes of one
of us, who used it as a lecture room, and that this was sufficient gradually
to inoculate the solutions. As the laboratory building, apparently con-
taminated with the dust of the more stable form, seemed to offer little
chance of preparing more of tlie form melting at 58°, we postponed pub--
lishing this paper for some years, in the hope that one of us (who left
Cambridge at the end of the year) might prepare more of the form melt-
ing at 58° by working in entirely new surroundings. Unfortunately, the
pressure of other duties has prevented the carrying out of this work,
and we have decided that it is wiser to publish now the results already
obtained rather than to postpone the appearance of the paper to a still
later date.
The conversion of the form melting at 58° into that melting at 64°
has been brought about by us in the following ways. By its conversion
into the ammonium salt and setting free the ester by acidification ; * by
inoculating a saturated alcoholic solution with a crystal of the form melt-
ing at 64° ; by melting the less stable form it was partially changed into
that melting at 64°, and, if the melting was repeated, the change became
com[)lete. It is possible, however, that this change might have been due
to inoculation from dust. On the other hand, we have not succeeded in
converting the form melting at 64° into that melting at 58° in s[)ite of
many experiments ; even melting the more stable form and stirrinaf it
with a rod tipped with the modification melting at 58° did not have the
desired effect.
The substance melting at 64° was purified by crystallization from alco-
hol, and after being dried in vacuo was analyzed with the followino-
results : —
I. 0.2050 gram of the substance gave on combustion 0,3149 gram of
carbonic dioxide and 0.0736 gram of water.
* In my earlier work with Socli, the sodium salt yielded on acidification the
ester melting at 58°. — C. L. J.
106 PROCEEDINGS OP THE AMERICAN ACADEMY.
II. 0.2007 gram of substance gave 0.3078 gram of carbonic dioxide
and 0.0789 gram of water.
III. 0.2040 gram of substance gave 20.5 c.c. of nitrogen at a tempera-
ture of 21° and a pressure of 759.6 mm.
Calculated for Tound.
CaH2(N02)3CH(COOC2H5),. I. II. HI.
Carbon 42.05 ' 41.89 41.82
Hydrogen 3.50 3.98 4.36
Kitrogen 11.32 11.41
The combustion of this substance must be carried on as slowly as pos-
sible, as it has a strong tendency to explode.
The molecular weight of the substance was also determined by the
method of Raoult with the following results : —
I. Weight of benzol used 19.2120 grams, substance 0.3291 gram.
Reduction of freezing point 0°.230.
II. Weight of benzol used 15.7244 grams, substance 0.3428 gram.
Reduction of freezing point 0°.306.
III. Weight of benzol used and weiglit of substance the same as in II.
Reduction of freezing point 0°.297.
Calculated for Found.
C6H2(N02)3CII(C00Cjn3)j. I. II. III.
Molecular Weight 371 365.1 349.0 359.6
From these results there can be no doubt that the preparation melting
at 64° was trinitropheuylmalonic ester.
It is worth while to point out here that the form melting at 58° could
not have owed its peculiarities in melting point and crystalline form to
the presence of a small quantity of some impurity, since it was repeatedly
brought to a constant melting point both by Dr. Soch and by us ; and, if
this explanation were correct, this form should have been the first pro-
duct of the preparations in our later work, whereas, after we had once
obtained the form melting at 64°, we could not prepare the other. Dr.
Soch's analyses of the form melting at 58° also show that it was pure.
Calculated. Found.
Nitrogen 11.32 11.75 11.38
We are therefore led to the conclusion that the trinitropheuylmalonic
ester exists in two isomeric forms.
The isomerism of these two forms is probably due to the same cause
as that of the red and yellow anilidotrinitrophenyltartronic esters (melt-
JACKSON AND PHINNEY. — PICRYLMALONIC ESTER. 107
ing at 143° and 122°) discovered by W. B. Bentley * and one of us. It
seems more doubtful whether it is related to the isomerism of the formyl-
phenylacetic esters discovered by W. Wislicenus,t of the dibenzoylace-
tones t and the benzoates of oxymethylenacetone § of Claisen, and the
benzalanilinacetacetic esters of R. Schiff,1[ since both of our isomeres can
be dissolved in alkaline solutions without alteration, whereas one of the
most marked differences between the isomeres in these other cases con-
sists in the fact that one dissolves in alkalies and the other does not.
Our present knowledge of these substances is not sufficient to allow us
to make a final statement in regard to the cause of this isomerism, but
the following discussion may be given for what it is worth. The fact
that each form has been converted into a salt, and set free from this un-
altered upon acidification, would indicate that it is a case of chemical
rather than physical isomerism ; and, if this is true, as the symmetrical
nature of the substance would forbid stereometric isomeres, the only
probable explanation which we have been able to find is that the isom-
erism depends on differences of structure in the malonic group, as indi-
cated by the following formulas : —
I. CeH2(N02)3CH(COOCoH5)2.
n. c.h,(no,).c1^Xoc:h..
In other words, that the isomerism is analogous to that of the bodies
studied by W. Wislicenus, Claisen, and R. SchifF. There are, however,
objections to this explanation, and, as we have said already, the whole
subject needs further investigation before any theory can be found
satisfactory.
Properties of the Trinitrophenylmahmc Ester melting at 64°. — Crys-
tallized from alchohol, it forms white plates, usually bounded by four or
six sides. In the four-sided form there are two unlike opposite angles,
one obtuse the other acute, which give a very characteristic appearance
to the crystals. The six-sided form is produced by two planes trun-
cating the similar angles in the four-sided form. Both these forms
resemble a kite in shape. Occasionally much more complex crystals
w^ere observed, having a parallel-sided projection arising from the middle
of the obtuse end, and also terminated by an obtuse angle ; this form had
a general resemblance to the "spade" in playing cards. The plates
* These Proceedings, XXVI. 82. § Ber. d. chem. Ges., XXV. 1785.
t Bar. d. chem. Ges., 1895, p. 767. T Bar. d. cliem. Ges., XXXI. 601.
% Ann. Chem., CCLXXVII. 188.
108 PROCEEDINGS OF THE AMERICAN ACADEMY.
were often thick, especially when having the complex form last described.
The substance turns yellow on long exposure to the air, and imparts to
alcohol a marked crimson color, which has also been observed on the
ground glass stopper of the bottle in which it was kept. In this case the
color may have been due, however, to some alkali from the glass, but this
was not the case in the alcoholic solution, as the color was only partially
discharged by acidification. The modification melting at 58° forms
similar crimson solutions. It melts at 64:° to a red TKUiid. Its solubility
in the organic solvents is essentially the same as that of the form melting
at 58° ; that is, it is very soluble in chloroform, ether, benzol, or glacial
acetic acid ; somewhat less so in carbonic disulphide ; soluble in cold
alcohol, freely in hot, rather more soluble in methyl than in ethyl alcohol ;
insoluble in ligroine, or cold water, slightly soluble in hot water.
The action of acids upon this more stable trinitrophenylmalonic ester
is described in some of the following paragraphs of this paper. Alkalies
give with it dark red salts, which crystallize well. The ammonium salt
is especially characteristic ; it was made by adding an excess of aqueous
aramonic hydrate to a wanm nearly saturated alcoholic solution of the
trinitrophenylmalonic ester. The solution took on a deep purplish red
color, and in a few seconds the whole solidified to a thick semisolid mass
havino- the purplish red color and consistency of clotted blood. A micro-
scopic examination showed that this mass- was made up of long very
slender hair-like crystals, which after drying had a rich golden brown
color and a silky lustre. The salt decomposed below 100°, at first turn-
ing black, but afterward melting to a clear yellowish liquid, which ex-
ploded at higher temperatures. It is rather sparintrly soluble in water;
soluble in alcohol, chloroform, or acetone ; insoluble in ligroine. The
solutions have a dark red color, but if the solution in water or alcohol is
boiled for some time it turns yellowish brown ; unfortunately no product
could be obtained from such solutions in a state fit for analysis.
An aqueous solution of the ammonium salt gave the following charac-
teristic precipitates.
With salts of barium an abundant crystalline amethystine precipitate.
With salts of strontium a dark red precipitate.
With salts of calcium a brick-red precipitate.
With salts of zinc an abundant crystalline scarlet precipitate, turning
to reddish brown on standing.
With salts of cadmium a granular scarlet precipitate.
With salts of copper a heavy flaky precipitate, varying from reddish
yellow to brownish red.
JACKSON AND PHINNEY. — PICRYLMALONIC ESTER. 109
With salts of lead heavy dark red flocks.
Salts of the other metals gave precipitates as a general rule, but they
were not characteristic. The behavior with argentic nitrate, however,
should be especially mentioned, as this gave no precipitate in dilute solu-
tions, and only a slight cloudy dark red one when the solutions were strong.
The barium, zinc, and copper salts were analyzed.
Barium Salt. — This salt was made by adding a solution of baric
chloride to the aqueous solution of the ammonium salt. The heavy
purple flocks thus obtained were washed with water, dried at 100°, and
analysed with the following result : —
0.4864 gram of the salt gave 0.1236 gram of baric sulphate.
Calculated for
[C8Hj(N0j)3C(C00CoIl5)2JjBa. Found.
Barium 15.62 14.95
The salt is so hygroscopic that two tenths of a gram will gain as' much
as four milligrams during the time of weighing, if it is in an open watch
glass. In evaporating it with sulphuric acid for analysis, care must be
taken not to apply the heat too suddenly at first, or an explosion may
result. The best plan is to heat the mixture for some time to 100", as
then the decomposition goes on quietly.
Properties. — The barium salt appears as a purple obscurely crystalline
mass, essentially insoluble in water. It explodes if heated to 120°.
Zinc Salt. — This salt was made by mixing solutions of zincic sulphate
and the ammonium salt of the ester. It was purified by washing, dried
at 100°, and analyzed with the following result: —
0.4869 gram of the salt gave 0.0478 gram of zincic oxide.
Calculated for
[Coll, NOolsCICOOOall,-, o „Zrx. Found.
Zinc 8.07 7.88
Properties. — The zinc salt forms a reddish brown crystalline mass,
essentially insoluble in water, but very hygro>copic. Like the barium salt
it explodes easily, if heated with strong sulphuric acid, and therefore the
same precautions must be used in analyzing it which were recommended
in the case of the barium salt.
Copper Salt. — This salt was made by mixing aqueous solutions of
cupric sulphate and the ammonium salt of the ester. It was purified by
washing with water, and after drying at 100° gave the following result
on analysis : —
0.3012 gram of the salt gave 0.0291 gram of cupric oxide.
110 PROCEEDINGS OF THE AMERICAN ACADEMY.
Calculated for
[C6H2(N02)3C(COOC2Hj)2]2Cu. Pound.
Copper 7.92 7.72
Properties. — The copper salt is a brownish red powder, essentially
insoluble in water, it resembles the barium and zinc salts in being very
hygroscopic, and exploding easily, when heated with strong sulphuric
acid.
Action of Nitric Acid on Trinitrophenylmalonic Ester.
The action of nitric acid upon this substance is similar to its action with
the broratrinitrophenylmalonic ester,* that is, the nitrite or the substituted
tartronic ester is obtained according to the length of the treatment.
Nitrite of Trinitrophenylmalonic Ester,
C6H2(N02)3CONO(COOC2H5)2.
Two grams of the trinitrophenylmalonic ester were heated on the
steam bath with about 15 c.c. of common strong nitric acid. After three
minutes the ester had melted to a clear red oily globide, and the entire
liquid had taken on a reddish color. If the process was continued for
two minutes more, the globule went into solution. The heating should
not be continued beyond this point ; in fact, it can be stopped to advan-
tage even as soon as the globule of the melted substance is formed. If
all the organic matter had gone into solution, reddish crystals separated
in large quantity, as the liquid cooled, which were purified by washing
with water and several crystallizations from alcohol, until they showed
the constant melting point 109°. The globule, if the process was stopped
before it disappeared, was allowed to solidify, and then purified in the
same way. The substance was dried in vacuo, and analyzed with the
following result : —
0.2190 gram of the substance gave 25.8 c.c. of nitrogen, at a tempera-
ture of 20° and a pressure of 760.8 mm.
Calculated for
C6H2(N02)3CN0o(C00C2H5)2. Found.
Nitrogen ' 13.46 13.46
Properties of the Nitrite of Trinitrophenylmalonic Ester.
It crystallizes in long flat white prisms, apparently belonging to the
monoclinic system, usually terminated by two planes at an obtuse angle to
each other, but sometimes only by a single plane. It melts at 109° with
* These Proceedings, XXVI. 72.
JACKSON AND PHINNEY. — PICRYLMALONIC ESTER. Ill
decomposition, as shown by the appearance of a red color. It is easily
soluble in ethyl, or methyl alcohol, or in chloroform, ether, benzol, glacial
acetic acid, or acetone ; nearly insoluble in carbonic disulphide ; essen-
tially insoluble in water, when cold, very slightly soluble in hot water,
^vinw a pink solution. The three strong acids have no apparent action
with it in the cold; strong sulphuric acid, when hot, decomposes it, as
shown by the dark color produced and the evolution of gas; strong nitric
acid, when hot, converts it into the corresponding tartronic ester. Alka-
lies have no action upon it at first, but gradually decompose it in the cold
with the formation of a red solution ; this action is much hastened by
heat, but we have made no attempt to isolate the uninviting product.
In all these respects the nitrite of trinitrophenylmalonic ester resembles
the nitrite of bromtrinitrophenylmalonic ester,* but we have not succeeded
in converting the former into the corresponding tartronic ester by decom-
position by heat, as was done with the latter. No crystalline substance
could be obtained from the viscous red product of the fusion.
Trmltrophenrjltartronic Ester, C6Ho(N02)3COH(COOC2H5)2.
This substance was obtained by long continued action of hot nitric acid
on triniti'ophenylmalonic ester, or by the action of the same reagent on
the nitrite just described. Two grams of trinitrophenylmalonic ester were
warmed in a porcelain dish on the water bath with about 'ib c.c. of com-
mon strong niti'ic acid, more acid being added from time to time to take
the pliice of what evaporated. To secure the complete formation of the
tartronic ester, the heating should be continued for five hours. If the
acid solution was not too concentrated, it deposited on cooling long nee-
dles of a pink color arranged in rosettes ; stronger solutions gave a more
or less red solid cake of the same compound. It was purified by washing
with water, and two recrystallizations from alcohol by cooling, when it
showed the constant melting point 117°, and after drying in vacuo was
analyzed with the following results : —
I. 0.2129 gram of the substance gave 20.8 c.c. of nitrogen at a tem-
perature of 21° and a pressure of 7G9.7 mm.
II. 0.2051 gram gave 20.3 c.c. of nitrogen at a temperature of 25° and
a pressure of 750.3 mm.
Calculated for Found.
CoH2(N02)3COH(COOC2H5)j. I. II.
Nitrogen 10.85 11.24 10.87
* These Proceedings, XXVI. 74.
112 PROCEEDINGS OF THE AMERICAN ACADEMY.
The analysis of this substance gave much trouble, because of the ease
with which it exploded, but the ditiiculties were overcome by mixing it
with a large amount of cupric oxide, and heating very carefullv.
P ruperties of TrinitropJienijltartronic Ester. — It forms white fluffy-
needles in clusters like sheaves or even circles. Under the microscope
the crystals are seen to be slender prisms terminated by a single plane at
an oblique angle. Its melting point is 117°. It is easily soluble in
ethyl, or methyl alcohol, in fact the hot concentrated solution in ethvl-
alcohol is so strong that it solidifies nearly completely on cooling. It is
also easily soluble in ether, benzol, chloroform, or acetone; soluble in
glacial acetic acid; somewhat soluble in carbonic disulphide ; soluble with
difficulty in ligroine ; insoluble in cold water, slightly soluble with decom-
position in hot. forming a reddish solution. The three strong acids have
no apparent effect on it in the cold, but hot strong sulphuric aciJ decom-
poses it, giving a dark color and an evolution of gas. Alkalies give with
it at once dark blood-red solutions probably containing its salts, but these
decompose rapidly, turning dirty brown, and we have not succeeded in
isolating the salts themselves, or any definite compounds from their
decomposition products.
Trin itrophenyJacett/ltartron I'c Ester,
C6H.,(NOo)3C(OCOCH3)(COOCoH5),.
Acetylchloride dissolves trinitrophenyltartronic ester easily, but does
rot react with it either at ordinary' temperatures, or when heated on the
steam bath in open vessels. We accordingly proceeded as follows :
Three grams of trinitrophenyltartronic ester were heated with 15 to 20 c.c.
of acetylchloride in a sealed tube to 110° for five or six hours. Care was
taken that the temperature did not go too high, as at 140° a blackish
decomposition product was also formed, which made the purification of
the product more difficult. On evaporating off the excess of acetylchlo-
ride white crystals separated, which were purified by crystallization from
alcohol until they showed a constant melting point of 125°, when they
were dried in vacuo, and analyzed with the following result: —
0.2201 gram of the substance gave 19.4 c.c. of nitrogen at a tempera-
ture of 24° and a pressure of 761.6 mm.
Calculated for
C6HJ(NOJ)3COC2H3O^CO0CJH5),. Fonnd.
Nitrogen 9.79 9.88
Properties of Trinitrophenylacefyltnrtronic Ester. — It forms white flat
rather broad prisms, or perhaps they should be called tables, terminated
JACKSON AND PHINNEY. — PICRYLMALONIC ESTER. 113
by two planes, and apparently belonging to the monoclinic system. It
melts at 125°, and is easily soluble in ethyl or methyl alcohol, benzol,
chloroform, or acetone ; somewhat soluble in ether, less so in carbonic
disulphide ; soluble in glacial acetic acid ; insoluble in ligroine ; insoluble
in cold water, slightly soluble in hot, forming a pink solution. Strong
hydrochloric acid does not dissolve it. Nitric acid when hot dissolves it
apparently without decomposition. Sulphuric acid dissolves it, and de-
composes it, when heated. Alkalies have no effect on it in ths cold, but,
if heated with it, form the red salts of its decomposition product.
The benzoyl derivative of trinitrophenyltartronic ester was also made,
but the close of the academic year prevented us from getting a satisfac-
tory analysis of it. It was made by heating the tartronic ester with
benzoylchloride in a sealed tube to 110° for four or five hours. On
evaporating off the excess of benzoylchloride, it was obtained in white
crystals, which were purified by recrystallization from alcohol until they
showed the constant melting point 152°.
Saponification of Trinitrophenylmalonic Ester.
In all these experiments the modification melting at 64"^ was used.
The method of saponification adopted was that which had given such
excellent results with the other substituted malonic esters studied in this
laboratory, that is, the action of sulphuric acid diluted to a specific grav-
ity of 1.44. Five grams of trinitrophenylmalonic ester were mixed with
100 c.c. of this acid, and the mixture boiled in a flask with a return con-
denser. The ester melted almost immediately to a straw-colored oily
liquid, which gradually decomposed with evolution of gas, and finally
was completely dissolved. This usually took place in from one and a
half to two and a half hours, and showed that the reaction had come to
an end. During the earlier part of the boiling the liquid appearing in
the condenser had a pinkish color, but this disappeared later in the pro-
cess. As soon as all the ester had dissolved, the liquid was allowed to
cool, when it deposited a bulky precipitate made up of yellowish white
needles. This substance consisted of the trinitrophenylacetic acid, and
was purified in the way described in the following paragraph.
Trinitrophenylacetic Acid, C6H2(N02)3CH2COOH. — The purification
of this substance cannot be effected by crystallization from alcohol or
water, as either of these solvents converts it into the corresponding sub-
stituted toluol C6ll2(N02)3CIl3. We tried at first as a solvent water
containing a small amount of sulphuric acid, since this had given excellent
VOL. xxxiv. — 8
114 PROCEEDINGS OF THE AMERICAN ACADEMY.
results with bromdinitrophenylacetic acid ; * but, whereas with that sub-
stance a few drops of sulphuric acid were sufficient to prevent the for-
mation of the substituted toluol, we found that even one per cent of acid
did not produce this effect with the triuitrophenylacetic acid, and that
the amount must be raised to five per cent to insure perfect safety. For-
tunately it was not necessary to use this rather strongly acidified water,
as benzol proved to be an excellent solvent for the substance. Accord-
ingly the crystals, which formed the product of the saponification (see
the preceding section), were filtered from the sulphuric acid through
glass wool, washed with water, dried, and recrystallized from benzol
until they showed the constant melting point 161°, when they were dried
at 100°, and analyzed with the following i-esults : —
I. 0.2020 gram of the substance gave 28.2 c.c. of nitrogen at a tem-
perature of 22° and a pressure of 750.8 mm.
II. 0.2093 gram of the substance gave 28.4 c.c. of nitrogen at a tem-
perature of 20° and a pressure of 758.3 mm.
Calculated for Found.
CoU,(NO„)3CU.,COOH. I. II.
Nitrogen 15.50 15.59 15.46
Properties of Trinitrophenylacetic Acid. — It crystallizes from benzol
in short needles arranged in sheaf-shaped bunches, and melts at 161°.
As at first prepared it is white, but it turns pink on standing, even when
in a corked tube, and this change is due to decomposition, as it is accom-
panied by an alteration in the melting point. It dissolves in ethyl or
methyl-alcohol with a pink color, but this solution produces a decompo-
sition, the trinitrophenylacetic acid losing carbonic dioxide and forming
trinitrotoluol, which is left in a nearly pure state on the evaporation of the
solvent. It was recognized by its melting point, 81°. Willbrand f gives
82°, Mills I 78°.84 and B0°.52. (It is possible that further cr\stalliza-
tion miglit have raised the melting point of our specimen to 82°.) Ihe
trinitrophenylacetic acid, therefore, is less stable than the bromdinitro-
phenylacetic acid, as two evaporations with alcohol were necessary to
convert this entirely into the corresponding bromdinitrotoluol. One crys-
tallization of the trinitrophenylacetic acid from boiling water was suf-
ficient also to form the trinitrotoluol, and, as has been already said, the
addition of five per cent of sulphuric acid was necessary to prevent this
action. The aqueous solution had a pink color like that of the alcoholic
solution, but we have not succeeded in discovering the cause of this color-
* These Proceedings. XXIV. 241. t Ann. Chem., CXXVIII. 178.
X Phil. Mag., [5], XIV. 27.
JACKSON AND PHINNEY. PICRYLMALONIC ESTER. 115
ation, which disappears from the solutions, as the substituted toluol is
purified by recrystallizatiou. The solvents which follow dissolve the tri-
nitrophenylacetic acid without decomposition ; easily soluble in ether,
glacial acetic acid, acetone, or ligroine ; soluble in chloroform, less so in
benzol or carbonic disulphide. The best solvent for it is benzol. The
three strong acids seem to dissolve it without decomposition; it is more
soluble in nitric acid or sulphuric acid than in hydrochloric acid. Alka-
lies dissolve it easily, forming deep blood-red solutions of its salts, but we
have not yet succeeded in bringing any of these salts into a state fit for
analysis.
Attempts to Make Ditrinitrophenylmalonic Ester.
Dittrich,* in his work on the action of picrylchloride on sodium aceta-
cetic ester, obtained without ditficulty a ditrinitrophenylacetacetic ester ;
in fact this substance occurred as a secondary product in his preparations
of trinitrophenylacetacetic ester, when he used the reagents in molecular
proportions. It seems strange, therefore, that we have never observed
the formation of a ditrinitrophenylmalonic ester as a secondary product
in any of our preparations of the mono substituted ester. In the hope
of preparing this substance (ditrinitrophenylmalonic ester) we converted
5 grams of the trinitrophenylmalonic ester into its sodium salt by treat-
ment with the sodic ethylate from 0.31 gram of so<lium, and then treated
it with 3.4 grams of picrylchloride. The mixture was allowed to stand
at ordinary temperatures over night, and, as no apparent chanjie had
taken place, it was divided into two portions, one of which was heated
on the steam bath until it turned brown, and the other allowed to stand
two weeks, when it had also turned brown, and deposited crystals of
sodic chloride. The brown solutions, whether obtained by heating or by
standing, yielded sodic picrate and a brown viscous mass, from which
nothing fit for analysis could be isolated. A repetition of the experiment
under other conditions gave the same result. It seems, therefore, that
picrylchloride brings about a deep seated decomposition, when it acts on
the sodium salt of the trinitrophenylmalonic ester, as the ditrinitrophe-
nylmalonic ester would be without doubt a well crystallized compound.
The products of this reaction recall Dittrich's description of the substances
obtained by him from the action of picrylchloride on sodium malonic
ester (pikrinsaures Natrium neben schmierigen Zersetzungs-producten des
Malonsaureester), and it may be that the presence of an excess of picryl-
* Ber. d. chem. Ges., XXIII. 2720.
116 PROCEEDINGS OF THE AMERICAN ACADEMY.
chloride was the cause of his faihire to obtain trinitrophenylmalonic ester,
although this does not appear from the statemeut in his papers.
Some experiments on the action of aniline on trinitrophenylbromraa-
louic ester yielded only viscous unmanageable products ; and boiling it
with water gave no more promising results.
We have also tried several times to detect the presence of trinitro-
benzol in the secondary products of the action of sodium malonic ester
on picrylchloride, but without success. It seems, therefore, that the re-
placement of the chlorine of the picrylchloride by hydrogen does not take
place under these conditions to any great extent, if it occurs at all.
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. No. 6. — December, 1898.
CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.
ON THE ACTION OF SODIC ETHYL ATE ON
TRIBE OMDINITR BENZOL.
By C. Loring Jackson and Waldemar Kocii.
CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF
HARVARD COLLEGE.
ON THE ACTION OF SODIC ETHYLATE ON
TRIBROMDINITROBENZOL.
By C. Loring Jackson and Waldemar Koch.
Presented October 12, 1898. Received October 20, 1898.
The principal object of the work described in this paper was to
determine the constitution of the dinitroresorcine diethylether melt-
ing at 133°, formed by the action of a hot solution of sodic ethylate
on tribromdinitrobenzol (Br3l,3,5(N02)22,4), or on the bromdiuitrore-
sorcine diethylether melting at 184°. This determination was of interest
to us, as it formed part of an investigation of the replacement of bromine
by hydrogen under the influence of sodium malonic ester, sodic ethylate,
and similar reagents, which has now been in progress for several years in
this Laboratory. In all the cases studied here, with a single exception,
the atom of bromine (or iodine) has stood between two other radicals,
each in the ortho position to it, and it was necessary to determine whether
this arrangement also existed in this case, since a consideration of the
possibilities shows that this dinitroresorcine diethylether might have either
the symmetrical structure, if the bromine between the two nitro groups
was replaced by hydrogen, or the adjacent structure, if this replacement
occurred with one of the other atoms of bromine.
The constitution of the diethylether was determined by saponifying it
to the corresponding dinitroresorcine, which proved to be the symmetrical
compound (OH)ol,5(N0.2)22,4, melting at 212°.5, according to Typke.*
The proof that this body has the structure assigned to it has been given
by Nietzki and Schmidt,t who converted it into diimidoresorcine, which
in turn yielded dioxyquinone by treatment with sodic hydrate, and this
dioxyquinone gave with nitric acid nitranilic acid, which is paradinitro-
dioxyquiuone, a result that is in harmony only with the symmetrical
structure for the dinitroresorcine. It follows therefore that the bromine
replaced by hydrogen in tribromdinitrobenzol is the one between the
* Ber. d. chem. Ges., XVL 552. t Ibid., XXL 2374.
120 PROCEEDINGS OF THE AMERICAN ACADEMY.
two nitro groups, as we had expected from the results of previous experi-
ments, and that the bromdinitroresorcine diethylether melting at 184°
has the constitution (OCoH5)2l,5,Br3.(NOo)o2,4. It is an interesting
fact in this connection that an isomeric bromdinitroresorcine diethylether
(having the constitution (OCoIT6)2l,3,Bro,(N02)22,4), in which the bro-
mine is not in the ortho position to two nitro groups, gave no replace-
ment of the atom of bromine by hydrogen, but, when treated with sodic
ethylate, gave dinitropliloroglucine triethylether by the replacement of
the bromine by an ethoxy group. Beilstein, in the third edition of his
Handbuch, provisionally places the diethylether C6H2(OC2lT5)2(N02)2
melting at 133°, and the corresponding dimethylether C6H2(OCH3)2(N02)2
melting at 167°, prepared from tribromdinitrobenzol by W. H. Warren
and one of us,* under the adjacent dinitroresorcine. The determination
of the constitution of the diethylether just given necessitates their trans-
fer to the symmetrical dinitroresorcine, as there can be no question that
the methylether has the same structure as the ethyl compound. It also
shows tliat the dimethylether CgH2(OCII3)2(N02)2 melting at G7°, pre-
pared by Honigt by the direct action of nitric acid on a solution of re-
sorcine dimethylether in glacial acetic acid, cannot have the symmetrical
structure which Beilstein assigued to it, as this now belongs to the body
melting at 167°.
In addition to this determination of the constitution of dinitroresorcine
diethylether, we have made a mi»re careful study of the products of the
reaction between sodic ethylate and tribromdinitrobenzol melting at 192°
both at ordinary temperatures and when aided by heat. These reactions
had been already studied by W. H. Warren and one of us, J with the result
that in the cold the bromdinitroresorcine diethylether melting at 184°
was formed, whereas, when hot, the dinitroresorcine diethylether melting
at 133° was the product isolated. As, however, subsequent work§ upon
the tribronitriniti'obenzol showed that in that case the action was far from
simple, it hardly seemed probable that the dinitro compound had given
only one organic product in each of thesfe cases, and this inference was
confirmed by our first new experiment m this field, as we found sodic
nitrite as well as sodic bromide among the products of the reaction, so
that it was evident that some other compounds were present besides those
recognized by Warren and one of us. The result of this later work was
* These Proceedings, XXV. 170, 178.
t Ber. d. chem. Ges., XI. 1041.
J These Proceedings, XXV. 166.
§ Jackson and Warren, Ibid., XXVII. 283.
JACKSON AND KOCH. TRIBROMDINITROBENZOL. 121
the isolation of the following products of the action of sodic ethylate
on tribromdinitrobenzol, melting jioint 192°, when absolute alcohol and
benzol are used as the solvents, and the mixture is allowed to stand at
ordinary temperatures : —
(1) Broradinitroresorcine diethylether CoHBr(OCoH5)o(N02)2, melting
at 184°, (OCoH5).3l,5,Br3,(N02)22,4.
(2) Broradinitroresorcine diethylether CcHBr(0C2H5)2(N02)>, melting
at 92°, (OC2H5)2l,3,Bro,(N02)22,4-
(3) Tribromnitrophenol C6HBr3(N02)OH, melting at 90°, Bra 1,3,5,
^0.2,0 114.
(4) Dinitroresorcine monoethylether C6H2(OC2H5)OII(NOo)2, melting
at 77°, (OC2H5)l,OH5,N022,4.
(5) Dinitrophloroglucine triethylether Con(OC2ll5)3(N02)2, melting at
105°, (OCori5)3l,3,5,(N02)22,4.
These products indicate that there are three primary reactions, when
sodic ethylate acts on tribromdinitrobenzol. These are, Reactions I. and
II., consisting in the replacement of two atoms of bromine by two ethoxy
groups, producing the two isomeric broradinitroresorcine diethylethers
(1) and (2), and Reaction III., in which one uitro group is replaced by
one hydroxyl group (or by an ethoxy group followed by saponification)
giving the tribromnitrophenol (3). There are also two secondary reac-
tions owing to the further action of sodic ethylate on the products of
Reactions I. and II. ; first, the replaceraent of the bromine in the brora-
dinitroresorcine diethylether (1) by an atora of hydrogen (followed by
partial saponification) giving the dinitroresorcine monoethylether (4) ;
and secondly, the replacement of the bromine in the bromdiuitroresorcine
diethylether (2) by an ethoxy group giving the dinitrophloroglucine tri-
ethylether (5). The saponification mentioned in this paragraph might
be brought about by the sodic hydrate formed from the sodic ethylate by
the water added in the course of the purification, but it seems to us more
probable that the phenols were formed by some sodic hydrate acting
directly on the tribromdinitrobenzol, since the ethers of the phenols
seemed decidedly stable, when treated with an alkali.
The following estimations of the approximate yields of the products
give a rough idea of the extent to which each of the three primary
reactions ran : —
Reaction I., forming substances (1) and (4) . . . 28 per cent.
Reaction II., forming substances (2)and (5) . . . 38 per cent.
Reaction III., forming substance (3) 1 per cent.
Total G7 per cent.
122 PROCEEDINGS OF THE AMERICAN ACADEMY.
As most of these substances were purified by crystallization, too much
weight must not be given to these yields, but we can safely infer from
them that Reaction II. ran to the largest extent, Reaction I. came next,
and Reaction III. took jilace only to a very limited extent.
These results are capable of a satisfactory theoretical explanation, but
before giving this it will be well to consider a parallel case studied by
W. H. Warren and one of us,* in which the conditions are simpler and
rather, better marked. This is the action of sodic ethylate on sym-
metrical tribromtrinitrobenzol CgBr3(N02)3, Br3l,3,5,(N02)32,4,6, which
consisted of two simultaneous reactions, first, the replacement of two nitro
groups by two ethoxy groups, giving C6Br3NOo(OC2H5)2, and second, the
replacement of the three bromine atoms by three ethoxy groups, giving
Cg(OC2H5)3(N02)3. These two reactions ran to nearly an equal extent
(about 45 per cent of each product), if the solvent was alcohol only.f
In attempting to explain these experimental results we must consider
the agencies which tend to loosen the nitro groups on the one hand and
the atoms of bromine on the other from the benzol ring. We have
designated these agencies by letters, and arranged them in the order of
intensity, beginning with the strongest.
A. The loosening effect of the three nitro groups on the three atoms
of bromine. This is a strong agency, as shown by the fixct that
picrylchloride (in which the chlorine is similarly disposed to the tliree
nitro groups) is decomposed easily and completely, even at ordinary
temperatures. J
H. The loosening effect exerted by the three nitro groups on each
other. Lobry de Bruyn § has shown that symmetrical trinitrobenzol is
converted into dinitroanisol by sodic methylate, even at ordinary tem-
peratures ; this loosening effect is therefore a powerful one, but not quite
so strong as A, since the action with picrylchloride is complete in a few
minutes, whereas the reaction with trinitrobenzol needs several days, if
carried on in the cold.
C The loosening effect of the three atoms of bromine on the threo
nitro groups. This is a much weaker action than B, since with the
tetrabrombenzol || (Br4l, 2,4,0) the atom of bromine (which stands in the
* These Proceedings, XXVII. 283.
t If the solvent was benzol and alcohol, less of the nitro groups were replaced ;
but this case need not be considered in this discussion.
} These Proceedings, XXXIII. 176.
§ Rec. Trav. Chim. Pays Bas, IX. 208.
II Jackson and Calvert, These Proceedings, XXXI. 132.
JACKSON AND KOCH. — TRIBROMDINITROBENZOL. 123
same position toward the three atoms of bromine as the three nitro groups
do in our body) was removed only after long boiling of the benzol and
alcohol solution, and then incompletely.
D. The loosening effect of the three bromine atoms on each other.
This is a feeble effect, as Blau * found it necessary to heat symmetrical
tribrombenzol to 120°-130° with sodic ethylate to obtain much effect.
Calvert and one of us f found that with sodic ethylate a reaction took
place at the boiling point of alcohol in open vessels, but the reaction ran
more slowly, and was less complete than that witli tetrabrombenzol.
We have then the most powerful loosening effect A, and the weakest
loosening effect D, exerted on the atoms of bromine, while the two inter-
mediate effects B and G influence the nitro groups, so that A -{■ D '\?,
very nearly, if not quite, equal io B -\- (J; that is, the attack of the sodic
ethylate will be directed about equally upon the atoms of bromine and
upon the nitro groups, which was the result of our experimental work as
stated above.
A similar discussion of the action of sodic ethylate on tribromdinitro-
benzol gives the following results. The loosening effects are marked by
letters, and arranged in order of intensity, as before.
E. The action of the two nitro groups in loosening the three atoms
of bromine. This is a strong effect, as shown by the ease with which the
bromdinitrobenzol Brl(NOo)22,4, is decomposed by potassic hydrate. $
F. The effect of the three atoms of bromine on the two nitro groups
already discussed.
G. The effect of the three atoms of bromine on one another, which has
also been considered in the previous discussion.
H. The effect of the two nitro groups on each other. So far as we
can find, this effect is too feeble to produce a replacement of a nitro group
by an ethoxy group, but it should not be left out of account, as Lobry de
Bruyn § has found that potassic cyanide in alcoholic solution converts
metadinitrobenzol into CgHsNOjOCaHsCN, which shows a certain loos-
ening of one of the nitro groups.
Here then we have only one strong loosening agency (^E), and this
acts on the atoms of bromine, while of the other three by far the feeblest
(^H) is one of those acting on the nitro groups. It is evident there-
fore that
E+ G> F+ H,
* Monatsh. f. Chem., VII. 630. J Clemm, Journ. Pr. Chem., [2], I. 145.
t These Proceedings, XXXI. 134. § Rec. Trav. Chim. Pays Bas, II. 205.
124 PROCEEDINGS OF THE AMERICAN ACADEMY.
or, putting it into words, the attack upon tlie nitro groups will be insignifi-
cant compared to tiiat on the atoms of bromine. This is in accord with
our experimental results already given, since the attack on the nitro
groups (Reaction III.) was to the attack on the atoms of bromine (Re-
actions I. and II.) as 1 to 6G.
From the yields of the two bromdinitroresorcine diethylethers, given
earlier in this paper, it is possible to draw conclusions in regard to the
effect of the position of the nitro groups in loosening the atoms of bro-
mine, but there is some question whether such conclusions are valid,
since the difference between the yields of the two bromdinitroresorcine
diethylethers is only ten per ceut, and this amount is within the probable
limit of error in this case, where the purifications were made by crystal-
lization. On the other hand, the product melting at 92° (2) was obtained
in the larger quantity, and this is the one where we shovdd expect the
greatest loss, since in its purification the crystallizations were the most
numerous. We feel, therefore, justified in giving the following discussion
with all necessary reserve. In the two reactions (I. and II.) one of the
atoms of bromine replaced occupied in each case the same position (ortho-
para) toward the nitro groups, and therefore the difference between the
reactions depends on tiie position of the second atom of bromine replaced.
In Reaction I. this atom of bromine was ortho to one nitro group and
para to the other, and this reaction gave 28 per cent of the product com-
pared to 38 per cent from Reaction II., in which the atom of bromine
was in the ortho position to both nitro groups. It would seem, therefore,
that the diortho position of the nitro groups exerted a stronger loosening
effect upon the bromine than the orthopara position.
When the reaction between tribromdinitrobenzol and sodic ethylate is
carried on at 70°, the products isolated by us were the dinitroresorcine
diethylether melting at 133°, a little of the bromdinitroresorcine diethyl-
ether melting at 184°, and much of the isomeric substance melting at
92°. These were the only products we have succeeded in identifying,
although we obtained indications of the presence of the tribromnitro-
phenol. There was also a crystalline substance melting at 112°, but
in too small quantity for identification, and a great deal of tarry matter.
The absence of the dinitrophloroglucine triethylether is surprising ;
we should account for it by supposing that the reaction ran for so
short a time (ten minutes) that the bromdinitroresorcine melting at 92°
did not undergo decomposition. The hypothesis that the phloroglu-
cine ether formed was converted into tarry substances by the hot sodic
ethylate seems to us less probable. The approximate yields of the two
products of the reaction at 70° were : —
JACKSON AND KOCH. — TRTBROMDINITROBENZOL.
125
Dinitroresorcine diethylether, melting point 133° .
Bromdinitroresorcine diethylether, melting point 92^
16 per cent.
9 per cent.
Total 25 per cent.
So that in this case 75 per cent of the substance was unaccounted for.
Most of this undoubtedly went into the tarry products.
The most striking phenomenon in the action of sodic ethylate on
tribromdinitrobeuzol at high temperatures is the replacement by hydro-
gen of the atom of bromine in the otho position to both nitro groups.
The fact that this reaction only takes place to a very limited extent in the
cold may perhaps be accounted for by the sparing solubility of the brom-
dinitroresorcine diethylether melting at 184° in the cold alcohol and
benzol, so that most of it is precipitated, and therefore removed from the
action of the ethylate. On the other hand, as this ether is easily soluble
in these liquids, when hot, it would be brought into the sphere of the
reaction under these conditions, and the dinitroresorcine diethylether
would be formed. The bromdinitroresorcine diethylether melting at 92°
is soluble in the cold solvents, and this probably accounts for the fact
that a considerable amount of it was converted into the dinitrophloroglu-
cine triethylether even in the cold.
It is perhaps worth while to call attention to the fact that in the re-
actions described in this paper the action is confined either to the atoms of
bromine or to the nitro groups; that is, if it has started in one set of
radicals in the meta position to each other, it does not extend to the other
set of radicals in the meta position to each other and ortho or para to
the first set. This observation has been made frequently in the course of
the investigation of which this is a part, and the principle has sometimes
given valuable aid in interpreting experimental results. In only a single
case has such an extension of the reaction beyond the limits of the first
set of trimeta positions been observed ; this was in the action of hot sodic
ethylate on tribromtrinitrobenzol,* which gave first tribromnitroresorcine
diethylether, and by further action bromnitroresorcine diethylether
Br Br Br t
NOo
N0.3
C2H5O
OC2H5
C2H5O
OC0H5
Br
Br
Br
Br
H
H
\y
\/
\y
ND2
NO,
NOo
* These Proceedings, XXVII. 315.
t The position of the atom of bromine in this substance has not been estab-
lisiied experimentally.
126 PROCEEDINGS OF THE AMERICAN ACADEMY.
We are not, however, inclined to be too sure of the general occurrence
of this restriction of reactions to a single trimeta zone, since the excep-
tional action just mentioned, in which the second trimeta zone was in-
vaded, was accompanied by the formation of much tarry matter, and
therefore it is possible that the tarry products so frequent in these reac-
tions may have been formed by similar invasions of the second trimeta
zone. The bearing of these observations on the benzol formula of J. N.
Collie* is obvious.
Preparation of Symmetrical Tribromdinitrobenzol.
The description of the preparation of tribromdinitrobenzol is scattered
through a number of papers from this Laboratory, so that it would be
a matter of some difficulty to find it, and there would be danger that any-
one looking it up might not find the latest and best form of the process ;
we have thouglit it well, therefore, to give here a connected account of
this process.
To make tribromaniline, 60 grams of aniline were treated with dilute
hydrochloric acid, and dissolved in four litres of water. Then a rapid
stream of air saturated with bromine vapor was drawn through the
liquid by means of a Bunsen pump, until it assumed a distinct yellow
color, which did not change for several minutes. The amount of bro-
mine required was about 320 grams. The precipitate of tribromaniline
was filtered out through cheese-cloth, washed until free from acid, and
dried by pressing with a screw press, followed by iieating to about 60°
over a steam radiator.
To convert this tribromaniline into tribrombenzol, 100 grams of it
were dissolved in 600 c.c. of alcohol and 150 c.c. of benzol by the aid of
heat ; 40 c.c. of common strong sulphuric acid were then added to the
hot solution from a pipette, and any precipitate formed was dissolved by
longer heating, or even adding more of the solvents. Forty grams of
finely powdered sodic nitrite were next sifted into the hot solution as
rapidly as the violence of the reaction permitted, and the product heated,
until there was no more effervescence, after which it was allowed to stand
at ordinary temperatures over night, and then filtered and dried.
The tribrombenzol was converted into tribromdinitrobenzol as fol-
lows. The perfectly dry tribrombenzol was mixed with four or five
times its weight of fuming nitric acid of specific gravity 1.51, and gently
heated over a low flame for two hours in a flask closed with a porcelain
* 2 Proc. Chem. Soc, 1896-1897, p. 143.
JACKSON AND KOCH. — TRIBROMDINITROBENZOL. 127
crucible, taking care that the temperature was kept just below boiling.
It was then allowed to stand over night at ordinary temperatures, when
most of the tribromdinitrobenzol crystallized out, while the rest of it was
obtained by pouring the supernatant acid liquid into a large excess of
water. The dried product was purified by crystallization from about
eight times its weight of a niixture of three parts of alcohol with one
of benzol.
Action of Sadie Ethylate on Tribromdinitrohenzol at 70°.
The tribromdinitrobenzol used in all this work, made in the way de-
scribed in the previous section, melted at 192°, and had the constitution
lir3l,3,o,(N02)22,4. Forty grams of this tribromdinitrobenzol were dis-
solved in 80 c.c. of benzol, and mixed with the solution of sodic ethylate
obtained from 6.8 grams of sodium and 180 c.c. of absolute alcohol,
wliich gave the proportion of three molecules of sodic ethylate to each
molecule of tribromdinitrobenzol. The flask containing the mixture was
then placed in a water bath, which was heated until a thermometer im-
mersed in the mixture rose to 70°, at which temperature the solution was
kept for ten minutes. The liquid turned dark brown during this heating,
a color which had been found to be characteristic of the reaction with
the aid of heat. At the end of the ten minutes the liquid was poured
into a large evaporating dish, and the solvents allowed to evaporate spon-
taneously. The residue thus obtained was washed with water until tha
wash waters became colorless, and the products purified as follows.
Residue Insoluble in Water.
This residue was dried and extracted three times with boiling ligroin.
The portion insoluble in ligroin was purified by repeated crystallization
from a mixture of ligroin and benzol, during which it was frequently
treated with boneblack. The product was the diuitroresorcine diethyl-
ether already obtained in this way by W. H. Warren and one of us;*
it was recognized by its melting point, 133°, and its solubilities and
appearance.
The ligroin solution on standing deposited a flocculent precipitate in
too small quantity for identification, which was filtered out, and the fil-
trate evaporated to dryness ; the crystalline residue was spread on an
ungluzed plate to remove oily impurities, after which it was purified by
crystallization from hot ligroin. After four crystallizations it showed
* These Proceedincrs, XXV. 170.
Calculated for
CeHBrlOCsH^yNOj);
Bromine
23.89
Nitrogen
8.36
128 PROCEEDINGS OF THE AMERICAN ACADEMY.
the constant melting point 92°, when it was dried in vacuo, and analyzed
with the following results : —
I. 0.2043 gram of the substance gave hy the method of Carius 0.1136
gram of argentic bromide.
II. 0:2885 gram of the substance gave 20.6 c.c. of nitrogen at a tem-
perature of 12° and a pressure of 745 mm.
Found.
I. II.
23.69
8.32
The substance is therefore a bromdinitroresorcine diethylether isomeric
with the one obtained by W. H. Warren and one of us * by the action of
cold sodic ethylate upon the tribroradinitrobenzol and melting at 184°.
As the constitution of this substance has been established (in this paper)
as OC2H5l,OC2ll55,Br3,NOo2,NOo4, it follows that our new bromdi-
nitroresorcine diethylether must have the constitution
OC2H6l,OC2H53,Br5,N022,N024,
as this is the only other body with this composition which can be obtained
from this tribromdinitrobenzol.
Properties of Bromdinitroresorcine Diethylether. Melting Point 92°.
C6HBr(OC2H5)2(N02)2. (OC2H,)2l,3,Br5,(NOo)22,4.
This substance crystallizes in white needles thickly crowded into
bunches shaped like hour-glasses ; when better developed, they form
prisms with blunt ends, which seem to be made up of a basal plane
modified by verv minute planes of a pyramid. On long exposure to the
air, especially in bright light, it turns brownish, and this change is ac-
companied by decomposition. It melts at 92° ; and is freely soluble even
in the cold in benzol, ether, chloroform, acetone, glacial acetic acid, car-
bonic disulphide, or acetic ester ; somewhat soluble in cold alcohol, freely
in hot; sparingly soluble in cold ligroin, freely in hot; essentially
insoluble in water whether cold or hot. Ligroin is the best solvent for
it. A hot solution of sodic hydrate has little or no action on it. On
heating it with sodic ethylate it is converted almost quantitatively into
the triethylether of dinitrophloroglucine formed by the replacement of
its atom of bromine by an ethoxy group; the product was recognized
by its melting point 105°, and its other properties. The wash waters
* These Proceedings, XXV. 166.
JACKSON AND KOCH. — TRIBROMDINITROBENZOL. 129
gave no test for a nitrite, showing that the nitro groups had not been
attacked.
Other Products of the Reaction. — During the purification of the dini-
troresorcine diethylether a small quantity of the bromdinitroresorcine
diethylether melting at 184" was isolated. Its presence shows that the
heating for ten minutes used by us was not sufficient to convert the whole
of it into the dinitroresorcine diethylether.
From the ligroin mother liquors of the bromdinitroresorcine diethyl-
ether melting at 92°, a few milligrams of a body were obtained, which
melted constant at 112°, contained bromine, and gave no test for nitro-
gen after fusion with sodium.* AVe supposed, therefore, that we had
the tribromresorcine, which melts according to Benedikt at 111°, and
could be formed by the replacement of the two nitro groups by ethoxy
radicals, followed by saponification : but upon comparing our specimen
with some tribromresorcine made for this purpose, it was found that,
although the two substances melted at the same point, (as we found
the melting point 11 2*^-1 13° for the tribi'omresorcine,) they differed
markedly in solubility in alcohol, our substance being much the less solu-
ble, and also in the method of crystallization, although the forms were
not incompatible. That our substance was not tribromresorcine was
proved by the action of sodic hydrate, which did not affect our body,
whereas it dissolved the tribromresorcine instantly, giving a solution
which turned black almost at once. The amount of our substance was
not enough for analysis, so we can make no statement in regard to its
nature. We add a description of its crystalline form, which may lead to
its identification. It forms when crystallized from alcohol white very
sharp needles, which when better developed appear as slender prisms
with square ends or bluntly sharpened by several planes at very obtuse
angles to the sides and terminated by a basal plane. When less well
developed, it forms felted masses of needles or bunches, or long sheaves
of needles. None of the arborescent forms were observed which seemed
to be characteristic of the tribromresorcine. This substance was also
characterized by its solubility in ligroin. It was only formed in very
minute quantities under the conditions of the reaction used by us.
The Products of the Reaction soluble in Water.
The wash waters obtained from the product of the action of sodic
ethylate on tribromdinitrobenzol at 70° were highly colored, but the
* We do not feel that this proves the substance is free from nitrogen.
VOL. XXXIV. —
130 PROCEEDINGS OP THE AMERICAN ACADEMY.
substances obtained by acidifying them with acetic acid were so tarry th^t
we could not isolate any body fit for analysis. The filtrate from this tar,
which was also colored, was evaporated to small bulk, and treated with
baric chloride, when a gelatinous precipitate was obtained, which resem-
bled the barium salt of tribromnitrophenol described later in this paper.
We did not however succeed in getting enough of it to purify for
analysis.
The approximate yields of the principal products of the action of sodic
ethylate on tribromdinitrobenzol at 70° were determined as follows : —
Dinitroresorcine Diethylether 16 per cent.
Bromdinitroresorcine Diethylether melting at 92° . 9 per cent.
Total 25 per cent.
From which it appears that 75 per cent of the theoretical yield had been
converted into the tarry mass, from which we could isolate nothing fit
for analysis.
Action of Sodic Ethylate upon Tribromdinitrobenzol in the Cold.
The proportions used were the same as those used in the experiment
at 70°, that is, 40 grams of the tribromdinitrobenzol dissolved in 80 c.c.
of benzol and mixed with the sodic ethylate from 6.8 grams of sodium
and 180 c.c. of absolute alcohol. The mixture was allowed to stand in
a corked flask from three to five days with frequent shaking, after which
it was poured into an evaporating dish, and allowed to evaporate spon-
taneously. The residue was washed with water, until the washings
were colorless, and the products were then purified as follows.
Residue Insoluble in Water.
This was dried, and extracted three times with boiling ligroin, which
left a crystalline residue, and this after recrystallization from a mixture
of alcohol and benzol showed the constant melting point 184°, and was
therefore the bromdinitroresorcine diethylether discovered by W. H.
Warren and one of us in the previous work on this reaction.
The ligroin extract was allowed to stand over night, when it deposited
crystals, which were filtered out, and strangely enough proved to be
essentially insoluble in ligroin. As they could not have been formed in
the ligroin solution, we can explain the extraction of them by hot ligroin
only by supposing they are soluble in a mixture of ligroin and the
other (soluble) product of the reaction. The crystals after purification
JACKSON AND KOCH. — TRIBROMDINITROBENZOL. 131
by crystallization from a mixture of alcohol and benzol showed the con-
stant melting point 105^, which suggested that they were the dinitrophlo-
roglucine triethylether discovered by W. R. Lamar and one of us.* To
settle the nature of the substance, it was dried at 100° and analyzed with
the following result : —
0.2297 gram of the substance gave 19.4 c.c. of nitrogen at a tempera-
ture of 23° and a pressure of 750 mm.
Calculated for
CoU(,OC2H5>s{NOj),. Found.
Nitrogen 9.34 9.42
It is therefore the dinitrophloroglucine triethylether.
Tlie ligroin mother liquor, from which the dinitrophloroglucine tri-
ethylether had been deposited, was evaporated to dryness, and spread on
a porous plate to remove oily impurities, after which it was diss^olved in
hot ligroin, and allowed to crystallize, when in addition to the square
prisms of dinitrophloroglucine triethylether long needles with square
ends were observed, which resembled the bromdinitroresorcine diethyl-
ether melting at 92°, obtained by the action of a hot sodic ethylate solu-
tion on tribromdinitrobenzol, as described earlier in this paper. The iso-
lation of this substance from the mixture of crystals proved a matter of
great difficulty. Hot ligroin did not accomplish this, and a number of
other solvents were tried with no better success. Finally, on soaking the
mixture for some time with cold ligroin, it was found that a considerable
amount of ths square prisms remained undissolved; these were filtered
out, and the filtrate on evaporation gave a residue, which was once more
extracted with cold ligroin in the same way. This second extract
yielded crystals, which could be purified by crystallization from hot
ligroin, when their constant melting point 92° and their solubilities and
crystalline form proved that they were the bromdinitroresorcine diethyl-
ether (OaH5),l,3,Br5,(NO,),2,4.
Although the dinitrophloroglucine triethylether was formed freely in
the cold, we obtained none of it when we carried on the reaction at 70°.
This was probably due to the short duration of our* experiments (ten
minutes), as there is little doubt that this substance is produced by the
further action of the sodic ethylate on the bromdinitroresorcine diethyl-
ether at first formed. It follows that the best way of making the die-
thylether melting at 92° is the process described under the action of sodic
ethylate at 70°.
* Am. Chem. Journ., XVIII. 670.
132 PROCEEDINGS OF THE AMERICAN ACADEMY.
Products of the Reaction Soluble in Water.
The first step in purifying tlie products of the reaction of soclic ethyl-
ate on tribromdiuitrobenzol in the cold consisted in washing with water ;
the highly colored wash waters were concentrated, and acidified with
acetic acid, which produced a flocculent precipitate ; this was dissolved
in dilute alcohol, filtered, and the hot filtrate treated with an aqueous
solution of baric hydrate. On cooling, beautiful leaf-like lemon-yellow
crystals of a barium salt separated, which, after purification by recrystal-
lizing from water, were dried at 100°, and the barium determined with
the following result : — -
0.3916 gram of the salt gave 0.1446 gram of baric sulphate.
Calculated for
[Con.(OC,U3)(NO,)20]oBa2noO. Found.
Barium 21.78 21.71
The retention of the two molecules of water at 100° is certainly strange,
especially as the salt changed from lemon-yellow to orange on heating,
evidently from loss of water, and, as therefore this analysis was not suf-
ficient to establish the identity of the substance, we prepared the free
phenol by decom[)osing the barium salt with acetic acid. It was recrys-
tallized from alcohol, until it showed the constant melting point 77°,
when it was dried in vacuo, and analyzed with the following result : —
0.2125 gram of the substance gave 22.1 c.c. of nitrogen at a tempera-
ture of 15° and a pressure of 752 mm.
Calculated for
C(iH2OUO0.,Il5(NO2)3. Found.
Nitrogen 12.28 12.11
The body is therefore a dinitroresorcine monoethylether, and seems to
be identical with that discovered by Aroidieim* by treating nitrosoresor-
cine ethylether with nitric acid, as the descriptions of the appearance
and solubilities of the two bodies coincide, but Aronheim gives the melt-
ing point 75°, whereas we found 77°. It seemed necessary therefore to
establish the identity of our body more firmly, and this was done by
saponifying a specimen of it by boiling for half an hour with a sulphuric
acid of specific gravity 1.44. As the dinitroresorcine thus obtained
melted at 212° our body has the symmetrical structure
OHI.OC2H55.NO22.NO24,
* Ber. d. chem. Ges., XII. 32.
JACKSON AND KOCH. — TRIBROMDINITROBENZOL. 133
and is the monoethylether corresponding to the diethylether formed by
the action of hot sodic ethylate on tribromdinitrobenzol (see later in this
paper). Whether our substance is identical with that discovered by
Aronheim, or whether the difference of two degrees in the melting point
is caused by a difference in constitution, we are unable to determine with
our present knowledge.
The filtrate, from which the dinitroresorcine monoethylether had been
precipitated by acetic acid, still contained a phenol to judge by its color,
and to obtain this hydrochloric acid was added, which produced a white
precipitate. Without filtering the liquid was extracted with ether, and,
as this removed almost all of the color, the aqueous liquid was thrown
away. The ether on evaporation usually left a red oily mass, which
obstinately refused to crystallize ; on one occasion, however, we succeeded
in obtaining about half a gram of a solid, which after recrystallization
from alcohol showed the constant melting point 90°, and agreed in its
properties with those described for the tribromnitrophenol which melts
at 89° according to Daccomo.* To characterize the substance more
thoroughly some of the oily mass was treated with a solution of baric
hydrate, and the barium salt formed, which liad a gelatinous consistency,
was washed with cold water, dissolved in cold alcohol, from which it was
allowed to crystallize, and after being dried at 100° analyzed with the
following result : —
0.1816 gram of the salt gave 0.0485 gram of baric sulphate.
Calculated for
(CoHBr3NO.O)2Ba. Found.
Barium 15.45 15.70
There can be no doubt, therefore, that the phenol is tribromnitrophenol
formed from tribromdinitrobenzol by the replacement of one of the nitre
groups by one hydroxyl radical. Our substance also shows two charac-
teristic properties, which have been observed in the case of the tribrom-
nitrophenol; these are the solubility of its barium salt in alcohol, and the
fact that the sodium salt is not decomposed by organic acids. The con-
stitution of the phenol follows from its mode of formation from tribrom-
dinitrobenzol melting at 192°. It is Br3l,3,5,OH2.N024.
The following statement of the yields of these various products of the
action of cold sodic ethylate on tribromdinitrobenzol will give a general
idea of the extent to which each of the reactions takes place, but it
should be remembered that these numbers are approximate estimates
* Ber. d. chem. Ges., XVIII. 1167.
134 • PROCEEDINGS OP THE AMERICAN ACADEMY.
rather than strict determiuations, as must necessarily be the case where
the products are purified by crystallization. The numbers are percent-
ages of the theoretical yield in each case : —
C6HBr(OC2H6)2(N02)2, melting point 184° . . 22 per cent.
C6HBr(OC2H5)2(N02)2, melting point 92° . . 19 per cent.
C6H(OC2H5)3(N02)2, melting point 105° . . 19 per cent.
C6H2(OC2H,)OH(N02)2, melting point 77° . . 6 per cent.
C6HBr3(N02)OH, melting point 90° ... . 1 per cent.
Total 67 per cent.
It appears from this that the principal reaction is that which forms the
bromdinitroresorcine diethylether melting at 92°, since the dinitrophloro-
glucine triethylether must be formed from this by a secondary reaction,
and the yield of the two together is 38 per cent. Next after this comes
the reaction forming the bromdinitroresorcine diethylether melting at
184°, and its decomposition product, the dinitroresorcine raonoethylether,
as these together make up 28 per cent of the theoretical yield ; while
the third primary reaction, that which forms the tribromnitrophenol, is of
very inferior importance, yielding only one per cent. These results also
make it probable that no other product was formed in any considerable
amount, as the percentage unaccounted for (33) is no greater than the
loss which would be expected from such numbers of crystallizations, and
other wasteful methods of purification, as were necessary in isolating the
substances enumerated. Small quantities of oily products were observed
at various points in the work, as stated in the description of the purifica-
tion, but except for these all the products were recognized.
Constitution of the Dinitroresorcine Diethylether melting at 133°.
The dinitroresorcine diethylether was boiled for some time with sul-
phuric acid of specific gravity 1.44 in a flask with a return cooler. The
reaction runs smoothly, and gives an almost quantitative yield of the
symmetrical dinitroresorcine (0C^ll,\\,b,(1^0^).^A. This was recog-
nized by its melting point 210° -211° ; Typke * gives 212°.5, Schiapa-
relli and Abellif 214°.5, both of which are higher than that observed by
us ; but, as only two dinitroresorcines could be formed from tribromdini-
trobenzol, the symmetrical one, whose melting point is given above, and
the adjacent, which melts at 142°, there can be no doubt about the iden-
* Ber. d. chem. Ges., XVI. 552.
t Ibid., XVI. 872.
JACKSON AND KOCH. TRIBROMDINITROBENZOL. 135
t'lty of our substance. Many of its other properties also coiucided with
those given for symmetrical diuitroresorciue. It formed vitreous yellow-
ish prisms apparently belonging to the monoclinic system when crystal-
lized from acetic ester, but crystallized from alcohol in the spear-head
forms, which were obtained by Typke from sublimation. The sodium
salt was orange-red, the ammonium salt yellow prisms, the silver salt a
red precipitate which soon turned brown. The acid barium salt con-
sisted of rather thick yellow needles. We did not succeed in getting
the carmine red neutral barium salt. The free phenol decomposes
carbonates.
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. iS^o. 7. — December, 1898.
CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.
ON CERTAIN DERIVATIVES OF SYMMETRICAL
TRIGHL ORBENZOL.
By C. LoRiNG Jackson and F. II. Gazzolo
CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF
HARVARD COLLEGE.
ON CERTAIN DERIVATIVES OF SYMMETRICAL
TRICHLORBENZOL.
By C. Loring Jackson and F. H. Gazzolo.
Presented October 12, 1898. Received October 20, 1898.
In a paper by Sidney Calvert and one of ns,* the behavior of tri-
bromiodbenzol Br3l,3,5,r2, and of tetrabrombeuzol Br4l,2,3,5, with
sodic ethylate was studied, and it was shown that the atom of iodine
(or the atom of bromine in the corresponding position) was replaced by
hydrogen under these conditions, giving the symmetrical trihroml^enzol.
It seemed of interest in this connection to study the corres[)onding tri-
chlor compounds, that is, the trichloriodbenzol and the trichlorbrorabenzol,
to see whether the loosening effect of the three chlorine atoms might
not be even greater than that of the three atoms of bromine, and thus
make it possible that these substances would react with other. agents
beside sodic alcoholates, which were the*only reagents that had any such
effect upon the bromine compounds.
At that time it was not worth while to undertake tlie work, because of
the great difficulty in preparing symmetrical trichloraniline, but since
this obstacle has been removed by the beautiful method of Victor Meyer
and Siidborough,t we have prepared these compounds and studied some
of their relations. Trichloriodbenzol, Cl3l,3,o,I2 melts at 55°; and by
treatment with sodic elhylate dissolved in alcohol and benzol yielded
symmetrical tricldorbeuzol recognized by its melting point 04° and two
analyses. It therefore behaved like the tribromiodbenzol under these
conditions. All our other attempts to obtain a simple replacement of
the iodine in this substance failed, so that it is no more reactive than the
tribromiodbenzol. Fuming nitric acid converted the trichloriodbenzol
into the trichlordinitrobeuzul melting at 129°, iodine being set free.
* These Proceedings, XXXI. 123.
t Ber. d. chem. Ges., XXVII. 315L
140 PROCEEDINGS OF THE AMERICAN ACADEMY.
The tricblorbrombenzol Cl3l,3,5,Br2 melts at 65°, that is, at nearly
the same temperature as the symmetrical trichlorbenzol 63°. 4 (Korner),
and what is strange, higher than the corresponding trichloriodbenzol.
With sodic ethylate it lost bromine, but the reaction was not specially
studied. When treated with fuming nitric acid it gave a trichlorbrom-
dinitrobenzol melting at 175°, which was a decidedly reactive substance.
Aniline replaced the three atoms of chlorine, giving the trianilidobrom-
dinitrobenzol CG(C6rT5NH)3Br(NOo)2, melting at 175° to 176°, discovered
by W. D. Bancroft and one of us.*
Sodic ethylate also acts upon it, probably giving a number of products,
to judge from analogy and the fact that both sodic bromide and sodic
nitrite were detected among them, but we have only succeeded in iden-
tifying one of these with certainty ; this is a bromdinitroresorcine diethyl-
ether melting at 81° to 82°, and probably having the following constitution,
(OC2ll5)2l?3,Br2(N02)24,6, although it may be that one of the ethoxy
groups stands at 5 instead of 3. It must have been formed by the re-
placement of two atoms of chlorine by two ethoxy groups, and the third
by hydrogen. As in most of the replacements of a halogen by hydrogen
it has been found that it stood between two uitro groups, we think there
can be little doubt that the first constitution assigned to this body is the
correct one. Its formation is interesting, as it is the first case we have
found iu which chlorine has been replaced by hydrogen under these
conditions. In all the otlier cases studied the clilorine has remained
unaltered, or has entered into some sim[)le reaction : thus picrylchloride
gave picryl ether,t picrylmalcftiic ester.J or picrylacetacetic ester, §
according to the reagent used ; chloranil gave dichlorquinonedimalonic
ester, || or, so far as the replacement alone was concerned, dichlordi-
ethoxyquinone ; If and trichlordinitrobenzol gave d;chlordinitrophenyl-
malonic ester,** or chlordinitroresorcine diethylether, and diuitrophloro-
glucine triethjdether.ft These last cases are especially striking, since the
corresponding tribromdinitrobenzol showed a replacement of bromine by
hydrogen, when treated with sodium malonic ester or sodic ethylate.
* These Proceedings, XXIV. 293.
t Jackson and Boos, These Proceedings, XXIII. 176.
t Jackson and Soch, Ibid., XXX. 401.
§ Dittrich, Ber. d. chem. Ges., XXIII. 2720.
II Stieglitz, Am. Chem. Journ., XIII. 38.
IT Kelirmann, J. prakt. Chem., [2], XL. .367 ; Jackson and Grindley, These Pro-
ceedings, XXX. 430.
** Jackson and Lamar, Am. Chem. Journ., XVIII. 775.
tt The same, Ibid., p. 668.
JACKSON AND GAZZOLO. — TRICHLORBENZOL. 141
This bromdinitroresorcine diethylether, melting at 81°-82°,
(OCoH5)2l,3, Br2,(N02)24,6 ?
is isomeric with those melting at 184°,
and at 92°,
(OC,UrJA,3,Bv5,(^0,),2,4,
made by the action of sodic ethylate on the symmetrical tribromdinitro-
benzol.
The trichlorbroradinitrobenzol also reacted with sodium raalonic ester
but we were unable to bring the product into a state fit for analvcis.
When the dry substance was treated with sodic ethylate, it gave a deep
vermilion product, which probably belongs to the class of colored com-
pounds formed by sodic ethylate and certain nitro bodies,* as it was de-
composed by water and some organic solvents. This is the first case, so
far as we can find, in which one of these substances has been observed
derived from a compound of benzol with all its atoms of hydrogen
replaced by other radicals.
TrichloriodhenzoJ ^ Cq H2CI3I.
To prepare this substance 17 grams of sublimed trichloraniline (made
by the excellent method of V. Meyer and Sudborough f) were mixed
with moderately dilute sulphuric acid in the proportion of one molecule
of trichloraniline to each molecule ot sulphuric acid, and after thoroutdi
cooling powdered sodic nitrite was added in small successive quantities,
until the nitrous fumes generated were no longer absorbed. After each
addition of the sodic nitrite the flask was corked and vigorously shaken,
until all the red fumes were absorbed, taking care that the contents were
kept cool throughout the operation. When a sufficient amount of sodic
nitrite had been added, the mixture was filtered, the cooled filtrate freed
as completely as possible from the excess of nitrous fumes by vigorous
shaking, and then treated with a distilled aqueous solution of hydriodic
acid, until there was no further action. The brownish precipitate thus
obtained was washed first with a solution of potassic iodide to remove
free iodine, and finally with water, after which it was purified by subli-
mation, or by crystallization from hot alcohol, until it showed the con-
stant melting point of 55°. It was dried in vacuo, and gave the following
results on analysis : —
* These Proceedings, XXXIII. 173, and Amer. Cliera. Journ., XIX. 199, where
a, complete list of the papers on this subject is given,
t Ber. d. cliem. Ges., XXVII. 3151.
142 PROCEEDINGS OF THE AMERICAN ACADEMY.
0.2188 gram of the substance gave by the method of Carius 0.4746
gram of a mixture of argentic chloride and iodide. After wash-
ing this precipitate with ammonic hydrate 0.1672 gram of argentic
iodide were left undissolved.
Calculated for CgUjClsI.
Found.
Chlorine and Iodine
75.94
76.10
Chlorine
34.69
34.73
Iodine
41.30
41.29
The substance is therefore trichloriodbenzol, and as it was made from
common trichloraniline its constitution must be Cl3l,3,5,I2.
Properties of Trichloriodbenzol. — It crystallizes from alcohol in white
slender needles terminated by one plane at a very acute angle ; these
needles are often a centimetre or more long, and are much branched, the
branches forming a sharp angle with each other and developing into
forms like feathers. It melts at 55°, and sublimes easily. It is freely
soluble in ether, benzol, chloroform, acetone, carbonic disulphide, or
ligroiu ; soluble in ethyl or methyl alcohol, when cold, more freely
soluble when hot; soluble in glacial acetic acid; insoluble in water, cold
or hot. A mixture of alcohol and chloroform is the best solvent for it.
It is apparently unaflfected by the strong acids, or by sodic, potassic, or
ammonic hydrate.
Behavior of Trichloriodbenzol with Sodic Ethylate.
Two fframs of trichloriodbenzol dissolved in anhydrous benzol were
mixed with 20 c.c. of an alcoholic solution of sodic ethylate made from
one gram of sodium, and the mixture was allowed to stand over night.
The liquid turned dark brown, and a precipitate began to separate soon
after adding the ethylate. To make certain that the reaction was com-
plete, the mixture was heated on the steam bath in a flask with a return
condenser, which rendered the brown color much darker. The product
was then evaporated to dryness; during the evaporation an odor like that
of an aldehyd was observed, but the presence of one could not be deter-
mined by other tests. The dry residue was treated with water, and the
insoluble portion separated from the solution, which gave tests for an
iodide. The portion insoluble in water, which was very dark brown,
was washed thoroughly, and then purified by crystallization from alco-
hol, until it showed the constant melting point 64°, which proved that it
was the symmetrical trichlorbenzol. This was confirmed by the following
analyses of the substance dried in vacuo : —
JACKSON AND GAZZOLO. — TRICHLORBENZOL. 143
I. 0.0906 gram of the substance gave by the method of Cariiis 0.2154
gram of argentic chloride.
II. 0.0702 gram of the substance gave 0.1862 gram of argentic
chloride.
Calculated for Found.
CcU-iClg. I. II.
Chlorine 58.68 58.78 58.13
Behavior of Trichloriodhenzol with Other Reagents.
"With aniline even at its boiling point trichloriodhenzol showed no signs
of action, except that the color of the mixture became darker, and a
certain amount of turbidity appeared, but no test for an iodide could be
obtained, and the trichloriodhenzol was recovered" unaltered.
When trichloriodhenzol was heated on the steam bath for three hours
with an aqueous solution of sodic hydrate, the liquid took on a chrome-
yellow color, but this must have been due to a very slight reaction, as
after acidification it gave no precipitate with argentic nitrate, and essen-
tially the whole of the trichloriodhenzol was recovered unaltered.
Melted sodic hydrate, on the other hand, seemed to act upon it, as a
brownish mass was obtained, which after solution in water gave a slight
precipitate on acidification and a reddish solution ; a good test for an
iodide was obtained, but the yield of the new organic substance was so
small that we did not study this reaction further, since at best it seemed
to us of slight interest.
Sodium malonic ester had little or no action on the trichloriodhenzol,
most of which was recovered unaltered from the product, so that the
hope of obtaining enough of a substance (if one were really formed) for
analysis was so small that we did not continue work in this line.
From these experiments it appears that the trichloriodhenzol is no
more reactive than the tribromiodbenzol, from which exactly similar re-
sults were obtained by Sidney Calvert and one of us.*
Wlien trichloriodhenzol was mixed with nitric acid of specific gravity
1.50 and strong sulphuric acid, and the mixture gently heated, the solid
went into solution. It was allowed to stand at ordinary temperatures
over night, and then precipitated with a large quantity of water, when
a mixture of a white body and scales of iodine was thrown down. The
iodine was recognized by its crystalline form, color, smell, and purple
fumes. The white body was purified by crystallization from alcohol,
* These Proceedings, XXXI. 128.
144 PROCEEDINGS OF THE AMERICAN ACADEMY.
when it showed the melting point 129°, and is therefore the trichlor-
diuitrobenzol Cl3l,3,o,(iS02)22,4.
Trichlorhrombenzol CgHjClsBr.
Twenty grams of trichloraniline dissolved in 100 c.c. of hot glacial
acetic acid were mixed with 90 c.c. of hydrobromic acid (boiling at 125°),
and, disregarding the heavy grayish yellow precipitate, the mixture was
thoroughly cooled in an ice bath, and then treated with powdered sodic
nitrite, until red fumes were given off freely. The sodic nitrite was
added in small quantities at a time, and the flask containing the mixture
shaken vigorously after each addition, the liquid behig kept cool through-
out. The heavy precipitate already mentioned went into solution during
the addition of the sodic nitrite, forming a dirty yellow liquid. After
standing for a few hours to complete the reaction, the contents of the
flask were poured into an evaporating dish, and heated for an hour on
the steam bath. In this way a dark oily product floating on the surface
of the liquid was obtained, which solidified on cooling. The mother
liquor deposited more of this product in a semi-crystalline condition, and
an additional amount was obtained from it by heating it again with more
hydrobromic acid. The product was purified by sublimation followed by
recrystallization from alcohol, until it showed the constant melting point
64°_65°. As this is the same as the melting point of trichlorbenzol, we
supposed at first that our product was this body, but the following analy-
ses of the substance dried in vacuo showed that it was the desired
trichlorhrombenzol.
I. 0.1618 gram of the substance gave by the method of Carius 0.3850
gram of the mixture of argentic chloride and bromide.
II. 0.1116 gram of the substance gave 0.2656 gram of the mixture of
argentic chloride and bromide.
III. 0.1276 gram of the substance gave on combustion 0.1274 gram of
carbonic dioxide and 0.0108 gram of water.
Calculated for
Found.
C.-.H^ClsBr.
I.
II.
III.
Chlorine and Bromine *
71.60
71.76
71.74
Carbon
27.64
27.23
Hydrogen
0.77
0.94
* By a curious coincidence the results of analyses I. and II., if calculated
as argentic chloride, give numbers agreeing excellently with the percentages calcu-
lated for CfiHgClg, so that the combustion was necessary to determine whether our
compound was this, or the CeHgClsBr which has the same melting point.
JACKSON AND GAZZOLO. — TRICHLORBENZOL. 145
The yield of trichlorbrombenzol was 12 grams, instead of the 26.5
grams which should have been obtaiued from 20 grams of trichloraui-
line, that is, over 45 per cent of the theory.
Properties of TrichlorhrombenzoJ. — It crystallizes from alcohol in white
radiating needles, which develop into long slender blunt ended prisms.
It melts at 64°-65°, that is, one degree higher than the symmetrical
triclilorbenzol, which melts according to Korner at 63°. 4; not only is
this coincidence striking, but it is also surprising that it should melt at
a higher temperature than the trichloriodbenzol, which melts at 55°. It
sublimes easily. It is freely soluble in ether, benzol, or acetone; solu-
ble in cold alcohol, more freely in hot ; soluble in glacial acetic acid, or
ligroin ; less soluble in methyl alcohol, and still less in carbonic disul-
phide. The best solvent for it is alcohol. The three strong acids have
no apparent action on it, but fuming nitric acid converts it into trichlor-
bromdinitrobenzol, as described later. When a benzol solution of the
trichlorbrombenzol was treated with sodic ethylate, the atom of bromine
was removed, as was shown by testing the wash waters from the pro-
duct for bromine, when a strong reaction for it was observed.
Trichlorbromdinitrohenzol CeClgBr ( NO2) .^
To prepare this substance the trichlorbrombenzol was mixed with
nitric acid of specific gravity 1.52 and one third the quantity of strong
sulphuric acid, and the mixture heated gently for an hour. At first the
solid dissolved, but later the nitro compound was deposited from the
solution. After the mixture had stood over night, the supernatant acid
was poured into a large quantity of water, which gave an additional
amount of the j)roduct. It was purified by crystallization from a mix-
ture of alcohol and benzol, until it showed the constant melting point
175°, when it was dried in vacuo, and analyzed with the following
result : —
0.1654 gram of the substance gave according to the method of Carius
0.2914 gram of the mixture of argentic chloride and bromide.
Calculated for
CsClaBrcNO,)^.
Found.
53.21
53.10
Bromine and Chlorine
The constitution of this substance is settled by the method, in which
it was made as Cl3l,3,5,Br2,(NOo)24,6. The yield is essentially
quantitative.
Properties of Tri clilorhromdinitrohcnzol. — It forms, when crystallized
VOL. XXXIV. — 10
146 PROCEEDINGS OP THE AMERICAN ACADEMY.
from alcohol and benzol, thick rather blunt rhombic plates, the obtuse
angles of which are frequently bevelled by two planes. They show a
tendeocy to form groups with the members superimposed, and have a
slight yellowish tinge. The substance melts at 175°; and is very sol-
uble in chloroform, acetone, or carbonic disulphide ; slightly soluble in
ethyl or methyl alcohol either cold or hot ; insoluble in cold, soluble in
hot glacial acetic acid ; essentially insoluble in ligroin, oi* in hot or cold
water.' The best solvent for it is a mixture of benzol and alcohol.
Strong hydrochloric or sulphuric acid has no apparent action on it.
Fuming nitric acid dissolves it when hot. It suldimes easily, and in this
way feathery ivory-white crystals are obtained sometimes over a centi-
meter in length.
Action of Aniliue on Trichlorhromdinitrohenzol.
When one gram of trichlorbromdinitrobenzol was warmed gently with
a slight excess of freshly distilled aniline, it went into solution forming
a cherry-red liquid, the color of which became deeper on longer heating.
To obtain the product the liquid was poured into a large quantity of
water, acidified with hydrochloric acid, and the crimson precipitate formed
in this way thoroughly washed, and crystallized from a mixture of alcohol
and benzol, until it reached the constant melting point 175°-176°.
This showed that the substance was the bromdinitrotrianilidobenzol
CGBr(NO.,)o(CGH6NH)3 obtained by W. D. Bancroft and one of us*
from tetrabromdinitrobenzol and aniline. In this case it was formed by
the replacement of the three atoms of chlorine by three anilido groups.
Behavior of Trichlorhromdinitrohenzol ivith Sodic Ethylate in the Cold.
Five grams of trichlorbromdinitrobenzol dissolved in anhydrous ben-
zol were mixed with the sodic ethylate made from twenty-five grams of
absolute alcohol and one gram of sodium, which gave the proportion of
three molecules of the ethylate to each molecule of trichlorbromdinitro-
benzol. The two substances reacted at once, since the liquid took on a
bright scarlet color as soon as they were mixed, and tliere was also a
slight evolution of heat. To complete the reaction the mixture was
allowed to stand three days at ordinary temperatures, during which time
a heavy precipitate was deposited, and the color changed to a yellowish
red. The precipitate was filtered out, washed with alcohol, and then dis-
solved in water. This solution gave a strong test for sodic nitrite with
* These Proceedings, XXIV. 293.
JACKSON AND GAZZOLO. — TRICHLOEBENZOL. 147
potassic iodide, starch paste, and dilute sulphuric acid, and also a heavy-
white precipitate with argentic nitrate and nitric acid. To obtain the
organic products of tlie reaction the reddish alcoholic filtrate was allowed
to evaporate spontaneously, and the residue treated with water ; the in-
soluble substance thus obtained was purified by crystallization from alco-
hol, until it showed the constant melting point 81°-82°, when it was
dried in vacuo, and analyzed with the following result : —
I. 0.1236 gram of the substance gave by the method of Carius 0.0684
gram of argentic bromide.
II. 0.2500 gram of the substance gave 18.4 c.c. of nitrogen at a tem-
perature of 23° and a pressure of 754.7 mm.
Calculated for Found.
CuHBrtCaHjOjCNOj),. I. II.
Bromine 23.89 23.56
Nitrogen 8.36 8.24
The substance is therefore a broradinitroresorcine diethylether formed
from the trichlorbromdinitrobeiizol by the replacement of two atoms of
chlorine by ethoxy groups, and of the third by hydrogen. Certain
points in regard to its constitution are settled, since the two ethoxy groups
must be in the meta position to each other, and the atom of bromine and
the two nitro groups are in the symmetrical position to each other. The
radicals therefore are probably arranged as follows,
(C2ll50)2l,3,Br2,(N02)24,6,
but it is possible that one of the ethoxy groups instead of the atom of
hydrogen stands at 5 between the two nitro groups. It is isomeric with
the bromdinitroresorcine diethylether melting at 184°, and made by
Warren and one of us * from tribromdinitrobenzol and sodic ethylate in
the cold, which has the constitution (C2H50)2l,5,Br3,(N02)2,4, and also
with that melting at 92° obtained by Koch and one of us, f as another
product from the same reaction, which has the constitution
(C2H50)2l,3,Br5,(N02)22,4.
The yield of the bromdinitroresorcine diethylether was one gram from
five of tiie trichlorbromdinitrobenzol, that is, about 21 per cent of the
theoretical yield.
* These Proceedings, XXV. 1G6. t These Proceedings, XXXIV. 128.
148 PROCEEDINGS OF THE AMERICAN ACADEMY.
Propel ties of Bromdinitroresorcine Dletliylether, melting at 81°-82°,
CGH(C.,H50)oBr(NOo)2.
It forms when crystallized from alcohol white ueedles or slender prisms
terminated by two planes at an obtuse angle to each other, which turn
brown on exposure to the light. It melts at 8l°-82°, and is very soluble
in benzol, or chloroform ; soluble in methyl alcohol, acetone, glacial acetic
acid, or carbonic disulphide; slightly soluble in cold ethyl alcohol, more
soluble in hot ; slightly soluble in ligroin. Alcohol is the best solvent for
it. It is not acted on ap|)arently by strong hydrochloric acid, either hot
or cold ; strong sulphuric acid does not act in the cold, but when warm
dissolves it with a brownish red color; strong nitric acid does not act in
the cold, but gives a colorless solution with it when hot.
That tliis bromdinitroresorcine diethylether was not the only organic
product of the reaction of sodic ethylate on trichlorbromdinitrobenzol was
shown by the fact that sodic nitrite as well as sodic chloride was formed.
Our attempts to isolate these other bodies, however, were far from suc-
cessful. Some experiments with the aqueous wash waters seemed to
indicate that they contained a phenol melting at 111°, and having pei-
haps the formula Cg(OC2lI)20II(N02)2lir> but the analytical data
obtained were too imperfect to justify a description of the body ; and we
did not succeed in bringing any of the other products into a state fit for
analysis.
The trichlorbromdinitrobenzol seemed to give colored compounds with
sodic ethylate similar to those given by picrylch oride and several other
nitro compounds, since upon adding an alcoholic solution of sodic ethylate
to the dry substance it took on a strong vermilion color, which was in-
stantly destroyed by w'ater, and more slowly by benzol or ligroin. Tiie
fact that the color disappeared on the addition of water indicates that it
is one of the colored compounds under discussion, and not a salt of a
phenol. Another sample of the color was allowed to stand for half an
hour exposed to the air, at the end of which time the red color had given
place to yellow. As the colored compound was less stable than several
others which have been studied, its investigation was not carried further.
The trichlorbromdinitrobenzol is acted on by an alcoholic solution of
sodium malonic ester. Our first experiments gave a crystalline product,
but in too small quantity for analysis. Our later experiments have
yielded only viscous masses, from which we have not succeeded in ob-
taining anything for analysis in spite of a very large expenditure of time
and material.
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. No. 8. — January, 1899.
Proceedings of the American Academy, Vol. 32.
ERRATUM.
On page 333, line five, of the Memoir of Francis James Cliild, for
"The boy was the youngest," etc., read: "The boy was the third in
a family of eight brothers and sisters."
148 PROCEEDINGS OF THE AMERICAN ACADEMY.
Properties of Bromdinitroresorcine DietJnjlether^ melting at 81°-82°,
C6H(C2H50)2Br(NO.)2.
It forms when crystallized from alpohol white needles or slender prisms
terminated by two planes at an obtuse angle to each other, which turn
brown on exposure to the light. It melts at 8l°-82°, and is very soluble
in benzol, or chloroform ; soluble in methyl alcohol, acetone, glacial acetic
acid, or carbonic disulphide; slightly soluble in cold ethyl alcohol, more
soluble in hot ; slightly soluble in ligroin. Alcohol is the best solvent for
it. It is not acted on api)arently by strong hydrochloric acid, either hot
or cold ; strong sulphuric acid does not act in the cold, but when warm
dissolves it with a brownish red color; strong nitric acid does not actiu
the cold, but gives a colorless solution with it when hot.
That this bromdinitroresorcine diethylether was not the only organic
product of the reaction of sodic ethylateon trichlorbromdinitrobenznl wmc
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. No. 8. — January, 1899.
SHORELINE TOPOGRAPHY.
By F. p. Gulliver.
SHORELINE TOPOGRAPHY.*
By F. p. (iULLIVER.
Presented by W. M. Davis, October 12, 1898.
Received October 2'J, 1898.
" When I have seen the hungry ocean gain
Advantage on the kingdom of the shore,
And the firm soil win of the watery main,
Increasing store witli loss and loss with store ;
When I have seen sucii interchange of state,
Or state itself confounded to decay ;
Ruin hath taught me thus to ruminate."
Shakespeare, Sonnet LXIV.
INTRODUCTION.
Physiographic Standpoint. — The present paper deals with the devel-
opment of coasts from a geographic standpoint, and attempts to workout
the criteria by which we may determine whether a given coastal area
stands now at a relatively higher or lower level with reference to the
level of the sea than it did in some previous cycle or portion of a cycle.
Particular emphasis will be laid upon the stages of development, which
follow each other according to dynamic laws in a systematic succession,
both after uplift and after depression. The time since the beo-iuuino- of
the cycle or epicycle t is found to have a very important bearing upon
the question of continental oscillation. The dynamic forces of nature do
not leave the initial forms produced by uplift or depression, but produce
a successive series of sequential forms, which may be used, when the
order of the normal succession is apprehended, as criteria to determine
the time since the cycle began.
Omitted Phases of the Subject. — The shoreline, the line formed by
the intersection of the plane of the sea with the land, is in a geographic
sense a most inconstant line. Though for the geographic minute, a
* This paper was written for the Doctorate of Philosophy at Harvard Uni-
versity, and was presented in June, 1806. It lias been condensed and slightly
revised for publication. September, 1898.
t See p. 154.
152 PROCEEDINGS OF THE AMERICAN ACADEMY.
generation of men, it is practically in the same position, yet even in the
short period of historic time records show that villages have been sub-
merged, or that seaport towns have been turned into inland places.
The historical side of this problem is not to be here discussed, nor is
the cause of the secular movements of the earth's crust, including the
question of isostasy, here considered.
As dwellers upon the land, we look at the change in the relative posi-
tion of land and water as it affects our position. Thus " the land rises,"
" the coast sinks," are the common expressions of man. If the point of
view was that of the inhabitants of the sea, the expressions would be
reversed, the sea sinks when the land rises and rises when the land is
depressed. It will be convenient in this paper to use the terms elevation,
uplift, emergence, raised, etc., and their opposites, depression, submer-
gence, sunken, etc., in reference to the land. Such use is not intended
to imply "a limitation of movement to the land, excluding that of the sea
floor, but is to describe the geographic efTects from the standpoint of
man, who lives upon the dry land.
The shrinking of the mass of oceanic waters will also cause the land
apparently to rise to the same amount all over the world.*
By the draining of lakes many characteristic forms of shore develop-
ment will be exposed, which are here classed with forms following up-
lift; while the increase of water in a lake for any cause will give the
same forms as are produced by a depression of the land.
The relation of the accumulation of glaciers to changes of elevation,
and the evidence afforded by coral islands to show rising or sinking
regions, are two problems for solution which are not attempted by the
present writer.
Use of Terms. — Throughout this paper the author uses shoreline for
the line of intersection of the sea with the land. The region immedi-
ately to the landward of the shoreline is called the coast, and seaward
from this line the shore. Thus cliffs and deltas are coastal features, while
waves advance and retreat along the shore.
In the figures the older mainland is cross-hatched, while forelands are
left blank. The observer is supposed to look from the point of view of
the sea as it attacks the land, therefore the two sides of the figures will
be spoken of as the right and left respectively as seen from the sea look-
ing toward the land.
* The effect in inland seas with imperfect outflow has been discussed by Prof.
Suess, Anzeiger d. k. Akad. d. Wiss., 1887, XXIV. 180-182.
GULLIVER. — SHORELINE TOPOGRAPHY. 153
Initial is here used as the technical term to define the form at the
beginning of any geographic cycle or epicycle. Any dynamic process
which produces a change in the relative position of land and sea may
interrupt a cycle at any stage of development, and introduce a new
cycle. Later stages and forms will be called sequential. These terms
are offered to avoid the misconception, on account of their vernacular
meaning, of the terms constructional and destructional, sometimes used
for the identical ideas.
Previous Work on Shorelines. — Since the days of Strabo and Aristotle,
two of the greatest observers among early geographers, much has been
added to the science of geography. Passing over the work * of the car-
tographers, explorers, and speculative writers, mention must be made of
the great mass of facts collected by Ritter and Humboldt and of their
use by Guyot ; but the great outdoor observer, De la Beche, whose work
was the stock in trade of the next generation, first interpreted many of
the coastal forms. He in 1834,t and Dana more fully in 1849,$ recog-
nized land-carved forms under water, or drowned valleys, as proof of
depression of the land. Robert Chambers recognized raised beaches and
associated coastal forms, and showed that the Atlantic coastal plain indi-
cated elevation. § Lyell with his doctrine of uniformity, Ramsay with
the tlieory of marine denudation, the Geikies, LeConte, Darwin, and
many other geologists, have worked out the changes in form of coasts
here grouped under various sequential stages.
Members of the United States Surveys, Bache, Mitchell, Gilbert,
Shaler, Whiting, Davidson, and others, have worked out many of the
details of coastal forms and their changes, and a large number of obser-
vations recorded upon maps and charts have been the basis of much of
the work in this paper.
In 1879, Dr. Halm discussed the rising and sinking of coasts, but he
did not consider the ratios between activities nor take into account the
time since which a given movement took place. Weule, Cold, Keller,
and Sandler have also studied shorelines, but the fullest discussion of
coastal and shore forms has been made by von Richthofen and his pupil.
Dr. Philippson.il
* See Lyell, Prin. Geol., 11th ed., 1872, 22, 57 ; Woodworth, Am. Geol, 1894,
XIV. 210.
t Theoretical Geology, 192-194.
X Geology of the Wilkes Expedition, 1849, 677.
§ Ancient Sea Margins, 221, 253, 270, 276, 299.
II See list of references for these and other papers.
154 PROCEEDINGS OF THE AMERICAN ACADEMY.
PART I. INITIAL FORMS.
1. The Geographic Cycle.
Systematic Sequence of Forms. — Before the consideration of the initial
forms themselves is undertaken, the position of the initial stage in relation
to geographic cycles of development must be clearly understood. Conse-
quently at the outset some of the general facts of cycles will be discussed
in their bearing upon the problem of stages in the development of shore-
lines, and particularly as regards the initial or first stage of a cycle.
In this paper many facts from different sources are brought together,
and the attempt is made to show some of the laws of coastal development.
After the inductive study of coast forms upon the better mapped areas of
the world had been made, and the deductive scheme of development
worked out, gaps in the scheme were found which are not filled by ex-
amples. This lack of facts to fit ideal cases may be because they do not
exist upon our small earth, or because they have not been reported, as
well as on account of a defective scheme. By showing w here such gaps
in our theoretical scheme of development occur at present, the eyes of
future field workers may be sharpened to look for the expectable facts.
Succession on Land. — Land forms go progressively through a series of
successive stages of development, to which have been applied names taken
from various stages of life, thus suggesting that forms as seen to-day
began as something else, and will as time advances become systematically
still further developed. Stages of the cycle follow one another from birth
to death in the ideal case, where the land stands still long enough for the
completed development. The initial stage, or birth, is succeeded in turn
by infancy, youth, adolescence, maturity, past-maturity, old age, and
finally by death.
A new cycle is inaugurated by each oscillation of any considerable
amount, minor changes of level being included as epicycles, or divisions
of a cycle. Land forms advance successively from infancy toward old
age in each cycle, while any stage of development may be arrested by
elevation or depression of the land and a second cycle begun. An essential
conception is that a region will be finally reduced to a peneplain if the
baselevelling action of the streams, and the other forces of subaerial
degradation, be allowed to continue long enough to reduce the land
forms to extreme old age. Insequent, consequent, subsequent, and ob-
sequent streams all play their part in the development of the land forms,
captures of one stream by another follow unequal chances, while super-
GULLIVER. — SHORELINE TOPOGRAPHY. 155
posed streams often corae unexpectedly upon a difficult piece of work.
Any one unacquainted with the details of this scheme, as wox'ked out bv
Professor Davis, will be referred to the articles given below.*
Succession on the Coast. — At the beginning of a cycle the subaerial
forces of degradation enter upon a new piece of work. Similarly the sea
has to begin anew its attack upon an initial coast. A series of coastal
forms would be expected to result, and these may be grouped in stages
analogous to those of land forms. On account of the many variables which
control topographic form, it would not be expected to find the inland
area and the coast of the same region in homologous stages of develop-
ment. The general surface of a coastal plain may be in youth or ma-
turity when its coastline has advanced to adolescence. Because the
coastline has reached an adolescent stage of development, it does not
follow that the surface of the coastal plain further inland is also in
adolescence.
The initial stages of coast and inland surface begin together, for both
are controlled by a relative change in position of land mass to sealevel.
In making out the initial and sequential stages of shore development from
an inductive study of coasts and shores, the writer has tried to follow
the principles used by Professor Davis in his studies of the stages of land
forms, aided by his critical suggestions during the progress of the work.
The thesis of this paper is : The forms of any coastal bp:lt may
BE GROUPED IN THE APPROPRIATE STAGES OF A CYCLE. TlIESE FORMS
WILL BE CONSISTENTLY RELATED TO THE ASSOCIATED LAND AREA OX
THE ONE HAND AND TO THE SEA BOTTOM ON THE OTHER. WhEN CON-
SIDERED TOGETHER, THE FORMS OF A COASTAL BELT INDICATE THE
RELATIVE TIME SINCE THE LAST CONSIDERABLE UPLIFT OR DEPRES-
SION, AS WELL AS THE RATIO EXISTING BETWEEN THE SEVERAL ACTIV-
ITIES, IN THEIR DYNAMIC EFFECT UPON THE FORMS OF THE COAST AND
THE SHORE.
Rising, raised ,' sinking, sunken. — In considering any of the conse-
quences of continental oscillations, care must be taken to discriminate be-
tween the movement of the land during historical time or the geographic
to-day, its movement during the immediate past of geographic time, and
the last movement of any considerable amount. Because there is good
evidence of either a geologic or a geographic character, that a given land
has moved either up or down during the period of more careful observa-
* Nat. Geog. Mag., 1889, I. 12-26, 183-253. Proc. B. Soc. Nat. Hist., 1889,
XXIV. 365-423. Am. Nat., 1889, XXII. 566-583. Bull. G. S. A., 1891, II. 541-
586. Geographical Illustrations, Harvard University, 1893.
156 PROCEEDINGS OF THE AMERICAN ACADEMY.
tion of the last century, it does not follow, from such observations 'per se,
that the land has moved in that same direction for any length of time
previous to the earliest of said observations ; and moreover, if in addition
to such demonstrated recent movement there exists geologic or geographic
evidence to show earlier motion in the same direction, such cumulative
evidence of motion in one direction is no valid argument for continued
motion in the same direction, for any period of time longer than that re-
quired by the nature of the evidence itself The converse of this propo-
sition is also true, viz. that evidence which shows that a country has
been raised or sunken does not prove that the region has been rising or
sinking in recent time, or is to-day rising or sinking.
That the time element has been left out of the majority of previous
considerations of shorelines, in the discussion of their elevation or depres-
sion, will be clearly perceived by any one who vs^ill go over the literature
of the subject.
Areas of Elevation and Depression as Mapped. — A comparison of
three maps of " rising " and "sinking" regions will show how different
points of view have led to opposite conchisions. Dr. G. R. Credner
considers the presence of large deltas a proof of slow rising, therefore in
his map of changes of level * he regards all regions of great deltas as
rising. In contrast with Credner's map, compare. that of upheavals and
depressions by Reclus.f In this map, " drawn after Chas. Darwin," all
regions where the coral growth is prevailingly of the atoll and barrier
reef type are given as sinking, while those regions where the fringing
reefs occur are mapped as regions of upheaval. Thus the hypothesis
of Darwin, t who regarded the form of coral construction as evidence of
" probable subsidence " and " probable elevation," when given such defi-
niteness as upon the map of Reclus, makes a striking contrast to the map
by Dr. Credner; many "rising areas" on the one are "sinking areas'"'
upon the other. Neither of these two criteria can be safely used to show
present or recent movement, nor are they more than hypothetical sug-
gestions of earlier changes of level.
Scandinavia on both of these maps is given as rising, because of its
raised beaches and " water-marks." This peninsula, as will be shown
later, is in the larger geographic sense in a cycle of development follow-
ing depression.
An interesting historico-geographic study could be made by a com-
* Pet. Geog. Mitt., Erg. Nr. 56, 1878, Tom. III.
t La Terre, 1872, I. 702, PI. XXIV.
t The Structure and Distribution of Coral Reefs, 3d ed., 1889, PI. III.
J 1 I L
Kilometers.
FiGCRE 1. Casco Bay, Maine. Drowned Topography in Youthful Stages of
Development. (Sheet 315, U. S. C. G. S.)
158 PROCEEDINGS OP THE AMERICAN ACADEMY.
parison of the facts used as criteria of rising and sinking in central and
southern Europe. For our present purpose, however, it will suffice to
indicate how loosely criteria have been interpreted. Compare the two
maps referred to above with that of France by Girard.* Each of the
three differs from both of the others. Southwestern France, the Landes
coast, is given by Credner as sinking, by Reclus as rising, and by Girard
as sinking in some places and rising in others.
Algebraic Sum of Movements : Maine. — The position of the land in the
present cycle is determined by the algebraic sum of all past oscillations.
The form is due to development in (n + 1) cycles. Cycles and epicycles
previous to tlie present may be recognized in inverse proportion to the
time since their close, and in direct proportion to the stage of develop-
ment reached in said cycle or epicycle. A region is classed in this paper,
as in a certain stage of development following elevation or depression,
according to the larger facts of form prevailing in the region.
For example, the coast of Maine (Figure 1) is on the whole a depressed
region. It has numerous islands, bays, etc., showing drowned topography
in a youthful stage of development. At Ogtniquit,| however, as well as
in other parts of this area, are seen criteria of elevation, elevated shore-
line, narrow coastal plain, nip, lagoon, and enclosing ofTshore bar. Since
the greatest depression, there has been an episode of elevation. The
development of the sequential features, following the initiation of a new
cycle of depression, has been interrupted by this episode or epicycle of
elevation.
Figure 2. Diagram sliowing Mutual Relations of Cycle, Epicycle, and Vibration.
Cycle, Epicycle, and Vibration, Neio Jersey ; Scandinavia. — It is only
to the larger movements of the land to which the term cycle is applica-
ble. The minor ups and downs of the coast are but portions of a cycle,
each of which may be called an epicycle, which in turn may be made
up of various smaller swings or vibrations. The relations of these
various movements to one another is shown in the accompanying diagram
(Figure 2).
* Soulevements et depressions du sol sur les cotes de France, Bull. Soc. Geog.,
1875, X. 225, etla carte,
t See pp. 185, 188.
GULLIVER. — SHORELINE TOPOGRAPHY, 159
Professor Salisbury has given us a very pretty example of such epi-
cycles in' his Beacon Hill and Pensauken subdivisions of the Yellow
Gravel of New Jersey, and of what we may term vibrations in the
Jamesburg and later subdivisions.* He has found it possible to deter-
mine from the geographic form and position of the deposits the change
of level of the coast, though the changes are relatively so small that the
evidence of movement cannot be traced far inland. It is possil)le, on the
other hand, to trace the Tertiary peneplain for a considerable distance into
the interior, where for instance it is seen in the floor of the Great Valley ;
while the Cretaceous peneplain is the great surface of reference for
geographic features in the eastern United States. These cycle features
must not be regarded as the result of some sudden massive uplift, but
rather as the summation of minor vibrations and epicycles, during which
the average position of the land was such as to cause the Tertiary and
Cretaceous peneplains.
Scandinavia (Atlas Univ., 29, 30) is a good example to show the
differences between cycle, epicycle, and vibration. Taken as a whole,
the peninsula is a depressed region, some portions being deeper drowned
than others. Two typical areas will illustrate this. The form of the
region around Stockholm (Swe., 67, 68, 75, 76, 77, and adj*ent sheets,
Swe. Geol., 50, 51, 52, 53) indicates that it was maturely dissected in
the previous cycle, and is now submerged to a greater and greater amount
out from the shore, as is shown by the large islands near shore, the
smaller islands off shore, and the minute islets and skerries out in the
Baltic. Baron de Geer makes the axis of greatest uplift in the recent
episodes of elevations in the central portion of Scandinavia.f This tilting,
at whatever time it occurred, is indicated by the increasing relief in certain
directions of "the topography of this area.
The second region is in central Norway (Nor., 45, C, D ; 46, C, D ;
48, B; 49, A, B, C, D; 50, A, B, C, D ; 52, B, D ; 53, A, C, D; 56,
A, B). This area shows adolescent dissection of the upland, the land
being more continuous than in the first region mentioned.
While in a large geographic way the Scandinavian peninsula is a de-
pressed area, there have been epicycles of elevation in which terraces
have been cut. | Recent vibrations are also shown by changes of water
level at the established water-marks (R. Seiger).
* Ann. Rep. State Geol. N. J., 1893, 3&-328.
t See references.
t See papers by Briigger, Chambers, de Geer, Hogbom, Kjerulf, Lyell, Miller,
Mohn, Munthe, Pettersen, Reusch, Sandler, Sexe, and Sieger.
160 PROCEEDINGS OP THE AMERICAN ACADEMY.
Episodes of depression occurring after those of elevation or alternating
with them, if they occurred, must have been of short duration, as well as
those of elevation, for there is no indication that the development of
coastal features has continued for a great length of time at any level
since that at which the adolescent to mature dissection took place. A
possible exception to the above is the short cycle represented by the rock
bench called by Dr. Reusch "the coast plain " (Nor., 6, B; 45, C, D ;
46, C; 48, B ; 49, A; 53, C ; 56, A, B). From the form it is impos-
sible to tell whether this was cut before or after the deepest valley
dissection, shown by the present fjords. In his English summary (the
writer is not able to read the Norwegian paper) Dr. Reusch says, " It
has been worked out in periods previous to the glacial period, and in the
intervals of that time." * If it is later than the deeper dissection some
traces of the material filling the bays should be found, though the glaciers
would have carried off most of the loose detritus.
Volcanic and Climatic Accidents. — In this paper the shore features
that result from the accidents, volcanic and climatic, which are not an
essential part of the normal cycle, are not considered in detail. With the
general scheme of the normal development of shorelines following eleva-
tion and depression in mind, a study of the accidental interruption of the
normal succession can profitably be made. The volcanic features as
shown in Etna (Italy and Sicily, 269, 270, etc.), and Santorin (Fouque,
Santorin et ses eruptions, Paris, 1879) and the glacial features as seen
on Oland island (Swe., 17, 22), in Boston harbor (C. S., 337), Green-
land, and Alaska ; and the arid coasts of Arabia and the shores of the
Red sea, etc., all furnish an attractive field for special study.
Geographic and Paleontologic Criteria. — By the emphasis laid on ge-
ographic criteria for the recognition of change of level and time since the
initiation of a new cycle, it must not be inferred that the writer implies
any lack of confidence in the value of evidence from the position of life
forms. Geography and paleontology should go hand in hand in showing
past changes of level, as where one fails the other may avail. While his-
torically paleontology has had the lead, perhaps the more natural leader
would be geography ; then the indications, given by the inductive study
of the form of a region, may be confirmed by its contained fossils.
Ideal Areas. — Two areas of strongly contrasted conditions are taken as
types. In each area the development of coastal forms has been consid-
ered to have advanced to late adolescence or into maturity in the previous
* See references.
GULLIVER. — SHORELINE TOPOGRAPHY. 161
cycle. In the first area the land is supposed to have risen with respect
to the water far enough to bring all the features of shore development
of the previous cycle above baselevel ; while in the second area these
features are depressed beneath baselevel. Criteria are worked out for
these two normal or average conditions, and later other possibilities will
be considered in connection with actual regions.
2. Uniform Uplift.
Initial Stage of an Ideal Area. — Let it be conceived that a region be
elevated as a unit to a certain distance above sealevel. The geologic
cause of such uplift need not be considered here, as this paper treats of
Figure 3. Ideal Block in Initial Stage following Uniform Uplift. Such a Re-
gion shows Smooth Bottom, Simple New Shoreline, Smootli Coastal Plain,
Elevated Former Shoreline, and the Dissected Oldland.
VOL. xxxiv. — 11
162 PROCEEDINGS OF THE AMERICAN ACADEMY.
the geographic results of continental movements. Enough that it be
granted as a possibility of geology that such uplift of a land mass may
take place.
The form of the land at the initiation of a new cycle of development is
a most important consideration, and it is one which is most frequently
left out of the discussions of elevation and depression. The form of the
land at the beginning of a cycle depends upon the stage of development
reached in the previous cycle, as well as upon the amount and rate of up-
lift. The ideal case here considered is taken where a coast of homogene-
ous structure had been developed to late adolescence or early maturity in
the previous cycle, and the uplift was supposed to have been sufficient in
amount to bring all the coastal and shore forms, developed in the pre-
vious cycle, considerably above sealevel ; and this uplift took {dace, not
suddenly, but steadily, so that the sea did not have time to appreciably
attack the land while it progressively rose. A diagrammatic represen-
tation of the resulting form is given in Figure 3. The forms of the
shoreline and of the inland and seaward areas will each be separately
considered.
(1) Smooth Bottom. — The waste from the land, brought down by
the streams or worn off the coast by the waves, would have been spread
out by the currents in the previous cycle, thus causing the bottom to be
smooth out from the new shoreline. In the ideal case which we are here
considering, the bottom would consist of the finer waste of the previous
cycle, where the sea currents had built it up into the continental delta, at
a depth below the deepest wave attack. Such sedimentation in the pre-
vious cycle would have filled any irregularities then existing, so that the
bottom offshore from the initial shoreline would be monotonously level
or gently undulating.
(2) Simple New Shoreline : Buenos Ayres. — Where the ocean or
other large body of water now intersects the land there will be initial
shore features. At first before the waves have had time to attack the
coast the outlines will be simple, the land gently sloping toward the sea
and ending in broad, undulating curves, probably convex where large
rivers enter. As the initial land surface would have but a slight dip
seaward, it having been formed under water, the sea would leave ex-
posed at low tide a wide area of flats. The most marked feature in
this new born shoreline is its slight crenation and long curves. It would
take but a faint convexity of the land surface to give a convex shoreline.
The Ai'gentine Republic southeast of Buenos Ayres has a shoreline
upon a gently sloping land, very nearly flat. Before the present chan-
GULLIVER. — SHORELINE TOPOGRAPHY. 163
nel was dredged to enable ships to reach the city, the steamers had to
discharge their freight into lighters, and these in turn to wagons driven
into the water. This region appears to be the one least advanced at
present beyond its initial stage, and is therefore given as the best ex-
ample known to the writer of an initial shoreline following elevation.
There is no good account of this coast, the fragramentary hints given by
travellers being the only descriptions which we have ; and the poor maps
(H. C, 616, 930) show little else of coastal and shore forms besides the
gently swinging shoreline.
(3) Smooth Coastal Plain: Texas. — A coastal plain ought to be
found along the margin of the uplifted area, wider where there had pre-
viously been an extensive continental shelf, narrower where less waste
had been deposited in tlie previous cycle ; but with its inner margin at
practically the same height on all sides of the elevated mass. Conse-
quent drainage would characterize this uplifted shelf, while extended
rivers from the oldland would flow across the coastal plain as master
streams.
The coastal plain of Texas, according to the account given by Professor
Penrose,* is a flat plain with the streams lying almost upon the surface,
which has a gentle seaward slope. This plane surface appears to be
nearly in its initial stage of development. The surveyors report that there
is " nothing to map " in this coastal plain area. The shoreline is not
consistently related to the surface of this coastal plain, for it has suffered
since the elevation a sHght episode of depression, as is indicated by the
narrow bays, where the sea has entered the lower portion of the valleys,
which the coastal plain streams had begun to widen.
(4) Elevated Former Shoreline : San Clemente, Figure A. — At the inner
margin of this coastal plain we should find shoreline features younger or
older according to the conditions of development of the region before its
uplift, but at a practically uniform elevation above the sea at the present
time. Reasonable variations in the height of the beach as formed must
be expected, but such variations will have to admit of explanation as
formed by one water level, as under this head of uniform uplift no
differential elevation is understood.
Any of the sequential coastal or shore forms, which will be discussed
in Part II., may be found at the level of the former shoreline, and the
stages to which these several forms had advanced in the previous cycle
should now be found consistently related to each other and to the old-
* First Ann. Rep. Geol. Sur. Texas, 1889, 5-lOL
164
PROCEEDINGS OP THE AMERICAN ACADEMY,
laud. The cliff, rock bench and terrace, beach, bar, etc., may all be
found elevated above the sea in this initial stage of the new cycle. A
good example of a recently elevated shoreline is not known to the writer.
Figure 4, giving a portion of San Clements island, California (C. S.,
607, 671, now 5100, 5127), shows several elevated former shorelines,
the last formed being nearly in their initial stage. For detailed descrip-
tion of these cliffs and terraces, resulting from periods of comparative
quiet in a series of progressive uplifts, consult the account of the island
by Professor Lawson.*
FiGDRE 4. Elevated Former Shorelines on San Clemente Island, California.
One of the features characteristic of progressive uplift is clearly shown
upon this map, namely, the more advanced stages of stream development
farther and farther inland from the shore. The valleys widen as one
ascends from the western shoreline, and are shallower on the lower ter-
races, as is seen on the map. The streams have had more time in which
to dissect the higher terraces.
The sequential forms, developed at each level on this island, indicate
extreme youth at the time of each uplift ; therefore the coastal plain
* Bull. Dept. Geol., Univ. of Cal., 1893, I. 128-133.
GULLIVER. — SHORELINE TOPOGRAPHY. 165
and offshore deposits, characteristic of an area uplifted after mature de-
velopment, which was the condition assumed in the discussion of the
ideal area, are not here found. The gently sloping terraces have but
scanty covering of" waterworn pebbles.*
(5) Dissected Oldland. — Upon either hand, as one stood at such a
raised beach in the ideal area under consideration just after it had been
elevated, strongly contrasting regions would present themselves. Below,
the faintly seaward-sloping plain ; while above would be seen the dis-
sected oldland. No general criteria for all regions can be given, for the
aspect which a given country at this time will present depends entirely
upon what stage of development was arrested by its change of position
with respect to baselevel. Young, mature, composite, or forms of almost
any other possible stage, may be found. The question for the observer
to ask is, Where in its path of life did this country stand ?
Variations from Ideal Scheme. — Many variations from this ideal scheme
will at once sufrgest themselves. The land may have been depressed but
a short time before the uniform uplift occurred, and then the bottom
would not have been smoothed over. The coast may have been so steep
that all the waste from the cliff cutting was dumped immediately offshore
from the rock bench, and only a narrow terrace was formed in continua-
tion of the bench. This is practically the case in San Clemente and in
the raised beaches of Scandinavia, already referred to (pages 159, 160),
where no broad coastal plain, simple new shoreline, nor smooth bottom
is found.
Variations in structure will cause great differences in the coastal and
shore forms. A mountain region with its structure transverse to the
shoreline, as is the case in Brittany, will show, after an uplift following
adolescent dissection, a much more irregular elevated shoreline than in
the ideal case of homogeneous structure considered above. A region of
longitudinal mountain structure, like the Austrian coast (page 1G8),
would show its characteristic features of development in its elevated
shoreline.
Time since the last considerable movement is however the most im-
portant factor to be considered in regard to variations from the ideal
scheme. If the previous cycle had advanced only to youth, the coastal
and shore forms, seen after the uplift in the elevated shoreline, would have
the characteristic forms of youthful development. In this case it would
be easy to tell whether the second cycle previous to the present was one
* Loc. cit., p. 132.
166 PROCEEDINGS OP THE AMERICAN ACADEMY,
following elevation or depression, for as will be shown later the forms in
the various stages are quite different, as far at least as into maturity.
After maturity is reached in the development of forms of the coast and
shore, the distinction between a cycle following uplift and one following
depression is not so marked.
Slow and rapid Movemsnt. — The initial criteria for the ideal case have
heen given as if the land were raised at once to a certain height and then
stopped, and as if its form were exactly as it had been when developed
at a lower level. Such a conception is of course admissible in an ideal
scheme, but in the consideration of actual examples the sea will gener-
ally be found to have done some work while the movement was in pro-
gress. A series of halts may be made in the upward movement, as has
been shown in San Clemente (Figure 4). Any speed of uplift may be
found in a given locality, and the above criteria must be modified to fit
the case under consideration.
Beffioncd and Continental Uplift. — From the uplift of a limited area
we may extend the conception to a whole continent, but we must be
careful that the criteria are found throughout the whole of the area in
which the uplift is inferred. If a whole continent was uplifted bodily,
the new shoreline, the coastal plain, and the elevated former shoreline
should be found all round its margin, unless some local reason could be
given for the absence of one or more of these criteria in a given locality.
Continental movements have been inferred from local phenomena, par-
ticularly by writers who have discussed the relations between elevation
and glaciation, so that the term as found in the literature is used in a
very loose way.
3. Uniform Depression.
Initial Stage of an Ideal Area. — As in the case of uniform uplift an
ideal case will be first considered in the study of the initial forms follow-
ing uniform depression. Tiie ideal case is taken of a region of homo-
geneous structure, which was developed to early maturity in the previous
cycle, and the depression was sufficient to entirely submerge all the forms
of the coast and shore developed in the previous cycle. The depression
is regarded as having been continuous, though not necessarily rapid.
The sea action upon the land during the slow depression was not suf-
ficient in such a short space of time to materially change the mature
forms of the previous cycle.
(1) Utteven Bottom. — If a region be submerged for a certain amount
beneath the sea, the vertical distance being the same on all sides, the
GULLIVER. — SHORELLNE TOPOGRAPHY. 167
subaerially carved topography would be partly under water. The in-
equality would be proportionate to the relief of the land still exposed,
the change from the more even offshore bottom of the former sea area
to the uneven floor of the submerged area being less abrupt the more
gradual the depression.
Criteria of submarine form have been very loosely used by writers in
the past. In some cases the same facts have been used to prove diamet-
rically opposed theories. Compare the use of inequalities of the bottom
by Dr. Spencer and M. Bertrand, the one to prove subaerial denudation
in the West Indies at a former greater elevation, while the other con-
siders all such irregularities in the English channel as the result of
warping. *
All along the Atlantic shore of the United States, from Maine to
North Carolina, submerged channels have been revealed by the detailed
soundings of the Coast Survey ; and on the Pacific shore Professor
Davidson has shown many channels which are not continuations of present
river systems. IVIany of these are however undoubtedly the result of
warping, and all have been more or less cloaked over with land waste,
so an example surely in an initial stage following uniform depression
cannot be given. An example, which comes as near as any known to
the writer to being still in a very youthful condition since depression, is
in the bay of Maine (C. S., 103, 104, 105, 106), where the soundings
indicate very marked submarine channels, wliich are continuous with
land valleys. A small portion of this area is sliown in Figure 1.
(2) Irregular New Shoreline : Scandinavia. — The intersection of the
sea with the uneven land surface produces an irregular shoreline, pos-
sessing many drowned valleys or rias t and arms of the sea between
headlands and islands. The degree of irregularity depends upon the
strength and variety of relief of the submerged area and on the amount
of submergence.
For any given area, it is probable that there is a certain medium
measure of submergence which will give a maximum irregularity of
shoreline. The slopes above and below the water level will be essen-
tially identical, inasmuch as the shoreline lies at a level independent of
the form of the land. ,
The excessive irregularity of a drowned shoreline is well illustrated
by the coast of Scandinavia. The coast of Maine (Figure 1) is less irreg-
ular, both on account of a less mature dissection before drowning and also
* See references. t See p. 22a
168 PROCEEDINGS OF THE AMERICAN ACADEMY.
because this coast is further removed from its initial stage. Puget sound
(C. S., 6450, 6460) shows an irregular shoreline with many branching
bays, but much more work has been done in this locality since the
drowning to simplify the shoreline. Mr. Willis has shown* that this
Region is complicated by faulting.
(3) Dissected Land comparable with Submerged Topography : Austrian
Coast. In from the coast the land would have for its initial form one
which is appropriate to the stage of the former cycle, which was inter-
rupted by the relative depression of the land with respect to the sea.
The whole region has been supposed to move together, so the streams fit
their valleys ; therefore if it were not for the many streams now pointing
into the same bay, " betrunked " and entering the sea independently, it
could not be told from their individual action in the present cycle that
their work had been diminished by the submergence.
The type example of drowned longitudinal topography, now in an
exceedingly early stage of development since the initial submergence,
is the Adriatic coast of Austria (Austr., Zone 24, col. IX, X, XI;
25, IX, X, XI, XII; 26, IX, X, XI, XII; 27, X, XI, XII; 28,
XI, XII, XIII; 29, XI, XII, XIII; 30, XII, XIII, XIV; 31,
XIII, XIV).
The cliffs on the more exposed land are older than where better
sheltered. It is a region of Mesozoic and Eocene strata of the Jura or
Appalachian type of folding, maturely dissected when drowned, into
whose longitudinal valleys tlie sea has entered, forming characteristic
drowned valleys of the longitudinal type, t In many places the slopes
intersect the sea level without a trace of having been attacked by the sea
since the depression. Following these slopes under water we sometimes
find them continuous with the unsubmerged portion, wliile in other places
the soundings indicate a rapid change from steep to gentle grades. A
detailed geological map with sections showing the structure of the Juras-
sic, Cretaceous, and other strata is needed to show whether these rapid
changes of slope under water are due to structure, to baselevelling, or to
aggradation during a slow depression. The central portions of the sounds
and channels bordered by the inner shoreline have broad flat areas
ranging in depth from 70 to «)5 meters.
The general accordance of level of these bottoms suggests as the most
* Chi. Jour. Geol., 1897, V. 99.
t See various articles in Austrian journals by the following geologists: Bittner,
A.; Hauer, Franz Hitter v.; Hilber, V. ; Petermann, A.; Stache, Guide; Tietze,
Emil; and Toula, Franz.
GULLIVER. — SHORELINE TOPOGRAPHY. 169
probable explanation of their origin that they represent the areas reduced
close to baselevel during tlie previous cycle, when the land mass stood
higher and was dissected to maturity. Such lowlands would be slio-htlv
cloaked over during a gradual submergence. Such slow depression is indi-
cated by the bays almost filled by deltas with no bay-bars at their mouths.
A period of gradual sinking, slow enough to allow delta growth to fill the
valleys as they went under water, and fast enough to prevent much cutting
of cliffs and building of bars, w ould account for the existing combination
of initial shoreline with bays nearly delta tilled.
Infantile islands, minutely irregular shoreline, projecting headland, and
unfilled bays are characteristic of the southern portion of this area. The
depression has no doubt varied slightly in time and amount in different
portions of this region, but as a whole it is a remarkably good examjile
of drowned topography close to its birth.
Other Exaniples. — A few other examples of drowned topograjohy that
have advanced but slightly from their initial stages are here given, with
but a word of comment in the several cases. Special features in these
areas which show an advance from the initial condition are considered
later under the several headings in Part II. All these regions taken
together with Austria, the type of longitudinal drowned topography, give
an idea of the various types of forms resulting from the drowning of sub-
aerially carved topography. In several cases the depression may not
have been absolutely uniform.
Tlie beautiful Cliristiania river system devclopcrl to adolescence before drowning
(Nor., P, A, B, C, D ; 10, A, B, C, D ; 14, B, I) ; 15, A, C ; 19, B, D ; 20, A).
The meandering valley form of Kolding fjord argues strongly for submer-
gence of subaerially carved topography (Uenm., Fredericia, Bogense, Skamllngs
Banke).
The meandering valley above Haderslebener lake is continued in Haderslebencr
fjord with swings of proportional radius of curvature (Germ., 7, 12, 13),
The drowned valley of tlie Warnow river below Rostock is about the same size
as that above the city (Germ., 8G).
Greece and the coasts of the iEgean sea (Atlas Univ., 40; Attica; maps in Der
Peloponnes). Dr. Philippson has shown in his monograph on tlie Peloponnesus
that tliis region is dissected into many blocks by diastrophism.* This causes rocks
of differing resistance to be near one another; thus on this account, and also
because of the stronger sea action in certain places, one finds adolescent develop-
ment replacing the more common youthful forms upon the coasts to the north.
Clarence strait, Revillagigedo channel, and Portland canal, Alaska, show the
typical ramifications of subaerially carved topography (C. S., 8100, 700).
* Der Peloponnes, 418-432.
170 PROCEEDINGS OP THE AMERICAN ACADEMY.
Variations. — As in the case of uniform uplift (p. 165) there will be
great variations from this ideal scheme of criteria for uniform depression.
The stage of development interrupted by the drowning, the steepness
and structure of the coast, and the rate of submergence, all have impor-
tant bearing upon the form of the depressed coastal and shore forms.
Slow sinking while the sea cuts into the land will materially aid the for-
mation of a planation surface. Professor von Richthofen goes so far as
to consider all regional plains of abrasion, " Abrasiontlachen/' as neces-
sarily the work of the sea aided by slow submergence.*
The gradual depression and cloaking over of a region are the normal
results of the isostatic return to a condition of equilibrium. Stripping
in one area and loading in another causes a lack of balance, which will
be restored by a rising of the stripped, and a sinking of the loaded area.
One of the best examples of isostasy is seen in the Mississippi basin. f
Now while the principle of isostasy explains some regions of slow
depression with concomitant sedimentation, it does not account for the
more pronounced changes of level, introduced by secular elevation or
depression. Geographic cycles are not introduced by isostatic move-
ments. The suggestions of cause are numerous, but these geological
questions are not considered in this paper. The subject is here dismissed
with the statement, made by Major Button, that " the nature of the process
is, at present, a complete mystery." %
4. Diverse Movements.
Tilting ; Position of Pirotal Axis. — Uniform uplift and depression
have been considered, and the resulting initial forms contrasted in the
two cases. If, instead of a uniform uplift throughout the area, the move-
ment is diverse, we have tilting, warping, or crumpling and faulting. If
the change of quantity proceeds at a constant rate, we have rigid tilting;
if at a variable rate, but of moderate variety, we have warping ; while if
much irregularity of rate appears, we have disorderly crumpling or
faulting.
With the exception that the topographic forms are elevated or de-
prissed to different amounts in various places, the criteria of tilting are
the same as those already discussed. Tilting may be of such a character
* Fiibrer fiir Forschungsreisende, 1886, 354.
t See the following artides : McGee, A. J. of S., 1802, XLIV. 177-192; Bull.
G. S. .\., 1804, VI. 5.5-70 ; Keves. Bull. G. S. A., 1894, V. 231-242.
t Pliil. Soc. Wash., 1889, XI. 03, G4.
GULLIVER. — SHORELINE TOPOGRAPHY. 171
as to give criteria of uplift in one portion o£ the tilted region and those
of depression in another.
The former shoreline in a tilted region, unless the axis of tilting was
parallel to the general direction of the coast, would not be level, as it
was found to be in a region uniformly uplifted. It will be progressively
higher away from the axis on the side of elevation, and will be more
irregular in height the more sinuous the shoreline before the tilting took
place. The raised beaches around lake Ontario, taking Dr. Spencer's
elevations of the Iroquois beach, show a very nearly even tilt.
The position of the pivotal axis, as pointed out by Professor Shaler,*
gives differing results, and thus the criteria differ for the several cases.
The pivotal axis may lie parallel to the coast, at right angles to it, or
in any intermediate position. This axis may be at the shoreline, inland
from the coast, or seaward from the shore. The tilting itself may be of
two kinds; either the seaward slope maybe increased, or diminished.
These various possibilities will cause many variations in the quantity and
quality of the criteria.
Topography of Tilted Regions : California ; New England. — A two-
cycle history of a region, in which an uplift occurs between the first and
the second, causes the development of composite topography. "When,
however, the uplift is not uniform, a new element comes in ; the topo-
graphic forms developed after a tilt are not only composite, but are also
inclined with respect to baselevel. Those forms of land, which were
developed with reference to one spheroidal plane when it coincided with
baselevel, are tilted, so that this spheroidal plane of the first cycle forms
throughout the region a constant angle with the plane of the sea in the
second cycle. The first cycle of course may be in any stage of develop-
ment when the tilt is made, but the recognition of the tilt will be pro-
gressively easier the later the stage reached before tilting.
A peneplain extends north for a hundred miles from about the fortieth
parallel to the great bend of Pit river, California. f Tliis plain is tilted
at an inclination of 100 feet to the mile toward the east, and is canyoned
by streams 300 to 400 feet deep, which have not yet reached grade.
"The caiions in general are deepest to the westward and gradually run
out to the Sacramento river in the newer deposits which fill the valley.
It is evident that since the baselevel was formed, it has been affected by
* Mem. B. Soc. Nat. Hist., 1874, II. 337.
t J. S. Diller, Jour, of Geol., 1894, II. 32-54; 14tli Ann. U. S. G. S., 1892-93,
Pt. II. 429 ; W. Lindgren, Bull. G. S. A., 1893, IV. 257-298.
172 PROCEEDINGS OF THE AMERICAN ACADEMY.
differential elevation in the uplifting of the Coast range and Klamath
mountains, just north of the fortieth parallel, to the extent of over
2,000 feet." *
A slope of small angular value, viz. 0° 8' 5", across the State of Mas-
sachusetts carries the southern New England peneplain to an eleva-
tion of twenty-five hundred feet in a distance of one hundred and sixty
miles. As one stands upon the peneplain in the western part of Massa-
chusetts, he may look to the southeast across an almost even surface of
denudation with liere and there a mouaduock rising above it, a monument
of resistant rock.
Warping: New Brunswick, N.J. — The definition of a warped surface
here adopted is that given in geometry, namely, a surface generated by a
straight line moving so that no two of its consecutive positions shall be
in the same plane. Various cases under warping may occur, the marked
characteristic of them all being the variability of the criteria.
In the depression or uplifting of the Schooley peneplain t there appears
to have been a warp, which causes the portion of the Cretaceous pene-
plain near New Brunswick to be lower than the rest.
Santa Catalina Depression. — Professor Andrew C. Lawson has de-
scribed t a very beautiful instance of differential movement between San
Pedro hill on the mainland and San Cleraente island. Upon the south-
ern California coast and also upon San Clemeute are many well marked
sea-cliffs rising one above another to an elevation of some 1500 feet.§
These show pauses in a progressive series of uplifts. But between San
Clemente island and San Pedro hill lies Santa Catalina island (C. S.,
5100), whose land sculpture shows subsidence and not elevation. Upon
this island (C. S., 5128, old number 613) there is a good example of a
divide almost submerged. Professor Lawson says that the sea-cliffs
show more rapid recession than is usually found in stationary or rising
coasts. He considers this Santa Catalina depression an erogenic, or local
movement, which occurred at the same time or later than the epeiro-
genio or general uplift, shown by many observations along the coast of
California.
Crumpling and Faulting. — Cycles and epicycles caused by uplift or
depression merge through tilting, crumpling, and faulting into those in-
augurated by mountain-building. A graded series of forms may be con-
* Jour, of Geol., 1894, II. 45.
t Messrs. Davis and Wood, Proc. B. Soc. Nat. Hist., 1889, XXIV. 380.
X Bull. Dept. Geol, Univ. of Cal., No. 4, 1893, I. 122-139.
§ See p. 164 and Figure 4.
GULLIVER. — SHORELINE TOPOGRAPHY. 173
ceived, and largely filled in with examples, beginning with the area
uniformly uplifted and ending with a highly complicated mountainous
region. This interesting subject falls outside the province of this
paper.
PART IL SEQUENTIAL FORMS.
5. Sea Attack and Transportation.
Differential Abrasion. — Varying hardness of rock is an important
factor in subaerial degradation, and it must also have considerable to do
with the attack of the sea upon coasts. The two ways of formation of
plains discordant with the rock structure have been contrasted thus : " A
subaerial baselevel plain is gradually completed by the action of ordinary
forces on all parts of its surface," while " a submarine platform is essen-
tially completed strip by strip, once for all, as far as it goes." * Pro-
fessor Shaler has recently called the monadnocks, the residual masses of
harder rock rising above the New England upland, " the most enduring
evidences of marine action." f
Without entering into the discussion whether the New England
monadnocks were formed by subaerial or submarine denudation, it is the
purpose of the writer to use tliese contrasting interpretations of the same
phenomenon as an introduction to the discussion of the effect of relatively
hard and soft rock upon marine denudation. Waves will attack softer
rock more rapidly than its more resistant neighbor. A promontory of
hard rock may thus be formed where the less resistant rock on either
side has been eroded by the sea. The ocean, however, tends to convert
irregular to straight or gently swinging coasts.
If the land therefore remains at the same level there will come a time
when the increased cutting upon the exposed promontory will equal the
lessened wearing of the softer material in the re-entrants on either side.
After such equilibrium is reached the shoreline will march inward, practi-
cally strip by strip. If, on the other hand, there is a gradual sinking of
the land, decided inequalities of surface due to differential marine erosion
may be covered by the offshore deposits. This has been pointed out
both by Professor Shaler and by Professor Davis in the papers quoted
above.
Monadnocks versus Marine Remnants. — A distinction should be sought
* Messrs. Davis and Wood, Proc. B. Soc. Nat. Hist., 1889, XXIV. 375.
t Bull. G. S. A., 1895, VI. 149.
174 PROCEEDINGS OP THE AMERICAN ACADEMY.
between the remnants above a submarine platform and the monadnocTia
rising above a subaerially carved peneplain. The burying by slow sub-
mergence would tend to protect from decay the sea cliffs and benches, so
that when re-elevated and divested of their sedimentary protection, the
marks of sea action would show marine origin, at least in part. A de-
pressed peneplain with its monadnocks would also show cliffs and benches,
if it remained in its descent at one level for a time sufficient for cutting.
Features of shore development must not then be considered as distin-
guishing between monadnocks and marine remnants.
The vital question is how far the cover extended inland, and what point
the former shoreline reached. Inside this limit all differential erosion
remnants will have been formed entirely by subaerial degradation, while
on the seaward side of the line the sea will have had more or less to do
with their formation. After the form of the old shoreline has disap-
peared, and the coastal plain sediments been more or less completely
stripped off, the evidence for the former greater inland extension of the
cover will lie in the arrangement of the streams. The area formerly
covered will show superposed streams and less perfect adjustment of
rivers to structure than is found beyond the limits of the former
shoreline.*
Coastal Inequalities. — IMany writers have ascribed all inecpialities of
the coast to differential erosion of the sea. Even as late as 1882, Prof.
A. II. Green implies that all bays and other coastal inequalities are due
to " the hardness and structure of the rocks." f The tendency in America
of later years has been to ascribe all inequalities of the shoreline to the
drowning of subaerially carved forms. "While submerged topography will
account for the greater i)art of such irregularities, we must not entirely
leave out of the consideration the action of the sea.
The agents of the sea are the waves. $ tides, and currents. Writers
differ widely in what they attribute to each of these three agents, and a
discriminating study of the work of the three should be made by some
careful observer. The present writer is inclined to attribute the attack
of the sea largely to the waves, and its transporting action largely to the
tides and currents. ,
* See Messrs. Davis and Wood, Proc. B. Soc. Nat. Hist., 1889, XXIV. 399-410;
Professor Davis, Lond. Geog. Jour., 1895, V. 128-138.
t Physical Geology, 577.
} For the method of wave attack see Gilbert, Mon. I., U. S. G. S., Chap. II.,
•with references ; Lyell, Principles of Geology, 11th ed., 1872, 1., Chaps. XX.-XXII. ;
LeConte, Elements of Geology, 2d ed., 1882, 31-43; Penck, Morphologic der Erd-
oberflache, II. 460-497, with references.
GULLIVER. — SHORELINE TOPOGRAPHY. 175
Wave-cut Islands. — On the Marblehead coast of Massachusetts we see the more
rapid erosion of the trap dikes which intersect the more resistant granite.
Sir Charles Lyell has given instances of differential marine erosion in the drongs
of the Shetland islands.* The granite and other harder rocks longer resist the
waves than the schists. The many veins of porpliyry in Hillswick Ness, Lyell
shows, will also in time similarly be etched.
The Orkneys and Shetlands are exposed to violent sea action, and since the
shore evolution is here considerably below grade, this is the place where differential
abrasion might be expected. The best maps of these islands t show many outlying
islets, high stacks, and low skerries, many of which are probably due to abrasion
of the sea since the drowning of this region.
Wave-cut islands are typically seen along the west coast of Ireland. Probable
occurrences are in the following areas (Ireland, 9, 51, 83, 93, 103, 160, 171, 204).
In the Southern rapids of Peril straits, Alaska (C. S., 8259), the sea is now ac-
tively eroding. The current, according to the Coast Survey, is often running ten
knots an hour, and the tides between Pinta head and Eureka ledge run with
terrific velocity. All the conditions are here favorable for the production of
wave-cut islands, and an examination of the charts shows many islands, rocks, and
ledges entirely isolated from each other. Tlie sea has here made no attempt to
simplify the irregular shoreline by connecting bars.
"Approached by sea, the Aleutian islands seem gloom}' and inhospitable. . . .
An angry surf vibrates to and fro amid outstanding pinnacles." J
Off cape Tsciiipnuski, Kamchatka, numerous rocky islets, stacks, and skerries
are seen upon the map, and in the sketch of Lieutenant Rogers (II. O., 54).
At Blanca and Concon points (H. O., 12.32), and at Guacache, Cobija, and Gua-
silla points (H. O., 1181) on the coast of Chile.
Algodonales point, west of Tocopilla, Chile (H. 0., 12G5).
Submarine Platform. — The late old-age of shore development, where
the laud has stood approximately at the same elevation for a period of
time sufficiently long for the sea to have carried out its intention, is the
submarine platform, the plain of marine denudation. This plain will not lie
as far below the surface of the sea as it did in its maturity. The broader
expanse of the submarine platform beneath the ocean will prevent the sea
from so actively attacking the coast. From birth to maturity the sea
pushes its zone of maximum action farther and farther inland, while fiom
maturity to old-age the atmospheric agencies will supply more waste than
the shore currents can take care of, and the offshore depth will gradually
decrease, though the shoreline will move landward at a lessening rate.
The steep cliffs of maturity will diminish in height as old-age cornea
* See figures, Principles of Geology, 11th ed., 510, 511.
t Ivoy. Scot. Geog. Soc, Atlas of Scotland, Edinburgh, 1895, sections XLII,
XLIII, XLIV, XLV.
t W. H. Dall, Sci., 1896, III. 44.
176 PROCEEDINGS OF THE AMERICAN ACADEMY.
on, and at a late stage will show as little elevation as was seen in the
youthful nip.
While the sea has produced the submarine platform, the land has been
worn down by subaerial degradation to a peneplain.* The controlling
plain for the production of the peneplain surface is baselevel, " the level
of the sea . . . below which the dry lands cannot be eroded." f The
surface will never reach baselevel, but will approach it, " in an infinite
series of approximations like the approach of an hyperbola to tangency
with its asymptote." t A possible qualification of the above statement
may be, that where the surface is near baselevel, the wind may excavate
a portion down to or even below sealevel.
American and English Views. % — Major Powell, Major Button, Mr.
Gilbert, and other geologists who worked upon our western interior
region, saw the great importance of sea-level as the controlling baselevel
down toward which the laud is worn. The action of the sea did not
enter into their considerations to any extent. The English geologists on
the otlier hand saw upon tlieir island the great destruction wrought by
the waves, and the lower level of wave action was their most important
plane of reference. Professor Ramsay included the subaerial forces as
aids in marine denudation, while later Dr. Geikie || made sea cutting of
less importance than subaerial denudation in the production of the plain
of marine denudation.
Figure 5. S L = sea level. \VB = wave-base. P = peneplain.
S P = submarine platform. CD— continental delta.
Wave- Base. — The twb planes of control should be distinguished, and
the almost plains produced by subaerial and submarine degradntion be
given separate names. Figure 5 shows the relation of the peneplain
surface with its controlling baselevel to the submarine platform and its
* W. M. Davis, A. J. of S., 1889, XXXVII. 430.
t J. W. Powell, Exploration Colorado River of the West, 1875, 203.
t C. E. Dutton, Tertiary History of the Grand Canon District, Mon. II.,
U. S. G. S., 1882, 76.
§ Since the following section was written, Professor Davis has made a more
extensive comparison of the American and English schools; Bull. G. S. A., 1896,
VII. 377-398.
II Scenery of Scotland, 1887, 137.
GULLIVER. — SHORELINE TOPOGRAPHY. 177
controlling wave-base. The term wave-base is here introduced as a com-
parable term to river baselevel or hard stratum baselevel. It is another
local baselevel, which ought to be distinguished from the grand baselevel
of the sea.
Thus, at a late stage of development, the peneplain and the submarine
platform almost merge into each other; indeed, so murh do the forms
resemble each other, that the one process or the other has been given by
many writers as explaining the total degradation toward a plain. The
plain of marine denudation, perhaps better called the submarine platform,
is distinguished from the peneplain by its cover of offshore deposits, and
the limits of this cover, even after it is partly stripped off, can be found
from the arrangement of the drainage.
The need for a separate term for the controlling plane from that of the
surface, down to which the forces of degradation are able to reduce the
land, is shown when o'ne examines recent writings upon this subject. To
speak of the deformation of the baselevel* is like saying a bent cone in
conic sections. Both terms imply abstract mathematical surfaces that
cannot suffer distortion. The peneplain may be uplifted, tilted, warped,
or folded, but not the baselevel. In the same way it is helpful to distin-
guish between the submarine platform and the wave-base. The offshore
erosion surface will approach the depth to which the maximum wave
action is possible, but the submarine platform will be cut to that depth
only in the zone of maximum wave activity.
Sea Transportation. — When the sujiply of waste has increased beyond
the power of the various currents to immediately deposit it offshore,
transportation alongshore will become more important, and aggradation
may take place in certain places. The tendency of shore currents is
undoubtedly to form curves in the shoreline wliich will be satisfactory
to the particular current acting.
The writer makes the following distinction between the sea action
upon the inner shoreline, which includes the more protected coasts of
bays, drowned valleys, sounds, channels, etc., and its action upon the
outer t shoreline, which is that of the exposed coasts of the ocean. The
ocean currents have little direct effect upon the inner shoreline, and the
wind has not opportunity to develop, by the formation of waves, current
eddies of large radius of curvature upon inland waters. In these narrow
arms of the sea the tidal currents are the preponderating force, for here
* Diller, 14th Ann. Rep. U. S. G. S., 1892-93, Part II., 406; Jour, of Geol., 1894,
IL 45.
t See Penck, loc. cit., II. 551.
VOL. XXXIV. — 12
178
PROCEEDINGS OP THE AMERICAN ACADEMY,
the ocean current and the local wind current do not have a chance to be
relatively so effective. It may be stated as a general principle that the
most effective agent of shore development upon the inner shoreline of
drowned topography is the tidal current. Broad bays form a middle
ground where any of the three forces may be the strongest. Upon the
outer shoreline the ocean eddy currents are the most effective, while
upon lakes and inland tideless seas the local wind currents are the most
important factor. The movement of the land waste is in all three cases
due largely to the action of the waves.
Offset; Orerlap ; Stream Deflection, Figures 6, 7,8. — The three
criteria of form by which the dominant current alongshore may be in-
ferred are offset, overlap, and stream deflection. The three usually
occur together, but each is found alone.
Figure 6. Offsets. Figure 7. Overlaps.
V
Figure 8. Stream
deflection.
Types of offset without accompanying overlap are given in Figure 6.
Overlaps are commonly accompanied by offsets of the shore curves in the
same direction, as is markedly the case in Fire Island inlet, Long island
(C. S., 119). One shore curve offsets another when the curve itself or
the continuation of the same passes to seaward of the next succeeding shore
curve. "When this offset is slight, it may be perceived by looking along
the shore curve, putting the eye close to the map.
The typical example of offset without overlap is on the west coast of
Jutland (Denm., Thisted), where the currents are known to be from the
GULLIVER. SHORELINE TOPOGRAPHY. 179
south, which is in this case the right.* The right shore curve syste-
matically offsets the left along all the western coast of Denmark.
Many examples of similar offsets are known along the coasts of the
world, and wherever the dominant current is known from observation
the offsets follow this law : The current flows from the outer
CURVE TOWARD THE INNER ONE. On account of the number of cases
ill which the offsets agree with the observed currents, it is pretty safe to
conclude when offsets occur systematically in one direction that the domi-
nant movement alongshore is in all probability from the curves which
offset toward tliose which are offset.
Figure 7 shows typical overlaps. The right hand curve of the outer
shoreline la|)3 over the next succeeding curve of the outer shoreline.
A curve which overlaps the succeeding one generally offsets it as well,
though in places, as is shown in the lowest example in Figure 7, the
up-current curve may intersect the down-current one if extended far
enough. This occurs where the factors of alongshore transportation are
probably changing, and the down-current curve is really made up of two
curves, and the up-current curve offsets the down-current one in each
case.
The overlap is an intermediate form between the offset and the
deflected stream. A graded series of examples might be given from
simple offset through various combinations of overlap to a case of stream
deflection without any offset.
Along coasts which are formed of unconsolidated materials, it is fre-
quently observed that rivers, brooks, or tidal channels aim toward the
sea for a certain distance and then turn and run along nearly parallel to
the shoreline, and finally empty to the right or the left of the point which
would have been their direct course to the sea. The river's intention to
reach the sea as quickly as possible is evidently not carried out where
such deflection is seen. Some disturbing force has come in. There
seems little doubt that this force is the current alongshore, which has
turned the outlet of the stream. Such has been the explanation of
many authors, f Figure 8 shows the relation of current to deflection
of streams.
Dominant Current. — There is probably wave movement in both direc-
tions along the shore at different times, and the form shows in which
* H. Mohn, The North Ocean, Norwegian North Atlantic Expedition, 1876-78,
2, XVIII. 168, Plate XLIII.
t De la Beche, Geological Notes, 1830, II. 11, Plate I. Fig. 3; Reclus, La Terre,
1870, I. 447 ; Sir A. Geikie, Textbook, 3d ed., 399.
180
PROCEEDINGS OP THE AMERICAN ACADEMY.
direction the dominant movement has taken place. The dominant move-
ment may not always correspond to the prevailing movement alongshore.
A few severe storms causing a strong
current from the right during one
month might determine forms, which
a weak current from the left prevail-
ing for eleven months of the year
would not be able to efface.
Current Cuspate Forelands: Type,
Figure 9.* — In adolescence, w'hen the
currents have more load than they can
carry, it is deposited in forelands of
various forms. A characteristic one
is the cuspate, of which a typical draw-
ing is given. In it are combined those
features of the three Carolina capes f
and cape Canaveral (Figure 10) which
the author deems important to show
the method of growth. Former posi-
tions of the shorelines are indicated
by the ridges of dunes built by the
wind along the shore.
Such former positions are beautifully
Figure 9. Typical Current Cus- indicated in Canaveral (C. S., 160, 161),
pate Foreland. wiiere three or four successive positions of
tlie outline of the cusp, each farther to tlie
left than the preceding, are delineated, besides many lines of aggradation in each
position (Fig. 10). Simihir lines of grovvtli are seen at cape Fear, where the present
right slioreline cuts off tlie eastern ends of the four dune ridges extending cast-
southeast from tlie lighthouse and curving sympathetically with the left shoreline.
Cape San Bias, on the west coast of Florida (C. S., 183, 184), shows four stages
on the right side and nine successive stages of aggradation on the left side.
A more striking example of aggradation lines is seen in the cusp of Dars cape
in the Baltic (Germ., 61, 62, 63), where thirty-eight systematic and successive
shorelines are indicated by dune ridges (Fig. 11). The dominant current is from
the right, according to the offsets and hook at the point of the cusp; but the
thirty-eight successive shorelines suggest a gradual aggradation of strips, and a
change from an earlier condition when the current was from the left. The tidal
flats, east of Zingst, point to present transportation and growth toward the left.
* For fuller account see Bull. G. S. A., 1896, VII. 399-411 ; also see references
for papers by Abbe and Tarr.
t See p. 242.
1 I I I I l_
/o
I Kilometers.
Figure 10. Canaveral Foreland, Florida. (Sheet 161, U. S. C. G. S.)
182
PROCEEDINGS OF THE AMERICAN ACADEMY.
The topography and geolog}' both imply that this offshore bar is built up of several
islands tied together. A very pretty problem for field study is here presented.
A rounded cusp projects into the Baltic north of Wismar bay (Germ., 85).
Markelsdorfer Huk on the northern end of Felimarn island is a foreland of ap-
parently this same general type of formation (Germ., 40).
/^:^^>'
Figure 11. Dars Foreland, Germany.
Jaederens point on the Norwegian coast (Nor., 6, B) probably owes its projection
in part to current aggradation.
The point del Faro on tlie northeast of Sicily shows action of tidal currents
combined with the large eddy of tlie T^yrrhene sea. The cusp does not grow at
right angles to tlie direction of the currents through the straits of Messina, but lies
between the Messina current and the Tyrrhene current (Ital. and Sicily, 254).
North of Sousse there are tAvo solid cuspate forelands, covered with dunes,
which are apparently built by a progressive series of additions to the coast
(Tunis, 57).
GULLIVER. SHORELINE TOPOGRAPHY. 183
In small water bodies, lakes and seas nearly without tides, the winds would
cause waves which in turn would originate currents of smaller radius of curvature,
which should produce smaller cuspate forelands. The cuspate points in the Danish
waters are probabl}' such forelands. Tliese are seen on the topographic maps of
Denmark in the following localities : Roskilde fjord (Denni., Hilderod), Suen Mel-
lem Smalandene (Denm., Saxkjobing, Vordingborg), Limfjorden (Denm., Logstiir),
on Langeland and the islands to the west (Denm., Svendborg, Nakskav, Gulstav,
Faaborg), and in other localities along the Danish and German coasts.
The Bonneville cuspate forelands are proportional to the currents which existed
on the old lake and are similar in size and outline to the Danish cusps. Professor
Russell also reports V-bars upon the fossil shores of lake Lahontan.* These
cusps seem to have been built upward as the waters of the lakes rose, but the
water level never remained constant long enough for the lagoons to have become
filled, forming solid forelands, since Mr. Gilbert reports only a partial silting up.t
6. Offshore Bar.
Shelving Shore. — When the sea takes a new position of attack after
elevation, if the shore is shelving, wave-base intersects the smooth bottom
at some distance from the simple new shoreline, and the point of maxi-
mum wave abrasion is out from the coast at some point on the shelving
shore. This condition would obtain in the ideal case assumed in Part I.
From the forms of observed shores which slope gently beneath the water,
the action of the sea appears to be somewhat as follows. The waves at
first beat upon the coast and cut a faint cliff or nip. The point of maxi-
mum wave action is offshore, and there the waves heap up sand from the
bottom and a bar is formed alongshore. The waves abrade rapidly until
the offshore bottom to seaward of the bar approaches wave-base. Dur-
ing this deepening the waves have broken farther and farther offshore,
so that the bar has gradually moved seaward. When now the bottom to
seaward of the bar has been abraded almost to wave-base, a condition of
shore-grade is reached : the sea is able to transport and build into the con-
tinental delta whatever waste is supplied from the bottom and offshore
bar. As soon as material is taken from the bar it will retreat toward
the land.
Stages. — The period of upbuilding and seaward growth of the offshore
bar has been regarded as the youth of the shoreline, and the period of
cutting back as adolescence, since the latter is a graded condition. Dur-
ing youth the seaward growth of the bar leaves long marshy strips, or
'' slashes," between the successive dune ridges formed along the shoreline.
These become overgrown with bushes, peat, etc. The lagoon behind the
* Mon. XI., U. S. G. S., 93.
t Lake Bonneville, 121, PI. XVIII.
184 PROCEEDINGS OF THE AMERICAN ACADEMY.
bar also is frequently converted into marsh. In the landward retreat of
the shoreline, this vegetable layer is discovered at or beneath sealevel,
covered by the beach sands, as on the New Jersey coast.
When the offshore bar has been completely cut back, the nip has been
extinguished, and the sea is actively cutting into the coastal plain, leaving
a more or less pronounced sea cliff, maturity is reached.
No Offshore Bar. — When the initial slope of the coastal plain is so
steep that the sea is able to begin the production of the submarine plat-
form immediately offshore, shore-grade is quickly attained, youth and
adolescence are of short duration, and the coast reaches a mature stage
of development without the production of an offshore bar. This has
probably been the case in eastern Italy (page 18G).
Youth : Texas. — Tlie offsliore bars on the Texas coast are very marked features
in tliat elevated region (C. S., 205, 206, 207, 208, 209, 210, 211, 212). An apparent
earlier position of the bar is shown by the string of small islands inside the present
offshore bar. The sea is here building apparently from the bottom in great meas-
ure. The transportation alongshore as indicated by many offsets, Cavallo pass,
Galveston entrance, etc., is dominantly from the left, caused doubtless by the eddy
circulation in the gulf of Mexico. There are occasional stream deflections to the
left, as Cedar bayou and San Bernard river, which are caused possibly by backset
eddies from the main circulation. The littoral forms in this region are complicated
by an episode of slight drowning. A great variety of dune forms are shown
on this bar.
The map of Costa Rica by Dr. Frantzius* shows a characteristic offshore bar.
The scale of the map is too small to show indications in which direction the bar is
moving, so this may be an adolescent coast.
The eastern coast of Corsica (Fr., 2G1, 2G3, 265) shows an offshore bar, but
whetlier advancing or retreating, the writer does not know.
Off the (Jgunquit- Wells Beach coast there is an offshore bar upon which the
writer could find no evidence as to which way it is moving.
The offshore bars on the Atlantic slope are furtlier advanced on the whole than
the Texan bars. Youtliful bars prevail in Texas, and adolescent ones from North
Carolina to Long island. Field study of these bars is needed to bring out more
fully the history of the sequential forms. The following quotation shows tlie
meagre character of existing descriptions.
The offshore bar opposite Beaufort harbor, N. C, "is mostly covered with a low
pine and mixed growth, and its average width is about half a mile ; the sand hills
and ridges upon it are from 20 to 35 or 40 feet high." f
Adolescence : Southern New Jersei/. — The Geological Survey of New Jersey re-
ports that the sand dunes overlie a layer of black soil along the shoreline, at
differing heights at different localities. The lagoon along the southern coast of
New Jersey is largely converted into marsh, while that along the central portion of
• * Pet. Geog. Mitt, 1869, XV. 81, Tom. V.
t H. L. Whiting, U. S. C. G. S., 1851, Appen. 28, 483.
GULLIVER. — SHORELINE TOPOGRAPHY. 185
the State is little filled, indicating an earlier stage. In tlie northern portion of the
State the offshore bar merges into the southern wing of the Long Branch behead-
land (page 213).
Upon the south side of Long island the sea has devoured the land to an appre-
ciable extent during the historical period. Meadows and cultivated lands have
been covered with sand, wagon tracks in peat have been found on the ocean side
of the dunes, wliile peat, cedar stumps, and tangled roots occur to-day between the
sand hills and the sea. These traces of land life seaward of the dunes indicate a
march of the dunes landward,* and a general pushing of the offshore bar inland.
7. Dissected Coastal Plain.
Surface Form. — On page 155 it was shown that the stage of develop-
ment of the surface of a coastal plain may not be the same as that of
the coastline of the same region. This subject comes more properly
under the cycles of development of land forms ; but, since the coastal
plain is one of the main criteria of uplift, the sequential forms will be
briefly sketched.
Mr. W. Lindgren shows a characteristic section of a Quaternary coastal
plain lying on a granite oldland,t but he does not use its stage of dis-
section to show the time since the elevation of the region around San
Diego.
Youthful Dissection : Ogunquit, Maine. — In southern Maine the forms indicate that
there has been a recent episode of uplift revealing a narrow coastal plain, which
fills in the irregularities of the coast made by a previous depression. The streams
have only begun to intrench themselves upon this late deposit.
The Monopoli coastal plain on the " heel " of the Italian boot shows youthful
dissection of a marine plain (Ital., 190, 191). The Pliocene strata t present a
surface gently rising from the sealevel to heights of from 100 to 200 meters at tlie
foot of an abrupt slope of Jurassic and Cretaceous rock. This slope rises from 75
to 250 meters above the plain, and has the form of an elevated former sea cliff now
slightly dissected. A problem for field study is the cause of the minutely ragged
outline of the present shoreline. § There is no ofTshore bar shown with this coastal
plain, which may be accounted for by the fact that the slope of the surface of the
coastal plain is considerable, 100 meters in 5 kilometers, and therefore it is probably
steep enough for the direct attack of the sea. The coastal plain character of the
heel of Italy is well shown on the topographic sheets by the radial arrangement of
roads (Ital., 202, 203, 204, 213, 214, 215, 223). The towns are like the hubs of
wheels, the spokes of which are the highways. The distribution of infaces,
streams, and outcrops suggests that the area has been developed in several cycles,
a study of which in the field would be most attractive.
* A. G. Pendleton, U. S. C. G. S., 18-50, Appen. 8, 80, 81.
t Proc. Cal. Acad. Sci., 1888, L., PI. III.
J Carta Geologica d' Italia, 1 : 1,000,000, Roma, 1889.
§ See page 239.
186 PROCEEDINGS OF THE AMERICAN ACADEMY.
The Pyrgos coastal plain on the western end of the Peloponnesus shows char-
acteristic intrenching consequent streams.*
The Sykonia coastal plain, south of the gulf of Corinth, is much complica-
ted by faults, t The streams are frequently lost in crossing the gravel of the
youngest step.
Adolescent dissection: Eastern Italij. — Eastern Italy, nortli of the " spur," from
Pesaro to Termoli, is a coastal plain of Pliocene strata \ dissected by consequent
streams now aggrading (Ital., 140, 111, 147, 148, 155). There is indication of cap-
tures, particularly in tlie Pescara, Saline, Vomano, and Sangro rivers, but study
upon the ground is needed for proof. Several streams siiow cutting of the right
bank more than the left, Biferno, Fortore, Sangro, Pescara, and Tavo.
Many portions of the Atlantic and Gulf plains of tlie United States, and of the
North German plain show characteristic adolescent dissection.
Mature Dissection: Eastern Virginia. — The form of the surface of the dissected
Neocene strata east of Riclimond, Virginia, indicates mature dissection. Tlie
slight drowning of the streams indicates that since dissection there has been an
episode of depression.
A portion of the coastal plain of southern Sicily where there is the least defor-
mation shows quite typical mature dissection (Ital. and Sicily, 272).
Adjustment of Drainage. — A characteristic feature of maturity is the
adjustment of streams according to the structure of the region. The
most perfect mature adjustment will result from (1) considerable diver-
sity in the size of the initial consequent streams ; (2) considerable alti-
tude of the land-mass ; (3) considerable diversity of resistance in the
strata that are cut through by the streams; aud (4) a significant amount
of inclination in the strata. § Two successive cycles of uplift will give
more complete adjustment than a single cycle.
8. Fading Elevated Shoreline.
Lake Shorelines. — Although some of the finest known examples of
initial elevated former shorelines occur upon shores abandoned by lake
waters, nevertheless these forms as seen to-day have entered their se-
quential stages and are fading away. This fact has not been forced upon
the reader's attention in the articles upon lake shorelines, and he is left
to infer that an elevated former shoreline remains as it was left by the
retreating water. Of course the shoreline of a lake, whose water has
* Dr. A. Pliiiippson, Der Peloponnes ; topograpliical and geological charts,
sheet I. section IV. ; text, 321-323.
t Loc. cit., sheet II. 118, 153; also see Der Isthmos von Korinth, Z. d. G. f. E.,
1890, XXV. 1-98.
\ Carta Geologica d' Italia, 1 : 1,000,000, Koma, 1889.
§ W. M. Davis, Lond. Geog. Jour., 1895, V. 133, 134.
GULLIVER, — SHORELINE TOPOGRAPHY. 187
abandoned its former stand on account of the removal of its barrier or
dowu-cuttiiig of its outlet, is in the topographic sense as truly an elevated
former shoreline as if the land had been raised. The relative position
of land and water is changed.
Typical forms of Bonneville. — Many of the illustrations of shore forms
of the Bonneville, Provo, and levels intermediate iu position between
lake Bonneville and Great Salt lake serve as types of elevated former
shorelines, in youthful stages. The deltas, terraces, embankments, clifFs,
V-bars, bay-bars, and the tying of islands to the mainland are all charac-
teristically shown. The stratigraphic and paleontologic proof of the
relative age of the shorelines is brought out by Mr. Gilbert, but the
fading features of the older shorelines are not dwelt upon to show
relative ages.
Lake Agassiz. — The descriptions of the shore forms of the ice-dammed
glacial lake Agassiz are given iu this same manner, as if the forms were
formed once for all and would forever remain as constructed. Gen. G.
K. Warren set aside the hypothesis of an ice-barrier and argued for an
actual change of level, depression to the south accompanied by a rise to
the north. Mr. Upham has traced the various beaches formed by the
different water levels and shown them to have been the result of an ice
mass to the north gradually retreating toward Hudson bay. These ele-
vated former shorelines rise from south to north and from west to east,
in the direction of the former ice-fields, the amount of slope varying from
zero to one and one third foot per mile. Since these old shores must
have been horizontal when formed, their present position shows a tilting
since the time of lake Agassiz.
Marine and Lake Terraces. — Early writers used the beach form to show
elevation,* but they often did not distinguish between the seashore forms
and those which had been produced by water above the sealevel. One
of the most fruitful sources of error has been in regarding the terraces
of ice-dammed lakes as produced by marine action. The classical exam-
ple is that of the Lochaber terraces, the Parallel Roads of Glen Roy.
For an historical discussion of the change of view from tiie detrital
dammed lake to the action of the sea and finally to the present hypoth-
esis of an ice-dammed lake, see " The Great Ice Age," by Professor
Geikie. t The geographic criteria for the differentiation of the similar
forms produced by these two processes are these. At the level of the
* R. Chambers, 1847, and many later writers.
t 3d ed., 1895, 282-285.
188 PROCEEDINGS OP THE AMERICAN ACADEMY.
supposed ice-dammed lake the terraces will be approximately continuous
except wliere tlte ice stood. On the drift barrier hypothesis it was very
difficult to explain how the terraces remained while the barrier was
removed, but the ice-barrier would disappear by simple melting, and
therefore the terraces would remain in nearly their initial form after the
retreat of the ice and removal of the water. The upper terrace level
would correspond in elevation with the height of the col over which the
lake discharged. This relation to the col has been worked out in con-
siderable detail in the study of the Great Lakes.* A reliable geologic
criterion is the occurrence of marine shells, which are found in marine
beaches but not in ice-dammed lake terraces.
Examples. Raised beaches are found in Ireland on the north coast, in Killary
harbor, along Kenraare and Glengarriff bays, and elsewhere, according to Mr.
HuU.t
For the raised beaches of Great Britain reference will be made to the papers by
the following authors : Ansted, Cliambers, De la Beche, A. Geikie, J. Geikie, God-
win-Austin, Prestwick, Richardson, Trevelyan.
Tlie literature on Scandinavian X raised beaches is extensive, and there are manj'
fine examples of fading elevated shorelines upon that coast. The features do not
show distinctly enough upon the topographic maps for purposes of illustration.
Old beaches at various levels above the water of Pechora bay in the Great
Tundra region of northern Russia appear to be former shorelines. § Mr. Jackson
does not mention less perfect terrace forms the further he went from the present
shoreline, but he proved the progressive stages of uplift by the less perfect preser-
vation on the more elevated beaclies of the pine tree trunks, which he considers as
brought down by the Pechora river.
There is a former shoreline near Ogunquit, Maine, and also farther to the north-
east, upon which little work has been done since uplift.
There is an 800 foot cliff six miles east of San Roque point. Lower California,
which should be examined in tiie field to determine whether it is a former sea cliflF
or not (H. O., 1208).
Another case suggestive of uplift is seen in Santa Rosalia bay. Lower California
(H. 0., 1100, 1193). The lack of more accurate information about this region
makes it impossible to use it as surely showing uplift.
Cliffs 75 feet high are seen along the Sonora coast, Mexico, near the mouth of
the Colorado river. They are so distinct as to indicate a recent elevation (H.
O., 800).
* See papers by Fairchild, Gilbert, Lawson, Leverett, Newberry, Schott, Spen-
cer, Taylor, Upham, and Warren.
t Physical Geology and Geography of Ireland, 1878, 107 ; see also paper by
Kinahan.
X See pp. 158-160.
§ F. G. Jackson, The Great Frozen Land, Macmillan & Co., 1895, 129, 262.
Map.
GULLIVER. — SHORELINE TOPOGRAPHY. 189
Superposed Drainage. — No attempt has been made in this study to
work out the sequential stages in the fading of elevated shorelines. This
problem is intimately connected with the dissection of the land, and de-
pends largely upon the factors which control such dissection in any given
locality. The shore deposits being coarser would probably remain longer
than those finer materials laid further out from the old shore. When
however all the shore deposits themselves are eroded away, the amount
of the former coastal plain overlap may frequently be inferred from the
arrangement of the streams. Where there never had been a cover, the
adjustment of the drainage to the structure would be more perfect than
where the streams had taken consequent courses over an uplifted coastal
plain. The coastal plain sediments would overlie uncomformably what-
ever structures happened to occur in the oflTshorc region of the previous
cycle, and thus the streams in cutting through the cover would have
many chances to become superposed upon unexpected dithculties be-
neath.* The line between the region of well adjusted drainage and the
region in which superposition of streams is found represents a former
shoreline, now elevated and in a late sequential stage.
9. Islands.
Consumption hy the Sea. — As a part of the sea's work to reduce all the
land to a submarine platform just above wave-base, the islands formed by
the depression of a region are some of thirst forms to be demolished.
Very small islets are quickly reduced to skerries and to submarine reefs.
Large islands are more continental in character, and their coasts may
become mature long before the islands themselves are consumed. These
larger islands are not as a rule tied to the mainland by bars. But
islands, which range in size from an area of one third of a square mile
up to some two hundred square miles, are very frequently tied to the
mainland by bars in the process of their demolition by the sea.
Upon the coast of Italy where island-tying in its various stages is beau-
tifully shown, such a bar is called a tombolo.f For convenience in
distinguishing island-tying bars from those of other kinds, the writer
proposes to call every bar of this kind a tombolo, giving an English
plural tombolos.
Loop-bar: Shnpka, Figure 12. — An island at some distance from the
mainland may be so large that the sea cannot dispose of all the detritus
* See fuller statement by Professor Davis, Lond. Geog. Jour., 1895, V. 128-138.
t See Figure 16.
190
PROCEEDINGS OF THE AMERICAN ACADEMY.
woru from the island, and then in youth this waste tails oif on right and
left toward the mainland or in the direction of the quietest water. K
the island is so far from the mainland that these spits cannot reach land,
they are most likely, in swinging back and forth with varying currents,
to join each other and thus form a loop-bar.
Sliapka I.
700 Ft.
Nautical MUe
Figure 12. Loop-bar : Sliapka Island, Alaska.
The form of Sliapka island^^laska (Figure 12), indicates that it had two spits
formed on its lee side from the waste of the eastern cliff face, and tliat these two
have now joined, forming a looped bar enclosing a lagoon (C. S., 8881).
Cup butte, Utah, is an example in the fossil condition.*
San Juan Nepomezino island, Lower California, has a salt lagoon at its southern
end evidently similarly inclosed (H. O., 42).
Cockenoes island has two long stringing bars pointing toward the coast at Nor-
walk, Conn., but the bars have not as yet joined (C. S., 116, 3039).
Endelave island (Denm., Bogense) is being consumed on the east and south sides
and the material is transported around the north and west ends. This is shown by
the hooked spit on the west end and by the five lines of slaslies, or narrow lagoons,
inclosed by successive outgrowing beaches. If this process is continued a little
farther we shall here see another Shapka island with enclosed lagoon.
Flying-bar : Sable Island. — When an island is completely reduced to
a submarine condition, the bar formed from its waste may still remain.
A case like Shapka, when the former island was completely consumed,
would give a flying-bar.
* Gilbert, Lake Bonneville, p. 55, PI. VL
GULLIVER. — SHORELINE TOPOGRAPHY. 191
Sable Island, composed of unconsolidated materials, is rapidly disappearing. *
The map shows an enclosed lagoon, which was formerly nearly twice its present
length. t Its form and structure suggest that it represents a flying-bar, after the
island, from which its materials were derived, had been completely destroyed.
Simple Cases of Island-tying and their Stages. — One of the features of
shore development following depression which shows in most clear and
decisive terms the relative time since depression, is the formation of tom-
bolos connecting islands with each other and with the mainland. When
the sea is able to do the work given it to perform, shore-grade is estab-
lished and littoral transportation occurs along the base of the cliffs, which
are cut on the more exposed portions of the island and mainland, and
deposition begins along the edge of the currents in the comparatively
dead water. Such dead water naturally occurs upon the protected side
of the island between it and the mainland, and a tombolo is begun usually
upon that side. According to the direction of transportation the bar
may grow from the island, from the headland, or from them both. The
essential point to bear in mind is this : the currents will seek to alter the
shoreline better to satisfy their conditions of work.
Numerous examples from various localities are given of the seven
stages into which island-tying has been divided. The lists under this
and other headings of the present article are however not at all exhaus-
tive, enough examples being given in each case to bring out the successive
stages of development and to show the play of the variable elements
within the limits of each stage.
I. Initial Island (Birth) : Austria ; Sweden. — Tlie fii-st stage in the
life of an island is where no work whatever has been done upon it by
the sea. Great variety of form and size will occur, depending largely
upon internal structure and pre-natal development. The longitudinal
structure of Austria, the transverse structure of Casco bay (Figure 1),
and the concentric structure shown on the Vaxholm sheet of Sweden,
give markedly different island forms. The mature dissection of Scan-
dinavia gives many small islands, while the more youthful dissection in
the Puget sound region shows but few islands, and these much larger.
II. Nipped Island (Iniancy) : Sweden; Maine, Figure 1. — The sea
first attacks the coast and makes a nip all around the island, but cuts
more upon the exposed side. The sea at first can dispose of all the
waste from the island.
* Patterson, Trans. Roy. Soc. Can., 1894, XII. (2) 1-50. Map.
t Loc. cit., p. 37.
192
PROCEEDINGS OP THE AMERICAN ACADEMY.
Many islands along the east coast of Sweden (Swe., U, 17, 22, 29, 37, 46, 67, 68,
76, 85, 86, etc.) ; also on the west coast (Swe., 18, 24, 25, 32, etc.).
Maine (C. S., 101, 102, 103, 104, 105, 106).
Numerous exiimples on the coast of Norway (Nor., 5, B ; 48, B ; 49, C ; etc.).
Many islands among the Orkney, Shetland, and Hebrides on the north and west
of Scotland (Scot., 58, 59, 101, 104, etc.).
Off Marseilles (Fr., 247) there are several very young islands.
Lipari islands, Tyrrhene sea (Ital., 244).
Capo Passero island, Sicily (Ital., 277).
Gemini and Corbella, south of Elba (Elba).
Islands west of Fosana, Austria (Austr., 26, IX).
Numerous islands in the Adriatic where ideal possibilities for future tying exist
(Austr., 31, XIII, XIV ; 34, XVII ; etc.).
III. Uncompleted Tomholo (Youth). — AVheii more waste is supplied
than the sea can deposit offshore, transportation alongshore begins, and
there is a tendency to aggrade the surplus load of detritus. Such build-
ing- would naturally be expected to occur in the comparatively quiet
water between the island and the mainland. This is found to have taken
place in many localities. According to the several conditions of the
variable factors in the problem, the tombolo may begin to grow from the
mainland, the island, or from both.
a. Attached to Mainland onli/ : Gigha. — The typical example is seen on the west
;;oa.st of Scotland, where Rhunahaorine point projects as a cuspate foreland from
the mainland toward the island of Gigha (Scot., 20). This foreland is of the nature
of the tidal forelands described on
page 214, and a tombolo may never
be completed across the deep chan-
nel.
A shoal extends from the large
island, Berneray, toward the rocky
stacks, Sgeir a' Ciiail (Scot., 89).
This islet is fast being consumed by
the sea, and probably never will be
tied.
Lingay strand extends below high
tide level toward Lingay island
(Scot., 89).
Callao, the seaport of Lima, Peru,
is built on the tombolo growing toward San Lorenzo island (Stieler, 94; Midden-
dorf. Das Kiistenland von Peru, 1894,36).
Angel island in San Francisco bay, California (C. S., 5581).
b. Attached to Island only : Tuno. — From Tunc") island (Denm., SamsJ)) there pro-
jects toward Sams(3 island a lanceolate cusp, showing the attempt to tie the smaller
to the larger island.
Figure 13. Diagram of Uncompleted
Tombolo.
GULLIVER. — SHORELINE TOPOCxRAPHY. 193
Another cuspate foreland projects from Taransoy island toward Harris island
(Scot., 98).
Vigso island has two curved spits not quite able to reach the mainland (Denni ,
SaxkjiJbing).
c. Attached to Mainland and Island : Aebelo ; Figure 13. — A bar is forming from
Aebelo island and from a smaller island close to the mainland (Denni., Bogense).
Another example is Spectacle island in Boston harbor (C. S., 337), where the
"nose-piece" of the spectacles consists of two cusps almost joined. Upon the
Coast Survey chart these two islands are not joined, but in 1896 the writer saw
from a steamer that the tombolo was completed.
Between Pabbay and Berneray islands a tombolo has begun to grow which con-
sists so far of a cuspate projection from each island (Scot., 89).
The tombolo connecting North rocks with the Irish coast (Ireland, 49, 50) is not
completed, and it is very probable tliat these rocks will be completely consumed
before tying on is accomplished.
The flats between Barra and Fiaray islands represent the attempt of the sea to
tie islands together (Scot., 58, 59, 68, 69). The flats surround three other islands.
Tombolo growth is indicated from both Ibiza and Formentera islands, the ad-
vance from each island being made toward the other (Spain, Bol. VII, Lam. B).
There are several small islands in the line of probable future growth, which will
be surrounded by the completed tombolo.
Marrowstone island, Washington, at present is detached from the peninsula to
the west, Kilisu harbor having communication across the bars at both its north-
ern and southern ends (C. S., 64.50 and 647).
Several of the islands in Sitka harbor, Alaska, are soon geographically to be-
come land-tied ; as, for example. Cannon island, Beardslee islands, and The Twins
(C. S., 725).
Isia de Apics, Mexico (H. O., 878), is now connected at low water.
Rush and Ackerman islands, Costa Rica, are nearly in this stage (11. 0. 1028).
Redonda and Siriba islands, Brazil (H. O., 486).
A tombolo largely of mechanical construction though there is some coral growth
in it, is attempting to connect Ceylon with the mainland.*
IV. Completed Tomholo (Adolescence). — As a rule when islands along
a stretch of coast are completely tied to the mainland by tombolos, the
coast as a whole is graded, and may he regarded as in adolescence. Oc-
casional youthful features will persist after the region has reached adol-
escence, and in the same way completed tombolos will sometimes be found
where the other features of the coast are indicative of youth. There
are three classes of tombolos : single, Y-shaped, and double.
a. Single Tombolo : Nahant, Figure 14. — Nahant is tied to the Massachusetts coast
at Lynn by a single tombolo, which is typical, with the exception that the island
itself is made up of Big and Little Nahant, which are themselves joined by a
tombolo.
* Map by J. Walther, Pet. Geog. Mitt., Erg. 102, 1891.
VOL. XXXIV. — 13
194
PROCEEDINGS OF THE AMERICAN ACADEMY.
Many other cases of single tombolos occur in Boston harbor (C. S., 337 ; G. S.,
Boston Bay, Mass.). Among these may be mentioned Winthrop head, point Siiir-
ley, Peddocks island, Hull, and the islands tied by Nantasket beach.
Little Koniushi island, Alaska (C. !S.,
I)-
Biorka island, Alaska, is made up of
two islets tied together by a bar (C. S.
724).
George island, Alaska, sliows com-
posite building (C'. S., 741).
Aniaknak island, Alaska, has three
component parts (C. S., 8901).
Morro Ingles island, Paz point, and
San Vicente island, Mexico (II. 0.>
640).
Mare island (C. S., 5524, old number
C25) is a case in San Pablo bay where
an island has been tied by a tombolo to
the mainland.
Spider island, Alert harbor, Chile
(II. O., 92G).
Mt. Division, 1880 feet high, is con-
nected with the mainland of Peru by
a low sandy isthmus (II. O., 1178, 1185,
11G2), which is probably a tombolo.
Morro of Barcelona, Venezuela (H.
O., 374).
An island off the Bonneville shoreline
near George's ranch was tied by a tom-
bolo, in which three attempts at tying
are figured by Mr. Gilbert.*
Gilsay island in the sound of Harris
(Scot., 80).
Taransoy is apparently built up of
three islands tied together (Scot., 98).
Ilowth peninsula has the form of an
island tied to the mainland northeast
of Dublin (Ireland, 112). Transporta-
tion is indicated both from the cliffs of
the mainland toward the island and from the island along the tombolo. Broad
tidal flats, cut with runways, occur on the right and left of the tombolo.
Mweenish island is made up of three drumlin-shaped portions connected by
narrow necks, presumably tombolos (Ireland, 115).
Illaunatee or Straw island, one of the Aron islands (Ireland, 113).
The Chesil bank t connects the isle of Portland with the mainland of Dor-
setshire (Eng., 17).
* Mon. I., U. S. G. S., 113, Fig. 23.
t For the literature on this tombolo consult De la Beche, Geol. Notes^ 1830, IL
p. ix ; Geikie, Textbook, 3d ed., 1893, 451.
Kiloineteri
I
Figure 14.
Single Tombolo : Nahant,
Massachusetts.
GULLIVER. — SHORELINE TOPOGRAPHY, 195
The rock of Gibraltar (Q.J. G. S., XXXIV., 1878, VL 23; Brit. Ad., U4, 1448)
was an island, and is now tied.
Sermione, Italy, is on an island in Garda lake connected by a bar three kilo-
meters in length, three times the length of the island (Ital., 48).
Cape Milazzo (Ital. and Sicily, 253).
Penisola Magnisi, and that on which Augusta is built (Ital. and Sicily, 274).
Monte P>nfola (Elba).
Tlie peninsula southwest of Vari, on which Zoster cape is situated (Attica, VIII).
Koronl (Attica, XI).
Probable tying of islets to Samsu island (Denm., Samsu).
Faejo island is composed of two parts tied with a tombolo (Denm., Saxkjobing).
Knudshoved point (Denm., Saxkjobing).
Bogo island has Faro tied to its northwest point by a long tombolo (Denm.,
Vordingborg).
Avernak island (Denm., Faaborg).
Drejo island is composed of two tied by a narrow tombolo (Denm., Svendborg).
Two or three islands were apparently tied together to form the hook north of
Aeriiskjubing (Denm., Svendborg).
b. Y-tombolo : Morro del Puerto Santo,
Figure 15. — The type of the Y-tombolo,
where one bar from the island unites
with two from the mainland, is found in
Puerto Santo bay, Venezuela (H. 0.,
374).
Northeast point on St. Paul island,
Alaska (C. S., 8990, old number 88G) is
connected by a Y-tombolo enclosing a
lagoon.
JIaliedia, Tunis, is figured by Reclus
as tied in this manner.* t^ ir it-. i i ^» i i
,^. , ,. ,,.,x,-r^ liGCRE 15. 1 -tombolo: i^Iorro del
Nicolaos (Attica, XVII). n . o it i
^ , , ~_ , , T^ , . , -I . Puerto Santo, V enezuela.
c. Double 1 ombolo. — if the island is com-
paratively near to the mainland, and if it
has considerable extension alongshore, there will generally be formed a tombol<o
from either end, enclosing a lagoon. Aebelo and Nahant are too far from thecoas t
to have a double tombolo, but Marblehead neck and Monte Argcntario (Fig. IG),
are near enough and large enough to have a bar at each end. >
(1) Only one Bar comjikted : Marblehead Neck (C. S., 335). — Only one bar is here
built, and that in recent geographic time, for the shore is not graded outside of the
tombolo, either on tlie right at the southern end of the island, or on the mainland
at the left. The tombolo has probably been built largely from the bottom, since
botli ends form nearly a right angle where they join the island and mainland.
Stony island in lake Ontario, New York, is composed of two islands, probably
drumlins, joined by a bar at the southern end, while a second bar is nearly com-
pleted at the northern end (G. S., Stony Island).
The island east of Port Townsend, Washington, is joined by one bar at the
head of Oak bay (C. S., 6405).
* La Terre.
196 PROCEEDINGS OF THE AMERICAN ACADEMY,
Bodega bead, California (C. S., 630), appears to be an island tied to tbe main-
land by a bar, probably broadened by elevation since it was built and now baving
its surface niucb diversified with dunes. A second bar is almost completed, a spit
extending from the mainland nearly across Bodega bay. Bodega head like Toma-
les point to the south is of resistant granite,* east of wliicli the longitudinal valley,
now shown by Bodega and Tomales bays, was carved in the weaker sandstones,
along a probable fault according to Professor Lawson.
Point Galero, Me.xico, is tied by San Juan beacli, and a second bar enclosing Cha-
cahua lagoon is being built (H. O. yo-3).
Copenhagen is ap.parently built upon a bar connecting Amager with Seeland,
ami the buildings and fortifications of the city have much altered tlie former ap-
pearance of the bar, harborage being gained by maintaining water communication
across tlie tombolo (Denra., KjiJbenhavn).
Helnes island is joined by a bar at its northern end (Denm., Vissenbjerg).
Overlap, offset, and stream deflection all indicate a current from the riglit, so that
the tombolo probably grew from tiie island to tlie mainland.
A small island south of Faaborg (Denm., Faaborg).
Several cases along the east shore of the Cattegat (Svve., 18, 24, 32, 41, 51, 01).
Kekenis is tied to the larger island of Alsen and tlie second tombolo is now be-
ginning as a spit on the other end of tiie island (Germ., '24; Denm., Faaborg).
An island north of Gliicksburg in tiie Flensburger tjord (Germ., 23).
Halbinsel Wustrow (Germ., 85).
Pulitz is almost tied (Germ., 04).
Several islands are strung together at the southeast point of Riigen island
(Germ., G4).
Eye peninsula is apparently tied to Lewis island (Scot., 105), but rocky ledges
are shown in the tombolo, and separation may never have been complete.
Vatersay island in the Hebrides is composed of two liigli portions connected by
3 lower neck (Scot., 58).
Peninsula point, California (C. S., 5581), is tied by one bar and a second is
)flearly comple'fed.
Conanicut isUnd (C. S., 353) in Narragansett bay is made up of two portions
joined by a b»r.
(2) B^fh Bars completed : Monte Argentario, Figure 16. — Monte Argentario, Italy,
is an I'nstvuctive example in explaining the method of tying islands.
In the interior of Orbetelio lagoon a bar extends from the mainland toward the
/Sidnd. This tombolo was probably the first built, from the mainland to the point
V(Were the village of Orbetelio now stands. Meanwhile a bar further north, Tom-
t)olo delln Giannella, was growing from the mouth of the Albegna river toward
Monte Argentario. At a little later staije shore-grade was established along the
southeast coast of the island and the Tombolo di FenifjUn grew toward the main-
land. Tiie growth of this third tombolo prevented the extension of the Orbetelio
tombolo.
The reasons for the above interpretation are as follows. The tidal inlet across
the Tombolo della Giannella is close to the island while that of the Tombolo di
Feniglia is next to the mainland. With such simple bars as these are, where there
» J. D. Whitney, Geol. Sur. Cal., 1866, I. 84, 85.
GULLIVER. — SHORELINE TOPOGRAPHY.
197
Figure 16. Torabolos : Monte Argentario, Italy.
198 PROCEEDINGS OF THE AMERICAN ACADEMY.
has evidently been no complete closing of lagoon and then a later reopening, this
would indicate the direction of growth, particularly when it accords with tlie evi-
dence from the shore curves, as it does in this case. This example then apparently
combines the features of single and double tombolos.
Monastir (Tunis, 57) is built on an island tied by two tombolos. This example
is worthy of special field studj* to bring out the relations of the several uncom-
pleted tombolos, apparently built from the mainland toward the island before the
formation of tlie present tombolos which enclose tlie others.
Jasmund is tied to Riigen island by two beautifully curving tombolos (Germ.,
42,64). At Lietzow there is a third connection with the mainland across a narrow
portion of the enclosed lagoon, but this in part at least is artificial. Transportation
is indicated as slightly stronger from the right, wliile the squareness of the bar
suggests that it was built largely from the bottom.
San Juan Nepomucino island, Lower California, is composed of two parts con-
nected by bars completely enclosing a salt lagoon (H. 0., 1223).
Margarita island, oft the coast of Venezuela, consists of two individuals joined
by two bars enclosing Laguna Grande (H. 0., 374).
Presqu'ile de Giens (Fr., 248).
V. Lngoon-marsh-meadow (Adolescence) : Colchester Point. — After
formation of a lagoon by a Y-tombolo or a double tombolo, the wind
blows in sand from the beaches and streams, and tides deposit silt, so that
in time the lagoon is converted into marsh and the marsh in turn into
meadow, if the island is not first consumed by the continued attack
of the sea.
On the Plattsburg, N. Y., sheet of the Geological Survey, at Colchester point,
Vermont, are two cases of filled lagoons, each having an almost extinguished pond.
The western pond still maintains connection witb the lake, while the eastern pond
has no outlet.
The lagoon between Cumberland head and the mainland is two thirds filled,
Woodruff pond overflowing across the last built bar into lake Champlain (G. S.,
Platt.sburg, N. Y.).
" The Bonnet" on the west side of Narrangansett bay appears to be an island
tied to the mainland (C. S., 353). Wesquage pond is the lagoon between tlie
tombolos.
Sachuest point, east of Newport (C. S., 35::!, 3044), has the lagoon between its
two connecting tombolos almost completely filled.
Monte Circeo south of Home is completely tied (Ital., 170).
Tiree island (Scot., 42) appears to be composed of two islands connected by
" The Reef." Further study is here needed.
Between San Franeisquito and Santa Teresa bays, Lower California, there is a
low dune-covered tract connecting land 300-600 feet high with the mainland. Tlie
only trace of a lagoon is the bed of a pond, half a mile in diameter, which is said
to contain fresh water during four months of the year (H. 0., 638).
Three islands are tied together and to the mainland west of Sacrificios island,
Mexico. Two of the lagoons are completely filled, and the third one is more than
half filled (H. 0., 875).
GULLIVER. — SHORELINE TOPOGRAPHY.
199
Tlie lagoon between the three or four individuals of Santa Maria island, Chile,
is completely converted into marsh (H. O., 1209).
Alki point, Washington (C. S., 651 ; G. S., Seattle). This point may never have
been separated from the mainland.
The northern portion of Unalashka island from cape Kalekhta to Constantine
bay, Alaska, has been tied by two bars to the main island. The enclosed lagoon
is a long narrow one, extending the whole distance between the bars. The map
indicates considerable filling on the sides of the lagoon (C. S., 821).
Massoncello point, Italy, upon the southern end of which Piombino is situated,
is an example where the enclosed lagoon has been completely aggraded by a river
delta, that of the Cornia Eiver (Ital., 119, 127).
VI. Vanishing Island (Adolescence). — After an island has become
land-tied, it continues to waste away by the action of the sea and subaerial
forces, until a stage is reached when the island is gone and nothing but
the tombolo which connected it to the mainland remains. This stage
must of necessity be a short one, for the unconsolidated tombolo will be
rapidly consumed. This feature would be one of late adolescence.
Theoretically we should expect to find single, Y, cuspate, and double
tombolos remaining after the islands had been
consumed. The three examples which appear
to be surely in this stage are all cuspate.
This form is probably the one which best re-
sists the sea, and each of the others is easily
converted into the cuspate.
Cuspate Tomholo: Block island, Figure 17. — The
type cusp whose position has been determined by a
former island is Sandy point. Block island (C. S.,
356).*
At the southern end of Revere beach (C. S., 337;
G. S., Boston Bay, Mass.) there is a cuspate projec-
tion where a drumlin was formerly tied on and has
now been consumed.
Uvita point on the western coast of Costa Rica
(H. O., 1035) is a cuspate foreland whose position is
apparently determined by rocky islets off the point.
This seems to be a case where the cusp is completely
tying the island when the island itself is practically
destroyed.
Figure 17. Vanishing Isl-
and ; Diagram of a Cus-
pate Tombolo. Similar
Stage found in Sand}'
Point, Block Island,
Rhode Island.
VII. Straight Coos^ (Maturity). — The mature stage of island-tying is
where the islands and their connecting tombolos are completely consumed
* See Livermore's History of Block Island, 1877, 175.
200 PROCEEDINGS OF THE AMERICAN ACADEMY.
by the sea. Therefore the straight coasts of Jutland, Italy, etc., as given
on page 246, are the forms of the stage next succeeding that of the
vanishing island.
Complex Cases of Tying. — Many of the actual examples of island-
tying are not simple. A great variety of combinations occur in nature,
but only three will be here considered, viz. where several tombolos unite
a group of islands, where rivers surround islands with their waste, and
where slight movements of the land have assisted tying.
Samso Island is composed of two higher portions joined by a lower narrow neck
(Denm., SamsiJ). The central portion of this neck is lieath and forest, presumably
overgrown marsh, bounded on either side by gently curved shores. These curves
also indicate tying and complete filling, for the coast farther north and south has
not such smooth outlines, indicating that the present position of the land has not
been maintained for a time sufficiently long to develop such curving shorelines.
The left hand end of the eastern bar is complicated on account of numerous small
islands.
Three islands in Lenox cove, Tierra del Fuego (II. O., 455').
West of Magdalena bay an island between cape Corso and Entrada point is tied
to another island at cape Lazaro (II. ()., G21, G44).*
]Marambava mountain, 20GG feet liigli, has a twenty-mile tombolo extending to
the mainland of Brazil, which close to tlie sliore is broken by a tidal opening. A
spit from the tombolo inside of Sapetiba bay is growing toward Jaguanao island
(H. 0., 488).
In Boston harbor there are three groups of islands tied by numerous tombolos
(C. S., 337; G. S., Boston, Boston Bay, Mass.), viz. the Winthrop, tlie Quincy,
and the Nantasket groups. Marshes occur in all three, indicating adolescent
development.
Sidi bon Said (Tunis, A'll, VIII, XII, XIV, XX, XXI) is on an island which
is tied to tlie Tunis mainland by three bars. The central one is 5 to 8 kilometers
broad, 10 kilometers long, and has an elevation in places of 10 or 12 meters.
Whether this broad istlimus was originally two tombolos enclosing a lagoon, or was
made land by elevation, is not certain from map inspection. Later, however, two
tombolos have been built from eitiier end of the island, enclosing between them
and the earlier built isthmus two lagoons, Sebkhat er Riana and Lac de Tunis.
Leucate (Fr., 255) is an island of Oligocene strata, tied by one tombolo to the
mainland, and has also a wing-like bar on both the right and left sides. These wing-
bars are built up from the bottom in large measure, according to the indications
given by the right-angled abutment of the left end of the right bar against Leucate,
and a similar abutment of the left bar against the older land near Port Vendres
(Fr., 258). Tlie stream deflections indicate alongshore motion to the right, which
is also suggested by the offset of the right wing-bar by the left.
The island of Cette (Fr., 23.3) , of Jurassic strata, is tied by the wing-bar from the
left side of the Hhone delta. To the southwest the volcanic knob of Agde is more
* In regard to dislocation as a probable cause of these islands, see W. Lindgren,
Proc. Cal. Acad. Sci., 1888-90, I. 173, II. 1, III. 26, and references there given.
GULLIVER. — SHORELINE TOPOGRAPHY. 201
completely surrounded by the aggraded detritus of recent times (Fr., 244, 245).
The latter may never have been an island.
Berneray island in the Hebrides is made up of several higher portions connected
by sandy areas (Scot., 89).
A group between Brandon and Tralee bays (Ireland, IGI).
Eddy island is a composite island, in which there are two examples of lagoons
almost included by tombolos (Ireland, 114, 115).
Several islands at the head of Galway bay seem to belong to this class
(Ireland, 115).
The type of islands tied to the mainland by delta growth is seen in lougli Swiliy
(Ireland, 11), where Inch Top island, 702 feet high, appears to be joined to the
mainland on the east side of the bay by the detritus borne down by the small
streams. Other knobs to the southeast of Inch Top hill were possibly tied in a
similar manner.
The islands of Paleozoic, Mesozoic, Tertiary, and eruptive rocks are surrounded
by the delta deposits of the Danube (Taf. III. Jahrb. k. k. Geol. Keiclis., XL., 18^0).
From San Pablo point to Richmond point is an island completely joined by
marshland (C. S., 5581). In this case streams have evidently aided the tidal cur-
rents in filling in between a former island and the mainland.
North head, McKensies head, and various other islands at the mouth of the
Columbia river (C. S., 681", 640), are seen upon inspection of the more detailed
charts and Mr. Davidson's sketches* to be tied together to form cape Disappoint-
ment, which is in turn tied to the mainland at Chinook point.
An example in which slight elevation may have helped island-tying is seen at
the mouth of the Medjerda river (Tunis, VII, XIII, XIV, XX).
Oland and Gjol are becoming land tied by river and tidal deposits, probably
more tidal than river, since the elevation of the land from which the streams come
does not exceed 75 meters (Denm., Nibe).
An example of complex island-tying is seen on the chart of San Quentin bay.
Lower California (II. O., 1043). It would appear that the earliest tying was done
wlien the land stood lower than at present, for some of the bars outside of the salt
lakes are cliffed.
10. Bat-bars.
An Adolescent Feature. ■ — Shore development of a submerged region
has been studied as regards island-tying ; a second important feature is
now to be considered. It has been shown that when sliore-frrade is
attained detritus will be transported along the beach at the foot of cliffs,
and tombolos connect many islands with the mainland. As the head-
lands are attacked faster than the bay heads on account of their more
exposed position, wing-bars will frequently be formed of the detritus
from the cliffs. This special form of bay-bars will be considered under
Winged Beheadlands.f
* Pacific Coast Pilot, 1889, 451. t See page 2i:l
202 PROCEEDINGS OF THE AMERICAN ACADEMY.
The sea also erodes the bottom and supplies material for the bar.
The proportion of bottom to side supplied detritus will vary exceedingly.
With a deeply dissected, steep coast, the proportion of material from
the headland will be large, while in a slightly drowned region, devel-
oped to past-maturity in the iirevious cycle, there will be more material
under water above wave-base, and therefore a greater proportion of
bottom detritus. There are so many variables which enter into this
problem, — viz. initial form, prevailing winds, strength of currents,
height of tides, radius of curvature of eddies, structure of laud, etc., —
that it is difficult to predict where a bar will be built across a bay.
It may be said, however, that the sea is not satisfied with an irregular
shoreline, and in its attempt to reduce the land to a submarine platform
it will straighten the shoreline in order better to attack the land. The
curve that a given shore will take depends upon the forces acting at
that point.
In one place wings will extend from the projecting headland, in another
the currents will build a bar across the mouth of the bay, in a third
the bar will grow from a point between headland and bay-head, while in
a fourth [)lace the alongshore action may be so weak or the bay so broad
that the sea will begin to fill at the head. In this fourth case any delta
filling will go on at the same place as the accumulation by sea action.
When the bay-bar is completed, and there is transportation of material
practically all along the shore, shore-grade is attained, and the period of
adolescence in shore evolution is reached. Tlie narrow and broader
bays behind the bars are gradually filled by river, tide, and wind. "Where
the river activity is strong enough, it pushes a delta beyond the bar.
Maturity is reached when the bays are filled and the headlands cut back
so that the initial shoreline is lost. From this time forward the sea,
satisfied with the shore curves, eats farther and farther into the land with
the intention of reducing all that stands above wave-base to a monotonous
submarine platform.
The classification of bay-bars here given is not a satisfactory one.
The separation into stages of development is only partial, for more facts
of observation are needed. The location of the bar in the bay, which
depends upon the ratio of alongshore to on- and offshore currents, as well
as upon the form of the bay, has been used to make three types of bay-
bars. Under each of these, stages of filling occur, all centring about
the attainment of shore-grade, and therefore bay-bars may be limited as a
class to a period extending from late youth to early maturity. Bay-bars
are characteristic of adolescence.
GULLIVER. SHORELINE TOPOGRAPHY.
203
A. Bar across Mouth of Bay: Lake Ontario, Figures 18, 19. —
Several bays on the eastern shore of lake Ontario are closed by bars,
whose form indicates more bottom than alongshore action, there being
3
_J Kilometers.
Figure 18. Bay-bar across Mouth of Bay, Lake Ontario.
no dominant offset and overlap. This indication is confirmed by the
observations of the currents, many of the courses of observed bottle
drifts having ended on these bars.*
Instead of looking at these sandbars as barriers to keep the sea out of
the bays,t let us regard them as built by the sea in order to prevent the
wasting of its force dashing into the indentation, where the delta growth
will finally be victor, and in order that the sea may be able to concentrate
* Surface Currents of the Great Lakes, U. S. Dept. of Agriculture, Weather
Bureau, Bull. B, Washington, 1895.
t See Geikie, Scenery of Scotland, 1887, 186.
204 PROCEEDINGS OP THE AMERICAN ACADEMY.
its force upon the more exposed coasts, by having a simpler coastline
upon which to work.
(1) Bay but little Jill ed (Youth-Adolescence). — Bay-bars are forming across
tliree bays on the Oldenburg sheet (Germ., GO). Two of tliese are on the southern
side of Fehmarn island and tlie third is south of Grossenbrode. In Orther bay
transportation is from the left, but in the other examples it is about equally from
right and left headlands.
Gruber bay is enclosed b^- a bar (Germ., 60, 84). The deflection of the outlet,
Dahmer-See, to the left indicates a dominant current from the riglit.
Stettiner bay is closed by a bar wiiich shows a very beautiful series of aggrada-
tion shorelines (Germ., 89, 90, 91, 92, 120, 121, 122, 154, 155, 187). Tlie dominant
current is indicated by offsets, overlaps, and stream deflections to be from the left.
The contest between river and tidal currents on tlie one hand and alongshore cur-
rent on the otlier is clearly shown. A study of details on the ground in connection
with these expressive general maps ought to bring out many features of the pro-
gressive steps in the formation of bay-bars. Islands are included in this bar and
thus complicate its form. Usedom island is made up apparently of several indi-
viduals, and WoUin island is in large part a portion of the drowned mainland and
not the later built foreland.
Three bays formerly arms of Hochwachter bay are enclosed (Germ., 59). The
indications are of a dominant current flowing from the left.
Several examples from Kiel nortliward (Germ., 39, 58).
Warneniiinde is built on a bay-bar (Germ., 86).
Kurische and Frisciie bays (Germ., 1, 3, 8,9, 15, 16, 29, 30, 48,49, 50, 71,72,73).
Carder, Dolgen, Leba, Sarbsker, and Zarnowitzer are enclosed to form lakes or
lagoons (Germ., 25, 26, 44, 45, 46).
Vietzker lake (Germ., 43).
Vitter lake (Germ., 66).
Jamunder and Buckower lakes (Germ., 65).
Kamper lake (Germ., 93).
IIorst-Eiersberger lake (Germ , 92).
Bankel-damm is shut in by a bay-bar (Germ., 13). Transportation is about equal
from right and left according to the map indications.
Schlief-see is closed by a bar growing from the right, for on that side the
curve of the cliff is continued in the line of the bar, while on the left the bar
abuts abruptly against the oldland, forcing the stream under the left hand bluff
(Germ., 13).
The Sejrslev headland has a right and left wing growing across bays (Denm.,
LugstiJr).
A cuspate bar extends from the left hand side of Ilorsens fjord toward a rock
near Alro island (Denm., Skanderborg).
Across the mouths of some ten drowned valleys, between tlie Dnieper and the
Danube rivers on the Black sea, bars have grown (Rus., .33; Atlas Univ., 38),
More than half of them are completely closed by the sea action. Tlie low mean
annual rainfall in this region, 1.5.83 inches at Odessa,* would cause weak stream
* E. Loomis, Contributions to Meteorology, revised ed., 1889, 151, PI. XXIII.
GULLIVER, — SHORELINE TOPOGRAPHY. 205
action. The waves and currents, though weaker on the inland sea than on the
open ocean, are relatively stronger than the streams, for they are able to close
these bays. The absence of ocean tides, which tend to keep open inlets into bays,
aids this shutting up.
Across the western end of the sea of Azov a bar has been formed (Atlas Univ.,
38 ; Rus., 48, 62) which has nearly rectangular junctions with tiie mainland, indi-
cating that it has been built largely from the bottom of the sea.
Lituya bay, Alaska, has the right spit offsetting the left. Glaciers now descend
to sea level in various arms of this fjord (C. S., 8451).
Thomas bay, Alaska, has bar forming between Vandeput and Wood points
(C. S., 7.33).
On Amaknak island, Alaska, the spit has grown southward from Ulakhta head
more than half way toward Rocky point (C. S., 821).
Coburg peninsula west of Esquimalt roadstead, Vancouver island (H. 0., 1306).
Tomales bay, California (C. S., 631), has a bay-bar extending three quarters of
the way from the left toward the right side of the bay, thus indicating a dominant
current from the left.
Willapa bay, Washington (C. S., 6185), shows incurving spits at the end of the
bar, pointing up the river.
(2) Baji more or /ess ^//ec? (Adolescence), Figure 19. — Silting up of the bay
enclosed by a bar progresses rapidly, as engineering works testify. Streams, tides,
and winds fill this quieter water with waste. The changing conditions of along-
shore transportation will be shown by the advance or retreat of the shoreline
of the bay-bar.
A bar has grown from the right across the mouth of Mobile bay (C. S., 187,
188). Successive positions of this bar are indicated by some eighteen sympatheti-
cally curving dune ridges with intervening stream or marsli. The offset here indi-
cates a dominant current from the right.
Tampa bay, Florida (C. S., 176, 177), shows overlap from right to left, thus
indicating a dominant current from the right.
The overlap of the right bar at the mouth of the harbor of Rio Grande do Sul,
Brazil, indicates a prevailing current from the riglit (H. 0., 1191).
A bar is built from Palmia point to Gorda point across the drowned valley of
San Jose river. Lower California. Its southern point overlaps and offsets the
northern portion, indicating a current from the left (H. 0., 635).
Several bays on Monte Gargano, Italy, show nearly complete filling (Ital., 157)
South of Soby on Arii island Vidsii bay is closed and the lagoon is considerably
filled with marsh. A nip decreasing in height toward the bay -head is clearly
shown on the German map (Germ., 24; Denm., Faaborg). These two surveys
differ decidedly as to tlie amount of filling.
On the eastern side of the southern point of Falster island (Denm., Gjedser)
the sea has a curving shoreline of large radius upon a low sandy coast behind
which lies an enclosed lagoon, to the west of which is an area of higher land.
The offsets and stream deflection indicate a prevailing current from the right
Botonor lake Jias a bordering belt of marsh, and there are other patches of marsh
between this lagoon and the eastern shore. All the above facts sug'jest the growth
of a bar across the mouth of an open bay. In front of the artificial sea wall,
built to protect the coast, the map shows a belt of sand, as if the sea was even
now building out in places.
206
PROCEEDINGS OP THE AMERICAN ACADEMY.
On the western coast of Langeland and at the soutliern point are several enclosed
bays partly filled (Denm., Svendborg, Gulstav). The offsets indicate movements
in both directions.
The bays on the southwest side of Chirikof island, Alaska, sliow several stages
of filling (C. S., 9191, old number 796). The sea built five bars from headland to
SOUTHWEST ANCHORAGE
CHIRIKOF ISLAND
ALASKA
Figure 19. Nearly Mature Filling of Bays.
headland, in swinging curves satisfactory to itself. In the broadest bay there still
remain two lagoons, but the others are completely filled (Fig. 19).
Kiska harbor on Great Kiska island, one of the Aleutian islands, has three hays
showing three stages of filling. In the central one there is a large lagoon com-
pletely enclosed, in the northern one a small lagoon remains, and in the southern
one the area back of the bar is completely converted into marsh (C. S., 9191).
The alongshore current is here probably from the left, witness overlap of bars to
the right, and streams crowded to right side of bays.
(3) Bay filled (Maturity). — The last few examples in the previous section are
as near to mature forms of this kind of bay-bars as have been found in the present
study. Mature stages of bay-bars merge into mature ria-deltas so closely that one
can hardly separate them upon the map. When the bay is filled, the stream
mouthing in the bay will attempt to push forward its delta, and the shore form
GULLIVER. SHORELINE TOPOGRAPHY. 207
then depends more upon the ratio between stream and current, than upon the at-
tempt of the sea to bridge across the bay. The day of tlie bay-bar is over.
B. Bar in Middle of Bay. — The form of the bay or the strength and
position of the currents may cause a bay-bar to grow from the side of
the bay, at some point between the head and mouth of the bay. If the
growth of the bar is due largely to alongshore action, a spit will extend
from the side of the bay ; but if the bottom action is dominant, the bay-
bar will abut nearly at right angles against the coast of the bay.
(1) Bay but to/e^/ec? (Youth-Adolescence). — Yakutat bay, Alaska, has a bar
forming at point Turner, about half way between mouth and head (C. S., 8451).
Ciiignik bay, Alaska, has spit growing from right side of bay only (C. S., 8891).
Kachemak bay, Alaska, has spit grown about half way across from its left side
(C. S., 766).
Salinas bay, Lower California, has a bar enclosing a salt pond into which there
is little or no drainage (H. ()., 850).
Inverness or Moray firth (Scot., 83, 84, 93, 94) has a spit growing from either
side, of which the left one is considerably near the mouth of the bay. The Dor-
noch firth (Scot., 94, 103) has spits in similar positions.* The writer questions
whether the position of these spits indicates a dominant current in these two bays
from tlie left.
Hagios Nikolaos bay (Attica, XVII). Current probably from riglit.
The small bay, Hejisminde, on the eastern end of the boundary line between
Denmark and Schleswig is shut in by a bar whose curve is continuous with that of
the coast to the north, indicating a current from the right. On the left side of the
bay there is nearly a right angle between bar and coastline, indicating that the sea
builds here mainly from the bottom (Germ., 7; Denm., Skamlings Banke).
(2). Bay more or less Jilled (Adolescence). — Marathon bay in Greece (Attica,
XVIII, XIX). The former courses of the Marathon river and the overlap and
stream deflection to the right indicate the prevailing current to be from the left.
On the right side of the bay the bar abuts against the oldland forming nearly a
right angle with Kynosura point. This fact indicates that the bar is built mainly
from the bottom.
Mosvig bay (Denm., Skamlings Banke) shows a smoothly curved bar continuous
to the right and left with the shore curves, indicating filling from either side. The
dominant direction of alongshore current is indicated as from the right by the
stream deflection to the left. Behind the bay-bar the lagoon is almost completely
filled with marsh.
The sea has built Tent moor and Barry links upon the two sides of the firth
of Tay (Scot., 49) and the rectangular junction of the bay filling with the bay sides
indicates that the building has been done from the bottom. Deflection is toward
the central channel, where the relatively strong tidal currents interrupt the for-
mation of a typical bay-bar.
Dornoch firth is a similar example (Scot., 93, 94, 102, 103).
* Geikie, Scenery of Scotland, 1887, 180. 187.
GULLIVER. SHORELINE TOPOGRAPHY.
209
(3) Bay filled inside Bar (Adolescence). — Under this heading are included several
cases in whicli it is impossible to tell whether the bar originally was found where
it now stands or whether it began near the bay head.
Procchio and Biodola bays (Elba).
Tiie gulf of Salerno has a bar from Salerno across to Agropoli, inside of which
it has been filled by river and tidal action, forming a rich delta plain with only a
few remaining marshy places (Ital., 185, 197, 198).
Gulf of S. Eufemia (Ital, 241). Current from the left is indicated by deflec-
tion of streams to right.
From Palmi, Italy, across to Nicoteca a bar extends from the right headland to
a point across the bay half way between the left headland and the bay head (Ital.,
215, 246). Direction of current is probably from the right.
Bay of Phaleron southwest of Athens (Attica, III). The bar is largely pro-
duced by the action of the sea on the bottom, for the shore curves at its two ends
are not continuous with curve of the bar.
Vari bay (Attica, VIII). Tliis also shows sea bottom action.
Hanu bay (Swe., G). Dominant current from the right.
C. Bar near Head of Bay: Drakes Bay, Figures 20, 21 — At the
head of Skelder bay (Swe., 8j a bur is built chiefly from the bottom,
since the bar curve abuts sharply against the two sides of the bay, and
streams are deflected to the right on the right hand side and to the left
on the left hand side of the bay, thus indicating currents in either direc-
tion from the centre (Fig. 20). There is evidence of very trifling trans-
portation along the sides of Skelder bay.
Figure 20. Diagram of Bay-bar at Head of Bay ; dominant Bottom Action :
Skelder Bay, Sweden.
VOL. XXXIV. — 14
210 PROCEEDINGS OF THE AMERICAN ACADEMY.
Tlie indications from riglit-angled abutment at both right and left sides of
Laholms bay (Swe., 8, 13) are that the bar is built mainly from the bottom with but
little transportation along the sides of the bay. The stream deflections from either
side of the bay toward a probable oldland island now included in the bar indicate
currents toward the centre. This is in marked contrast to the last example, taken
from the same coast.
Nearly as typical an example is in Dingle bay (Ireland, 172) where the bottom
action is dominant, though there is indication of some transportation from the
right side of tlie bay.
There are many cases of bay filling of this type along the Irish coast (Ireland,
7,9, 15, 51,62, etc.).
Dundrum and Dundalk bays on the west shore of the Irish sea (Ireland, 61,
70, 71, 81, 82).
Drakes bay, California (Fig. 21), is a broad open bay, with a bay-bar growing
near its head. Delta filling is now progressing between the nip and the bar.
Another typical example of the head of a broad bay being aggraded by sea
and streams working together is seen in the gulf of Taranto, Italy. From tlie
swinging outline of the shore the sea action is seep to be stronger tiian the river
action, the deltas are either rounded or stunted. The dominant direction of current
is seen by the offsets to be from the left (Ital., 212, 201, 202).
At the southwest corner of this gulf where the Crati river enters, the shore is
also aggrading. Here the several streams form a great confluent delta, whose form
is modified by the sea (Ital., '221, 222, 229, 230).
In Wachusett cove, Alaska, the curve of the shore at the head of the bay was
nearly satisfactory to the sea forces, acting therefore sea and river fill at the same
point. The great rise and fall of the tides, 18 feet, doubtless prevents the forma-
tion of a bar across the mouth of the cove from Bluff point (C. S., 734).
Nateckin bay, Alaska (C. S., 821).
In the upper portion of tlie valley of San Francisquito bay, Lower California
(H. 0., 638).
Harbor of Acapulco, Lower California (H. O., 872).
San Juan del Sur, Nicaragua (11. O., 934).
Manzanilla and Santiago bays, Mexico (II. O., 915).
Filling is shown on west coast of Central America (H. 0., 1016, details in
1025-1033).
Todos Santos bay, Lower California (II. 0., 104G).
Several bays on the east coast of Scotland are of this tj'pe, Lunan, Montrose,
Aberdeen, Cruden (Scrot., 57,77, 87). The beaches abut sharply against headlands
of irregular shorelines, an indication of bottom building.
Sinclairs and Dunnetbays (Scot., 116) similarly indicate growth from the bottom.
Several beaches on Tiree island (Scot., 42) are of this type.
Magilligan Foreland, Figure 22. — Magilligan point (Ireland, G, 12) is
a bay-bar which combines some of the characteristics of tidal cusps witli
the normal bay-bar ^features. From the cliff on the left of the bay a
gently curving beach extends to the end of the point, indicating growth
by alongshore transportation from the left. McKinneys bank points up
GULLIVER. — SHORELINE TOPOGRAPHY.
211
Figure 22. Magilligan Foreland, Ireland.
s
_| Kilometers.
212 PROCEEDINGS OP THE AMERICAN ACADEMY.
lough Foyle from near the tip of Magilligan point, and its straightness
and length suggest a fairly constant and strong tidal current through the
inlet, three fourths of a mile broad, between the point and the right side
of the bay.
This foreland has the cuspate outline appropriate to a tidal cusp, but
on the bay side it shows no line of growth. Back of the present ocean
beach, however, are seventeen roads which curve sympathetically witli
the shoreline of to-day. These roads are not parallel lines, but each
curve is nearly parallel with the two on either side, departing just enough
from parallelism to make a systematic series, changing gradually from a
direction a few points south of west to one nearly northwest and south-
east. Between almost every pair of adjacent roads is a ditch, whose
course is systematically accordant with the direction of the roads. These
culture lines are not constructed in a haphazard manner, and their orderly
arrangement suggests a series of successive curves of higher and lower
ground. Such a systematic series indicates lines of growth. If this
indication be true, then this cuspate bay-bar began some five miles farther
up the bay. and has advanced by some seventeen steps to its present
position.
As said above, the bay side or right side of the cuspate bay-bar shows
no airirradational line of growth. Each of the roads and ditches ends
at the liigh tide line, forming transverse features in respect to the bay
shoreline, while the same are longitudinal with respect to the ocean
shoreline. This abrupt ending, together with the continuous curve of
the bay shoreline, suggests that the sea may have shaved off the ends of
these presumable former beaches.
As confirming the above hypothesis for the formation of Magilligan
point, it is observed that the cliffs, some 200 feet high, where the along-
shore action changes from cutting to aggrading, continue as a nip behind
the Magilligan foreland. This cliff is progressively lower toward the
supposed earlier formed portion of the foreland. This would be expected,
since the portion of the cliff which was exposed for a shorter time to sea
action, other thiniis being equal, would have the less height.
The dominant current from the right, which is indicated by the form
of this bay-bar, is possibly due to a backset eddy between Islay island
and Ireland. The ocean current flowing from the southwest along the
west coast of the Britiah islands * could cause such a backset eddy in
this locality.
* Eriimmel, Uebersichtskarte der Meeresstromungen, 1886.
GULLIVER. — SHORELINE TOPOGRAPHY. 213
11. Winged Beheadland.
A Combination of Bay-bar arid Cliff. — One of the striking features
of adolescent shore development is the winged beheadland. Where the
projecting headland has been cut back and transportation has taken place
in both directions, spits extend to the right and left from the headland.
The winged beheadland is made up of a cliffed headland and two bay-
bars, one extending to tlie right and the other to the left; and this com-
bination of wiug-bars with a headland beheaded is so striking a form,
and one so characteristic of depressed regions developed to adolescence,
that these forms have been grouped separately.
Type: Long Branch, N. J. — From Sandy point to Barnegat inlet is
one of the finest winged beheadlands to show the method of formation.
The records since the Europeans came to America show the cutting back
of the cliffs on the headland, and the shifting of the wings on either side.
From the slope of the land above the cliffs upon the headland, it is seen
that the land probably projected several miles beyond the present cliffs.
The consumption of this land supplied more waste than the sea could
immediately carry offshore, and it was temporarily deposited, forming
Sandy hook and Island beach. The left wing was probably built in large
part from the bottom, and represents the offshore bar of the low New
Jersey shore. The cliffs upon Navesink highlands represent the nip,
cut before the barrier of Sandy hook was formed.
Other Examples. — Cape Cod is an example of a winged beheadland which pro-
jects far into the sea, so tliat the wings do not extend across bays on either side.
The series of changes of the right or Provincetown wing, as indicated by tlie form,
has been worked out by Professor Davis.*
Another typical example is seen on the island of Laaland (Denm., Nakskov,
Maribo, Dannemare).
Scjrslev headland (Denm., LiJgstur) shows the same form.
S3'lt island off the west coast of Schleswig has a right and left wing extending
respectively toward Amrum and Hum islands (Germ., 11, 20, 21, 35, 36). The
Miocene headland (Geol. Eu., 24) is separated from the rest of the oldland by the
drowning, and is therefore somewhat different from the type example.
Insel Poel and Halbinsel Wustrow north of Wismar have both developed wings
(Germ., 85, 116). These wing-bars have been curved strongly back toward tiie
mainland.
Similar wings from islands are seen farther north (Germ., 41, 63).
Usedom island is made up of two main headlands with wings (Germ., 89, 90,
121, 122).
WoUin island (Germ., 91, 121, 122). »
* Proc. Amer. Acad., 1896, p. 303.
214 PROCEEDINGS OF THE AMERICAN ACADEMY.
Samland with its wings protecting Kurische and Frisclie bays (Germ., 3, 8, lo,
16, 28, 29, 48, 49, 71, 72).
The Pomeranian headland has its right wing enclosing several lagoons, while its
left wing is growing into Danziger bay (Germ., 25, 26, 27, 44, 45, 4b, 47).
West of tlie Crimean peninsula there is a very striking winged beheadland
(Rus., 33). The initial shoreline is very evidently changed by the cutting back of
a projecting headland and the extension of wing-bars, beliind which the less altered
shoreline is seen. Although the map is not of the highest quality, the scale is
small, and the geology of the region is unknown to the writer, he has no hesitation
in classifying this feature on account of its typical winged outline.
North of the left-hand wing of the last example is a somewliat similar area,
whose origin is, however, not so clear. The outline is similar to a winged behead-
laml, but the headland is full of small lakes, and it seems probable that it is a i)art
of the Dnieper delta drowned by a late episode of depression. A submergence is
indicated by the drowned rivers to the northwest (Rus., 33; Stelier, 48).
12. Tidal Cuspate Forelands.
Location and Description. — In regions of drowned valleys, long inlets,
or narrow sounds, where the two opposite shorelines are roughly parallel
to each other, cuspate deposits of sand frequently occur, when shore
development has readied an adolescent stage and transportation along-
shore has begun.
These forelands project from one quarter to three quarters of a mile
into the sea, and vary in breadth between the same limits. In some
cases the cusps are long and narrow, while in others they are short and
broad. Frequently they more or less completely enclose lagoons, but in
some instances tliere is no included water body, or if there was one it has
become filled. The curve of the two outer edges of these sand deposits
is concave toward the water, and is a continuation of the curve at the
base of a shore cliff. These two concave curves intersect in a marked
cusp, which is sometimes typically pointed. thou<;h in other cases the tip
is rounded. The axis of these forelands projects approximately at ri^ht
angles to the shoreline, and also at right angles to tlie general direction
of the tidal currents in the inlets.
T>/pe, Figure 23. — West point, north of Seattle, Washinston, will
be taken as the type, and, after giving its description and discussing the
method of its formation, others differing in details of form will be con-
sidered. Magnolia bluff, two miles northwest of the city of Seattle, has
a gently swinging curve, doubtless quite satisfactory to the current here
prevailing. This curve continued forms the right boundary of the West
point cusp. The curve on the west side of the foreland is in like manner
a continuation of the curve of another cliff (C. S.. 658 ; G. S., Seattle).
GULLIVER. — SHORELINE TOPOGRAPHY.
215
On the inside of the cusp there is a faint cliff where the coast was nipped
after the initial drowning. The central lagoon is nearly all converted
k--"-'^
Figure 23. Filled Stage of Tidal Cuspate Foreland : West Point. Washington.
into marsh, a small tidal inlet remaining on the left side with a few small
ramifying branches. The cusp is very perfectly formed by the intersec-
tion of the two curves in a sharp point.
Sand point in Narragansett bay is nearly as typical in form and
position as West point. This point projects from the eastern side of
Prudence island (C. S., 353) into a channel less than two miles broad
and from 10 to 17 fathoms deep, 5.^ fathoms off the point of the foreland.
This cusp is smaller than the average tidal cusp, and it shows no included
Isgoon or marsh. Mr. J. B. Woodworth says that the ice in winter
overrides this cusp, and thus any indications of embryonic form would be
obliterated. The secondary cusp on the left side of the foreland appears
to be due to the collection of sand about rocks or piles driven into the
sand.
A profile has been drawn from another typical cusp, point Wilson
(C. S., 6405), north of Port Townsend, Washington. This drawing (Fig-
ure 24) shows the relation of the foreland to the older mainland. The
broken line indicates the probable initial form of the land following the
depression which inaugurated the present cycle of shore development.
The " foreland" quality of the cusp is here clearly seen ; it is constructed
216
PROCEEDINGS OP THE AMERICAN ACADEMY.
by transportation and accumulation in front of the nipped oldland.
Although plotted from the soundings and contours about point Wilson,
this figure will serve as a general profile of all the cusps of this class.
Where the initial slopes were less steep, less contrast is seen between the
oldland and the foreland.
.SL
Kilometers.
Figure 24. Profile of Tidal Cuspate Foreland : Point Wilson, Washington.
Tidal Hypothesis. — Before considering other cusps which differ some-
what from West point, let us look for a moment at what might be
expected to result in narrow channels with sides nearly parallel. Waves
would attack this inner shoreline to a greater or less extent at all points.
When adolescence is reached in the process of shore development, and
waste is supplied faster than it can all be carried offshore, it will be
transported and deposited somewhere* The great system of ocean eddy
currents is not able to affect this inner as it does the outer shoreline.
Local winds must pro-
duce small currents pro-
portional to the size of
the water bodies, but
these will be so weak
in narrow channels that
their effects will be lost
in those of even mod-
erately strong tidal cur-
rents. Thus it seems
safe to conclude that the
probable agent of trans-
portation in such chan-
nels is the tidal ebb and
flow.
An ideal scheme of inflowing tide, with the eddies which would prob-
ably accompany it, is given in Figure 25. Where the movement is
least in the triangles of comparatively dead water between the several
Figure 25. — Ideal Scheme of Tidal Inflow : Port
Discovery, Washington.
* Compare pages 176-178.
GULLIVER. — SHORELINE TOPOGRAPHY.
217
Gaspee Point
members of the circulatory system, the deposition would take place. In
the majority of places the outflowing tide would reverse the direction of
flow and transportation of shore waste, therefore the combined action of
ebb and flow would shape the tidal foreland so that its central axis would
be at right angles to the general direction of tidal flow.
The cuspate forelands, which are mentioned under the three following
heads, are arranged in three stages of progressive development, — the
V-bar stage, the lagoon-marsh stage, and the filled stage.
V-bar Stage. — A much younger stage than that of West point is seen on the
same sheet at Meadow point. Here the bars surround a relatively large lagoon,
which apparently has hardly begun to fill. The form of this bar is what Mr. Gilbert
has called V-shaped.*
Various examples on the east
coast of Port Discovery (C. S.,
648) show V-shaped bars enclos-
ing lagoons. Tlie majority of
forelands in this bay have their
greater extension alongshore.
Beckett point, however, has its
length normally at right angles
to the shoreline.
At point Monroe, near Port
Madison, Washington (C. S.,
663), a looped bar encloses a
lagoon somewhat similar to those
just mentioned. The shore drift
is here all from the left, and the
curve of the bar is convex sea-
ward. At point Jefferson farther
north on the same sheet there is
another convex bar enclosing a
lagoon where the drift has been
from the left, as shown by the
continuation of the cliff curve in
the bar. These two examples
do not give the typical cuspate
form.
Laqoon-marsh Stage, Figure
26. — Various stages of lagoon
filling are shown on the Port
Townsend sheet (C. S., 6405, old number, 647). Walan point foreland has consid-
erable area of lagoon, and still maintains open connection with Port Townsend.
At point Hudson there remains an unfilled lagoon, but its connection with the sea
is lost. At point Wilson a small lagoon now exists, while at Kala point the lagoon
Figure 26. Lagoon-marsh Stage of Tidal
Cuspate Foreland : Gaspee Point, Narra-
gansett Bay, Rhode Island.
* Lake Bonneville, 57, 58, PI. VII.
218 PROCEEDINGS OF THE AMERICAN ACADEMY.
is practically converted into a marsh. On the foreland at Marrowstone point the
sand dunes have almost obliterated tlie marsh.
On this same Port Townsend sheet the rounding of the point of the cusp may
be studied. At point Wilson the concave curves intersect in a slightly rounded
cusp, while at Kala point the cusp is more blunt, and VValan point is decidedly
rounded. The curves at point Hudson have a long radius, so the sides of the
cusp are nearly straight, and since they meet at nearly right angles the foreland
has a broad flattened appearance. The curve on the right side of Marrowstone
point changes from a concave to a convex form, so that it gives that side of this
foreland a snubbed look.
Sand point, projecting into Popof strait, Alaska (C. S.,8891), is a fairly typical
example of a cusp with enclosed lagoon. The point is here somewhat blunted,
more on the southern than the northern side. This foreland as mapped is evidently
a piece of made land, built forward in the process of shore development. ^
A typical example is seen in New Dungeness harbor, Washington (C. S., 046),
where inside of the beautiful hooked spit forming the harbor tlie foreland projects
with a very sharp point.
Gaspee point (Fig. 26) in Narragansett bay (C. S.,3047) may be taken as a typical
example of this lagoon-marsh stage.
A rounded cusp with completely enclosed lagoon occurs near the mouth of
llorup bay (Germ., 24 ; Denm., Faaborg). Upon the same sheet there is a typically
sharp pointed cusp projecting from the north end of Aru island. This projects
at right angles to the general shoreline, but the belt of water is iiere so wide that
the wind-made currents probably have as much controlling influence as the tidal,
possibly more.
Filled Stage, Figure 23. — West point, Washington.*
Dungeness point t on Romney marsh, England, is a cuspate projection into the
English channel (Eng., 4).
In tiie eastern entrance to Magellan strait, South America, is one of the largest
known forelands of tliis class. Westward from cape Virgins and south of a nipped
cliff 100 to 300 feet in height projects from five to six miles a second Dungeness,
named by some British seamen in recognition of a form similar to that of tlie great
Englisli sand cusp (H. 0., 443, profile in View A).
Sandy point, Magellan strait, Soutii America (H. O., 450"), is another example.
On Douglas island opposite Juneau, Alaska, is a tidal cusp at low water, while at
high tide it is covered (C. S., 734). The rise and fall of tide at tins point is 18 feet.
Sextant point, San Quentin bay, Lower California, is apparently a cusp built
out between two rocky headlands (H. O., 1043).
Estauques point, Venezuela, is a long narrow cusp (H. O., 1087).
Alice point, on tlie bottom of the foot of the Italian boot, is a foreland which
sliows no included marsh. Its axis if projected across the gulf of Taranto would
touch the extremity of the heel, as if its existence showed the attempt of the sea
to close the gulf (Ital., 231).
* See page 214.
t Topley, Geol. of the Weald, 1875, 211, 303; F. Drew, Romney Marsh, Mem.
Geol. Sur. Eng. and Wales, 1864 ; F. P. Gulliver, Dungeness Foreland, London
Geog. Jour., 1897, IX. 536-546.
GULLIVER. SHORELINE TOPOGRAPHY. 219
South of Rettin there is a somewhat irregular cusp (Germ., 84).
A cusp projects into Der Bodden from the southeastern point of Eiigen island
(Germ., 89).
There are several cusps inside of Frische and Kurische bars (Germ., 3 8 15,
16, 2y, 48, 49, 71, 72).
In Vejle fjord (Denm., Fredericia), tiiere are several cuspate projections, often
called " Hage " or hook, whose lorm and position indicate eddies in the tidal in
and out flow.
At the mouth of the Elbe river, west of Cuxhaven, where fortifications now
stand, is a low projecting point, a foreland of this class (Germ., 110).
Two broad, completely filled cuspate forelands occur in the Kieler and Eckern-
forder bays respectively (Germ., 58). Friedriclisort is built upon the former, while
the latter lies six kilometers east of Eckernforder.
Cebii, on an island of the same name among the Philippine islands, is built on
a point apparently of this same nature (Spain, Bol. XIII, 1886, Lam V, 1 : 100,000).
A cusp with irregular outline of " Schaaf Land " (Pasture?) is built in front of
one of the Holland dikes (Holl., 8).
Landskrona (Swe., 4) is built upon a low cuspate point, which appears to be a
foreland of this class. As there are shoals ofE the point it may be that this is
a cusp resulting from the tying on of an island which is now cut away.
Between Ilelsingborg and Raa there is a rounded foreland, an embryonic tidal
cusp (Swe., 4).
A cusp near the head of Skelder bay (Swe., 8) curves around from the usual
position at right angles to the tidal currents and points toward the mouth of the
bay. The presence of oldland may account for this, but the form of the cusp
indicates a change in direction of growth from the normal position at right angles
to the shoreline to one where a spit is growing from the cuspate point toward
the right.
Methods of Growth. — It would seem from inspection of the maps that
it was the more common thing to enclose lagoons, though in some places
the growth has evidently begim at the mainland and progressed out-
wards. In False Dungeness harbor some of these cuspate deposits are
seen which do not appear to have ever enclosed any lagoons (C. S., 646).
Three of the- cusps on the inside of the Coatue Spit, Nantucket, have no
lagoons, but as the other two have, and since they are nearer the end of
the spit and hence probably later formed, it is quite likely that the earlier
formed forelands also began with lagoons (C. S., Ill, 343 ; G. S., Nan-
tucket, Mass.).
Professor Shaler has ascribed these Coatue cusps to tidal whirlpools.
He says : '' From a superficial inspection it appears that the tidal waters
are thrown into a series of whirlpools, which excavate the shores between
these salients and accumulate the sand on the spits," *
* Bull. U. S. G. S., No. 53, 1889, 13.
220 PROCEEDINGS OF THE AMERICAN ACADEMY.
Among these filled cusps are included doubtless those which have
pa&sed through the V-bar stage as well as those which have grown by
gradual out-building, since from present knowledge it is impossible to
separate the two groups. With better maps and descriptions of the
cusps a later classification will make closer distinctions.
Theory confronted loith Fact. — After this general survey of the varying
forms of cuspate forelands selected from the many examples in the nar-
row water bodies of the world, the following generalization may be
made. However varied the form resulting from the local conditions,
tides, relief of oldland, etc., the axis, or a line drawn from the point of
the cusp through the centre of the foreland, is always at right angles to
the general direction of flow of the tidal currents.
Where there are strong tides, as in Paget sound, Chesapeake bay, and
Narragansett bay, there are numerous and typical cuspate forelands ;
while in Albemarle sound the range of tides is less than one foot, and
here few sandy points of a cuspate form occur.
Thus tlie facts of observation seem to correspond with the principal
requirement of the theory. Studies of the existing currents in regions
where these forelands are found are now needed to further test the tidal
hypothesis. From present knowledge this seems to be the best working
hypothesis.
Two methods of growth are suggested. Jn one the outline of the
foreland is early given by a V-bar, and later this enclosed lagoon is pro-
gressively filled. In the other the foreland grows by successive additions
to the mainland. The first appears to be by far the larger class ; though
the examples of the latter are liable to be confused with the filled stage
of the first class.
Between the narrow channels and the open sea there are all gradations
in size of water bodies, so we should expect to find forelands built by
combination of tidal and wind currents in different proportions. Such
cases have been referred to in Del Faro point, Aro island cusp, and
Alice point.
13. Bat-deltas.
History of a Drowned Valley. — Bay-deltas fill drowned valleys. The
term ria, from the Spanish, may be advantageously used to cover all
types of subaerially carved troughs, including von Richthofen's fjord, ria,
dalmation, and liman types.* After depression, the stream in youth
* Fiihrer, 305-312. Compare use by Pencik, Morphologie der Erdoberfliiche,
II. 562-582.
GULLIVER. — SHORELINE TOPOGRAPHY.
221
Figure 27.
build.s a delta at the head of the ria or drowned valley, in adolescence it
pushes it well forward, and in maturity it has completely driven the sea
out of the valley and thus obliterated the initial shoreline of depression.
Bay-deltas, or ria-deltas will here be grouped under the three heads,
young, adolescent, and mature.
Type Young Bay-delta : Loch Fine, Figure 27. — The type is seen in
loch Fine (Scot., 37) where Fine river has begun to fill the drowned
valley. Shira river farther
west has also begun to fill, and
later the two streams will join
their deltas, filling up the lower
portion of the ria. At the head
of this long bay the delta form
is practically free from compli-
cation due to sea action, which
in so many cases influences the
form of filling.
Other Young Bay-deltas. — Many
occur on the northwestern coast of
Spain, from which locality the gen-
eric term " ria " was taken (Stieler,
33; Spain, Bol. VII, 1880, Lam. D).
Strath Beag at the head of loch EireboU (Scot., 114).
Loin water at the head of loch Long (Scot., 38).
At the heads of loch Lomond (Scot., 38, 46); loch Awe (Scot., 45); loch
Striven (Scot., 29); loch Etive (Scot., 45) ; loch Maree (Scot., 92); loch Broom
(Scot., 92) ; loch Diiich (Scot, 72) ; etc.
Tomales bay,* California, a well marked " ria," shows delta filling at its head
(C. S., 631).
Drago delta, Austria (Austr., 25, IX). This stream is small, and has not yet
filled much of the ria, Canale di Leme.
Type Adolescent Bay-delta : Dwamish. — The Dwamish river pushing
forward its delta to fill Elliott bay is the type. Seattle, built upon the
mainland along the margin of both the delta and the unfilled bay, has a
splendid combination of elevated locations for residences, flat delta land
for future business blocks, and a water front on deep tide water. As the
delta grows forward the city will occupy it, probably accelerating its
advance, and transfer the shipping interests farther down the bay (C. S.,
651; G. S. Seattle).
Young Bay -delta : Loch Fine,
Scotland.
* See page 196.
222 PROCEEDINGS OP THE AMERICAN ACADEMY.
Other Examples. — The larger streams in the Puget sound region, notably the
Skokomish, de Cliate, Nisqually, Puyallup, and Snohomish (C. S., 6450, 6460,
690, 644).
The Tay river has filled about half of the valley which the firth occupied after
the depression (Scot., 48).
Kyle of Durness and kyle of Tongur (Scot., 114).
Euel strath (Scot, 29, 37).
Sachaig strath (Scot., 37).
Clyde delta (Scot., 38).
Carron delta (Scot., 92).
Torridon delta (Scot., 91, 92).
Scotland gives also many other examples.
Bilbao ria is partly filled (Spain, Vizcaya, 1892, Lam. 1; Spain, Bol. 111,1865,
Lam. D).
Mobile bay is being rapidly filled by waste from the Alabama rivers, although
tlie delta front is some 30 miles from the mouth of the bay (C. S., 188).
The ria-delta in Irondequoit bay is not so far advanced as the Dwamish, and is
perhaps on the border line between the periods of youth and adolescence. The
mouth of tlie bay is closed by a bay-bar, so the stream will be able to push
forward its delta undisturbed by tlie waveS of lake Ontario.
Halkjaer (Deiim., Nibe).
Kanders (Denm., Mariager).
Vejle is built on the delta growing forward into Vejle fjord (l)enm., Fredericia).
The stream emptying into Kands fjord (Denm., Fredericia).
Several streams entering into Odense fjord (Denm., Ilindsholm, Nyborg).
Quieto and Dragogna rivers, Austria (Austr., 24, IX).
Arsa delta, Austria (Austr., 25, X).
Bado delta, Austria (Austr., 26, X).
In St. Jordals fjord (Nor., 47, C).
Ler river (Nor., 15, C).
The Dniester (Rus., 33).
Tlie Gedis delta in Asia Minor (Brit. Ad., 1523, Map II, Dr. C. Cold, Kusten-
veranderungen im Archipel, Miinchen, 1886) has grown, even in historical times,
into the gulf of Smyrna, and will soon geographically cut in two the harbor of
Smyrna and leave that city without communication with the ^gean sea.
The above mentioned catastrophe has happened to Ileraclea and other places
south of Smyrna on the coast of Asia Minor, where the Ma;ander has nearly filled
its ria (Brit. Ad., 1555; Cold, loc. cit.. Map III).
Type Mature Bay-delta : Simeto, Figure 28. — South of the volcanic
mass of Etua, the three rivers Simeto, Dittaiuo, and Gorna Lunga have
built a common flood plain and delta of recent alluvium. This is a
beautifully mapped illustration of the ideal form of a delta-filled bay
(Ital. and Sicily, 269, 270, 273, 274). The filling may have been
contemporaneous with a slow sinking of the region, or the space now
occupied by the flood plain may be simply what was not filled with the
lava. The form is so typical however that it is given as the best mapped
GULLIVER. — SHORELINE TOPOGRAPHY.
223
example of a mature bay-delta, although there is some doubt as to the
early stages of this example.
Other Examples. — In Cliehalis river, Wasliington, the delta has filled not only
the drowned valley, but has also considerably filled North and South bays between
the nipped coast and the bar (C. S., 643).
A similar case is WiUapa river, Washington (C. S., 681", 642, and 6180).
Solkjoer river ha.s completely filled its ria (Denm., Skamliiigs Banke), except
for a small pond. The deflection of the mouth is to the left, indicating a current
from the right.
/v
Figure 28. Mature Bay-delta : Simeto. Sicily.
Taubaek (Denm., Nibe).
Kastbjerg (Denm., Mariager).
Narenta river has nearly filled the drowned portion of its valley (Austr. 33,
XVII).
The Guadalhoree river has filled its ria; also the Ve'Iex (Spain, Bol. XVII,
1991, Lam. A ; XVIII, 1892, Lam. A).
Lake Bay-deltas. — Lakes whose basins are portions of river valleys
frequently show, at the end where the stream enters, deltas of similar form
to those mentioned above. These deltas show almost complete river
intention, since ocean currents and tides do not affect them, and the
narrow water bodies do not permit the winds to stir up very destructive
waves.
The delta of the Ticino river in lake Maggiore, Italy, is a typical example
(Pet. Geog. Mitt., Erg. XIV., Nr. 65, 1881, Taf. III.).
See also the delta of the Rhone in lake Geneva (Carta Geologica d' Italia,
Roma, 1889) ; that of the Aare in lake Neuchatel (Swiss, VIII, and Carta dell'
Italia Superiore, by 11. Lcnzinger, Ziirich) ; etc.
224 PROCEEDINGS OF THE AMERICAN ACADEMY.
Bonneville Bay-deltas. — Almost all the streams entered lake Bouue-
ville through momitain gorges, aud the detritus of the Bonneville epoch
was deposited iu these narrow estuaries forming bay-deltas, the water
at the Bonneville level having entered the previously eroded valleys.
When the water fell the detritus was carried away, leaving no deposit
to mark the Bonneville shoreline in these larger streams.*
14. Deltas.
Ratio between Sea and River Activities. — The shore changes caused
by delta growth depend on the ratio between sea and river activity more
than upon any other factor ; and deltas would therefore be typically
developed either after uplift or depression. The area of a delta is a
measure of time since the beginning of a cycle. A large river will soon
build a great delta, a precocious infant ; but the delta will attain its
maximum area at some period in late adolescence or maturity, after which
the delta will diminish in area by the degrading of the river. A smaller
stream has a similar maximum area, though its dimensions at any given
stage are always less.
The life of a river is in a sense to be considered apart from the cycles
of coastal plain development and also as distinct from the other shore
changes, though its life is intimately connected with and a most important
part of both sets of processes. The river's aim is to convey the load
given it by the laud to the sea. Of itself it would build forward a lobe
for each distributary, the shifting of these distributaries on account of the
upbuilding causing in time a broad fan-shaped deposit, so well shown in
the confluent delta of , the Hoang and Yangtze rivers.
The sea, on the other hand, desires a straight shoreline. The delta
intention is opposed to its attack upon the land, and therefore the sea
aims to cut off the front of the lobes and carry the delta waste out to
sea, depositing it beyond wave attack or below wave-base.
The form of a delta front does not indicate sharply the stage of the
cycle in which a given re<rion stands. The relative strength of sea and
river may cause a given form of delta front at many stages in the life
history of a region. The river activity is increasing from birth toward
maturity, so that in the case of any given river there will come a time of
maximum activity, when the river will be best able to push forward and
build a lobate delta into the sea. This time, however, may not be the
time when there is the greatest likelihood of the formation of such a
* Gilbert, Lake Bonneville, 154.
GULLIVER. — SHORELINE TOPOGRAPHY. 225
lobate delta, for the activity of the sea is also a variable, and it may
happen that the ratio between the two activities is more strongly in
favor of the river at some time before or at some time after its period of
maximum activity. For example, if the time of maximum activity ot
the sea on a given shore occurs at a later stage than the time of the
maximum activity of a certain river, the largest ratio in favor of the
river will probably occur considerably before its maximum ; while if the
sea's maximum occurs at an earlier stage of shore development, and is
decreasing at a more rapid rate than the river's activity, the ratio in
favor of the river will be greater after its maximum is passed.
It may be that the sea action is so strong off any river's mouth that
the river never is able to carry out its intention. Indeed, this seems to
be the case in a large proportion of the rivers of the world. The sea is
relatively stronger than the river in all cases except where the volume
of the river is exceptionally great, as in the Mississippi, or where the
mouth is protected from the stronger sea attack, as is the case of the Po
at the head of the comparatively narrow Adriatic.
Delta Stages. — In the initial stage deltas do not exist. At any time
after the beginning of a cycle, a delta may be built, whose size will
depend far more on the volume and drainage area of the river than upon
the time since uplift.
In the cycle following uplift a delta of a certain frontal outline may
occur at various stages, and the forms appropriate to successive stages
have not been worked out, because of the many complications of tlie
problem, some of whose factors are indeterminate. Deltas doubtless
follow a normal succession of forms under the various conditions. This
has been shown to be true in the case of bay-deltas. A delta foreland
of any considerable size would not be found projecting from an initial
coast, where the valleys of all the larger river systems had been sub-
merged. Until maturity of shore development has been reached, large
delta forelands would consequently not be expected upon depressed
regions.
Credner. — Dr. Credner's monograph on Deltas* is today, nearly
twenty years after its publication, the most complete source of informa-
tion about the deltas of the world. While his descriptive portions are
classics, the theoretical conclusions of the text seem open to question.
Dr. Credner apparently looked at deltas as phenomena requiring some
common cause which would account for their presence or absence. He
* Die Deltas, Pet. Geog. Mitt., Erg. XII., Nr. 56, 1878.
VOL. XXXIT. — 15
226 PROCEEDINGS OF THE AMERICAN ACADEMY.
found that deltas were not due necessarily to a large amount of sedi-
ment ; that they are not explained by the greater or less velocity of
current ; that their presence is not determined by a deep or shallow sea
in front of the mouth of the river ; that they are not explained by the
presence or absence of an offshore bar ; that they occur even where tides
"are strong ; that " the presence of a controlling ocean current (Tiber,
Rhone) does not alone suffice to prevent the formation of deltas " ; and,
finally, that deltas are not prevented by the wind. He then goes on
to show that delta growth is aided by slow elevation and hindered or
prevented by gradual depression. He concludes that relative elevation of
the land with respect to the water is the controlling cause of delta growth.
Such slow elevation of the land is surely an aid to delta extension, but
it is only one of the factors which work together in the determination of
delta growth ; and cannot be considered necessary, in the opinion of the
present writer, for the aggradation of a coast line by rivers.
Further Study. — The subject of deltas offers a very attractive field for
investigation. The writer has not been able to make out nearly as much
as he had expected to in regard to the stages of development shown by
deltas. He is convinced that each delta goes through appropriate stages,
but the variables are so many, and vary between such wide limits, that
the laws of development are not clearly seen. Vigorous advancing
deltas are characteristic of maturity, following both uplift and depres-
sion. But the maximum of delta growth may be either before or after
this period, as shown above. Bay-deltas have been separated from the
rest, as they show characteristic stages following depression. When
better understood, other deltas will fall into their appropriate stages.
Classification. — Deltas are here classified, not according to stage in
cycle, but according to ratio of activity between river and sea. The
examples of shore development by delta growth are arranged in the
following series.
1. Lobate deltas : (a) unilobate ; (b) multilobate.
These show the river intention successful.
2. Cuspate-lobate deltas.
The river intention is in these deltas predominant, but the sea action
prevents typical lobes.
3. Cuspate deltas.
The river mouths at the point of intersection of two shore curves,
concave seaward.
4. Rounded deltas.
The shore currents prevent the cuspate extension.
GULLIVER. — SHORELINE TOPOGRAPHY. 227
' 5. Stunted deltas.
The stream iu this case is able to alter but slightly the shore curve.
6. Blocked streams.
The sea here builds a bar closing the mouth of the river. This ratio
was recognized by Dana,* who used the term blocked.
1. Lobate Deltis. (a) Unilobate type. — In tliis class are those deltas in which a
single lobe is formed by a single stream, showing pure river intention. No very
typical example has been found.
The delta of the Masander (Brit. Ad., 1555 ; Dr. C. Cold, Kiisten veranderungen
im Archipel, Miinchen, 1886, Map III) has now a typical unilobate front, project-
ing from the nearly straight front of the almost tilled ria.
The Ebro is of the general form of the unilobate delta, slightly modified toward
a muitilobate type (Credner, loc. cit., 17 ; Spain, Bol. XVI, 1889, Lam. A).
(b) Muitilobate Tijpe. — The type is the Mississippi. This classic example is
so well known, and has been so frequently given to illustrate successful river
intention, that a description is here unnecessary. a
The Kilia or northern distributary of the Danube (Rus., A; Credner, loc. cit.,
23).
The Volga (Rus., 114 ; Credner, loc. cit., 16 ; Pet. Geog. Mitt., IV., 1858, Tom. V.).
The Atrato (Credner, loc. cit., 5).
The Po (Stieler, 23) shows the river intention dominant, with partial cutting off
of the more exposed lobes.
The Gedis delta in the gulf of Smyrna (Brit. Ad., 1523; Cold, loc. cit., Map II ;
Credner, loc. cit., 11).
Tiie Rhone (Fr., 233, 234, 235, 246, 247) has a form midway between the Mis-
sissippi and the Tagliamento.
2. Cuspate.-lobctte Deltas. — The type is the Tagliamento at the head of the Adri-
atic (Austr., 22, VIII, IX ; 23, VIII, IX). The river intention is plainly seen in
the form of the delta front, but the alongshore currents prevent the formation of
typical lobes. Former positions of the delta front are indicated by the lines of
villages on the higher ground. Three of these lines are seen west of Palmanova
(Austr., 22, VIII). On the earliest shoreline are situated Gonars, Castions, Flam-
bro, Rivolto, Codroipo, and intermediate places ; on a later and less clearly marked
line are Castello, Paradiso, Torsa, Ariis, and Rivignano; while a third stage in the
growth of the delta is indicated by the road connecting St. Giorgio with Latisana.
These lines are more or less perfect divides across whicli but few streams cut, and
these few gather the many small watercourses from the areas which represent the
filled lagoons. The transition from lagoon to marsh, to wet meadow, and finally
to a rich lowland plain capable of high cultivation, is beautifully shown in going
inland from the Adriatic east of the Tagliamento.
The Danube delta (Rus., A) shows many former channels with recent changes.
That the river has been pushing forward is also indicated by the marshy, reed-
covered surface, but thinly populated. t
Frnser river, British Columbia (H. O., 961).
* Manual, 3d ed., 1880, 683.
+ Draghicenu, Jahrb. k. k. gcol Rcichs., 1890, XL. 409.
228
PROCEEDINGS OP THE AMERICAN ACADEMY.
Holland (Atlas Univ., 26) is in large part a great confluent delta formed largely
by the Rhine and adjacent streams.
The Nile is a raulti-cuspate-lobate delta (Brit. Ad. 2673, 2630; Credner, loc.
cit., 2).
3. Cuspate Deltas* — The typical example of a cuspate delta is given in Figure 29.
The two gently swinging shore curves, concave seawards, with their dune-lined
beaches, are the work of the sea.
At the point of intersection of these
two curves the river empties. The
form of the land siiows that, if it
were not for the river, there would
not be any cusp here, as there is
no projecting point in the oldland
to cause eddies in the currents.
The evidence from tiie turning of
the mouths of the small streams
both to right and left indicates
that the direction of current motion
alongshore is probably sometimes
in one direction, and sometimes in
the reverse. The smaller streams
on each side of the river's mouth
are deflected away from the point
of the cusp, indicating that the
delta mass divided an onshore
current and turned it to the right
and left, carrying the river sediment
from the river along the shore.
Farther from the river, both on
the right and left sides, there are
streams deflected toward the mouth
of the main stream. There is here
evidently no dominant movement
in either direction along the shore.
A former stage of the delta is
indicated by the ridge of geologi-
cally older material, which is repre-
sented in the figure by the broken
line. This earlier stage of the delta
front is seen to have a rounded
— outline. This suggests that for-
Tiber, merly there was a dominant move-
ment alongshore. Back of this
former shoreline are seen areas of
marsh, filled lagoons, or lowland behind the old beach. Since this leap from
some still earlier position of the shoreline, the forward growth seems to have been
Kilometers
Figure 29.
Typical Cuspate Delta
Italy.
* For fuller discussion, see Bull. G. S. A., 1896, VII. 417-421.
GULLIVER. — SHORELINE TOPOGRAPHY. 229
gradual, for no long slashes of swamp are shown. This type is the Tiber (Ital.,
149; Carta Geologica della Campagna Romana, Roma, 1888).
The Angitola delta (Ital., 241) extends a small cuspate point beyond the curve
of the bay-bar, as if the stream crossing the bar was relatively strong enough to
divide the alongshore current. It has been found impossible to pick out from the
other examples of cuspate deltas given below, any which were later stages of this
embryonic type. The maps give little more than the form of the latest stage of
development. Tiie forms should be studied on the ground, in order to see what
was the embryonic condition. This study is analogous with what is done by the
paleontologist when he peels ofE the outer shell of an Ammonite in order to dis-
cover its embryonic form.
In both the Biferno (Ital., 155) and the Ofanto rivers (Ital, 165) the deflections
indicate a current from the right at present, though formerly the deflection was in
the opposite direction, according to the indications from inland form.
In the two following examples of deltas, Volturno (Ital., 171, 172, 184) and Om-
brone (Ita)., 127, 128, 135), the streams are deflected in both directions, thus indi-
cating no dominant current alongshore.
The current is probably from the left in front of Alento delta (Ital., 141) and
from the right at Neto delta (Ital., 238).
Diina river (Hus., 13) has a cuspate foreland projecting into the Gulf of Riga.
The deflection of the Aa river to the left indicates a strong current from the right
at the head of this gulf.
The Aa de Livonie, east of the Diina, also has a cuspate outline (Rus., 13).
Punta Arenas, a Chilean settlement. South America, is built on a foreland made
by combined action of river and sea (H. O., 450a). Deflection is to the left.
A variation from the typical form is seen in the hooked j)oint of A usable delta
in Lake Champlain (C. S., 554 ; G. S., Plattsburg, N. Y.).
In the Voistrap at Aaeby (Denn)., Frederikshavn) the southward deflection of
the mouth indicates a prevailing current from the right.
The Danzig mouth of the Vistula (Germ., 70) shows deflection to the right.
Kolberg is built on the cuspate delta of the Persante (Germ., 98). The evidence
along this coast is for a current from the right.
Many of the discharge sluices emptying into the Zuyder Zee have built cuspate
deltas, and though aided by artificial means, the form is so typically cuspate that
they are included in this category (Holl., 15, 16, 21, 26, 27, 32).
4. Rounded Deltas. — The type of this class of deltas is that of the Arno (Ital.,
104, 111). The delta front is bounded by a curve convex seaward, changing into
the shore curves concave seaward. In contrast with the case of the Tiber the river
is here not strong enough to separate the alongshore current into two eddies. The
current swings around the delta, and gives it a smoothed outline in place of
a cuspate.
Fortore delta, Italy (Ital., 155). This sheet shows a portion of the former
channel, fiume morto, on the left wing of the delta.
The Sinni, Agri, and Basento deltas (Ital., 212, 201). Dominant current from
left.
The Sele delta, Italy (Ital., 197, 198). Dominant current from right.
Sangro and Trigno deltas, Italy (Ital., 148). Currents uncertain.
Vomano, Saline, and Pescara deltas, Italy (Ital., 141). Dominant current from
right.
230 PROCEEDINGS OP THE AMERICAN ACADEMY.
Savuto delta, Italy (Ital., 236).
Crati delta, Italy (Ital., 222). Current from left.
Trionto delta, Italy (Ital., 230). Currents in both directions.
The Fiunienica delta, Italy (Ital., 231). Mouth deflected to right.
The Esk delta upon which Musselburgh is built (Scot., 32) shows no dominant
current.
The deflection of over 2,000 meters of the ditch-like stream near Lyngsaa
suggests current from rrght (Denm., Frederiksliavn).
The northward deflection of the mouth of the Liver river with its rounded delta,
indicates a current from the right (Denm., Hirshals).
Maratlion delta, Greece (Attica, XVIII). Current from the left.
Rega (Germ , 92) ; VVipper (Germ., 66) ; and Stolpe (Germ., 44). All three in-
dicate a shghtly dominant current from right.
Rion delt.a, at the eastern end of the Black sea (Rus., 80; Stieler, 49). Streams
are deflected to the right and to the left from the mouth of the river.
The Rio Grande (C. S , 212) has filled a considerable portion of the lagoon, en-
closed fartlier north by Padre island or the great Texan offshore bar, and is now
advancing in front of the bar at Bagdad, having in recent geographic time aban-
doned the distributary running toward Boca Chica inlet.
The Llobregat (Spain, Barcelona).
5. Slitnted Deltas. — In a fifth class of deltas the relative strength of the stream
is so much less than that of the sea that tiie shoreline curves around the front,
making almost no change in its curvature at the mouth of the river.
An example is seen in Figure 30, where the Cavone empties into the Adriatic
(Ital., 212), making a very slight convexity. The dominant current in the gulf of
Taranto is from the left.
The Soltane makes but faint projection into the gulf of Tunis (Tunis, XXI).
Easterly current.
The Simeto delta, Sicily (Ital. and Sicily, 270, 274). The deflection of the
mouth is toward the right. The bottom cutting of the sea is judged to be greater
than its alongshore action, because the beach abuts nearly at right angles against
the lava at Catania and Agnone (Fig. 28).
The Simeri river is cut off without any delta growth (Ital., 242). Deflection is
about equal in both directions.
Acate delta, Sicily (Ital. and Sicily, 272, 275). Dominant current from the right.
The delta of Garigliano river, Italy, makes but a faint projection against the
sea (Ital., 171). Deflection to right.
Lama d'Arco delta (Ital., 201) and Lato and Lenna deltas (Ital., 202) on the
gulf of Taranto, Italy. Dominant current is from the left.
F. Alento, Italy, is not allowed by the sea to build forward its delta (Ital., 209).
Current from left, as deflection is to right.
F. Oliva, Italy (Ital., 2.36).
Several streams on the Catanzaro sheet (Ital., 242). Currents about equally
from right and left.
The ditch-like stream 4,000 meters north of Saeby enters the sea with no deflec-
tion and producing no apparent alteration in the shore curve ; a similar stream six
kilometers south of Saeby causes as little alteration in the shore curve, but it ia
deflected 800 meters to the south, which indicates a current from the right ; a third
GULLIVER. — SHORELINE TOPOGRAPHY.
231
variety of stunted delta shown on tlie same sheet is that of another ditch-like
stream, one kilometer north of Vorsaa, which produces slight alteration in the
shore curve and is not deflected, but the land on the north of the stream decidedly
offsets that upon the south, indicating again a current from the right (Denm.,
Frederikshavn).
Kolkjaer and Brendelsig give no indication as to direction of current (Denm.,
Aalborg).
Haslevgd is deflected to the south, suggesting a current from right ^Denm.,
Mariager).
Kjul, Ugerby, and Tversted deltas suggest current from right (Denm., Hirshals).
Guadalaviar 6 Turia (Spain, Valencia).
Si-ala cliiioinetrira di I a lOO.OOa
Figure 30. Typical Stunted Delta: Cavone, Italy.
The Tinto river is strongly deflected to the right and shows very little projection
of a delta (Spain, Huelva; Stieler, 35).
The Tet delta (Fr., 255).
6. Blocked Streams. — If the sea action is relatively stronger than in the last
case it may close up completely the mouth of the rivers (Fig. 31). The water
from the typical ponded streams on the Oceanside sheet, California, reaches the
sea by percolation through the beach.
The stream intention is often completely blocked by the sea when the water is
carried in a deflected course far to the right or left, and such examples have been
included in this category.
232
PROCEEDINGS OF THE AMERICAN ACADEMY.
An example of a blocked stream of a different type from the Oceanside case is
seen in the Vistula (Germ., 70, 71, 99, 100). The sea here blocks the course of the
river and carries out its intention of a concave shoreline. The main work of the
river at present seems to be filling up Frische lagoon into which the main distribu-
taries have been turned, since the aggradation of the drowned valley.
4
t Kilometers.
Figure 31. Blocked Stream : Oceanside, California.
The delta of El Rincon river on the north side of the ridge south of which the
stream flows is an anomalous feature, until one perceives that in the contest be-
tween river intention and sea intention the latter has been victorious. A current
from the left flowing up the gulf of Dulce is indicated, such current having forced
the stream to turn a right angle around the point (H. O., 1036).
Sea intention versus river intention is prettily shown at Coos bay, Oregon (C.
GULLIVER. — SHORELINE TOPOGRAPHY. 233
S., 5984). The spit formed by the southward flowing current has crowded the
river as far south as possible, close to the cliffs on Coos head.
Another good example of a river forced against a rocky headland by the sea
is seen in Garcia river at point Arena (C. S., 661).
Bezirk river is lost in the sand dunes before reaching the gulf of Tunis (Tunis,
XXI). Current is indicated from the left.
15. Tidal Scour.
Action in Bays. — Tides as abradiug agents are most effective in
drowned valleys. The destructive effects of the bore have been much
discussed, and more work has been ascribed to tliis inrush of tidal waters
than that for which it is probably responsible. The depth to which tides
may scour a submarine channel is still a very problematical question, and
the amount of wearing of the shores by the tides is a subject needing
study.*
Tides are not here taken up at any length, because the relation of
their products to stage of cycle is not as yet shown. The forms of shores
as determined by the changing ratios between tidal on- and offshore and
alongshore currents is the only point upon which emphasis is here laid.
As in the consideration of deltas, this point is dwelt upon because it
shows so clearly the importance of perceiving tlie varying ratios between
the several factors that determine shore forms.
Runways. — On Hat coasts where there are broad surfaces covered at
high and bare at low tide or wide stretches of tidal marsh, there is op-
portunity for much tidal work. When the main body of ocean water
retreats during ebb tide, that portion lying upon the flats must flow off
down the easiest path, and thus runways are formed dissecting the sur-
face. Such runways may be broad or narrow, deep or shallow, short or
long, etc., according to the values of the varying factors which obtain in
any given case. The scouring action may continue below low tide level,
to greater depths, the greater the range of tide and volume of water
passing through the runways.
The tidal scour if strong would tend to prevent the tying of islands
and closing of bays, which are normal features of shorelines in an adoles-
cent stage of development.
Such prevention of island-tying is seen on the Schleswig coast, where
from the stage of development indicated by the long wings on Sylt
island (page 213), the development of tombolos would be expected. It
* For the discussion of the scouring of tides in estuaries, see papers by the fol-
lowing authors : Bache, Branner, Dana, Ferrel, Mitchell, Shelford, and Sollas.
234 PROCEEDINGS OF THE AMERICAN ACADEMY.
seems therefore justifiable to infer that iu this locality strong tidal runs
have prevented the growth of tombolos.
The form of the tidal runway is of the indefinite type of channels,
called inseqaent by Professor Davis. A broad tidal flat will be cut by
runways forming a dendritic drainage pattern (Germ., 5, 36, 37, 79, 80).
Where there are large rivers, the pattern of dissection will show the con-
trolling influence of the master stream (Germ., 109, 110, 111).
Ratio betiveen Tides and Currents. — The relative strength of tidal on-
and offshore action and alongshore current action is a most important
consideration in the determination of the form of coasts. The form of
the North Carolina coast indicates that back-set eddies * are relatively
stronger than the tides. The prevailing extension of water bodies is
along the shore. These alongshore channels or lagoons are connected
with the ocean by tidal inlets which cross the bars, and whose position
is constantly shifted by alongshore transportation. These inlets repre-
sent the weaker tidal intention working at right angles to the general
shoreline, while the lagoons indicate the stronger alongshore action.
Something of the insequent pattern is seen in the ramifying channels
inside the offshore bar at Bogue and New River inlets, North Carolina
(C. S., 148). Southward along the Carolina coast the alongshore action
would appear from the shore forms to diminish in strength in relation to
the tidal in- and outflow.
Series of Forms. — It is possible to arrange shores in a progressive
series according to the ratio between tidal on- and offshore and along-
shore currents. This series is not one following stages of development,
but one which is determined by the ratio between two variables, whose
average directions of activity are at right angles to each other.
The normal development of shorelines as affected b}^ the sea should
however be kept in mind, and allowance made in each example for the
stage of development indicated.
When the ratio is in favor of the alongshore current, the forms
developed should show extension in the general direction of the shore-
line ; when on the other hand the ratio is in favor of the tides, the most
pronounced shore feature should be development at right angles to the
shoreline.
Western Florida Type. — On the western shores of Florida, although
the tides are weak, the ratio between tide and current action is inferred
to be preponderantly in favor of the tidal, as indicated by the form of
* See page 180, and Bull. G. S. A., 1896, VII. 405.
GULLIVER. — SHORELINE TOPOGRAPHY. 235
the shore (C. S., 180, 181). These two sheets show almost no iudica-
tion of alongshore work. The shoreline is minutely irregular from the
dissections of the tidal runways. The bottom is very shallow, the three-
fathom line extending on an average eight miles from the shoreline.
The average rise of the tide is here 2.5 feet.
The runways off this coast are not so deep nor so markedly dendritic
as in the succeeding case, but the stream pattern is very irregular.
Schleswig-Holstein Type. — On the west coast of the Schleswig-Holstein
peninsula (Germ., 5, 11, 20, 21, 35, 36, 37, 55, 56, 79, 80, 109, 110,
111) occurs an example of marked tidal scour. The west coast of the
Schleswig-Holstein peninsula from the mouth of the Elbe to the Danish
boundary is low and flat, with many outlying islands of the same char-
acter. The spaces between islands and mainland are occupied by broad
flats, bare at low tide, with steep-sided channels dissecting them. Some
of these channels are continuations of existing valleys on the mainland,
and were possibly cut when the land stood higher. Others however
head upon the flats, and appear to be runways cut by the tide. The
volume of water covering the broad flats at high tide must have con-
siderable scouring power when confined in these narrow channels, and it
has probably cut many new channels and deepened previously existing
inequalities.
Offsets, overlaps, and stream deflections indicate a dominant current
from the right in this region, whose existence is proved by observation.*
Generalizations need to be followed by detailed study and observation
of localities. Wherever possible to include facts of local observation, it
has here been done, but of many localities there are no descriptions. In
the present case it is possible to compare the rate of alongshore current
and the range of tides.
The resultant for the year 1880-81 of the northward flowing current
along the west coast of Jutland was eighteen nautical miles in twenty-four
hours or 0.75 mile per hour.f The rate of flow is probably not so
great along the less exposed coast immediately north of the Elbe river.
The range of the tides off this west coast of the Schleswig-Holstein
peninsula is from 2.75 meters to 3.50 meters (Germ., 20).
The volume of water which flows off these flats must be large on
account of the breadth of the area flooded at high tide. The form
plainly indicates that with the above ratio between alongshore currents
* H. Mohn, The North Ocean, p. 166, PI. XLIII.
t Loc. cit., p. 168.
236 PROCEEDINGS OF THE AMERICAN ACADEMY.
ami the tides, and the volume of tidal waters, the tidal in and out flow
largely determines the forms developed.
The stage of development of this shoreline is one which shows char-
acteristic features of adolescence. The projecting headland of Sylt
island is graded and has characteristic wings on right and left. The sea
is now carrying this shoreline landward. The islands of Fnno, Rom,
and Amrum were at an earlier period graded, as is indicated by their
form, and are now locally aggrading, as is indicated by the outlying sand
banks built in front of probable former shorelines. This geographic
interpretation is found to accord with the geology as given by Dr.
L. Meyn.*
The relation of such marked tidal scour to the adolescent stage of
development of this coast is not clear. The sequence of tidal forms
during successive stages of cycles is a subject needing much furtlier
study.
Georgia-South Carolina Type. — In the case of portions of the Georgia-
South Carolina coast (C S., Iu2, 153, 154, 155, 156) the ratio is less in
favor of the tides, although they are still the controlling factor in the
development of coastal forms. The shore curves are not continuous for
long distances, nor are the offsets arranged systematically. Many tidal
channels interrupt the sea beaches. These tidal runways are prevail-
ingly at right angles to the general direction of the shoreline, but there
are many connecting channels which run alongshore roughly parallel to
the beaches.
With the exception of Bulls bay (C. S., 153), where there are bro;id
flats covered at high and bare at low tide, the runways drain great areas
of salt marsh. The detritus brought down by streams from tlie land, the
sands blown by the wind, and the growth of swamp vegetation, as well
as the action of salt water upon mud-laden waters, all tend to convert
flats into solid land.f The tidal scour is opposed to this filling, and
carries off what it can by its runways into deeper water, to be finally
built into the continental delta.
The control of tidal runways b}' large rivers is shown in this region by
the Savannah, Broad, and Winyah. Tidal channels are turned toward
the stream current in some places, while in others the river fills up the
runways with detritus causing the water to flow away from the river, in
* Geologische-Uebersichtskarte der Provinz Schleswig-Holstein, 1 : 300,000,
Berlin, 1881.
t See discussion by Prof. Shaler, Gtli Ann. Rep. U. S. G. S., 1884-85, 360,
361 ; and 10th Ann. Rep. U. S. G. S., 1888-89, 2G1-2G1.
GULLIVER. — SHORELINE TOPOGRAPHY. 237
the same manner that a river higher up in its course causes streams to
flow away from the main channel down the slope of the alluvial plain.
The gathering is seen in the case of the Broad, and the filling up in the
Savannah (C. S., 155).
These tidal runways, which open to the sea in a direction away from
the main river, are often no doubt former distributaries of the river, at
present ke[)t open by the tides. The river delta phenomena merge into
the tidal very intimately in this region, and features are due usually
to more than one cause. River, tidal, and current action are here
blended, with the indications that the tides are the dominant factor in
tlie determination of the shore forms.
The average height of tides at cape Romaiu is five feet, while the
highest observed tides in this area rise from eight to nine feet.
An example of a ratio similar to the South Carolina type occurs on
the north coast of Holland (Holl., 2, 3, 4, 5, 6, 9 ; Atlas Univ., 33).
Southern New Jersey Type. — Northward from Winyah bay the ratio
is in favor of alongshore action. Areas showing some tidal in and out
flow, controlled by alongshore action, are seen on the following sheets
(C. S., 152, 149, 148, 123).
When the lagoons are nearly filled, as in southern New Jersey, the
longitudinal feature of the shore is not so marked. But in this case,
the time element must be considered, a stage later than that in North
Carolina probably exists, and the ratio between tide and current is not
necessarily changed.
North Carolina Type. — The next ratio taken to illustrate this pro-
gressive series is where the youthful shoreline shows continuous offshore
bars, broken only at intervals of several miles by tidal inlets. Examples
are seen on the following sheets (C. S., 122, 138, 145, 14G, 147, 150).
The tides in the region of Ilatteras rise from one to three feet, while
the rate of the currents is various, being much affected by storms.
Texas Type. — A final example in this series may be taken from the
gulf of Mexico, on tlie opposite side from which the first was taken.
On the east side of the gulf the ratio seemed to be in favor of pure
tidal action ; on the west, however, the alongshore action a[)pears to be
practically uninterrupted by the tides, and has determined the form of
the Texas bar, continuous for 102 miles (C. S., 210, 211, 212).
"The tide is almost always less than a foot, and its time is very
variable and uncertain. Storm tides are the only important ones."
(C. S., 211.)
238 PROCEEDINGS OF THE AMERICAN ACADEMY.
16. Cliffs.
Nip. — A very characteristic feature of tlie early stages following both
uplift and depression has been shown to be the first cut, or uip, made in
the initial coast, before the formation of an offshore bar succeeding eleva-
tion or foreland succeeding depression, the presence of either of which
protects the coast for the time from further attack.
Examples of Nips. — Back of an offshore bar a nip is usually observed, tliougli
the scale of many maps is too small to show so faint a cliff.
Nips are also seen in many regions which have been depressed. Drakes estero
(Fig. 21).
Brackenridge bluff and Stearns bluff are nips on the initial coast of Grays
harbor, Oregon (C. S., 6-13).
Back of Willapa bar, Washington, the irregular coast was nipped (C. S., 642
and 6185).
There is a nip north of Empire City, Oregon (C. S., 637).
In Chit^nik bay, Alaska, inside of spit (C. S., 8891).
Behind the marsh in Brown cove, Alaska (C. S., 704).
Both east and west of the delta of the Dwamish river, Washington.
Powder point, Du.xbury, Captains hill, and High cliff, Plymouth, Mass., were
nipped before the sea built Duxbury and Long beaches (0. S., SoS).
Back of the bar in San Rafael bay and behind the dunes in the filled valley of
San Francisquito bay, Lower California, nips are seen (II. 0., 638).
Todos Santos bay. Lower California (H. 0., 1046).
Behind tiie marsh on Santa Maria island, Chile (II. O., 1209).
In Frische and Kurische bays (Germ., 3, 8, 29, 30, 49, 72, 73).
Irregular Cliffs (Infancy- Youth). — .Cliffs occur along ungraded shores,
where there is no protection afforded by bars or other forelands. These
are characteristically jagged and irregular in youth, becoming more and
more gently curved as the graded shoreline of adolescence approaches.
The actual height of these cliffs upon ungraded coasts depends almost
entirely upon the character of the country submerged. The 1,000-foot
cliff of North cape, Norway, where the waves dance up and down and
accomplish but little abrasion, is young ; while the low cliflTs upon the
islands east of Stockholm are also youthful.
Caves are characteristic of this stage of development. Fingals cave
is cut by the sea in the sheets of igneous rock in the drowned western
coast of Scotland.*
Cliffs cut in the older Paleozoics of the southern uplands of Scotland, 200 to 300
feet high (Scot., 33, 34, 41).
* Geikie, Scenery of Scotland, 1887, 218, 219.
GULLIVER. — SHORELINE TOPOGRAPHY. 239
The western side of Lewis island has a very jagged outline (Scot., 98, 104).
Many fine examples occur on the west coast of Ireland. Particularly fine ones
are on Acliill head (62), southwest of the Bloody Foreland headland (9), Brandon
head (160, 171), and Bray head (182, 190).
George island, Alaska, on the west side of Granite cove (C. S., 741).
Point Colorado to cape Haro, Mexico (H. O., 640).
Sixty per cent of the cliffs on the eastern third of the island of Elba (Elba).
A portion of the Adriatic coast of Austria (Austr., 24, IX; 25, IX).
Curzola island (Austr., 34, XVI).
West coast of Brittany (Fr., 57).
Many good examples in Sweden (Swe., 11, 23, 29, 32, 37, 41, 46, 52; 18, 24, 25,
32,41,51,61).
And tiiere are numerous similar youthful cliffs elsewhere.
Minutely Irregular Shoreline. — The minute irregularity of the shore-
line of certain regions is a feature to which field study should be directed.
As a working hypothesis may be offered the idea that these irregulari-
ties are due to minor differences in resistance in beds at the coast. The
irregularity in the cases given below does not appear to be due to alter-
nations in the resistance in successive strata.
(Italy, 177, 178).
(Attica, X).
(Denm., Hilderod).
(Fr., 28, 29), irregular rocky ledges below high water.
Tlie east coast of Oland island (Swe., 17, 22, 38, 30).
Tiie east coast of Gotland (Swe., 23, 31).
South of Warberg (Swe., 18).
Gently Curved Cliffs (Adolescence). — Drakes Bay (Figure 21), Cali-
fornia, shows for about one half of its extent transportation alongshore.
Many of the cliffs retain youthful irregularities (C. S., 629).
Captains bay, Unalashka island, Alaska (C. S., 821).
Chirikof island, Alaska (C. S., 796).
Similar conditions prevail on either side of Acapulco harbor, Lower California
(H. O., 872).
Transportation is seen practically all along the shore from Chipequa point to
Ventosa bay, Mexico (H. 0., 876).
The cliffs on seventy-five per cent of the central portion and fifty per cent of the
western portion of the island of Elba (Elba).
Many portions of the Baltic coast of Germany (Germ., 84, 85, 12, 23, 24, 39, 68,
59, 60, 115, 42, 64, 28, 26, 27, 92, 93).
At many points along the east shore of the Cattegat (Swe., 4, 8, 13, 18).
Straighter C'///fs (Maturity). — When the initial shoreline following
depression is all cut back by the sea, the cliff line of the depressed
11 -
GULLIVER. — SHORELINE TOPOGRAPHY. 241
region is nearly straight, as is also the case with the mature cliff upon
an elevated region. Distinction between the two must be sought not in
the cliff, but in some other sequential feature. In a depressed region
the heads of filled bays may still give witness to the drowning. If a
coastal plain region was dissected, then depressed, and then its coast was
developed to the middle of the stage called maturity, the evidences of
the episode of depression would be lost. The northwestern coast of
France is probably an example of a region which has gone through some
such history. For otlier examples of straight coasts consult references
on pages 243 and 246.
17. Spits.
A Characteristic of Adolescence. — A spit is formed by currents carrying
waste from an attached end into open water, where the unattached point
of the spit may be shifted by varying conditions of water motion. Spits
are found characteristically tilong adolescent coasts, and may be found
wherever there is transportation alongshore, and are particularly marked
at the stage of adolescence.
Straight Spits. — Port Angeles, Wasliington, is enclosed by Ediz hook, which
shows an attempt to form a barb (C. S , 630:5, old number G4G). This may be
considered as an earlier stage of the condition seen in New Dungeness harbor
to the east.
Putziger spit (Germ., 27, 47).
A spit has grown northeast into the Zuyder Zee from the dike surrounding Urk
island (Holl., 20).
Examples of straight spits occur in the Limfjorden (Denm., Logstilr).
The type example of a broad spit is Skagen point, upon the northern extremity
of Jutland (Denm., Hirshals, Skagen). The prevailing motion of currents is indi-
cated by offsets and stream deflection to be toward the point of the spit on the
right and from the point on the left. In this example we are able to confirm the
indications given by the geographic form, as to the direction of the forming cur-
rents, by actual observations. The prevailing current along the west coast of
Jutland is from the right, the surface water having been shown to flow from the
south toward the point of Skagen spit.* The lines of growth are beautifully
shown by the curving strips of marshland ; even the artificial ditches follow the
same curves as if their location was determined by dune ridges. The direction of
growth of this spit has been toward the nearest land on the Swedish coast.
Curred Spits. — T)n\.c\\ point (Fig. 32), lake Michigan, has grown from right to
left, looking from the lake to the shore. Storms from the opposite direction have
" turned its end toward the land and the successive recurvements are clearly dis-
* H. Mohn, The North Ocean, Norwegian North Atlantic Expedition, 1876-78,
[2,1 XVIII. 168, PI. XLIII. ; Danish, Meteorologisk Aarbog, 1880, 1881.
VOL. XXXIV. — 16
242 PROCEEDINGS OP THE AMERICAN ACADEMY.
cernible near the apex. The last of these is the greatest ; and it is possible that
the spit has acquired permanently the form of a hook." *
Hooked Spits. — Cape Lookout t is cliaracterized by a spit projecting from the
point of the cusp vvhicli has a recurving barbed hook on its left side. Tlie curve
of the right side of tlie spit is continuous with tlie curve of the rigiit offshore bar,
and tliere is an offset from tlie right near the point of the cusp proper. On the
opposite side of the spit there are tliree minute offsets, also from the right to the
left. The offsets therefore indicate currents flowing in opposite directions upon
the two sides of the spit, botli moving from tlie right to the left. The form of the
barbed hook is evidence for a current moving from tiie sea toward tlie land at this
point on the left side of the cusp, because for its extension material must be carried
toward the point of the hook from some other locality, and since the hook curves
in toward the land and has a smootii contour on the outside and an irregular one
on the inside, transportation is inferred along tlie graded and not the ungraded
path. The form of the Lookout recurved spit indicates a current from the land
toward the sea on the riglit and one from tlie sea toward the land on the left of
the cusp.
Capes Obitotchnaia and Berdianska'ia in the sea of Azov (Atlas Univ., 38) are
markedly hooked even on a small scale map.
Messina spit, Sicily, is another hooked spit (Ital. and Sicil}-, 254).
On the south side of Hjarnii island (Denm., Skanderborg), and southeast of
Nyborg (Denm., Nyborg), are two otlier examples. One forms the harbor of
Marstal, aided by artificial breakwaters, another lies five thousand meters to
the northwest of the city, and a third hooked spit is about the same distance to
the east (Denm , Svendborg).
Another encloses New Dungeness harbor (C. S., 646).
Serpent Spits. — Li certain places the currents may be so variable or periodically
shifting that the spits do not grow in a straight line or simple curve, but take a
serpentine course. In many cases this current irregularity may be due to the form
of the bottom, reefs, submerged ridges, etc.
The type example of such spits is that of cape Etolin, Nunivak island, Bering
sea (C. S.. 896).
Spelmo island has a somewhat similar serpent spit growing northward toward
the mainland (Denm., Faaborg).
18. Stages of Development of Various Coasts.
Average Stage. — Taking the criteria as g-iven in this paper as a basis
of comparison, the maps of various regions have been studied and com-
pared with any available deseriptions ; and the regions have then been
classified according to the prevailing criteria shown. A few criteria in
any region may be in advance or behind the average development of
features, and such a region has been classed according to the majority
of its features. This classification is not as complete as could be desired,
* G. K. Gilbert, 5th Ann. Rep. U. S. G. S., 96, PI. IX.
t See page 180.
GULLIVER. — SHORELINE TOPOGRAPHY. 243
nor is it free from inaccuracies. The sources of information are in many
cases very meagre.
All the reasons for the classification of each example are not dis-
cussed ; and many of the most important features are omitted, because
they already have been considered. Under examples of depression is
given some idea of the various kinds of lands which were depressed, and
also some hint as to the different stages of development which had been
reached before the new cycle was inaugurated.
Uplift: Youth. — The coast of the Argentine Republic (page 162).
Texas (p. 89), and other parts of the Atlantic and Gulf plains.
Maine (C. S. and G. S. sheets), (pages 158, 185, 188).
Corsica (Fr., 261, 263, 265).
Parts of Attica (page 186).
Probably the coast of Brazil from cape Benevento to cape Frio should be here
included. The ciiarts sliow offshore bars enclosing lagoons (H. 0., 470, 471),
probably following uplift.
Also from cape Santa Marta to Tramandaky bar (H. 0., 477).
The soutliern half of Lower California, Mexico, appears on its western slope to
liave a coastal plain, described as " low hills and rising plains," with offshore bars
enclosing narrow lagoons (H. O., 621).
The western coast of Mexico, east of the gulf of California, also shows a similar
coast (H. O., 621 and 622 as far south as San Bias).
A third example in the same general region is seen on the west coast of Guate-
mala, extending a little on either hand into Mexico and San Salvador (H. O., 931,
932). The detail of bar and lagoon is shown on the chart (H. 0., 873). The lack
of definite information makes a positive statement in regard to this region unsafe.
A study of the maps indicates that they belong in this stage. Mr. J. J. Williams
says in regard to this region : " The tertiary clays, gravels, and beds of detritus
which cover up so much of the Isthmus along the line of survey, extend on tlie
north side almost to the summit-level, and the base of the hills which lie east and
west of it. Those deposits being found pretty uniformly spread, even to the depth
of thirty feet in some places, as at a point north of the sunnnit-level, and between
it and the river Almoloya, are evidences of the slow and tranquil elevation of this
portion of the Isthmus above tiie sea." *
Uplift.: Adolescence. — Southern New Jersey (page 184), and other portions of
the Atlantic coastal plain.
Uplift: il/atH;v«y. — Eastern Italy (page 186). The southeastern and southern
coast of the "toe" of Italy (Ital., 246, 247, 255, 263, 264).
The southern coast of Sicily (Ital., and Sicily, 265, 266, 271, 272, 275, 276).
The Tertiary strata are considerably deformed and therefore the shoreline is not
as straight as in the case of a more simple coastal plain. In a few places along
this line tlie development has hardly reached maturity, as for example west of
Pozzallo (276).
Northwestern France (page 241).
* The Isthmus of Tehuantepec, J. G. Barnard, New York, 1852, 149.
244 PROCEEDINGS OF THE AMERICAN ACADEMY.
Tlie west coast of Hollanfl from the northernmost outlet of tlie Rhine to Helder
at the uilet to the Zuyder Zee (Atlas Univ. 26; Holl., 14, 19, 24, 25, 30, 37). Tliis
is a porti'^n of tlie confluent delta of the Rhine and adjacent streams.
The coast of Belgium (Alias Univ., 25; Belg., 4, 5, 11, 12, 19).
Depressioii : Youth. — Southwest Ireland is a typical example of youthful shore
evolution upon a vigorous coast (Ireland, 150, 151, 160, 161, 162, 171, 172, 17o, 182,
183, 184, 190, 191, 192, 197, 198, 199, 203, 204). A region of strong relief, with
transverse trends, dissected to about early maturity, was deeply drownt'd and
exposed to the strong attack of the open sea. Far up into the bays the waves
attack the coast and the offshore currents carry away the waste from tlie jagged
cliffs. Grade is reached only in the bay-bars near the heads of the bays.
The southern coast of Curzola island (Austr., 34, XVI).
The west coast of Central America, San Juan del Sur to gulf of Nicoya (II. 0.,
1016, details in 1025-1034).
Soledad bay and Santo Tomas anchorage. Lower California (II. 0., 1044).
Port Islay, Peru (H. 0., 1183).
Brayza island, Austria (Austr., 32, XV).
Meleda island, Austria (Austr., 34, XVII).
The southernmost portion of the Austrian coast, in places becoming adolescent
(Austr., 3(i, XIX ; 37, XX).
Many portions of the coast of Greece (Attica, III, VIII, XVI, XXI, XXII, and
XXIII).
The youthful shoreline of the low coast of Saitiiolm is markedly contrasted with
the adolescent shoreline north and south of Copenliagen (Denm., Kjilbenliavn).
The eastern coast of Schleswig (Germ., 7, 13, 14, 23, 24). The development has
advanced to adolescence in many of the more exposed portions of this low coast.
The steep eastern coast of Sweden from Ilanil Bay northward to a point on the
mainland opposite the north end of Gland island (Swe., 6, 10, 11, 12, 17, 22, 29).
It is worth notice how sligiitly the glacial accidents have here modified the forms
due to drowning.
The Stockholm district (page 159).
The irregular cliffs of eastern Scotland indicate yonth'"iil shore evolution (Scot.,
57, 67, 77, 87).* The development has gone a little farther toward adolescence
near Rattray head (Scot., 97), but very jagged cliffs are seen to the west of this
head (Scot., 95, 96). The north coast of Scotland (Scot., 113, 114, 115, 116) shows
almost no transportation alongshore, althougli several bays have been partly filled.
The west coast of North and South Uist (Scot., 08, 69, 78, 79, 88, 89) shows
a nearer approach to adolescent simplification of outline than their irregular
eastern coast.
Taken as a whole the western coast of Scotland, where the sea attack is stronger
though the rocks are more resistant, is nearer adolescence than the eastern, where
the weaker attack of the North sea has not done so much work upon the less meta-
morphosed rocks. The time since the beginning of the present cycle may not have
been the same in the two areas. As the division lines have been drawn in this
scheme, this Atlantic coast of Scotland is about on the border between the two
stages. Youthful and adolescent features both occur. Two typical areas, from
■ * See Geikie, Scenery of Scotland, 2d ed., 1887, 56-59.
GULLIVER. SHORELINE TOPOGRAPHY. 245
the northwest and southwest respectively, are given (Scot., 71, 72, 81, 82, 91, 92.
100, 101 ; 19, 20, 21, 22, 27, 28, 29, 30, 35, 36, 37, 38, 43, 44, 45, 46).
The shores of Kristiania fjord (Nor., 9, A, B, C, D ; 10, A, C ; 14, B, D).
The Bergen region with structural northwest to southeast trend (Nor., 16, G,
D; 22, A, B; 23, A).
Central Norway (page 159).
Soutliern coast of Finland (Rus., 11, 25).
The eastern shore of the Cattegat northward from Warberg (Swe., 18, 24, 25,
32, 41, 51, 61).
Coast of Chile (H. 0., 445, 446, 446«, 447, 447', 38; also Piano Topografico y
Geologico de la Republica de Chile, sheet 12).
The California coast near San Francisco (C. S., 5500, 5520, 5521, 5599).
Depression : Adolescence. — The type example of adolescent sliore development
following depression is in Germany on a coast of moderate relief upon the south-
ern shore of the Baltic (Germ., 1, 3, 8, 15, 16, 25, 26, 27, 28, 29, 30, 43, 44, 45, 46,
47, 48, 49, 50, 65, 66, 70, 71, 72, 73, 92, 93). Beaches occur at the foot of the
cliffs, and the cliff lines are gently curving. The transportation of material takes
place practically all alongshore, wings have grown out from the headlands, and the
bays are nearly all enclosed by bars. Deltas occur at tlie bay-heads and are
growing forward, but the bays are not as yet filled by laud waste and sand blown
in from the bars. Upon the inside of the bars there are cuspate projections of
sand, while nips are seen upon the mainland itself. At a distance of from 3 to 10
kilometers offsliore there is a depth of from 20 to 25 meters, which seems to repre-
sent the submarine platform. Offsets, overlaps, and stream deflections are not
strong in either direction, but show a slight alongshore action toward the east or left,
thus indicating a dominant current from the right. This dominance appears to be
stronger toward the eastern side : witness the wing growing eastward from Putzig
headland, the inlet to Frische bay nearer the northeast end of the bar, and the
inlet to Kurische bay crowded way over to the northeast end close to the mainland.
One feature very typical of adolescence which is not well shown along the
northern coast of Germanj' is island-tying. There are several areas, which were
possibly isolated portions of the mainland at the beginning of the cycle, that are
now completely connected with the mainland by sea and river aggradation, but
there are no typical tombolo-tied islands,* so common elsewhere. Tlie reason for
the absence of such islands seems to be that this shoreline is one developed upon
a drowned coastal plain, not deeply dissected and somewhat masked by glacial
aggradation. The writer's interpretation of the late history of this region is that,
after the elevation of tlie Tertiary and Pleistocene strata of the North German
plain, the dissection of the land advanced to adolescence. Then followed a mod-
erate depression by which the adolescent valleys were drowned, but the land was
not sufficiently dissected to allow the formation of many islands.
Another example is the west coast of Central America, gulf of Nicoya to
Burica point (H. 0., 1016, 1017).
Blanca and Falsa bays. Lower California (H. O., 1115).
Playa Maria bay to Rosalia point, Lower California (H. O., 1118).
Bay of Avatcha and approaches, Kamchatka (H. 0., 54).
* See page 189.
246 PROCEEDINGS OF THE AMERICAN ACADEMY.
San Juan Bautista bay, Juan Fernandez island, from Salimas point to Bacalao
point (H. 0., 1267).
The western coast of Mexico from San Bias to Tehuantepec {H. 0., 622, 932,
933; details 876, 915, 936, and 938).
Tuget sound (C. S., 6450, 6460).
Transportation occurs along nearly the whole firth of Forth shore (Scot., 32,
33, 40, 41).
The northeast coast of Ireland has entered adolescence (Ireland, 8, 14, 20, 21, 29).
The area of moderate relief in the region of Dublin, from Dundruni bay to
Wicklow head (Ireland 60, 61, 70, 71, 81, 82, 92, 102, 112, 121, 130), shows con-
tinuous transportation alongshore and other typical features of adolescence.
The eastern coast of Jutland from Skagen to Horsens (Denm., Skagen, Frede-
rickshavn, Aalborg, Mariager, Stavnshoved, Skanderborg). The offsets and river
deflections indicate a prevailing current from the left. The shoreline is here much
simplified, and for the most part in long swinging curves, but the initial outlines
are still seen in the fjorded bays, therefore the shoreline is classed as adolescent
approaching maturity.
The east coast of Kiigen island (Germ., 42, 64). See also hypsometric map by
Dr. R. Credner which shows the simplification of a very irregular shoreline.*
The southeast coast of MiJen and Falster (Denm., Moensklint, Vordingborg,
Gjedser).t The offsets and stream deflection indicate a prevailing current from the
northeast.
Part of the northeast coast of Schleswig-Holstein (Germ., 39, 58, 59, 60).
Southern coast of Sweden (Swe., 1, 2, 3). Tiie streams are deflected to the left,
which indicates a dominant current from the right.
The eastern shore of the Cattegat southward from Warberg (Swe., 4, 8, 13, 18).
Portions of the coast of Greece (Attica, VIII, XI, XVII).
The southeast coast of Arabia on the map by S. B. Haines.|
The coast of Tunis (Tunis, VII, XIV, XXI, etc.).
The northwest shore of the Black sea (Rus., A, 19, 33 ; 2\tlas Univ., 38).
See page 214.
Depression : Maturity. — The west coast of Italy from Punta Rianca southward
to the land-tied island, Massoncello (Ital., 96, 104, 111, 112, 119).
The west coast of the Italian "foot" (Ital., 228, 229, 236).
The western coast of Jutland (Atlas Univ., 30; Denm., Thisted, L(>gstur, Lok-
ken, liirshals, Skagen). A prevailing current from the right is indicated by the
offsets.
» Rugen, Forsch. z. deutsch. Landes- u. Volkskimde, 1893, VII. 373-494.
t See also maps and sketches in Geologie der Insel Muen, by Dr. C. Poggaard,
Leipzig, 18-52 ; and for sketches of disturbed strala in cliffs see account by F.
Johnstrup, Deutsch. geol. Gesell. Zeit., 1874, XXVI. 533-585.
t J. R. G. S., 1845, XV. 104.
GULLIVEB. — SHORELINE TOPOGRAPHY. 247
SHORELINE REFERENCES.
A few of the most important authors for consultation in regard to coastal and
shore forms are marked with an asterisk.
Papers not seen by the writer are marked with a dagger.
Abbe, Cleveland, Jr. Remarks on the Cuspate Capes of the Carolina Coast.
B. Soc. Nat. Hist.. 1895, XXVI. 489-497.
Ackermann, Karl. (1) t Eine ausfiihrliche Morphologie des Strandsees.
(2) Beitrage zur physischen Geographie der Ostsee. Hamburg, Otto Meiss-
ner, 1883.
Andre'ws, Edmund. The North American Lakes considered as Chronometers
of Pos^glacial Time. Trans. Chi. Acad. Sci., 1870, IL 1-23.
Ansted, D. T. Channel Islands. London, 1862. Raised Beaches, 280-296.
Babbage, Charles. (1) Observations on tiie Temple of Serapis at Pozzuoli, near
Naples ; with Remarks on certain Causes whicli may produce Geological
Cycles of great Extent. Proc. Geol. Soc, 1834, II. 72-76. (2) Q. J. G. S.,
1847, III. 186-220.
Bache, A. D. New York Bay and Sandy Hook. U. S. C. G. S., Appen. 27,
1858, 197-203.
tBarrois, Ch. (1) Brittany, Finisterre Raised Beaches. Ann. Soc. Geol. du
Nord, 1877, IV. 186. (2) Raised Beaches, Sangatte, etc. Ann. Soc. Geol. du
Nord, 1880, VII. 182.
Beaumont, Elie de. Lerons de geologie pratique. Paris, 1845, I. 221-253,
Sliorelines.
tBerendt, G. Die Geologie des kurischen HafEes, u. s. w. Umgebung. Konigs-
berg, 1869.
Bertrand, Marcel. Continuite du phe'nomene de Plissement dans le Bassin de
Paris. Paris, Bull. Soc. Ge'ol. de France, 1892, XX. 118-165.
Bezzenberger, A. Die Kurische Nehrung und ihr Bewohner. Stuttgart, 1889.
Branfill, B. R. Physiographical Notes, etc., on Tanjore. Jour, of the Asiatic
Soc. of Bengal, 1878, XLVII. (2), 179-190.
Branner, John C. The " Porordca," or Bore, of the Amazon. Sci., 1884, IV.
488-492.
Brogger, W. C. Uber die Bildungsgeschichte der Kristianiafjords. Christiania,
Nyt. Mag. Nat., 1886, XXX. 1-135.
Bruckner, Ed. Uber Schwankunger der Seen und Meere. Wien, Verb. d.
deutsch. Geographentages, IX. 209-223.
Chambers, Robert. (1) Ancient Sea-margins as Memorials of Changes in the
relative Level of Sea and Land, 1847. (2) Tracings of the North of Europe,
1850. (3) Tracings in Iceland and the Faeroe Islands, 1856.
248 PROCEEDINGS OF THE AMERICAN ACADEMY.
Cold, Conrad. Kusten-veranderuiigen im Archipel. Miinchen, Knorr u. Hirtli,
1886.
*Cornaglia, M. P. Sul regime delle spiaggie et suUa regolatione del porti.
Turin, Ra. Tipografla Paravia, 1891. Review, Nature, 1892, XLV. 362.
*Credner, G. R. Die Deltas. Pet. Geog. Mitt., Erg., Nr. 56, 1878.
Dall, W. H. Bering Sea. U. S. C. G. S., Appen. 16, 1880, 297-340. References.
*Dana, James D. Geology. Wilkes U. S. Exploring Expedition. Pliiladel-
piiia, C. Siierman, 1849.
Darwin, Charles (Robert). (1) Elevation and Subsidence in the Pacific and
Indian Oceans ; Coral Formation. Proc. Geol. Soc, 1838, II. 552-554, 654-660.
(2) Geological Observations on the Volcanic Islands and Parts of South Amer-
ica, visited during the Voyage of H. M. S. " Beagle." New York, od ed., 1891.
*Davidson, George. (1) An Examination of some of the Early Voyages of Dis-
covery and Exploration on the Northwest Coast of America, from 1539 to 1603.
U. S. C. G. S., Appen. 7, 1886, 155-2-53. (2) The Submerged Valleys of the
Coast of California, U. S. A., and of Lower California, Mexico. Proc. Cal.
Acad, of Sci., 1897, I. 73-103.
Davis, Charles Henry. (1) Upon the Geological Action of the Tidal and
other Currents of the Ocean. Boston, Mem. Amer. Acad., 1849, IV. 117-156.
(2) A Scientific Account of the Inner Harbor of Boston, with a Synopsis
of the general Principles to be observed in the Improvement of Tidal
Harbors. Boston, Mem. Amer. Acad., 1855, V. 93-110.
Davis, "Wm. M. (1) Geographical Illustrations : Southern New England.
Harvard University, 1893. (2) The Outline of Cape Cod. Proc. Amer.
Acad., 1896, XXXI. 303-3.32
*De la Beche, Henry T. (1) Geological Notes. London, Trenttel & Wiirtz,
1830. (2) A Geological Manual. London, 2d ed., enlarged, 1833. (3) Re-
searches in Theoretical Geology. London, Chas. Knight, 1834. (4) The
Geological Observer. London, 2d ed., 1853.
Desor, E. Sur les deltas torrcntiels anciens et modernes. Nice, 1880.
Diller, J. S. (1) Tertiary revolution in the topography of the Pacific Coast.
14tli Ann. Rep. U. S. G. S., 1892-3, II. 397-434. (2) A Geological Reconnais-
sance in Northwestern Oregon. 17th Ann. Rep. U. S. G. S., 1895-6, 1. 441-520.
Dinse, P. Die Fjordbildungen. Berlin, Z. d. G. f. E., 1894, XXIX. 189-259.
Drew, Frederic. Alluvial and lacustrine deposits and glacial records of the
Upper-Indus Basin. Q. J. G. S., 1873, XXIX. 441-471.
Dubois, Marcel. Role des articulations littorales. Etude de Geographic com-
paro'e. Ann. de Gcog., I. 131-142.
Elias, Ney. Notes of a journey to the new course of the Yellow River, in 1868.
J. R. G. S., 1869, XL. 1-33.
Fergusson, James. Recent changes in the Delta of the Ganges. Q. J. G. S.,
186.3, XIX. 321-354.
*Fischer, Theobald, t (1) Kiistenveranderungen im Mittelnieer. Z. f. E., 1878.
(2) Zur Entwickelungsgeschichte der Kiisten. Pet. Geog. Mitt., 1885, XXXI.
409-420.
Geer, Gerard de. (1) Pleistocene changes of level in eastern North America.
Proc. B. Soc. Nat. Hist., 1892, XXV. 454-477. (2) Om Skandinaviens niva-
filrandringer under quartarperioden. Aftryck nr Geol. Foren. i Stockholm
Forhandl., 1888, X. 5, 366-379; 1890, XII. 2, 61-110.
GULLIVER, — SHORELINE TOPOGRAPHY. 249
*Geikie, A. (1) Tlie Scenery and Geology of Scotland. London, Macmillan &
Co., 1st ed., 1865, 2d ed., 1887. (2) Textbook of Geology, 3d ed., 1893,
Upheaval and Depression, 281-296. References.
Gesner, Abraham. Elevations and Depressions of the Earth in North America.
Q. J. G. S., 1861, XVIL 381-388.
*Gilbert, G. K. (1) The topographic features of Lake Shores. 5thAnn.Eep.,
U. S. G. S., 1883-4, 75-123. (2) Lake Bonneville. U. S. G. S., Mon. I., 1890.
(3) On certain Glacial and Post-glacial Phenomena of the Maumee Valley.
A. J. of S., 1871, CI. 339-345. (4) Post-glacial changes of level in the basin
of Lake Ontario. Sci. 1886, VI. 222. (5) The History of the Niagara River.
6th Ann. Rep. Com. State Res. at Niagara, 1889.
Gilbert, S. A. Coast of Texas, embracing the shores of Espiritu Santo, San
Antonio, and Aransas Bays. U. S. C. G. S., Appen. 32, 1859, 324-328.
Gordon, R. Report on the Irrawaddy. Rangoon. 4 Pts. 1879-80.
Giittner, Paul. Geographische Ilomologien an den Kiisten niit besonderer'Be-
riicksichtigung der Schwemmlandkiisten. Leipzig, Mitt. V. f. Erdk., 1894,
39-95.
Haas, H. Studien iiber die Entstehung der Fohrden (Buchten) an der Ostkiiste
Schleswig-Holsteins, sowie der Seen und des Flussnetzes dieses Landes.
Mitteil. a. d. mineral. Indt. d. Univ. Kiel, 1888, I. 13-32.
Hahn, F. G. (1) Untersuchungen iiber das Aufsteigen und Sinken der Kiisten.
Leipzig, 1879. (2) Inselstudicn. Leipzig, Veit & Co., 1883.
Haig, M. R. The Indus Delta Country. London, Kegan Paul & Co., 1894.
Hansen, Andr. M. Strandlinje-studier. Archiv for math, og Naturvidenskab,
XV. 1-96. Kristiania og Kjubenhavn.
Haupt, Lewis M. The Physical Phenomena of Harbor Entrances. Proc. Am.
Phil. Soc, 1S88, XXV. 19-41 ; 1889, XXVI. 146-171.
Hilgard, E. W. On the Geology of the Delta, and the Mud-lumps of the Passes
of the Mississippi. A. J. of S., 1871, CI. 238-246, 35G-3G8, 425-435.
Hogbom, A. G. Om sekulara liiljningen vid Vesterbottens kust. Stockholm,
Geol. Foren. Forhandl., 1887, IX. 19-25.
Humphreys and Abbot. Report on the Physics and Hydraulics of the Mis-
sissippi River. 1861.
Hunt, A. R. The formation and erosion of beaches. Nature, 1892, XLV.
415-416.
*Keller, H. Studien iiber die Gestaltung der Sandkiisten und die Anlage der
Seehafen im Sandgebiet. Berlin, Zeit. f. Bauwesen, 1881, XXXI. 189-210,
301-318, 411-422 ; 1882, XXXII. 19-36, 161-180.
Kinahan, G. H. Irish Tide Heights and Raised Beaches. Geol. Mag., 1876, III.
78-83.
Kjerulf, T. Terrassen in Norwegen. Berlin, Z. deutsch. geol. Ges., 1870, XXII.
1-14 ; Geol. Mag., 1871, VIII. 74-76.
Koons, B. F. High Terraces of the Rivers of Eastern Connecticut. A. J. of S.,
1882, XXIV. 425-428.
Krummel, Otto. Uber erosion durch Gezeitenstrome. Pet. Geog. Mitt., 1889,
XXXV. 129-138.
*Lawson, A. C. (1) Coastal Topography of the northern side of Lake Supe-
rior. Ann. Rep. Geol. & Nat. Hist. Sur. of Minn., 1891, XX. 181-289. (2) Ge-
250 PROCEEDINGS OF THE AMERICAN ACADEMY,
ology of the San Francisco Peninsula, loth Ann. Rep. U. S. G. S., 1893-94,
399-476.
Lehmann, P. (1) Pommerns Kiiste von der Divenow bis zum Darss. Breslau,
1878. (2) Das Kustengebiet Hinterpommerns. Berlin, Z. d. G. f. E., 1884,
XIX. 332-404.
Lehmann, R. Uber-ehemalige Strandlinier. Halle, 1879.
Leverett, Frank. Raised Beaches of Lake Michigan. Wis. Acad. Sci., 1889,
VII. 83-87.
Lyell, Charles. (1) Rise of Land in Sweden. Phil. Trans., 1835, CXXV.
1-38. (2) Antiquity of Man. London, John Murray, 4th ed., 1873, 322-336,
Changes of Level of England in Pleistocene Period.
Margerie, Emm. de and G. de la Noe. Les Formes du Terrain. Paris,
Imprimerie Nationale, 1888.
MaTV. Notes on the comparative structure of surfaces produced by Subaerial
and Marine Denudation. Geol. Mag., 1866, III. 439-451.
McGee, W. J. The Lafayette Formation. 12th Ann. Rep. U. S. G. S., 1890-91,
347-521.
Merrill, F. J. H. Barrier Beaches of the Atlantic Coast. Pop. Sci., 1890,
XXXVII. 736-745.
Miller, Hugh. Some Results of a Detailed Survey of the Old Coastlines near
Trondhjem, Norway. Geol. Mag., 1885, II. 518-520. Rep. Brit. A. A. S.,
Aberdeen, 1885, 1033-1035.
*Mitchell, Henry. (1) Reclnmation of Tide Lands, and its Relation to Navi-
gation. U. S. C. G. S., Appen. 5, 1869, 75-104. (2) Nauset Beach and
Monomoy Peninsula. U. S. C. G. S., Appen. 9, 1871, 134-143. (3) A Report
on Monomoy and its Shoals. U. S. C. G. S., Appen. 8, 1886, 255-261.
Mohn, H. (1) The North Ocean, its Depth, Temj)erature, and Circulation.
Norwegian North Atlantic Expedition, 1876-78, XVIII. 1-212, Cliristiania,
1887. (2) Cm Gamle Strandlinier i Norge. Nyt. Mag. Nat., 1877, XXII.
1-53.
Munthe, Heur. Preliminary Report on the Physical Geography of the Litorina
Sea. Bull. Geol. Inst. Univ. of Upsala, 1895, IL 1-38.
Patterson, George. Sable Island : Its history and phenomena. Trans. Roy.
Soc. Can. 1894, XII. 2, 1-50.
*Penck, Albrecht. Morphoiogie der Erdoberflache. Stuttgart, J. Englehorn,
1894.
Pettersen, Karl. (1) Om de i fast Berg udgravede Strandlinier. Christiania,
Arch. Math. Nat., 1878, III. 182-223. (2) The Rise and Fall of Continents.
Geol. Mag., 1879, VI. 298-304. (3) Tromsii Museums Aarshefter, III., 1880.
(4) Sitz. Acad. Wien., XCVIII., 1889. (5) Vestfjorden og Salten. Christi-
ania, Arch. Math. Nat., 1885-6, XI. 377-492.
*Philippson, Alfred. (1) tJber die Kiistenform der Insel Riigen. Sitzber.
Naturh. v. Rheinl., 1802, 63-72. (2) Uber die Typen der Kiistenformen, insbe-
sondere der Schwemmlandkiisten. Berlin, Richthofen Festschrift, 1893.
Pillsbury, J. E. Gulf Stream Explorations. U. S. C. G. S., Appen. 10, 1890,
461-620.
Prest-wick, Joseph. (1) On the Raised Beach of Sangatte. Bull. Soc. Geol.,
1880, VIII. 547. (2) Raised Beaches, etc., of the South of England. Q. J. G.
S., 1892, XL VIII. 263-343.
GULLIVER. — SHORELINE TOPOGRAPHY. 251
*Ramsay, A. C. (1) On tlie Denudation of South Wales and the Adjacent
Counties of England. Mem. Geol. Sur. of Gr. Brit., 184G, I. 297-335.
(2) The riiysical Geology and Geography of Great Britain. London, Ed.
Stanford, 5th ed., 1878.
Ramsay, A. C, and J. Geikie. Geology of Gibraltar. Q. J. G. S., 1878,
XXXIV. 505-541.
Reusch, Hans. (1) Strandfladen, et nyt track i Norges geografi. Norges
Geologiske Undersogelse, No. 14, 1892-93, 1-14. English Summary, 144-
145. (2) The Norway Coast Plain. Jour, of Geol, 1894, II. 347-349.
Richardson, Ralph. Terraces Occurring on the Banks of the Tay and its
Tributaries. Edinburgh, Trans. Geol. Soc, 1885, V. 56-70.
*Richthofen, F. F. von. Fiihrer fiir Forschungsreisende. Berlin, Robert
Oppenheim, 1886.
Rothrock, J. T. Elevated Seabeach on Grand Cayman between Cuba and
Jamaica. Am. Nat., 1801, XXV. 81(5.
Sandler, Christian. (1) StranJlinien und Terrassen. Mitteil. Ver. f. Erdk.
Leipzig, 1888, 191-201. (2) Strandlinien und Terrassen im Romsdalsfjord.
Pet. Geog. Mitt., XXXVI., 1890, Taf. 16. (3) Zur Strandlinien- und Terrassen-
Litteratur. Leipzig, Wiss. Veroffentl. d. Ver. f. Erdk., 1891, L 295-313.
Schott, Arthur. Die Kiistenbildung des nordlichen Yukatan. Pet. Geog. Mitt.,
1866, XIL 127-130.
Schott, C. A. Fluctuations in the level of Lake Champlain and average lieight
of its surface above the Sea. U. S. C. G. S., Appen. 7, 1887, 165-172.
Sexe, S. A. On Rise of Land in Scandinavia. Christiania, Index Scholaruiu
of Univ., 1872.
Sieger, Robert. Seenschwankungen und Strandverschiebungen in Skandina-
vien. Z. d. G. f. E., 1893, XXVIII. 1-106, 393-488. References.
*Shaler, N. S. (1) Recent changes of Level on the Coast of Maine. Mem. B.
Soc. Nat. Hist., 1874, II. 321-.340. (2) Seacoast swamps of the eastern
United States. 6th Ann. Rep. U. S. G. S., 1884-5, 353-398. (3) The geo-
logical history of harbors. 13th Ann. Rep. U. S. G. S., 1891-2, 93-209.
Bhelford, 'Williani. Rivers flowing into Titleless Seas — Tiber. Min. Proc.
Inst. Civ. Engin., 1885, LXXXII. 2-68.
Sokol6w, N. A. Die Diinen, u. s. w. Berlin, Julius Springer, 1894.
Sollas, W. J. The estuaries of the Severn and its tributaries. Q. J. G. S.,
1883, XXXIX. 611-626. References on Estuaries.
Spencer, J. W. (1) Reconstruction of tlie Antillean Continent. Bull. G. S. A.,
1895, VI. 103-140. (2) Geographical Evolution of Cuba. Bull. G. S. A.,
1896, VII. 67-94. (3) High Level Shores in the region of the Great Lakes
and their Deformation. A. J. of S., 1891, XLL 201-211.
Tarr, R. S. Wave-formed Cuspate Forelands. Am. Geol , 1898, XXII. 1-12.
Taylor, F. B. (1) A reconnaissance of the abandoned shoreline of Green Bay
and of the southern coast of Lake Superior. Am. Geol., 1894, XIII. 316-
327, 365-383. (2) A short history of the great lakes. Inland Educator,
1896, II. 138-145. (3) Correlation of Erie-Huron beaches with outlets and
moraines in southeastern Michigan. Bull. G. S. A., 1897, VIII. 31-56.
Trevelyan, W. C. Indications of Recent Elevations in the Islands of Guernsey
and Jersey and on the Coast of Jutland, and on some Tertiary Beds near
Porto d'Anzio. Proc. Geol. Soc, 1837, II. 577-578.
252 PROCEEDINGS OP THE AMERICAN ACADEMY.
Tylor, Alfred. On the Formation of Deltas, and on the Evidence and Cause
of Great Changes in the Sealevel during the Glacial Period. Geol. Mag.,
1872, IX. 392-399, 485-501.
Upham, Warren. (1) The Glacial Lake Agassiz. Mon. XXIV. U. S. G. S. ;
also Canada, Geol. & Nat. Hist. Sur., IV., 1888-9, E. (2) Late Glacial or
Cliamplain Subsidence and Re-elevation of the St. Lawrence River Basin. A.
J. of S., 1895, XLIX. 1-18. References on Glacial Lakes.
Walther, Johannes. Die Adamsbriicke und die Korallenriffe der Palkstrasse.
Pet. Geog. Mitt., Erg., 102, 1891, 1-40.
Warren, G. K. (1) Survey of Upper Mississippi River. Rep. U. S. Engrs.,
War Dept., 45 pp., 1867. (2) On Certain Physical Features of the Upper Mis-
sissippi River. Rep. U. S. Engrs., 1868, 307-314 ; Am. Nat., 1868-9, II. 497-
502. (3) Valley of the Minnesota River and of the Mississippi River to the
Junction of the Ohio; its Origin considered. A.J. of S., 1878, CXVI. 417-
431.
Weule, K. Beitrage zur Morphologic der Flach-Kiisten. Weimar Z. f wiss.
Geog., 1891, VIII. Pts. 6-8, 211-256.
Whiting, H. L. (1) Progress of Sandy Hook from 1848 to 1860. U. S. C. G. S.,
Appen. 9, 1850, 81-82. (2) Report of changes in tlie shoreline and beaches
of Martha's Vineyard, as derived from comparisons of recent with former
surveys. U. S. C. G. S., Appen. 9, 1886, 263-266.
Woods, J. E. Tenison-. Raised Beaches, Southern Australia. Proc. L. Soc.
N. S. W., 1880, V. 645 ; 1882, VII. 383.
GULLIVER. — SHORELINE TOPOGRAPHY. 253
CONTENTS.
INTRODUCTION.
Page
Physiographic Standpoint. — Omitted Phases of the Subject. — Use of Terms.
— Previous Work on Shorelines 151-153
PART I. INITIAL FORMS.
1. The Geogkapiiic Cycle.
Systematic Sequence of Forms. — Succession on Land. — Succession on the
Coast. — Rising, raised ; sinking, sunken. — Areas of Elevation and Depres-
sion as mapped. — Algebraic Sum of Movements : Maine. — Cycle, Epicy-
cle, and Vibration : New Jersey ; Scandinavia. — Volcanic and Climatic
Accidents. — Geographic and Paleontologic Criteria. — Ideal Areas . 154-161
2. Uniform Uplift.
Initial Stage of an ideal Area. — (1) Smooth Bottom. — (2) Simple new Shore-
line: Buenos Ayres. — (3) Smooth coastal Plain: Texas. — (4) Elevated
former Shoreline: San Clemente. — (5) Dissected Oldland. — Variations
from ideal Scheme. — Slow and rapid Movement. — Regional and Continen-
tal Uplift 161-166
3. Uniform Depression.
Initial Stage of an ideal Area. — (1) Uneven Bottom. — (2) Irregular new
Shoreline: Scandinavia.— (3) Dissected Land comparable with submerged
Topography : Austrian Coast. — Other Examples. — Variations . . 166-170
4. Diverse Movements.
Tilting: Position of Pivotal Axis. — Topography of Tilted Regions: Califor-
nia ; New England. — Warping : New Brunswick, N. J. — Santa Catalina
Depression. — Crumpling and Faulting 170-173
PART IL SEQUENTIAL FORMS.
5. Sea Attack and Transportation.
Differential Abrasion. — Monadnocks versus Marine Remnants. — Coastal
Inequalities. — Wave-cut Islands. — Submarine Platform. — American and
Englisli Views. — Wave-base. — Sea Transportation. — Offset ; Overlap;
Stream Deflection. — Dominant Current. — Current Cuspate Forelands 173-183
254 PROCEEDINGS OP THE AMERICAN ACADEMY.
6. Offshore Bar. ^
Page
Shelving Shore. — Stages. — No Offshore Bar. — Youth : Texas. — Adoles-
cence: Southern New Jersey 183-185
7. Dissected Coastal Plain.
Surface Form. — Youthful Dissection : Ogunquit, Maine. — Adolescent Dis-
section : Eastern Italy. — Mature Dissection: Eastern Virginia. — Adjust-
ment of Drainage 185-186
8. Fading Elevated Shoreline.
Lake Shorelines. — Typical Forms of Bonneville. — Lake Agassiz. — Marine
and Lake Terraces. — Examples. — Superposed Drainage 18G-189
9. Islands.
Consumption by the Sea. — Loop-bar: Shapka. — Flying-Bar: Sable Island.
— Simple Cases of Island-tying and their Stages. — I. Initial Island (Birth) :
Austria ; Sweden. — II. Nipped Island (Infancy) : Sweden ; Maine. —
III. Uncompleted Tombolo (Youth). a. Attached to Mainland only:
Gigha. b. Attached to Island only : Tunij. Attached to Mainland and
Island: Aebeli). — IV. Completed Tombolo (Adolescence). a. Single
Tombolo : Nahant. h. Y-tombolo : Morro del Puerto Santo, e. Double
Tombolo. (1) Onlj- one Bar completed : Marblohead Neck. (2) Both Bars
completed : Monte Argentario. — V. Lagoon Marsh-meadow (Adolescence) :
Colchester Point. — VI. Vanishing Island (Adolescence). Cuspate Tom-
bolo: Block Island. — VII. Straight Coast (Maturity). — Complex Cases
of Tying 189-201
10. Bay-Bars.
An Adolescent Feature — A. Bar across Mouth of Bay : Lake Ontario. —
(1) Bay but little filled (Youth-Adolescence). — (2) Bay more or less filled
(Adolescence). — (3) Bay filled (Maturity). —B. Bar in Middle of Bay. —
(1) Bay but little filled (Youth-Adolescence). — (2) Bay more or less filled
(Adolescence). — (3) Bay filled inside Bar (Adolescence). — C. Bar near
head of Bay : Drakes Bay. — Magilligan Foreland 201-212
11. Winged Beheadlaxd
A Combination of Bay-bar and Cliff. Type: Long Branch, N. J. — Other
Examples 213-214
12. Tidal Cuspati: Forelands.
Location and Description. — Type. — Tidal Hypothesis. — V-bar Stage. —
Lagoon-Marsh Stage. — Filled Stage. — Methods of Growth. — Theory
confronted with Fact 214-220
13. Bat-deltas.
History of a drowned Valley. — Type young Bay-delta : Loch Fine. — Other
young Bay-deltas. — Type adolescent Bay-delta : Dwamish. — Other Exam-
GULLIVER. — SHORELINE TOPOGRAPHY. 255
Page
pies. — Type mature Bay-delta : Simeto. — Other Examples. — Lake Bay-
deltas. — Bonneville Bay-deltas 220-224
14. Deltas.
Ratio between Sea and River Activities. — Delta Stages. — Credner. — Fur-
ther Study. — Classification. — 1. Lobate Deltas, (a) Unilobate Type.
(b) Multilobate Type. — 2. Cuspate-Lobate Deltas. — S. Cuspate Deltas. —
4. Rounded Deltas. — 5. Stunted Deltas. — 6. Blocked Streams . . 224-233
15. Tidal Scour.
Action in Bays. — Rmiways. — Ratio between Tides and Currents. — Series of
Forms. — Western Florida Type. — Schleswig-Holstein Type. — Georgia-
South Carolina Type. — Southern New Jersey Type. — North Carolina
Type. — Texas Type , 233-237
16. Cliffs.
Nip. — Examples of Nips. — Irregular Cliffs (Infancy- Youth)- — Minutely
irregular Shoreline. — Gently curved Cliffs (Adolescence). — Straighter
Cliffs (Maturity) 238-241
17. Spits.
A Characteristic of Adolescence. — Straight Spits. — Curved Spits. — Hooked
Spits. — Serpent Spits 241-242
18. Stages of Development of Various Coasts.
Average Stage. — Uplift : Youth. — Uplift : Adolescence. — Uplift : Maturity.
Depression : Youth. — Depression : Adolescence. — Depression : Matu-
rity 242-246
Shoreline References 247-252
256 PROCEEDINGS OF THE AMERICAN ACADEMY.
LIST OF FIGURES.
1. Casco Bay, Maine : Drowned Topography in Youthful Stage of Develop-
ment (Sheet 315' U. S. C. G. S.).
2. Diagram sliowing Mutual Relations of Cycle, Epicycle, and Vibration.
3. Ideal Block in Initial Stage following Uplift, sliowing Smooth Bottom,
Simple New Shoreline, Smooth Coastal Plain, Elevated Former Shoreline, and
Dissected Oldland.
4. Elevated Former Shorelines on San Clemente Island, Cal.
5. Diagram showing the Relation to Baselevel of the Peneplain, the Sub-
marine Platform, Wave-base, and the Continental Delta.
6. Offsets.
7. Overlaps.
8. Stream Deflections.
9. Typical Current Cuspate Foreland.
10. Canaveral Foreland, Fla. (Sheet IGl, U. S. C. G. S.).
11. Dars Foreland, Germany.
12. Loop-bar : Sliapka Island, Alaska.
13. Uncompleted Tombolo Stage.
14. Completed Tombolo Stage. — n. Single Tombolo : Nahant, Mass.
15. " " " b. Y-Tombolo : Morro del Puerto Santo,
Venezuela.
IG. " " " c. Double Tombolo : Monte Argentario,
Italy.
17. Vanishing Island Stage : Sandy Point, Block Island, li. I
18. Bay-bar across Mouth of Bay : Lake Ontario.
19. Nearly Mature Filling of Bays on Chirikof Island, Alaska.
20. Bay-bar near Head of Bay ; dominant Bottom Action : Skelder Bay,
Sweden.
21. Bay-bar near Head of Bay : Drakes Bay, California. Nip in Drakes Estero,
and infantile Cliffs on Reyes Point.
22. Mngilligan Foreland, Ireland.
23. Filled Stage of Tidal Cuspate Foreland. Type: West Point, Washington.
24. Profile of Tidal Cuspate Forclaml : Point Wilson, Wash.
25. Ideal Scheme of Tidal Inflow: Port Discovery, Wash.
26. Lagoon-marsh Stage of Tidal Cuspate Foreland : Gaspee Point, Narragan-
sett Bay, R. I.
27. Young Bay-delta : Loch Fine, Scotland.
28. Mature Bay-delta : Simeto, Sicily.
29. Type Cuspate Delta : Tiber, Italy.
30. Type Stunted Delta : Cavone, Italy.
31. Blocked Stream : Oceanside, California.
32. Curved Spit: Dutch Point, Lake Michigan (PI. XXIX., 13th Ann. Rep.
U. S. G. S.).
GULLIVER. — SHORELINE TOPOGRAPHY, 257
LIST OF ABBREVIATIONS.
Periodicals.
A. A. A. S. = American Association for the Advancement of Science : Salem,
Mass.
Am. Geol. = Tlie American Geologist : Minneapolis.
A. J. of S. =: The American Journal of Science : Kew Haven.
Am. Nat. = The American Naturalist : Pliiladelphia.
Atlan. Mon. = The Atlantic Montiily.
BulL G. S. A. = Bulletin of the Geological Society of America: Rochester, N. Y.
BulL Soc. Geog. = Bulletin de la Societe de Geographie : Paris.
Bull. U. S. G. S. = Bulletin of the United States Geological Survey : Washington.
C. R. — Coniptes Rendus de I'Academie des Sciences : Paris.
C. R. Soc. Gfeog. = Compte Rendu des Seances de la Socie'le de Geographie :
Paris.
Geol. Mag. = London Geological Magazine.
J. A. G. S. = Journal of the American Geographical Society: New York.
Jour, of GeoL = The Journal of Geology : Cliicago.
J. R. G. S. = Journal of the Royal Geographical Society: London.
Mem. B. Soc. Nat. Hist. = Memoirs of the Boston Society of Natural History.
Mill. Proc. Inst. Civ. Eugin. = Minutes of Proceedings of tlie Institution of
Civil Engineers : London.
Nevr Phil. Jour. = New Philosophical Journal : Edinburgh.
Pet. Geog. Mitt. = Peterraanns Mitteilungen aus Justus Perthes' geographischcs
Anstalt : Gotha.
Pet. Geog. Mitt., Erg. = Ditto, Erganzungs-Heft.
PhiL Trans. = Philosopliical Transactions of the Royal Society : London.
Pop. Sci. = Popular Science Monthly : New York.
Proc. Acad. Nat. Sci. Phil. = Proceedings of the Academy of Natural Sciences :
Philadelphia.
Proc. B. Soc. Nat. Hist. — Proceedings of the Boston Society of Natural
History.
Proc. Geol. Soc. = Proceedings of tlie Geological Society : London.
P. R. G. S. = Proceedings of tiie Royal Geograplncal Society : London.
Q. J. G. S. = Quarterly Journal of the Geological Society : London.
Rep. Brit. A. A. S. = Report of the British Association for tiie Advancement of
Science.
Rev. de Geog. = Revue de Geographie : Paris.
Sci. = Science : New York.
Scot. Geog. Mag. = Tlie Scottish Geographical Magazine : Edinburgh.
Smith. Cont. Kn. = Smithsonian Contributions to Knowledge : Washington.
Trans. GeoL Soc. = Transactions of the Geological Society : London.
VOL. XXXIV. — 17
258 PROCEEDINGS OF THE AMERICAN ACADEMY.
U. S. C. G. S. = United States Coast and Geodetic Survey : Washington.
U. S. G. S. = United States Geological Survey : Washington.
Z. d. G. f. E. = Zeitschrift der Gesellschaft f iir Erdkunde : Berlin.
Z. f. E. = Zeitschrift fiir Erdkunde : Berlin.
Maps.
Atlas Univ. = Atlas Universel par Vivien de St. Martin : Paris, Hachette &
Compagnie. References follow revised numbering of the continuation of the
series by Fr. Schrader.
Attica = Karten von Attika, von E. Curtius und J. A. Kaupert, 1 : 25,000 : Ber-
lin, 1881-1897.
Austr. = K. u. K. militar-geographisches Institut : Austria, 1 : 75,000.
Belg. = Carte Topographique de la Belgique, 1 : 40.000.
Brit. Ad. = Charts published by Order of the Lords Comaiissioners of the Admi-
ralty : London, various scales.
C. S. = United States Coast and Geodetic Survey, various scales.
Deum. = Geiieralstabens Kort over Danniark, 1 : 100,000.
Elba = Carta Geologica dell' Isola d' Elba, 1 : 25,000; Rome, 1884.
Eug. = Ordnance Survey of England, 1 : 63,860.
Fr. = Carte de la France, au Depot general de la Guerre, 1 : 80,000.
Geol. Eu. = Carte geologique internationale de I'Europe, 1 : 1,600,000. Beyrich
und Hanchecorne, Berlin, 1894.
G. S. = United States Geological Survey, scales, 1 : 62,500, 1 : 125,000.
Germ. = Karte des deutschen Reiches, K. Preuss. Landes-Aufnahnie, 1 : 100,000.
HoU. = Topographische en militaire kaarte van het koningrijk Nederlanden,
1 : 50,000.
H. O. = Hydrographic Office, United States Navj', various scales.
Ireland = Ordnance Survey of Ireland, 1 : 63,360.
Ital. = Instituto geografico militare : Italy, 1 : 100,000.
N. J. = Atlas of New Jersey, Geological Survey of New Jersey.
Nor. = Topografisk Kart over Kongeriget Norge, 1 : 100,000.
Rus. = Special Map of European Russia, 1 : 420,000.
Scot. = Ordnance Survey of Scotland, 1 : 63,360.
Sicily = Carta Geologica dell' Isola di Sicilia, 1 : 100,000.
Spain = Meniorias de la Comision del Mapa Geologico de Espana. Mapas,
1 : 400,000.
Spain Bol. = Boletin de la Comision del Mapa Geologico de Espana. Mapas,
1 : 400,000.
Stieler = Adolf Stieler's Handatlas, Gotha, Justus Perthes, various scales.
Swe. = Generalstabens Karta ofver Sverige, 1 : 100,000.
STve. Geol. = Sveriges Geologiska Undersiikning, 1 : 50,000.
SawIss = Eidgenossisches Militair Archiv, Switzerland.
Tunis = Carte topographique de la Tunisie, Service geographique de I'Armee :
France, 1 : 60,000. •
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. Xo. 9. — January, 1S99.
JAPANESE COLLEMBOLA.
By Justus Watson Folsom.
With Three Plates.
JAPANESE COLLEMBOLA. Part U*
By Justus Watson Folsom.
Received November 1, 1898. Presented by Samuel II. Scudder, November 23, 1S98.
The interesting collection upon which this paper is based was made
by Prof. C. Ishikawa, of the Agricultural College at Komaba, Tokyo,
and was most generously given me to study by Professor Packard, to
whom my sincere thanks are due. All that has hitherto been published
upon the Collembola of Japan is a short article by myself, in which three
new species are described. The present paper deals with eleven species,
of which six species and one variety are new. In accordance with the
wishes of Professor Packard, a series of specimens has been given to the
United States National Museum at Washington, D. C, and another to
the Museum of Comparative Zoology at Cambridge, Mass.
My friend, Dr. C. Shaffer, of Hamburg, has materially assisted my
studies by sending identified examples of European Collembola, as well
as valuable notes. Mr. Samuel Henshaw has given me many useful
suggestions, and freely permitted me to study the collection of the
Cambridge Museum.
List of Japanese Species.
1. Aphorura inermis Tull. 7. Cremastocej/halus affinis, n. sp.
2. Xeni/lla longicnuda Folsom. 8. Seirajaponica Folsom.
3. Achorntes communis Folsom. 9. Tomocerus varius, n. sp.
4. Achorutes rp-acilis, n. sp. 10. Papirius denticulatus, n. sp.
5. Isotoma nitida, n. sp. 11. Sminthurus hortensis Fitch.
6. Entomobrya straminea, n. sp. 12. Sminthnrua viridis L. var. annulatus, n. var.
* Part I. was publislied in Bull Essex Inst., Vol. XXIX., (1897) 1898.
262 PROCEEDINGS OF THE AMERICAN ACADEMY.
Fam. APHORURID^.
Genus Aphorura MacG.
Aphorura inerinis TuU.
(Plate 1, Figs, 1-5.)
Tullberg, p. 18.
■ia (L.) Lubbock, p. 303, PI. 22, Figs. 27, 28.
Tullberg, p. 154.
" pp. 55, 56.
Lubbock, pp. 191-193, PI. 46, PI. 56, Figs. 24-26.
Packard, pp. 28, 29.
Parona, pp. 609, 610.
Oudeinans, p. 90.
Uzel, p. 75.
Schott, p. 24.
" pp. 88, 89.
MacGillivray, p. 313.
Loiinberg, p. 165.
Schott, p. 128.
Reuter, p. 33.
Schott, p. 187.
Schaffer, pp. 161, 163, Taf. 2, Figs. 17-21.
Lie-Pettersen, p. 21,
White (Fig. 1) ; the contents of the stomach appear as a broad black-
ish stripe. Eyes absent. Postautennal organ (Figs. 2, 3) elongated,
parallel-sided, of eight or nine elements. Pseudocelli of the head ten in
number (Fig. 2), of which three lie behind the base of either antenna,
and four occupy the posterior border of the head. Antennae (Fig. 2)
shorter than the head, of four segments, related in length as 4 : 5 : 6 : 7 ;
antennal organ (Fig. 4) consisting of a dorso-lateral group of four chiti-
nous, finger-like processes, accompanied by setae, on the anterior part of
the third segment. Body elongated, clothed with short bristles and tuber-
culated more finely than the head; the number of pseudocelli on the
dorsum of each successive segment is, respectively, 0, 4, 4, 6, 6, G, 6, 4, 0.
Legs stout, bristly ; superior claws (Fig. 5) stout, curving and tapering
uniformly from a broad base, untoothed ; inferior claws slightly shorter,
slender, gradually attenuating into a fine filament. Ventral tube present ;
furcula absent. Length, 1.8 mm.
One hundred and forty-two specimens examined, which were collected
at Komaba, Tokyo, October 9, 27, and November 14, ,1894.
1869.
Lipura inermis
1868.
" Jinietaria {
187L
" inermis
1872.
<( (I
1873.
" Jinietaria
1873.
"
1878.
"
1890.
"
1891.
"
1891.
'' inermis
1893.
"
1893.
A .,;,„-,.^..
1894.
Lipura i/iennis
1894.
"
1895.
"
1896.
"
1896.
Aphorura "
1897.
Lijiura "
FOLSOM. — JAPANESE COLLEMBOLA." 263
Except for the number of elements in the postantennal organs, the
Japanese specimens agree perfectly with North American representatives
of the species ; the examples from Massachusetts which I used for com-
parison are of the same lot of which some had been sent to Dr. SchafPer,
who pronounced them to be A. inermis Tull., and the equivalent of
Lipura Jimvtaria (L.) Lubb. Packard's specimens oi L. jimetaria in the
Museum of Comparative Zoology are the same species; in fact, Packard
hiiDself wrote ('73, p. 24), "It appears that on comparison I can find no
difference between European and American specimens of Lipnra jime-
taria." It is at least questionable, however, whether L.fimetaria (L.)
Lubb. is the Linnaean species.
A. inermis is a widely distributed form, having been recorded from
Sweden (Tullberg, Schott), Norway (Lie-Pettersen), Finland (Renter),
Germany (Schaffer), Bohemia (Uzel), Italy (Parona), England (Lub-
bock), Sumatra (Oudemans), and in North America from Massachusetts
(Packard), California (Schott), and Florida (Lonnberg, Schott).
Fam. PODURID^.
Genus Achorutks Tempi.
Achorutes communis Folsom.
1898. Achorutes communis Folsom, pp. 52, 53, Figs. 1-9.
One hundred and thirty specimens, of all sizes, from Komaba, Tokyo,
differ from the types only by having longer and more slender anal spines,
in many cases.
Achorutes gracilis, n. sp.
(Plate 1, Figs. G-13 )
General color, indigo blue (Fig. 6) ; legs and furcula pale ; sternum
yellow; the disposition of the hypodermal pigment is shown in Figure 7.
Head clothed with stiff setae ; ej'es eight on either side (Fig. 8), upon
black patches ; postantennal organ (Fig. 9) of four elements. Antennae
subequal to the head in length, with long bristles and with segments
related in length as 3 : 4 : 4 : 6. Body cylindrical-ovate in dorsal aspect
(Fig. 6), and sparsely clothed with short reclinate setse (Fig. 10). Legs
stout, basally spotted with blue; superior claws (Fig. 11) uniformly ta-
264 PROCEEDINGS OF THE AMERICAN ACADEMY.
pering and curving, uuidentate one third from the apex ; inferior claws
half as long, lunate, acuminate ; tenent hairs, two on the fore feet and
three on the others. Dentes stout, slightly tapering, with stout reclinate
satEe and a few extra long bristles ; mucrones a third as long as the
dentes, in form as represented in Figure 12. Anal spines (Fig. 13) two,
small and conical. Length, 1.5 mm.
Described from twenty-four types, from Yanaka, Tokyo, November
18, 1894.
This species closely approaches the European A. purpurascens Lubb.,
as well as A. theelii Tull. of Nova Zembla, both of which species I have
received from Dr. Schiiffer ; from these A. gracilis is distinguished espe-
cially by the form of the mucrones, together with the coloration.
Fam. ENTOMOBEYID^.
Genus Isotoma Bourl.
Isotoma nitida, n. sp.
(Plate 1, Figs. 14-18.)
Bluish gray, with a slight greenish tinge ; antennjc darker ; sides mot-
tled with pale spots ; sternum pale. Head, body, and appendages clothed
with short dense bristles (Fig. 14) ; genje gibbous. Eyes (Figs. 15, 16)
eight on either side, upon black patches ; postantennal organ absent.
AntennjB two fifths as long as the body, stout, with segments related in
lenf th as 2 : 4 : 4 : 5, and with the last three segments petiolate. Supe-
rior claws (Ficr. 17) broad basally, slightly curved, untoothed, with two
filiform pseudonychia ; inferior claws broadly lanceolate, without teeth ;
tenent hairs absent. Furcula slender, exceeding the ventral tube, with
segments related as 4 : 16 : 1 ; mucrones (Fig. 18) four-toothed. Length,
1.4 mm.
Described from seventy-two types, of which seven were collected at
Komaba, Tokyo, November 16, 1894, and the remainder at Miyagi,
Boshyu, November 9, 1895,
This species is, upon the whole, most nearly allied to 1. pahistris Miill.,
especially in the form of the mucrones (cf. Schott '93, Taf, 6, Fig. 5).
T find no species of Isotoma, except that now described, which possesses
filiform pseudonychia.
FOLSOM. — JAPANESE COLLEMBOLA. 265
Genus Entomobrya Rond.
Entomobrya straminea, n. sp.
(Plate 2, Figs. 19-23.)
Pale straw yellow throughout. Head, body, and appendages densely
clothed with barbellate bristles ; the vertex and basal antennal segment
bear also stout, clavate sette (Figs. 19, 20). Eyes three on either side,
black, arranged as in Figure 21. Antennae almost half as long as the
body, segments cylindrical, related in length nearly as 1 : 2 : 2 : 3. The
body segments, measured along the median dorsal line, are related as
4:17:12:8:11:11:31:5:3; a cluster of clavate setje arises from
the anterior border of the mesonotum, and a similar dorsal cluster occurs
upon each succeeding segment. Legs slender; superior claws (Fig. 22)
almost straight, tapering to a sharp point, in lateral aspect showing a
small tooth on the outer margin and two on the inner margin ; one of
the latter is comparatively small, situated near the middle, and overhung
by the greatly developed second, or basal tooth ; inferior claws over
half as long as the others, straight, broadly linear, acuminate, bearing on
the outer margin a broad, acute, hyaline lamella ; a single stout but
unknobbed tenent hair is present. Furcula with segments related as
28 : 49 : 3 ; mucrones (Fig. 23) broadly falcate, with a prominent erect
tooth near the middle and surrounded by three or four stout barbellate
bristles which project from the dentes. Length, 1.9 mm.
Seven types, from Komaba, Tokyo, November 16, 1894.
E. straminea agrees with E. sexoculata Schott ('9G, pp. 180, 181, PI. 17,
Figs. 30-32) in the number of eyes, but differs in the formation of
the claws and mucrones, as well as in other respects. It also bears
much resemblance to Sinella hijfti Schaffer ('96, pp. 192, 193, Taf. A,
Figs. 103-105). P"'or reasons already urged by Schott ('91, p. 20,
'96, p. 180), I follow that author in uniting the genus Sinella with
Entomobrya.
Genus Cremastocephalus Schott.
Cremastocephalus affinis, n. sp.
(Plate 2, Figs. 24-27.)
Color, chrome yellow ; the lateral borders of segments two to six
inclusive, the posterior borders of the last two abdominal segments and
266 PROCEEDINGS OF THE AMERICAN ACADEMY.
the apex of each antennal segment are frequently dark purple ; the
antennre, legs, and furcula are pale yellow. The head hangs down
(Fig. 24), is elongated, and clothed with numerous proclinate bristles
interspersed with extra long, slender, erect bristles ; similar long bristles
occur also on the body, antennge, legs, and furcula. Eyes eight on
either side (Fig. 25), arranged in two longitudinal rows, on black
patches. Antennae one fourth longer than the body, bristly, with seg-
ments cylindrical or slightly dilated and related to each other in length
as 25 : 31 : 27 : 35. The body (Fig. 24) is elongate-cylindrical, clothed
with reclinate bristles and scaleless ; its segments, measured along the
median dorsal line, are related as 2 : 12 : 8 : 6 : 7 : 2 : 34 : 4 : 4. The
thorax curves downward ; the mesonotum almost covers the prothorax
and bears clavate bristles on its anterior border. Legs long, slender,
and bristly; superior claws (Fig. 26) stout, but little curved, with a
tooth on the inner margin, one third from the apex, and a second tooth
near the base ; inferior claws half as long as the others, broad, with
acuminate apex, convex outer margin and a single tooth, borne upon an
obtuse angle at the middle of the inner margin ; a single tenent hair is
present which gradually expands to a broad truncate apex. Furcula
five eighths as long as the body and bristly ; manubrium cylindrical,
slightly shorter than the dentes ; dentes gradually tapering, each bearing
a large oval scale near the apex (Fig. 27) ; mucrones oblong, somewhat
curved, with three terminal lobes, which are subequal, rounded, and
surrounded by barbelhite bristles projecting from the dentes. Length,
2 mm.
Described from seven types, collected at Komaba, Tokyo, October
25, 1894.
This curious form is closely related to the Mexican Cremastocephal us
trilobatus Schott ('96, pp. 175-178, Plate 16, Figs. 20-23, Plate 17, Figs.
25, 26), which has hitherto been the only representative of its genus.
The specific distinctness of the two species is evident when my figures
are compared with those of Schott ; the chief differences exist in colora-
tion and the form of claws and mucrones ; the dental scale is elliptical
in trilobatus, but oval in the species now described. I may mention that,
although Schott states tliat the upper claw of trihbatus is "provided
with two teeth," there are three represented in his figure.
FOLSOM. — JAPANESE COLLEMBOLA. 267
Genus Seira Lubb.
Seira japonica Folsom.
1898. Seira japonica Folsom, pp. 55, 56, Figs. 15-18.
Many of Dr. Ishikawa's specimens differ from ray types by being
larger, attaining a length of 3 mm. The antennal segments are more
slender, are related to each other nearly as 3 : 5 : 4.5 : G, and are purple
throughout. Clavate hairs are few in number and the scales have dis-
appeared. The second abdominal segment is usually yellow ; the meso-
uotum is laterally bordered with blackish blue and occasionally each side
of the head bears a stripe of that color. A single example is yellow
throughout, excepting the antennae, lateral borders of meso- and meta-
notum and the posterior border of the fifth abdominal segment, all of
which are purple. In all other respects the specimens agree perfectly
with the types, which are manifestly younger individuals.
Sixteen specimens, large and small, were taken at Komaba, Tokyo,
October 27 and November 16, 1894. I omitted to mention in my pre-
vious paper that the types are dated June 24, 1897.
Genus ToMOCERus Nic.
Tomocems varius, n. sp.
(Plate 2, Figs. 28-30, Plate 3, Figs. 31, 32.)
Color with scales, plumbeous ; without them, dull yellow. Eyes
(Fig. 28) six on either side, on black patches. Antennne almost as long
as the body, with purple segments, related in length as 3 : 4 : 27 : 5.
From under the anterior margin of the mesonotum project many stout
stiff seta3 (Fig. 29). Superior claws (Fig. 30) nearly straight, rather
stout, with from two to five teeth which successively become more obscure
toward the apex of each claw ; in the kw specimens at command, the
fore claws bear two or three teeth, the mid claws from two to five, and the
hind claws two ; the right and left claws of the same pair of feet often
differ in the number of teeth. The inferior claws are a little more than
half as long as tlie others, broadly lanceolate and unidentate. A sin-
gle tenent hair is present. Furcula seven tenths the length of the body,
with segments related nearly as 3 : 4.5 : 1. The dental spines (Fig. 31)
268 PROCEEDINGS OF THE AMERICAN ACADEMY.
are simple and vary in number from eight to ten ; the proximal spines
are smallest and the two most distal are largest. Each mucro (Fig. 32)
bears a single blunt tooth near the middle. Length, 2.5 mm.
Described from three types, collected at Komaba, Tokyo, November
16, 1894, and Oji, Tokyo, November 18, 1894. These had unfortunately
dried and shrivelled.
This species is intermediate between T. mhiutus Tull. ('76, p. 32, Taf.
8, Figs. 9, 10) which has 2-3 teeth on the superior claws and 10-11 den-
tal spines, and T. arcticus Schott ('93, pp. 43, 44, Taf. 3, Figs. 8, 9) which
possesses 4-5 teeth and 7-8 spines.
Fam. SMINTHURID^.
Genus Papiuius Lubb.
Papirius denticulatus, n. sp.
(Plate 3, Figs. 33-36.)
Pale chrome yellow, with purple markings (Fig. 33). Head sparsely
clothed with stout, stiff setai ; eyes upon black patches. Antenna;
purple, three fourths as long as the body, with segments related nearly
as 1 : 4 : 7 : 2 ; third segment with at least nine distal subsegments and
dilated apex ; terminal segment lanceolate in outline, with the basal
half composed of four subsegments. Legs yellow, banded with purple,
with stout bristles ; superior claws (Fig. 34) slender, nearly straight, the
outer margin unidentate one third from the apex, both inner margins
bidentate ; inferior claws (Fig. 34) over half as long, with acute apex,
almost parallel sides, long, knobbed, subapical hair, and two unequal,
perpendicular teeth on the rounded basal portion of the inner margin ;
the smaller tooth is occasionally absent ; a single, slender, unknobbed
tenent hair is present. Abdomen elongate-ovate, sparsely clothed with
short, stitf seta;, which become longer and numerous posteriorly. Color-
ation as shown in Figure 33 ; two broad paramedian stripes occur upon
the anterior half of the dorsum ; several oblique wedge-shaped bands
extend upward and backward from either side ; a median U-shaped
mark is conspicuous on the posterior part of the dorsum. Furcula
yellow, bristly, with segments related as 2 -. 4 : 1 ; manubrium broadly
oblong in dorsal aspect ; dentes slender, with long, stout, lateral bristles
(Fig. 33) which become barbellate toward the apices of the dentes (Figs.
FOLSOM. — JAPANESE COLLEMBOLA. 269
35, 36) ; rauci'ones (Fig. 35) oblong-lauceolate, serrate upon both edges.
Length, 2 mm.
Three types, from Komaba, Tokyo, November 16, 1894.
P. denticulatas' most nearly approaches the North American P. mar-
moratus Pack. (73, p. 42), of which I have examined the types and with
which I am inclined to regard P. lymculosus Schott ('91, pp. 14, 15, Taf.
3, Figs. 1-3) as synonymous.
Genus Sminthurus Latr.
Sminthurus horterisis Fitch.
(Plate 3, Figs. 37-40.)
1863. Smynthurus hortensis Fitch, pp. 6G8-673, Figs.
1841. Harris, p. 125.
1842.
1844. Smynthurus sp. " Fig.
1969. " [cHCumeris] " p. 362.
1871. Sminthurus pruinosus Tullberg, p. 145.
1872. " " " p. 31, Taf. 3, Figs. 13, 14.
1873. Smynthurus quadrisignatus Packard, p. 44.
1876. Sminthurus lineatus Reuter, p. 83.
1891. Smynthurus hortensis MacGillivray, p. 271.
1891. " frontalis Uzei, p. 37, Taf. 1, Fig. 3, Taf 2, Figs. 3-5.
1893. Smint/iurus pruinosus Schott, pp. 28, 29, Taf. 2, Figs. 13-16.
1895. " " Reuter, pp. 10-12.
1897. " " Schaffer, p. 26.
1897. Smynthurus albamaculata Harvey, pp. 124-126, Figs. 1-5.
1898. Sminthurus pruinosus Sclierbakow, p. 60.
General color dark purple, spotted with pale yellow ; antennae, legs,
and furcula paler purple. Head densely clothed with short proclinate
settB. Eyes (Fig. 37) upon large black patches, broadly surrounded
with pale yellow ; vertex yellow ; geiife with several circular yellow
spots. AntennjE over half as long as the body, with segments related
as 2 : 4 : 17 : 15 ; terminal segment composed of seven subsegments :
the apical two thirds of the terminal segment (Fig. 38) consists of six
subsegments, of which five are subglobose, while the last is elongate-
conical and itself represents three subsegment?, which however are not
distinct as such. Legs bristly ; segments darker apically ; superior claws
(Fig. 39) tapering, slightly curved, unidentate near the middle of the
inner edge ; inferior claws three fifths as long, entire, apex acuminate,
270 PROCEEDINGS OP THE AMERICAN ACADEMY.
outer margin straight, inner margin roundly dilated near the base ; tenent
hairs two or three, clavate. Coloration rather variable ; dorsum marked
anteriorly with transverse rows of pale yellow circular spots (Fig. 37) ;
sides of abdomen with many regular rows of minute, circular, yellow
spots ; sternum posteriorly yellow ; body well clothed with bristles,
which are short and reclinate, except upon the posterior segments, where
they become longer and more numerous. Furcula long and stout ;
dentes scarcely tapering; mucrones (Fig. 40) almost one third as long
as the dentes, oblong-lanceolate, with rounded apex. Length, 1.2 mm.
Two specimens with no locality given.
The Japanese examples agree satisfactorily with North American
representatives of S. hortensis ; in fact, I found some of our specimens
which match them quite closely in coloration. A study of Packard's
types leaves no doubt that quadrisignatus Pack, is a synonym of hor-
tensis Fitch. Professor Harvey kindly sent me numerous specimens of
his albamaculata, which also prove to be hortensis ; he regards the last
antenna! segment as being composed of nine subsegments, evidently
considering the last subsegment as three. The identity of the European
pruinosus Tull. with the American hortensis was called to my attention
by Dr. Shaffer, to whom I had sent examples of our species, and he has
since sent me ten South American specimens of the same form.
S. hortensis has been found in various parts of Europe : in Sweden
(Tullberg), Finland (Renter), Russia (Scherbakow), and Bohemia (Uzel).
In the United States it is recorded from New York (Fitch), Maine
(Packard, Harvey), Massachusetts (Harris), and Ohio (MacGillivray).
Finally, it is reported from subantarctic America (Schaffer).
Sminthnrus viridis Linn., var. annnlatiis, n. var.
(Plate 3, Figs. 41-43.)
1758. Podura viridis Linnaeus, p. 608.
1762. " " GeoflFroy, p. 007.
1781. " " Schrank.
1793. " " Fabricius, p. 65.
1804. Smyntliunis viridis Latreille, p. 82.
1806. " " " p. 166.
1835. " " Lacordaire et Boisduval, p. 115.
1835. " " Templeton, p. 97, PI. 12, Fig. 7.
1839. " ** Burmeister, p. 451.
1841. " " Nicolet, p. 82, Pi. 9, Fig. 9.
1842. " " Boiirlet.
FOLSOM. — JAPANESE COLLEMBOLA. 271
Lucas, p. 567.
Gervais, p. 401.
Lubbock, pp. 295, 296, PI. 21, Figs. 1-3.
I TuUberg, pp. 144, 145.
Tullberg. p. 30, Taf. 2., Figs. 16-20, Taf. 3,
Figs. 1-5.
Lubbock, pp. 100, 101, PI. 1, PI. 55, Figs. 1-4.
Reuter, p. 79.
Tullberg, p. 30.
TumiJsvary, pp. 37, 38.
Dalla Torre, p. 149.
Reuter, p. 227.
Uzel, pp. 34, 35.
I ScliGtt, pp. 22-24, Taf. 1, Figs. 1-5.
Parona, p. 696.
Reuter, pp. 9-10.
SciiiiEfer, pp. 209, 210, Taf. 4, Figs. 122, 123.
Poppe und Schaffer, p. 271.
Lie-Pettersen, p. 8.
Scherbakow, pp. 60, 65.
Pale yellow, spotted with blackish purple (Fig. 41) ; most of the spots
are approximately ring-like, being polygonal with pale centres ; one
individual is pale yellow throughout, except for faint purple spots on the
posterior part of the abdomen. Head densely clothed with proclinate
bristles. Eyes upon black patches. Antennae three fifths as long as
the body, yellow basally, purplish apically, with segments related as
1:4:6:13; terminal segment composed of about seventeen subsegments.
Body clothed with long reclinate bristles, numerous upon the anal
tubercle. A median dorsal purple streak occurs on the anterior half of
the body. Legs pale yellow, with long setas. Superior claws (Fig. 42)
stout, broad, slightly curved, mucronate at apex and unidentate on the in-
side, two fifths from the apex ; inferior claws over half as long, broa<llv
triangular, with a subapical hair and a single tooth upon the inner,
rounded margin ; tenent hair single, slender, and unknobbed. Furcula
pale yellow with long setae ; mucrones (Fig. 43) elliptical, with entire
margins. Length, 2 mm.
Described from two types, found at Komaba, Tokyo, November 16,
1894.
1842.
Smytithurus viridis
1844.
" "
1868.
"
1871.
Sminthurus " vars. cine-
reoviridis, nigromaculata
1872.
Sminthurus uiridis
1873.
Smynthurus "
1876.
Sminthurus "
1876.
»< (t
1883.
Smynthurus "
1888.
« <;
1891.
Smi)ithurus viridis var. tri- |
punctatus )
1891.
Sminthurus viridis,
1893.
'■ " vars. spe-
ciosus, dorsovittatus
1895.
Sminthurus viridis,
1895.
" " var. 1
infuscutas )
1896.
Sminthurus viridis, var. (
multipunctata )
1897.
Sminthurus viridis
1897.
"
1898.
var. )
lineata
272 PROCEEDINGS OF THE AMERICAN ACADEMY.
Upon comparing the Japanese variety with several examples of S.
viridis L. var. cinereoviridis Tull., which Dr. Schaffer gave me, I find
but few structural differences ; the European variety has more slender
mucrones and is clothed with longer bristles.
S. viridis is extremely variable in coloration and the present variety is
the ninth to receive a name. The species ranges throughout Europe,
having been recorded from Sweden (TuUberg), Norway (Lie-Pettersen),
Finland (Renter), Russia (Scherbakow), Nova Zerabla (Tullberg),
France (Bourlet, Geoffroy), Switzerland (Nicolet), Germany (Renter,
Schaffer), Bohemia (Uzel), Tyrol (Dalla Torre), Hungary (Tomosvary),
Italy (Paroua), England (Lubbock,) and Ireland (Templetou). Outside
of Europe, it is known from Tunis (Parona), Siberia (Reuter), and South
America (Parona).
FOLSOM. — JAPANESE COLLEMBOLA. 273
LITERATURE CITED.
BouRLET. 1842. Meinoire sur les PodureDes. Mem. Soc. Agric, etc. Nord.
78 pp., 1 PI.
BuRMEiSTER, H. 1839. Ilandbuch der Entomologie, Bd. 2. Berlin, pp. 443-458.
Dalla Torre, K. W. v. 1888. Die Thysanuren Tirols. Zeits. Ferd. Tirol Vor-
arl., Folge 3, Heft 32, pp. 145-160.
Fabricius, J. C. 1793. Entoniologia systematica, Tom. 2, pp. 63-68.
Fitch, A. 1863. Eighth Report on the noxious and other insects of the State
of New York, pp. 668-675.
FoLSOM, J. W. 1898. Japanese Collembola. Bull. Essex Inst., Vol. 29 (1897),
pp. 51-57, 1 PI.
Geoffroy. 1764. Histoire abrcge'e des insectes, Tom. 1, pp. 605-614, PI. 20.
Gervais, p. 1844. In Walckenaer, Histoire naturelle des insectes apteres, Tom.
3, pp. 377-456, PI. 50-52.
Harris, T. W. 1841. A Report on the insects of Massachusetts injurious to
vegetation. Cambridge, p. 125.
Harris, T. W. 1842. A Treatise on some of the insects of New England which
are injurious to vegetation. Cambridge, p. 125.
Harris, T. W. 1844. Cucumber skippers. Mass. Ploughman, Vol. 3, No. 42.
Harris, T. W. 1869. Entomological correspondence of Thaddeus William
Harris, M. D., ed. by Samuel H. Scudder. Boston Soc. Nat. Hist., Occas.
Papers, p. 362.
Harvey, F. L. 1897. Twelfth Ann. Rep. Maine Agric. Exp. Station (1896),
pp. 124-126, 1 PI.
Lacordaire et Boisduval. 1835. Faune entomologique des environs de Paris.
Paris, Tom. 1.
Latreillb, p. a. 1804. Histoire naturelle, generale et particulicre, des Crus-
tace's et des Insectes. Paris, Tom. 8, pp. 67-82, PI. 78.
Latreille, p. a. 1806. Genera Crustaceorum et Insectorum. Paris, Tom. 1,
pp. 164-167.
Lie-Pettersen, O. J. 1897. Norges Collembola. Bergens mus. aarb., 24 pp.,
2 Pis.
LiNNiEus, C. 1758. Systema naturae. Ed. 10. Holmiae, pp. 608, 609.
Lonnberg, E. 1894. Florida Aphoruridae. Can. Ent., Vol. 20, pp. 165, 166.
Lubbock, J. 1868. Notes on the Thysanura, Part 3. Trans. Linn. Soc, Vol. 26,
Pt. 1, pp. 295-804, PI. 21, 22.
Lubbock, J. 1873. Monograph of the Collembola and Thysanura. London,
255 pp., 78 Pis.
Lucas, H. 1842. Histoire naturelle des animaux articules, annelides, crustace's,
arachnides, myriapodes et insectes. Paris, pp. 553-568.
MacGillivray, a. D. 1891. A Catalogue of the Thysanoura of North America.
Can. Ent., Vol. 23, pp. 267-276.
MacGillivray, A. D. 1893. North American Thysanura. — IV. Can. Ent.,
Vol. 25, pp. 313-318.
VOL. XXXIV. — 18
274 PROCEEDINGS OF THE AMERICAN ACADEMY.
NicoLET, H. 1841. Recherches pour servir a I'liistoire des Podurelles. Nouv.
mem. soc. Helv., etc. 84 pp., 9 Pis.
OuDEMANS, J. T. 1890. Apterygota des Indischen Archipels. Weber's Zool.
Ergeb., Bd. 1, pp. 73-92, Taf. 6, 7.
Packard, A. S. 1873. Synopsis of the Thysanura of Essex Couuty, Mass., with
descriptions of a few extralimital forms. Fiftii Ann. Rep. Trust. Peab. Acad.,
pp. 23-51.
Parona, C. 1878. CoUembola. Saggio di un Catalogo delle Poduridi italiane.
Atti soc. ital. sc. nat., Vol. 21, pp. 559-611.
Parona, C. 1895. Elenco di alcune CoUembola dell' Argentina. Ann. mus. civ.
St. nat. Genova, ser. 2, vol. 14 (34), pp. 69G-700.
PopPE, C. A., UND ScHAFFER, C. 1897. Die CoUembola der Umgegend von
Bremen. Abhdlgn. nat. Ver. Bremen, Bd. 14, Ileft 2, pp. 265-272.
Reuter, O. M. 1876. Catalogus prsecursorius Poduridaruni Fenniae. Medd.
Soc. Faun. Flora Fenn., Heft. 1, pp. 78-86.
Reuter, O. M. 1891. Podurider fr&n nordvestra Sibirien, samlade af J. R.
Salilberg. Ofv. finsk. vet. soc. forh., Bd. 33, pp. 226-229.
Reuter, 0. M. 1895. Apterygogenea Fennica. Acta Soc. Faun. Flora Fenn.,
Bd. 11, No. 4, pp. 1-35, Taf. 1, 2.
Schaffer, C. 1896. Die CoUembola der Umgebung von HamVjurg und benach-
barter Gebiete. Mitt. Naturb. Mus. Hamburg, Jiig. 13, pp. 147-216, Taf. 1-4.
Schaffer, C. 1897. Apterygoten. Hamb. Magalh. Sammel.
Scherbakow, a. 1898. Einige Bemerkungen liber Apterygogenea die bei Kiew
1896-1897 gefunden warden. Zool. Anz., Bd. 21, pp. .57-65.
ScHOTT, H. 1891. Beitrage zur Kenntniss Kalifornisclier CoUembola. Bih. k.
Sven. vet. akad. hand., Bd. 17, afd. 4, No. 8, 25 pp., 4 Taf.
ScHOTT, H. 1893. Zur Systematik und Verbreitung pala»arctischer CoUembola.
Kongl. sven. vet. akad. hand., Bd. 25, No. 11, 100 pp., 7 Taf.
ScHoTT, H. 1894. Lipurider frdn Florida. Ent. tidsk. Srg. 15, p. 128.
ScHoTT, H. 1896. North American Apterygogenea. Proc. Cal. Acad. Sc, Ser. 2,
Vol. 0, pp. 169-196. PI. 16-18.
Sciirank, F. de p. 1781. Enumeratio Insectorum Austria indigenorum, pp.
494-499.
Templeton, R. 1835. Tiiysanurae Hibernicas. Trans. Ent. Soc. Lond., Vol. 1,
Pt. 2. pp. 89-98, Pi. 11-12.
TomosvArt, O. 1883. Magyarorszagban talalt Smynthurus-fajok. Termcs. fiiz.
Magyar nem. miiz., Bd. 7, pp. 31-38, Fig.
TcLLBERG, T. 1869. Om skandinaviska Podurider af underfamiljen Lipurinae.
Akad. afii. Upsala.
TuLLBERG, T. 1871. Fiirteckning ofver Svenska Podurider. ufv. k. vet. akad.
fr.rh. Srg. 28, No. 1, pp. 143-155.
TcLi.BERG, T. 1872. Sveriges Podurider. Kongl. sven. vet. akad. handl., Bd. 10,
No. 10. 70 pp.. 12 Taf.
ToLLBERG, T. 1876. CoUembola borealia. Ofv. k. vet. akad. forh. Srg. 33, No. 5,
pp. 23-42, Taf. 8-11.
UzEL, J. 1891. Thysanura Bohemia?. Sitzber. k. boh. Gesell. Wiss., Bd. 2, pp.
3-82, Taf. 1, 2.
FoLSOM. — Japanese Collewibola.
Plate I
15
.Z2.
"~^^
10
''^^wMd
^^^
Oo
On
o
oO
16
</. W. FOLSOM. DEL.
PLATE 2.
Fig. 19. Entomohrija straminca, n. sp. X OG.
Fifr. 20. " " " Bristle from head, X 424.
Pig. 21. " " " Eyes of left side, X 352.
Fig. 22. " " " Left aspect of left fore foot, X 540.
Fig. 23. " " " Left aspect of left nnicro, X 540.
Fig. 24. Cirmaatocephalus affinis, n. sp. X 44.
Fig. 25. " " " Eyes of left side, X 269.
Fi(T. 20. " " " Left aspect of right fore foot, X 540.
Fig. 27. " " " Mucro, X 540.
Fig. 28. Tomorerus varitia, n. sp. Eyes of left side, X 124.
Fig. 29. " " " Bristles of mesonotum, X 424.
Fig. 30. " " " Right aspect of right mid foot, X 424.
FoLSOM. — Japanese Collembola.
Plate 2.
d. W. FOLSOM, DEL.
• PLATE 3.
Fig. 31. Toinoccrus varius, n. sp. Dental spines, X 65.
Fig. 32. " '• " Right aspect of right mucro,
X 269.
Fig. 33. Pupirius denticulatus, n. sp. X 26.
Fig. 34. '■ " " Right aspect of right mid foot,
X 26y.
Fig. 35. " " " Left aspect of left nmcro, X 136.
Fig. 36. " " " Bristle froai dens, X 605.
Fig. 37. Smiutliurus hoitensis, Fitch. X 68.
Fig. 38. " " " Terminal antennal segment, X 65.
Fig. 39. " " " Left aspect of left mid foot, X 605.
Fig. 40. " '• " Mucro, X 269.
Fig. 41. Sminthurus viridis, var. annukitus, n. var. X 26.
Fig. 42. " " " " " Riglit aspect of right hind foot,
X 269.
Fig. 43. " " " " " Left aspect of left mucro, X 136.
FoLSOM. — Japanese Collembola
Plate 3.
</. W. FOLSOM. DEL.
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. No. 10. — January, 1899.
CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.
THE USE OF THE TRANSITION TEMPERATURES OF
COMPLEX SYSTEMS AS FIXED POINTS IN
THERMO METR Y.
Bv Theodoue William Richards and Jesse Briggs Churchill.
CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF
HARVARD COLLEGE.
THE USE OF THE TRANSITION TEMPERATURES OF
COMPLEX SYSTEMS AS FIXED POINTS IN
THERMOMETRY.
By Theodore William Richards and Jesse Briggs Churchill.
Received November 14, 1898. Presented November 23, 1898.
In a brief paper upon the transition temperature of sodic sulphate,* we
have recently shown that this non-variant point is cajmble of being repro-
duced in practice with great certainty, and that it is therefore admirably
suited for use as a standard of reference in thermometry. We pointed
out the fact that many other systems composed of two or more com-
ponents might answer equally well, and declared our intention of fixing
as many points as possible in order to simplify the accurate measurement
of temperature. The subsequent appearance of a hastily written note by
Messrs. Meyerhoffer and Saunders,! claiming for themselves a part of
this scheme, has prompted the present paper, which has as its object a
more detailed statement of the plan.
It is obvious that, while any number of components might be employed
simultaneously for this service, the simpler systems will be on the whole
the most useful. Water is so omnipresent as to be difficult to exclude
from any kind of experiment, hence the investigator is almost forced to
adopt it as one of the components. The choice is then of the other
material or materials, and the first step was obviously to study all com-
mon substances with a view to discover the probable usefulness of the
transition temperatures of their aqueous crystals. If a complete tem-
perature scale could not be built up from such simple data it would
obviously become necessary to investigate quintuple points, of which a
very great number could be devised. This additional complication could
not but be regretted, however; for it involves the necessity of preparing
two salts instead of one in a pure state, it renders less easy the use of
* American Journal of Science, VI. 201 (1898); also Zeitschr. phys. Chem.,
XXVI. 690.
t Zeitschr. phys. Chera., XXVII. 367, October, 1898.
278
PROCEEDINGS OP THE AMERICAN ACADEMY.
the same material over and over again, and it introduces a complication
which might seriously retard the speed of attaining equilibrium, and
hence the constancy of the desired point.
After studying with care the published records of all the field of inor-
ganic chemistry, about two dozen salts were selected as probably suitable
lor the work in hand. Of these magnesic and nickelous sulphates, calcic
and nickelous nitrates, nickelous chloride and borax were rejected as
being too inconstant in their indications. The lack of constancy was due
sometimes to the smallness of the latent heat of transition, and sometimes
to a superabundance of crystalline hydrates. On the other hand, at least
eight of the new salts gave results comparable in certainty with those
given by the ever trustworthy Glauber's salt. These salts are given in
the table below.
TRANSITION TEMPERATURES OF NINE SALTS.
Salt.
By Jlercury
Thermometer.
By Hydrogen
Tliermometer.
Sodic Chromato, Na^CrOi . IOH2O . .
19.7
19.6
Sodio Sulphate, NaoSO^ . lOII^O . . .
32.484
32.379
Sodic Carbonate, Na^COs . lOHoO . .
35.2
35.1
Sodic Thiosulphate, Na^S^Oa . 5IL0
48.1
48.0
Sodic Bromide, NaBr . 2IL0 ....
50.8
50.7
Manganous Chloride, MhCIq . 4IL0 . .
57.9
57.8
Strontic Chloride, SrCl . GILO . . .
Gl.l
61.0
Sodic riiosphate, NasPO^ . 12^,0 . .
73.5
73.4
Baric Hydroxide, Ba(0H).2 . SH.^O . .
78.0
77.9
The temperatures given above are only approximations, and are sul)-
ject to future revision ; at this stage of the work our effort has been only
to determine if the points were constant, and not to fix their absolute
value. Besides these nine, several other salts promise well. Baric
hydrate, given above, was our first trial of three components, for the car-
bonate was naturally allowed to be present. The study of such quintuple
points has now been interrupted by Messrs. Meyerhoffer and Saunders's
claim.
RICHARDS AND CHURCHILL. — TRANSITION TEMPERATURES, 279
These gentlemen studied in a very hasty fashion the transition tem-
perature of Glauber's salt in the presence of an excess of common salt.
Although theoretically sound, such a system labors under a serious prac-
tical disadvantage ; for the addition of heat to it means the dissolving
of common salt as well as the melting of crystallized sodic sulphate and
the depositing of anhydrous material. The first of these processes is
obviously less speedy than the others, and must surely occupy appreci-
able time even if the solid is finely powdered. Such a " lag " inevitably
affects the temperature ; our own experience with this mixture as well
as with other similar ones su[)ports this inference, and is anything but
reassuring. Indeed, we found that Glauber's salt itself did not give
absolutely accurate results if it was allowed to " freeze " instead of to
" melt," for a similar reason.
We agree with Messrs. Meyerhoffer and Saunders as to the great
desirability of uniting upon some normal temperature for the graduating
of flasks, etc., but we cannot conclude with them that 18° is the best
temperature. In America the steam-heated winters and sun-heated
summers raise the average temperature of our laboratories at least to
20°, and indeed this temperature is more comfortable than 18° unless
one is performing active manual labor. Hence at Harvard we have set-
tled upon 20° as the normal room temperature. Sodic chromate (19.G°)
clearly gives us very nearly the standard of reference which we desire.
In determining the specific gravities of liquids, a temperature above that
of the room is preferable to one beloio, — for the expansion of the liquid
during the drying of the exterior of the pycnometer is otherwise apt to
be troublesome, hence 18° is not suitable for this purpose. The authors
before mentioned suggest the use of a bath of mixed salts as a means of
keeping the temperature constant during determinations of electrolytic
conductivity ; but it should be pointed out that in such work the neigh-
borhood of a very large amount of a good electrolyte is necessarily risky,
except in the best of hands. This is especially the case when the sub-
stance effloresces to form a fine powde"", easily wafted around by currents
of air. In short, while for some work with closed vessels demanding the
greatest accuracy such a bath may be invaluable, the Ostwald thermostat
is the safest and most convenient appliance for preserving a constant
temperature in the laboratory. The baths of " melting " crystals will
find their greatest use in the standardizing of thermometers at fixed
points ; and these thermometers will continue to serve as the most
handy means of attaining and registering any desired temperature. It
is obvious that if a thermometer is standardized under exactly the
280 PROCEEDINGS OF THE AMERICAN ACADEMY.
conditions imposed upon it during its use, the correction for the cool
column projecting into the air may be omitted. This correction, by the
way, may account for the fact that Meyerhoffer and Saunders found
the transition temperature of sodic sulphate to be only 32.35° * instead
of 32.38.°
We are much pleased that the idea should have been grasped with
such eagerness in the laboratory of Professor van't Hoff, for no better
proof could be found of its unquestionable utility. We feel too that
constants of this sort, like atomic weights, should be studied by more
than one set of investigators, and that they should be finally investigated
with the utmost care in the Bureau Internationale des Poids et des
Mesures, and the Reichsanstalt ; hence we are glad to accord to Messrs.
Meyerhotfer and Saunders the right which they demand to investigate
quintuple jjoiuts involving sodie sulphate or sodic carbonate. At the
same time, we feel that our undoubted priority (our preliminary paper
having been finished early in June) allows us to study any desired
portion of this field ; and for the present, feeling that the simpler sys-
tems are the better ones, we shall investigate primarily the salts named
above.
Cambridge, November 14, 1898.
* By a clerical error, Rimbach's table for this correction was stated in our last
paper to be on page 143 of Landolt and Burnstcin's Tables (1894). It is really on
page 95.
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. Xo. 11 —February, 1S99.
ON THE THERMAL CONDUCTIVITY OF CAST IRON.
By E. H. Hall and C. H. Ayres.
With Two Plates.
Investigations on Light and IIeat, made and pdhlished wholly or in part with Appropriation
FROM the RUMFORD Fu.ND.
ON THE THERMAL CONDUCTIVITY OF CAST IRON.
By E. H. Hall and C. H. Ayres.
Presented October 12, 1898. Received December 6, 1898.
Two or three years ago an article entitled " On the Thermal Conduc-
tivity of Mild Steel " was published by one of the authors * of the pres-
ent paper. The method described at length in that article made use of
a disk of soft steel, about 0.3 cm. thick and about 10 cm. in diameter,
coated on each face with an electrolytic deposit of copper about 0.05 cm.
thick. Thin copper wires attached electrolytically to these copper coat-
ings led to a sensitive galvanometer, the deflections of which depended
upon the thermo-electromotive force of the couple made by the steel of
the disk and the copper of its coverings, and indicated the difference of
temperature existing between the two faces of the steel itself.
"Water of a known temperature was made to flow across one copper
face of the disk and water eight or ten degrees warmer across the other
copper face. The water delivery of one stream was measured, and its
change of temperature between entering and leaving the vessel con-
taining the disk was determined by means of two copper and German
silver thermo-electric junctions.
The apparatus containing the disk was surrounded by a water jacket
having a temperature near that of the disk, so that the radiation to or
from the exposed convex surface of the disk could be neglected.
This hasty review shows that, if all the measurements indicated were
correctly made, the thermal conductivity could be found by a simple cal-
culation based on the data affoi-ded by the experiments. In fact, the
experiments described in the article under discussion left something to
)e desired ; for they showed that different parts of the same face of the
lisk were not at the same temperature, and the process of calculation
necessary to deduce from the observations the mean difference of tem-
perature of the two faces of the disk was laborious, and perhaps to the
casual reader not entirely convincing. This difficulty and the desira-
bility of certain changes in the apparatus used were recognized in the
article itself.
* E. H. Hall, These Proceedings, Vol. XXXI. p. 271, 1896.
284 PROCEEDINGS OF THE AMERICAN ACADEMY.
The authors of the present paper have used the same method, applying
it to measure the thermal conductivity of a certain grade of cast iron ;
but they have employed a thicker disk and thicker layers of copper.
The result of these changes has been to produce such uniformity of tem-
perature over each face of the disk, that calculation of the mean differ-
ence of temperature between the two faces of the disk has become an
exceedingly simple matter.
The Iron Used,
The disk was made from a slab of cast iron the origin and description
of which are well set forth in the following extract from a letter written
by Mr. A. C. Colby, the metallurgical engineer of the Bethlehem Iron
Company : —
"Dear Sir, — In response to the instructions contained in your letter of
the 30th ultimo, I send to-day . . . the casting which has been made at
these works, and for which no charge will be made to you. I selected a high
silicon iron so as to make the casting free from any chill, and it is smooth
as can be obtained in a sand lined mould, and, I think, very close to the
dimensions you desire, namely, 12" X 4" X 1"-
" In the following composition of the iron entering into this casting, the
sulphur and silicon determinations were made on a gate of the casting.
The other determinations are approximate, based on our daily analyses from
the furnace from which this casting was made : —
Carbon
3.40 -
3.00
Manganese
.50 -
■ .55
Phosphorus
.0.33-
■ .058
Copper
.050-
- .055
Sulphur
.106
Silicon
1.40."
Dimensions and Treatment of Disk.
The diameter of the disk made from this casting was 10. OG cm.; its
mean thickness was called 1.787 cm.; the largest of nine measurements
at different places indicating 1.798 cm., and the smallest 1.776 cm.
The copper plating of the disk was effected by giving it first a thin
coating from a cyanide of copper solution, such as is used by nickel-
platers in preparing iron to receive the nickel, and then finishing the
operation by use of a sulphate of copper solution. Much preliminary
experimenting in this process was done on a block of the cast iron be-
fore the disk was taken in hand, in order to make sure that a good
HALL AND AYRES. — HEAT CONDUCTION IN IRON. 285
and strongly adherent coating would be obtained upon the latter at the
first trial. The final procedure, which worked well, was as follows : —
Two electrolytic baths were prepared, one consisting of a cyanide of
copper solution purchased ready made from a nickel plater, the other
being an ordinary sulphate of copper solution of specific gravity about
1.10, acidulated by about one drop of strong sulphuric acid to ten cubic
centimeters of the solution. Each solution contained two vertical plates
of copper, somewhat broader than the disk to be coated, placed several
centimeters apart.
A hole about O.o cm. in diameter was bored a short distance into the
curved side of the disk, and in this was fixed one end of a steel rod,
which was to serve the double purpose of a handle and a conductor of
the current from the disk. After being rubbed tolerably bright the disk
was boiled in a strong solution of caustic potash for ten minutes, then
rinsed in flowing water, then scoured with powdered pumice and water
by means of a bristle brush, then dipped some seconds in a 20% solution
of hydrocliloric acid, then rinsed again in flowing water, then dipped
again in the acid solution, then rinsed again, then placed between the
two plates of copper in the cyanide of copper solution, which was at a
temperature near 70° C, and kept there half an hour with a current of
about 3.5 amperes flowing through it. At the end of this time the sur-
face of the disk, including its curved side, was well coated with copper.
Accordingly, the disk was taken from the solution, rinsed, covered as to
its curved surface with a rubber band to prevent further deposit of copper
there, then jilaced in the sulphate of copper solution between two copper
plates about 8 cm. apart; and a current of 3 amperes or more was set to
flow through it.
After a number of hours, beads of copper were seen to have formed at
the edges of the two flat faces of the disk, and the disk was removed
from the solution long enough to allow these beads to be broken or filed
off". It was then rinsed, dipped in the hydrochloric acid solution, rinsed
again, then replaced in the sulphate bath and again subjected to the cur-
rent. This course of operations was continued for several days, about
135 hours of current use, until the layer of copper on each face of the
disk was about 0.2 cm. thick. At one stage of the procedure it was
found necessary to resort to the cyanide bath again for a short time, the
filing off of the copper beads at the edges having exposed the iron at
certain parts of the convex surface.
The coatings when completed were somewhat tWcker near the edge
than in the middle. Accordingly, they were turned off in the lathe to a
286 PROCEEDINGS OF THE AMERICAN ACADEMY.
nearly uniform thickness, though one of them was left slightly thicker at
the edge. The final thickness of each coating was not far from 0,2 cm.
The convex surface of the disk was turned off sufficiently to leave a
good surface and show clearly the junctions of the copper coatings with
the iron body. The diameter of the disk was thus reduced to 9.94 cm.
Each coating was now channelled at the edge to a depth of about
0.1 cm. and a width of about 0.17 cm,, as iu Figure 1 of Plate I., and
a brass ring, Ji or H', 0.3 cm. thick, was shrunk into the channel iu
each coating. R' is cut through in Figure 1. The object of this detail
will appear presently.
Mounting and Use of the Disk.
Figure 2 shows how the disk was placed and used in the experiments
on conductivity. In tliis figure, the scale of which is -^-, I represents the
disk; c and c' are the copper coatings; and the rings just described can
be seen set into the edges of the coatings. The lower ring is shown cut
through by horizontal passages. There are, in fact, in this ring 24
horizontal slots, each about 1 cm. long and 0.2 cm. wide, the object of
which is to allow the water entering vertically beneath the middle of the
disk to flow out horizontally from beneath the disk, thus carrying away
the air-bubbles which warm water inevitably contains, and which would
accumulate beneath the disk if an immediate downward escape of the
water through small passages were required. Upon passing from be-
neath the disk the water enters a groove cut in a hard rubber ring, h'h',
and covered by a brass flanged ring n n fastened to h'h'. Thence it
passes downward and out of the apparatus by several passages of 4 or
5 millimeters iu diameter, only two of which are shown in the figure.
The slotted brass ring through which the water flows carries the iron
disk, and rests in a groove in the hard rubber ring h'h', a soft rubber
tube at the bottom of this groove making a water-tight packing. The
ring h'h' has at the bottom another groove, which receives the top of the
brass ring r'r', which rests upon a wooden support to which it is firmly
attached by means of a horizontal flange. Soldered within r'r' near its
top is the brass plate p'p', which carries the hard rubber block H' H', in
the centre of which is fixed the tube that carries the water to the bottom
of the disk. This tube is enclosed for a part of its length by another,
which extends downward from p'p' ; but this is an unimportant detail.
Encircling the iron disk is a soft rubber band b h, which was intended
partly as a protection of the iron against rusting, and partly as a dam to
prevent leakage of water upward past n?i. Another similar band, not
HALL AND AYRES, — HEAT CONDUCTION IN IRON. 287
shown, rested its lov.'er edge upon n n ; but such precaution agaiust leak-
age was perhaps hardly necessary. The downward escape of water from
the groove in h'h' was so free that there was little tendency for it to over-
flow n n.
Starting again at the iron disk and now proceeding upward, we find
hh, H H, and pjo, corresponding in material and in general shape and
position to hJh! ^ H'H', and p'p'., already mentioned. The mere weight
of the apparatus being insufficient to prevent vertical movement and
dislocation under the pressure of the water within, a retaining device was
used, which is described as follows. A flat ring of brass, not shown in
Figure 2, was provided with three internal radial offsets, each of which
bore upon a block of wood resting upon the narrow external flange of
pp. Three brass bolts led from tliis ring to the brass base-plate of the
apparatus, enabling the experimenter to apply to the plate pp any neces-
sary amount of downward pressure.
Certain other parts in the upper portion of the figure require expla-
nation. The parts there shown in dotted outline do not lie in the median
section of the apparatus, and are to be regarded as behind the plane of
the rest of the figure. For example, the vertical tube indicated above
Ji does not rise directly from t/j, but from a horizontal offset extending
from Ji as in Figure 3. Another horizontal offset from J^ receives, as tlie
same figure shows, one end of a plug. Pi, consisting of two semi-cylin-
drical pieces of hard rubber jiressing between them a strip of soft rubber
packing, which packing separates the wires of the copper-German silver
junction j\. J^, Figure 2, is similar to Ji, and contains a similar junc-
tion. More will be said of these junctions later.
Water entering at A flows vertically past the bulb of the thermometer
Ti, which gives a rough indication of its temperature, then horizontally
past the junction in ./j, then by a brass tube into the funnel-shaped
passage FF, then downward through numerous holes near the edge of
pPf and so on, as the arrows show, under ////, upward through the ver-
tical brass tube ^j, which touches the enclosing brass tube U only near the
ends of the latter, past the other copper-German silver junction within
J2, past the bulb of TT,, thence out by means of a rubber tube to the lower
part of the jacket KK, around and upward through this jacket to the
main outlet at 0. The jacket has a supplementary outlet at S ; and the
water from both outlets is collected and weighed below.
Leading upward from the apex of the funnel FF is a small tube,
w, through which a slight waste flow of water is maintained in order to
carry off air-bubbles from FF. Two openings in the top of the water
288 PROCEEDINGS OF THE AMERICAN ACADEMY.
jacket give escape for air, and allow the use of thermometers for taking
the temperature of the water in the jacket.
This water flows over as well as around the enclosed apparatus. The
opening in the double top of the jacket, through which extend the tubes
shown above Jx and J 2 in Figure 2, is about 7.5 cm. by 2.5 cm. Below
Ji and J.2, down to the hard rubber ring h h, the tubes and funnel were
thickly wrapped with cotton to lessen radiation between these parts and
the jacket. The space between h h and h'h', as well as that around and
below hHi', was carefully and fully packed with the same material.
Figure 2 shows two fine copper wires leading out from the coating C,
and passing through holes in the hard rubber ring h h, where they are held
in place by means of hard rubber plugs, k^ and X-o, with soft rubber pack-
ing. There are, in fact, see Figure A, p. 290, thirteen such wires, 0.018 cm.
in diameter, attached to C by electrolytic deposit of copper by a process
sufficiently described in the article referred to in the beginning of this
paper. Wire no. 13 is attached at the centre of C ; nos. 3, 6, 9, and 12
are attached symmetrically about 2 cm. from the centre ; nos. 2, 5, 8, and
11, symmetrically about 3.2 cm. from the centre; nos. 1, 4, 7, 10, sym-
metrically about 4.4 cm. from the centre. Similar wires, nos. l'-13', are
similarly attached to the coating C, no. 1' being immediately beneath
no. 1, no. 2' immediately beneath no. 2, etc. These wires pass through the
ring h'h' exactly as nos. 1-13 pass through the ring /^A. To prevent
deposit of copper upon the free parts of the wires during the process of
attachment, and to prevent illegitimate metallic contacts between the
wires and the coatings C and C during the experiments on conduction,
the wires were coated with shellac between the points of attachment to
the coatings and the places of exit through the hard rubber rings. Out-
side the rings each wire was led to a point on a wooden platform, where
it was, by means of a screw and copper washers, held in firm copper con-
nection with a larger copper wire. The twenty-six larger wires thus
brought into connection led to a like number of small mercury wells in
a board j^laced at some distance from the apparatus shown in Figure 2.
Determination of the Difference of Temperature of
THE Two Faces of the Disk.
The mercury wells were so arranged that by means of copper con-
nectors reaching from one well to another any point of junction on the
upper coating of the iron disk could be thrown into circuit with the cor-
responding point on the under coating and with an astatic galvanometer.
HALL AND ATRES.
HEAT CONDUCTION IN IRON.
289
Care was taken to make the thirteen circuits which could be, one at
a time, thus formed very nearly equal in resistance. It was possible to
use such copper connectors between the mercury wells as to throw all
the wires leading from the upper coating of the disk into multii)le arc
with each other, and all those leading from the lower coating into multiple
arc with each other, and to connect the two sets of wires in one circuit
with the galvanometer. The latter mode of connection was finally used
in the conductivity experiments ; but certain preliminary observations with
the single circuits were made in order to find whether the various pairs
of junctions on the disk were enougli alike in performance to justify con-
necting them in multiple. Tiie method of testing was to run a stream
of water at constant temperature through the apparatus on the under
side of the disk, and another stream at a different constant temperature
through on the upper side of the disk, and to note the galvanometer
deflections obtained from each of the pairs of junctions in turn. The
following table shows the result of the observations : —
Deflections.
Junctions.
13 and 13'
1 and 1'
4 and 4'
7 and V
10 and 10'
2 and 2'
5 and 5'
8 and 8'
11 and 11'
Oct. 23.
4.91
5.31
5.08
4.93
4.71
5.06
5.03
4.76
4.66
Oct 26.
4.38
4.70
4.70
4.47
4.33
Oct. 26.
4.46
4.90 4.87
4.78 4.70
4.55 4.56
4.53 4.40
4.51
4.63
4.40
4.30
Nov 3.
4.86
5.25
5.20
4.80
4.75
5 15
5.06
4.75
4.76
Nov. 3.
4.74
5.19
5.00
4.60
4.55
5.05
5.00
4.68
4.68
Mean.
4.67
5.10]
4.95 '
4.69
4.59
)>4.83
4.89]
4-88 I ^„
4.61,^^-^^
4.55 I
3 and 3'
6 and 6'
9 and 9'
12 and 12'
4.90
4.70
4.66
4.60
4.67
4.67
4.40
4.33
4.59
4.59
4.40
4.35
5.02
4.99
4.76
4.65
4.98
4.98
4.70
4.70
4.83 ]
4.79 '
4.58
4.53 j
}^4.68
An examination of this table, in connection with Figure A, leads to the
conclusion that the mean difference of temperature between the two sides
of the disk increases from the centre to the circumference about 3 per
cent. It appears, too, that the mean difference of temperature between
top and bottom is greater along the radius 1-2-3-13 than along the
VOL XXXIV. — 19
290 PROCEEDINGS OF THE AMERICAN ACADEMY.
radius 4-5-6-13, greater along the latter than along 7-8-9-13, greater
along the last than along 10-11—12-13. The mean difference of tem-
perature along 1-13 apparently exceeded that along 10-13 about 6 per
Figure A.
cent. A partial explanation of this difference probably is that, on one
face or the other of the disk, the flow of water was freer in the region
1-13-4 than in the region 7-13-10. This inequality of flow might be
caused by a slight tilting of HH ox of NH' (Fig. 2).
Two short sets of observations were made to compare the deflections
given by all the pairs of junctions in multiple with the mean of the
deflections given by the pairs used separately. The pair 5-5' was now
found defective, and was omitted from the comparison. The results,
after allowance for the smaller resistance of the multiple arc, were as
follows : —
Nov.
The test of November 20 was the more careful of the two; but even
in this test the difference between the ratio found, 1.014, and unity, was
within the possible limits of error of observation.
The tests which have been described were considered to justify using
the junctions in multiple and treating the current obtained from the
combination as representing the mean thermo-electromotive force of the
From Multiple Arc.
From Single Pair.'.
Ratio.
16
4.47
4.37
1.023
20
4.37
4.31
1.014
HALL AND AYRES. — HEAT CONDUCTION IN IRON. 291
whole disk, and therefore the mean difference of temperature of the two
faces of the iron. There is doubtless some inaccuracy in this conclusion.
Strictly, somewhat greater weight should be given to the indications from
the outer circle of junctions than to those from junctions nearer the centre ;
for an inspection of Figure A will show that each of the outer junctions
represents a somewhat greater area than one of the inner junctions.
The multiple arc method of operation makes no allowance for this fact,
but the error from this cause was probably very small. It is to be ob-
served, moreover, that an error of 1%, for example, in the absolute
value of the thermal conductivity of a particular piece of iron is of no
great consequence in the present state of investigation, provided the
change of conductivity with change of temperature can be determined
with some degree of accuracy.
Before the experiments upon conductivity were made, a number of the
fine wires leading from the faces of the disk having failed, a complete
new set of wires, from the same piece as the first set, was put in by the
same method and in the same positions as before. The apparatus was
then set up once more, in its former condition as nearly as possible.
The strength of the electric current coming from the disk was meas-
ured bj' means of an astatic galvanometer, the sensitiveness, or figure of
merit, of which was determined frequently by sending through it a known
fraction of a current measured by a good tangent galvanometer. The
resistance of the circuit containing the disk and the galvanometer being
known, the thermo-electromotive force producing |he current from the
disk was found. But before this e. m. f. could be translated into differ-
ence of temperature between faces of the iron, it was necessary to deter-
mine by experiment the e. m. f. arising from some known difference of
temperature between two junctions made of copper and of iron like the
iron of the disk. For this test a piece about 10 cm. long and 0.16 cm.
in diameter was cut from the same slab of cast iron from which the disk
had been taken ; and to each end of this slender bar a copper wire, from
the same piece as the wires attached to the coatings of the disk, was
fixed by electrolytic deposit of copper. The bar was set in a hard rub-
ber holder, about 2.5 cm. projecting at each end, and the whole was
mounted between two brass tubes in such a way (see Fig. 4) that water
flowing through either tube would flow over one end of the bar. Thus
water entering at Ai ran past the bulb of the thermometer 7\ along the
end /i of the iron, and out at Ei. An alternative exit for the water is
indicated by the dotted lines below I^. A coating of shellac was used to
protect the iron and the copper from the chemical action of the water.
292 PROCEEDINGS OF THE AMERICAN ACADEMY.
In order to eliminate various sources of i^ossible error, including espe-
cially disagreement of the two thermometers, sets of observations were
made iu pairs, the stream entering at Ai being warmer than the other in
the first set of observations and colder than the other iu the following
set. The ditference of temperature was usually between 5° and 10° C.
Some doubt was felt at first as to whether the ends of the bar would
have the same difference of temperature as the thermometers. It is evi-
dent that use of exits £Ji and j^o, whereby the streams were made to flow
a considerable distance along the bar, would be more effective than the
use of the dotted exits. The latter were used upon occasion, with the
idea that, if they gave about the same effect as £i and SJ2, the latter
could be regarded as satisfactory. The dotted exits gave a result some
four or five per cent less than that given by the other exits. It ap-
peared unlikely that any considerable error would be made in assuming
that, when J^i and iJ.. were used, the difference of temperature of the
ends of the bar was the same as the difference of temperature of the
thermometers.
It will be observed, however, that the iron bar used in this test was
like a piece cut parallel to a certain diameter of tlie disk, not parallel to
the thickness of the disk. It was a matter of very grave doubt whether
the thermo-electric quality of the bar, taken lengthwise, could be re-
garded as identical with the thermo-electric rpuxlity of the disk taken
thickness-wise. It was the latter quality that came into play in the con-
ductivity experiment^ ; and some way of determining it was to be found.
The method which was finally adopted is described in Appendix I. of
this paper. It showed that scarcely an appreciable error would have
been made by using the results obtained from the first method, above
described.
With the information thus obtained it was easy to calculate with con-
siderable accuracy the difference ii» temperature of the two faces of the
iron disk iu the conductivity experiments, the indications of the astatic
galvanometer being readily interpreted. The deflections of the galva-
nometer, upon reversal of the current, were usually about 9 cm. ; and
the difference of temperature of the faces of tlie iron was usually rather
more than 1° C. The difference of temperature of the streams, above
and below the disk, was usually about 8° C.
HALL AND AYRES. — HEAT CONDUCTION IN IRON. 293
Determination op the Difference op Temperature of the Ingo-
ing AND OuTCOMING WaTER AT THE ChAMBEK ABOVE THE DlSK.
The method of making this measurement has already been indicated.
Two copper-German silver junctions, each like that shown in Figure 3,
were used, one at Jx in Figure 2, the other at J^ in the same figure. The
German silver wire used, about 0.015 cm. in diameter, was continuous
from /i, Figure'3, to the corresponding junction in the other plug. Its
length was perhaps 30 cm. The fine copper wire, 0.018 cm. in diameter
(from the same piece as the wires attached to the coatings of the disk),
leading from t/j, Figure 3, did not extend completely through the hard
rubber plug, or holder, but was soldered carefully, some distance from
the outer end of the plug, to a larger copper wire, which led off toward
an astatic galvanometer. The arrangement of copper wires at the other
plug was quite similar. The fine wires of each junction were coated
thinly with shellac.
The two copper-German silver junctions thus described, or similar
ones,* were tested, or " calibrated," by means of streams of water, of a
known difference of temperature, flowing past the junctions according to
the arrows in Figure 3. The ditierence of temperature of the streams was
found by means of the same pair of thermometers that are indicated in
* In accorrlance with my advice, I\Ir. A3Tes made 07ily such experiments in the
calibration of liis copper-German silver junctions as to show that their performance
differed but little from that of similar junctions used previously by myself. After
this, in all his calculations of the conductivity, he took his values of the thermo
e. m. f. of copper-German silver from Figure 8 of my previous paper, already re-
ferred to, "On the Conductivity of Mild Steel." In preparing the present paper, I
have had some misgivings as to the accuracy of these values, and therefore in
October, 1898, I made more experiments upon a pair of junctions quite similar to
those used by Mr. Ayres. Tlie results are given in the second column below.
The third column gives values, for the same temperatures, taken from the figure
used by Mr. Ayres : —
Mean Temn Electromotive Force, in Volts, per I'' C. DifiFerence of
Temperature of Junctions.
20° 3 .00001743 .00001732
37°.9 .00001826 .000018.30
58°.2 .00001944 .00001942
From the old observations and the new combined a curve representing the ther-
mo e. m. f. at temperatures ranging from 15° to 65° was constructed, and the values
of the conductivity found by Mr. Ayres were revised accordingly. The resulting
changes of conductivity were slight, but they had considerable effect upon the
estimated temperature coefficient of conductivity. — E. H. II.
294 PROCEEDINGS OF THE AMERICAN ACADEMY.
Figure 4, and the same method of alternating the hotter and cooler streams
was used here that was used in the test of the copper-iron junctions.
The difference of temperature of the streams in the calibration tests was
usually about 4° or 5° C.
In the conductivity experiments proper, the usual difference of temper-
ature of the copper-German silver junctions, the usual difference of tem-
perature, that is, of the ingoing and outgoing water, was probably rather
more than 0°.5 C.
The Flow of Water.
The method of controlling and heating the streams of water was
essentially the same as that described in the previous paper,* and illus-
trated in Figure 5 of that paper. Powerful gas-burners, of a type
manufactured by the Buffalo Dental Comijany and expressly intended
for heating streams of water, were used. Each stream flowed through the
conduction apparatus from the base of an overflowing standpipe, which
device answered the double purpose of insuring a constant flow, and
allowing air bubbles to escajie from the water before reaching points
where they would do harm. A supplementary air-vent was provided
for the upper stream near its entrance at A, Figure 2.
The stream which flowed above the disk, the only one upon which
careful measurements were made, ran into a covered barrel standing
upon a platform balance. The time of flow was noted, and the amount
of water accumulating in the barrel during that time was determined by
weighing. The rate of delivery of the stream ranged, during the whole
course of the investigation, from about 15 grams per second to about
25 grams per second. The stream flowing beneath the disk was of the
same order of magnitude.
Results and Discussion.
A (ew sets of observations were made at low temperatures without the
use of the jacket. A few others were made at various temperatures with
use of jacket, but without the cotton packing within and below it. These
observations were preliminary, and none of them will be used in dedu-
cing the final results. The detailed results of subsequent observations,
made with jacket and cotton packing in use, are given below in chrono-
logical order. None are omitted, although some are placed in brackets
for reasons to be given later.
* Tliese Proceedings, Vol. XXXI., 1896.
HALL AND AYRES. — HEAT CONDUCTION IN IRON. ,295
The " meau temperature " is the mean between the temperature of the
upper stream upon entering the apparatus (as indicated by the ther-
mometer Ti in Fig. 2), and the lower stream upon leaving * the apparatus.
Neither of these temperatures was taken with great accuracy, aud any
one of the mean temperatures given may be wrong to the extent of 0°.5,
Date, 1897.
Mean Temp.
Ki
A's
K
May 11
21°
0.1471
0.1495
0.1483
" 13
19°.6
0.1489
0.1558
0.1524
'' 15
20°.4
0.1503
0.1511
0.1507
«' 18
39°.l
0.1485
0.1489
0.1487
" 20
40° .9
0.1512
0.1520
0.1516
" 25
22°
0.1522
0.1515
0.1519
" 26
40°.2
0.1482
0.1494
0.1488
" 28
20°.6
0.1533
0.1507
0.1520
" 29
35°.7
0.1523
0.1494
0.1509
June 7
74°
0.1421
0.1309
0.1365
" 8
72°.9
0.1523
0.1559
0.1541
" 21
77°.3
0.1382
0.1400
0.1391
" 23
2r.7
0.1536
0,1465
0.1501
" 26
56°.2
0.1506
0.1407
0.1457
[July 1
55°.5
0.1741
0.1443
0.1592]
[ " 2
61°.6
0.1584
0.1188
0.1386]
[ " 8
58°.5
0.1564
0.1425
0.1495]
" 23
54° .2
0.1518
0.1487
0.1503
" 24
57°.3
0.1485
0.1423
0.1454
'' 26
28°.4
0.1496
0.1504
0.1500
[ " 30
74°.6
0.1368
0.1588
0.1478]
Aug. 3
56^5
0.1427
0.1441
0.1434
4
27°.5
0.1557
0.1528
0.1543
" 7
27°.0
0.1519
0.1515
0.1517
8
59°.2
0.1470
0.1446
0.1458
Under K^^ are given values of the conductivity obtained from observa-
tions made when the vrarmer stream ran above the disk. Under Ko are
given values of the conductivity obtained from observations made when
the warmer stream ran below the disk. A' is the mean of K^ and /iTo.
In the calculation of the values here given, no account was taken of the
* The temperature of the lower stream on entering the apparatus was not taken;
but the change of temperature within the apparatus was slight.
296 PROCEEDINGS OF THE AMERICAN ACADEMY.
variation of the specific heat of water with variation of temperature, this
specific heat being called 1 for every temperature used. This inaccuracy
will be referred to again.
These figures show a considerable range of temperature, and from
them it should be possible to derive an approximate value at least of the
temperature coefficient of K. The numbers given in brackets, however,
will not be used for this purpose. The numbers for July 1, 2, and 8
exhibit great differences between Ki and iu, and also between the values
of K. On those days, and those only, the pair of copper-German silver
junctions, used to determine the change of temperature of the upper
stream, were covered with shellac melted on. The coating thus obtained
was too thick, so that the junctions did not take the temperature of the
water with sutficient readiness. There are in the table above given
other values of K obtained at temperatures not very different from
those at which these rejected values were found. The values obtained
for K ai all temperatures above 70° differ much among themselves; but
it hardly seems best to reject them altogether in the attempt to arrive
at an approximate value of the temperature coefficient of K. The great
variation observed among them was probably due to unsteadiness of
temperature of the water streams when very hot, or to possible impair-
ment by the hot water of the shellac coating on the copper-German
silver junctions.
All the values of K not contained within the bracketed lines will be
used in some fashion in estimating the temperature coefficient ; but they
will be used in two divisions, one for May and June, the other for July
and August. The reason for this division is that on August 5 some of
the lines of wire leading from the copper coatings on the iron disk were
found to be out of condition. The pairs of wires affected were 1, 2, 11,
and 12, the other nine pairs remaining in good condition. When this
partial breakdown began it is impossible to determine ; June 30 was
the last date on which all the pairs of wires were known to be in good
order. It has been shown in the early part of this paper that each pair
of wires gave about the same effect as any other pair ; therefore, as all
were joined in multiple, the failure of a few of them should affect the
total current but little, the resistance of the remaining pairs being but a
small part of the total resistance of the circuit. The failing pairs lay,
one in the outermost circle, two in the next, and one in the next. It
appears, from a comparison of the values of K obtained near 21° and
near 39° in May and June with the values obtained near 28° in July
and August, that the impairment of the wires or some other unknown
■ HALL AND AYRES. — HEAT CONDUCTION IN IRON. 297
cause made the later values at a given temperature about one per cent
greater than they would have been had they been obtained at the same
temperature in May or June. In these earlier months sets of observa-
tions were made at various temperatures from near 20° to near 75°. In
July and August sets near 28° were intermingled with sets near 57°. It
is possible, therefore, to make for each period an independent determina-
tion of the temperature coefficient of K.
We have from the May and June division : —
Date.
Mean Temp.
A-.
K,
K
May 11
21°
0.1471
0.1495
0.1483
" 13
19°.6
0.1489
0.1558
0.1524
" 15
20°.4
0.1503
0.1511
0.1507
« 25
22°
0.1522
0.1515
0.1519
« 28
20°. 6
0.1533
0.1507
0.1520
June 23
2r.7
0.1536
0.1465
0.1501
20° .9
0.1509
0.1509
0.1509
May 18
39°.l
0.1485
0.1489
0.1487
" 20
40°.9
0.1512
0.1520
. 0.1516
" 26
40°.2
0.1482
0.1494
0.1488
" 29
35°.7
0.1523
0.1494
0.1509
38°.9
0.1501
0.1499
0.15U0
June 26
56° .2
0.1506
0.1407
0.1457
June 7
74°
0.1421
0.1309
0.1365
« 8
72°.9
0.1523
0.1559
0.1541
« 21
77°.3
74°.7
0.1382
0.1400
0.1391
0.1442
0.1423
0.1432
From July
and August we hav
e : —
Date.
Mean Temp.
K,
K,
Mean K
July 26
, 28°.4
0.1496
0.1504
0.1500
Aug. 4
27°.5
0.1557
0.1528
0.1543
" 7
27°.0
0.1519
0.1515
0.1517
27°. 6
0.1524
0.1516
0.1520
July 23
54°.2
0.1518
0.1487
0.1503
" 24
57°.3
0.1485
0.1423
0.1454
Aug. 3
56°.5
0.1427
0.1441
0.1434
" 8
59°. 2
0.1470
0.1446
0.1458
56°.8 0.1475 0.1449 0.1462
298 PROCEEDINGS OF THE AMERICAN ACADEMY.
The single set of observations, made July 30, at a temperature near
75° is hardly worth taking into account here, the uncertainty of observa-
tions at such a temperature being great, as we have seen.
According to the evidence thus far we have, from the May and June
observations.
at 20°. 9
K= 0.1509
" 38°.9
" = 0.1500
« 56°.2
« = 0.1457 (?)
" 74°.7
" 0.1432 (?)
August observations.
at 27°.5
^=0.1520
" 56°.8
" = 0.1462
As the change of A' with change of temperature appears to be small in
any case, it becomes important to consider the change of specific heat of
water with change of temperature ; for all values which precede are
given on the assumption that the specific heat of water is 1 at all tem-
peratures used.
Winkelmann, in Part II. of Volume II., p. 340, gives a table of the
specific heats of water, which he has deduced from a formula proposed
by himself after a discussion of the results obtained by numerous experi-
menters. Til is table gives : —
Temp.
Sp. Heat.
Temp.
Sp. Heat.
o°c.
1.0000
50° C.
0.9939
10°
0.9944
60°
0.9992
15°
0.9924
70°
1.00G7
20°
0.9910
80°
1.0164
25°
0.9901
90°
1.0283
30°
0.9898
100°
1.0424
40°
0.9907
Revising, in accordance with this table, the values of K last given
above, we get,
at 20°.9 K= 0.1494
" 38°.9 « = 0.1485
" 56°.2 " = 0.1453 (?)
" 74°. 7 " = 0.1447 (?)
" 27°.6 " = 0.1505
" 56° .8 " =0.1458
HALL AND ATRES. — HEAT CONDUCTION IN IRON. 299
The difference between the value of ^ found for 20°. 9 and that found
for 38°. 9 is so slight that very little importance can be attached to it, in
view of the much greater differences between successive measurements
of K at or near any one temperature. Making the formal calculation,
however, from these values as they stand, we get, as the temperature co-
efficient of K between 20°.9 and 38°.9,
0.1494 — 0.1485 ^ ^^^„„
— —0.00033.
0.1494(38.9 — 20.9)
Taking the mean of 20°.9 and 38°.9 and the mean of 0.1494 and 0.1485,
we have, at 29°. 9, A'= 0.1490. Taking this as a starting point, we
find, for the tempeiature coefiicieut of iiT between 29°.9 and 56°.2,
0.1490-0.1453 __o.ooo94.
-0.00064.
0.1490(56.2 — 29.9)
Similarly, we find between 29°.9 and 74°.7,
0.1490 — 0.1447 _
0.1490 (74.7-29.9;""
So much for the IMay and June numbers.
From the July and August numbers we get, between 27°. 6 and 56°. 8,
0.1505—0.1458
= —0.000107.
0.1505(56.8 — 27.6)
The mean of all these estimates of the temperature coefficient of K is
— 0.00075, according to which the thermal conductivity of the cast iron
disk diminishes about 1% for each 18°. 3 rise of temperature within the
limits of the observations above recorded. According to experiments
described in Appendix 11. following this paper, the temperature coefficient
of electrical conductivity of the same cast iron, between 17° and 67°,
is — 0.00118; which means that the electrical conductivity between
these limits diminishes at the rate of 1 % for each 8.°5 rise of tempera-
ture. At one time during the preparation of this paper it appeared that
the two temperature coefficients were very nearly equal. This led to a
more careful examination of the evidence than had been made before,
and a repetition of certain measurements, with the result given above.
It may yet be that the two coefficients are equal. Where both are so
small the question of equality or inequality is difficult to settle, although
300 PROCEEDINGS OF THE AMERICAN ACADEMY.
a new series of experiments with same cast iron disk would probably
give results much more concordant than those set down in this paper. In
wrought iron the temperature coefficient of electrical conductivity is
much greater than in cast iron, and if the tempei*ature coefficient of
thermal conductivity is correspondingly large in wrought iron, fairly ac-
curate measurements of this latter coefficient should be attainable with
this material. A disk of wrought iron will probably be put to the test
before long. The disk of mild steel used in the experiments described
in a preceding paper was very like wrought iron in many respects ; but
it has already been stated, in the first part of this paper, that the experi-
ments with this disk were not entirely satisfactory, the disk itself and its
copper coverings being too thin for the best effect.
The experiments of this paper have given a larger value of K, for the
piece of cast iron dealt with, than was expected. It is much larger than
the value, about 0.105, found some years ago for two specimens of cast
iron near 115°C. by one of the authors * of this paper, using the method
of Forbes. It is much larger than the values found by Kohlrausch and
by AViedemann and Franz for soft steel near 15°. Nevertheless, there
seems to be no good reason for doubting the substantial accuracy of the
value of K found in this paper. The most novel, and perhaps the most
doubtful, feature of the method here described is the use of the iron
itself as part of a thermo-electric element. How carefully the thermo-
electric behavior of the iron with respect to copper has been considered
will be apparent to the reader of Appendix I.
Another subject of possible doubt is the amount of error caused by
neglect of radiation or convection between the water jacket, Figure 2, and
the apparatus surrounded by it. The value found for K is affected, 1st,
by such interaction as occurs between the jacket and the disk ; 2d, by
that between the jacket and those surfaces which lie above the disk and
below Jx and J<,_. The mean temperature of the curved surface of the
disk was probably four or five degrees below the temperature of the
jacket when the warm stream ran above, and nearly an equal amount
above that of the disk when the cold stream ran above. The area of
this surface between the two hard rubber rings h h and //// was about
50 sq. cm. Preston, "Theory of Heat," p. 4G1, gives, as found by McFar-
lane for a blackened sphere suspended within a water jacket 5° cooler
than itself, ^''heaf. emitted per second, per degree difference of temperature^
* E. H. Hall, in these Proceedings, 1892, p. 202.
HALL AND AYRES. — HEAT CONDUCTION IN IRON. 301
per square centimeter in water-gram units " equals 0.000252. Assum-
ing this rate of emission or absorption for each of the 50 sq. cm. of the
curved surface of the disk, we should get for the passage of lieat per
second 50 X 0.000252 X 5 = 0.0G3 units. The heat passii.g per second
through the disk from face to face was usually about one hundred times
as much as this. If we as&ume, for the moment, that the passage of heat
through the curved surface of the disk is equal to 1 % of that which Hows
from face to face, we may thereupon reason as follows. When the warmer
stream flows above the disk, the disk takes in heat from the jacket, and
the total amount passing out through the lower face of the disk exceeds
by 1% the amount flowing in at the upper face. The inflow at the
curved surface distorts the isothermals and lines of flow within the disk
in such a way that, with a given difference of temperature between the
faces, the flow from fixce to face, which is equal to the inflow at the upper
face, is less than it would be if the flow within the disk were adiabatic,
that is, if there were no inflow at the curved surface. On the other
hand, the outflow at tlie lower face is greater than the adiabatic flow
from face to face would be. We may conclude that, under the conditions
assumed, the actual inflow at the upper face is about 0.5% less than the
adiabatic flow would be with like temperatures at the faces, and that the
outflow at the lower face is about 0.5% greater than the adiabatic flow
would be. Our method of calculating ^assumes adiabatic flow, while
our observations give us the inflow at the upper surface. Accordingly,
under the conditions here assumed, the value obtained for K would be
about 0.5% too small. A similar error, in the same direction, would be
made with the colder stream flowing above the disk, so that we could
not eliminate it by combining two sets of observations. In fact, however,
the assumption that the rate of transmission at each square centimeter
of the curved surface of the disk, thickly wrapped with cotton, is the same
as that found by McFarlane for a bare blackened surface of copper,
gives a very large overestimate of the possible error from this source.
Turning now to the surfaces between the top of the disk and the parts
Ji, Jo, Figure 2, we find the area of these surfaces to be about 300 sq.
cm. No systematic observations of the temperature of the jacket were
kept during the experiments of which this paper gives an account, but
from previous observations it appears probable that the mean tempera-
ture of the jacket was about 1° C. lower than the temperature of the
parts which we are now considering, when the warmer stream ran
above, and 0°.5 or less higher than the temperature of these parts when
the colder stream ran above. Assuming the difference of temperature
802 PROCEEDINGS OF THE AMERICAN ACADEMY.
to have been 1° C, and assuming, for the moment, the same rate of sur-
face transmission which we have used above, we get, as the amount of
heat passing per second between the parts considered and the jacket,
300 X 0.000252 = 0.0756 units. This is, jjerhaps, rather more than
1% of the heat carried from face to face through the disk, and if it were
a fair estimate of the actual transmission between the jacket and the
surfaces considered, the neglect of this transmission would make K, as
calculated, about 1 % too large. This error would not be eliminated by
combining sets of observations, some with the warmer stream above, and
some with the colder stream above. But in this estimate, as in that
relating to the action at the curved surface of the disk, the rate of
transmission assumed is no doubt much too large, the surfaces enclosed
by the jacket being, for the most part, well wrapped with cotton.
It seems, therefore, unlikely that any considerable error was made
by neglecting the intercliange of heat between the water jacket and the
apparatus within it.
There is little doubt that much more concordant values of K than
those given in this pa])er can be obtained by a somewhat more careful
control of the temperature of the water, and by making each set of
observations longer than the sets, often very brief, which were made in
the investigation which has here been described.
Summary.
The thermal conductivity of the cast iron used is about 0.1490 at
30° C. The temperature coefficient of thermal conductivity, if Winkel-
mann's rule for the change of specific heat of water with temperature is
correct, appears to be about —0.00075 between 20° and 75°, so that
a rise of about 13° C. corresponds to a fall of \% in conductivity.
If the change of specific heat of water between 30° and the higher
temjjeratures up to 75° were neglected, the value found for the temper-
ature coefficient would be about —0.0010.
The electric conductivity of this cast iron is about 112,200 in c. g. s.
units. (See Appendix II.)
The temperature coefficient of its electric conductivity between 17° and
67° is about —0.00118.
The method used appears to be capable of giving better results than
have yet been obtained by it.
HALL AND AYRES. — HEAT CONDUCTION IN IRON. 30i
APPENDIX I.
Measurement of the Thermo-Electric Quality op
Short Iron Bars.
"When it became necessary to determine, relatively to copper, the
thermo-electric quality of the cast iron disk thickness-wise, the problem
appeared to be one of some difficulty. The thickness of the disk was
about 1.8 cm. The thickness of the slab from which the disk had been
taken was such that bars 2 cm. long could be cut from it thickness-wise;
but to make satisfactory thermo-electric measurements upon a single bar
of this length appeared to be impracticable. The device of putting a
number of such bars end to end, so as to make a column of considerable
length, and placing this column lengthwise between two blocks of copper
of different temperatures, seemed a hopeful one; but it had to be put
to the proof before it could be used with confidence.
Accordingly a very soft magnet core rod, about 0.16 cm. in diameter,
was taken, and from it were cut one piece 15 cm. long and ten pieces
each 2 cm. long. Copper wires were soldered to the ends of the 15 cm.
piece, and this piece was then mounted very much as the piece I^L is
mounted in Figure 4. The parts exposed to the streams were now, how-
ever, some 5 cm. long, about twice as long as the exposed parts in sim-
ilar preceding tests. A thin coating of paraffine was now used to protect
these parts from the chemical action of the water.
The ten 2 cm. pieces, after being carefully cleaned and polished at the
end surfaces, were placed end to end in a wooden tube, which in all its
dimensions was much like the wood of a common pencil from which the
graphite has been taken out. The iron, corresponding in position to the
graphite of a pencil, projected from the wood about 0.2 cm. at each end.
R R in Figure 5 (Plate II) shows in diagonal lines a section of the
wooden tube, or rod, the iron witliin being indicated by a heavy black
line ; the scale of the figure is \. Water jackets, ^ and J^ in Figure 5,
surrounded R R for the greater part of its length. The iron column
projecting from R R was pressed between two copper blocks B^ B^ and
B^ B^, through which flowed streams of water at any temperature re-
quired. The pressure was applied by means of a wooden plunger p, sup-
ported in the block 5, and pushed against ^2 -^2 by ^ fairly constant force.
The blocks B^ B^ and B^ B„ were of the same diameter as the jackets ji
and J2, and all of these objects rested in a slot cut lengthwise in a piece
304 PROCEEDINGS OP THE AMERICAN ACADEMY.
of hard wood. S, S, S, in the figure, are parts of this wooden support
which would have been cut through by a vertical longitudinal section
through the middle of the apparatus. Certain edges of the support which
do not form part of the section shown are indicated in the figure by
light dotted lines. Behind, to the left of, the block Bi B^^ is shown a
copper wire TFj, about 0.1 cm. in diameter, which extends through the
centre of a wooden rod r and bears against Bi B^, thus making a back-
stop for the pressure exerted at the other end of the apparatus. From
the block B^ B.2 another copper wire, W^, held in firm contact with Bo B,,
leads away. The wires Wi and W2 are parts of the thermo-electric cir-
cuit of the apparatus, and are in metallic connection with the terminals
of a galvanometer. B^ Bi is provided with a water jacket Ji J^, the
construction of which is indicated by certain lines in Figure 5 and in
Figures 6, 7, and 8. Thus, Figure 6 shows a vertical cross-section
through Ji Jx near the left end of ^1 B^, the dotted lines indicating cer-
tain edges not lying in this section. Figure 7 shows a vertical cross-
section through t/i ^1, through B^B-^. and through the thermometer T^
(Fig. 5). Figure 8 shows a horizontal section through J^ J^ and Bi By.
The block B2 B2 is protected by a jacket quite similar to Ji J^. Wads of
cotton were used to protect certain parts of each block which were not
covered by the jackets.
The course of the water through the apparatus is indicated by arrows.
Thus, at the left hand the streams enters at A^ passes down along the
bulb of Ti through B^ i?i, thence by a rubber tube, longer in fact than
in the figure, to J\J\^ thence by another rubber tube to j^-, and out at
E]^. The flow of the right hand stream is strictly analogous. Each
stream usually carried 20 or more grams of water per second. The
thermometers T-^ and To were the same that were used with the 15 cm.
bar of iron and in previous tests of thermo-electric junctions. Sets of
observations at a given mean temperature were made in pairs, one set
having T^ the warmer, the other set having To the warmer.
It was necessary to give careful attention to the electrical resistance of
the column of short iron bars ; for it could not be safely assumed that
this resistance would be either small or constant. It was found, naturally,
to depend somewhat upon the magnitude of the pressure applied at the
ends of the column. In the experiments upon soft iron which we are
just now considering, the pressure was exerted by means of a compressed
piece of india-rubber tubing, not shown in Figure 5. In later experi-
ments, with cast iron, it was applied through a lever as in Figure 5,
the force F being exerted upon the end of the lever by means of a
HALL AND AYRES. — HEAT CONDUCTION IN IRON. 305
spring balance at a point too far down to be shown in place. "With the
rubber tube in use the pressure against the end of the column was per-
haps 1.8 kilograms. When the lever and balance were used, it was some-
times about three kilograms and sometimes less.
The various jacketing and protecting devices shown in Figure 5 were
not all used at first, and in the early experiments on soft iron the e. m. f.
obtained from the column of short bars was several per cent less than
that obtained from the 15 cm. soldered bar, with a given difference of
temperature between the two thermometers. This discrepancy gradually
diminished as the method of experimentation was improved, until at last
it became little or nothing, as the following numbers, obtained with the
system of jacketing shown in Figure 5, will testify.
Wrrn Soldebed Bar.
With Column
OP Shoht Bars.
Date, 1898.
Mean Temp.
E. M. F.
Per Decree
DilT. of Temp.
Mean Temp.
E. M. F.
Per Degree
Diff. of Temp.
April 9
14°.o C.
18°.4
[214.7]
214.2
14°.5
18°.4
[209.6]
212.6
" 20
20°. 9
213.2
20°. 9
211.0
U ((
16°.9
215.8
17°.0
214.3
" 22
18°.4
214.2
18°.4
214.4
The e. m. f. is here given in terms of a purely arbitrary unit. The
values in brackets were obtained under conditions of special uncertainty
as to resistance. Considering the final trials of the end-to-end short bar
method satisfactory, I prdceeded to apply it to cast iron. From the end
of the slab that furnished the conductivity disk a slice was cut crosswise,
about 10 cm. long, 2.5 cm. wide, and 0.3 cm. thick. This was cut up
into 26 parts, and each of these parts was turned down to a thickness
of about 0.16 cm.; or, rather, 18 of them were so treated, the other
8 being broken at some stage of the operation. They were then boiled
for about 20 minutes in a strong solution of caustic potash, partly
to free them from oil, partly because the disk had been thus heated
before its conductivity was tested. In all of this work an attempt Avas
made, and I think a successful one, to keep the bars in the same order
with respect to each other that they had before being cut from the slice,
so that I could at the end tell what bars had been taken from near the
end of the slice and what ones from near the middle.
When ready for the test of thermo-electric quality, I rubbed the flat
ends of each little bar bright with infusorial earth, and wiped them care-
fully ; for it is evident that a particle of dirt or of vegetable fibre left
VOL. XXXIV. — 20
806 PROCEEDINGS OF THE AMERICAN ACADEMY.
upon one of the ends may break altogether the electric circuit of which
the column of bars should form a part. In later work it seemed better
to rub the ends with fine emery paper, and then wipe them upon smooth
hard-finished paper to remove adhering particles of dust.
Bars 1, 3, 5, 7, 9, 11, 14, 17, 19, and 23, the numbers indicating their
order from one end toward the other of the slice from which they had
been cut, were placed in the order just given, end to end in the appar-
atus shown by Figure 5. The resistance of the column, which in the
case of soft iron bars had been about 2 ohms under a pressure of 1.8
kgm., was now found to be surprisingly large. It diminished with in-
crease of pressure, but even with a pressure of 3 kgm. was at first, June
30, about 16.5 ohms. Under a nearly continuous application of this
pressure it gradually grew less, until, on July 2, it was about 5 ohms,
after which it changed but little, although it appeared to be somewhat
greater on July 4.
With this set of bars, and with the method already described, the follow-
ing results were obtained : —
Date. P'? °^ "^^^P- Mean Temp. E. m. f. in Volts per Degree,
between Lads. "^ r o
June 30, 1898, 9°.84 29°.3 ) .00000549 1 ......,,
July 4, " 10°.21 30Mr--' .00000549 P^^^^^^*^
" « " 11°.97 45°. 2 .00000593
" 2, " 13°.40 C3°.2) •^00^0^491
" 4, « 15M3 63°.0i^^-^ .00000043 r^^^^^^^*^
On November 4, 1898, observations were made in the same way with
ten cast iron bars taken from the same set as those used in June and
July ; whether the same bars or not, could not be told. The result was :
Date. Diff. of Temp. Mean Temp. E. m. f. per Degree.
Nov. 4 10°.24 29°.7 .00000550
It should be remembered that the end to end method of experimenta-
tion with short bars of cast iron was adopted because of a doubt as to
the availability of the thermo-electric test made by a different method on
a 10 cm. bar cut crosswise from the cast iron slab. This earlier method
had given : —
Mean Temp. E. m. f. per Degree.
14° .00000507
18°.6 .00000518
40°.6 .00000576
63°.3 .00000647
HALL AND AYRES. — HEAT CONDUCTION IN IRON. 307
A comparison of these results with those obtained by the end to end
method with short bars, cut thickness-wise from the shib, shows that the
two methods gave almost identical results. Of course, it is possible that
the bars used in the two methods differed considerably in thermo-electric
quality, and that some error in one or the other method compensated for
and obscured this difference of quality ; but it is much more reasonable
to conclude that the slab from which all of the bars were cut had practi-
cally the same thermo-electric quality crosswise as thickness-wise, and
that the accuracy of each method of testing this quality is affirmed by its
concordance in results with the other method. The results of both
methods were used for plotting a line from which values of the copper-
iron thermo e. m. f. could be derived for purposes of interpolation. This
line is a curve ascending with increase of temperature, and slightly con-
cave upward. The divergence of this line from true rectitude is prob-
ably not very significant. There is in the corresponding curve for the
thermo e. m. f. of the copper-German silver junctions, described in
the preceding pages, a divergence of about the same relative amount in
the same direction. It is possible that this peculiarity of both lines is
due to some idiosyncrasy of the thermometers used in the thermo-electric
tests. The same thermometers were used in all these tests ; and tliere-
fore, as the method of calculation of conductivity involved the ratio of
tlie e. m. f. of copper-iron and copper-German silver, no final error as
to conductivity results from any small imperfections of these thermom-
eters. E. H. II.
APPENDIX II.
Measurement of Electric Conductivity of the Cast Iron.
One of the 2 cm. bars described in Appendix I. was used for this
determination. Four copper wires were attached to this bar by elec-
trolytic deposit of copper. Two of the wires were about 0.08 cm. in
diameter ; these were attached to the flat ends of the bar, and served to
carry in and out an electric current of about 0.25 ampere. The other
two wires were much finer, about 0.018 cm. in diameter; these were at-
tached at two points about 1.7 cm. apart, each being about 0.15 cm. from
one end of the bai", and were used for making connection with a poten-
tiometer. The bar was submerged in oil during the measurements.
The temperature of the oil was controlled by water flowing through
308 PROCEEDINGS OF THE AMERICAN ACADEMY.
a lead tube bent into solenoidal form. The bar was placed horizontal
within the solenoid, the axis of which was vertical.
The electrical resistance in absolute c. g. s. measure was about 112,200
at 17°. 4.
From the observation of September 20, 1898, the temperature-coefRcient
of conductivity between 20°.9 and 61°.2 appeared to be —0.00120. The
observations of September 27, between 17°.4 and 67°. 4, gave —0.00116.
We may take the mean, —0.00118. In both cases the coefficient was
calculated by the formula
C e-ffi ' Cond. at high temp. — cond. at low temp,
Cond. at low temp. X (high temp. — low temp.)
without reference to 0°.
E. H. H.
Hall and Ayres. — Heat Conduction in Iron.
Fig. 6
Plate ii.
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. No. 12. — February, 1899.
ON THE OPTICAL CHARACTERS OF THE VERTICAL
ZONE OF AMPHIBOLES AND PYROXENES; AND
ON A NEW METHOD OF DETERMINING THE
EXTINCTION ANGLES OF THESE MINERALS BY
MEANS OF CLEAVAGE PIECES.
By R. a. Daly.
With Three Plates.
ON THE OPTICAL CHARACTERS OF THE VERTICAL
ZONE OF AMPHIBOLES AND PYROXENES: AND ON
A NEW METHOD OF DETERMINING THE EXTINC-
TION ANGLES OF THESE MINERALS BY MEANS OF
CLEAVAGE PIECES.
By K. A. Daly.
Presented by John E. Wolff, December 14, 1898. Received December 15, 1898.
It is evident that, in an optical examination of monoclinic minerals
with coincident optical and crystallographic planes of symmetry, the
angle of extinction on (010) is a highly important datum, for it is indis-
pensable to a knowledge of the shape and orientation of the ellipsoid of
elasticity. The difficulty of preparing sections in the plane of symmetry
of amphiboles and pyroxenes, and the practical impossibility of doing so
in the case of very small crystals, lead the student of these species to
revert to those good natural sections, cleavage plates, and inquire as to
the relation subsisting between the extinction on sections parallel to (010)
and that on prismatic cleavage flakes. This relation is not simple, and
it was long ago demonstrated that, accepting Fresnel's optical theorem,
the extinction on (110) is dependent in a complex way on the angle of
extinction on (010) and on the optical angle. Michel Levy made it clear
that, for pyroxenes, the extinction on (110) would always be less than
on (OlO), since the latter is the maximum possible value of extinction
read against cleavage cracks on any section in the vertical zone. On
the otlier hand, it was shown for the negative amphiboles that among
the infinite number of possible sections made by a plane revolving in
the vertical zone from (010) to (100), there is one which has the highest
value of extinction in that zone, and that this value decreases as the
revolving plane moves toward (010) or (100).* It is interesting to
determine whether an amphibole with this property of showing a max-
imum of extinction for positions of the rotating plane between (010)
and (100) could have an extinction-angle on (110) greater than that on
* Fouque and Michel Levy, Mineralogie Micrographique, p. 368.
312
PROCEEDINGS OF THE AMERICAN ACADEMY.
(010). It seems to have been taken for granted by some writers that
this is not true.*
It is the object of this paper to make a systematic statement of the
relation between the extinction on (110) and on (010), and to indicate
under what conditions the former may be used to determine the latter.
First, there will be deduced a formula to express the extinction on any
plane of the vertical zone when the optical angle 2 Fand p, the angle of
extinction on (010), are known, and a graphic representation of possible
extinctions in that zone will be attempted. Then specific tables will be
introduced to show how 2 V, p, and 6' are related for any variety of
pyroxene or negative amphibole, where 6' represents the extinction on a
cleavage plate of that variety. Secondly, a method is proposed for the
determination of p by means of 6' and a new angle of extinction, 6",
found after turning the cleavage piece through a certain angle about the
vertical axis.
Figure 1
In Figure 1, let ECB be the plane of symmetry of any monocHnic
crystal characterized by parallelism between (010) and the optical plane.
Let E C G hQ any plane in the vertical zone cutting the plane of sym-
* Rosenbusch, Physiographie der petrographisch wichtigen Mineralien, 3te
Auflage, 1892, p. 550. Zirkel, Lehrbuch der Petrographie, 2te Aufl., Bd. I. p. 302.
DALY. AMPHIBOLES AND PYROXENES. 313
metry in C E, and B C G a, plane cutting the vertical zone at right
angles. Let P represent the vertical crystallographic axis, D the
direction of extinction on (010), A (9 and B the optic axes. Let 0' P^
be the rectangular projection of P on B C G, 0' representing the sim-
ilar projection of on B G. Further, let O'D' be the direction of
extinction on BOG, and A' (/ and B' 0' respectively the lines of inter-
section of the planes A 0' and BOO' with the plane B C G. Let
the angle B C B' ^ C, ZP A = a, ZP B = /3, ZP'O'A' = a',
ZP'd'B' = /3'. We have also ZP'O'B' = 6 and ZPOB^p.
This construction represents an application of Fresnel's fundamental
law that, for a biaxial mineral, the direction of extinction in any section
is found by taking the intersection of the plane of the section with that
which bisects the angle between the two planes respectively containing
an optic axis and the line (ray of light), perpendicular to the section.
It follows that
a + ^ a — ft
and 6 =
tan a ■=
tan/3 =
2 2
2
tan a
cos G '
tan^'
(1) tan 2 p =
cos O
tan a — tan /?
1 + tan a tan /3 '
(2) tan2^=,^^"'"^-/""^>"l^.
^ ^ 1 + tana tan ^' cos^ O
Equation (2) is a special case of Michel Levy's general formula on
page 65 of the " Mineralogie Micrographique," and also of the new for-
mula by Cesaro for expressing the extinction on any face of a zone.*
The mathematical treatment of these expressions shows that, as the sec-
tion moves continuously from the position (010) to the position (100),
it will pass through an intermediate position of maximum extinction if
the angle 2 F be less than 90°, or the angle (a + /3) greater than 90°.
Thus, has a maximum when cos c = .f This is true of
V tan a tan ft
* Me'moires de TAcademie Royale des Sciences, etc. de Belgique, 1895, Vol. LIV,
p. 26. Kef. in Zeit. fiir Kryst., etc., 1897, p. 181.
1 The position of tiie section where an extinction may be observed between
314 PROCEEDINGS OP THE AMERICAN ACADEMY.
all amphiboles so far as known, except pargasite and a very few other
varieties ; with these exceptions, each has a maximum in a section far
removed from the plane of symmetry. A striking example is to be found
in an interesting amphibole rich in ferrous oxide from Philipstad, Sweden.
It has a pronounced zonal structure ; all the zones extinguish together in
(010) at 15° 9', but at different angles on (110), the latter varying from
21° to 17°, corresponding to different and unusually small optic angles
in the respective zones of from 50° to 60°.
Plates I., II., and III. represent diagrammatically the variation which
may be observed in the behavior of certain negative amphiboles that
have such maxima of extinction, — namely, those with optical angles of
50°, 60°, 70°, and 80°, and each characterized by extinctions on (010)
of 10° (Plate I.), 15° (Plate IL), and 20° (Plate III.). The abscissa rep-
resents the angle of rotation of the section out of the plane of symmetry,
the ordinate indicates the corresponding angle of extinction. Arrow-
heads show which plane of the vertical zone possesses the maximum
extinction peculiar to each curve, and also the value of that maximum.
The diagrams clearly show that the maximum extinction observable in a
rock-slide examined for one of these amphiboles would be far from repre-
senting the extinction-angle on (010), and data regarding extinctions and
pleochroism derived from the study of thin sections would be worthless, if
not controlled by this principle. For the sake of comparison, the anal-
ogous curves for amphiboles with 2 V= 90° and p respectively equal to
10°, 15°, and 20°, appear in the plates. It will be seen that there is no
position of maximum extinction between (010) and (100).*
Now, among the i^ossible positions of the movable plane, there is one
which surpasses all others in interest except that of the plane of sym-
(010) and (100) just equal to that on (010) may be found from the following un-
publislied formula by Dr. A. C. Lane of Houghton, Mich. : —
1 — cos- p — cos'2 V
sin X — — — — >
cos- V — cos- p
where x is the angle made with (100) hy the required section.
* It is characteristic of all the curves, that the angle of extinction changes very
slowly in passing out from (010). This is important in the study of rock-slides,
since a section may be removed several degrees (even 30° when the optical angle is
large) out of the plane of symmetry, and but small error would be made in using
its value of extinction as equivalent to that on (010). It would, in that case, be
only necessary to be sure that the section is really in tlie vertical zone, as ascer-
tained by the parallelism of cleavage cracks. That it is near tlie position (010)
can, of course, be proved by the absence of a well defined hyperbola in con-
vergent light.
DALY. — AMPHIBOLES AND PYROXENES.
315
TABLE I.
s
B
2F =
= 50°.
2F =
.60°.
2V =
= 70°.
2 V =
= 80°.
2V
= 90°.
o'
o"
50
o"
o'
rHhH
CD
o"
o'
S5
o*
o"
CD
•J2 C
II
11
II
II
1!
II
II
II
II
II
W
O
o
o
o
Ci
o
o
o
O
O
/
/
O '
'
'
O '
O '
o '
O '
o '
2
2 37
2 2
2 16
133
158
1 13
143
1
132
51
3
3 55
3 8
3 23
2 19
2 57
1 50
2 35
130
2 18
1 16
4
5 14
4 6
4 32
3 7
3 56
2 28
3 27
2 1
3 4
142
5
6 33
5 9
5 40
3 55
4 56
3 6
4 20
2 31
3 51
2 7
6
7 52
6 15
6 49
4 44
5 56
3 44
5 12
3 3
4 37
2 33
7
9 12
7 22
7 58
5 34
6 56
4 23
6 5
3 34
5 24
3
8
10 32
8 32
9 8
6 26
7 57
5 3
6 58
4 6
6 11
3 26
9
1153
9 44
10 18
7 20
8 58
5 44
7 52
4 39
6 59
3 54
10
13 14
11
1129
8 15
10
6 26
8 46
5 13
7 47
4 21
11
14 36
12 19
12 40
9 12
11 2
7 10
9 41
5 47
8 36
4 50
12
15 58
13 42
13 52
10 12
12 5
7 55
10 37
6 23
9 24
5 18
13
17 20
15 9
15 5
11 15
13 9
8 42
1132
6 59
10 14
5 48
14
18 44
16 42
16 19
12 21
14 14
9 31
12 29
7 37
11 4
6 18
15
20 7
18 19
17 33
13 31
16 19
10 22
13 27
8 16
1155
6 50
16
2131
20 2
18 49
14 45
16 26
11 16
14 25
8 57
12 47
7 22
17
22 56
2152
20 5
16 3
17 33
12 12
15 25
9 39
13 39
7 56
18
24 21
23 48
2122
17 25
18 41
13 11
16 25
10 24
14 32
8 30
19
25 46
25 45
22 39
18 53
19 50
14 14
17 26
11 10
15 26
9 6
20
27 12
27 58
23 58
20 27
21 1
15 21
1.8 28
1159
16 21
9 44
21
28 37
30 12
25 17
22 6
22 12
16 31
19 31
12 50
17 17
10 23
22
30 3
32 30
26 36
23 52
23 24
17 47
20 36
13 45
18 14
11 4
23
31 29
34 53
27 57
25 44
24 37
19 7
2141
14 42
19 12
1147
24
32 54
37 18
29 18
27 43
25 51
20 32
22 47
15 43
20 11
12 32
25
indet.
indet.
30 39
29 48
27 7
22 2
23 55
16 48
21 11
13 19
metry itself. I refer to the plane of prismatic cleavage. Table I.
contains the results of calculations (repeated applications of equation 2),
intended to find the value of 0, (6'), for amphiboles when Cis equal to half
the cleavage angle (taken at 62^°), 2 Fat successive values of 50°, 60°,
70°, 80°, and 90°, and p with all values of whole degrees from 2° to 25°
inclusive. Thus, if any two of the three terms 2 V, p, or 6' be known,
the third can be determined. The values of $' are tabulated in the col-
316
PROCEEDINGS OF THE AMERICAN ACADEMY.
umn headed " 62|°." This table is regarded as capable of more general
application than the one published by Harker,* inasmuch as his refers
only to that small class of amphiboles which have the acute optical angle
situated about the axis of elasticity lying next to the vertical axis (op-
tically 2)ositive, represented by pargasite). It will be observed that for
amphiboles with small optical angle (up to 60°), any one of the three
variables may be determined if the other two are known, but that, con-
sidering the instrumental errors of reading, the larger values of the
optical angle cannot with much safety be determined on account of the
slow variation in the corresponding angles of extinction on (010) and
(110). When 2 Fis equal to 70°, p and 0' are for any negative species
nearly equal.
The following is the similar table (Table II.) for pyroxenes, where the
cleavage angle is taken at 92° 54'. It is simply a slight extension of
that of Harker. t
TABLE IT.
Angle of
extinrtion
on (010).
Values of fl'.
Angle of
extinction
on (010).
Values of 6'.
2V=5lO°
2F=60°
2r=70°
2F=50°
2F=603
2V=10°
o
o
o
o
o
35
29i
30i
sn
45
89
401
41|
36
30i
3H
32f
46
40
411
42f
37
m
32^
33|
47
41
42i
44
38
32J
33J
34|
48
42
43i
45
39
33i
m
35f
49
43i
44i
46
40
34J
35i
36|
50
44i
45^
47
41
35
36i
37|
51
451
46|
48
42
36
37i
38f
52
461
47f
491
43
37
38^
39f
53
471
48|
501
44
38
39J
40|
54
481
49f
5H
We are not yet, however, in a position to make universal use of cleav-
age pieces for the purpose of finding the value of p. There are many
* Extinction-angles on cleavage-flakes. Mineralogical Magazine, 1894, Vol. X.
p. 2.39.
t Op. cit., p. 240.
DALY. — AMPHIBOLES AND PYROXENES. 317
rock-forming pyroxenes and amphiboles which, owing to the small size
of the crystals or their very friable nature, it is extremely difficult, if
not impossible, to cut in the directions necessary to obtain p and 2 V.
On the other hand, good cleavage flakes are almost always to be had, and
it is by the use of these that I propose the following method of finding
the extinction on the clinopinacoid of an amphibole or pyroxene.
A perfectly flat cleavage piece, thick enough to give the greatest pos-
sible definiteness to the position of extinction and showing clearly marked
cleavage cracks, is laid on an object-glass with the broadest face down.
It is then carefully mounted on the stage of a two-circle FedorofF table.
With the vertical circle set at zero, the stage of the table is turned so
as to bring the cleavage cracks of the specimen into a position parallel
to the axis of the vertical circle.* This axis should be parallel to the
principal section of either polarizer or analyzer. By taking the average
of a number of good readings, the extinction angle is now obtained.
Following this operation, the vertical circle is turned in such a direction
that the plane of symmetry of the crystal is more oblique to the polarized
ray and by an angle nearly approaching that at which the specimen
would begin to slide on the object glass. I have found that 15° is a con-
venient amount of rotation, and that angle will be used in the following
discussion. Extinction is again read in this new orientation with the
greatest possible care.
We have in this way determined two special angles of extinction (0'
and 6"), corresponding to two jjlanes in the vertical zone, which are at
different angles (C and C" = C + 15°) to (010). It is now possible
in a simple way to eliminate a and 13, and thus permit of the determina-
tion of p directly from 6' and 6". To this end, we have the following
series of transformations, which I owe to Mr. J. K. Whittemore of
Harvard University.
Substituting in (2) the special values of and C, we have,
- ., (tan a — tan B) cos C"
(3) tan 2 6' = ^, "^ ^-^7/ ; and
^ ^ 1 -f tan a tan )8 cos^ 6" '
(4) tan 2 6
Then
(5)
(tan a — tan ft) cos C
1 + tan a tan /3 cos" C"
tan 2 p 1 + tan a tan /? cos'^ C"
tan 2 (^' ~ (1 -f tan a tan /3) cos C
* The Nachet form of the table or the simpler model by Fuess is best for the
purpose.
318 PROCEEDINGS OF THE AMERICAN ACADEMY,
tan 2 p 1 + tan a tan /8 cos^ C"
(6)
tan 2 6" (1 + tan a tan {3) cos C
1 — M cos C" 1 — w cos C"
(7) tan a tan ^ = P^t — -^ = j^, ^-^ .
^ ^ MOOS 6^— cos' C rcosC — cos^C
tan 2 6' — cos C tan 2 p _ tan 2 9" — cos (7'' tan 2 p
^ -^ tan 2 p cos C" — cos'^ C tan 2 6' ~ tan 2 p cos (7" — cos^ C" tan 2 6'
Let tan 2 p = x, tan 2 6' = x' , tan 2 Q" = x' . We have
x' — X cos C" x" — X cos C"
^^^ a; cos C - x' cos'-' C ~ x cos (7" - x" cos^ C" '
Hence
(10) X {x' cos C" sin^ C - x" cos C" sin^ C") = x' a:" (cos'^ C" - cos^ (7') .
Substituting the values of x, x\ and x",
_ tan 2 g^ tan 2 &' (cos'' C^^ - cos' C)
(11) tan 2 p - ^^^ ^ ^, ^^^ ^,„ ^.^, ^„ _ ^^ 2 ^„ ^^^ ^, ^.^^2 ^„ .
Reducing to a form more convenient for the use of logarithms, we have
' _ tan 2 & tan 2 &' sin ( C + C^'Q sin ( C - C")
(12) tan 2 p _ ^an 2 6/' sin^ C cos C" - tan 2 6/" sin^ 6'" cos C '
We now have an expression by means of which the extinction can be
calculated even when the optical angle is not known. The theory of
the same problem was elaborated by Liebisch in his discussion * of the
general determination of the optical angle and of the direction of the
optic axes in biaxial minerals, using sections whose cry stall ographic
orientation is known, and on which the planes of vibration of the two
refracted rays are known. Of immediate interest to us is the case in
the monoclinic system where the optical plane is also the plane of sym-
metry, but the orientation of the two axes of elasticity is not determined.
He finds the problem soluble when two sections can be employed, and
this is what we practically have for optical purposes in the cleavage
piece placed at two positions with respect to the incident ray from the
polarizer.
But it is evident that there must be some degree of error in the instru-
mental readings of 0' and 6"; we have next to inquire if equation (12)
is so sensitive to changes in 6' or 6" or to simultaneous changes in both
as to make any determinations of p by means of it valueless. By ac-
* tJber die Bestimmung der optischen Axen durch Beobachtung der Schwings-
richtung ebener Wellen. Neues Jahrbucli fur Min., etc., 1886, Bd. I. p. 155.
DALY. — AMPHIBOLES AND PYROXENES.
319
tually trying certain cases, I have found that p changes for small errors
in 6' and $", but that the rate of change is not sufficiently rapid to make
the method of no practical use. Indeed, the errors in p may in certain
cases be little greater than the original errors of reading 6' and 6". A
Zillerthal actinolite may be taken as an example.
Let C = 62° 15', and 2 r=: 80° (an actinolite). Then 6' = 13° 27';
and, if C" — 77° 15' (15° of rotation of the vertical circle), then 6" =
8° 16'. In the attempt to find p from 0' and 6", errors of various
magnitudes in 0' and 0" were introduced as follows.
TABLE in.
0' = 13" 27'
+
Error of
0" = 8^ 16'
Error of
p deter-
miued as
Error
in p.
1
0°
+ 10'
14° 31'
-29'
2
0°
+20'
14° 8'
-57'
3
0°
+30'
13° 39'
- 1° 21'
4
+10'
+10'
15° 1'
+ ¥
5
+20'
+20'
15° 2'
+ 2'
6
+30'
+30'
15° 5'
+ 5'
7
+ 1°
+ 1°
15° 11'
+ir
8
+10'
-10'
16° 8'
+ 1° 8'
9
+20'
-20'
17° 28'
+ 1° 28'
10
+80
-30'
19° 2'
+ 2° 2'
The limit of error in reading extinction-angles for colorless cleavage
flakes, or for those that do not display very deep absorption, may be
placed at 20'. At this limit, in our example, there is seen to be an error
of rather less than 1° for an error in 0", with 0' exactly determined.
When 6" and 6" show errors of reading in the same direction (in the
example, both plus), they may be large without making an equally large
error in p. This is important, since any fault in the construction of
the microscope will thus equally affect both readings and need not greatly
influence the value of p, even though found by means of so sensitive an
320 PROCEEDINGS OF THE AMERICAN ACADEMY.
equation. On the other hand, if & and 0" vary from the truth in oppo-
site directions, p quickly changes, a fact which is evident from an inspec-
tion of the expression for tan 2 p. The table shows an error of 1° 28'
in p when 6' is 20' too large and 6" 20' too small ; and again an error
of +2° 2' in p for corresponding errors of +30' and —30' in 6' and 6".
If the rotation of the vertical circle had been in the opposite direction
through the same angle, so as to make C" = 47° 15', the errors for cases
1, 2, 3, 8, 9, and 10 of the foregoing table would have been considerably
less ; those for cases 4, 5, 6, and 7, on the contrary, somewhat greater.
The curves of Plates I., II., and III. show, however, that the extinction
angles for each of the different amphiboles in sections cut at respective
angles of 47" 15' and 62° 15' to the plane of symmetry would be nearly
the same, and that the variation in the extinction-angle at 47° 15' in pass-
ing from one amphibole to another would be much slower than that
peculiar to the 77° 15' position.* Hence I have chosen the latter as the
more useful ; in Table I., in the columns headed " C = 77° 15'," will be
found the values of extinction angles characteristic of the same amphi-
boles whose extinctions on cleavage pieces have already been calculated.
By the use of the whole table, a first approximation to the value of the
extinction-angle on (010) can be rapidly made without the necessity of
going through the rather tedious application of equation (12).
Analogous results characterize the introduction of errors into the 6'
and 6" of pyroxenes. I have chosen an Ala diopside with C = 43° 33'?
2 K=: 59°, and p = 36°. Then, revolving the cleavage-face (110) 15°
away from (010), we have C" = 58° 33'. 6' was calculated to be 31° 16',
and 6" = 26° 8'. Introducing arbitrary errors in 6' and 6", we obtain
the results of the accompanying Table IV.
Generalizing from the two error tables, supplemented by the inspection
of equation (12), we can reach certain conclusions regarding the in-
fluence of instrumental errors. Equation (12) is least sensitive for
errors in 6' and 6" when these are either both plus or both minus and
equal or nearly equal. When equal, p can be more accurately deter-
mined than by direct measurement on a section in the plane of sym-
metry (given the use of the same microscope in both cases, as well as
equal thicknesses, absorption, etc. for the cleavage piece and cut section).
* It is, of course, evident that both readings (at 47^° and 77|^°) can be taken on
the same cleavage piece ; and also that amounts of rotation other than 15° may be
advantageously employed. Experiment shows that the cleavage piece will not
slide on the object-glass even at an angle of 20°, and thus 6 may be determined for
the 421° and 82^° positions.
DALY.
AMPHIBOLES AND PYROXENES.
321
TABLE IV.
9' = 8V1&
+
Error of
0" = 26=8'
+
Error of
p deter-
mined as
Error
iu p.
1
0°
10'
35° 46'
-14'
2
0°
20'
35° 32'
-28'
3
0°
30'
35° 18'
-42'
4
10'
10'
36° 8'
+ 8'
5
20'
20'
36° 18'
+18'
6
30'
30'
36° 27'
+27'
7
1°
1°
36° 57'
+57'
8
10'
-10'
36° 36'
+36'
9
20'
-20'
37° 15'
+1° 15'
10
30'
-30'
37° 52'
+1° 52'
"When the errors of 6 and 6" are unequal, the resulting error in p is
more significant and is most unfavorable when the errors of reading the
extinction in the two positions are of opposite sign. When equal
numerically, the error in p for actinolite and diopside is about four times
that made by directly reading the extinction on a plate cut parallel to
(010). The necessity for accurate measurement of 6' and 6" is evident.
In actual practice, moreover, several independent determinations of p
should be carried out and an average taken to secure the safest results.
Two examples may suffice to show the kind of results obtained by
the author in the application of the method here outlined. The Philip-
stad hornblende already referred to was studied with reference to the
extinction on (010), first by the use of two carefully oriented sections
made by M. Werlein of Paris, then by the use of cleavage pieces. The
former gave an extinction on (010) of 15° 5' (white light). Extinction
was then read on a cleavage flake and found to be 20° 53', as the result
of averaging many readings. This flake was then turned into the posi-
tion where the plane perpendicular to the ray of light from the polarizer
made an angle of 77;^° with the plane of symmetry of the crystal. Ex-
tol. XXXIV. — 21
822 PROCEEDINGS OP THE AMERICAN ACADEMY.
tinction was read in this position against the cleavage trace, at an angle
with the latter of just 20° (average of ten readings). Substituting this
value and the value of extinction on (110) in equation (12), I obtained
15° 3' as the corresponding value of p. This very close correspondence
with the determination of p by the use of the oriented sections is not
less than accidental, but the example clearly shows that cleavage pieces
may be made to yield information concerning the value of extinction on
the plane of symmetry as useful for practical purposes as that derived
from a section in that plane.
A second example was found in a hornblende, which was given me for
examination by Professor J. E. Wolff of Harvard University. It occurs
associated with much pyroxene and biotite in the classic coarse theralite
of Theralite Peak in the Crazy Mountains. Optical study of the horn-
blende in the numerous slides which have been made of the rock is
difficult, on account of the scarcity of the mineral in large individuals.
It would be quite impossible to make oriented sections even if the grains
could be freed from the much more abundant pyroxene, with which they
are commonly intergrown. Apart from the absolute small amount of
the hornblende present in the rock and from the fact of intergrowth with
another silicate, this seemed to be a particularly unfavorable case for the
application of our method, in that the absorption is strong, and errors in
6' and 0" should appear rather larger than characterize readings on such
an amphibole as Zillerthal actinolite, for example. The rock was pul-
verized and, after some search, suitable cleavage pieces were discovered
in the powder, and manipulated as in the last example. This time, the
curve of extinctions was examined on four points besides that at the
position (010); namely, at the positions, 42:^°, 47J°, 771°, and 82^°.
Extinctions were read for these at respective values of 29°, 29° 30', 31°,
and 30°. The cleavage position gave 34°. The average value of p now
determined by substituting the readings in the general equation is 28° 5'.
Now, I was fortunate enough to find in one of the thin sections a longi-
tudinal section of the hornblende, evidently in the vertical zone since the
cleavage cracks were rigorously parallel to one another, and very near
the plane (010), inasmuch as there was practically no trace of a hyper-
bola in convergent light. Careful reading of the extinction afforded
an astonishingly close approximation to the value of p just determined,
viz. 28°. It is possible that the true angle is half a degree or more
greater or less than that, but the calculated value is in any case near that
observed in* the rock -slide. It may be noted in passing that this is a
rather remarkable hornblende, from the fact that its extinction angle is
DALY. — AMPHIBOLES AND PYROXENES. 323
abnormally high ; the optical angle is also unusual, for calculation shows
that it must be about 55°. Unfortunately, for reasons speciQed above,
the mineral cannot be separated and chemically analyzed.
In the process of working out these illustrations, the author has come
to the conclusion that, with the proper safeguards, the method can be
safely applied to the amphiboles in general ; and that the pyroxenes can
be similarly treated. It may thus prove to have more than mere theo-
retical interest.
R. A. Daly.
0° 10° 20° SO
40°
Amphiboles and Pyroxenes.
50° 60° 70° 80° 90°
2 V= 00
2 F = 80
2 r= 70'
ihlj j!i
2 v=m'-
2 r= 50
Plate I. — Showing the curves of extinction in the vertical zone of Amphiboles
for which the extinction on (010) is 10° and the optical angle is 50°, 60°, 70°, 80°, or
90°. The abscissa represents tlie amount of rotation of a plane in the zone away
from the plane of symmetry. The ordinate is the corresponding angle of extinc-
tion for that position of the movable plane. The position and amount of maximum
extinction are represented by arrow-heads. Tiie values of extinction for 10° inter-
vals in the zone are entered in the diagram.
R. A. Daly. Amphiboles and Pyroxenes
0° 10° 20° 30° 40° 50° 60° 70° 80° 90°
2 V= 90°
2 V:
2V= 70°
2 F=60o
2 V= 50°
0^ 10° 20° 30° 40° 50° 60° 70° 80° 90°
Plate II. — Sliowing curves formed on the same plan as those of Plate I. The
extinction on (010) is here 15°.
R. A. Daly. Amphiboles and Pyroxenes.
0'^ 10° 20° oO° 40^ 50-" (i()^ 70° 80° 'J0°
2 I' = 90
2 1' = 80
2 V = 70^
2V= eo-^
2 V = 50^
0" 10' 20" oO° 40° 60° 50° 70° 80° 90'
T-7TTT— r-rn-rrrTT-i
Plate III. — Showing curves formed on the same plan as those of Plate I.
The extinction on (010) is here 20°.
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. No. 13. — February, 1899.
CONTRIBUTIONS FROM TPIE CHEMICAL LABORATORY
OF HARVARD COLLEGE.
A REVISION OF THE ATOMIC WEIGHT OF NICKEL.
SECOND PAPER. — THE DETERMINATION OF THE NICKEL IN
NICKELOUS BROMIDE.
By Theodore William Richards and Allerton Seward Cushman.
CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.
A REVISION OF THE ATOMIC WEIGHT OF NICKEL.
SECOND PAPER. — THE DETERMINATION OF THE NICKEL IN
NICKELOUS BROMIDE.
By Theodore William Richards axd Allerton Seward Cushman.
Presented October 12, 1S98. Received December 29, 1S98.
In a recently published investigation upon the atomic weight of nickel
we gave three series of results, depending upon the ratios of silver and of
argentic bromide to nickelous bromide.* These led us to the conclusion
that the atomic weight in question could not be far from 58.09 if oxygen is
16.000. Since however no single method is ever convincing, and since the
results above mentioned did not represent all that might be done even with
one method, it was evidently advisable to pursue the matter to a more defi-
nite conclusion. Accordingly, we determined to continue the work with a
twofold object in view: first, to study further the preparation of pure
material, and secondly, to complete the analysis of nickel bromide in such
a way as to determine directly the amount of nickel as well as the amount
of bromine in the salt. No surer test of a quantitative result than such
a complete analysis is known. The careful study of all the conditions
and results has led us to the conclusion that the preparation of pure
nickelous bromide is an unusually difficult problem ; but the problem has
been so nearly solved that we know exactly the precautions and correc-
tions necessary to make this substance serve as an accurate basis for the
determination of the atomic mass of nickel. In the light of this later
knowledge we find that the results of the last paper need a slight cor-
rection, but it is satisfactory to note that the change amounts to not
much over one unit in the second decimal place of the atomic weight.
While carrying out the work, we had continually in mind the various
controversies which have arisen over the atomic weights of nickel and
* These Proceedings, XXXHI 97.
328 PROCEEDINGS OP THE AMERICAN ACADEMY.
cobalt ; and feeling that the " Gnomium " question raised by Kriiss and
Schmidt had never been conclusively laid at rest, v/e naturally dwelt espe-
cially upon it. Unfortunately, one cannot enter into a discussion of this
subject without directly antagonizing many views which have been ex-
pressed on one side or the other ; but of course the following results are
recorded solely in the interests of truth, without controversial bias.
The balance and weights used in the work described were the same as
those described in the former paper. The weights were re-standardized,
with results very similar to those found a year before. All weighings of
nickelous bromide were reduced to the vacuum standard by the addition
of 0.114 milligrams per gram to the observed weight. The specific
gravity of nickel (about 8.7) is so near the specific gravity of brass (8.4)
that the correction of the nickel to the vacuum standard is less than one
part in a hundred thousand, hence it may be omitted.
Experiments concerning the Purity of the Materials.
The purification of our nickelous material has already been described
at length.* Most of the work described below was done with nickel
which had been purified by Mond's process and many subsequent opera-
tions (Sample Ill.t), but two analyses were made with a somewhat less
pure sample (No. II. t) made from commercial material. Since we had
proved that further protracted treatment produced no effect on the com-
bining weight, evidently these specimens were quite pure enough for
our purpose.
Kriiss and Schmidt used glassware in their preparation work, there-
fore it seemed worth while to make an exact observation of the well
known danger involved in this practice. To this end two very carefully
treated specimens of spongy nickel were prepared, one having been made
wholly in platinum, and the other wholly in the best Bohemian glass,
which had been thoroughly steamed. The former of these preparations,
which was supposed to be absolutely pure, left upon sublimation as bro-
mide in a stream of bromine vapor only a very minute siliceous residue.
The specimen of nickel which had been prepared in glass vessels was
totally different in appearance from the one prepared in platinum ; in-
stead of being metallic and coherent, it was black and powdery. Upon
the conversion of about ten grams of this dark powder into nickelous
bromide, a beautifully iridescent voluminous residue, weighing about five
milligrams and consisting mainly of silica, was left in the boat. Evi-
* These Proceedings, XXXIII 102. t Ibid , 105.
RICHARDS AND CUSHMAN. — ATOMIC WEIGHT OF NICKEL. 329
dently this silica had prevented the cohering or " sintering " of the metal
during its original reduction from the oxide, and hence caused the pul-
verized state of the impure metal, — a fact which is interesting as
showing the great change in properties produced by a small amount of
impurities.
These two experiments emphasize the well known facts that glass is
wholly unsuitable for accurate work, and that material prepared even in
platinum is extremely difficult to I'ender wholly free from silica, unless
it is vaporized.* While however our purest nickel sometimes contained
traces of silica, the bromide prepared from it by sublimation was un-
doubtedly as free as possible from this impurity. Further light upon
this question will be given in a following paper upon cobalt.
Although a possible contamination with silica was thus little to be
feared, some other constituents of glass or porcelain were much more
dangerous. All the material actually used in our analyses had been
prepared wholly in platinum vessels at every stage excepting at the very
end, when it had been sublimed in a porcelain tube. Obviously this
tube might be attacked by the hot mixture of hydrobromic acid, bromine,
and nickelous bromide vapor ; but since nothing beside sodic bromide
would probably sublime with the nickelous salt, and the "equivalent" of
sodic bromide is almost equal to that of nickelous bromide, the slight
impurity could produce no important effect upon our last year's work.
This was realized at the time ; and the discovery of the presence or
absence of this error was one of the prearranged objects of the present
paper. Since the work recorded below was completed, Professor Wink-
ler, by a kind personal letter as well as by a recent article, wisely called
attention to this flaw ; and it will be seen that his objection had been both
substantiated and answered before he wrote about it.
The easiest method of detecting sodic bromide in nickelous bromide
is obviously to reduce the latter and then to extract the former with
water from the spongy metal. Moist hydrogen easily divorces the halo-
gen from its none too stable metallic union at a temperature of not much
over 300°, at which temperature sodic bromide is essentially non-volatile.
In this way repeated experiments showed that all our nickelous bromide
had contained on the average not far from one tenth of one per cent of
sodic bromide. The particulars concerning the determination of this
serious impurity naturally form an essential point in the method of anal-
ysis of the nickel salt, hence they will be found later under that head.
* Compare Stas's " Untersuchungen " (Aronstein), pp. 269 and 279.
380 PROCEEDINGS OF THE AMERICAN ACADEMY.
The presence of an unreduced bromide was first detected by Mr. Baxter
in the course of his work on cobaltous bromide, and some interesting
details involved in its discovery will be recorded in the paper upon that
subject. The amount present varied with the temperature used in the
sublimation, but was otiierwise surprisingly constant. The concurrent
sublimation of the two salts is undoubtedly similar to the distillation of
organic substances with steam, sodic bromide possessing a small constant
vapor-tension at the constant temperature of about 900° used in the
sublimations.
After the completion of a series of reductions of nickelous bromide
containing this impurity of sodic bromide, a final attempt was made to
obtain the salt of nickel in a state of absolute purity. We expected
that platinum would be attacked by the mixture of bromine vapor, hydro-
bromic acid, and nickelous bromide, which exists in the red-hot tube
during the sublimation of the salt, but platinum is the last resort in cases
of this kind. In order to sacrifice as little of the precious metal as pos-
sible in our desperate experiment a large porcelain tube was lined with
platiimm foil,* and inside of this was placed a platinum boat containing
the metal to be converted into bromide. In each of two separate speci-
mens of nickelous bromide made in this apparatus merely a trace of
sodium was found, but unfortunately enough platinum was present to
render the results valueless. They are not included in the tables below.
Only a small strip of the foil was injured, the very hot parts and the cool
parts being alike untouched. After these exj)«riraents we abandoned
the attempt to prepare absolutely pure nickelous bromide, and returned
to the use of the porcelain tube for the sublimations ; for sodium
is an impurity much more easily weighed than platinum, under the
circumstances. The only method of obviating the difliculty would have
been to use a tube of nickel for the sublimation ; but the obtaining and
moulding of a large amount of the metal in a perfectly pure state prom-
ised to be so troublesome that we have not yet attempted this improve-
ment.
The arrangement for supplying and purifying the large volumes of
hydrogen needed in this research was gradually evolved from the simple
form used in the first experiments to an elaborate piece of apparatus
which will be described in detail in the paper upon cobalt.
The preparation of bromine and of the other materials has been de-
* This idea was suggested by Professor H. B. Hill. Compare also Penfield,
Zeitschr. Anorg. Chera., VII. 22.
RICHARDS AND CUSHMAN. — ATOMIC WEIGHT OF NICKEL. 331
scribed in sufficient detail in other papers,* so that no further words
need be wasted upon these points. It is almost unnecessary to state that
in these simple operations no loophole was left open through which an
error might creep in to destroy the value of the more dithcult under-
taking before us. We are indebted to the Cyrus M. Warren Fund for
Chemical Research in Harvard University for some of our more expen-
sive pieces of apparatus.
Thb Method of Analysis.
At first many attempts were made to determine nickel by electrolysis,
with the hope that nickelous bromide might be analyzed in this simple
and direct fashion. In order to test the metliod, weighed amounts of the
purest spongy metal were dissolved and reprecipitated electrolytically.
The spongy metal had been prepared by boiling the purest platinum-
made amraonio-nitrate with much water, igniting and reducing the pre-
cipitate with pure ammonia, and heating the metal in a vacuum. The
weight of nickel deposited by electrolysis always exceeded that of the
pure nickel taken, hence the electrolytic method was abandoned as un-
suitable for work of the highest accuracy. The excess of weight, which
was noticeable even when the film was heated to 120° before weighing,
and often exceeded two tenths of one per cent when it was dried at 50°
after the method of Winkle i",t was traced to inclusion of mother liquor
between the film and the dish, and to the probable presence of occluded
hydrogen in the nickel. t Since the deposit was beautifully metallic
and coherent in appearance, one might well have expected a better
result. It is possible that these observations may help to explain Wink-
ler's high values for the atomic weight of nickel and cobalt, since he
used the electrolytic method. On the other hand, the spongy metal
which had been used in our experiments was probably purer than that
prepared in any other way, for solid impurities had been rigorously
excluded, and the traces of gas present had been pumped out.
These preliminary experiments showed that the best method of deter-
mining the amount of nickel in the bromide would be to reduce it in a
stream of hydrogen, provided that the reduction could be accomplished
without the loss of any of the bromide by volatilization. Following in
the footsteps of Mr. Baxter's work with cobalt, it was found that moist
* These Proceedings, XXXIII 106 et seq.
t Zeitschr. Anorg. Chem., IV., 22.
X Raoult, Compt. Rend., LXIX. 826 ; Bottger, Dingler's Polytech. Journ.,
CCI. 80 (1871).
332
PROCEEDINGS OF THE AMERICAN ACADEMY.
hydrogen answered the purpose ; for the temperature at which the reduc-
tion takes place is so low that no trace of nickel was found outside of the
boat which originally contained the bromide, if a rapid current of the
gas was maintained. It is obvious that according to the law of mass-
action, the presence of a large proportion of hydrobromic acid resulting
from the reduction would tend to prevent the desired reaction, and hence
to facilitate the undesired sublimation ; therefore a large excess of hydro-
gen must be present. Of course the hard glass tube used for this pro-
cess was always afterwards treated mternally with nitric acid, and the
liquid was examined with minute care for traces of nickel.
Fig. 1. Apparatus for igxitixg Nickelous Bromide in ant desired
Mixture of Gases.
The use of rubber was confinefl to the first part of tliis train, where it could do
no harm (A B C D E F and A M N O P).
The nickelous bromide to be analyzed was contained in a platinum
boat, and the method of drying and weighing it was in every respect the
same as that described in detail in the previous paper upon this subject.
After having been weighed, the boat was carefully placed in a hard glass
tube, in which the bromide was cautiously reduced to the metal. When
the reduction was completed and the apparatus had just cooled, the boat
and its contents were returned to their weighing bottle, where they were
enclosed in an atmosphere of dry air. The weight of the residue was
RICHARDS AND CUSHMAN, — ATOMIC WEIGHT OF NICKEL. 333
found after half an hour, and then at the end of many hours. Since the
boat was cool before having been introduced into the bottle, two suc-
cessive weighings thus made never differed from one another by amounts
beyond the limit of error of weighing. Spongy nickel evidently does
not oxidize in dry air.
It is obviously a matter of great importance to discover whether or not
the material prepared in this way contains weighable amounts of occluded
hydrogen. According to the experiments of Neumann and Streintz*
who worked with reduced metals, two cubic centimeters of the gas were
occluded by each gram of spongy nickel. This amount would alter the
observed atomic weight by only about the fortieth of one per cent ; but
since we are aiming at even greater accuracy, the matter should evidently
be probed to the bottom.
In the first place, carefully weighed nickel remaining from one of the
analyses recorded below was ignited in a Sprengel vacuum at perhaps
550°, For fear of losing some sodic bromide a higher temperature
could not be employed. No appreciable loss of weight occurred and no
gas was evolved during this process, which was repeated with several
specimens ; hence there seemed to be good reason to believe that no
hydrogen was occluded by the metal in our experiments. In order to
prove the matter, one and a half grams of spongy nickel reduced from
the bromide and allowed to cool in hydrogen was oxidized by heating in
a current of dry air, which was subsequently passed over red hot cupric
oxide, and through a carefully weighed tube containing phosphoric oxide.
Since the absorption tube did not gain in weight, no water could have
been formed during the combustion, and hence no hydrogen could have
been occluded.
Treated in exactly the same way, four grams of nickel prepared by
the reduction of the oxide yielded about three milligrams of water, or
about half the amount found by Neumann and Streintz. This agreement
is sufficient to show that these investigators were not mistaken in their
conclusions, and that the permeability of nickel is enormously modified
by minor circumstances.
Is the presence of the sodic bromide, otherwise so objectionable, the
agency which prevents the occlusion in our case, or does the volatility of
nickelous bromide allow its metal to be deposited in a form more cohe-
rent than that remaining from the oxide ? The attempt to answer these
* Monatshefte fur Chemie, XII. 6i0 (1891). Berichte der d. ch. Gesell., XXV.,
1872.
334
PROCEEDINGS OF THE AMERICAN ACADEMY.
questions experimentally would be far from uninteresting, but it is suffi-
cient for the present purpose to prove in the manner described above that
nickel treated as we have treated it in the following determinations does
not occlude an important amount of hydrogen, and does not oxidize in
dry air. Confirmatory evidence will be found in the paper upon cobalt
which follows.
Fig. 2. Apparatus for Redccing the Bromide.
Before it was possible to use the data thus obtained for the calculation
of the desired atomic mass, the weight of sodic bromide existing as impu-
rity within the spongy metal must obviously be found. Accordingly the
residue was digested with successive portions of pure water in a platinum
dish, and the bromine in the filtrates was precipitated and weighed as
RICHARDS AND CUSHMAN. — ATOMIC WEIGHT OP NICKEL. 335
argentic bromide. Since the spongy metal (after slow solution in very
dilute nitric acid in a platinum dish) was found to contain no bromine, it
is safe to infer that all the soluble impurity had been leached out by the
water. This process was repeated with every analysis, the result being
always the same. From the weight of argentic bromide thus obtained,
the weight of the impurity of sodic bromide was alculated, and wlien
this quantity was subtracted from both the original weight of the nickel-
ous bromide and the weight of the spongy metal, data suitable for the
calcuhxtion of the atomic mass were obtained.
Iq order to show that tliis method of correcting the results is really
exact, it is necessary to prove, first, that no impurity other than a bro-
mide remains behind in the nickel, and secondly, that no impurity beside
sodic bromide is dissolved by the water. The first point has been already
partially considered; we have shown that silica at least was absent.*
Since silica was the only non-volatile and insoluble acid likely to have
been present, and all the bromine had been dissolved out by water, the
only probable impurities were other metals capable of being reduced
fiom their bromides by hydrogen. But these, even if they had been
present in unsuspected and undiscoverable traces, could have exercised
no appreciable effect upon the atomic weight unless their equivalents
were widely different from that of nickel. Hence this possibility of error
need cause no anxiety. Finally, it has been already stated that no
weighable amount of hydrogen was ever found in the metal, hence we
are justified in the assumption that the washed out spongy nickel is a
safe material upon which to base the calculation of the atomic mass in
question.
The question as to the purity of the sodic bromide in the wash-waters
was a matter less easily settled. Careful qualitative and quantitative
analysis of the liquid were alone capable of deciding the point, and with
only infinitesimal amounts of material such elaborate examination was
difficult. To make a very long story short, nothing was found in any of
the wash-waters beside sodium, bromine, nickel, and in some of the earlier
analyses traces of sulphuric acid. This last impurity may have crept in
from the air during the evaporation of the aqueous solutions, or possibly
from the towers used for drying the nitrogen and air. In the later analy-
ses sulphuric acid was not used in these towers, and was proved to be
absent from the nickelous bromide. As its amount was in any ease
very small, we felt justified in neglecting it as one neglects an infinitesimal
of the second order in the course of mathematical reasoning.
* See p. 328.
836 PROCEEDINGS OF THE AMERICAN ACADEMY.
On the other hand, the amounts of nickel in the wash-waters were
distinctly weighable, and not to be overlooked. That it was nickel, and
not a new metal, there could be no room for doubt, for it gave a black
sulphide and a green sulphate, the characteristic pink coloration with
potassic thiocarbonate, as well as a beautiful rose-colored flame test,
which we have found to be characteristic of nickelous halides. Since
the nickel salts are much more easily reduced than those of cobalt (see
the next paper), it is harder to obtain a satisfactory flame test in the
former case than in the latter. Only when both salts and gas are dry,
and the vaporization proceeds in the inner flame, are the best results to
be obtained. We have found no reference to this flame reaction of
nickel halides in chemical literature,* and one cannot but believe it
to have been unknown to Krliss and to Winkler at the time of one
of their disputes.!
The question now arises. Whence came this nickel ? Was it occluded
as nickelous bromide in the interior of crystals of sodic bromide, and
thus protected from reduction, or was it dissolved as hydroxide from the
spongy metal ? The large quantity of nickel present seemed to over-
throw the former alternative, but the possibility of the latter is emphati-
cally denied by Winkler,t and rejected after some hesitation by Kruss.§
Since the equivalent of nickelous bromide is very near that of sodic bro-
mide (109.3 : 103), and the impurity was calculated from the amount of
bromine present, the point has no important bearing on the immediate
problem; but nevertheless it is always interesting to settle a mooted
question of this sort.
Winkler's experiments were made with coherent nickel coated on the
inside of a platinum dish, while Kriiss's experiments were made with spongy
metal in porcelain vessels. The former found no trace of any substance
in the water in which his nickel had been digested, while the latter found
a large residue (most of which must have come from the porcelain), and
hastily ascribed this residue to the presence of an unknown element. It
* Vogel, Spectral Anal, irdisclier Stoffe, pp. 246, 262.
t Kriiss and Schmidt, Zeitschr. Anorg. Cliem., II. 249; Winkler, Ibid., IV. 17.
Kriiss observed the appearance of a pale rose-colored flame during tiie ignition
of his nickel in a Rose crucible in hydrogen, accompanied with a loss of weight.
If the hydrogen was r/r//, and especially if, as is often the case, it contained traces
of hydrochloric acid, minute traces of nickelous chloride might have been sublimed
and have caused this phenomenon, which Kriiss ascribed to "gnomium" and
Winkler ascribed to potassium. This question is worthy of further attention.
\ Winkler, Zeitschr. Anorg. Chera., IV. 12.
§ Kriiss, Zeitschr. Anorg. Chem., II. 2i{8.
EICHARDS AND CUSHMAN. — ATOMIC WEIGHT OF NICKEL. 337
is not at all impossible that, in spite of the well known permanence of
the smooth bright surface of a nickel-plated object, the spongy metal re-
duced by hydrogen might be easily oxidized and dissolved.* The differ-
ence in this respect between polished iron and the same metal reduced by
hydrogen is well known. Hence Winkler's experiments prove nothing with
regard to the behavior of finely divided nickel. The position of nickel
on the positive side of hydrogen in the electro-chemical series leads one to
expect that it must behave as zinc and iron do, although to a less degree. f
In order to decide the matter so far as the present work is concerned,
we sought to determine experimentally, first, if the purest water is capa-
ble of acting on the purest nickel, and secondly, if the hydroxide thus
formed is sliglitly soluble in water. To settle the first point, some very
pure reduced nickel was thoroughly washed with water, and then digested
at 20° for some time with repeated portions of water just purified. In
every case nickel could be found in the filtrate, both by means of po-
tassic thiocarbonate and by evaporation to dryness. Many repetitions
of the experiment with new samples of metal brought always the same
result. It is quite possible that galvanic action hastened this oxidation,
for of course the nickel was contained in a platinum dish. Since
Winkler's dish was evenly coated, the water probably did not touch the
plathium in his case, and hence this possible cause of acceleration was
absent ; but the difference between the smooth surface and the finely
divided surface alone is amply sufficient to explain the difference in
the speed of the reaction.
It will be remembered that Kriiss, in one of his experiments, digested
a mass of nickel for a year on the steam bath with water, and obtained a
white residue which was the chief basis of his alleged discovery. Our
work shows that this residue must have confined not only dissolved por-
celain, but also enough nickel to yield the black sulphide, the pale green
color, and the electrolyzed metal which led Kriiss so far astray.
* It is a peculiar household fact that cold water faucets plated with nickel are
usually less brilliant than their hot water comrades; this difference may well be
ascribed to the slow action of condensed water, even upon polished nickel. Tlie
greater tendency to rust shown by cobalt may well be due to the fact that in its
case the hydroxide is converted into a higher state of oxidation immediately upon
being dissolved, thus giving opportunity for the solution and hence for the forma-
tion of more hydroxide. In the case of nickel the thin permanent film of Ni(0H).2
probably protects the metal, although its solubility may be no less than that of
Co(OH)2.
t See a paper entitled " Autoxydation," by R. Ihle, Zeitschr. phys. Chem.,
XXII. 114.
VOL. XXXIV. — 22
338
PROCEEDINGS OP THE AMERICAN ACADEMY.
It is clear that, since no acid was present in the water, the nickel must
have been dissolved in the form of hydroxide. In none of our experiments
with the purest metal did the solution give an alkaline reaction with
phenol phthalein, hence hydroxyl ions must be absent, and tlie hydroxide
must be dissolved in a colloidal form. The alkaline reaction sometimes
observed by Kriiss and Winkler must have been due to impurities, as
Winkler pertinently suggests. Traces of alkali would surely be found in
nickel oxide precipitated in glass vessels, even when mercuric oxide was
used as the precipitant.
Nickelous hydroxide, precipitated by alkali and thoroughly washed,
possesses at least as great solubility as the hydroxide which is formed by
the slow oxidation of nickel. Upon suitable electrolysis, fifteen cubic
centimeters of such a cold solution yielded 0.00062 gram of nickel, while
another similar portion yielded 0.00057 gram, or about 0.04 gram per
litre. Since the solubility is a colloidal one, its limit is indeterminate;
hence no elaborate attempt was made to discover its exact amount. It
is well known that the presence of other salts in the solution diminish
this kind of solubility, but the small amount of sodic bromide present in
this case was insufficient to produce any considerable effect. Undoubtedly
the fact that the solubility is colloidal and uncertain is responsible for
the conflicting statements of the various authorities and handbooks.*
This solubility of nickelous hydrate is the circumstance which obliged
us to determine the residue of sodic bromide through argentic bromide,
instead of simply by evaporating the wash-waters poured off from the
nickel and weighing the residue. This necessity is made clear by the
foUowius table.
»T c Tf ■ I... c Weight of Residue
No. of Weight of extracted bv Water
Aual.
Nickel,
grams.
0.805
6
7 1.488
8 0.G07
and dried at 105-'
grams.
0.00330
0.00895
0.00400
Weight of
AgBr.
grams.
0.00410
0.01398
0.00568
Weight of NaBr
calculated
from AgBr.
grams.
0.00225
0.00767
0.00311
Weiglit un-
accounted for.
grams.
0.00105
0.00128
0.00089
In order to prove that this large percentage of unexplained residue
* Finkener (Handbuch der Anal. Cliem. von Rose, 6te Auflage von Finkener,
II. 136); Busse (Zeitschr. Anal. Chem., XVII. 60); Fresenius (Quant. Anal.,
1877-1887, II. 393, 823) ; Roscoe and Schorlemmer (A Treatis^e on Chemistry, Vol.
II. Part II. p. 149); Winkler (loc.cit.); Kriiss {loc.cit.); etc. Temperature is a
circumstance which produces great effect on this kind of solubility, heat being apt
to coagulate the dissolved material.
RICHARDS AND CDSHMAN. — ATOMIC WEIGHT OF NICKEL. 339
consisted chiefly of nickelous hydrate, two more determiuations were
made with new material. Since enough data for the atomic weight had
been obtained, and time pressed, the bromide of nickel was not weighed
in the first place. In these analyses the nickel dissolved as hydroxide
was also weighed.
Weight of
Residue ob-
tained, dried
at 105^.
Argentic
Bromide
louud.
Nickel
depo.«ited elec-
trolytically.
Sodic
Bromide
calculated.
Nickelous
Hydroxide
calculated.
Total calcu-
lated Weight
of Residue.
Weight
unaccounted
fcr.
grams.
0.00925
0.00480
grams.
0.01405
0.00630
grams.
0.C0092
0.00084
grams.
0.00770
0.00345
grams.
0.00146
0.00133
grams.
0.00916
0.00478
grams.
0.00009
0.00002
Besides this quantitative proof that the residue consisted of nothing
but sodic bromide and nickelous hydrate, many qualitative analyses had
shown the absence of lime, alumina, silica, and even potash, from the so-
lution poured off from the reduced nickel. Traces of sulphuric acid
were found, as has been said, only in the first specimens.
In the light of all these results, no doubt seemed to remain as to the
proper mode of correcting the direct gravimetric results of the reduction.
Obviously the weight of the sodic bromide must be subtracted from the
weights of both the nickelous bromide and the metal formed from it by
the action of the hydrogen ; for the salt existed in each. On the other
hand, the nickelous hydroxide was simply to be neglected ; for this sub-
stance was formed after the last weighing had been finished. The tem-
perature used for the reduction was so low that no sodic bromide was
vaporized ; at least none could be found in the cooler parts of the
reduction tube.
Analyses 5 and 6 were made with nickelous bromide of the grade of
purity represented by the numeral II. in the former paper, while all the
others were made with that labelled III., which had been made through
the carbonic oxide process. The headings of the various columns will
show with sufficient clearness the meanings of the figures given below.
In this series of results, the lowest is 58.696, while the highest is
58.719, a variation ±0.012 from the mean. Adding all the determina-
tions together, 24.31725 grams of nickelous bromide yielded 6.52286
grams of pure washed nickel and 0.02778 gram of sodic bromide, calcu-
lated from 0.0507 gram of argentic bromide. These two weights, taken
in connection with the amount of bromine found a year ago, furnish a
340
PROCEEDINCxS OF THE AMERICAN ACADEMY.
THE ATOMIC WEIGHT OF NICKEL.
Fourth Series. — = 16. — Ratio = NiBrj : Ni.
No.
of
Anal.
B
CO
Weight of
Nickel Bromide
in Vacuum.
Weight of
Nickel
in Vacuum.
Weight of
Sodium Bromide
to subtract.*
Weight of
Nickel Bromide
corrected.
Weight of
Nickel
corrected.
Atomic
Weight of
Nickel.
1
III.
grams.
2.83610
grams.
0.76366
grams.
0.00285
grams.
2.83325
grams.
0.76081
58.705
2
III.
3.21908
0.86641
0.00283
3.21625
0.86358
58.696
3
III.
2.31578
0.62431
0.00337
2.31241
0.62094
58.703
4
III.
2.88330
0.77707
0.00377
2.87953
0.77330
58.710
5
II.
2.29843
0.61872
0.00193
2.29650
0.61679
58.719
6
II.
2.99118
0.80497
0.00225
2.98893
0.80272
58.714
7
III.
5.52058
1.48823
0.00767
6.51291
1.48056
58.716
8
Ilia.
2.25280
0.60726
0.00311
2.24969
0.60415
Average
68.710
58.709
complete analysis of the best nickelous bromide which we were able to
prepare. It must be remembered that lo.olooG grams of the halide
contained, according to two different methods, 11.34985 and 11.34979
grams of bromine.
COMPLETE ANALYSIS OF NICKELOUS BROMIDE.
Per cent.
= 26.824
( Bromine combined
73.151% = ] with Nickel
Nickel . .
Total Bromine
Total Impurity = 0.114
} = 73.
062
) Bromine actually )
) found in impurity f ~~
Sodium
0.089
0.025
Total = lOO.OOOt
* The weights of argentic bromide from wliich the sodic bromide was calcu-
lated were re.spectively 0.00520, 0.00516, 0.00615, 0.00687, and 0.00352 grams in
analyses 1, 2, 3, 4, and 5. For analyses 6, 7, and 8, the figures are given on p. 338.
t As has been detailed, ^me of the earlier preparations contained traces of sul-
phuric acid in addition. Since this is not taken into account in the average, the
sura given is slightly too small, and is perhaps deceptive in its accurate showing.
The flaw was eliminated as soon as it was discovered, therefore we had no means
RICHARDS AND CUSHMAN. — ATOMIC WEIGHT OF NICKEL. 341
It has been already pointed out that this impurity of ^'^j per cent of
sodium would make no difference in last year's results if the " equivalent "
of sodium equalled that of nickel. Since, however, it is somewhat less,
slightly too much argentic bromide was obtained last year, and the
atomic weiglit of nickel appeared lower than it really is. Assuming the
amount of impurity to have been the same in last year's preparation as
in this, 0.0 15 should be added to the atomic weight, in order to correct
this error. The results of the four series, which represent the sum and
substance of the present research, are then as follows. Ratios (e) and
(/) were obtained by cross-reckoning from the earlier ratios. The
reason for thus restating the results is because this restatement uses the
weight of the nickelous bromide only as a constant, and not as a basis of
calculation.
Atomic Weight of
Nickel ifO = 16.000.
(a) Preliminary: 2AgBr
: NiBrg . .
. . . [58.695]
(^)
2AgBr :
: NiBra . .
. . . 58.703
(.c) ■
2Ag
: NiBr^ . .
. . . 58.704
(d)
(NiBr2-Ni) ;
; Ni
. . . 53.709
Ue)l
2AgBr ;
; Ni
. . . 58.706
1(f)]
2Ag ;
; Ni
. . . 58.707
59.706
If any assurance is needed that this average indicates very nearly the
true atomic weight of nickel, the assurance may be found in the review
of older work which follows. This review has been postponed until the
close of the paper, in order that the methods might be judged in the light
of our own experience with the subject.
A Brief Criticism of Earlier Work.
The atomic weights of nickel and cobalt have each been the subject of
a score of different researches since the i^roblem was first attacked by
Rothoff in 1818. These investigations have not only led to exceedingly
discordant results, but have also given rise to several interesting and
important controversies. As a chronological list is given in our former
papers,* there is no need of repeating it here. For the purpose of
of finding the exact amount of the sulphuric acid, but in any case it was so small
as to produce only a negligible effect on the result. From some experiments of
Mr. Baxter's it is safe to assume that this error could not have exceeded 0.005 per
cent in tiie worst cases, and in the average it must be still much less.
* These Proceedings, XXXIII. 07, 115.
842 PROCEEDINGS OF THE AMERICAN ACADEMY.
criticism one should rather classify the investigations according to the
methods used in them.
The most direct method of determining the atomic weight of nickel is
obviously the reduction of nickelous oxide, for in this way the ratio
between nickel and oxygen, the usually accepted standard of atomic
weights, is settled at once. Russell, Zimmermann, Mond Langer and
Quincke, Schlitzenberger, and Krtlss and Schmidt * used this method
with varying degrees of success. Of these five investigations the third
was hastily undertaken only to show that nickel which has been vapor-
ized as nickelcarbonyl is essentially similar to the ordinary material ;
the fourth included only two determinations made with oxide undoubt-
edly containing traces of sulphate, and the last was hopelessly faulty for
reasons already discussed. Hence as exact criteria we may reject these
three at once, and turn back to the much more carefully executed work
of Russell and Zimmermann.
Russell showed that when the higher oxides of nickel and cobalt are
ignited in an inert atmosphere, oxygen is driven off and the monoxide re-
mains. His materials were contained in a Rose crucible of platinum ; and
after igniting the oxides in a stream of carbon dioxide, he reduced tliem in
a stream of hydrogen. His nickelous oxide was very carefully freed from
all extraneous matter except the insidious impurities derived from his
glass vessels, against which no precautions were taken. A careful study
of his work shows that this is the most serious cause of error likely to
affect his final result, but that it was probably in part counteracted by the
presence of occluded gases in the oxide ; l^ence we have good reason to
believe that this result is probably somewhat, but not much, too high.
Zimmermann followed essentially the same method as Russell. From
ten exceedingly concordant analyses he deduced the value 58.694t for
nickel, a value slightly lower than Russell's 58.743. His work also was
carried out with great care, and the effort was made to avoid the occlu-
sion of alkaline impurities by precipitating the final hydroxide of nickel
Result.
* 1863 Russell, Jour. Cliem. Soc, [2], I. 51, Ni = 58.743
1886 Zimmermann, Annalen (Liebig's), CCXXXII. 324, Ni = 58.694
1890 Mond, etc. Jour. Cliem. Soc, LVII. 753, Ni = 58.580
1892 Scliutzenberger, Compt. Rend., CXIV. 1149, Ni = 58.515
1892 Kriiss and Schmidt, Zeitschr. Anorg. Cliem., 11. 235, Ni = 57.5 to 64. (!)
t Kriiss and Alibegoff, who published Zimmermann's result, after his death,
unwisely omitted to apply the correction to the vacuum standard. This omission
has been supplied above.
RICHARDS AND CUSHMAN, — ATOMIC WEIGHT OF NICKEL, 343
with pure oxide of mercury, thus doing away with the use of au alkaline
precipitant. This improvement also lessened the danger of Russell's other
chief error, and at the same time introduced yet another with au opposite
tendency, the inclusion of mercury ; hence it is not surprising that his
result should be somewhat the lower of the two. An unprejudiced
critic cannot but consider Zimmermann's work as the best among the
older researches, and it is pleasant to call attention to the fact that Zim-
mermann's result differs by only the fiftieth of one per cent from ours.
The reason why this method gives a result more satisfactory in the case
of nickel than in that of cobalt is probably because cobaltous oxide is so
much more readily raised to the higher stage of oxidation. Both Rus-
sell's and Zimmermann's work may have been slightly vitiated by the
presence of occluded hydrogen in their nickel ; but it is impossible now
to appraise the error involved, because the phenomenon is so irregular.
It is convenient to class together five more investigations which ap-
peared between 1857 and 1871.* In the light of present knowledge
concerning the possibilities of accurate quantitative work these con-
tributions may be dismissed with few words. Marignac showed that
Schneider's oxalate contained occluded impurities, and Schneider showed
that Marignac's chloride could hardly have been both anhydrous and
free from oxide. Our own experience entirely confirms both of these
criticisms. Dumas's lack of ability to determine chlorine with accuracy
throws out his analyses of the chloride at once, even if one is credulous
enough to believe that the choride itself was pure. Sommaruga precipi-
tated sulphuric acid from nickelous potassic sulphate as baric sulphate,
a method now ostracised except for crude work. Lee's work was won-
derfully accurate, considering the small quantities of materials which
he used, but these quantities were so microscopic, and his compounds
were so complex, that one could not have been expected to improve
much upon his error of one per cent without a radical reformation of
method.
Within this same period appeared another paper by Russell, elaborated
as carefully as his previous one, but depending upon a less satisfactory
process. The hydrogen evolved by the action of nickel upon hydro-
chloric acid was measured. Many uncertainties combine to make this
* (1857) Schneider, Pogg. Annal., CI. 387, CVII. 61G, 58.07
(1858) Marignac, Arch. Sci. Nat., (nouv. sen), I. 375, 58.90
(1860) Dumas, Annalen (Llebig's), CXIII. 25, 59.02
(1866) Sommaruga, Sitzber. Wien. Acad., LIV. [2], 50, 58.0-3
(1871) Lee, Am J. Sci., [3], II. 4-1, 58.01
344 PROCEEDINGS OF THE AMERICAN ACADEMY.
method of little value, so that the result (58.77 ?) does not carry with it
much weight.
Tiiree chemists in three different decades, Marignac, Baubigny, and
Schiitzenberger,* have attempted to solve the question by the quantita-
tive ignition of nickelous sulphate. The three investigations agreed
fairly well upon an average result, 58.71,t Baubigny's being by far the
most satisfactory. This method is one involving two errors which
nearly counterbalance each other: — the sulphate has a tendency to
retain water, while the oxide almost invariably retains sulphuric acid.
For this reason, the method gives results which approximate closely to
the truth.! Here again we have a support for the conclusion that the
value iu question cannot be far from 58.7.
All the published work upon the subject has now been referred to
except some early work of Eothoff, Erdmann and Marchand, and
Deville § (which deserves no more than a passing mention), and the
more recent researches of Winkler. Kriiss's misguided work has been
sufficiently dissected by Winklei''s able but unsparing criticism || and in
the experimental part of this paper. The only points not covered by
Winkler, — the rose-colored flame test and the solubility of nickelous
hydrate, are explained in the foregoing pages.
The work of W^inkler is surprising in its variety and in the ingenuity
of his methods, but unfortunately it is equally surprising in the wide
range of one per cent between his several results. His earliest work,**
depending upon the reduction of sodic aurochloride by nickel, giving the
extremely high value 59.45, is obviouslj' at fault. Winkler himself
ignores it in his discussion,tt so that further criticism of it may well
be omitted.
Winkler's two later investigations, carried out only a few years ago,
gave results much lower and more satisfactory. In his first revision he
weighed nickel, converted it into chloride, and determined the chlorine
Greatest Difference
from Mean.
* (1858) Marignac, Arch. Sci. Nat., (nouv. ser.), I. 374, Ni = 58.70 ± 0.15
(1883) Baubigny, Compt. Rend., XCVII. 951, Ni = 58.73 ±0.002
(1892) Schiitzenberger, Compt. Rend., CXIV. 1149, Ni = 58.65 ± 0.075
t Clarke, Recalculation, top of page 302.
I See " A Table of Atomic Weiglits," These Proceedings, XXXIII. 297, 298.
§ See Clarke, Recalculation, p. 291.
II Zeitschr. Anorg. Chcm., IV. 10.
** 1867. Zeitschr. Anal. Chem., YI. 18. Ni = 59.45 (Clarke).
tt Zeitschr. Anorg. Chem , IV. 10, VIII. 1.
RICHARDS AND CUSHMAN. — ATOMIC WEIGHT OF NICKEL. 345
both gravimetrically and volumetrically.* He wisely regulated his opera-
tions in such a way as to avoid the use of alkalis ; but in his iirdor
to escape this danger he encountered others as serious. It is highly
unlikely that the electrolytic nickel, dried at only 50° while adhering
to the dish, could have been free from impurities, as we have already
shown in the experimental part of our paper. Indeed, he confessed in a
later paper f that the electrolysis of cobalt at any rate is very far from
being as accurate a process as it is sometimes supposed to be. The
report of the acid reaction of the chlorides of these metals would have
had more significance if the indicator had been named, for the salts
destroy the magenta of phenol phthalein only because they remove the
hydroxyl ions from the solution. Both chlorides are perfectly neutral to
methyl orange. A loss of chlorine during the drying of either chloride
would of course raise the observed atomic weight ; and while in the
case of nickel the loss was so small as not to have produced a visible
cloudiness, no proof is offered that no loss took place. In the case of
the cobalt " eine gewisse, aber so schwache Triibung, dass sie, wie man
zu sagen pflegt, nicht ' blank ' erschien " involved a loss of as much as
one per cent of material, and it is well possible that a smaller but still
important amount of basic salt may have escaped notice in the case of
the nickel. Our own experience with the halides of both metals con-
vinces us that it is quite impossible to obtain them pure and dry by evap-
orating to dryness in moist air. Turning now to the determination of
the chlorine in the salts, we find other grave flaws. No account was
taken of the solubility of argentic chloride in the gravimetric work, and
several milligrams must have been washed away by the hot dilute nitric
acid used as a washing fluid. One is surprised, too, to find that the anti-
quated process of burning the filter was adopted, instead of Gooch's
admirable substitute. In the volumetric work again the solubility of
argentic chloride was overlooked, although it produces a most injurious
effect on the method of Volhard.t It is indeed surprising to see so emi-
nent a chemist using volumetric methods at all in this way, for every one
knows the difficulty of obtaining results of a very high grade of accuracy
by their aid. In this laboratory the burette is only called into use when
at least ninety-nine per cent of the material has been weighed out, and
then only a few cubic centimeters of a very dilute solution are added to
* Zeitschr. Anorg. Chem., IV. 10, 1893.
t Zeitschr. Anorg. Chem., VIII. 4.
t These Proceedings, XXVI. 34, and XXIX. 67.
346 PROCEEDINGS OP THE AMERICAN ACADEMY.
complete the quota. This is quite a different story. It is evident that
nearly all the errors mentioned tend to make Winkler's result too high.
Some singular oversights appear also in the calculation of the results.
For example, in one place Winkler compares 0.1 G62 gram of nickel with
0.6079206 gram of silver. The small amount of nickel was deposited
in a large platinum dish, and its weight could certainly not have been
determined more accurately than within 0.1 milligram, hence at least
three decimal places of the recorded weight of silver were superfluous,
even if the volumetric solution could have been prepared with an error
of only one part in six millions. It is perhaps well to mention also that
his final results, varying in the case of cobalt from 59.5996 to 59.7480
(if O = 15.96) are given as far as four decimal places.
While a review of this work is necessary in order to explain why the
results should be too high, perhaps one should not be severe in one's
criticism of it, for Professor Winkler himself rejects it, as well as some
later work on cobalt,* in his most recent contribution upon the subject. f
In this new paper he pins his faith to another series of determinations
made in 1894, with a very ingenious method adopted after sundry fruit-
less attempts in other directions. It behooves us then to consider this
later work with great care. J
Evidently many of the errors which render the older investigation
untrustworthy were eliminated from that of the subsequent year. The
nickel was separated from the platinum dish and afterwards ignited in an
atmosphere of hydrogen, and the solubility of argentic chloride does not
enter into the question. On the other hand, the unfortunate use of vol-
umetric operations and the misuse of figures remained, while to these were
added other dangers not present in the older work. The ingenious pro-
cedure was as follows: pure nickel was acted upon by pure iodine, and
the excess of iodine was determined by sodic tliiosulphate. Many text-
books upon volumetric analysis name the process of iodometry as one
of the most accurate of titrimetric methods simply because the end
point is an extremely sharp one. In reality, the lack of permanence
of the necessary solutions render it distinctly unsuitable for very ac-
curate work even under the best conditions. When the iodine must
* Zeitschr. Anorg. Chem., IV. 462.
t Zeitschr. Anorg. Chem., XVII. 236.
i 1894 (1895), Winkler, Zeitschr. Anorg. Chem., VIII. 1,291; Ni = 58.85. It
must be borne in mind in referring to Winliler's papers that he uses the old stan-
dard = 15.96. His values have all been translated into the more convenient nota-
tion (0 = 16.000) in this paper.
RICHARDS AND CUSHMAN. — ATOMIC WEIGHT OP NICKEL. 347
remain in solution for twenty-four hours after weighing and before
titration, and when this circumstance is complicated by the presence
of a metal capable of acting to a slight extent even upon pure water
in the presence of air, one can hardly contend that the conditions are
the best. The chance of side reactions seems to be too great to admit
of infallibility in the results. One is surprised, indeed, that Winkler's
results approach as near to those of Zimmermann as they do, and this
close approach is evidence of great accuracy of manipulation on Wink-
ler's part. In short, viewed from the standpoint of ordinary analytical
experience, Winkler's last work is admirable, while from the standpoint
of atomic weight research it is inadmissible. In justice to Professor
Winkler it is only fair to add that he realizes this fact himself.* One
need not dwell upon possible inaccuracies, however; for Winkler himself
has furnished us with data for computing the error of his method. In a
short paper he uses the same method for determining the atomic weight
of iron, and finds for this quantity the value 5G.174, if O = IG.OOO.f
Now according to the fairly consistent work of Berzelius, Erdmann and
Marchand, Svauberg and Norlin, and Maumene, the atomic weight of
iron cannot be far fi-om 56.02 ; and there is no contradictory evidence
of serious value.t Winkler's method then gave him a result 0.275 per
cent § too high in the case of iron, and it is fair to conclude that the
error could not have been far different in the case of nickel. Making
the corresponding subtraction, Winkler's corrected result approaches
astoundiugly near to those obtained by Zimmermann and by us.|l
Winkler's corrected value 58. G9
Zimmermaun's value 58.694
Richards and Cushman's value 58.706
Average 58.70
Owing to a slight uncertainty in the atomic weight of iron, as well as
to the possibility that iron may behave somewhat differently from nickel
* Zeitschr. Anorg. Chem., XVII. 239.
t Zeitschr. Anorg. Chem., VIII. 291.
} Clarke's recalculation, p. 289. The atomic weight of iron is now being further
studied in this Laboratory.
§ It is possible that a small part of this error is due to the omission of the
reduction to the vacuum standard, which would afiect the final value by about
0.01 per cent. This correction may have been applied, but there is no evidence
of such application.
li Mr. Baxter first called oiir attention to this remarkable unanimity.
848 PROCEEDINGS OF THE AMERICAN ACADEMY.
in iodine solutions, this comparison is less significant than it seems to be ;
but certainly it does not militate against our value for the atomic weight
of nickel. It is of interest to note that Clarke's mathematical method
of selecting from among the older values led to the number 58.687.
Professor Winkler's sixth and last paper upon this subject appeared
only last summer, after the work described in this paper had been com-
pleted.* In it he kindly points out several possible flaws and omissions
in our earlier paper. This criticism will be discussed at length in the
next paper on cobalt.
Cambridge, Mass., October 22, 1898.
* Zeitschr. Anorg. Chetu., XVII. 236.
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. No. 14. — February, 1899.
CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.
A REVISION OF THE ATOMIC WEIGHT OF COBALT
SECOND PAPER. — THE DETERMINATION OF THE COBALT IN
COBALTOUS BROMIDE.
By Theodore William Richards and Gregory Paul Baxter.
CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF
HARVARD COLLEGE.
A REVISION OF THE ATOMIC WEIGHT OF
COBALT.
SECOND PAPER: — THE DETERMINATION OF THE COBALT IN
COBALTOUS BROMIDE.
By Theodore William Richards and Gregory Paul Baxter.
Presented October 12, 1898. Received December 29, 1898
In a recent paper upon the atomic weight of cobalt,* we began the
analysis of cobaltous bromide by the determination of its bromine, with
results which seemed to show that the atomic weight of cobalt is very
nearly 58.99 (0 = 16). Although very great care was taken in purifying
the cobaltous bromide used in this work, it was never certain that during
the sublimation of the bromide in the porcelain tubes small amounts of
impurity had not crept in. The situation here was exactly the same as
in the research upon nickel, carried on at the same time in this Labora-
tory,! and here also the simplest and most convincing way of settling the
question was to determine directly the amount of cobalt present in the
salt, thus obtaining its total percentage composition.
Of the three methods of procedure which presented themselves, —
precipitation, electrolysis, and reduction of the bromide, — reduction by
hydrogen was chosen as being the least complicated and most certain.
The slight hope offered in the nickel research that electrolysis might be
used for the determination of the metal was here lacking on account
of the greater difficulty in obtaining a satisfactory electrolytic deposit.
A conceivable objection to the use of the reduction of an oxide as an
accurate quantitative method is the possibility that this reaction may not
be capable of absolutely complete fulfilment. It is well known, for
example, that one is rarely able to reoxidize wholly a metal once reduced,
* These Proceedings, XXXIIL 115.
t The preceding paper describes this work. This Volume, p. 327.
352 PROCEEDINGS OF THE AMERICAN ACADEMY.
even if the material is finely divided. In the case of the oxide there is
no means of deciding whether or not the last traces of oxygen have been
removed, while in the case of the bromide a residue of halogen is easily
detected. The complete reduction of nickelous bromide described in the
preceding paper shows that some such operations are in fact possible.
It is easy to see how the increase in volume involved in oxidation might
cause a part of the metal to be permanently protected by the growing
coat of oxide ; on the other hand, an oxide or salt which is being reduced
leaves only its skeleton behind, the innermost meshes of which may be
penetrated by the reducing agent.
Cobaltous bromide when heated in a current of dry hydrogen begins to
be reduced at a temperature of about 350°. At higher temperatures the
reduction goes on more rapidly, but is then accompanied by partial sub-
limation of the bromide. Schiitzenberger's * observations on the sublima-
tion of nickelous chloride were similar to these, but he seemed to think
that some strange compound must have been the medium of the change,
instead of realizing that in the presence of hydrochloric acid a trace of
the vapor of the metallic salt might easily exist even in an atmosphere
of hydrogen.
If the hydrogen is moist, however, the action begins at a lower tem-
perature, about 250° ; but even under these conditions sublimation can be
completely avoided only with the greatest difficulty. The reduction of
nickelous bromide offered fewer difficulties ; this process could be effected
in a current of dry hydrogen, and without danger of sublimation of a
trace of the material. The fact that moist hydrogen is more efficient
than the dry gas is easily explained by the hypothesis that the oxide is
formed as the first step in the reaction.
Purification of Materials.
Pi'eparation of Cobaltous Bromide. — The cobaltous bromide used in
this work was prepared by methods similar to those used in our previous
investigation, and for details the previous paper should be consulted.
Pure metallic cobalt was heated in a current of pure bromine and hydro-
bromic acid, and the sublimed bromide was preserved in weighing bottles
contained in desiccators until used for analysis. Samples I. and II. were
essentially the same as in the earlier investigation, even greater pains
having been taken, however, in the purification of the reagents and
water used in their preparation. Sample III. likewise was purified by
* Compt. Rend., CXIII. 177.
RICHARDS AND BAXTER. ATOMIC WEIGHT OF COBALT, 353
essentially the same method as before, but was then further treated by
six additional recrystallizations as the purpureo-chloride, with the help
of very pure redistilled ammonia and hydrochloric acid. The oxides
obtained from these preparations were finally reduced by means of pure
hydrogen, instead of by ammonia as before.
Preparation of Bromine. — Here also the process of purification did
not differ from that previously employed. The purity of the bromine is
sufficiently proven by two analyses in which a known weight of silver
was precipitated by a slight excess of ammonic bromide made from the
halo£:en.
g in Vacuum.
AgBr in Vacuum.
AgBr: Ag
grains.
2.91386
grams.
5.07226
57.447
2.97097
5.17170
Average
57.447
57.447
Stas
found
57.445
The balance and weights used in this work were the same as those
described in our previous paper. The weights were carefully restan-
dardized, the values differing from those previously found by only a few
hundredths of a milligram. Since the balance was wholly free from
iron, no inaccuracies could have arisen from magnetic attraction, either in
this work or in that upon nickel.
Owing to the fact that the specific gravity of cobalt is almost identical
with that of brass, no correction was required to reduce the weight of
the cobalt to a vacuum standard.
The correction of +.00010 gram per gram of cobaltous bromide was
applied in each case to the weighings of that material. All weighings
were made by substitution, as usual.
We are indebted to the Cyrus M. Warren Fund for Chemical Research
in Harvard University for some of our more expensive pieces of
apparatus.
Method of Analysis.
By means of the glass apparatus described in our earlier paper, cobalt-
ous bromide, contained in a platinum boat, was dried in a current of pure
dry nitrogen and hydrobromic acid gas in a hard glass tube heated to
about 400° ; and after the tube had been thoroughly swept out with
nitrogen and then by dry air, the boat was transferred to a weighing
bottle in this safe atmosphere. The bottle was then weighed, and the
VOL. XXXIV. — 23
354
PROCEEDINGS OF THE AMERICAN ACADEMY.
boat carefully placed in the reduction tube, where it was heated in a
current of moist hydrogen until the fumes of hydrobromic acid ceased to
come off. The tube was then swept out with dry hydrogen, and when
cool the boat was quickly replaced in the weighing bottle coutaining dry
air, and was thus weighed after a suitable delay. In most analyses this
process was repeated until the weight of the cobalt ceased to change.
Cobalt reduced from the bromide is less constant in weight than nickel,
gaiuing several tenths of a milligram in weight in twenty-four hours.
Fig. 1. Apparatus for ignitixg CoBALTOrs Bromide in ant desired
iNIixTCRE OF Gases.
Tlie use of rubber was confined to the fir.st part of this trnin, where it could do
no harm (A B C D E F and A M N O P).
For this reason the metal was always allowed to become thoroughly cool
in the atmosphere of hydrogen, and the weighing bottle was allowed to
come to perfect equilibrium with the atmospheric conditions inside a desic-
cator at the room temperature before receiving the boat. After half an
hour nickel treated in the same way had been found to come to constant
weight, and in half an hour the opportunity for oxidation of cobalt is so
slight as to be negligible ; hence this interval was the one which always
elapsed between the bottling and the weighing of the cobalt. In one
analysis, to see if exposure to the air affected the weight of the cobalt,
the boat was bottled in dry nitrogen in the bottling apparatus, after
RICHARDS AND BAXTER. — ATOMIC WEIGHT OF COBALT. 355
being heated to constant weight in the usual way. Upon correcting the
weight of the bottle for the difference in weight of nitrogen and air, the
weights obtained by both methods agreed perfectly.
The resulting cobalt was in the form of a gray metallic sponge, which
showed no traces of oxidation upon standing in air. Since previous
work by other experimenters * is not unanimous as to the occlusion of
hydrogen by cobalt under these circumstances, it seemed desirable to us
to obtain more evidence on this point. Accordingly, in several analyses
the boat containing the cobalt was placed in a hard glass tube sealed at
one end. After the air had been exhausted by means of a Sprengel
pump, the tube was heated to about 500°, the highest temperature used
in the reduction. In no case was a measurable quantity of gas evolved,
and the cobalt did not lose in weight ; hence it would appear that cobalt
prepared from the bromide does not possess the property of occluding
important amounts of hydrogen when heated in the gas.
To avoid any possibility of error, two and a half grams of cobalt,
freshly reduced from the bromide aud allowed to cool in hydrogen, were
subjected to quantitative combustion in the manner already described in
the paper upon nickel. Only five tenths of a milligram of water were
formed ; and even the repeated reduction and combustion of the residual
oxide yielded only a milligram. Evidently the amount of hydrogen
occluded was very small. On the other hand, cobalt reduced originally
from the oxide, when treated in the same way, was found to contain
about fifteen times its volume of hydrogen and when allowed to remain
in hydrogen several hours, it was found to have absorbed amounts com-
parable to those found by Neumann and Streintz.f A fuller statement
of the experiments will be reserved for a future paper upon the nature
and causes of these singular irregularities ; for the present, it is sufficient
to have shown that cobalt, like nickel, reduced from the bromide, does
not retain enough hydrogen to vitiate the results recorded below.
It is perhaps worth while to state also that the empty platinum boat
was tested as to its power of absorbing weighable amounts of hydrogen.
After ignition and cooling in the gas, and bottling in dry air as usual, it
was found to have gained 0.02 milligram when compared with its weight
after ignition in air. Evidently the occlusion of hydrogen, if measurable
at all, is balanced by adsorption of air ; hence for our purpose it may be
neglected.
* Neumann and Streintz, Monatshefte fiir Chem., XIT. 642; Berichte d. d. ch.
Gesell., XXV., 187R ; Henipel and Thiele, Zeitschr. Anorg. Chem., XI. 93.
t Loc. cit.
356 PBOCEEDINGS OF THE AMEEICAN ACADEMY.
A slight sublimation of the cobaltous bromide took place during the
reduction in almost every case. The amount of this sublimed material
was determined by washing out the tube with a few cubic centimeters of
nitric acid, and evaporating this liquid to dryness. After the solution of
the residue in water and the addition of an excess of ammonia, a very
dilute standard solution of potassic permanganate was run in until a pink
color appeared. This method of Winkler's is applicable only when ex-
tremely small amounts of cobalt are present, because the brown color of
the cobaltic salt interferes seriously with the end point in the presence
of large amounts of cobalt. The weight of the cobaltous bromide thus
sublimed never amounted to more than three tenths of a milligram, and
seldom exceeded one tenth of a milligram.
The platinum boat used in the earlier work served to contain the
bromide in these experiments also. Although the metallic cobalt alloyed
itself with the surface of the boat to a slight extent, we were able to
remove completely the alloy by treating the inside of the boat with aqua
regia. After this treatment and scrubbing with round sand, the boat
showed no trace of darkening upon ignition. Evidently, then, the cobalt
had been completely removed. Of course a gradual loss of weight took
place, owing- to solution of small amounts of platinum, but this loss
amounted to only half a gram in the course of the work.
In the first two determinations the hydrogen was generated from hydro-
chloric acid by means of pure zinc. It was purified by passing through
bulbs containing silver nitrate, potash, silver nitrate again, then through
a hard glass tube heated to redness. From this point the gas came in
contact only with glass, being conducted through three towers containing
glass beads moistened with silver nitrate, then by means of T-tubes,
either directly in a moist state, or through a long drying tube containing
stick potash, into the reduction tube. The reduction tube was connected
with the rest of the apparatus by means of a ground glass joint.
A small amount of white sublimate, which appeared beyond the boat
during each of the preliminary ignitions, proved to be ammonic bromide.
The source of the ammonia was not at first apparent, as it was hard to
believe that the cobaltous bromide could retain ammonic bromide at a
temperature between 400° and 500°. Upon examination of the silver
nitrate columns it was found that reduction*had taken place there, metal-
lic silver being precipitated upon the beads.* Of course the reduction of
* The fact that silver nitrate is reduced by molecular hydrogen has already
been noted by other experimenters : Russell, Jour. Chem. See, [2], XII. 3, (1874) ;
Pellet, Compt. Rend., LXXVIII. 1132,
mCHAEDS AND BAXTER. — ATOMIC WEIGHT OP COBALT. 357
the silver alone could do no harm; but unfortunately it was attended by
a reduction of the nitric acid also. This was proved by passing the re-
sulting gas through a hot tube, when traces of ammonia were formed,
capable of easy detection by Nessler's reagent. We had come face to
face with one of those frequent cases where an attempt at purification
had introduced a flaw as serious as the one it eliminated. The very
common use of argentic nitrate as a means of purifying hydrogen is
obviously a pernicious one, if accurate results are desired.
The hydrogen apparatus was then entirely remodelled. Owing to the
fact that the amount of hydrogen required for the completion of a reduc-
tion was very much larger than the amount actually necessary to combine
with the bromine,* a gasometer was constructed which should collect the
hydrogen after it had passed through the tube and deliver it repeatedly
to the apparatus, after removal of the hydrobromic acid. The hydrogen
was generated by a primary battery consisting of zinc amalgam, hydro-
chloric acid, and platinized platinum. The gas delivered by this appa-
ratus is pure, except foi- the presence of a little hydrochloric acid, which
can be removed easily by means of potash. The following cut shows
the apparatus in its improved form.
The bottle B is filled with pure hydrogen generated by the battery
C. From the bottle A water is siphoned into B, forcing the hydrogen
by way of the stopcock e through the column D, filled with beads
moistened with aqueous cupric sulphate to remove sulphur compounds
taken from the rubber; through the columns E and .^ which contain
dilute sodic hydrate, then either directly through g or through a potash
tube G into the reduction tube H. After being freed from hydrobromic
acid in the bottle K containing potash, the gas is conducted through the
open stopcock h into A. When B is full of water the process can be
repeated by interchanging A and B, and opening the stopcocks c and d
after closing h and e. The generator C served to keep the pressure
always outward. The current of gas could be regulated by a pinchcock a
on the rubber siphon tube. This apparatus proved entirely satisfactory,
and was not altered during the investigation.
As in the case of nickel, it was found impossible by reduction alone
to remove the last traces of bromine from the spongy cobalt. Even
long continued heating to a temperature much above the subliming
point of cobaltous bromide failed to give complete reduction, the solu-
tions of the reduced cobalt giving decided tests for bromine. In the first
* See the preceding paper, p. 333.
358
PROCEEDINGS OP THE AMERICAN ACADEMY.
Fig. 2. Apparatus for Reducing the Bromide.
analysis this bromine was precipitated with an excess of silver nitrate
and weighed. In analyses 2, 3, and 4 the cobalt was leached with pure
water, and after filtration the bromine was determined in the filtrate.
The solutions of the cobalt in dilute nitric acid proved almost entirely
free from bromine, that obtained from analysis 4 being entirely so. The
silver bromide from these solutions was combined with that obtained
from the aqueous extracts and weighed. As it was hard to believe that
cobaltous bromide could be enclosed by the reduced cobalt in such a way
as to remain unreduced and yet to go completely into solution with ap-
parent ease, it was obvious that some foreign bromide must be present.
Accordingly the filtrate from one of the precipitates of silver bromide
RICHARDS AND BAXTER. ATOMIC WEIGHT OP COBALT. 359
was treated with an excess of hydrobromic acid in order to remove the
silver, and the retiltered solution was evaporated to dryness. Upon a
careful qualitative analysis tests were obtained for nothing but sodium
and traces of cobalt. The sodium evidently came from the porcelain
tube used in the sublimation, and we were dealing with an impurity
precisely similar in source and nature to that described in the case of
nickelous bromide. As the two investigations progressed side by side,
the discovery was almost simultaneous in the two cases. We have
already said that this outcome was not an unexpected one.
Since it would be impossible to calculate the weight of the soluble
bromides from the weight of the silver bromide without a quantitative
analysis of the two bases present, the effort was made to evaporate the
aqueous extract from the reduced cobalt and to weigh the dried residue
directly. In the work upon nickel this method was necessarily rejected
because the spongy nickel was oxidized and went into solution as nickel-
ous hydrate which could not be removed by filtration. Our spongy
cobalt oxidized much more rapidly than the nickel upon treatment with
water ; but the decantate, upon filtration and evaporation in the air,
deposited most of its cobalt as cobaltic hydroxide. The presence of salts
of the alkalis greatly increases this oxidation, cobalt which has once been
leached being oxidized but little. Heat also increases the oxidation, and
so probably does the galvanic action with the platinum dishes which were
used wherever possible through the whole course of this investigation.
In view of the colloidal solubility of nickelous hydrate in water, it is
probable that cobaltous hydrate possesses the same property. When
cobalt is treated with water in the presence of air, the metal oxidizes and
goes into solution as cobaltous hydrate ; this is then further oxidized by
exposure of the solution to the air and thrown out of solution as cobaltic
hydrate, which can be filtered off. Since the dissolved cobalt is almost
completely removed by this process, it is obviously legitimate to weigh
the residue obtained by evaporating the aqueous extract of the reduced
cobalt, and thus to obtain directly the amount of impurity present in the
metal. This difference of procedure in the two cases is an interesting
example of the way in which subordinate side reactions may influence
two researches otherwise unusually analogous.
In the next four analyses (5 to 8) the cobalt was leached with the purest
hot water in a platinum dish and the solution, after filtration, was evapo-
rated to dryness. The residue was taken up with water, filtered from
the deposited cobaltic hydroxide into a weighed platinum crucible, again
evaporated, heated to 130°, and weighed. That these residues contained
360 PROCEEDINGS OF THE AMERICAN ACADEMY.
cobalt was undoubtedly true from the fact that they were colored a pale
blue both after evaporation and after heating. During the heating
Iiowever a slight blackening took place which was due to the oxidation of
traces of unoxidized cobaltous hydrate. In order to make sure that all
unreduced bromides had been renaoved from the cobalt by the process of
leaching, in each analysis the metal was dissolved by cold very weak
nitric acid* and treated with silver nitrate. As not even the faintest
cloudiness was ever again visible in these solutions, it was assumed that
the soluble matter had been completely removed.
In analysis 5 the bromine in the residue was determined and the result
found to be considerably too low to correspond with the weight of the
residue, if calculated as sodium bromide. In order to discover the cause
of this discrepancy an elaborate series of experiments was carried out at
the expense of much time and labor. To describe these experiments in
full would result only in confusing the mind of the reader. It is sufficient
to say that 2:)ure spongy cobalt was treated with varying amounts of pure
sodium bromide in solution, the conditions being regulated so as to be as
nearly as possible like those in the analyses. The following conclusions
drawn from these experiments are of great importance.
In the first place, no bromine is lost by the residues either during
evaporation or in heating to 130°. The cobalt in the residues may
be present in three forms, as cobaltous hydrate which has escaped oxi-
dation, as cobaltic hydrate, and as unreduced cobaltous bromide,t each
in exceedingly small amounts. The doubt as to the quantity of each
present makes it impossible to apply the correction for dissolved cobalt
with any degree of accuracy and in the table of results no attempt has
been made to do so. In the later analyses this cause of uncertainty was
removed. t This correction within a correction is however an infinitesi-
mal of the second order ; neglecting it can produce no serious effect upon
the accuracy of the final result. § In this respect the research upon
cobalt differs from that upon nickel, where the amount of hydroxide in
the residue was relatively great, owing to the fact that it had not been
chiefly eliminated by oxidation during the evaporation.
* In a platinum dish this solution takes place with great ease, and with no danger
of a loss of bromine. The galvanic action is a great assistance.
t Cobaltous bromide could only have found its way into this residue by having
been protected from reduction by enclosure in crystals of sodic bromide. It must
have been exceedingly small in amount, if present at all.
f Compare page 3G3.
§ This extra cobalt may partly explain why the sum of the total analysis slightly
exceeds 100.000 per cent. See page 365.
RICHARDS AND BAXTER. — ATOMIC WEIGHT OP COBALT. 361
The possibility of the presence of some other acid than hydrobromic
acid in the residues led us to make tests in this direction. Silicic acid
Avas of course the first one to suggest itself. Early in the work one test
liad been made by subliming about two grams of pure cobaltous bromide
from a platinum boat in a current of hydrobromic acid gas. After the
sublimation of the bromide the boat was perfectly bright, and gave not
the slightest evidence of the presence of any silica. One of the residues
was now treated with pure strong hydrochloric acid, heated to 130°,
again treated with hydrochloric acid, and filtered. The amount of silica
found, three one-hundredths of a milligram, is a negligible quantity. Al-
though the purity of the phosphoric anhydride used for drying our gases
had been proven by passing air through a tube filled with the pentoxide
into aqua regia, which upon evaporation gave no precipitate with ammo-
nic molybdate, nevertheless one of the soluble residues from the reduced
cobalt was tested with the same reagent with a negative result.
Upon examining one of the earlier residues for sulphuric acid, how-
ever, a slight precipitate of baric sulphate was formed. The source of
this sulphuric acid was hard to discover, but finally it was found that the
strong sulphuric acid in some of the drying columns had become dis-
colored in places by organic matter. This must have led to a slight
decomposition of the acid and formation of sulphur dioxide, which was
subsequently oxidized to sulphuric acid by the bromine. The amount
of sulphuric acid present in the bromide was very small, 3.88 grams of
cobaltous bromide giving only 0.00036 gram of baric sulphate in one
case. As in the instance of nickel, however, this cause of error was
wholly eliminated in the later experiments, aod no trace of sulphuric
acid could be detected in the material used in the last series.
The conclusions to be drawn from these experiments seem to be : —
First, that our cobaltous bromide was almost if not completely reduced.
Secondly, that the impurities, which consist of alkaline bromides (with,
in some cases, a minute trace of sulphates), can be completely removed
by leaching the cobalt.
Thirdly, that the residue obtained by evaporating the water extract of
the cobalt after reduction represents within an exceedingly small amount
the weight of the impurities.
In the fourth series, during which the truth of this third conclusion was
not realized, and a method similar to that used in the nickel research was
adopted, the weight of the residue had to be calculated. The basis of
calculation was the knowledge obtained from analysis 5, Series V. Un-
fortunately this is the only analysis of material similar to that used in
362
PROCEEDINGS OF THE AMERICAN ACADEMY.
Series IV., where both the weight of tlie residue and the silver bromide
obtained from it were determined. Since, however, the use of this
analysis causes the average of Series IV. to approach within one part in
thirty thousand of the average of Series V. and VI., we may safely
assume that the rather meagre data represent with great exactness the
real weight of the impurity contained in this Sample I. of cobaltous
bromide. In Series V. and VI. the residues were weighed directly, so
that this factor was not needed. Below are given the results of the first
two series of analyses.
THE ATOMIC WEIGHT OF COBALT.
O = 16 ; Br = 79.955.
Fourth Series (Preliminary). CoBfj : Co.
No. of
Anal.
Sample
of
CoBr,.
Observed Weight
of Cobaltous
Bromide
in Vacuum.
Observed Weight
of Cobalt in
Vacuum.
Weight of
AgBr found
from Residue.
Weight of Resi-
due calculated
from AgBr.*
Atomic
Weight of
Cobalt.
grams.
grams.
grams.
grams.
1
I.
5.59216
1.50873
0.00309
0.00193
59.007
2
I.
4. G 1914
1.24807
0.00081
0.00426
58.996
3
I.
3.75291
1.01713
0.01207
0.00793
58.989
4
I.
3.00G15
0.81409
0.00815
0.00510
59.007
Average .
. 59.000
Fifth Series. CoBfo : Co.
No.
of
Anal.
Sample
of
CoBrj.
Observed
Weight of
Cobaltous
Bromide in
Vacuum.
Observed
Weight of
Cobalt
in Vacuum.
Weight of
Residue.
Corrected
Weight of
Cobaltous
Bromide.
Corrected
Weight of
Cobalt.
Atomic
Weight of
Cobalt.
grams.
grams.
grams.
grams.
grams.
5
I.
5.32955
1.44189
0.00761*
5.32194
1.43428
58.996
6
I.
7.51430
2.02965
0.00644
7.50786
2.02321
58.989
7
II.
2.32910
0.62957
0.00280
2.32630
0.62677
58.973
8
II.
7.45336
2.01378
0.00642
7.44694
2.00736
59.011
A
verage .
58.992
* From the residue in Series Y., analysis 5, was obtained 0.01210 gram of
RICHARDS AND BAXTER. — ATOMIC WEIGHT OF COBALT. 363
It seemed highly important to us at this point to prepare cobaltous
bromide which should be free from every impurity. In the first place
the drying apparatus was slightly modified, no strong sulphuric acid
being used except in drying the air necessary for sweeping the nitrogen
out of the weighing bottle. Dilute sulphuric acid was substituted in
every case, and two columns of stick potash followed by one of phosphoric
anhydride were inserted beyond this dilute acid. As the porcelain tubes
had evidently been the source of the alkaline impurities found, a platinum
lining, made by bending a large piece of platinum foil into the form of a
cylinder, was provided for the outside porcelain tube.* The smaller
porcelain tube was not used at all, the sublimed matei'ial being removed
by taking out the foil and unfolding it. Cobalt from Sample III. was
then sublimed in the remodelled apparatus from a platinum boat. The
material obtained in this way gave results for the atomic weight alto-
gether too high, a circumstance due to large quantities of ^ilatinum
actually found in the sublimed bromide. Even here a small amount of
alkaline impurity existed, having crept in through the crack in the pla-
tinum foil. Hence, no more work was done with this material ; but
renewed precautions were taken to prepare by the older method cobaltous
bromide which should contain the smallest possible number and quantity
of impurities.
Four analyses were made with this new material, which proved in
spite of all our care to contain as much soluble matter as before. The
water extracts from these analyses were evaporated in a flat platinum
dish, which exposed a large surface of the solution to the air. This served
to oxidize completely the dissolved cobaltous hydrate, for the residues did
not become gray when heated, and upon the addition of water gave per-
fectly clear solutions. These residues were faintly blue, the color being
due doubtless to a trace of unreduced cobaltous bromide. For some un-
discovered reason, the amount of silver bromide obtained from the
residues was still too small to correspond to the weight of the residue,
if calculated as sodic bromide.
Tests were repeated for sulphuric, phosphoric, and silicic acids with
the greatest possible care, but still with negative results. In one analysis
the cobalt was determined in the residue and found to be only 0.00013
gram. It is possible, however, that these few tenths of a milligram dis-
argentic bromide. Hence 0.0010 gram of argentic bromide corresponds to 0.000626
gram of residue. This factor is used in calculating the results in Series I.
* Compare the preceding paper on Kickel, page 331.
364
PROCEEDINGS OF THE AMERICAN ACADEMY.
Weight of Residue.
Weight of AgBr
obtained from Kesidue.
NaBr iu Residue,
calculated from AgBr.
Unidentified.
grams.
.00300
.00643
grams.
.00439
.00978
grams.
.00240
.00536
grams.
.00066
.00107
crepaucy are due to the presence of all or at least several of the above
mentioned acids combined with sodium, each iu quantity too minute for
detection ; and for the present this will have to rest as the explanation.
Sixth Series. CoBr.i : Co.
No.
of
Anal.
Sample
of
Colirj.
Observed
Weight of
Cobaltous
Bromide
in Vacuum.
Observed
Weight of
Cobalt
in Vacuum.
Weight of
Residue.
Corrected
Weight of
Cobaltous
Bromide.
Corrected
Weight of
Cobalt.
Atomic
Weight of
Cobalt.
grams.
grams.
grams.
grams.
grams.
9
III.
5.11197
1.38027
0.00306
5.10891
1.37721
59.016
10
III.
6.41822
1.73333
0.00483
6.41339
1.72850
58.999
11
III.
6.60707
1.78778
0.00902
6.59805
1.77876
59.021
12
III.
3.03497
0.82249
0.00643
3.02854
0.81606
58.982
A
verage .
. 59.004
Average of Series V. and VI,
Co = 58.998.
This final average differs from that published before by only about
one part in ten thousand ; but iu comparing the two one must remember
that the material used last year must have been contaminated by the
same impurities which have been discussed in this paper. If the im-
purity contained as much bromine as cobaltous bromide contains, it would
have had no effect upon last year's results. In the case of nickel, where
the impurity consisted wholly of sodic bromide, the effect of correct-
ing the observed results in the paper of 1897 was to raise the atomic
weight of nickel from 58.688 to 58.703. The impurity from our cobalt-
ous bromide, on the other hand, contained unknown substances in quau-
tities so small as to elude detection, but large enough to change the
sign of the corresponding correction. Thus 22.63 grams of cobaltous
bromide in Series V. (the series in which the materials most nearly
RICHARDS AND BAXTER. ATOMIC WEIGHT OP COBALT. 365
resembled those used ia Series II. and III. of last year's work) were
found to contain 0.02327 gram of impurity. If, as we may reasonably
suppose, all the residues obtained from this sample of material resembled
that found in analysis 5, Series V., this residue would have yielded
0.0372 gram of argentic bromide. We may now correct last year's
results by subtracting from the several weights of cobaltous bromide
proportional weights of impurity, and also subtracting from the several
weights of argentic bromide amounts of this substance corresponding to
the impurity. Making this correction, the atomic weight of cobalt would
be loioered 0.008, the averages of Series II. and III. becoming 58.987
and 58.979. Our uncertainty regarding the nature and amount of the
impurity thus involves an uncertainty of about one part in six thousand in
last year's results. In the light of all the circumstances, it is perhaps
safest not to attempt any correction of these values, but to accept them
uncorrected as subject to this possible error. The results are accord-
ingly giveu below in an uncorrected form.
The data just discussed obviously afford a basis for recording the total
percentage composition of the cobaltous bromide analyzed.
COMPLETE ANALYSIS OF COBALTOUS BROMIDE.
Based upon Series II., III., and V.
Per cent.
Cobalt (Series Y.) = 26.923
Total Bromine -
' (II.) 73.050% '
(III.) 73.053%
Aver. 73.051%
Bromine combined , -r. «^,
J . , „ , , y— 72.981
with Cobalt
/ Bromine actually [ ^y^.
r found in impurity C
i Remainder of -s
Total impurity . . — 0.103% =z^ impurity '= .033
( (chiefly Sodium) j
Total r= 100.007
Series VL, perhaps the best of all, is not included in this table because
the material used iu it was not quite identical with that used in the three
other series. Obviously it is possible to calculate two more ratios in-
volving the atomic weight of cobalt, in which the values for the bromine
in the bromide are compared with the cobalt found in it. Into this calcu-
lation the weio^ht of the bromide itself enters simply as a constant, and
an indifferent impurity (such as water) would be eliminated from the
2 AgBr
: CoBr^
2 Ag
: CoBr
CoBr„
Co
2 AgBr
Co
2 Ag
Co
366 PROCEEDINGS OF THE AMERICAN ACADEMY.
result. The table below includes all of the five possible ratios obtainable
from our work, series I. and IV. being rejected because they were merely
preliminary : —
(Series II.) 58.995
(Series III.) 58.987
(Series V. and VI.) 58.998
(Series II. and V.) 58.994
(Series III. and V.) 58.992
Average .... 58.993
This table, although giving an interesting statement of the possible
combinations, does not yield a fair average, — for Series V. is introduced
three times, Series II. and III. each twice, while Series VI., which is at
least as accurate as the others, appears only once. A fairer method
would probably be to avoid all hypotheses and combinations, and assign
to each of the four series equal weight, as follows : — ■
Series II. (uncorrected) Co = 58.995
Series III. (uncorrected) Co = 58.987
Series V. (corrected) Co = 58.992
Series VI. (corrected) Co = 59.004
Final Average . . . 58.995
Obviously it makes but little difference which method we adopt : the
averages are essentially identical. The highest individual experimental
result among all these determinations was 59.021, and the lowest 58.955,
the average variation from the mean 58.995 being 0.012. Because
these results are less concordant than one could wish, and the conclusion
is somewhat less positive than that reached in the case of nickel, the
atomic weight of cobalt is being further studied by radically different
methods in this Laboratory.
In a recent article,* Professor Winkler calls attention to some possible
errors, in the work upon both nickel and cobalt published last year.
That the disagreement between his results and ours is due to the metliods
employed by him in his work upon these two elements has been suf-
ficiently shown in the preceding paper upon the atomic weight of nickel.
In addition, however, his several suggestions concerning our own work
should obviously be reviewed and discussed in detail.
His specific criticisms are four in number. First, he suggests that the
* Zeitschr. Anorg. Chem., XVII. 236.
RICHARDS AND BAXTER. ATOMIC WEIGHT OF COBALT. 367
porcelain tubes might have been attacked during the sublimation of the
bromides, with the introduction of foreign bromides into the nickelous
and cobaltous bromides. He goes on to state that the bromides were
dried in an acid atmosphere, and probably retained hydrobromic acid
after this gas had been displaced by air. His third criticism is that the
nickelous bromide in the earlier analyses contained nickelous oxide
which had to be determined and subtracted ; and his final objection
applies to the use of the Gooch crucible.
Answers to the greater part of his criticisms can be found in the very
articles which he criticises. In one case only can his view be substan-
tiated; — the porcelain tubes are really attacked. That this flaw was
a possibility we realized at the time ; but we also realized the smallness
of the error introduced by even a comparatively large amount of such
impurity. This matter has been already discussed in detail, both in this
paper and in the paper upon the atomic weight of nickel.
There are two possible ways in which hydrobromic acid could have
been retained: — by adsorption and by inclusion. At the high tempera-
tures employed the adsorption must have been very slight, and the
long process of washing witli an indifferent gas was favorable to the
elimination of any tendency in that direction. "While the inclusion of
liquids is a very serious possible cause of error, that of gases is usually
negligible because of the small mass involved. For this reason crys-
tallization from solutions is far less satisfactory than sublimation as a
means of purification.
That as a matter of fact our bromides were neutral there is no lack
of evidence. The possibility of acidity had occurred to us also, but rea-
soning from analogy we had decided that this possibility was rather an
improbability. Bromides and chlorides of barium and strontium, heated
in the same way in a dry acid atmosphere, after the acid has been dis-
placed by dry air, give absolutely neutral reactions with methyl orange.*
With cobaltous bromide the end point is not as easy to detect as with the
before mentioned halides on account of the color of the dissolved salt,
but colorimetric comparison makes it possible to distinguish the change
very accurately. A solution of our cobaltous bromide containing methyl
orange perceptibly changed color upon the addition of the minimum
amount of hundredth normal acid necessary to produce a change of color
in pure water containing methyl orange, showing that the salt must have
been very near if not at the turning point. As a final test, potassium
* These Proceedings, XXIX. 59, XXX. 373.
368 PROCEEDIiNGS OP THE AMERICAN ACADEMY.
bromide was sublimed iu the apparatus which had been used for the prep-
aration of the cobaltous bromide. The sublimate was then heated in a
current of dry nitrogen and hydrobromic acid gas, and finally, when cool,
the nitrogen and acid were displaced by dry air, just as in preparing the
cobaltous bromide for analysis. This potassic bromide upon solution
gave an absolutely neutral reaction with methyl orange. Taking into
consideration these three points, one cannot believe that enough hydro-
bromic acid was retained to have had an appreciable influence on our
results.
The third criticism, objecting to the ftict that in some analyses a small
amount of nickelous oxide was found in the bromide, is an unfortunate
one. A careful perusal of the work would have shown that only in the
preliminary series of results was this the case, and that this series does
not enter into the final average, although its results differ only by a very
small amount from those subsequently obtained with material free from
oxide. As far as the nickel is concerned, a conclusive proof of the ab-
sence of acid is afforded by this very fact that the earlier determinations,
in which it was necessary to filter off a residue of finely divided nickelous
oxide, gave results no higher for nickel than the later results.* Hence
the second and the third criticisms are obviously inconsistent with one
another.
The advantaces of the Gooch crucible are too well known to need
mention. Professor Winkler's specific objection to the collecting of dis-
placed asbestos upon an ordinary filter affects only an amount of a few
tenths of a milligram ; and a i^roof that no error was introduced in this
way lies in the fact that in every case the amount of silver bromide found
agreed very closely with the amount of silver necessary to complete the
reaction, where all but a few tenths of a per cent of the silver was
weighed out, and the remainder was added volumetrically. The following
brief table will make this clear.
Tlius, in the cobalt work, 18.1G302 grams of silver yielded 31.61642
grams of silver bromide, — a ratio of 57.448 to 100.000, — while in the
case of nickel, 15.51556 grams of bromide gave 26.67078 grams of silver
bromide, — a ratio of 57.444 to 100.000, while Stas found 57.445.
From the cobalt work AgBr : Ag = 100.000 : 57.448
From the nickel work AgBr : Ag = 100.000 : 57.444
From Stas's work AgBr : Ag = 100.000 : 57.445
* The slight colloidal solubilitj' of nickelous hydroxide was evidently destroyed
by the presence of large amounts of nickelous bromide, as one would expect.
RICHARDS AND BAXTER. — ATOMIC WEIGHT OF COBALT. 369
Thus the last of Professor Winkler's criticisms is sufficiently answered.
After necessarily dwelling at such length upon disagreements, it is a
pleasure to emphasize other points in which we agree with Professor
Winkler. The evidence of our work, together with Dr. Cushman's,
strongly supports Winkler's contention that nickel and cobalt, as we
knew them of old, cannot contain more than an infinitesimal amount of
any unknown element. Several radically different methods of prepara-
tion and many fractionations invariably led us to constant atomic weights,
within a reasonable limit of experimental error ; and we are forced to
conclude that the familiar properties of these common and useful metals
are to be ascribed to elements as definite as any of the seventy-five. It
is needless to point out also that we agree with Professor Winkler in
assigning to cobalt a higher atomic weight than to nickel, in spite of the
conflict of periodicity with rhodium and palladium. According to our
results, the atom of cobalt, weighing almost exactly 59.00, is very nearly
half of one per cent heavier than that of nickel.
Cambridge, October 29, 1898.
VOL. XXXIV. — 24
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. No. 15. — March, 1899.
IV.— CONTRIBUTIONS FROM THE HARVARD MINERALOGICAL
MUSEUM.
A COMPARATIVE STUDY OF ETCH-FIGURES. THE
AMPHIBOLES AND PYROXENES.
By R. a. Daly.
With Fouk Plates.
A COMPARATIVE STUDY OF ETCH-FIGURES. THE
AMPHIBOLES AND PYROXENES.
By R. a. Daly.
Presented by J. E. Wolff, December 14, 1898. Received January 18, 1899.
CONTENTS.
PAGE
Introduction 374
Previous Notices of Etcli-Pits on Anipliiboles 374
Reasons for the Present Investigation 375
Terminology 375
Discussion of Etch-Figures in the Microscope. Tlie Illustrations 37()
Materials of Study • 378
Methods used in the Investigation 381
Advantage of using Cleavage Pieces 382
Necessity of fixing the Conditions of Etching; Nature, Concentration,
and Temperature of the Solvent 383
The Optimum Exposure for the different Species in the Case of Hydroflu-
oric Acid 385
Attackability of the different Species and of the different Faces on the same
Species 386
Description of the Etch-Figures 391
Etch-Figures on (110) 391
Actinolite Type 391
Hornblende Type 394
Wolfsberg Sub-type 394
Krageru Sub-type 398
Edenville Sub-type 399
Philipstad Sub-type. Especial Anomalies in the Behavior of the
Philipstad Hornblende 399
Etch-Hills on Hornblende (110) 402
Glaucophane Type 404
Riebeckite Type 404
Arfvedsoriite Type 404
Etch-Figures on (010) 405
Non-aluminous Aniphiboles. Experiments on the Dilution of the vSol-
vent (Hydrofluoric Acid), and on its Mixture with Sulphuric Acid 405
Aluminous Amphiboles 411
Etch-Figures on Faces other than (110) and (010) 412
The Orthopinacoid 413.
374 PROCEEDINGS OP THE AMERICAN ACADEMY.
PAGE
The Dome (101) . 413
Isomorphism in the Monoclinic Amphiboles 415
Holohedral Character of the Monoclinic Amphiboles 418
Comparison of the Amphiboles and Pyroxenes as to Etching Properties . . 419
Crystallographic Orientation of the Amphiboles 422
Optical Orientation of an Amphibole Crystal or Cleavage Plate by Means of
Etch-Pits 423
Etch-Figures on Anthophyllite and on Gedrite. Orthorhombic and Holohe-
dral Character of these Minerals 424
Etch-Figures on Aenigmatite 425
Summary of Conclusions 426
Introduction.
Among the larger groups of rock-forming silicates, there is none
perhaps which, in the present state of our knowledge, ofTers more dif-
ficulties in the determination of systematic relationships than the amphi-
bole family. Its importance for the petrographer needs no emphasizing
here, yet it is he who has to meet the difficulties of classification and
discussion under the most disadvantageous circumstances ; in general, by
reason of its association, the amphibole of an eruptive rock or of a crys-
talline schist lacks crystal form, and, because of numerous inclusions, it
may often be impossible to procure a reliable chemical analysis of the
mineral. Thus deprived of two principal aids to diagnosis, the worker
in rock-forming amphiboles must make the most of the other criteria
which offer themselves. In so doing, he may eventually be able to repay
the pure mineralogist for his services to the study of the crystalline rocks
and present new considerations that can lead to the interpretation of the
mineral species as such without relation to rock genesis or rock classifica-
tion. Of the methods which, so fnv, have been almost completely neg-
lected by petrographers in the investigation of amphiboles, is that of the
use of etch-figures on planes of the more important zones. I propose
in the following pages to record briefiy certain results I have obtained
while breaking ground in this new field of inquiry.
The first, and so far as I have learned, the only published reference
to actual experiments in etching an amphibole, occurs in Boricky's first
essay on microchemical methods.* Plate II. Fig. 7 of his work repre-
sents a hornblende etched with fluosilicic acid on the clinopinacoid. The
reference in the text to this drawing was occupied with the mention of
the chemical reaction, and especially of its products, — nothing further.
* Archiv d. naturw. Landesdurchforschung von Bohmen, III. Prague, 1877.
DALY, — ETCH-FIGURES ON AMPHIBOLES. 3T5
Sir David Brewster observed his " O^Jtical Figures " on a few hornblendes
characterized by natural pits of corrosion.* Since nothing has been
done towards a comparative review of etching phenomena with respect
to the amphiboles, I shall state some of the reasons why the present
reseai'ch was begun.
(1) In the first place, it has been considered a good opportunity to
test once more how far etch-figures are dependent on the method of
attack, and to devise a convenient and uniform method for the group in
question. (2) Will the etching process furnish any information as to
the attackability of amphiboles in hydrofluoric acid 1 (3) Are the etch-
figures variable in shape with the chemical composition of the mineral?
Will they throw any light on the problem of isomorphism among the
amphiboles ? (4) Will the amphibole figures by comparison with those
on the corresponding faces of pyroxene, tend to strengthen the parallel
between the two mineral groups ? (5) Can cleavage pieces and crystals
of amphibole be crystallographically oriented by means of etch-figures ?
(6) Will the latter give us any data on the vexed question of the best
standard orientation of amphiboles as a whole? Is Tschermak's or
Nordenskiold's recommendation better fitted to disclose the many va-
ried relationships of the group ? (7) Are the amphiboles holohedral ?
(8) Are anthophyllite and gedrite really orthorhombic ? (9) Inciden-
tally, in connection with the attempt to solve these problems, I have
compared etch-figures using hydrofluoric acid with those obtained with
the corrosive alkalies.
Now, in order to pave the way for a concise description, and perhaps
readier understanding of the following discussion, a few lines may be
taken to define a certain number of terms which have been introduced
with more or less technical import. Several of these ai-e literal or
slightly modified translations of the valuable German names or phrases
of Becke, Baumhauer, Leydolt, and others.
The etch-figure itself may be a cavity of corrosion, an " etch-pit "
("pit of corrosion," Aetzgriibchen, Aetzvertiefung), or it may be an
etch-hill (Aetzhiigel), a residual boss standing up in relief above the gen-
eral surface of the crystal. Either etch-pit or etch-hill is bounded by
•' figure-faces " f (Aetzflachen), in general manifestly plane, sometimes
* Phil. Mag., 1858, Vol. V., p. 16.
t Molengraaf's nomenclature seems unfortunate in relegating the short, useful
word " Aetzflache " to the comparatively unimportant curved surface which often
truncates a crystal edge when exposed to corrosive agents, and compelling us to
speak of the figure-face of a pit as an " internal etch-face " (innere Aetzflache). —
Zeit. fur Kryst, 1888, Bd. XIV. p. 174.
376 PROCEEDINGS OF THE AMERICAN ACADEMY.
apparently curved, faces. Many pits normally exhibit a figure-face
parallel to the plane etched ; it may be designated the " bottom-face."
An " etch-zone " is a zone containing two or more figure-faces (" Aetz-
zone" of Molengraaf, not equivalent with the " Aetzzone " of Becke).
The periphery of the pit where its figure-faces intersect the plane at-
tacked is here called the " outline " of the figure-, and each " edge " that
composes it may thus be a straight or curved line as its figure-face is
plane or curved. A " corner " is the point where the etched plane and
two adjacent figure-faces meet.
As the process of etching continues, a pit usually increases in size,
often (depending partly on the symmetry of the etched plane) changes in
shape of outline, and, in many cases, deepens as the result of replace-
ment of early formed figure-faces by others of different indices, accom-
panied by the necessary " diminishing " of the " bottom-face " if there be
one present (cf. pits on apatite, calcite, galenite, gypsum, zinc-blende,
etc.). These changes in the figures may be continuous, but often have
rather the look of being intermittent, the replacement of one figure-face
by another taking place as a momentary cliange, faces of intermediate
indices not appearing at all. The first stage of development of a pit
may be called its " initial " form. The development ends where the out-
line besfins to be seriously impaired by the solution of the surrounding
part of the etched surface. Just preceding this point in the history, the
pit may be called " mature," and the process intervening between the
initial and mature stages is that of " maturing." Von Ebner's " instan-
taneous " and " retarded " types are connected by transitions, but are
not easily to be compared to " initial " and " mature " figures, since his
types refer simply to the length of time required to develop the pits, and
are not restricted to the use of one solvent.*
Discussion of Etch-Figures in the Microscope.
It is believed that a description of the etch-figures as seen in reflected
light with vertical incidence would be, on the whole, of more value than
an account of the same figures examined under other conditions (trans-
mitted light, Lichtschimmer). Within certain limits, this method is
easily carried out with the aid of the modern appliances to be found on
the large models of most petrographical microscopes, and thus a new
etch-figure can in a few minutes be compared in its main features with
* Sitzungsber. der Akad. d. Wissen., Vienna, 1885, Bd. XCI. p. 776
DALY. — ETCH-FIGURES ON AMPHIBOLES. 377
those already established for the corresponding face and mineral group, or
with analogous figures belonging to other species. Furthermore, such a
description may be made in cases where any determination of the indices
of the figure-faces is impossible on account of the absence of " Licht-
schimmer," due to various causes, as fibrosity, minuteness of figures, cur-
vature of the surface studied, etc. But it is necessary to recognize that
a complete analysis of a figure is not j^ossible in many cases, even under
the most favorable circumstances. Tliis is true, for example, of the pits
on the prismatic faces of amphibole and in the vertical zone of most
monoclinic minerals. Relatively low powers of the microscope must
always be used, since contrasts of dark and light are speedily lost above
200 diameters, and thus it often cannot be decided whether an apparently
rounded figure-face may not really be one compounded of many small
faces, according to the well known examples of Becke, Baumhauer, and
others. Hence, inasmuch as it is not practicable to determine in the
microscope the elements (faces, angles, symmetry, etc.) with the same
precision and detail with which we can define a crystal, it becomes ad-
visable to choose certain elements of the figure that are sufficient to fix
its general shape. Such elements will be those which can be directly
measured in the microscope and with a maximum of exactness. They
will include straight sides and the angles between them as well as the
special angles between curved sides characteristic of each figure. These
elements, too, had best be such as can be recognized on very small figures
of a given category, since in some varieties it may be feasible to produce
figures only relatively very minute. Lastly, we must have a base-line of
reference for all measurements ; — in amphiboles, there is an excellent
one, the trace of the cleavage, which generally makes it unnecessary to
search out the directions of edges bounding the crystal-face.
For the convenient examination of figures on (110), it is well to use
prisms with sides as smooth as possible, so that the mineral will lie flat,
and the plane to be studied perpendicular to the axis of the microscope.
In the study of terminal planes, or of material with which such perfect
prisms are not obtainable, the crystal or cleavage piece may be readily
brought into the desired position by mounting it on an object-glass with
wax and then adjusting it so that the simultaneous reflection of a ray of
light from the glass and the plane may occur. "With lustrous faces, this
adjustment can thus be carried out with a close degree of accuracy.
The microscope used was a Nachet, provided with an apparatus for
vertical incidence of the light that illuminated the crystal-face. The
light was led through a collimator attached to an Auer lamp. Below
378 PROCEEDINGS OF THE AMERICAN ACADEMY.
the collimator, the metal casing of the lamp was pierced so as to allow
of a source of transmitted light for getting extinctions in cleavage plates.
Orientation could thus be eifected by means of extinction (when the
amphibole was known), as well as by using terminal planes.
For reasons explained further on, Tschermak's orientation is adhered
to throughout this paper (/3 = 73° 58').
The photographic illustrations I owe to the skill of M. Monpillard of
Paris. The difficulties in reproducing anything like the sharpness of the
etch-figures on amphiboles, especially on faces other than that of tlie
fresh cleavage flake, are very great and fully explain any lack of defi-
niteness that may be observed in the micro-photographs. The diagram-
matic figures were drawn by means of a camera lucida and bring out
more clearly than the photographs, the points of essential resemblance
and dissimilarity which need emphasis.
Both in the diagrammatic figures and the wash-drawings of Plate I.
the cleavage trace on each etched surface is represented by a straight
line, which the reader will immediately recognize. This line is replaced
in the photogravures by the longer edge of the page. The top of the
crystal will as usual be directed toward the top of the page in the case of
planes in the vertical zone ; the front of the crystal toward the bottom
of the page for terminal planes.
Materials of Study.
The work the results of which are embodied in the following pages was
begun in the laboratory of Professor Rosenbusch at Heidelberg, where
the initial experiments were carried on with crystals of Vesuvius horn-
blende obtained from the collection of the Mineralogical Institute, and
with Zillerthal actinolite, St. Gothard (?) tremolite, and Bohemian horn-
blendes from the private collection of Professor V. Goldschmidt of
Heidelberg. To this material were added 41 specimens from the Im-
perial Museum, Vienna, through the kindness of Professor Berwerth,
some 20 others from the collection at the Jardin des Plantes, Paris, due
to the liberality of Professor Lacroix, six fine crystals of aeuigmatite and
arfvedsonite from Professor Ussing, Copenhagen, classic glaucophane
from Professor Barrois, Lille, and much American material from Pro-
fessor Hobbs, Madison. To these gentlemen I should like to express
my hearty thanks for the privilege of securing so many specimens with
little trouble to myself, — material which in many cases is classic, and of
considerable value from the mere monetary point of view. Without the
use of so many representatives of the group, I should not have felt
DALY. — ETCH-FIGURES ON AMPHIBOLES. 379
enough confidence in the generalization from certain types to all types.
The systematic investigation was almost entirely pursued in the labora-
tory of Professor Lacroix, and I desire to acknowledge, in particular,
his generosity in placing at my disposal the apparatus necessary for
the etching.
The list of specimens is self-explanatory. For convenience I have
followed Dana's classification closely, without however implying an
absolute adherence to all its details. I have inserted the catalogue num-
bers of the Museum specimens, sometimes the date of collection, and,
where possible, leading references to original papers in connection with
those specimens that have furnished particular descriptions or material
practically identical with them. An asterisk denotes a specimen "pre-
sumably similar to the classic material from the same locality. The
source of each specimen is indicated in the general list by letters pre-
fixed to the number corresponding. H. — Heidelberg, V. = Vienna,
P. = Paris, C. = Copenhagen, M. — Madison, L. = Lille.
Amphiboles proper.
A. Orthorhombic amphiboles.
a Anthophyllite.
P. 1. Kongsberg, Norway : —
Des Cloiseaux, Nouv. Recherch., 1867, p. 541.
Michel Levy and Lacroix, Min. des Roches, 1888, p. 149.
P. 2. Nunangiast, Greenland. —20-125.
P. 3. Regardsheina, Norway.
b. Gedrite.
P. 4. Gedres, France : —
Dufre'noy, Ann. des Mines, 1836, Vol. X. p. 582, etc.
B. Monoclinic Amphiboles.
a. Non-aluminous.
(1) Tremolite.
H. 6. St. Gothard (?).
P. 6. Siberia. — " Grammatite." — 2-855.
P. 7. Faroe Islands. — 64-63.
V. 8. Newport, Bucks Co., Mass. (?) A. c. 4154.
*M. 9. Gouverneur, N. Y.
(2) Actinollte (witii certain allies).
*H. 10. Zillerthal.
*P. 11.
P. 12. Syra.
P. 13. Gellivara, Asia Minor, 68-31.
V. 14. Killaersarbik, Greenland. A. e. 924, 1818.
V. 15. Anaitsirksarvik, Greenland. A. u. 98.
*V. 16. Orange Co., N. Y. E. 5606, 1888.
V. 17. Arendal, Norway. A. e. 961, 1825.
380 PROCEEDINGS OF THE AMERICAN ACADEMY.
V. 18. Ottawa River, Canada. A. e. 968, 1826.
P. 19. (Smaragdite), Greenland.
*V. 20. (Richterite), Langban, Sweden. A. a. 509.
*V. 21. " " " G. 3122, 1894.
*V. 22. (Astochite), " " G. 4080, 1895.
(3) Cummingtonite.
*M. 23. Cummington, Mass.
(4) Grunerite.
P. 24. Collobrieres, Dep. du Var, France : —
Lacroix, Bull. Soc. Min., 188G, p. 40.
Mine'raux des Roches, 1888, p. 144.
b. Aluminous.
*P. 25. (Edenite) Edenville, N. Y. 64-145.
*V. 26. (Pargasite) Pargas, Finland. A. e. 928, 1848.
V. 27. " " " A. o. 469.
Berwerth, Sitzb. Akad. Vienna, 1882, Bd. LXXXV. p. 158.
*V. 28. (Pargasite). A. e, 929, 1826.
V. 29. (Carinthine) Saualpe, Carinthia. A. o. 464 and 465.
Tschermak, Min. und Petrog. Mittheil., 1871, p. 38.
*H. 30. (Syntagniatite of Breitliaupt), Vesuvius.
*V. 31. Krager5, Norway. G. 3287, 1894.
V. 32. Arendal, Norway. A. f. 2., 1827.
V. 33. " " A. e. 967.
V. 34. " " A. e. 891 and 892.
V. 35. " " A. e. 897.
V. 36. *' " A. e. 898.
V. 37. " " A. o. 4.32.
V. 38. " " A. o. 434.
V. 39. > " " A. f . 3, 1824.
V. 40. Norway. A. o. 459.
V. 41. Philipstad, Sweden. A. o. 458.
*V. 42. Kafveltorp, Sweden. 1889.
*V. 43. Wolfsberg, Bohemia (in basalt). A. a. 1860.
V. 44. Orbus, Kupferberg, Bohemia. A. o. 446.
*H. 45. Bilin, Bohemia.
V. 46. Mayenegg, near Kupferstein. B. e. 5378, 1838.
V. 47. Easton, Pennsylvania. A. e. 971, 1826.
V. 48. Worthington, Mass. A. e. 902, 1825.
V. 49. Kangerotvarsvik, Greenland. A. e. 922, 1818.
V. 50. Edenville, N. Y. (greenish-black). A. e. 913, 1829.
V. 51. ^ " " (dark green). A. e. 901, 1827.
V. 52. Ge'bel Gharib, Arabia. B. d. 6369, 1877.
P. 53. (Ganisigradite) Gamsigrad, Servia. 97, 351.
Lacroix, Bull. Soc. Min., 1887, Vol. X. p. 147.
*V. 54. Wolfsberg, Bohemia (twin). F. 3850, 1890.
P. 55. Eiveau Grand, Mont Dore.
Lacroix, Mine'ralogie de la France, 1893-95, Vol. I. p. 663.
DALY. — ETCH-FIGURES ON AMPHIBOLES. 381
Glaucophane.
*P. 5G. He de Groix.
L. 57.
Barrois, Am. Soc. Geol. du Nord, 1883, p. 19.
P. 58. Oulx, Savoy.
*P. 59. (Gastaldite) Champ de Praz, Val d'Aosta. 90, 197.
Crossite.
*P. GO. Berkeley, Cal.
Riebeckite.
P. 61. St. Peter's Dome, Colorado. 89, 6.
Arfvedsonite.
*P. 62. Kangerdluarsuk, Greenland.
*C. 63.
(Barkevikite.)
*P. 64. Barkevik.
Aenigmaiite.
*P. 65. Naujakasik, Greenland. 95-190.
*C. 66.
Bronzite.
*P. 67. Kraubat, Styria.
P. 68. Greenland. 9-25.
Hypersthene.
P. 69. St. Paul's Island. 35-2631.
Lacroix, Min. des Roches, 1888, p. 261.
Diopside.
*P. 70. Ala.
A ugile.
P. 71. Puy de la Rodde.
Gonnard. Cf. Lacroix, Mine'ralogie de la France, Vol. 1, p. 578.
Fowlerite.
*P. 72. Franklin, New Jersey.
Methods Used.
The researches of recent years on figures of corrosion have shown that
the most fruitful results are obtained by quantitative methods, that is, by
the use of reagents under definite specified conditions, and by close
measurement of the figures. It hardly needs mention that there is
much yet to be learned regarding the coliesional properties of the
species belonging to each of the great mineral groups, as well as re-
garding the similar relations which may exist between corresponding'
members of two different families. Just as we may describe as accu-
rately as possible the hardness or the fusibility, the specific gravity or
the optical properties, of a species, not only to fix it as an independent
type, but also to relate it to the other members of its own family and
to other mineral groups, so we believe it is possible to construct with
382 PROCEEDINGS OP THE AMERICAN ACADEMY.
some precision a scale by means of which crystal faces may be defined
as to molecular cohesions. This has, in fact, been accomplished in
certain cases ; but, in general, etch-figures have only been used quali-
tatively, so to speak, and as yet there has been not enough of continuity
of method from one investigation to another to make possible detailed
comparison of species with species in this matter of facial cohesion.
In the particular case of the group here considered, it has seemed
desirable to make it possible to reproduce the conditions finally selected
for etching, so as to permit of the discussion of new amphiboles with
the aid of data already in hand. May it not be possible that the
contradictory results of certain observations is simply due to difference
in methods ? Thus, Penfield, Meyer, and Bomer found the basal
plane of quartz characterized by etch-hills when hydrofluoric acid was
used,* while with the same solvent Gill obtained pits of corrosion.f
For these and other reasons noticed below, I have tried to establish
a constant method which would give good results with all varieties
of the great amphibole family, and one which can be extended to the
pyroxenes and other silicates. That such a method be wrought out, it
was necessary that some preliminary experiments should be made, for
reasons, some very obvious, others less so, all of which I shall summarize
in this connection.
1. It is generally advisable to choose prominent crystal-faces, usually
those of simple indices aud those parallel to cleavages. That one would
select such faces is to be expected, but I think this point should be espe-
cially in mind if comparisons are to be made as widely as possible. An
advantage in choosing cleavage pieces or the corresponding crystal-face
is evident in those groups where certain members appear only as allotrio-
morphic individuals in rock aggregates. Even in these cases, their etch-
figures may be produced on a good cleavage when figures on other planes
would ouly be possible on artificial faces. Fortunately, too, for the
discussion of rock-forming amphiboles as well as of the group as a whole,
the cleavage pieces give the sharpest and most regularly developed
figures to be obtained on any given crystal. J The prismatic cleavage of
amphiboles has thus a superior claim to attention, and I have accordingly
laid most stress on this important face in the course of the present inves-
tigation. Next to these, pinacoids will naturally give the most useful
* Trans. Connecticut Academy, 1889, p. 157. Neues Jahrb. fiir Min., etc., 1891,
Beil. Bd. VII. p. 534.
t Zeit. fiir Kryst., Bd. XXII. p. 111.
t Cf. Bauiuhauer, Resultate der Aetzmethode, p. 3.
DALY. ETCH-FIGURES ON AMPHIBOLES. 383
results. As a matter of fact, I have found (010) and (100) to be among
the interesting faces for etching on amphiboles. Moreover, not only are
they very common planes ; they are also those most likely to reveal the
fundamental features of crystal structure.
2. Since the production of etch-figures on a bisilicate is in large part
the result of a chemical reaction between mineral and solvent, it is clear
that the figure will, in every case, depend on the chemical nature of both.
Two varieties apparently isomorphous and differing little in composition
may afford figures markedly different from each other, although produced
on planes with the same symbols and with the same solvent. Examples
will be noted in the sequel. Such being the fact, it is natural to conclude
that the figures shall be similarly sensitive to small changes in the solvent
also.* A striking illustration is to be found in the series of etch-pits
formed on (010) of actinolite by hydrofluoric acid in various states of
dilution in water or mixture with sulphuric acid. There is a steady
change in the orientation of the figure as the state of purity of the hydro-
fluoric acid is affected. Details concerning this phenomenon are given
in the section devoted to a description of figures on (010).
I chose hydrofluoric acid as the universal attacking reagent on account
of its convenience and efficiency. Inasmuch as its working depends on
the degree of concentration of the acid, it becomes necessary to fix on some
particular grade of acid. A number of trials soon convinced me that the
concentrated commercial water solution is for general purposes the best.
Not needing special preliminary preparation, it is easily obtainable in any
desired quantity ; experiment showed that it gave the most satisfactory
figures just as the alums afford the best results with the solvent in an
active state, f There is one danger to be guarded against, namely, the
loss of concentration with prolonged heating of the acid ; hence the ad-
vantage of easy renewal of the reagent.
Temperature and its function, the duration of immersion, are now well
established to have a strong influence on the process and effects of etch-
ing. A new variable must thus be considered. As a result of a large
number of trials made both incidentally and with this distinct purpose in
mind, I found that good figures could be produced at many different tem-
peratures ; thus, Zillerthal actinolite will yield well developed pits of
corrosion when boiled one minute in HF, or at three minutes in HF
* Cf. Von Kobell, Sitzungsber. Miinchner Akad., 1862, p. 199 ; Ben Saude,
Ueber den Perowskit, Gottingen, 1881; Meyer, Neues Jahrbuch, 1883, Bd. I.
p. 77.
t Klocke, Zeit. f iir Kryst, 1877-78, Bd. II. p. 130.
384 PROCEEDINGS OF THE AMERICAN ACADEMY.
on a water bath, or, again, when immersed several hours in cold HF.
Among all these possibilities I have endeavored to secure for all com-
parative studies a temperature as nearly constant as possible, and this for
two reasons. In the first place, it would not be at all certain without
direct proof that the figures with any one reagent remain constant for all
temperatures. Bomer discovered that the form and orientation of the
figures on quartz produced by attack with HF were affected by the tem-
perature of the reaction.* It is reasonable to suppose that temperature
may have a corresponding effect on amphibole figures when the same
reagent is used. As will be noticed elsewhere, I have been able to deter-
mine no sensible variations in figures on (110) from this cause, but it
cannot be denied that they are present. Secondly, I wished to tabulate
the amphiboles with reference to their power of resisting solution in the
etching process. The standard temperature chosen for these reasons is
that of the water bath, one that is nearly constant, attained with no dif-
ficulty, and found to suit the necessities of the case very well.
To secure a standard temperature for the reaction repeatedly and expe-
ditiously demanded in addition a certain amount of arbitrary treatment,
since the amount and initial temperature of the acid have not yet been
allowed for. A platinum crucible of the usual slightly conical form
and with a diameter at the bottom of about 4 cm. is filled to a depth of
1 cm. with the cold acid ; the mineral, resting in a platinum net, is im-
mersed, and at the same time the crucible is placed 1 cm. deep in the
steam of a water-bath which is kept constantly at 100 degrees Centigrade.
The attack is readily checked at any moment by lifting the platinum net
and plunging it with the mineral into water. The coating of fluorides
could be readily removed by brushing the mineral in running water, or
by dissolving them in hot concentrated hydrochloric acid.
Another point of inquiry in connection with formation of an ideal
method would be the effect of increasing the energy of the reaction by
the agitation of the acid during attack. Klocke, in his classic research
on the alums, found that on agitating the solution in which corrosion pits
were forming, the figures grew larger rapidly, due, as he stated, to the
dissipat'on of the " Hof " (Losungshof) of liquor near the figure which
had become laden with the products of solution. t Experiment of the
same kind was carried on with basaltic hornblende, but no material im-
provement was effected on the sharpness of figures produced without
* Neues Jahrbuch, 1891, Beil. Bd. VII. p. 538.
t Zeit. fur Kryst., 1877-78, Bd. II. p. 298.
DALY. — ETCH-FIGURES ON AMPHIBOLES.
385
agitation. I am inclined to think that convection currents in the warmed
acid are sufficient to perform the same function.
Ill thus fixing on a universal solvent, its temperature, state of concen-
tration and of convention, we have narrowed down the variables of the
process to one, the duration of attack. This facility in arriving at the
conditions of a uniform method of etching is not possible when the caus-
tic alkalies are employed and hence these will be referred to only inci-
dentally in the following pages. Unless the contrary be stated, the
standard conditions of etching are to be understood in every experiment.
The Optimum Exposure of the Different Species.
Here the study was confined to cleavage pieces. It was naturally
found that as the figures increased in number, they also increased in size,
any changes in shape not being sufficient to remove them from the cate-
gory of "primary" figures (Becke). In most cases, the general cleav-
age surface showed no serious roughening as the figures grew and the
attack was allowed to continue nearly to the point where overlapping of
the pits (generally aggregated in groups) would occur. This length of
exposure usually gave the best figures for study ; hence I have called it
the " optimum " duration of attack. Since the determination of the
optimum length of immersion was a matter of considerable labor, the
result of several trials with almost all varieties, I have thought it worth
while to tabulate the results obtained with a certain number of speci-
mens. It is to be understood that the following table is only approxi-
mately accurate. Perfectly fresh acid was not used in every case, and,
of course, the longer the acid remains on the water bath, the weaker it be-
comes ; moreover, the amount of steam in the water bath is variable to some
extent, and thus the HF might become heated at different rates. Care was
taken to allow for such causes of variation from the true optimum.
Optimum Exposures.
Anthophyllite.
P. 11
P. 1 . . .
2 minutes.
P. 12
P. 2 . . .
2
le
P. 13
Gedrite.
V. U
P. 4 . . .
2
«
V. 15
Tremolite.
V. 16
H. 5 , . .
3
«
V. 17
V. 8 . . .
3
<(
V. 18
Actinolite and allies.
P. 19
H. 10 . . .
2i
«
V. 20
VOL. XXXIV.
-25
2J minutes.
91 "
2i.
1
4
n
2
2i
1
u
1 "
1| m. (Sraaragdite).
f m. (Richterite).
386
PROCEEDINGS OF THE AMERICAN ACADEMY.
V. 21 . .
. |m
(Richterite).
V. 44 . .
\ minutes.
V. 21 . .
Ifm
"
V. 46 . .
• 2i
(another specimen).
V. 48 . .
. 3
V. 22 . .
1 m
(Astochite).
V. 49 . .
. 2
Common and basaltic Hornblende.
V. 50 . .
. 1
P. 25 . .
2|m.
(Edenite).
V. 51 . .
. 1
V. 26 . .
2i-3 m.
(Pargasite).
V. 52 . .
• 1
V. 27 . .
" m.
"
P. 53 . .
. 1 (Gamsigradite)
V.28 . .
" m.
'(
Glaucophane.
V. 29 . .
2 m.
Carinthine).
P. 56 . .
. 2 minutes.
H. 30 . .
2 m.
(" Syntagmatite '
). P. 58 .
(^) li
«
V. 31 . .
Ij minutes.
P. 59 .
• H
((
V. 32 . .
1|
Riebeckite.
V.33 . .
H
P. 61 . .
• 1
«
V. 34 . .
1
Arfvedsonite
V. 36 . .
i
P. 62 .
• i
«
V. 37 . .
i
(Barkevikite
.
V. 39 . .
1
P. 64 . .
• i
"
V. 40 . .
1
Aeuigmatite.
V. 41 . .
H
P. 65 .
• \
"
V. 42 . .
1
C. 66 . .
■ i
«
V. 43 . .
n
The average optimum exposure of these groups as indicated by the
optimum exposure can, then, be expressed somewhat as follows : —
Anthophyllite 2 minutes.
Gedrite 2
Tremolite 3 "
Actinolite 2 "
Richterite 1 "
Astochite 1 "
Edenite 2\ "
Pargasite 2| "
Common and basaltic hornblende l^^ "
Glaucophane If "
Riebeckite f "
Arfvedsonite \ "
Barkevikite \ "
Aenigmatite ^ "
Attackabilitt of the Amphiboles.
Referring as they do to the appearance of best figures on the prismatic
cleavage, without regard to their .size, these tables do not express the
attackability of the various species. Thus the average time for the
development of the sharpest figures on anthophyllite and actinolite is in
DALY. ETCH-FIGURES ON AMPHIBOLES. 387
each case, two minutes, yet the actual amount of material carried off in
sohuion from the very minute pits of the anthophylHte is extremely
small when compared with that removed in the process of excavating
the much larger pits on the monoclinic mineral. While working out
optimum exposures, I generally had opportunity to observe the incipient
stages of attack on cleavage-cracks and of the roughening of tlie whole
surface of the crystal. In this way a general impression of the i-elative
attackability of these minerals was gradually made upon me. I give
the seiies for what it is worth, beginning with the varieties most resistant
to hydrofluoric acid, and naming the others in order of less resist-
ance: — 1. The orthorhombic araphiboles. 2. Actinolite. 3. Ti-emolite.
4. Glaucophane. 5. The light-colored aluminous monoclinics. G. The
common green hornblendes. 7. The basaltic hornblendes. 8. The
Richterites. 9. Arfvedsonite. 10. Riebeckite (?). 11. Aenigraatite.
It will be seen that the resistance to solution decreases with increase in
soda and in sesquioxide of iron.
But not only are the differences of attackability due to differences in
chemical composition and crystal system, they are also strongly affected
by physical conditions irrespective of species. The physical influences
may entirely mask the attackability resulting from the chemical reaction
alone. The theories which have been made to explain the irregular dis-
tribution of etch-figures on a given plane by corresponding irregularities
in the grouping of the chemically active part of the solvent in use, can
have no application to many cases that have come under my notice during
the course of tlie present research. They are often rather to be explained
as dependent on a loosening of the original molecular structure of the
mineral by mechanical action without at the same time being accompanied
by chemical decomposition. The presence of submicroscopic cracks or
planes of parting (a superficial capillary zone) will necessarily give the
acid greater surface by which to attack and permit of a readier dislodg-
ment of the molecules from the grip of physical cohesion, A good anal-
ogy is found in the hardness of certain pseudomorphs ; manganite with a
hardness of 6 forms a pseudomorph after polianite (pyrolusite), but has
then an apparent hardness of only 3.
A few typical examples of this differential resistance to solution will
suffice for our present purpose. Two intergrown crystals of V. 42 ap-
parently of etjual freshness, each bearing the plane (Oil), were simulta-
neously immersed for several periods and examined at the end of each
interval for the relative progress of attack. The plane (Oil) of one
crystal was seen to have been affected decidedly sooner than the same
388 PROCEEDINGS OP THE AMERICAN ACADEMY.
plane of the other. In this case, the ready yielding of the former may
have been caused by its being exposed to more active convection currents
than the second crystal, due to position in the acid. But this explanation
cannot apply to V. 46, where the optimum exposures for cleavage flakes
from different ciystals, though from the same hand specimen, varied from
two and one half to five minutes, and the pits were of nearly equal size
(area and depth) on all the pieces. V. 34 afforded some light on the
question in the behavior of two terminated crystals, the (Oil) of the one,
less lustrous than the same plane of the other, was the more rapidly at-
tacked. The suggestion that the phenomenon is a result of alteration
can hardly be avoided. That the alteration may be an almost, if not
quite, exclusively physical one and not associated with a serious change
in the original chemical molecule of the hornblende, seems clear from
the facts observed in an experiment on V. 31. The hand specimen is a
" Krystallstock " composed of well defined individuals, tipped with asbes-
tus and occasionally showing patches of an asbestiform substance on the
sides of the crystals. At two minutes' exposure, one of these gave sharp
but relatively few figures on (110) near the point of attachment at the
end of the crystal. The other three fourths of the surface of the prism
was characterised by the appearance of numerous cleavage cracks which
gradually increased in number in the direction of the free (in the druse
unattached) end, and there was a simultaneous increase in the number of
pits. The development of the latter was independent of the asbestiform
patches, and, from the uniformity of color of the general surface, I have
concluded that there has not been chemical decomposition sufficient to
explain the differing rates of etching. It may be noted that the much
attacked end was directed upward in the acid, and any heating by direct
contact with the platinum at the bottom of the crucible would tend to
dissolve the unaltered extremity of the crystal the more rapidly.
As a general rule, the crystal-face (110) was observed to be less resist-
ant to solution than the parallel cleavage plates of the same individual
crystal, but the converse was often true.* Thus at one and two thirds
minutes, the cleavage showed much stronger attack than the correspond-
ing crystal surface (HO) of a specimen of V. 32, although here again
the original face was facing downward, the cleavage plane upward, in the
acid solution. The mineral was perfectly fresh in appearance, and be-
trayed no alteration to the eye such as seemed best to explain the same
relations characterizing other examples: e.g. V.8,V. 14, V. 17,V. 26,V. 42.
* Cf. Becke's experience with zinc blende, Min. und Petrog. Mitth., 1883, Bd. V.
p. 485.
DALY. — ETCH-FIGURES ON AMPHIBOLES. 389
In these cases, actual separation had taken place along the cleavage
planes in question long enough before the specimens were collected in
the field to have j)ermitted the physical processes of weathering to de-
stroy to some extent the original cohesion of those layers of the crystals,
while, for some reason (probably the protection of surrounding minerals),
the crystal-face escaped such disintegration. That there may be other
and more obscure physical differences which can explain the grouping of
etch-figures was suggested by Baumhauer, who attributes the zonal ar-
rangement of the pits on etched fluorites to them rather than to chemical
variations in the zones.* The same principle is illustrated in the expla-
nation of " Aetzgriiben," f linear aggregations of pits following direc-
tions of weak cohesional control. Thus, on etched cleavage plates, there
are commonly to be observed a regular grouping of figures in directions
which are crystallographically fixed on hornblende ; long rows of pits
parallel to the trace of the second vertical cleavage, or to the trace of
the rudimentary basal cleavage, without, in either case, there being actual
cleavage cracks opened which would affect the etching directly. Viola
has recently described the zonal distribution of pits on gypsum, the zones
being regularly fixed with reference to tlie axis of symmetry.! Whether
these zones represent the original stratification of molecular deposition or
are the result of secondary physical change, acting on a homogeneous
crystalline mass, the behavior of gypsum is another example with those
already cited to show the dependence of attackability on physical co-
hesion, when the latter varies according to the laws of crystal structure.
Incipient weathering and the development of secondary strains will favor
an irregular grouping of etch-figures that are structurally accidental.
These have been noticed in connection with the amphiboles. These facts
lead us to suspect that Ebner's hypothesis of irregularities in the solvent
can no better explain the differing attackability of a cleavage piece of
calcite in its various parts than the hypothesis of varying physical con-
ditions in the crystal itself. §
* Resultate der Aetzmethode, p. G. Mere heating may render actinolite fibrous.
See Doelter and Hussak, Neues Jalirbuch, 1884, Bd. I. p. 24.
t Ibid., p. 6.
t Zeit. fur Kryst., 1897, Bd. XXVIII. p. 575.
§ That no one tlieoretical cause yet adduced in explanations will suffice is clear.
Minute fracturings cannot, for example, pave the way for unequal etching in the
case of a growing crystal of alum in a saturated alum solution. It is simply neces-
sary to dilute the solution slightly in order to etch the newly made surface ; here
there could be no reasonable supposition that the crystal had already undergone
any disintegration, yet the pits are irregularly dispersed on the surface. See
390 PROCEEDINGS OP THE AMERICAN ACADEMY.
Having traced tlie influence of chemical composition and of physical
conditions on the property of attackability, we may now proceed to in-
quire whether the single planes of an amphibole have different powers
of resisting solution in hydrofluoric acid. I shall once more use the infor-
mation gained in the process of etching ; the appearance of figures and
the loss of illumination using vertically incident light will be taken as
the criterion of attack. Actinolite, tremolite, and common hornblende
were thus examined.
A crystal of V. 42 was exposed first, 40 seconds, then one minute.
At 40 seconds (111) was visibly roughened but showed no figures, while
(110) remained practically intact; at the end of the second exposure both
planes were equally provided with pits. From the mere development of
figures, it would appear that (Til) and (110) were of about equal attack-
ability, but the roughening of surface gave much more reliable indications
of the fact. Combining with this, the observations on the other faces,
the series is correctly written iu the order of increasing resistance to the
( (100) 1
acid (Til), (011)-< (010) V. As a matter of fact, the maturing of pits
((.no))
on the various faces often occurred at intervals so far apart as to permit
of a pretty tolerable determination of the facial attackability by means
of their study alone. It is essential, of course, to distinguish etch-pits
from etch-hills, which may (as described by Becke) simultaneously ap-
pear on two different faces of the same crystal. A generalization was
made from the study of twelve crystals of aluminous hornblendes, as
follows: V. 26, V. 32 (2), V. 33, V. 35, V. 40 (2), V. 42 (3), V. 54, and
P. 55. Arranging the planes in order of increasing resistance to the acid,
we have, (Til), (Oil), (130), (100), (010), (110). But what has been
said about the phenomenon as affecting the corrosion on (110) applies
with equal force to the other planes. With V. 54, I obtained the series
(TOl), (010), (110), (Oil), in the same order as above, — an anomalous
order which I could see was correlated with the lustre and general state
of repair of each of the faces belonging to the crystal in question. By
reason of the peculiar geological conditions, and of the position of the
crystal in its druse, a prism face may suffer alteration while an end face
may escape that process and the latter can thus resist an attacking acid
longer, though, as we have seen, the terminal planes are regularly the
ones to yield first.
Klocke, Zeit fiir Kryst., 1877, Bd. II. p. 128. Cf. Retgers, Zeit. fur Phys. Chemie,
1895, Bd. XVI. p. 638.
DALY. ETCH-FIGURES ON AMPHIBOLES. 391
Similar results were found with crystals of Actinolite (V. IG), (V. 17),
and of tremolitic anaphibole (V. 8). Tlie series are, respectively, (TOl),
(Oil), (100), (110), and (Oil), (010), (110).*
These conclusions illustrate the now well established method of inter-
preting molecular structure by solubility in different crystallographic
directions.! The series of facial attackability correspond to the dimen-
sional variations in the pits on the various faces of an amphibole. They
explain, for example, the very general elongate form of the pits on planes
of the prism zone, an elongation in a vertical sense and coupled with a
greater resisting power in the prism zone itself. I may further note, in
passing, that here the different zones behave, as always, in a way to indi-
cate the holohedral character of all the amphiboles. The " Lusungsober-
fiiiche " (Becke) of amphiboles is monosymmetric and centrosymmetric.
Description op the Etch-Figures.
We may now proceed to the characterization of the etch-figures
themselves. It has been thought most convenient, and as tending towards
an easier survey of the facts, to group them with reference to crystallo-
graphic planes primarily and to consider in order the behavior of each
species on etching each of tliese planes with hydrofluoric acid. TVe may
hope thus to lessen the repetition of detail necessary in some degree ; at
the same time, the essential features of likeness or unlikeness of the dif-
ferent varieties will appear with most clearness. Following this more
or less bald statement of fact, which is abbreviated as far as consistent
with our immediate aims, there will be an attempt to correlate the facts
both in way of summary review and as related to certain others which
shall be especially introduced in the general discussion.
Etch-Figures on (HO), Actinolite Type.
Throughout the whole series of non-aluminous amphiboles (excepting
riebeckite) which I have yet studied (actinolite, tremolite, smaragdite, rich-
terite, astochite, etc.), cleavage pieces give figures that are practically
* One fresh crystal of Zillerthal actinolite etched with an alkali, caustic soda,
displayed greater resistance to attack on (010) tlian on (110). This may be another
illustration of tiie rule enunciated by Bauraliauer, that, in certain cases, the direc-
tions of rapid solution are reversed for acids and alkalies. Thus, he found it to be
true for Linneite when he compared its behavior in the process of etching using
nitric acid with its behavior in Becke's experiment of etching with caustic potash
(Resultate der Aetzmethode, p. 20). The same holds true for magnetite (Ibid., p. 21).
t Cf. Baumhauer, Resultate der Aetzmethode, p. v.
392
PROCEEDINGS OP THE AMERICAN ACADEMY.
indistinguishable from one another. They are the same as those on the
corresponding crystal-face and are, so far as observed, uniform in charac-
ters for all strengths of acid and times of exposure. This does not imply
that there are no differences in the figures, but these are so minute as, in
most cases, to defy measurement. Since I obtained the largest and best
pits on actinolite, I shall call theirs the " actiuolite " type of etch-pit.
In the diagrams (Fig. 1), (Plate I. Fig. 2), and Photograph 1, it will
be at once seen that in the claw-shaped figure, there are not many ele-
ments which can permit of precise measurement and of the comparison
of one figure with another, and with those of the aluminous amphiboles.
Yet the shape is so constant as to render possible an immediate recogni-
tion of these figures and of their orientation. The etch-faces are gen-
erally three in number (sometimes four, as in Plate I. Fig. 4), one plane,
the others more or less curved. The drawing and photograph convey a
far better idea of the arrangement of the etch-faces than could be given
in a verbal description, and it will only be necessary to note a few special
points. First, we have the corner at A (Fig. 1), which is always the
most clearly defined part of the figure. Often in the initial stage of
development this re-entrant angle appears before one can make out any-
DALY. ETCH-FIGURES ON AMPHIBOLES. 393
thing of the rest of the figure (Plate I. Fig. 1), a phenomenoa tliat
seems to have its correlative in the filling of the deeper parts, first of the
figures on a crystal of alum when it is placed in a concentrated solution
of its own substance (Ausheilen) .* The edge A £ is likewise the most
sharply outlined edge. The angle £ A i/^ which it makes with the trace
of the cleavage can be seen to alter as the figure gradually evolves in the
process of etching. Thus, on Zillerthal actinolite, this angle increases as
the figure grows more mature, from minus 2° to plus 10°, plus angles
being read on the right of the cleavage trace passing through A (compare
in Plate I. Fig. 1, Fig. 2, and Fig. 3). The limits of variation are the
same for the Syra actinolite. In the actinolite numbered V. 14, they are
0" and 12° ; in V. 16, 0° and 8°. During this jirocess of swivelling, the
edge AB remains straight ; this suggests an actual change in the indices
of the figure-face adjoining AJ3 as the pit is deepened, rather than a modi-
fication of a figure-face of constant indices by secondary solution. An
analogous feature characterizes the pits on aluminous amphibole, as will
be noted further on. Again, the corner at C, while not definite enough
to allow of exact location as a point, can be with low powers so fixed
with reference to A and the trace of the cleavage as to orient immediately
the whole figure. Measurements were made of the angle CAH, within
about 2° of accuracy in each case, as follows : —
Specimen. Angle CAH.
H. 5 12°.
H. 10 13°-16°.
V. 12 13°.
V. 14 10°.
V. 16 12°-14°.
V. 17 12i°
P. 19 130-15°.
V. 20 i . 13°.
V. 22 13°.
This angle is then seen to vary only slightly, and it has always the same
sign with respect to the trace of the cleavage. We may thus state the
orientation of the pit by means of its longer axis -4 C; — the point
of the " claw " is directed towards the positive hemij^yramids. Further-
more, the asymmetry of the figure expressed in the contrast of the
curved side ADC diad the side AB gives the orientation at a moment's
ijlance.
* Cf. Klocke, Zeit. fiir Kryst., 1877-78, Bd. II. p. 144.
394 PROCEEDINGS OF THE AMERICAN. ACADEMY.
Strongly attacked Zillerthal actinolite regularly exhibits normal etch-
hills on the prism, but, since they are not of immediate interest in
connection with our main j^urposes, I shall pass them over with this
mere mention.
MoNOCLiNic Alujiinous Amphiboles. Etch-Figures on (110).
As might have been expected, I have observed a considerably greater
variety in the aluminous amphiboles than in the relatively few non-
aluminous species that, so far, have yielded figures that can be discussed.
The pits of corrosion are, however, reducible to three types, respectively
characteristic of basaltic hornblende, glaucophanc, and arfvedsonite, and
it is by the names of these species that we shall know the types.
The Hornblende Type {\\Q).
To use an expressive German designation, the term " hornblende "
is a " Sachname," and signifies a large number of bisilicates which are as
yet not dignified by more specific names because of our ignorance of
their real nature. If the phenomena of etcliing are to have weight in
the discussion of the isomorphism or the orientation of cleavage pieces
of hornblendes, it is expedient to examine an extensive and representa-
tive suite of specimens from the different members of the group, and
determine what are the variations in the figures of corrosion along the
series. I have accordingly etched about one hundred crystals and
cleavage plates from the thirty localities mentioned in the list of mate-
rials used. The result was to show that all these species will give etch-
pits whose main characters are constant, but permit of the recognition
of at least four sub-tyi)es. From the localities of specimens that
illustrate tliem best, these may be called the Wolfsberg, the Kragero,
the Edenville, and the Philipstad sub-types. With the exception of
the last, crystal-face (110) and cleavage (110) give invariably the
same figure.
The Wolfsberg Sub-type of Corrosion Pits on (110).
The front positive prism-face on the lustrous basaltic hornblendes of
Norway, Bohemia, Vesuvius, etc., uniformly give an etch-pit, scalene-
triangular in shape, with the most acute angle (corner) pointing down-
wards and the next most obtuse angle (corner) pointing northeast (Fig. 2,
Plate I. Figs, b, 6, 7, and 8). In a mature figure (Fig. 2), the three figure-
DALY. — ETCH-FIGURES ON AMPHIBOLES. 395
faces and the edges are curved, AC more than AD and DC more than
AC. AB is a part oi AG which is practically straight, but of variable
length when coinjiared to the pit as a whole. A, B, 6', and D being the
principal angular points of the figure on the outside, i. e. marking the
more or less clearly defined corners formed by the meeting of pairs
of figure-faces and the prism-face etched, A', B', and D' are corre-
sponding corners at the bottom of the pit. Their position changes with
the maturing of the figure, but they are always analogous to A, B,
and D, because the bottom figure-face remains parallel to (110). To get
some idea of the relative dimensions of the j^its we shall define the
" length " of a figure as the distance, ^l H, from A to the foot of the
perpendicular running from C to the trace of the cleavage passing
through^. The "breadth" is, in like manner, the distance from this
trace to the line tangent to the curved side D C and parallel to the
cleavage AH. The maximum* length observed was about one tenth of
a millimeter and the breadth never far from one third of the length.
The angle D' A' B' gives an indication of the bluutness of the figure,
and is selected for measurement rather than D A B on account of its
greater sharpness, and hence the greater accuracy in measurement. It
is strikingly constant at from 72° to 73°. The angle C A His of course
variable with the elongation of the figure, but preserves a north by east
trend in all cases. Its value is usually from 13° to 15°.
The angle BAH also helps to orient the figure, and displays an
interesting relation to the deepening of the pit. Tiie shallow initial pits
of V. 46 characteristically had a large value for this angle (== 5°) ; as
they matured, it passed through intermediate values until parallelism
oi AB with the trace of the cleavage was reached; there the swivelling
o( AB seems to have been arrested in even much older figures, and thus
the angle BAIT never was observed to change its sign. A cause for
this swinging of ^ i? is problematical, but it looks as if Becke's* prin-
ciple of differential solution might be used in explanation. In the corner
A, there is likely to be more rapid saturation with the jjroducts of solu-
tion than along the medial part of the figure-face A C, since these are not
so readily whisked away in convection currents acting on the constricted
parts of the pit as in those affecting the more open region about the
point B, for example. Thus, the middle part of AC will suffer in a
unit of time the attack of a greater number of chemically active ions
than will the corners, and yield faster to solution accordingly. Such
* Min. und petrog. Mittheil., 1885, Bd. VII. p. 240.
396 PROCEEDINGS OF THE AMERICAN ACADEMY.
a hypothecated process of "secondary solution" was incidentally re-
ferred to in connection with the actinolite figures ; its effects can
often be seen where straight-edged outlines of a figure are replaced
by curved edges as the figure is undergoing destruction by prolonged
solution.
Plate I. Fig. 7 furnishes a noteworthy variation on the normal and
simple process of pit development. The figure is compound, and consists
of three pits, formed respectively at the bottom of the next oldest pit.
Each must represent an abrupt stage in the solution of this part of the
surface. The side AD remains sensibly parallel to itself in all three
steps, but the angle BAH grows larger from the first to the third num-
bered downwards; at the same time, the edges of the successive figure-
faces against (110) are seen to curve more in the first than in the second
and in the second than in the third. These facts accord with those
observed in the case of the pits that grew continuously, not intermit-
tently, to maturity from the initial and less advanced stages of previous
exposures to the acid. The stepjied figure seems, then, to show that
the formation of pits may (though not always) take place spasmodically,
so to speak (•' sprungweise," in the German phrasing), the attack the
affair of a moment and preceded and followed by longer periods of
almost perfect quiescence as regards other than "secondary" solution.
Klocke believed, similarly, that the formation of figures on the alums
is an instantaneous thing.* The stepped form is jiresumably not due
to a zonal structure in the hornblende, because such a hypothesis would
imply very considerable variations in the attackability in passing from
the exterior to the inner zones of the crystal, — variations improbable
* Zeit. fiir Kryst., 1877-78, Bd. II. p. 131. The rapidity of the reverse process,
that of healing over an etched surface, was early commented upon by Sir David
Brewster in connection with his studies on the instructive group of crjstalline
substances, the alums, especially' with reference to his now familiar " light-figures."
On immersing an etched crystal of an alum in a concentrated solution of its own
substance, he observed that, "in an instant," the pits of corrosion began to fill.
He says, "The singular fact in tliis experiment is the inconceivable rapidity with
which the particles in the solution fly into their proper places upon the disinte-
grated surface and become a permanent portion of the solid crystal." Phil. Mag.,
1853, Vol. V. p. 27. The intermittent character of the process of etching was
noticed by Rinne (Neues Jahrbuch, 1885, Bd. II. p. 15), who etched milarite with
dilute HF. At first, the base became covered with regular hexagonal pits placed
symmetrically on (0001). On further attack, these are suddenly modified by the
appearance in each case of a second hexagon in the bottom of the first, but now
turned through an angle of 30°.
DALY. ■
ETCH-FIGDRES ON AMPHIBOLES.
397
and not yet shown to exist. The interrupted phases of solution are
more reasonably connected with the lamellation of the cleavage parallel
to the plane attacked. Actual microscopic or submicroscopic separation
of layers parallel to this cleavage would present to the solvent action
of the hydrofluoric acid a series of thin plates each of which, strongly
resistant in a direction at right angles to the broad flat surface, would
readily yield, in directions in that surface, i. e. along the grain of the
mineral.
The above mentioned data regarding the highly important Wolfsberg
sub-type refer to conditions of etching described as standard for the
present investigation. Remembering Bomer's conclusion that the tem-
perature of the solvent has, in tlie case of HF and quartz, an influence
on the figures, I have recorded the facts from a number of experiments
intended to test the principle in its application to the group of the com-
mon and basaltic hornblendes. The experiments were made on V. 4G,
as follows : —
(1)
HF boiling on Bunsen
burner.
45
seconds.
(2)
"
"
90
"
(3)
Water bath at 100° C.
2 minutes.
Standard conditions.
(4)
U "
2^
minutes
(5)
(( (<
3
(G)
(( «
(7)
" "
6
(8)
« «
8
(9)
10)
Acid slightly warmed.
10
20
(11) Room temperature
16 hours
Result.
Many poorly defined pits.
Larger poorly defined pits.
Small sliarply defined pits.
Pits numerous, in every
case, some of great sharp-
ness and large enough to
measure.
Pits indistinguishable from
the last, but associated
with etch-hills.
Few but good figures of the
foregoing type.
Throughout the whole series there is such a close correspondence in the
forms and measurements of the figures that we must posit for the latter
an independence of the temperature of the hydrofluoric acid, so far as
any sensible differences are concerned.
Neither temperature nor concentration of acid, nor, indeed, any cause
known to me will explain a notable variation in the shape of the pits on
(110) of the amphiboles now under consideration. It consists in the
appearance of an adventitious fourth figure-face on the southeast side,
in addition to the three usual ones (see Plate I. Fig. 8). There is no
398
PROCEEDINGS OP THE AMERICAN ACADEMY.
discoverable rule governing its development, and I even doubt that it
is always the same face. We might suppose that B G is composed of a
large number of minute planes, and that, for some at present unknown
reason, there is a selection of now one, now another, of these multitu-
dinous facets, which grows rapidly and becomes prominent as a fourth
figure-face of the pit.
Pargasite, carinthine, gamsigradite, syntagmatite, and barkevikite (see
Plate I. Fig. 20), each from its classic locality, give figures that are equiv-
alent with those just described for basaltic hornblende. The differences,
if any, from the Wolfsberg sub-type are so slight as to prevent any
determination of these varieties from the shapes of their respective
corrosion pits. The other three sub-types merit a few remarks, inasmuch
as they undoubtedly owe their distinctive characters primarily to chemi-
cal composition.
Figure 3.
Tlie Kragero Sub-type, (110).
The pits on (110) of V. 31 were sotne-
what sharper than those usually obtainable
on basaltic hornblejide (Figure 3, Plate I.
Figs. 9 and 10, and Photograph 2). There
are again normally three, abnormally four,
figure-faces all oriented in the sense of the
corresponding faces of the pits of the Wolfs-
berg sub-type. The straightness of the edge
A D \% here characteristic. The angle D A H
is extraordinarily variable, having the value
of about 38° in the initial figures, aud in-
termediate values up to 80° in the large
matured pits. There is also a swivelling of
the edge A! , jy, but its amount is difiicult to
determine. Accompanying these changes in
the figure-faces, there is a tendency for the
figure to broaden out as it deepens; so that,
while in the initial figures the ratio of length
to breadth is about 3.5 : 1 (0.05 mm. : 0.014
mm.), that ratio is 2 : 1 (0.072 mm. : 0.03(>
mm.) in the matured pit. The angle C A H
ranges through limits about equal to the
Wolfsberfj readings.
DALY. — ETCH-FIGURES ON AMPHIBOLES. 399
The Kragero sub-type differs, then, from the Wolfsberg sub-type in
the hick of curvature in the figure-face A D D A\ in the acuteness of
the northeast angle at A, in the greater variation of shape as the figure
grows older, and in the stoutness of the figure in plan. The orientation
is the same in both, as also the occurrence of the adventitious fourth
figure-face adjoining G E in the figure (see Plate I. Fig. 10). V. 42 is
the only other amphibole that showed closely similar etching phenomena
(Plate I. Figs. 11 and 12, and Photograp»h 3), but their clearness has led
to the conclusion that we have here a new category of figures whose ex-
planation should be looked for in the internal structure of the varieties in
question. The lack of analyses forbids the extension of this hypothesis.
The Edenville Sub-type, (110).
Edenite gave abundant figures, apparently identical with those on the
Wolfsberg hornblende, but the habit of a cleavage piece of the former
was peculiar in exibiting a general predominance of the four-sided figures
already noted in connection with the previous two sub-types (Plate I.
Figs. 13 and 14). The triangular figures do occur, but their number is
quite subordinate. Even without their aid, it is very easy to orient cleav-
age plates of the mineral by means of fairly well developed quadrilateral
figures ; in them, the upper end is always recognizable.
The PkiUpstad Sub-type, (110).
The last of the hornblende sub-types which we have to notice is of
special interest, since it led at once to further investigation and the dis-
covery of a new and interesting variety of amphibole. Figures were
produced on four different crystals, with exposures of 2 min., 2 min.,
1| min., and 2 min., respectively, and on another crystal immersed in
hot HF (near its boiling point) for one minute. In all five, the result
was the same, two clearly defined figures, whose distribution was at first
a mystery ; on further study it was found that one type was confined to
the cleavage faces and the other to the crystal faces, or to cleavage surfaces
lying not more than 1 mm. beneath the latter. We begin with a de-
scription of the first mentioned sort of figures.
They vary in number of sides, in tint (proportion and distribution of
light and shade), and considerably in shape (Photographs 6 and 7).
Usually, they are six- or seven-sided and bounded by nearly or quite
straight edges (Photograph 6), and either uniformly dark or more or less
brilliantly illuminated on certain figure-faces. I could get no satisfactory
400 PROCEEDINGS OF THE AMERICAN ACADEMY,
idea as to the relative steepness of the figure-faces on account of the great
variability of the pits, on the one hand, and, on the other, the deep tint
characteristic of the pits. The largest mature pits are perfectly black,
blunt on the upper end because of the truncation by a long transverse
edge and bulging below but tapering rapidly to the lower end. The
angle made by the transverse edge with the cleavage trace was measured
at 83°, a value which lies within the limiting values of the angle D A H,
in Figure 2 ; the edge is thus oriented the same as the edge D A oi the
Wolfsberg sub-type, and the two are regarded as equivalent.
Now, these pits appear on all four cleavage faces and are also asso-
ciated on these according to the law of twinning after (100) ; but with
them were found a second set of j^its arranged enantiomorphously to
the first. Evidently, no system of twinning can explain their orienta-
tion, and it is all the more surprising from the fact that there is no
perceptible difference in the young forms of the two sets of figures.
Moreover, whole areas of a cleavage surface are covered with pits belong-
ing to a third category, in which the straight edges are the exception and
a curved outline the rule (Photograph 7). By an optical illusion, these
pits have the appearance of projecting from the general surface after the
manner of etch-hills ; in so doing, the form of the outline and a pro-
nounced shelly structure, exactly imitating the lines of growth in a mol-
luscan shell, make each pit extremely organic in look. Very often, the
shells are greatly elongated in a direction transverse to the cleavage
trace (Photograph 7). Only one probable explanation of these two
classes of abnormal pits has suggested itself, namely, that the crystals
are in a condition of internal tension, which interferes with the workings
of the usual molecular cohesions. If this hypothesis be correct, we
should expect all transitions between the normal type and the other two
abnormal types of figures. Such is the fact, typified in the observed
occurrence of pits intermediate to those of the first and second kinds,
where the transverse edge is seen to have all directions within the limits
set by the enantiomorphous pits (see Photograph 6). Moreover, we
might expect on this hypothesis that there should be distorted pits on the
cleavages etched with reagents other than HF. Using caustic soda, I
obtained normal pits on (110) illustrated in Photograph 9 ; occasionally,
though more rarely than in the case of crystals etched with the acid, the
beautifully marked shells were replaced by others elongated transversely.
A possible cause of this distortion by differential tensions is not far to
seek. Several carefully cut sections of the mineral showed that it is
strikingly zoned, each zone possessing its own tint of color which is
DALY. — ETCH-FIGURES ON AMPHTBOLES. 401
doubtless correlated with a special chemical constitution. It is not difficult
to believe that the molecular substance of the crystals may be put into a
condition of strain by varying rates of expansion and contraction in the
ditferent zones due to temperature changes. Again, there is also possible
a massive stress set up in the crystal caused by its contortion in the rock
from which it was derived. In this case, we are dealing with a druse-
mineral ; nevertheless, I have observed that at least two of the individu-
als are greatly twisted so that the plane of the cleavage is replaced by a
curved surface with from 8° to 10° of curvature. Attempts were made
to etch this jjarticular crystal on (110), without marked success, although
the mineral was strongly attacked ; yet the few pits actually obtained
seemed to be of the distorted types. May this massive distortion not be
the result of differential stress in the zones ? *
The Philipstad (cleavage) sub-type of etch-pit is thus analogous to the
other sub-types peculiar to hornblendes, but it is similar to no one of
them. It bears the same relation to the second principal kind of pit
observable on this mineral ; namely, that on the zone occurring on all the
individuals so far examined, just underlying the crystal face (Photographs
4 and 5). Most commonly, they are of the shape illustrated in Photo-
graph 4 ; that is, while the general habit of the figure is very like that
of the Wolfsberg sub-type, tliere is here a more pronounced blunting of
the lower end of the pit by an edge nearly as long as the transverse edge
of the upjier end. The rarer pits with a sharpened lower end are
portrayed in Photograph 5. The photographs show with sufficient
clearness the contrast existing between these pits and those character-
istic of the inner zones exposed on cleavages. There is never any inter-
mixture of the two kinds ; the former is confined to a light colored
exterior zone, which appears to be less strongly charged with iron than
the inner zones, which are always darker in color. Examination proved
that these pits on the outside zone could not be explained as natural
etchings ; they are manifestly the result of an interaction between the
acid and the mineral, aud probably differing from the figures on the
other zones on account of the fact that, while the latter agree in chemical
composition fairly closely among themselves, there is a strong chemical
peculiarity adhering to the outer zone. That the phenomenon is not
confined to the reagent used is evident from the comparative study of
* Sudden twisting of crystals does not seem to affect the form of the etch-
figures. Thus, I etched a cleavage plate of gypsum which had been bent through
an angle of at least 15°. The pits produced had perfectly normal characters, as
those recently described by Viola.
VOL. XXXIV. — 26
402 PROCEEDINGS OF THE AMERICAN ACADEMY.
Photographs 8 and 9. No. 8 indicates the result of etching (110) (crys-
tal face) with caustic soda. The pits are like those on the inner zones in
possessing a shelly structure, but have a different outline (compare
Photograph 9).
Before leaving this peculiar hornblende, it should be stated that it
stands also in a unique position with respect to the etching properties of
the clinopinacoid. I have called the angle A D H \n Figure 8 posi-
tive ; it varies in value with the different varieties of aluminous amphi-
boles from 1° to 10". But, in the analogous pit produced on (010) of
the Philipstad hornblende, this angle is always negative and averages 2^°
in value. The description of the optical and other characters of the
mineral is deferred to another occasion. (See following article.)
Etch-Hills on Hornblende^ (110).
A digression from the main subject of types of pits of corrosion may
be permitted in the form of a short discussion of another result of attack
with hydrofluoric acid, namely, etch-hills. The usual effect of dissolving
a cleavage piece is, in time, the disappearance of the pits formed at the
beginning, and their replacement by these residual bosses. When they
are numerous, the mineral has a characteristic raammillated look. Besides
the normal bosses left on the removal of the ridges between successive
pits, however, there often appear on aluminous amphiboles, when etched
rapidly (fresh acid and high temperature), a variable number of remark-
able etch-hills which, from their form, can have but little to do with
those just mentioned.
As examples, V. 46 furnishes some very striking specimens. I pro-
duced these peculiar etch-hills at 3 minutes, at '6\ minutes, at 7 minutes,
and also after 12 minutes' suspension in IIF gas evolved from a hot
(fuming) aqueous solution of that gas. In all four cases, the bosses were
on the whole similar in look, and, on account of their perfect development,
I shall describe the etch-hills on the cleavage piece last mentioned, as typ-
ical of all (Photograph 10). They are bizarre in form and arrangement;
in plan, triangles, irregular quadrilaterals of many shapes, trapezoids,
pentagons, etc., sometimes in groups of two, three, or a half-dozen, similar
to one another in outline, and even showing parallelism between cor-
responding sides. They are commonly bounded by straight lines that
have no definite relation to the hornblende crystal, and are thus in strik-
ing contrast to the pits which are oriented in the regular way on the
same cleavage face. Occasionally, small groups of the figures have the
same form and orientation ; thus, three scalene triangles were observed
DALY, ETCH-FIGURES ON AMPHIBOLES. 403
in one place, with sides mutually parallel. At the same time, other
aggregations could be found in which the individuals were bounded by the
same number of sides, and with angles sensibly equal, yet the corre-
sponding sides were not parallel and the orientation of the figures was
necessarily unlike.
For some time I was without a clue to the meaning of these myriad
extraordinary figures, but another of the many valuable suggestions in
Becke's writings afforded some light on the problem.* When galenite is
etched with hydrochloric acid, the resulting chloride of lead often crystal-
lizes out in areas of local supersaturation of the liquid, particularly in
regions where the jiits are most numerous. The individual crystals of
the chloride may be locally oriented in the same way, and will doubtless,
in certain cases, favor a skeleton growth. They serve as a kind of pro-
tection to the surface on which they lie ; the acid will thus dissolve the
intervening parts of the general crystal-surface not so protected, and the
substance of the galenite underneath the chloride crystals is left projecting
as residual hills on corrosion. The common orientation of these crystals
and their skeleton-crystallization (touching the galenite surface only
where the regular growth of skeleton crystals would permit) could
explain the accordant attitude of certain similarly arranged groups of
the bosses.
An analogous explanation is believed to apply to the curious etch-hills
on hornblende above noted. The chemical reaction is diflerent, the me-
chanical cause of differential attack is the same. Instead of hydrochloric
acid we have here hydrofluoric acid, and in place of a single resulting
compound, the chloride, there are probably several salts of hydrofluoric
and fluosilicic acids that are produced during chemical solution of the
bisilicate, and in the form of crystals or of skeletal aggregates might
serve as the protective caps in the lithographic process. What particu-
lar fluorides and silicofluorides would be most likely to play such a role,
it is perhaps not impossible to say. From their relative insolubility in
warm water, the prisms of fluosilicate of magnesium, the rhombohedrons
of the fluosilicate of iron, and the spindle-like crystals of the fluosilicate
of calcium, seem to be the most favorable to such action. The more
soluble octahedrons of the fluosilicate of sodium and hexagonal prisms of
the fluosilicate of potassium might also result in an atmosphere of hydro-
fluoric acid gas diluted with only a small proportion of vapor of water,
* Aetzversuche am Bleiglanz, Min. und petrog. Mittheil, 1884-85, Bd. VI.
p. 240.
404 PROCEEDINGS OP THE AMERICAN ACADEMY.
as in the case before us. The tendency of fluosilicate of magnesium to
crystallize out in the form of skeletal groupings is noted and figured by
Boricky in his classic work on microchemical methods.* He also de-
scribes the actual determination of these various fluosilicates on (010) of
an amphibole,t etched with hydrofluosilicic acid.
The Glaucophane Type, (110).
Glaucophane furnishes a new tyi^e of pit on (110). It is more elon-
gated than the Hornblende type, is characterized by a more pronounced
straightness of edges, and is unique by reason of the parallelism between
its longest edge (corresponding to A C \n Figure 2), and the trace of the
cleavage. (See Plate I. Figs. 15 and 16.) It is likewise triangular in
outline, possesses three figure-faces on the sides and a migrating bottom
face. Gastaldite from the Champ de Praz (P. 59) afforded pits in no
respect to be distinguished from those on the He de Groix glaucophane.
On the other hand, crossite gave figures decidedly differently and more
closely allied to the Hornblende type (Plate I. Fig. 17).
The Rieheckite Type, (110).
Figures of corrosion were obtained on riebeckite only with much dif-
ficulty, apparently due to its extreme attackability in concentrated acid.
They were alwaj's excessively small, often with imperfect development;
the upper end of the pit was the first to become clearly evident, in the
process of maturing. The figure has many points in common with the
sub-type noted above on Edenite ; it is usually quadrilateral, though
sometimes three-sided and analogous to the Wolfsberg sub-type. Put it
differs from both in its being much darker than they in vertically incident
light: — the figure-faces are steeper than in the same (110) pit on com-
mon hornblendes (Plate I. Figs. 18 and 19).
The Arfvedsonite Type, (110).
Quite an exceptional category of etch-figures is represented in the pits
generated on the prism-face of arfvedsonite by the use of hydrofluoric
acid. (Plate I. Figs. 21 and 22, and Photograph 11.) Their peculiar-
ities are so salient as to enforce the belief that, in the matter of cohesion
on this particular face, arfvedsonite is at least as far removed from the
* Archiv d. naturw. Landesforschung von Bohmen, III. 5, Prague, 1877, Plate I.
Fig. 12. Translated by Winchell, 19th Ann. Rep. Minnesota Geological Survey,
t Op. cit., Plate II. Fig. 7.
DALY. ETCH-FIGURES ON AMPHIBOLES.
405
other amphiboles as it has been proved to be optically. The pit is here
a spindle-shaped well defined figure, generally about six times, rarely only
four times, as long as it is broad. The sjtindle is usually ideally perfect,
and theu the axis can be seen to make an angle with the cleavage cracks
of two degrees east of north on (110), two degrees west of north on
(ITO). Many pits show that they are bounded by two curved figure-
faces of unequal steepness, and hence of unequal illumination in the
microscope. The narrower, darker one lies to the left on (110) ; it is
separated from the other by a narrow light streak that corresponds to the
keel of the unsymmetrical canoe. The relations are enantiomorphous on
(ITO). Occasionally, the spindle is blunted with what appears to be an
imperfectly formed third figure-face that would represent the upper figure-
face of the Hornblende type (Plate I. Fig. 22). The photograph does
not give an idea of the exceeding sharpness of these figures, at least
as compared with most other amphiboles ; there can be no doubt that
the type is a distinct one and stands alone.
The figures show the mineral to be holohedral and centrosymmetric
and a cleavage plate can be easily oriented in the absence of crystallo-
graphic data by observing the position of the adventitious third face, or
the direction of the spindle-axes with respect to the obtuse angle of cleav-
age (110 : 110), and one, say the darker, of the two longitudinal figure-
faces of a pit.
Both the symmetry and orientation of arfvedsonite are, however, better
made out by the use of figures resulting from exposure to molten caustic
soda. A cleavage piece was found after 25
seconds' immersion to be covered with three-
sided pits, as depicted in Figure 4. These
show once again the radical difference in
behavior between arfvedsonite and common
hornblende. (Cf. Plate I. Figs. 34 and So.)
It is, furthermore, an interesting case, in that
the directions of rapid solution are here
transverse to those of rapid solution by the
acid, and, secondly, the formation of etch-
figures is once again seen to be independent
of cleavage. Figure 4.
Non-Aluminous Amphiboles, Etch-Figures on (010).
The actinolite of Zillerthal was found to be typical of the whole
group of amphiboles not containing a sesquioxide as regards the facts
406
PROCEEDINGS OF THE AMERICAN ACADEMY.
of etching on the clinopinacoid. Good figures are obtained with ease.
(Plate I. Figs. 23a, 236, 23c.) They are remarkable in belonging to
two classes, analogous to those described by Pelikan for pyroxene with
the same reagent.* (Photo. 12, cf. Photo. 13.) The pit of the one
category is a quadrilateral in outline, with four pyramidal figure-faces
and a bottom-face that grows smaller and then disappears as the figure
matures and deepens. The other kind of pit is also four-sided in habit,
but may possess another pair of figure-faces in addition to the five cor-
responding to those of the first class. Figures 5 a and 5 b represent
diagrammatically the two kinds matured under normal conditions (water
bath, concentrated acid, etc.). The drawings are lettered in order to in-
dicate the elements chosen to fix the shape and orientation of the pits.
Figure 6.
Figure 5 6,
C Figure 7.
* Ueber den Scliichtenbau der Krystalle. Min. und petrog. Mittheil., 1896,
Bd. XVI. p. 16.
DALY. — ETCH-FIGURES ON AMPHIBOLES. 407
The smaller pits (Figure 5 a) are invariably four-sided and elongated
parallel to the side the more oblique to the trace of the cleavage. So
extremely minute as to be at times easily overlooked, they are, moreover,
inconspicuous on account of their shallowness and the consequent lack
of contrast with the rest of the crystal surface ; they may hence be called
the " light " figures to distinguish them from their darker fellows of the
second class to be described. The angle A^ D' H was determined at from
3° to 5°, the auirle B' A IX averaged 108°. A' B' and A' D' were
measured as under : —
-
A' B'
A' D'
Ratio,
A' B' :
A' D<
0.012 mm.
0.012 mm.
3
: 3.
0.011 "
0.009 "
11
: 9.
0.014 "
0.012 "
7
: 6.
0.014 ''
0.010 "
7
: 5.
On the average, A' B' : A' U is about equal to 8 : 5.
Quite different are the pits of the other kind (Figure 5 h) ; they are
larger, deeper, and often more numerous. The elongation is, in this
case, in the sense of the vertical axis of the crystal. The four-sided
figures are, as in the first class, parallelograms in plan, the acute angle
here lying, however, in the upper left hand corner. It is this corner and
the opposite one that are truncated by the accessory pair of figure-faces
already mentioned. The angle A D II is 8° instead of o°-5° in the
light figures. The angle £ A D varies from 82° to 88° ; the curvature
of the sides prevents a close determination. The limits of variation in
the ratio A B : AD were observed as 1 : 1 (0.04 mm. : 0.04 mm.) and
about 2:3 (0.032 mm.: 0.048 mm.).
According to the character of the figure-faces (always pyramids) these
larger pits can be subdivided into two sub-groups in one of which the faces
A B and C D are equally illuminated but darker, i. e. more steeply in-
clined to (010) than AD and B C, also of equal brightness, while in
the other group the reverse relation holds. (The figure-faces are here
indicated by their corresponding edges of intersection with (010).) Both
kinds have figure-faces belonging to the same crystallographic zones, and
they are probably allied as are the figure-faces described by Baumhauer
on apatite.* He noted — that which is evident from his photographs —
that, as the apatite matures (deepens) its faces become steeper and steeper
on (0001), (" Verschleppung " of Becke). In one case, I discovered a
large pit having A D steeper than A B, with the bottom occupied by a
* Resultate der Aetzmethode, p. 48.
408 PROCEEDINGS OF THE AMERICAN ACADEMY.
second typical pit manifestly of younger generation, but with its A B
steeper than A D. Brauns found on the etched surface of sylvite neigh-
boring pits bounded by figure-faces of different steepness, but he does
not seem to have connected the phenomenon with the difference of
maturity of the two pits. (See Nuues Jahrbuch fur Miu., etc., 1889,
Bd. I. p. 113.)
Either of these types of dark figures may show a keel at the intersec-
tion of a pair of figure-faces (Plate I. Fig. 23 b).
It was in connection with the study of the pits on (010) of actinolite
that I became convinced of the necessity of fixing a standard degree of
dilution of the hydrofluoric acid before beginning a series of comparative
experiments in etching the araphiboles. The problem could be here
more successfully attacked than in the examination of the figures on
(110) because of tlie greater likeliiiood of being able to observe differ-
ences in the shape or arrangement of the systematically straight-edged
pits on (010). Six cleavage pieces of Zillerthal actinolite were immersed
in II F, either in the form of pure gas or in different states of dilution
with water. The procedure and the results are synopsized in the accom-
panying table, which shows the effects on the angles A D H 3ind BAD
of the dark figures under the different conditions (Figure 6).*
Specimen. Solvent. Exposure,
No. Minutes.
(1) Pure HF (gas) 70
(2) Com. cone, water solution of HF 2\
(3) 75% HF, 25% water 4
(4) 50% HF, 50% water 15
(5) 25% HF, 75% water 25
(G) 10% HF, 90% water 20 -10
No. 1 was etched by hanging it above the surface of some concentrated
water solution of the acid, that was very gently heated far below its
boiling point, and thus only a small percentage of water vapor could be
present during the reaction. The other examples were etched in the
ordinary way on a water bath. The percentages of dilution are by
volume.
The effect of dilution with water is, then, to produce a rotation of
each dark figure about a line perpendicular to the crystal plane. The
direction of the rotation is opposite to that of the hands of a watch, its
amount (within the limits of these experiments) about 30°. Along
* For Figures 6 and 7 see page 406.
Angle ADII.
Angle BAD.
o
13-14
76
12
75
2
64
-2
Acute but indet.
-11
" "
DALY. — ETCH-FIGURES ON AMPHIBOLES. 409
with the rotation, there is a simultaneous distortion of the ontline so as
to make the angle B A D more and more acute. The latter could not be
measured closely in the cases of dilution of 50 per cent or more on ac-
count of the poor development of the pits. The longer edge preserved
its distinctness much better than did the shorter edge ; hence the angle
A D If was measurable throughout.
The light figures remained quiescent, neither changing in shape nor
orientation, so far as those rather unsatisfactory figures would permit of
measurement. In the same manner, I could discover no variations in
the pits on the prism-face (HO).
Bomer has described other examples of the dependence of the results
of etching with hydrofluoric acid on its state of dilution in water. He
found the form and attitude of the pits on the base of quartz to alter so
much with dilution that, whereas concentrated acid gives figure-faces
belonging to the right trigonal pyramids, with very dilute acid they are
negative rhombohedrons. Intermediate forms characterize decrees of*
concentration between these two extremes.* Baumhauer discussed a
similar anomalous behavior in the pits on apatite, using hydrochloric acid
as the solvent.f There is throughout a close analogy between the pits
on (0001) of apatite and those on (010) of actinolite. In both, we have
dark fijjures and two cateiiories of liiiht figures, and, finally, the same
tendency to rotate with decreasing concentration of the respective acids
in aqueous solution. |
It is well known that sulphuric acid, when added in small quantities,
will in certain cases intensify the solvent power of hydrofluoric acid. I
made one or two trials of a basaltic hornblende to see whether in this way
the figures of corrosion might be improved. They resulted in partial
failure, for, although the pits were a little larger than usual, they lost
considerably in definiteness of outline. It then occurred to me to try
mixtures of the two acids in various strengths on a more resistant sub-
stance, actinolite, primarily to determine what, if any, would be the
influence of the sulphuric acid on the etching. The problem was anal-
ogous to that just discussed for dilution with water, the procedure was
similar, the results just as striking. The table is so like the last as to
need no special explanation. (Figure 7.)
* Neues Jahrbucli fiir Min., etc., 1891, Beil. Bd. VII. p. 535.
t Eesultate der Aetzmetliode, p. 48.
t Becke states that concentration affects the position of the figure-faces in the
etch-zone. Min. u. petrog. Mittheil., 1883, Bd. V. p. 487.
410
PROCEEDINGS OP THE AMERICAN ACADEMY.
Specimen. Solvent.
No.
(1) Pure HF (gas)
(2) Commercial cone, solution
(3) 95% HF, 5% H2SO4
(4) 90% HF, 10% H2SO4
(5) 75% HF, 25% H2SO4
(6) 50% HF, 50% H..SO4
(7) 20% HF, 80% H2SO4
Exposure.
Angle JBH.
Angle BAD.
Minutes.
^
70
13-14
76
2i
12
75
5
11-12
75±
H
19
74-78
8
19±
76-78
31 and 8
23
79±
26± Indeterminable.
The mixtures of acids were proportioned by volume.
The table clearly expresses a rotation of the figures with increasing
percentages of PI2SO4 in a right-handed direction, that opposed to the
direction of movement in the water-dilution series. At the same time,
the angle BAD grows larger, and thus the parallelogram tends more
and more towards a rectangular outline. In both respects, the influence
of sulphuric acid is in the shaping and the arranging the pits on the
clinopinacoid, when etched by means of a mixture of that acid and
hydrofluoric acid, the reverse of that of pure water mixed in the same
manner with hydrofluoric acid. I am not prepared to offer any expla-
nation of this interesting phenomenon.*
It may be noted that the surface of cleavage (110) exhibits great
changes in the etch-figures as the proportion of the sulphuric acid in-
creases. With only 5% of the latter (Plate I. Fig. 24), there is a sen-
sible variation iu look from the normal type, and in the 50% solution a
strong suggestion of the Wolfsberg figure of common hornblende
(Plate I. Figs. 25 a and 25 b).
Tremolite and richterite afforded, on the clinopinacoid, etching phe-
nomena identical with those described for actinolite when exposed to
concentrated HF; the last mentioned experiments of mixture were not
essayed in connection with them.
* Baumhauer not only showed a rotation of the figures on the basal plane of
apatite with increasing dilution of tlie solvent, hydrochloric acid, but he also
established a rotation in the same direction when nitric acid was substituted for
the hydrochloric, and a rotation in the opposite direction if sulphuric acid be simi-
larly employed. Ref. in Bomer's article, Neues Jahrbuch fiir Min., 1801, Beil.
Bd. VII. p. 538. The accumulation of such facts as these makes it difficult to
follow Becke in the hypothesis that his " Hauptaetzflachen " have simple indices,
because of greater molecular density in planes having such indices. The molecu-
lar density evidently does not change with the solvent. (Min. u. petr. Mitth.,
1887, Bd. VII. p. 200).
DALY. — ETCH-FIGURES ON AMPHIBOLES. 411
Etch-Figures on the Clinopinacoid of Aluminous
Amphiboles.
The cohesional properties of the plane (010) on any one of the akimi-
nous amphiboles are simpler than those of the same face on actinolite, in
that we find the development of only one kind of figure on the former
though etched in exactly the same fashion. It corresponds to the light
type on the actinolites in several respects, being a parallelogram with
the obtuse angle situated in the upper left hand corner and the non-
appearance of the adventitious third pair of figure-faces. The equiva-
lence is not complete because of the slight elongation in a vertical sense,
and because of a greater obliquity of the figure-faces to the plane (010).
(See Fig. 8 on page 406, and Plate I. Fig. 26, and Photograph 14.)
We have seen that the pits on (110) of the hornblendes vary some-
what with the chemical composition ; the same is true of the (010)
pits. A number of specimens from different localities undoubtedly
with considerable, though unknown differences in the proportions of
the constituent oxides, were etched, and two of the elements determined
in each case, as recorded in the table : —
pecimen.
Exposure.
Angle A DH.
Angle BAD.
Minutes.
o
p
V. 26
H
5
95
V. 32
If
4
113
V. 33
1
5
98±
V. 34
li
7
95
V. 40
1
7
97
V. 42
1
10
100-105
V. 54
2
1
Not determinable.
P. 55
3
4
114
There can be no question, in view of the facts of the foregoing table,
that there is a lack of uniformity in the crystallographic zones into
which the figure-faces on (010) of hornblendes fall. The differences in
outline and orientation are so salient as to impress the eye at once on
seeing them in the microscope. Now those hornblendes (V. 26, V. 33,
V. 34, V. 40, V. 54) which had the edge A B not far from being sen-
sibly perpendicular to the trace of the cleavage were also characterized
by (110) pits of the Wolfsberg sub-type, and those so far studied (V. 32,
V. 42, P. 55) that showed important variations from that orientation
carried pits more allied to the Kragero sub-type, if not identical with it.
More detailed investigation in the future with carefully analyzed material
412 PROCEEDINGS OF THE AMERICAN ACADEMY.
may reduce the iagures and their individual characters to law, and make
the etch-pits of determinative value.
Owing to lack of material, etch-figures were not obtainable on
the clinopinacoid of barkevikite. crossite, glaucophane, riebeckite, or
arfvedsonite.
Amphibole Etch-Figures on Faces other than (110)
AND (010).
From the preceding sketch of the pits of corrosion on the prism and
clinopinacoid of the aluminous and non-aluminous groups of amphiboles,
respectively, it is evident that the general figure-types are modified by
variations in the chemical nature of the species. But the prime impor-
tant modifications are conditioned by the presence in the riiolecule of
alumina, or at least of a sesquioxide. That is, in the one class, we have
to do with amphiboles whose molecular constitution is similar, excepting
perhaps arfvedsonite ; whether it be lime, iron, magnesia, or soda, — any
or all of them, — that, together with alumina, compose the complex
silicate, the results of etching are always similar.* On the other hand,
so soon as the alumina (sesquioxide) molecule disappears, or is present
in only very small amouqts, there is a radical change in the etch-pits on
both faces. Tliis leads to the expectation that other planes will show
corresponding change. The few specimens which I have been able to
secure confirm that conclusion, and a brief account of the observations
thereupon may prove of interest.
Before proceeding directly to them, however, I shall state the negative
results characterizing the examination of three other faces (130), (Oil),
and (111) ; their indefinite etching phenomena did not allow of compar-
isons by measurement. V. 42 at one minute's exposure exhibited many
pits on (130). They were warped triangles, with the upper acute corner
pitching into the crystal mucli after the manner of the analogous hair-like
projections described by Tschermak on siderite. (See Baumhauer,
Eesultate der Aetzmethode, Microgram 20.) The same crystal of V. 42
* Mr. Walker's use of the term "similar" to denote an enantiomorphous rela-
tionship between etch-figures seems to me to be inadvisable since it is not needed
in that sense, and such a usage deprives us of a convenient designation for figures
that are not identical but differ from one another very slightly as in the case of
the hornblendes. (American Journal of Science, 1898, Vol.1, p. 181.) "Analogous"
might be employed in this connection rather more freely to mean similarity in
some one or more features, and would need supplementary statement to indicate
wherein the analogy consists.
DALY. ETCH-FIGURES ON AMPHIBOLES. 413
furnished arrow-headed etch-hills on (Til). The direction of the arrows
was iu the sense of, though not parallel to, the b axis inwards. On
(Oil), likewise, the figures of corrosion were etch-hills of tiiangular
shape, in the case of both actinolite and hornblende, but again so ill
defined that no certain statement of likeness or unlikeness could be made.
V. 16 (actinolite), V. 8 (tremolite)^ and V. 20 (richterite), at respective
exposures of two minutes, two minutes, and one minute, show strong
attack on the terminal planes and the development of numerous bosses
on (Oil) ; the same was true of the hornblendes, V. 33 at two minutes,
V. 35 at 1 min., V. 36 at 1^ min., and V. 42 at 1 min. Though charac-
teristic, they do not lend themselves to analysis, and I have been able to
accomplish nothing toward comparative detailed study.
Better success was had with the orthopiuacoid. Among the aluminous
amphiboles, crystals of V. 33, V. 34, V. 35, and V. 42 gave pits at 1:^^,
IJ, 1, and 1 minutes respectively. The result was unifoi'm, — a figure
of triangular outline, isosceles, the upper apical angle bisected symmet-
rically by the plane of symmetry of the crystal. The figure-faces cor-
respond to the edges ; two very steep pyramids, a dome at the base and a
bottom-face parallel to (100), which diminishes as the pit matures. Occa-
sional figures on V. 42 were notable for the replacement of the pyramids
by two pairs of positive and negative pyramids (Plate I. Fig. 28). A
much commoner variation is the symmetrical curving of the two lateral
edges, probably as an effect of secondary solution as the figure grows
older. Simultaneously, the figure tends to grow stouter; thus, a youno-,
light figure of V. 33 was found to have a ratio of altitude to base of
4:1, while on the same face a dark matured figure possessed a ratio
of 3 : 1. I'he stoutest figures measured on any hornblende (V. 34),
showed a ratio of 3 : 2.
Actinolite pits on (100) display one noticeable differonce, and only
one, from these last figures; they are more slender (Plate I. Fig. 27).
The ratio of altitude to base in the pits of V. 17 (at CA 2h min.)
changes from a value of 6 : 1 in the initial figure to a minimum of 3 : 1
in a matured pit. Whether this distinction applies to all the non-alumi-
nous amphiboles or not is a question that needs for its solving more
material than I have had the opportunity of studying.
Finally, we have to note the corrosion pits on the dome (TOl).
These are, on the whole, rather difficult to obtain on hornblende by
reason of the narrow limits between maximum and minimum times of
exposure necessary to bring about their maturing properly. Only three
of the species examined afforded good figures, and only one gave those
414
PROCEEDINGS OF THE AMERICAN ACADEMY.
capable of fairly accurate measurement. The crystals belonged to H. 30,
V. 35, and V. 54. The first yielded figures after 30 seconds' boiling
in cone. HF, the second on one minute's exposure on the water bath,
and the third after two minutes' immersion in a dilute (10%) solution of
HF. The plane (TOl) was in every case strongly affected, and penta-
gonal pits of corrosion were visible (Fig. 9, and Plate I. Fig. 30).
Figure 9.
Figure 10.
The measurements on the crystal of V. 35 were as follows : — ■
ZBAE 84°. ABCD 96°. /.ABC 132°.
The outline is symmetrical to the plane of symmetry of the crystal.
The acute angle BAB points toward the front (in Tschermak's orien-
tation). Each edge corresponds to a plane figure-face indicated in the
drawing, and there is a diminishing bottom-face parallel to (101). The
twinned character of V. 54 was clearly manifested in the attitude of
the pits on the two adjacent dome-faces intersecting in the twinning-
plane (100).
An actinolite (V. 16 at two minutes' exposure) and richterite (V. 20
at one minute) furnished the type of (TOl) pits characteristic of the non-
aluminous amphiboles (Plate I. Fig. 31, and Fig. 10). The figures on
the former were extremely sharp, but a photograph could not be pre-
DALY. — ETCH-FIGURES ON AMPHIBOLES. 415
pared ou account of the roughness of the general crystal-surface. They
are not pentagonal but triangular ; two edges, A and B C (Fig. 10),
meet at an angle of 80° or more, the apes of which points forward on the
crystal. Tlie edges A C and B O are quite straight, and their respective
figure-faces are plane. On the other hand, the figure-face A B B' A' is
curved and is predominant as a curve even when A A' C and BBC are
not well developed. The bottom plane of the pit is a plane parallel to
(TOl) and diminishing in size as the pit matures. AAG and BBC
are steeper (darker) than A BB A', but are of equal obliquity to (TOl),
thus agreeing with the other characteristics of the pit in supplying
perfect monosymmetry for the figure with respect to (010).
The differences in the etclvpits on the positive unit-dome of the acti-
nolitic species as contrasted with the corresponding pits on hornblendes
is striking in face of the fact that we have already seen exemplification
of even greater contrasts between the mineral groups in the behavior of
the unit prism and the clinopinacoid during the same process of etching.
Isomorphism in the Amphiboles.
Chemical crystallographers are of different opinion regarding the value
of etch-figures in determining isomorphism. Arzruni, on the one hand,
denies any necessary relationship between them and the fact of isomor-
phic mixture ; * in this he is supported by Baurahauer, who regards the
pits on dolomite, calcite, and siderite as not indicating lack of isomor-
phism, although their orientation on corresponding faces of all three
species is widely different, f The opposite view has been strongly main-
tained by Retgers in his recent and valuable research on the subject. J
Retgers gives his three critei'ia of isomorphism as follows: — 1. Mix-
tures of the constituent salts of an isomorphic series must take place
in all proportions. 2. There must be a lack of chemical combination
in the mixture : thus diopside (CaMgSiOg) is not a member of the
isomorphic series (CaSiOg and MgSiOg), but an independent body.
Likewise manganese augite (MnSiOg) and the manganolime augite
(MnCaSiOa) are not isomorphic. 3. Etch-figures produced on the
same crystallographic plane of all members of the series will be alike
both in shape and symmetry. To support the last statement, he cites a
* Phys. Chemie der Krystalle, 1893, p. 162 et seq.
t Resultate der Aetzmethode, p. 37.
X Beitrage zur Kenntniss der Isomorphismus. Zeit. fiir phys. Chemie, 1895,
Bd. XVI. p. 36.
416 PROCEEDINGS OF THE AMERICAN ACADEMY.
number of such isomorphic groups whose etch-figures are known, and in
every case thei-e is this similarity among the figures. Thus the experi-
ments of Baumhauer on the double sulphates,* the lime and strontium
hyposulphates,! ^he alums, $ the phosphate (arsenate) group,§ and the
apatite family ; || the studies of Baumhauer and Becke on the carbonates
of iron and magnesium ;1[ those of Wulff on the barium, strontium, and
lead nitrates,** and Becke's on spinel and its relatives ; tt all exhibit a
marked stability in the form of the etch-pits as the different members
of each group are attacked by the same reagent. The earlier objections of
Arzruni are considered to have been met by more recent observations of
Baumhauer (Res. der Aetzmethode, p. 40). Retgers does not contend
that similarity of etch-figures implies isomorphism, as in the two cases of
calcite and soda nitrate on the one hand, and the rutile-zircon-cassiterite
group, on the other; but regards the converse as a fixed law, "das
erhiirtete Gesetz," that genuinely isomorphic substances always show, on
corresponding surfaces, similar etching phenomena. He further states
that "a successful study of isomorphism, without continual controls by
means of the methods of etching, is no longer conceivable." tt At the
same time, he remarks that the etch-figures do vary, and that the limits
of variation in most groups have not yet been determined ; they are,
however, as far as known, always narrow limits.
Now, if this contention be valid, the application of the princijile to
the amphiboles will have important consequences. It will be remem-
bered that tremolite, actinolite, richterite, and astochite, from many
different localities and of various chemical composition, gave uniformly
the same etch-pits on (110), and, where the material was at hand to
determine the point, the same figures on (010), (100), and (101). With
respect to the same planes, the much greater group of common and ba-
saltic hornblendes, also of very variable composition, agreed among them-
selves as well as with barkevikite, glaucophane, crossite, and riebeckite ;
but each of these two sets of etch-figure types was so strikingly different
* Resultate der Aetzmethode, p. 46.
t Zeit. fiir Krystalloszraphie, 1877, Bd. I. p. 54.
t Ber. der k. bayr. Akad. d. Wissenschaften, 1874, and Res. der Aetzmethodcj
p. 45.
§ Resultate der Aetzmethode, p. 43.
II Ibid., p. 39, and earlier references therein.
1 Ibid., p. 67.
** Zeit. fiir Kryst., 1883, Bd. IV. p. 142.
tt Min. und petr. Mitth., 1885, Bd. VII. p. 200.
U Zeit. fur phys. Chemie, 1896, Bd. XX. p. 528.
DALY.
ETCH-FIGDRES ON AMPHIBOLES.
417
from the other, that, od etching (110), (010), or (TOl), one could tell
with ease whether he be dealing with an amphibole of the first class,
devoid of a sesquioxide, or with one of the second, probably aluminous,
but possibly one whose iron represented the total content of sesquioxide.
The testimony of the etch-pits, then, would be to establish perfect iso-
morphism in each of the two groups, but to refute the idea that there is
isomorphism between the two groups themselves. Whether or not this
hypothesis be justified by future investigation, the facts of the case seem
to have an important bearing on the theory of Retgers.
Arfvedsonite figures present relationships to the hornblende type, but
that species cannot be asserted from them to be in the same isomorphic
series with pargasite, for example.
It was thought to be of interest to test further the etching properties
of actinolite and hornblende by using caustic alkalies as the solvent in-
stead of hydrofluoric acid. The results were confirmatory of the division
just made in the amphiboles. A crystal of Zillerthal actinolite was im-
mersed in molten caustic soda for 15 seconds. A large number of sharply
outlined pits were produced on both (110) and (010). Plate I. Figs. 32
and 33 and Fig. 11, are diagrammatic
representations of them oriented on their
respective planes. Now a cleavage piece
of AVolfsberg basaltic hornblende (V. 54)
furnished splendid figures on (110) after
35 seconds' exposure in the same reagent,
and these were of the varieties shown in
Plate I. Figs, 34a and 345, and Photograph
16. Another crystal of the Kafveltorp
hornblende (V. 42) gave the pits of Plate
I. Figs. 35fl!, 35b, 35c, and Photo. 17, on
(110) and pits represented by Figure 12
on (010). Still a third aluminous horn-
blende (V. 33) from Arendal afforded
good figures at 30 seconds, this time prac-
tically identical with those of V. 42.
Without further analyzing these types, it
appears to be a legitimate conclusion that
the differing habit of the two may be
regarded as significant, not accidental, but indicative of a fundamental
difference in the two kinds of substance.
If the tremolite molecule represents one of the fundamental ingredi-
VOL. XXXIV. — 27
Figure 12.
Figure 11.
418 PROCEEDINGS OP THE AMERICAN ACADEMY.
ents in the actinolitic group, the formula of richterite (after Groth, Dana,
and Hintze) would have to be recast to show that the latter species is
the result of the mixture of CaMg3(Si03)4 with another metasilicate
molecule. It would be highly interesting to determine the figures on
grunerite (FeSiOg), cummingtonite [(FeMg)Si03], and dannemorite
[(FeMnMg)Si03], for purposes of similar comparison ; many trials with
material of the two former from the classic localities failed to produce
figures that could be discussed.
The similarity of the riebeckite pits to those on hornblende may go to
show that (if it be true non-aluminous riebeckite that was dealt with
during the examination) the strong influence of the alumina molecule
may be replaced by that of its common associate, sesquioxide of iron.*
In any case, the etch-figures in all thirty of the common and basaltic
hornblendes, as well as in the glaucophanes and barkevikite, can hardly be
explained except as the effect of an interaction of hydrofluoric acid and a
common molecule constituted with reference to one or other or both of
the two ses(|uioxides.
In summary, then, if we accept the law that isomorphic mixtures must
have similar etch-figures on corresponding crystallographic planes, we
have two divisions among the amphiboles, each of which is isomorphic in
itself, but not so related to the other group. If this theory be rejected,
we have still the facts remaining of an important difference in the struc-
tural plan of each division.
IIOLOHEDRAL CHARACTER OF THE MONOCLINIC AmPHIBOLES.
Throughout the whole suite of specimens which I have studied, the
evidence is convincing that the family of amphiboles belongs to the holo-
* Haefcke was so impressed with tlie importance of alumina in the amphibole
molecule that lie was led to consider it with the other oxides, to be in combination
with orthosilicic acid and thus helps to form salts constitutionally different from
the non-aluminous amphiboles, the metasilicates (Inaug. Diss. Gilttingen, Berlin,
1890). The masterful influence of alumina is further seen in Wiik's table of ex-
tinction angles showing the dependence of the angle of extinction of amphiboles on
the percentage of AUOg present. (Zeit. fiir Kryst., 1882-83, Bd. VII. p. 79.)
That the sum of the sesquioxides as well as the amounts of each should be con-
sidered in any comparative study of minerals containing them, is illustrated in an
analysis of Doelter's paper on the pyroxenes, in which he traces the influence of
alumina and of the sum (Al.^Og -f- Fe.iOg -f FeO) on the optical constant c : c of
pyroxene (Neues Jahrb. fiir Min., etc., 1885, Bd. I. p. 43). He finds that FeO alone
will not explain the position of c with respect to the vertical axis, nor will AlgOg
alone nor Fe.iOs alone ; bat he concluded that both the iron oxides added to the
DALY. — ETCH-FIGURES ON AMPHIBOLES.
419
hedral (Groth's '"prismatic") class of the monoclinic system. I ex-
amined practically all varieties with this point in mind : figures produced
with hydrofluoric acid, caustic soda, caustic potash, and bicarbonate of
soda, and the attackability of the different faces, all told the same story.
Besides the diagrams so often referred to, I shall introduce another to
represent etch-pits on (100) of a hornblende (V. 42) acted upon by
caustic soda. (Plate I. Fig. 29.)
CoaiPARISON "WITH THE PYROXENES.
Many years of comparative study, chemical and physical, have evolved
a vast body of evidence to restore something of the early belief in the
identity of the pyroxene and amphibole substance,* but it is, after all,
alumina furnished a serial correspondence to the variations in c : c. If, however,
the doubtful case of Mte. Rossi be rejected and it be noted that tlie extinction c : c
for the Siderao occurrence was read on a section not quite parallel to (010), it will
be seen that the sum (AI.2O3 + FcoOs) gives better results than any of the variables
considered by Doelter. Witness the accompanying table : —
Locality.
Extinction
Angle (010).
AUOs+FeoOs
FeO
Fe^Os
AUO3
FeO+Fe^Os-l-ALGs
Green Augite, Vesuvius
/
41 00
%
• 8.35
%
3.16
%
8.51
%
4.84
%
11.51
Greenwood Furnace
42 25
10.14
2.55
5.05
5.09
12.69
Aguas Caldeiras . .
43 35
11.40
4.81
3.51
7.89
16.21
Pedra Molar . . .
45 45
11.85
5.43
6.18
5.67
17.28
S. Vincent ....
46 45
13.40
5.20
5.25
8.15
18.60
Black Augite, Vesuvius
46 45
14.22
4.09
4.47
9.75
18.31
Garza
47 55
14.61
5.43
4.95
9.66
20.04
Fassaite (Toald. Foja)
47 10
15.11
2.09
5.01
10.10
17.20
Cuglieri
48 00
14.93
5.05
6.32
8.61
19.98
Ribiera da8 Patas .
51 00
22.13
5.95
7.89
14.24
28.08
Siderao
50 05
22.37
9.14
9.29
■e
13.08
31.51
P. da Cruz ....
51 50
31.34
2.23
15.37
16.97
84.57
* Tschermak would find in the relative size of molecule the probable ground
for difference in those striking external characters, crystalline development, and
cleavage. He hypothecates a size for the amphibole molecule just double that of
the pyroxene molecule. Cf. Lehrbuch der Mineralogie, 1894, p. 460.
420 PROCEEDINGS OF THE AMERICAN ACADEMY.
astonishing to observe the very close parallelism which even a cursory
study established between the etching properties of the two mineral
genera. This is one part of the present investigation which the author
regrets to leave in an especially incomplete stage. Thus, it would be
of moment to produce etch-figures on that prism of pyroxene (210),
which is nearly equivalent to (110) of amphibole, and compare them with
those of the last-mentioned face. This has been left undone for lack of
material. But even the facts in hand are wonderfully accordant. To
simplify matters, I shall enumerate some of the main conclusions we have
reached regarding the etch-pits on amphiboles and note the comparisons
with pyroxene in connection with each.
1. Actinolitic amphiboles give one class of etch-pits, aluminous am-
phiboles anotber, — especially evident on (010), (110), and (101). I
have been able to find a similar strong contrast between diopside and
augite in this respect: they were etched on (010) and (110).
2. Not only is such a double cleavage of the groups possible ; there
are positive similarities in the respective pits on the pinacoids (010)
and (100) of the non-aluminous amphibole actinolite and the non-
aluminous pyroxene diopside, and there are positive resemblances char-
acterizing the pits on the same planes of augite and hornblende. The
paper of Pelikan,* referred to in detail further on, along with those
of Wltlfing t and Baumhauer J show clearly that the etching phe-
nomena (outline of figures, number of pits and of figure-faces, orien-
tation of figures, etc.) on (010) of diopside, using hydrofluoric acid,
are hardly distinguishable from those on actinolite. (Compare Photo-
graphs 12 and 13.) Pelikan § has shown, in addition, that the augites
from Vesuvius, Laacher See, Wolfsberg, etc., are characterized by only
one sort of pit on the clinopinacoid and thus are in contrast to diopside,
and, as we now see, are analogous to hornblende. It must be confessed
that no amphibole which I have yet etched has yielded anything like so
perfect figures on the orthopinacoid as those readily procurable on the
same face of Ala diopside (see Photograph 15). While the general re-
semblance to the pits on (100) of actinolite is certainly great, I cannot say
whether or not the pair of figure-faces at the lower end of the diopside
pit is represented in the actinolite pit. (Plate I. Fig. 27, is somewhat
hypothetical as to the " primary " figure-faces.) I have not been able
* Min. u. petr. Mitth., 1896, Bd. XVI. p. 1.
t Die Pyroxen Familie, Heidelberg.
I Poggen. Annalen, Bd. CLIII. p. 75.
§ Op. cit, p, 21.
DALY. ETCH-FIGURES ON AMPHIBOLES. 421
to secure etchings of this plane for augite nor of (TOl) for any py-
roxene.
3. As was to be expected, the pits on the unit-prism of actinolite and
diopside are unlike in outline. I made the experiment of testing the
cohesional property of the actinolite substance on a plane making with
(010) an angle essentially equal to half the cleavage angle of diopside.
A new prism-face was thus ground on a crystal of Zillerthal actinolite,
then carefully polished and etched in the usual way, with cone, hydro-
fluoric acid after two minutes' exposure. The face was found to be
covered with etch-hills, the effect of strong attack (much quicker on the
artificial face than on any natural face in the prism-zone) ; but, among
them, a few pits which were surprisingly like those on diopside (see
Photograph 18), and on aegerine. Further comparison could be made
by etching an artificial face on diopside lying 62° 15' out of the plane
of symmetry, and also by using the caustic alkalies in this round of
experiment.
4. From the close association of the pyroxenes with our group, it is
important to recognise that recently attempts have been made to remove
the diopsides from the holohedral class into a hemihedral (" domatische,"
Groth) class of the monoclinic system. In 1889, G. H. Williams sug-
gested this hypothesis on purely crystallographic grounds, interpreting
the imperfect development of the planes about the extremities of the
vertical axis of crystals from Orange County, N. Y., and Canaan, Conn.,
as an evidence of hemihedrism.* In this, he was followed by Dana in
the " System " (1892, p. 352). But that the failure of planes about one
end of an axis need not mean true structural lack of symmetry is well
known, and has lately been exemplified by etch-figures on cuprite that
restore it to the holohedral category. Pelikan rightly rejected this argu-
ment, but still, on the basis of etching results, he considers it probable
that the diopsides are nevertheless hemihedral in Williams's sense.f He
was led to this conclusion chiefly by the study of the HF pits on (010).
He figures some of these from a Nordmarken specimen which are un-
symmetrical in that lines drawn from one side to another through the
centre of a pit would not be bisected at that centi-e ; there is, in other
words, a lack of that antimetric (dimetric) character necessarily charac-
teristic of (010) if diopside be holohedral. In particular, the asymmetry
of the face is supposed to betray itself in the fact that one corner of the
rhomboidal figure may be truncated by a fifth figure-face while the oppo-
* Amer. Jour. Science, 1889, Vol. XXXVIII. p. 115. t Op. cit., p. 19.
422 PROCEEDINGS OF THE AMERICAN ACADEMY.
site corner is not so truncated. But this phenomenon is a familiar one
in etching, as pointed out bj Becke with reference to siderite and mag-
nesite ; it is simply analogous to the unsymmetrical appearance of a
crystal, due to the non-development of faces, which, by the known sym-
metry of the crystal, should appear on it.* Furthermore, Pelikan states
that, even in the typical Ala diopside, the regular antimetric pits appear
in abundance, and my own observations on another crystal from the
same locality (P. 70, Photograph 13) confirm the statement. There is
no trace on my well etched specimen of pits that are not antimetric.
Secondly, he uses certain figures on the clinopinacoid of Ala diopside
as suggestions of hemihedrism because of the curvature of the edges be-
tween their respective figure-faces (see his drawing, Op. cit., p. 20). I
have not been able to establish the observation by reference to my Ala
specimen, and I am inclined to think the curvature must be a consequence
of the solution of planes no longer "primary." If secondary solution
really exists, (and the numerous experiments of Becke seem to prove it
incontestably,) we should expect it to warp the straight edges between
primary figure-faces with some such curves as those represented in
Pelikan's drawing.
The evidence seems to be perfectly convincing that diopsides as well
as augites, amphiboles as well as pyroxenes, are holohedral, and there-
with we may close this brief comparative sketch of their etch-figures.
Crystallographic Orientation of the Amphiboles.
The extraordinary resemblance between the amphiboles and pyroxenes
in the matter of etch-figures is certainly correlated with likeness in mo-
lecular structure, and is an effectual criticism of that mischievous con-
servatism which has not accepted the arguments of Tschermak, G. H.
Williams, and others, in favor of a change in the classic crystallographic
orientation of amphibole, introduced by Nordenskiold. The new differs
from the old simply by the rotation of the crystal about the vertical axis
* After these lines had been written, the paper by Baumhauer appeared in the
Zeit. fiir Kryst. (1898, Bd. XXX. p. 97), in which the author stated, as the result
of a careful examination of some of Pelikan's original material, that, in his opinion,
diopside is holohedral, and that the anomalous pits described by Pelikan are really
only imperfectly formed representatives of either of the two tj'pes of normal anti-
metric pits, or are the result of the combination or fusion of these two types (Op.
cit., p. 101). On similar grounds, Baumhauer regards Colemanite as monoclinic,
although certain etch-pits apparently indicate an asymmetric character for the
mineral.
DALY. — ETCH-FIGURES ON AMPHIBOLES. 423
by an angle of 180°. The reasons for altering the old orientation have
been so well expressed by Williams that they here need no more than
mere mention. (1) The base of pyroxene is an important plane on
account of the mineral's well known habit of twinning parallel to that
face, and, secondly, on account of the (probably resulting) planes of
parting that are so often developed parallel to (001). Amphibole shows
both phenomena with reference to the unit dome of the old orientation.
(2) Parallel intergrowths of the two minerals are much more intelligi-
ble if this unit dome of amphibole, sensibly parallel as it is, to the
base of the pyroxene, be really regarded as the base of the amphibole.
(3) There can be no doubt that the optical and other properties of the
two groups can be more easily compared in the new orientation.
Now, when we remember that the pits on (100) are in both families
boat-shaped figures with the bow of the boat headed in opposite direc-
tions if amphibole be placed in the old orientation, in the same direction
if in the new ; that the two kinds of joits respectively characteristic of
the clinopinacoid on diopside and actinolite are practically identical in
arrangement in the new position advocated ; that the cohesional relations
of the aluminous amphiboles as regards (010) witness to the same close
relationship; and that the analogy of the figures on the two sorts of
prismatic cleavage is so well brought out in the greater bluntness of the
upper end in each case ; — it is undeniable that the physico-chemical facts
of corrosion with the acid render it highly expedient to follow Dana
in his " System," and Lacroix in " La Mineralogie de la France," in
reversing, for purposes of systematic comparison, the orientation so
recently, and with so little reason, advocated by Hintze.*
Optical Orientation of an Amphibole Crystal or Cleavage
Plate by Means of Etch-Figures.
I cannot subscribe to the opinion of Pelikan,t that the etch-pits on
the orthopinacoid and on the clinopinacoid of diopside are valueless for
purposes of optical orientation. Whether it be because of different
methods of procedure or not, yet I have always found a minimum of
difficulty in applying Wiilfing's directions for the employment of etch-
figures to this end. The same facility of use characterizes the corre-
sponding figures on amphiboles. For their actual application, as well
as for that of the pits on (110), the reader is referred to the type
* Handbucli fiir Mineralogie, p. 1186.
t Op. cit., 1896, p. 12.
424 PROCEEDINGS OP THE AMERICAN ACADEMY,
diagrams and their accompanying descriptions. In every case, there is no
doubt which is the upper end of the cleavage piece, on which the (110)
etch-pits can be seen. This study of amphiboles was undertaken pri-
marily to determine the value of etch-figures on cleavage pieces, in order
to give a means of orienting those belonging to new varieties which are
being so often discovered in allotriomorphic development in crystalline
rocks. For this reason, I was especially glad to have access to the
many species named in the rather voluminous list of page 379. The con-
clusion is that, while etch-figures render possible a certain amount of
differentiation in the whole family of amphiboles, they yet have so much
constancy, so many analogies in outline, as to furnish a reliable means of
determining up and down, right and left, in a new variety.
Etch-Figures of Orthorhombic and Triclinic Amphiboles.
There remain two problems which I have set before me for solution
by means of etching; one, the crystal system of antliophyllite and
gedrite, the other, the comparison of aenigmatite and monoclinic amphi-
boles in the matter of cohesions on their prismatic cleavage-faces.
Etch-Figures on Anthophyllite and on Gedrite (110).
With no other amphibole did I find so much difficulty in producing
and discussing figures of corrosion as on anthophyllite and its near
relative, gedrite. By dint of some patience, however, pits were ob-
tained on (110) that fully served the purpose. The relatively great
resistance of these minerals to the solvent power of hydrofluoric acid
was illustrated in every specimen. Although recognizable pits could be
seei on P. 1, P. 2, and P. 3 at 3 minutes, they were often exceedingly
sn.al (longest diameter about equal to 0.002 mm.) and the amount of
material removed in solution was insignificant. They could not, however,
be much enlarged by longer immersion, as shown by a number of trials
at various exposures up to 20 minutes. Instead, they became gradually
lost in an indefinite confused surface of irregular solution. The best
figures were furnished by P. 2 after 2 minutes' exposure (see Plate I.
Fig. 36a and 366), most of them were very shallow, elliptical in shape,
with the longer axes of the ellipses uniformly parallel to the vertical
axis of the crystal; tliey were commonly aggregated along lines of
cleavage. As shown in the diagram, other pits were considerably larger,
rectangular, sometimes wholly black under vertical incidence of the
light (Plate I. Fig. 36c), at other times characterized by several visible
DALY. — ETCH-FIGURES ON AMPHIBOLES. 425
ligure-faces (Plate I. Fig. 3Gd). These, like the smaller elliptical pits,
possessed one plane of symmetry transverse to the trace of the cleav-
age, and were, of course contrary to expectation, sensibly symmetrical
to a plane parallel to the cleavage trace. The other three planes of the
unit prism showed figures that clearly indicated the holohedral ortho-
rhombic character of anthophyllite. Gedrite (P. 4 at 8 miimtes' exposure)
afforded precisely similar phenomena to those of anthophyllite.
These observations confirm the optical determinations of orthorhombic
symmetry by Des Cloiseaux, and meet the objection thereto by Ilintze,*
who referred to the possibility that both anthophyllite and gedrite may
really be monoclinic, but imitate the optical behavior of an orthorhombic
mineral. That these minerals are monoclinic for chemical reasons has
more recently been suggested by Retgers : f we have once again an
illustration of how etch-figures may come to the help of the chemical
theorist.
In passing, it may be noted that the plane of easy parting (010) in
hypersthene (P. 69) and bronzite (P. 67 and P. 68) gave with hydro-
fluoric acid extremely sharp pits with six-sided bisymmetric outline, that
betokened orthorhombic symmetry for these minerals, though I could
obtain no certain residts on etching (110).
Etch-Figures on the Prismatic Cleavages (110) and (ITO)
of Aenigmatite.
Without destroying the splendid crystals of aenigmatite which Mr.
Ussing was good enough to place at my disposal, I was able to etch and
orient cleavage pieces of that mineral. The crystals showed (Brogger's
orientation) the planes (110), (ITO), (010), (Tol). The standard con-
ditions of etching were again used, and, after some trouble, both cleavages
were finally attacked with successful outcome. Figure 13 and Plate I.
Fig. ola will give an idea of the pits. They are triangular in outline,
very analogous to the pits on the cleavage of common hornblende, but
with a decided peculiarity of orientation. The attitude of the figure-faces
shows a greater resistance to attack in a direction parallel to the edge
110 : ITO than at right angles to it, a phenomenon we have already no-
ticed in treating of the analogous position of the caustic soda figures on
arfvedsonite.J Neighboring pits may overlap, and thus the figure-faces
* Handbuch, p. 1180.
t Zeit. fiir phys. Chemie, 1895, Bd. XVI. p. 618.
I Baumhauer observed the independence of etch-figures and cleavage in the
case of several species. Poggen. Annalen, 1872, Bd. CXLV. p. 460.
426
PROCEEDINGS OP THE AMERICAN ACADEMY.
Figure 13.
which are nearly parallel to the trace of the second cleavage may run
into one another, forming a more or less straight edge in that direction
(Plate I. Fig. 37^). The small differences in the shape and orientation
of the pits on (110) and (ITO) are
such as to form another testimony
to the fact expressed by Forstner
that cossyrite (aenigmatite) closely
approximates a monoclinic habit.*
(See Figure 13.)
It is possible that the moot ques-
tion as to whether kolbingite is a
distinct species, or (after Brogger)
an intergrowth of arfvedsonite
and aenigmatite, might be settled
by etching cleavage pieces of the
mineral.
Zinc-bearing rhodonite (fowlerite, P. 72) is much nearer the mono-
clinic pyroxenes, as shown by etch-figures, than is aenigmatite like the
monoclinic amphiboles. Fowlerite at 45 seconds' exposure gave tri-
angular pits on (110) and (ITO), which are elongated in the sense of
the edge 110 : ITO, and are strongly suggestive of the pyroxene figures
on cleavage plates.
Summary of Conclusions.
Reviewing the ground over which we have come, we may make the
following brief resume of results : —
(1) It seems to be clear that, for the group of the amphiboles, a
special method of etching must be adopted, if a comparative study of
the etch-figures derived from the different species is to be instituted. A
universal solvent must, of course, be used, but its temperature, degree of
concentration, and facility of convection at the time of each attack must
be attended to if a strict control over the effects of corrosion be possible.
It is only by observing this principle, that the measurement of the out-
lines of pits and their elements of form will lead to the most valued con-
clusions ; and we have seen that measurement, i. e. discussion of the
figures by quantitative methods, serves this purpose much more perfectly
than does a mere statement of the kind of pit or etch-hill to be seen on
any given specimen. It has further been shown, with respect to the pits
* Zeit. fur Kryst., Bd. V. p. 350.
DALY. — ETCH-FIGURES ON AMPHIBOLES. 427
on the cleavages and on the clinopinacoid of amphibole at least, that the
action of hydrofluoric acid is seriously affected by its mixture with j^ure
water and also seriously affected, in the opposite sense, by its mixture
with sulphuric acid; the action of pure hydrofluoric acid gas is inter-
mediate to that of both kinds of mixture. Is it not always necessary to
guard the conditions of attack when etching of crystals by chemical cor-
rosion is to be the means of comparison among the substances represented
thus in the crystalline state ?
The relatively rapid process of studying etch-flgures in vertically in-
cident light can, under the circumstances just outlined, lead to results of
importance, not inferior to that attaching to problems where the rather
laborious method of determining the exact symbols of figure-faces by
means of the goniometer and the Brewster light-figures is necessary.
(2) A scale of optimum exposures for (110) under standard conditions
is recorded and the attempt is made to systematize the amphiboles as
regards their attackability on the same face. The comparative attack-
ability by hydrofluoric acid of the different faces on certain non-aluminous
amphiboles has been determined. It has been found that this property
is affected, in large measure, by the physical state of the specimen
attacked.
(3) In many instances it has been exemplified that a systematic com-
parison of etch-figures on different species will be of most service if the
observer recognizes the principle that there is a decided change in the
etch-pit characteristic of any face, in accordance with the stage it has
reached in the process of maturing from an initial figure to the often very
different figure peculiar to an advanced stage of corrosion. Chiefly for
this reason, it has been found difficult, if not impossible, to tell which of
the many successive figure-faces composing a given pit on amphibole are
the " primary " figure-faces of Becke's definition. It may have been, too,
partly for this reason, that I have found it as yet impossible to co-ordinate
perfectly the figure-faces and the related directions of easy and difficult
solution, so as to construct the " Losungsoberflache " characterizing any
amphibole.
(4) The curious adventitious etch-hills on hornblende, illustrated in
Photograph 12, have led us to suspect that they in no wise represent the
true cohesional property of the face considered, but have suggested the
hypothesis (following Becke) that they may be due to the unequal pro-
tection of the mineral surface by the solid products of the chemical reac-
tion, and that the parts so shielded from attack will project above the
general surface after corrosion has further advanced. It would be
428 PROCEEDINGS OP THE AMERICAN ACADEMY.
interesting to know if other studies of etcli-hills will confirm this
hypothesis.
(5) A point of considerable theoretical interest is raised by the be-
havior of the clinopinacoid of actinolite when etched by hydrofluoric
acid. It is another of those puzzling cases of the existence, side by
side, on the same face, of two quite different kinds of figures, an asso-
ciation for which valid explanation has not yet been vouchsafed.
(6) The amphibole family merits particular notice from the student of
etch-figures because it forms a test case for the theory that isomorphic
crystalline bodies must have similar etch-figures on corresponding faces.
If actinolite and common hornblende, for example, be isomorphous,
then their etch-jiits must be " similar " in a sense very different from that
adopted in the foregoing paper. By our usage, they can only be said to
be " analogous." In view of these facts, if this theory of the association
of isomorphism and etchings be of universal apj^lication, it will be neces-
sary to define more closely than has yet been done by an advocate of the
theory, the degree of variability that may occur in the etch-figures of
any isomorplious series.
(7) The amphiboles are throughout holohedral.
(8) A further proof of the extraordinary similaiity between the pyrox-
ene substance and the amphibole substance is aff"orded by the study of
the pits on crystals showing (010), (100), and (110). This is especially
true of the phenomena which can be observed when the clinopinacoid is,
in each case, etched with hydrofluoric acid.
(9) It is incontestable that, in spite of their different crystalline de-
velopment and angles of cleavage, pyroxene and amphibole are so
closely and so instructively allied that the standard orientation for both
should bring out as conveniently as possible their points of resemblance.
Taking etchings as particularly significant of what there be of genuine
likeness in the two kinds of substance, there can be no doubt that the
orientation proposed by Tschermak should be universally adojjted in pref-
erence to the older orientation.
(10) It is further possible to make etch-figures on amphibole of prac-
tical value by using them as a means of orienting cleavage flakes and
other crystal fragments. This use is parallel to that proposed by "Wiilfing
for the pyroxenes.
(11) Lastly, so far as etch-figures may be trusted to show relation-
ships, we have the following results of our survey as to the systematic
classification of the amphiboles.
The patent pronounced separation between non-aluminous and alumi-
DALY. • — ETCH-FIGURES ON AMPHIBOLES, 429
nous amphiboles, signalized by the two-group division of all our hand-
books, is once more confirmed. At the same time, attention has been
again called to the overwhelming importance of a sesquioxide, whether
iron oxide or alumina, in the mineral.
Glaucophane and gastaldite are the same species, and both isomorphous
with hornblende.*
Arfvedsonite appears to hold a more or less independent place in the
family of amphiboles. '
Barkevikite is more closely related to common hornblende than to
arfvedsonite.f
Anthophyllite and gedrite are plainly orthorhombic and holohedral.
Aenigmatite diverges considerably from the amphibole habit, but be-
trays a tendency toward symmetrical cohesional property, as it does
toward crystallographic symmetry. t
Lastly, it is believed that our present methods of determination of
species can be reinforced by the detailed study of mineral groups with
respect to etching. The peculiarities of the pits on the cleavages of
riebeckite, arfvedsonite, and barkevikite make it easy to say to which
of these a given cleavage flake belongs. Similarly, the differentiation
of crossite and glaucophane, difficult as it often is by purely optical
methods, is ready at hand if the mineral be etched on (110). The
striking characteristics of the Philipstad hornblende (V. 101) first became
evident in the process of etching cleavage pieces. Its description as a
new variety will form the sequel to this paper.
* Cf. Striiver's statement: "It is probable that glaucophane and gastaldite
are isomorphous with amphibole [proper], but it is not yet proved." Neues Jahrb.
fiir Min., etc., 1887, Bd. I. p. 217.
t The same opinion is held by Lacroix, chiefly on optical grounds (Mine'ralo-
gie de la France, Tom. I. p. 561) ; the opposite opinion by Brogger (Zeit. fiir Kryst.,
Bd. XVI. p. 414), followed by Dana (System, p. 403), and Hintze (Handbuch.
p. 1256).
f See Brogger, op. cit., p. 424.
PLATE I.
Fig.
1. Initial form of pit, Actinolite type (110), HF. X 300.
2. Mature pit of the Actinolite type (110), HF. X 300.
3. Another form of the mature pit of tlie Actinolite type (110), HF. X 300.
4. A rather exceptional variant on the normal Actinolite type (110), distin-
guished by a well defined fourth figure-face at tlie lower end, HF.
x300.
6. Initial form of the Wolfsberg sub-type (110), HF. x 1800.
6. The Wolfsberg sub-type (110), HF. x 300.
7. Compound stepped etch-pit of the Wolfsberg sub-type (110), HF. X 300.
8. Exceptional pit of the Wolfsberg sub-type, showing a fourth figure-face at
the lower end (110), HF. x 300.
9. Immature pit of the Kragero sub-type (110), HF. x 300.
10. Matured pit of the Krageru sub-type (110), HF. X 300.
11. Another form of the last on another Hornblende variety, showing the com-
mon occurrence of a fourth figure-face at the lower end of tlie pit. X 300.
12. An exceptional pit found on (110), along with the pits of Figure 11. X 300.
13. The Edenville sub-type (110), HF. x 600.
14. Another form of the last, where the individual figure-faces can no longer
be distinguished. X 600.
15 and 16. Two forms of the Glaucophane type (110), HF. X GOO.
17. The Crossite type (110), HF. x 1000.
18 and 19. Two forms of the Kiebeckite type (110), HF. x 1000.
20. The type of pit on Barkevikite (110), HF. X 300.
21 and 22. Two forms of the pits on Arfvedsonite (110), HF. X 175.
23a, 236, 23c. Pits on (010) of Actinolite, HF. 23a is the "light" etch-pit, the
other two the "dark" pits simultaneously occurring with the first,
HF. X 300.
24. The modification of the normal pit on Actinolite (110) by the admixture
of h% sulpliuric acid to the commercial hydrofluoric acid generally used.
X300.
25a and 256. The same as the last, except that the admixture is here 50% of
sulphuric acid. X 300.
26. The type of pits on (010) of common and basaltic (aluminous) Hornblendes,
HF. X 300.
27. A pit on (100) of Actinolite (non-aluminous amphibole), HF. X 300.
28. A pit on (100) of basaltic Hornblende (aluminous), HF. X 300.
29. Caustic soda pit on (100) of basaltic Hornblende. X 300.
80. Pit on (TOl) of basaltic Hornblende, HF. X GOO.
31. Pit on (fOl) of Actinolite and other non-aluminous amphiboles, HF. X 300.
32 and 33. Caustic soda pits on Actinolite (110). X 1200.
34a and 346. Caustic soda pits on basaltic Hornblende (110). X 300.
85a, 356, and 35c. Caustic soda pits on the Kafveltorp Hornblende (110). X 600.
36a, 366, 36c, and 36c?. The various types of pits produced on Anthophyllite
and Gedrite, (110), HF. X 3000.
37a and 376. Pits on Aenigmatite (110), HF. X 3000.
Daly— Etch-Figures on Amphiboles.
Plate I.
Proceedings of the American Academy.
THE HELIOTYPE PRINTING CO.. BOSTON
PLATE II.
DESCRIPTIOy OP THE SnCEOPHOTOGRAPHS.
Etch-figures are illustrated as follows : —
2^0. 1. Actiiiolite type on (110), Zillerthal Actinolite, HF. x 88.
No. 2. Kragen") sub-type on (110), Krageru Hornblende, IIF. X 78.
No. 3. Tlie same on another variety, Kafveltorp Hornblende, x 78.
No. .4. Philipstad sub-type on (110), (crystal face = outer zone), showing pits
with blunted lower end, riiilipstad Hornblende, HF. x 225.
No. 5. The same on another crystal, showing pits characterized by a sharpen-
ing of the lower end. X 05.
No. 6. Philipstad sub-type on (110), (one form on the cleavage surface of the
inner zones,) Philipstad Hornblende, HF. x 225.
Daly — Etch-Figures on Amphiboles.
Alii v.. t-^'
%.
' -r^
'r:.':\i
Proceedings of the American academy.
THE HELIOTYPE PRINTING CO.. BOSTON
PLATE III.
DESCRIPTION OF THE MICBOPHOT06RAPHS.
Etch-figures are illustrated as follows : —
No. 7. The same as No. 6 on another crystal, showing the common distortion
of the pits, HF. X 78.
No. 8. Caustic soda pits on Philipstad Hornblende (110), (crystal face). X 225.
No. 9. Caustic soda pits on Philipstad Hornblende (UO), (cleavage, inner
zones). X 225.
No. 10. Etch-hills on (110), basaltic Hornblende from Mayenegg (HF gas).
X71.
No. 11. Pits on (110), Arfvedsonite from Kangerdluarsuk, HF. X 78.
No. 12. Pits on (010), ZUlerthal Actinohte, HF. x 78.
Daly— Etch-Figures on Amphiboles.
PLATE III.
12
Proceedings of the American academy.
THE HELIOTYPE PRINTING CO.. BOSTON
PLATE IV.
DESCRIPTION OF THE MICR0PH0T0GRAPH3.
Etch-figures are illustrated as follows : —
No. 13. Pits on (010), Ala Diopside, HF. X 78.
No. 14. Pits on (010), basaltic Hornblende from Kafveltorp, HF. x 78.
No. 16. Pits on (100), Diopside from Ala, HF. X 78.
No. 16. Caustic soda pits on (110), Wolfsberg basaltic Hornblende. X 78.
No. 17. Caustic soda pits on (110), basaltic Hornblende from Kafveltorp.
X 225.
No. 18. Pits on (110), Diopside from Ala, HF. x 78,
Daly— Etch-Figures on Amphiboles.
Plate IV.
15
17
Proceedings of the American Academy.
THE HELIOTYPE PRINTING CO., BOSTON
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. No. 16. — March, 1899.
V. — CONTRIBUTIONS FROM THE HARVARD MINERALOGICAL.
MUSEUM.
ON A NEW VARIETY OF HORNBLENDE.
By R. a. Daly.
ON A NEW VARIETY OF HORNBLENDE.
By R. a. Daly.
Presented by J. E. Wolff, February 8, 1899. Received February 11, 1899.
The abnormal etching characters of the Philipstad hornblende referred
to in the foregoing paper have suggested further study of the mineral ;
the description of its otlier properties shows that it should be regarded as
an independent member of the amphibole group.
It is a variety given me for study by Professor Berwerth of the Royal
Museum in Vienna. It is catalogued in that collection as " A. o. 458,
Philipstad, Sweden." In addition to the original cleavage pieces for
etching purposes, Professor Berwerth has been kind enough to turn over
to me several fine crystals from the parent druse and enough extra
material to permit of chemical analysis. To him, in thus abundantly
supplying me with the mineral, my best thanks are due.
The crystals stand upon a compact mass of the same hornblende.
The usual planes (110), (010), (100), (130), (Oil), (Tschermak's orien-
tation), with normal interfacial angles, are well developed. The adjoin-
ing table shows the close correspondence of the observed angles with the
calculated angles (cf. Lacroix, Mineralogie de la France) : —
Observed. Calculated.
110:130 150° 9" 150° 6'
010 : 130 147° 23' 147° 29'
110:110 (faces) 124° 11 '-124° 17' 124° 11
" " (cleavages) 124° 27'
110:010 117° 50' 117° 54'
110:011 68° 33' 68° 46'
011:011 149° 12' 149° 11'
The unit prism is usually striated, owing to the presence of vicinal
planes.
The reflexes in the goniometer from the planes of the vertical zone
were often considerably displaced. This is probably due to the warping
of the crystals. In several of the latter, it is possible to see with the
VOL. XXXIV. — 28
434 PROCEEDINGS OF THE AMERICAN ACADEMY.
naked eye a marked flexure and even twisting of the prism-faces. I
consider that the curious distortion of the etch-tigures on (110) is due to
the warping and consequent molecular strain.*
Several of the crystals are twinned parallel to (100).
The different etching behavior of the crystal-face and of the surface
of cleavage has been explained as due to the zonal structure which is
a prominent characteristic of the mineral. Six oriented sections and
numerous cleavage pieces display the structure : it is illustrated in Fig-
ures 1, 2, and 3.
Figure 1. Figure 2. Figure 3.
Figures 1 and 2. — Two sections parallel to (010) of the same doubly termi-
nated crystal, showing zonal structure. In Figure I three zones, in Figure 2 four
zones, are indicated. The zone of deepest tint is the most closely shaded, that of
the lightest tint is left unsiiaded. The lines of shading run parallel to the cleavage
trace. The trace of the edge 010 : Oil slopes downward from right to left.
Figure 3. — A section parallel to (110), showing three zones represented as in
Figures Ijand 2. The extinction of the lightest zone is 17°, that of the intermedi-
ate zone is 19°, and that of the darkest zone is 22° 30'.
That this structure is rare in the amphiboles is clear from the recent
statement by Becke in his essay on the zonal structure of crystals in the
eruptive rocks.f Brogger describes one casein connection with his cato-
phorite series. He notes the fact that sometimes the core of a crystal
may consist of catophorite and the outer zone of arfvedsonite.J Tscher-
mak long ago noted another example in a Vesuvius hornblende. §
Palache has figured the structure in crossite. ||
The diagrams show that the bulk of each crystal is composed of
pretty uniform substance, in which a darker colored phase of the mineral
may be apparent, either without definite arrangement with respect to the
former or in the form of true hour-glass intergrowth with it. In the
* See These Proceedings, Vol. XXXIV. page 400.
t Min. und petrog. Mittheilungen, 1898, Bd. XVII. p. 101.
t Die Gesteine der Grorudit-Tinguait-Serie, pp. 27 et seq.
§ Min. und petrog. Mittheilungen, 1871, Heft I. p. 40.
II Bulletin, Department of Geology, Univ. of California, Vol. I. p. 187.
DALY. VARIETY OP HORNBLENDE. 435
latter case, the darker areas of the section generally occur at the ends of
the crystal, representing a late stage of growth, and are, with the rest
of the crystal, commonly covered with a thin mantle of still a third kind
of substance considerably lighter in tone than either of the other two.
Accurate sections cut in the appropriate directions by M. Werlein of
Paris enabled me to determine the chief optical properties of the horn-
blende. The optical plane is parallel to the plane of symmetry. The
axis of least elasticity lies in the obtuse angle (i (Tschermak's orienta-
tion), making an angle of 15° 9' with the vertical axis using yellow light
or 15° 5' using white light. The mineral is negative. In oil with an
index of refraction of 1.609, I found the optical angle (2 H) to be
53° 24'; in another oil with an index of refraction of 1.5011, I deter-
mined 2 H to be 57° 24'. The hyperbolas were not well defined, and,
on account of strong absorption, the readings had to be made in the
brightest white light procurable.
Owing to the extreme ease with which the mineral cleaves, it was
found impossible to cut oriented prisms for the purpose of finding the in-
dices of refraction ; nor was any other method feasible under the cir-
cumstances. The true optical angle cannot then be found from 2 H,
since the mean index of refraction is not known. It may, however, be
considered that this index lies withm the limits of 1.622 (tremolite) and
1.725 (hornblende, a high value). The first reading for 2 H (53° 24')
would give for b = 1.622, 2 V = 52° 56', and for 6 = 1.725, 2 V =
49' 42'. The second reading (57° 24') would give for the same values
of b, 2 V = 52° 46' and 49° 24'. The closeness of the agreement in the
respective calculated values of 2 V is rather fortuitous. The optical angle
for this section is, then, within a degree or so of 50°.
The double refraction seems to be low. The dispersion is weak
(p<^)-
On (110) the extinction varies with the zones, increasing with the
depth of tint. ' One dark zone gave in white light an average reading of
20° 53' ; other lighter zones afforded extinction angles as low as 17°.
The total range, so far as observed, lies between 22° 30' and 17°. This
can only mean that the optical angle for the different zones varies and
must have values between 42° and 60°. (See Figure 3.)
The pleochroism is very strong in characteristic colors : -
a = light brownish green.
b =1 dark yellow green.
t = dark blue green.
b >c>a
436 PROCEEDINGS OF THE AMERICAN ACADEMY.
This scheme of color and absorption applies to all the zones, the colors
being simply modified in intensity.
The specific gravity was determined in methyl iodide solution at 1 6° C.
The average for two unaltered crystals is 3.275. An outer light-colored
zone gave 3.195, and an inner darker zone 3.230. The difference be-
tween the last two was too small to permit of the separation of the light
and dark zones. I doubt that the lightest zone is more than one per cent
of the whole. There are no important inclusions in the mineral.
M. Pisani of Paris made an analysis of the hornblende ; it resulted as
follows : —
SiO. 45.20
TiOa 0.84
AloOg 7.34
Fefis 7.55
FeO 15.80
MnO 1.52
CaO 12.30
MgO 8.40
NaoO 0.80
KoO 0.37
Loss on ignition 0.70
100.82
The analysis does not lend itself to calculation in a satisfactory way.
There is considerable divergence in the proportions of the oxides from
an old analysis by Rammelsberg of a Philipstad hornblende with a spe-
cific gravity suggestively close to that of our hornblende.*
It will be seen that the most noteworthy feature of the analysis is the
high percentage of ferrous iron, a fiict which correlates the mineral with
hastingsite, which also has an unusually great proportion of this oxide,
as well as an extraordinarily small optical angle.f
That a high content of ferrous iron (plus MnO) always means a cor-
respondingly small optical angle cannot be asserted ; pargasite affords a
case sufficiently clear to invalidate any such claim. Yet it does seem that
there is some intimate relationship between the amount of the oxide and
the optical angle. The analogy of another group of allied silicates is in
striking corroboration of this conclusion. Thus Hiutze t gives a table
* See Hintze, Handbuch der Mineralogie, 1894, p. 1223.
t Cf. Adams, Canadian Record of Science, 1896, Vol. VII. p. 77.
} Handbucli, p. 9G4.
DALY. — VARIETY OF HORNBLENDE. 437
of fourteen chemical and optical analyses of enstatite wherein the optical
angle in oil continuously decreases from 133^ 8' (red light) to 59^ 20'
(green light) while the percentage of (FeO plus MnO) simultaneously
increases from 2.76% to 33.6%.
This hornblende is thus unique among the species yet described in that
it possesses the combination of properties including an unusually small
optical angle, an unusual pleochroism and absorption scheme, a well de-
veloped zonal structure, and quite anomalous etch-figures with hydro-
fluoric acid on the prism (110) and on the clinopinacoid.
For convenience of reference, this variety of amphibole may be called
philipstadite, from the name of the locality whence it was derived.
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. No. 17. —March, 1899.
THE ORTHOPTERAN GENUS SGHISTOCERCA.
By Samuel H. Scudder.
THE ORTIIOPTERAN GENUS SCIIISTOCERCA.
By Samuel H. Scudder.
Received February 20, 1899. Presented March 8, 1899.
ScHiSTOCERCA belongs to the Acridii, the typical group of AcridiinaB,
in which the fastigium is dellexed and passes insensibly into the frontal
costa, lateral carintE are wanting on the pronotum, the mesosternal lobes
are longer than broad and usually produced and strongly acutangulate
posteriorly on the inner side, the hind tibi;e have smooth margins with
numerous spines regularly disposed on both sides, but with no apical spine
on the outer margin, and the second tarsal joint is only half as long as
the first.
There is but one other genus in the group, Acridium, from which
Schistocerca was separated by Stal in 1873, on account of the apically
broader anal cerci of the male and the apically fissate subgenital plate of
the same sex. In doing this he also separated the Old World species of
Acridii from those of the New "World, for Acridium does not occur in
America and Schistocerca is found only in the New World, except for a
single species, which occurs both in South America and in Africa, but
which has also been found in such circumstances in mid-ocean as to render
it in the highest degree probable that Africa was originally colonized from
America.
Schistocerca is therefore normally an American genus. Like Acridium
it is composed of large species with a wing expanse usually reaching
nearly or quite a decimeter, though it contains more species of a moderate
size than does Acridium, and some much smaller than any Acridium
known to me. Two at least of the larger forms, including the species
common to the two worlds, are known to be both migratory and very
destructive ; but the greater number appear to do less harm than their
large size would lead us to expect. The species of Acridium are mostly
confined to Africa, southern Asia, and Australia, and many species are
apparently still undescribed.
A considerable number of species of Schistocerca have been described
from America, but many have received more than one name, even since
442 PROCEEDINGS OF THE AMERICAN ACADEMY.
Stal first brought together the synonymy of the species known to him, in
his Recensio Orthopterorum ; and a number of described forms are still
indeterminable with the considerable material I have brought together
for their study. Especially is this true of Walker's species, his genera
Acridium and Cyrtacanthacris being heterogeneous assemblages of species
of many genera, and usually quite impossible of determination apart from
the specimens themselves in the British Museum.
Unfortunately, a not inconsiderable part of the material on which the
species are here separated is not so good as one could wish, being speci-
mens dried after previous immersion in spirits, — a favorite mode of
collecting these bulky Acridians, but one which dulls the coloring and
often exaggerates the salient parts of the structure of the head and pro-
notum. I have therefore relied as little as possible on mere color, which
nevertheless plays a very important part in the distinction of species in
this genus. It takes a longer immersion to destroy the markings, but
these also are sometimes lost. On this account I have been obliged to
discard a small part of my material, which seemed to indicate additional
species coming from localities, from which one rarely obtains specimens.
Of the forty-four species here tabulated, eleven are known from the
United States, twenty-three from Mexico or Central America, six from
the West Indies, and twenty from South America (including the Gala-
pagos), besides one of which the provenance is unknown, but which
probably belongs to South America.
I am indebted to my friends, Messrs. S. Henshaw, A. P. Morse, and
L. Bruner, for loans from their collections, and have had all the material
from Central and South America in the Museum of Comi)arative Zoology
for study. The main portion of my material is found in my own collec-
tion. Only the new forms are described, but the following table will
enable one to determine any of the species I have seen, new or old.
Table of the Species of Schistocerca.
ai. Antennae of male (those of female always relatively sliorter) nearly or quite
one third, often one half, longer than liead and pronotum together.
M. Pronotum rectangulate behind, or in tlie female faintly obtusangulate, the
angle narrowly or very narrowly rounded.
ci. Tegmina distinctly maculate or if occasionally feebly and obscurely macu-
late, then the pronotum is unstriped above ; lateral lobes rarely with any
markings, and when present rarely separated obliquely.
d^. Pronotum with no, or at most very obscure, dorsal stripe, the lateral
lobes at most clouded with fuscous.
SCUDDER, — THE GENUS SCHISTOCERCA. 443
e^. Pronotum not or scarcely tectiform ; eyes normal.
/^. Relatively small and slender; prozona as long as metazona ; greatest
dorsal width of metazona not exceeding width at eyes . . 1. (jracilis.
/'-. Relatively stout and large ; prozona shorter than metazona ; great-
est dorsal width of metazona exceeding widtli at eyes.
r/i. Wings flavo-fuliginous 2. aurantia.
g~. Wings hyaline with fuscous veins.
10-. Maculation of tegmina delicate, obscure, in the distal half sub-
strigate 3. curinata.
Ifi. Maculation of tegmina coarse, distinct.
i^. Wings more or less distinctly infumate ; prosternal spine erect,
yi. Fusco-testaceous ; hind femora with hoary outer face ; hind
tibiae purple 4. cnhimbina.
j-. Olivaceo-testaceous ; hind femora and tibia flavo-oliva-
ceous 5. crocoluria.
P. Wings hyaline ; prosternal spine retrorse . . 6. interrita.
e2. Pronotum distinctly tectiform ; eyes more oblong than usual.
7. camerata.
d^. Pronotum with median dorsal stripe and diversified lateral lobes.
f-. Dorsal stripe of pronotum of equal width throughout; lateral lobes
longitudinally banded, the bands not or scarcely oblique.
f^. Prosternal spine slender ; wings more or less infumate ; hind femora
without fasciation.
9I. Males with no distinct dorsal stripe ; maculation of tegmina dis-
tinct; wings distinctly flavo-infumate throughout ... 8. mellea.
(j^. Males with a conspicuous pale dorsal stripe on head and prono-
tum ; maculation of tegmina often obscure ; wings very feebly infu-
mate at most 9. zapoteca.
f^. Prosternal spine stout; wings hyaline or nearly hyaline; hind
femora obscurely fasciate.
g^. Pronotum very coarsely carinate, the position of the lateral carinas
unmarked ; tegmina densely maculate 10. vaga.
g"^. Pronotum finely carinate, the position of the lateral carinae marked
anteriorly with fuscous ; tegmina sparsely maculate. 11. simulatrix.
c^. Dorsal stripe of pronotum narrowing from in front posteriorly.
yi. Lateral lobes of pronotum mottled and longitudinally variegated
with fuscous, the lower anterior margin not distinct from the rest ; costal
area of tegmina with no light streak 12. pi/ramldala.
/'-. Lateral lobes of pronotum with a strongly oblique anterior and in-
ferior light patch, edged with fuscous ; costal area of tegmina marked
with a pallid streak 13. desiliens.
c^. Tegmina immaculate, or if rarely feebly and obscurely maculate, then the
pronotum is marked with a dorsal stripe, narrowing posteriorly ; lateral lobes
of pronotum often marked below with dull yellow, separated very obliquely
from the darker parts above.
d\ Lateral lobes of pronotum with a very obliquely delimited inferior clay
yellow patch, often obscure in the female.
444 PROCEEDINGS OP THE AMERICAN ACADEMY.
e^. Median stripe of pronotum generally narrowing from in front back-
ward ; wings more or less and simply infumate.
/I. Antennse of male about one half as long again^as head and prono-
tum together ; tegmina immaculate, or feebly and obscurely maculate ;
wings feebly infumate 14. Jlavofasciala.
f'^. Antennae of male not more than one third as long again as head
and pronotum together; tegmina immaculate; wings distinctly infu-
mate 15. iiifumata.
e^. Median stripe of pronotum broad and equal throughout ; wings flavo-
infumate 16. cvqualis.
d'. Lateral lobes of pronotum without markings 17. maya.
Jfl. Pronotum distinctly obtusangulate behind, even in the male, the angle gen-
erally rather broadly and obtusely rounded at apex.
ci. Pronotum never tectate, or at most but very feebly ; tegmina more or less
distinctly, generally distinctly, maculate throughout (or, if without maculation,
the head, pronotmn, and closed tegmina have no light colored dorsal stripe);
wings generally hyaline, occasionally lutescent basally or even throughout, the
veins fuscous or ferruginous, rarely luteous ; lateral lobes ofpronotimi often
marked conspicuously with longitudinal fuscous stripes on a lighter ground,
but often unmarked.
d^. Pronotum so broadly and strongly rounded behind as to be rotundate,
rather than obtusangulate ; prostcnial spine somewhat retrorse.
18. aiistralis,
d'K Pronotum distinctly angulate behind, the angle more or less but not
greatly rounded, at least in the male; in the female it is sometimes rather
broadly rounded.
fii. Pronotum never tectate, generally distinctly striped on the lateral lobes.
f-. Hind femora not transversely fasclate.
7^. Prosternal spine erect or suberect, straight or almost straight;
lateral lobes of pronotum not or feebly marked, and tlien irregularly
blotched.
/(I. Dorsal stripe of pronotum obsolete ; maculations of distal half
of tegmina slight and scattered, or if at all regular it is to form
numerous narrow one- or two-celled transverse broken stripes.
19. gulosa,
Ifi. Dorsal stripe of pronotum moderately distinct ; maculations
of distal half of tegmina massed in several rather broad oblique
bands . 20. hogotensis.
(J-. Prosternal spine distinctl}' curved and retrorse ; lateral lobes of
pronotum distinctly banded longitudinally.
h^. Pronotum griseous, slenderly strigate with fuscous ; wings
nearly h^-aline 21. inscripia.
h-. Pronotum with a broad median light dorsal stripe between
fuscous bands; wings flavo-infumate 22. idonea.
f^. Hind femora distinctly trifasciate.
g^. Pronotum with no dorsal stripe; tegmina distinctly maculate
throughout; interspace between eyes distinctly narrower than nar-
rowest part of frontal costa 23. literosa.
SCUDDER. — THE GENUS SCHISTOCERCA. 445
^2. Pronotum with a median dorsal stripe; tegmina feebly or not
maculate, especially in the distal half; interspace between eyes not
or scarcely narrower than narrowest part of frontal costa.
24. melanocra.
e^. Pronotum feebly tectate, at least in tlie female, not striped or very
obscurely striped on lateral lobes,
yi. Metazona rugulose as well as punctate ; maculations of tegmina
when present quadrate or rounded ; male cerci of subequal breadth,
tapering only in distal half; subgenital plate of moderate length,
apically witli a ratlier shallow V-shaped emargination.
25. riihiginosa,
f". Metazona punctate but scarcely rugulose ; maculations of tegmina
distinctly elongate ; male cerci tapering throughout so as to be at apex
only two thirds as broad as at base ; subgenital plate long and slender,
apically deeply fissate 26. sonorensis.
c^. Pronotum generally more or less tectate ; tegmina immaculate, or at most
marked linearly with fuscous or yellow on lower half (or if, rarely, distinctly
maculate, then the head, pronotum, and closed tegmina are distinctly marked
by a light colored dorsal stripe) ; wings generally lutescent with luteous veins ;
lateral lobes of pronotum very rarely with conspicuous and definite markings,
generally clear or irregularly mottled.
d^. Pronotum with a distinct percurrent median light colored stripe.
fii. Lateral lobes of prozona immaculate, or with feeble light colored
stripe at or above the middle,
yi. Dorsum of metazona plane or nearly plane in both sexes, occasion-
ally faintly tumid anteriorly in the female; hind tibiae purplish, testa-
ceous, or very dull ferruginous.
g^. Hind femora usually not fasciate ; hind tibiae testaceous or red-
dish, sometimes basally purplish above ; subgenital plate of male with
a relatively shallow apical U-shaped fissure, but little deeper than
broad 27. alutacea.
92. Hind femora usually fasciate; hind tibise dark purple; subgenital
plate of male cleft narrowly, almost to the base ... 28. ohscura.
f'-. Dorsum of metazona distinctly tumid in the female and sometimes
in the male ; hind tibiaj coral red or purplish.
g'^. Flavo-testaceous ; fore and middle femora of male very stout;
hind femora generally conspicuously fasciate; hind tibias purplish or
ferruginous 29. Uneata.
g^. Fore and middle femora of male only moderately stout; hind
femora never fasciate ; hind tibiae coral red.
Ix^. Flavo-testaceous, the dorsum of prozona, except for stripe, much
infuscated 30. alhoJ'meata.
h~. Nearly uniform olivaceous, except for the yellow dorsal stripe.
31. t;enusta.
e^. Lateral lobes of prozona with a conspicuous black-edged pallid stripe
below the middle 32. mexicana,
d^. Pronotum with no median light colored stripe, or if a feeble one occurs,
it terminates with the prozona.
446 PROCEEDINGS OP THE AMERICAN ACADEMY.
e^. Wings nowhere roseate.
f^. Prozona neither tectate nor tumid, with an obscure broad median
stripe, not passing the prozona 33. sepa7-ata.
/2. Prozona more or less tectate, and, at least in the female, tumid,
without a median stripe.
f/^. Superior carinas of hind femora obscurely serrate ; inner spines of
hind tibiffi but little longer tlian deptli of tibiae.
h^. Prozona not arched longitudinally ; pronotum of male not more
than half as long again as greatest dorsal width of metazona, its
posterior margin distinctly obtusangulate ; hind tibije red.
34. shoshone.
li'. Prozona feebly arched longitudinallj' ; pronotum of male dis-
tinctly more than half as long again as greatest dorsal width of
metazona, its posterior margin rather feebly obtusangulate; hind
tibiae testaceous (?) 35. oUiquata.
(j^. Superior carina of hind femora distinctly serrate ; inner spines of
hind tibiae nearly twice as long as depth of tibiae . . 3G. jicrturbans.
e-. Wings roseate distally 87. hicittala.
aP'. Antennae of male (those of female always relatively sliorter) not or hardly more
than, often less tlian, one fourth longer than tlie iiead and pronotum together.
l^-. Prozona more or less strangulate (especially in female?), narrower than tlie
head exclusive of the eyes, the metazona somewhat abruptly and not gradually
expanded and bullate.
ci. Prozona transversely rotundato, though feebly and delicately carinate ;
metazona posteriorly rotundato-subrectangulate.
c/i. Anal area of tegmina at broadest one third broader than the interspace
between the eyes ; anal cerci of male apically rounded ; metazona about one
half wider than middle of prozona 38. peregrina.
d-. Anal area of tegmina no broader or scarcely broader than the interspace
between the eyes ; anal cerci of male distinctly emarginate apically, the
lower lobe the longer; metazona about one third wider than middle of
prozona 39. pamnensis.
c^. Prozona distinctly tectate and bluntly carinate ; metazona posteriorly very
obtusangulate and broadly rounded 40. exsul.
b". Prozona not strangulate, no narrower than the head exclusive of the eyes,
the metazona gradually and regularlj- expanding posteriorly to a greater or less
degree, never bullate.
ci. Of large size. Pronotum scarcely or not at all tectate, the median stripe
broad, the posterior margin obtusangulate and rounded ; tegmina distinctly
maculate or obliquely strigate ; male cerci tapering from base to apex.
</i. Tegmina feebly if at all pantherine in markings, the costal area immacu-
late,
fii. Markings of distal half of tegmina composed of longitudinal streaks by
the more or less interrupted infuscation of the longitudinal veins, enforced
by a partial infuscation of the adjoining cross-veins * ... 41. pallens.
* These interruptions, however, often occur at similar intervals on adjoining
veins, and so give rise also to a more or less noticeable transverse arrangement,
but this is less conspicuous than the longitudinal disposition.
SCUDDER. THE GENUS SCHISTOCERCA. 447
c2. Markings of distal half of tegmina composed of fuscous maculations,
generally feeble, arranged in obliquely transverse series at right angles
to the veins (much as in S. americnna), the transverse cross veins at such
points infuscated equally throughout 42. cancellata.
d-. Tegmina distinctly pantherine in markings, the costal area maculate.
43. americana.
(?. Of small size. Pronotum distinctly and strongly tectate, the median stripe
narrow, the posterior margin rectangulate or even acutangulate, at least in the
male ; tegmina immaculate or very feebly maculate ; male cerci of subequal
breadth 44. damnijica.
1. Schistocerca gracilis sp. nov. ;
One of the smallest and slenderest of the genus, fusco-testaceons, ob-
scurely marked with fuscous. Head fusco-testaceous ; frontal costa sub-
equal, a little expanded basally, strongly sulcate excepting above, coarsely
punctate, the margins flavo-testaceous ; eyes much longer than the gente
below them ; antennjB fulvo-testaceous. Pronotum compressed, subequal,
hardly expanding on the metazona, where the width does not exceed that
at eyes, fusco-testaceous with a ferruginous tinge, the metazona somewhat
infuscated on the disk, the lateral lobes immaculate but a little pallescent
centrally ; pi-ozona not tectate, of the same length as the metazona, ante-
riorly produced and well rounded, rather delicately scabrous, the median
carina slight but distinct and slender, more pronounced on the metazona,
which is delicately scabro-punctate, the posterior angle rectangulate,
hardly rounded, Prosternal spine rather slender, rather short, equal,
blunt, a little retrorse. Tegmina slender, much longer than the body,
fusco-testaceous sprinkled, especially in the distal half, with slight and
not very dark fuscous maculations, rather irregularly scattered throughout ;
wings apparently vitreous.* Fore and middle femora slender ; hind femora
rather small, but little surpassing the abdomen, fusco-ferruginous, with a
somewhat hoary outer face, the upper carinas scarcely serrate ; hind tibiae
ferruginous, the spines black tipped. Anal cerci fully twice as long as basal
breadth, tapering slightly, bent a little inward at the middle, the apex
truncate and feebly emarginate, the angles rounded ; subgenital plate a
little upcurved, scaphiform, but tapering regularly as seen from above and
compressed, apically acuminate and fissate half way to the base, the angles
acute.
* The specimen is in too fragile condition to be spread.
448 PROCEEDINGS OP THE AMERICAN ACADEMY.
Length of body, 26 mm.; antennae, 12.5+ mm.; tegmina, 28 mm, ;
hind femora, 16 mm.
1 (J. South America.
2. Schistocerca aurantia sp. nov.
Of moderate size and stoutness, fusco-testaceous somewhat obscured
with fuscous. Head rather large, nearly uniform fusco-testaceous, above
with feeble, more or less divergent, dull fuscous stripes ; frontal costa
subequal, not very broad, j^unctate, deeply sulcate below the ocellus ;
eyes narrow elliptical, very much longer than the infraocular portion of
the genje ; antennoe testaceous, the distal half infuscated. Pronotum well
arched but not tectate, expanding slightly on the metazona so as to be
about as broad as at the eyes, with a feeble median carina, merely indi-
cated on the prozona, the latter produced anteriorly and rounded, a little
shorter than the metazona, scabro-punctate, a little more coarsely than the
metazona, the whole pronotum uniform in color except that the metazona
is more or less ferruginous, posteriorly rectangulate. Prosternal spine
moderate, erect, cylindrical or feebly tapering, blunt. Tegmina consid-
erably longer than the body, moderately slender, testaceous becoming
subvitreous distally, maculate with moderately large quadrate or rounded
fuscous spots, darker proximally than distally and pretty uniformly dis-
tributed in the median area, the costal and anal areas more or less minutely
flecked with fuscous; wings hone3^-infumate, the veins flavous and the
cross-veins fusco-flavous, with a slight sprinkling of fuscous dots apically.
Fore and middle femora not inflated in the male ; hind femora scarcely
surpassing the abdomen, fusco-ferruginous, the outer face whitish, the
carinai with the serrations marked with fuscous ; hind tibia3 fusco-ferru-
ginous verging on puri)lish, the spines luteous, apically black. Male
cerci nearly twice as long as basal breadth, tapering only a little, apically
truncate and broadly emarginate ; subgenital plate narrowly and rather
deeply fissate apically.
Length of body, ^, 31 mm., 9, 46 mm.; antennte, <?, 15.5 mm., 5,
16 mm.; tegmina, S, 33 mm., 9^ 44.5 mm.; hind femora, J, 19 mm.,
9 , 27 mm.
1 (J, 8 9. Mexico, Packard; Yucatan, Schott ; Meriden, Yucatan;
Realejo, Nicaragua, April, McNeil.
The male, from Nicaragua, may not belong here ; the wings are almost
clear hyaline, but the specimen has been long immersed in spirits and is
decolored.
SCUDDER. — THE GENUS SCHISTOCERCA. 449
3. Schistocerca carinata sp. nov.
11 Acridium scuteUare Walk., Cat. Derm. Salt. Brit. Wus., III. 579 (1870). Cf.
No. 25, below.
Slightly above the ordinary size and moderately stout, fusco-testaceous
with an olivaceous tinge. Head rather large, dull testaceous, the promi-
nences iufuscated, with a slender suborbital, genal, obscure fuscous streak
and usually a pair of slender, diverging, fuscous streaks on the vertex;
eyes considerably longer than the infraocidar portion of the gena3, promi-
nent in the male ; antennai flavous or rufous, sometimes infuscated,
especially on distal half. Pronotum well arched and very distinctly
carinate so as almost to appear tectiform, the metazona enlarging so as to
be a little broader than width at eyes, at least in the female, the pos-
terior margin feebly more than rectangulate, the angle very narrowly
rounded ; prozona jjroduced and rounded anteriorly, somewhat shorter
than the metazona, more or less infuscated on the disk, with obscure pallid
quadrate patches on the lateral lobes. Prosternal spine moderate, sub-
conical, erect, blunt. Tegmina much longer than the abdomen, obscure
olivaceo-testaceous, very obscurely cloudy-maculate with fuscous through-
out, substrigate distally; wings vitreous, washed in the faintest manner
with aurantio-fuliginous, nowhere maculate. Fore and middle femora
scarcely enlarged in the male; hind femora ferrugineo-testaceous, the
outer face dull ivory white, the inner portion of the upper face often with
a pair of distant fuscous blotches ; hind tibia; purplish testaceous, the
spines luteous with black tips. Male cerci more than twice as long as
broad, straight, tapering slightly throughout, truncate and very feebly
emargiuate apically; subgenital plate rather small, elongate, apically
acuminate as seen laterally, the apical fissure U-shaped and moderately
deep.
Length of body, ^, 31 mm., 9? 55 mm.; antennre, ^, 14mm.; teg-
mina, (J, 33 mm., 9» 54 mm.; hind femora, ^, 17.5 mm., 9) 30 mm.
1 $, Q 9- ^^^ I^i^gO) Cal., Crotch; Sierra Nola, Mex., Dec. 3-6,
Palmer ; Orizaba, Mex., Jan. (Bruner) ; Vera Cruz, Mex., Heyde
(Bruuer).
4. Schistocerca columbina.
Gri/llus cegtjptius Thunb., Mem. Acad. St. Petersb., V. 247 (1815)t. Stul [misnomer].
Gryllus columhinus Thunb., Loc. cit., IX., 899, 425 (1824).
Acridium (Schistocerca) columhinum Stal, Rec. Orth., I. 67 (1873).
Schistocerca columbina Brunii.-Redt., Proc. Zool. Soc. Lond., 1892, 210 (1892).
Gri/llus occidentalis Thunb., Loc. cit., IX. 400, 429 (1824) t. Stal.
VOL. XXXIV. — 29
450 PROCEEDINGS OF THE ABIERICAN ACADEMY.
This species was originally described from St. Bartholomew in the
"West Indies, and has been i-eported from other islands, — St. Vincent's,
Grenada, Martinique, and Trinidad, as well as from Mexico, Costa Rica,
Nicaragua, Guatemala, Panama, Colombia, Venezuela, Surinam, Brazil,
and Peru. I have seen specimens only from Costa Rica, Underwood
(Bruner) ; San Mateo del Mar, Tehuantepec, Feb., in lagoons, Sumi-
chrast ; and Panama, besides one marked Central America.
5. Schistocerca crocotaria sp. nov.
Slightly above the ordinary size and moderately stout, olivaceo-testa-
ceous, more or less infuscated. Head rather large, olivaceo-testaceous,
with faint diverging fuscous stripes on the vertex, the front rather densely
punctate with pale ferruginous; frontal costa subequal, feebly sulcate
below the ocellus; eyes considerably longer than the infraocular portion of
the genre ; antennse luteous, apically ferruginous. Pronotum well arched,
scarcely subtectate, the metazona enlarging so as to be slightly broader
at its greatest dorsal width than at the eye?, distinctly shouldered later-
ally ; prozona very slightly produced and broadly rounded in front, bluntly
punctate, slightly shorter than the posteriorly rectangulate, rather finely
and sharply punctate metazona ; whole pronotum olivaceo-testaceous,
more or less ferruginous on disk, especially on metazona, obscurely mottled
with ferruginous on lateral lobes, the median carina distinct throughout,
but especially on metazona. Prosternal spine moderately stout, cylindrical,
very blunt, erect or suberect. Tegmina much longer than the body, rather
broad, olivaceo-testaceous, rather obscurely maculate with faint fuscous,
distally in irregularly oblique broken transverse bands ; wings subvitreous,
faintly aurantiate throughout, with very feeble signs of maculation apically
in anterior area. Hind femora reaching tip of abdomen, moderately stout,
flavo-olivaceous on outer, inner, and inner-superior faces on an olivaceous
ground, the hind tibiae flavo-olivaceous with a ferruginous tinge, the spines
luteous with black tips.
Length of body, 54 mm. ; antennjE, 19 mm. ; tegmina, 53 mm. ; hind
femora, 31 mm.
5 9 • Chon tales, Nicaragua ; Realejo, Nicaragua, April, McNeil.
6. Schistocerca interrita sp. nov.
Size and form of the last preceding, ferrugineo-testaceous. Head
moderately large, of the ground color with obscure fuscous markings;
SCUDDER. — THE GENUS SCHISTOCERCA. 451
frontal costa subequal, sulcate throughout ; eyes scarcely longer than the
infraocular portion of thegenoe; antennae ferruginous. Prouotum well
arched, in no way tectate but with distinct and sharp median carina,
ferruginous with feeble fuscous maculations on the lateral lobes (which
have also below the middle an obscure pallid spot), widening considerably
on the metazona, so as to be considerably broader than at the eyes, the
prozona scarcely produced anteriorly, distinctly shorter than the metazona,
which is faintly obtusaugulate behind, the angle narrowly rounded. Pro-
sternal spine moderate, cylindrical, blunt, retrorse but not arcuate.
Tegmina extending much beyond the abdomen, moderately broad, testa-
ceous or ferrugineo-testaceous, with distinct subpantherine fuscous mark-
ings of rather large size extending from base to tijj; wings vitreous, the
veins luteous. Hind femora ferrugineo-testaceous with hoary outer face,
the serrations of the upper carinte fuscous ; hind tibiae pale ferruginous,
the spines with black tips.
Length of body, 45 mm. ; antennae, IT-j- mm. ; tegmina, 50 mm. ; hind
femora, 28 mm.
2 ?. Peru, H. Edwards (Mus. Comp. ZooL, Scudder).
7. Schistocerca camerata sp. nov.
Compact and bulky, somewhat above the medium size, ferrugineo-
testaceous more or less iufuscated. Head ferrugineo-testaceous, mottled
posteriorly with faint plumbeo-fuscous and with a pair of divergent fus-
cous streaks behind upper edge of eyes ; frontal costa subequal, slightly
narrower at the ocellus, deeply sulcate at and below the ocellus ; eyes
narrower than usual, distinctly longer than the infraocular portion of the
genae ; antennte luteo-testaceous. Pronotum distinctly tectate, with very
coarse and jjrominent median carina, which is free from the very obscure
and faint infuscation of the remainder ; prozona scabro-punctate, produced
and subangulate anteriorly, but little shorter than the not very profusely
punctate metazona ; the latter enlarges but little, but is shouldered pos-
teriorly and just broader than at the eyes, the hind margin faintly obtus-
augulate, the angle rather narrowly rounded. Prosternal spine short,
stout, cylindrical, blunt, erect. Tegmina extending but little beyond the
abdomen, rather broad, ferrugineo-testaceous, rather profusely and some-
what obscurely maculate throughout, except the anal area, with fuscous,
the maculations in the distal half having a tendency to an obliquely trans-
verse direction ; wings impure vitreous, with luteous veins. Hind femora
452 PROCEEDINGS OF THE AMERICAN ACADEMY.
dull testaceous, the outer face faintly hoary, the cariuse punctate with
fuscous at the serrations, the genicular arc black ; hind tibiiB dull dark
purplish, the spines luteous with black tips.
Length of body, 49 mm. ; antennas, 14+ mm. ; tegmina, 44 mm. ; liind
femora, 28 mm.
3 9 . Sinaloa, Mex., Koels, Behrens.
8. Schistocerca mellea sp. nor.
Of medium size and moderately stout, fusco-testaccous with a ferrugi-
nous tinge. Head ferrugineo-testaceous, with the margins of the frontal
costa and facial caringe punctate with fuscous, a pair of diverging fuscous
stripes on the vertex and a genal stripe below the eyes ; frontal costa
slightly narrowed just below the ocellus and feebly sulcate, excepting
above ; eyes a little tumid in the male, much longer than the infraocular
portion of the geuse ; antennae ferrugineo-testaceous. Pronotum well
arched transversely, in no way tectate, the median carina faint excepting
on the metazona of the female, where it is slight ; prozona obscurely
ruguloso-punctate, slightly and scarcely angularly produced anteriorly, a
very little shorter than the rather finely punctate metazona, which enlarges
considerably jiosteriorly, so as to be considerably broader than at the eyes,
and is posteriorly rectaugulate, or in the female faintly obtusangulate;
the whole pronotum is ferrugineo-testaceous, the disk, at least in the
female, strigate and blotched with fuscous, leaving clear a broad median
stripe, the lobes irregularly maculate with fuscous, the middle with a
longitudinal fuscous bar, all of which is very obscure in the male. Pro-
sternal spine erect, moderately slender, conico-cylindrical, blunt. Teg-
mina extending considerably beyond the abdomen, not very slender,
ferrugineo-testaceous, profusely and distinctly maculate throughout with
fuscous, in the distal half arranged rather conspicuously in transversely
oblique stripes ; wings honey yellow with a smoky tinge, apically macu-
late in the upper area. Fore and middle femora not enlarged in the male;
hind femora about reaching the tip of the abdomen, ferrugineo-testaceous
with hoary outer face, which is punctate with black, as are also the serra-
tions of the carinas ; hind tibite ferrugineo-testaceous, the spines pallid
with black tips. Male cerci moderately slender, feebly tapering, about
twice as long as basal breadth, apically truncate and feebly and broadly
emarginate ; subgenital plate rather elongate, scaphiform, apically fissate
narrowly half way to the base, the flaps so formed acutangulate, the angles
rounded.
SCUDDER. — THE GENUS SCHISTOCERCA. 453
Length of body, ^, 30 mm., 9, 47 mm.; antennae, ^, 13.5 mm., 9,
12+ mm.; tegmina, <?, 32 mm., 9, 47 mm.; hind femora, ^, 18 mm.,
9 J 25 mm.
1 (J, 19- Vera Cruz, Mexico, Heyde (Bruner).
9. Schistocerca zapoteca sp. nov.
Of moderate size and stoutness, but in these respects with considerable
disparity between the sexes, fusco-testaceous with a slight ferruginous
tinge. Head rather prominent, testaceous, flecked and more or less ob-
scured with fuscous, with a pair of divergent fuscous stripes bordering
the flavo-testaceous median stripe which marks the vertex ; frontal costa
a little contracted at the ocellus, sulcate at and below the same, flavous at
the margins ; eyes prominent in the male, much longer than the infra-
ocular portion of the genaj; autennie flavo-testaceous, in the male a third
as long again as head and pronotum together. Pronotum somewhat
compressed, well arched, hardly tectate, feebly and very bluntly carinate,
the disk ferrugineo-fuscous, sometimes strigate on the metazona, with a
distinct, rather narrow, median testaceous stripe, broader in the female
than in the male, the lateral lobes broader than deep, testaceous, mottled
or obscured or occasionally vittate with fuscous ; jorozona produced ante-
riorly and rather strongly rounded, a little shorter than the metazona,
which is postei'iorly rectangulate with narrowly rounded angle, and ex-
pands but little even in the female so as hardly to exceed the width at the
eyes. Prosternal spine slender, feebly tapering, blunt, erect. Tegmina
extending far beyond the abdomen, moderately slender, ferrugineo-testa-
ceous, rather profusely but feebly maculate with fuscous, mainly by the
infuscation of cross-veins, occasionally disposed in obliquely transverse
stripes on the distal half ; wings vitreous or very faintly infumate apically,
occasionally faintly maculate apically in the anterior area, the veins luteo-
ferruginous. Fore and middle femora scarcely thickened in the male;
hind femora rather slender, reaching beyond the abdomen, luteo-testaceous,
flecked and punctate with fuscous, often tinged more or less with ferrugi-
nous, especially on the distal half, occasionally flavescent basally ; hind
tibias dark dull purple, the spines luteous with black tips. Male cerci
twice as long as broad, equal, straight, apically truncate and mesially
emarginate, the angles rounded ; subgenital plate tapering, scaphiform,
apically compressed, deeply cleft, the fissure closed.
Length of body, $, 28 mm.. 9, 40 mm.; antenna?, ^, 13 mm., 9,
14.5 mm. ; tegmina, ^, 29 mm., 9, 43 mm. ; hind femora, ^, 18 mm.,
9 , 2G mm.
454 PROCEEDINGS OP THE AMERICAN ACADEMY.
21 ^, 13 ?• Venis Mecas, Mex., Jan. G, Palmer; Mexico, April,
Sumichrast, Botteri ; Guatemala, Van Patten ; Costa Rica, Underwood
(Bruner) ; South America.
10. Schistocerca vaga.
Acridium vcnjum Scudd.!, Proc. Bost. Soc. Nat. Hist., XVIII. 269 (1876).
Schistocerca vaga Brun., Proc. U. S. Nat. Mus., XII. 187 (1890).
I have received this species from CaliJ^ornia, H. Edwards ; Fresno,
April 27 (Stanf, Univ.), Pasadena, June (Stanf. Univ.), Los Angeles,
March (Bruner), South Santa Monica, July 30, Morse, Colton, July 17,
Morse, San Bernardino, July 15, Morse, Palm Springs, July 12, Morse,
and San Diego, Cal., Edwards, Crotch, and Mohave Desert, Cal. ; Ft.
Whipple, Palmer, and Yuma, Ariz., July 7, Morse; Mesilla, N. Mex.,
Cockerell, Oct. 18, Morse; San Antonio, Sept. 18-27, Palmer, Uvalde,
July, Palmer, and El Paso, Tex., Aug., Dunn (Bruner); Guadalupe
Isl., off Lower California, Palmer; Cape St. Lucas, Lower Cal., Xantiis ;
Matamoras,Tamaulipas, Couch, Uliler; San Pedro, May 20, Palmer,
and Montelovez, Coahuila, Sept. 20, Palmer ; Sonora, Schott ; Bledos,
Mex., Oct. 1, Palmer; Sierra Nola, Mex., Dec. 3-6, Palmer; Mexico
City, Palmer; Jalapa, Mex., June 22 (Bruner), Durango, Mex., Palmer,
and Jalasco, Max., Berendt; and Realejo, Nicaragua, April, McNeil.
11. Schistocerca simulatrix.
? Ci/rtacanthacn's simulatrix Walk., Cat. Derm. Salt. Brit. Mus., IV. 010 (1870).
? Acridium simulatrix [sic] Thom., Rep. U. S. Geol. Surv. Terr., V. 230 (1873).
Originally described from San Domingo, I have a specimen which
appears to belong here, which comes from Inagua, Bahamas.
12. Schistocerca pyramidata sp. nov.
Of medium size and stoutness, fusco-testaceous. Head rather large,
pale testaceous, the face generally much infuscated, especially on the
prominent parts, with a ferruginous tinge, posteriorly striped with fuscous
and the vertex with a pair of diverging fuscous stripes running from the
front of the fastigium or even the median ocellus backward, leaving be-
tween them a broad clear luteo-testaceous median band; frontal costa
subequal, deeply sulcate at and below the ocellus ; eyes prominent in the
male, much, in the male very much, longer than the infraocular portion
of the gense ; antennas considerably more than a third longer than the
SCUDDER, THE GENUS SCHISTOCERCA. 455
head and iironotum together in the male, luteous. Pronotura well arched,
in no way tectate, but with a delicate percurrent median carina, in the
middle of a regularly narrowing but percurrent luteous or ferrugineo-
luteous median stripe, the rest of the disk fuscous, often longitudinally
strigate with testaceous on the metazoua ; lateral lobes ferrugineo-testa-
ceous, much mottled and sometimes longitudinally strigate with fuscous, a
paler spot generally appearing on the upper part of the prozona ; prozona
a little jiroduced anteriorly and rounded, a little shorter than the meta-
zoua, which is posteriorly rectangulate, the angle rounded not very nai"-
rowly. Prosternal spine rather slender, slightly tapering, blunt, straight,
slightly inclined. Tegmina extending considerably beyond the abdomen,
moderately slender, testaceous, in the male nearly immaculate, in the
female distinctly but not very profusely maculate with fuscous throughout
the median area, the maculations of the distal half small and obscurely
arranged in obliquely transverse lines; wings faintly tiuged, especially in
the anal area, with pale citron, apically very faintly fuliginous, nowhere
maculate. Fore and middle femora not enlarged in the male; hind
femora attaining (9) or surpassing (^) the end of the abdomen, dull
testaceous with the outer face dull ivory white, punctate with fuscous
along the carinas ; hind tibi;e dull jjurplish testaceous, the spines luteous
with black tips. Male cerci of subequal breadth, about twice as long as
basal breadth, apically a little obliquely truncate and considerably emar-
ginate mesially, the lobes thus formed rounded, the lower somewhat the
longer ; subgenital plate slender, elongate, scaphiform, apically very deeply
fissate, the acute angles only slightly rounded.
Length of body, ^, 37 mm., 9,53 mm.; antennae, ^, 15.5 mm., 9,
18 mm.; tegmina, ^, 36 mm., 9, 53 mm.; hind femora, J", 20 mm., 9?
31.5 mm.
2 c?, 5 9. Cuernavaca, Mexico, May, Sept., Barrett (Morse).
13. Schistocerca desiliens sp. nov.
Of moderately large size and moderate stoutness with considerable dis-
parity between the sexes, ferrugineo-testaceous, considerably infuscated.
Head not very large, ferrugineo-testaceous, testaceous posteriorly, all the
carinas marked with fuscous, and a pair of diverging fuscous stripes on
the vertex, enclosing a broad median luteo-testaceous band, as in the last
species; frontal costa subequal, deeply sulcate at and below the ocellus;
eyes somewhat prominent, especially in the male, distinctly (j) or
scarcely ( 9 ) longer than the infraocular portion of the genoe ; antennae
456 PROCEEDINGS OF THE AMERICAN ACADEMY.
more than lialf as long again as the head and pronotum together in the
male, luteous, faintly infuscated apically. Pronotum well arched, in no
way tectate, but with a delicate percurrent median carina in the middle
of a gradually diminishing but percurrent luteo-testaceous median stripe,
the remainder of the disk ferrugineo-fuscous, the lateral lobes the same
with a strongly oblique anterior and inferior luteo-testaceous patch, edged
above with fuscous fading superiorly ; prozona roundly produced anteriorly,
slightly shorter than the metazona, which is very slightly broader than
at the eyes, posteriorly rectangulate (^) or faintly obtusangulate (9),
the angle narrowly rounded. Prosternal spine moderate, cylindrical,
blunt, erect or suberect. Tegmina extending well beyond the abdomen,
slender, with a pallid luteous streak basally in the costal area, in the male
otherwise immaculate or nearly so, in the female sparsely and rather
feebly maculate in the median area and particularly along the median line,
all the maculations small ; wings vitreous, with an exceedingly feeble
infumation, most distinct apically. Fore and middle femora not enlarged
in the male ; hind femora ferrugineo-testaceous, with hoary outer face
and the carinas punctate with fuscous ; hind tibia3 ferrugineo-testaceous,
the spines luteous with black tips. Male cerci very small, straight, hardly
twice as long as broad, ta})ering to a blunt rounded tip ; subgenital plate
slender, haustrate, somewhat compressed, subacuminate, apically narrowly
and not very deeply fissate.
Length of body, <J , 31 mm., 9-1 55 mm. ; antennre, $, 14.5 mm., 9>
18 mm. ; tegmina, (^ , 31 mm., 9 5^0 mm. ; hind femora, ^ , 11 mm., 9>
32 mm.
1 (? , 4 9. Rio de Janeiro, Brazil, Nov. (Mus. Comp. Zool.) ; Vic-
toria, Brazil, May (Bruner).
I know of no species of Schistocerca in which the male cerci are so
narrow at apex as here ; the next species is the most closely allied in
that respect.
14. Schistocerca flavofasciata.
Acri/diumjfnvqfasciaium DeGeer, Mem., III. 488, pi. 40, fig. 8 (1773).
Acrldium {Schistocerca) flarofasciatum Stal, Rec. Orth., I. 67 (1873).
Gryllus nitens Thunb., Mem. Acad. St. Petersb., V. 236 (1815) t. Stal.
Gri/Ihisjimbrlntiis Thunb., Loc. cit., IX. 428 (1824) t. Stal.
Gnj/lus livichis Thunb., Loc. cit., IX. 428 (1824) t. StSl.
Acrldium hmrjipenne Burm., Handb. Ent., II. 632 (1838) t. Stal.
The only specimens I have seen are from Rio de Janeiro Nov. (Mus.
Comp. Zool.), and Corumba, Brazil, March, April (Mus. Comp. Zool.).
It was originally described from Brazil.
SCUDDER. — THE GENUS SCHISTOCERCA. 457
15. Schistocerca infumata sp. nov.
Of large size and moderate stoutness, dark olivaceo-fuscous. Head
not very large, olivaceo-fuscous, with a subocular genal fuscous streak,
and a pair of divergent fuscous streaks on the vertex, enclosing a median
luteo-testaceous stripe ; frontal costa subequal, deeply sulcate at and below
the ocellus ; eyes a little prominent in the male, distinctly ((^ ) or scarcely
( 9 ) longer than the infraocular portion of the geuoe ; antennae barely a
third longer than the head and prouotum together in the male, luteous
or luteo-testaceous. Pronotum well arched but feebly subtectate, with
distinct percurrent median carina in the middle of a posteriorly attenu-
ating luteo-testaceous median stripe edged with fuscous, the rest of the
pronotum dark olivaceo-fuscous with an obliciue inferior fuscous cloud or
stripe on the lateral lobes, the whole prouotum gradually enlarging pos-
teriorly, so that the metazona is considerably (9) or a little ( J" ) wider
than at the eyes; prozona angularly produced anteriorly, the angle rather
narrowly rounded, but little shorter than the metazona, which is posteri-
orly rectangulate d^) or faintly obtusangulate (9)? the angle not very
narrowly rounded. Prosternal spine slight, cylindrical, blunt, erect.
Tegmina extending far beyond the abdomen, moderately slender, olivaceo-
testaceous, immaculate ; wings distinctly infumate throughout. Fore and
middle femora not enlarged in the male ; hind femora dark testaceous, the
outer face and genicular lobe ivory white, the carinoe punctate with fus-
cous, occasionally the other faces more or less hoary ; hind tibife dull
ferruginous, the spines luteous (or in the female the outer spines luteous,
the inner ferruginous) with black tips. Male cerci very small for this
genus, half as long again as broad, the lower margin straight, the upper
rounded, and tapering so as to be at the truncate apex hardly a third as
broad as at broadest. Subgenital plate straight, long, scaphiform, hardly
compressed, apically deeply and still more widely emarginate to form a
U-shaped fissure with diverging sides, the angles subacuminate. Whole
body and legs pilose.
Length of body, ^ , 33.5 mm., 9, 58 ram. ; antennae, ^, 14 mm., 9,
18 mm. ; tegmina, ^ , 38 mm., 9 , 56 mm. ; hind femora, ^ , 21.5 mm.,
9 32 mm.
7 (J , 3 9 . Montevideo, Uruguay, Meyer-Diir ; Brazil, Janson.
458 PROCEEDINGS OF THE AMERICAN ACADEMY.
10. Schistocerca sequalis sp. nov.
Of fair size and moderately slender, fusco-testaceous. Head moderately
large, fusco-testaceous, the vertex fuscous except a broad median ferru-
gineo-luteous band ; frontal costa subequal, sulcate below the ocellus ;
eyes rather prominent, very much longer than the iufraocular portion of
the geuse ; antennae at least a third longer than the head and pronotum
together, ferrugineo-luteous. Prozona well arched, in no way tectate,
but with distinct and delicate percurrent median carina, in the middle of
a rather broad and equal percurrent ferrugineo-luteous stripe, the rest of
the disk fuscous, the lateral lobes fusco-testaceous, with an oblique inferior
fuscous stripe, below which they are testaceous ; prozona roundly produced
anteriorly, about as long as the metazona, which enlarges but slightly and
is narrower than at the eyes and jjosteriorly rectangulate or perhaps faintly
obtusangulate, the angle rounded. Prosternal spine rather long, erect,
cylindrical, bluntly tapering apically. Tegmina extending well beyond
the abdomen, slender, immaculate, ferrugineo-testaceous, the costal and
anal areas testaceous ; wings very faintly infumate with a slight citron
tinge. Fore and middle femora not enlarged ; hind femora rather slen-
der, slightly surpassing the abdomen, ferrugineo-testaceous, the outer face
hoary, feebly punctate with fuscous on the carince; hind tibiae dull pur-
plish or ferruginous, the spines luteous with black tips. Male cerci sub-
equal, about half as long again as broad, inbent at the middle, apically
angularly emarginate above the middle, the lower lobe projecting ; sub-
genital plate short scaphiform, almost haustrate, apically U-shaped, the
emargination deeper than broad, and the angles subacute.
Length of body, 35 mm.; anteunje, 13.25-1- mm.; tegmina, 36 mm.;
hind femora, 19 mm.
2 (J . Demerara, British Guiana.
17. Schistocerca maya sp. nov.
Below the medium size and moderately slender, testaceous, more or less
infuscated. Head moderately large, ferrugineo- or luteo-testaceous, the
vertex with a pair of fuscous stripes bordering a ferrugineo-luteous median
band ; frontal costa subequal, sulcate at and below the ocellus ; eyes
moderately prominent, much longer than tlie infraocular portion of the
genfB ; antennae more than a third longer than the head and pronotum
together, ferruginous. Pronotum well arched, in no way tectate, with a
delicate median carina in a (sometimes obscured) ferrugineo-luteous dorsal
SCUDDER. — THE GENUS SCHISTOCERCA. 459
stripe, the rest of the disk fuscous, the hxteral lobes ferrugineo-testaceous,
uumarked, or at most feebly clouded with fuscous ; prozona produced and
atigulato-rotuudate in front, as long as the metazona, which enlarges but
little, is narrower than at the eyes and posteriorly rectangulate and nar-
rowly rounded. Prosternal spine moderate, conical, blunt, feebly retrorse.
Tegmina much surpassing the abdomen, immaculate or very faintly and
most obscurely clouded with fuscous, ferrugineo-testaceous, the anal area
lutescent ; wings vitreous with luteous veins. Fore and middle femora
not enlarged ; hind femora ferrugineo-testaceous, with hoary outer face ;
hind tibine ferruginous, the spines luteous with black tips. Male cerei
subec^ual, fully half as long again as broad, a little obliquely truncate
apically, the lower posterior angle distinctly produced and rounded ; sub-
genital plate scaphiform, subacuminate, apically fissate to the base, the
fissure closed.
Length of body, 31 mm. ; autennte, 13 mm. ; tegmina, 28.5 mm. ; hind
femora, 18 mm,
3 ^. Venis Mecas, Mexico, Jan. 6, Palmer; San Mateo del Mar,
Tehuantepec, in lagoons, Feb., Sumichrast.
The description is based mainly on the Mexican specimen, and the
others may possibly not belong here. All have been immersed in alcohol.
18. Schistocerca australis nom. nov.
Acridium occidentale Scudd.!, Proc. Bost. Soc. Nat. Hist., XII. 3.30 (1869).
Acridium {Schistocerca) occidentale Scudd., Loc. cit., XVII. 274 (1875).
The name is here changed, as the name occidcntalis was given by
Thunberg to another species of Schistocerca, placed by liim in Gryllus ;
see No. 4, above.
I have specimens before me from Rio de Janeiro, U. S, Expl, Exp.,
Thayer Exp., Mrs. Davis (Mus. Comp, Zool.) ; Brazil, Linden (Mus,
Comp. Zool.); Santarem and Cudais, Brazil, Thayer Exp. (Mus. Comp.
Zool.) ; Paramaribo, Dutch Guiana, Richardson (Mus. Comp. Zool.) ; and
the Napo or Maranon River, eastern Peru.
19. Schistocerca gulosa sp. nov.
Large and bulky, ferrugineo-testaceous. Head large, ferrugineo-
testaceous, obscurely marked with fuscons and especially with a pair of
very divergent stripes on the vertex ; frontal costa subequal, faintly
broadening below, sulcate below the ocellus ; eyes much longer than the
460 PROCEEDINGS OP THE AMERICAN ACADEMY.
•
iufraocular portion of the gense ; antennte rufous. Pronotum well arched,
in no way tectate, but with a distinct though slight median carina, ferru-
gineo-testaceous, the front edge narrowly fuscous, the disk with exceed-
ingly obscure infuscation, leaving a clear dorsal stripe scarcely perceptible,
the lateral lobes obscurely mottled with fuscous and with a mesial longi-
tudinal fuscous stripe ; prozona produced and rounded anteriorly, a little
shorter than the metazona, which broadens considerably behind so as to
be very much broader than at the eyes, and is obtusangulate, the angle a
little rounded. Prosternal spine rather small, erect, subconical, blunt.
Tegmiua extending far beyond the abdomen, moderately broad, ferru-
gineo-testaceous, fiiintly feebly and sparsely maculate with fuscous in
one- or two-celled jiatches ; wings vitreous with a faint citron hue basally,
all the veins luteous or ferruginous. Hind femora ferrugineo-testaceous,
the outer face ivory white, the carinaj punctate with fuscous ; hind tibiai
dull purplish, the spines luteous with black tips.
Length of body, 52 mm. ; tegmina, 51 mm. ; hind femora, 28.5 mm.
1 9 . Demerara, British Guiana.
20. Schistocerca bogotensis sp. nov.
Below the average size and not very stout, ferrugineo-testaceous, much
infuscated. Head moderately large, ferrugineo-testaceous, the frontal
costa much infuscated, especially at the margins, whence a fuscous stripe
proceeds on either side backward across the vertex, leaving a broad sub-
ferruginous median band; frontal costa subequal, sulcate below the
ocellus ; eyes considerably longer than the iufraocular portion of the genae.
Pronotum well arched, in no way tectate, but with a sharp though slight
median carina in the middle of a broad subequal, but on the metazona
slightly attenuated, subferruginous stripe, the rest of the disk deeply in-
fuscated, the lateral lobes ferrugineo-testaceous, blotched with fuscous and
testaceous, but not longitudinally striped ; prozona a little shorter than
the metazona, a little and roundly produced anteriorly, the metazona but
little enlarged and hardly exceeding the width at the eyes, posteriorly
obtusangulate, the angle broadly rounded. Prosternal spine rather long,
erect, subconical, blunt. Tegmina extending well beyond the abdomen,
moderately broad, testaceous, maculate with fuscous in the median area,
the maculations in distal half obscurely massed in broad oblique bands;
wings citriuo-infumate, without maculations. Hind femora ferrugineo-
testaceous, the outer face dull ivory white ; hind tibiie vinous, the spines
vinous with black tijDS.
SCUDDER. — THE GENUS SCHISTOCERCA. 461
Length of body, 49 mm. ; tegmiua, 46.5 mm. ; hiud femora, 26 mm.
2 9 • Bogota, Colombia.
21. Schistocerca inscripta.
CijrtacantJiacris inscripta "Walk., Cat. Derm. Salt. Brit. Mus., III. 550 (1870).
Acridium inscriptum Thorn., Rep. U, S. Geol. Surv. Terr., V. 228 (1873).
Originally described from Jamaica. I have seen a single specimen
from Mandeville, Jamaica, April, Cockerell (Bruner).
22. Schistocerca idonea sp. nov.
Of fully average size and moderately slender, ferrugineo-testaceous,
much infuscated. Head rather large, luteo-testaceous, all the prominences
marked with fuscous, besides a distinct suborbital genal fuscous stripe,
and on the vertex a pair of divergent fuscous stripes on either side of a
broad testaceous median stripe, sometimes tinged with ferruginous ; frontal
costa broadly and rather shallowly sulcate, subequal ; eyes shorter than
the infraocular portion of the genai ; antennte rufous. Pronotum well
arclied, in no way tectate, very faintly strangulate, the median carina
delicate, percurrent, lying in the middle of a very broad and equal ferru-
gineo-testaceous stripe, bordered on either side by a slightly broader
posteriorly widening fuscous or ferrugineo-fuscous stripe, occupying the
rest of the disk ; lateral lobes testaceous with a very broad median slightly
oblique longitudinal fuscous stripe, often itself with a median testaceous
thread ; prozona strongly produced and well rounded in front, slightly
shorter than the metazona, which broadens so as to be slightly broader
than at the eyes, and is posteriorly obtusangulate, the angle generally
very broadly rounded but variable. Tegmina extending far beyond the
abdomen, rather slender, testaceous, the costal area with a long luteous
streak, the anal area wholly luteous or luteo-testaceous, the median area
profusely maculate with fuscous, more or less blended in the proximal
half, scattered and more feeble and generally subquadrate in the distal
half; wings rather faintly flavo-infumate, immaculate. Plind femora
testaceous, the outer face hoary or lutescent below, infuscated along the
middle and generally above, the carinas punctate with fuscous at the ser-
rations ; hind tibiae ferruginous, the spines luteous with black tips.
Length of body, 46 mm. ; antennae, 16.5 mm. ; tegmiua, 41 mm.; hind
femora, 22.5 mm.
3 9. Crapada, Brazil, July, Aug. (Mus. Comp. Zool.).
462 PROCEEDINGS OP THE AMERICAN ACADEMY.
23. Schistocerca literosa.
Acridium Uterosuni Walk./Cat. Derm. Salt. Brit. Mus., IV. 620 (1870) ; Butl, Troc.
Zo(j1. Soc. Lond., 1877, 88 (1877).
Schistocerca literosa Scudd., Bull. Mus. Comp. ZotU., XXV. 15, pi. 2, figs. 1, 3 (1893).
This species is known only from the Gahipagos, and has been found
on Chatham, Hood, Tower, and Charles Islands, the forms occurring on
each island, or on all but the last, being distinct enough to be regarded
as races, as I have pointed out in the paper above cited, where nine points
of distinction are tabulated.
24. Schistocerca melanocera.
Acridium melanoccrum Stal, Eug. Resa, Ins., Ortb., 326 (1860).
Acridium {Schistocerca) melanocerum Stal, Bee. Orth., I. 65 (1873).
Schistocerca vielanocera Brun., Proc. U. S. Nat. Mus., XII. 193 (1889); Scudd.,
Bull. Mus. Comp. Zoiil., XXV. 11, pi. 2, figs. 5, 6 (1893).
Acridium tibiale Walk., Cat. Derm. Salt. Brit. Mus., III. 582 (1870) t. Walker.
This is known to me only from the Galapagos Archipelago, where ii
has been found on Charles, Albemarle, Indefatigable, Chatham, Jervis,
Barrington, James, and Duncan Islands; but it is also credited by Walker
to the '* west coast of America." In my paper on the Orthoptera of the
Galapagos, quoted above, I have discussed at length the distinct types
which appear to be forming ou the different islands.
25. Schistocerca rubiginosa.
Acridium nd)iginosum Harr.!, MS., Scudd., Bost. Journ. Nat. Hist., VII. 467 (1862).
Schistocerca rubiginosa Morse, Psyche, VII. 105 (1894).
■?■? Acridium scutellare Walk., Cat. Derm. Salt. Brit. Mus., III. 579 (1870). Cf.
No. 3, above.
This insect is found along the entire Atlantic coast of the United States
from central Massachusetts to Key West, Florida, and in the interior,
east of the Great Plains, from as far north as Iowa and Minnesota to the
Gulf, and it extends into Mexico and even farther south.
My specimens come from Massachusetts, Sanborn ; Wellesley, Aug. 8,
Sept. 24, Oct. 10-11 (Morse), Dedham, Aug., Maynard (Morse), Prov-
incetown, Sept. 4-6 (Morse), and Hyannis, Mass., Scudder ; Kingston
and Wickford, R. I., Aug. 29 (Morse); Connecticut, Uhler, Norton;
Thompson, Aug. 6, 9 (Morse), Deep River, Aug. 24 (Morse), North
SCUDDER. — THE GENQS SCHISTOCERCA. 4G3
Haven, Aug. 23 (Morse), New Haven, Aug. 29 (Morse), Smith, Stam-
ford, Aug. 13-17 (Morse), and Greenwich, Coun., Aug. 27 (Morse);
Long Island, Sept. ; Sparkill, N. Y., Baird ; Maryland, Uhler ; Middle
States, Osten Sackea ; Jefferson, Iowa, Sept. 20, Allen; District of
Columbia (Bruner); Virginia, Oct. (Bruner); Smithville, Nov. 21,
Dingo Bluff, Nov. 15, Parker, Maynard, and Newbern, N. C. ; Georgia,
Morrison ; Florida, AVurdemann ; Biscayne Bay, Palmer, and Key West,
Fla., Morrison, Palmer, Maynard ; Texas, Aug. 19, Belfrage, Lincecum,
and Dallas, Tex., Boll; Inagua, Bahamas; Mexico, Schaum ; Yucatan,
Schott ; and Guatemala, Van Patten.
It has also been reported from Staten Island, Davis ; New Jersey,
Smith ; Kentucky, Garman ; Illinois, McNeill ; Mmnesota, Lugger ;
and Nebraska, Bruner.
26. Schistocerca sonorensis sp. nov.
Of medium size and not very stout, testaceous. Head testaceous with
none but the most obscure markings ; frontal costa subequal, deeply sul-
cate below the ocellus ; eyes very much longer than the intraocular portion
of the genjB ; autennge a third longer than the head and pronotum to-
gether, luteous. Pronotum distinctly subtectate with an excessively blunt
median carina, testaceous, without markings except a slight indication of
a quadrate fuscous patch on the lateral lobes ; prozona slightly produced
and rounded anteriorly, a very little shorter than the metazona, bluntly
rugulose, the metazona punctate but scarcely rugulose, broadening poste-
riorly so as to be a very little wider than at the eyes, posteriorly obtus-
angulate, the angle narrowly rounded. Prosternal spine luoderate,
cylindrical, erect, blunt. Tegmina extending far beyond the abdomen,
slender, testaceous, the median area profusely but obscurely maculate,
the maculations in the distal half mostly elongate ; wings hyaline, with
the very faintest possible apical infumation, immaculate. Fore and
middle femora not enlarged ; hind femora testaceous, the outer face hoary,
the inner with feeble fuscous clouds as the basis of fasciation, the carinas
punctate with fuscous on the proximal half ; hind tibise testaceous, the
spines luteous with black tips. Male cerci feebly incurved, tapering
gently by the slope of the upper margin, nearly twice as long as middle
breadth, apically truncate and minutely emarginate ; subgenital plate
rather long and slender, scaphiforra, apically acuminate on a side view
due to the slope of the inferior margin, hardly compressed, apically deeply
fissate, the fissure closed.
464 PROCEEDINGS OF THE AMERICAN ACADEMY.
Length of body, 35 mm.; antennjE, 13.5 mm.; tegmina, 38.5 mm.;
hiud femora, 20.5 mm.
2 ^ . Sonora, Mexico, Scliott.
The specimens have been long immersed in spirits.
27. Schistocerca alutacea.
Acridium alutaceum Harr.!, Ins. Inj. Veg., 139 (1841).
Cijriacanthacris alutacea Walk., Cat. Derm. Salt. Brit. Mus., IV. 609 (1870).
Schistocerca alutacea Brun., Publ. Nebr. Acad. Sc, III. 2G (1893).
Acridium emarginatum Uhl.!, MS., Dodge, Can. Ent., IV. 15 (1871).
This insect has much the same distribution as S. ruhiginosa, but is a
little less extended on the Atlantic coast, reacliing only from extreme
southern Massachusetts to northern Florida. It is not only more common
than that species at the west but has a wider range there, extending in
the north to Montana, Utah, and Nevada, and in the south to New
Mexico and even southern California, while it also occurs in northern
Mexico.
Specimens at hand come from "West Chop, Martha's Vineyard, Mass.
(Morse) ; Farmington, Norton, Deep River, Aug. 24 (Morse), New
Haven, Smith, North Haven, Aug. 23 (Morse), and Stamford, Conn.,
Aug. 10-22 (Morse); Long Island; Middle States, Osten Sacken ;
Maryland, July 11, Uhler; Newbern, N. C. ; Georgia, Morrison, Oem-
ler; Florida, Uhler, and Jacksonville, Fla., Priddey (Bruner) ; Indiana,
Oct. 2, Blatchley (Morse, Scudder) ; Illinois and southern Illinois, Uhler;
Colona, III., Aug. 12, McNeill; Minnesota, Bruner; Dallas Co., Aug.
20-24, Allen, and Jefferson, Iowa, Sept. 20, 26, Allen ; Nebraska, Dodge ;
Sidney, Nebr. (Bruner) ; Valley of the Platte, Hayden ; southern Black
Hills, Austin ; Upper Missouri River, Hayden ; Colorado, Baker (Morse),
Morrison ; Pueblo, Aug. 30-31, Scudder, Denver, Scudder, and Mani-
tou. Col., Aug. 24-25, Scudder; Texas, Pope, Belfrage, Aug. 19, Oct.
13 ; southwestern Texas, Schaupp (Bruner) ; San Antonio, Sept. 18-27,
Palmer and Dallas, Tex., Boll; Spring Lake Villa, Utah Co., Utah,
Aug. 1-4, Palmer; Reno, Nevada; Julian, San Diego Co., Cal., Palmer;
Mesilla, N. Mex., June 30, Morse ; and Sierra Nola, Mex., Dec. 3-6,
Palmer.
It has further been reported from Staten Island, Davis ; New York,
Beutenmiiller ; New Jersey, Smith ; Kentucky, Garman ; and Kansas,
Bruner.
SCUDDER. — THE GENUS SCHISTOCERCA. 465
28. Schistocerca obscura.
Grylhis ohscurus Fabr., Suppl. Etit. Syst, 194 (1798).
Acridiam ohscuruvi Burni.!, Haiidb. Ent., II. 632 (1838).
Acridium olivaceum Serv., Ortli., 666 (1839).
This species has a more southern range than the preceding, from which
it is witli difficulty distinguished, not being known on the Atlantic coast
north of North Carolina, though in the west it occurs as far north as
Nebraska and even Iowa. It is not known west of eastern Colorado,
except in the south, where it occurs in New Mexico ; and it is found
throughout Mexico.
I have specimens before me from North Carolina, Holder, Uhler, Shute ;
Dingo Bluff, N. C, Nov. 15, Parker, Maynard; South Carolina, Oem-
ler; Georgia, Morrison; Morris Isl., Geo., Akhurst; Florida, Uhler;
Biscayne Bay, Palmer, Green Cove Springs, Boardman, and Cedar Keys,
Fla., Palmer; Jefferson, Sept. 20, Allen, and Dallas Co., Iowa, Aug.
1-10, 20-23, Allen; Sidney (Bruner), and Platte River, Nebr., Hayden;
Pueblo, Aug. 30-31, Scudder, and Manitou, Col., Aug. 24-25, Scudder;
Texas, Lincecum, Belfrage; Dallas, Boll, Bosque Co., Oct. 23, Belfrage,
Eagle Pass, Schott, and Carrizo Springs, Tex., Wadgymar (Bruner);
White Sands, 30 m. south of Tularosa, Dona Ana Co., N. Mex. 3600',
Aug. 25, Wooton (Morse); Mexico, Uhler; Matamoras, Tamaulipas,
Couch, Uhler; Montelovez, Coahuila, Sept. 20, Palmer; Sonora, Schott;
Venis Mecas, Mexico, Palmer; Tepic, Mex., Cal. Acad. Sc. (Bruner);
and Vera Cruz, Mex., Heyde (Bruner).
I have examined Burmeister's species in the Halle Museum, and this
is also the species so named in the Berlin Museum.
29. Schistocerca lineata sp. nov.
Of large size and robust form, pilose, flavo-testaceous, marked with
fuscous. Head moderate, flavo-testaceous, often marked with fuscous
on the prominent parts, with a distinct, suborbital, genal streak of fuscous,
and the whole vertex more or less infuscated except for a median flavo-
testaceous stripe; frontal costa subequal, feebly sulcate; eyes somewhat
prominent in the male, distinctly longer than ((^) or of about the same
length as (9) the infraocular portion of the genos ; antenufe about half
as long again as the head and pronotum together in the male, flavous.
Pronotum feebly tectate, with a slight percurrent median carina, the disk
much, generally deeply, infuscated, with a not very broad median flavo-
VOL. XXXIV. — 30
466 PEOCEEDINGS OF THE AMERICAN ACADEMY.
testaceous stripe, the lateral lobes flavo-testaceous, more or less suffused
with fuscous, sometimes much iufuscated, rarely blotched with fuscous;
prozona slightly and roundly produced anteriorly, about as long as the
metazona, which is distinctly tumid dorsally in the female, and sometimes
in the male, expanding also so as to be about as wide as ((^) or consid-
erably wider than ( 9 ) the breadth at the eyes, posteriorly obtusangulate,
the angle broadly rounded. Prosternal spine moderate, erect, bluntly
acuminate. Tegmina extending somewhat beyond the abdomen, moder-
ately broad, testaceous or flavo-testaceous, the median area generally
heavily iufuscated next the flavous anal area, but otherwise immaculate
excepting for sometimes, especially in the female, the faintest signs of
quadrate maculatious transversely arranged in the distal half ; wings faint
flavous with luteous veins. Fore and middle femora considerably tumid
in the male; hind femora flavous, flavo-testaceous or testaceous, generally
conspicuously trifasciate with fuscous, the fasciations more or less broken
and occasionally considerably reduced; hind tibite purplish or ferrugi-
nous, the spines luteous with black tips. Male cerci less than half as
long again as broad, bent inward at the middle, tapering very little, api-
cally deeply and obliquely emarginate, the lower lobe the longer, more
extended and more broadly rounded; subgenital plate short, haustrate
rather than scaphiform, the apex deeply cleft in U -shape, the cleft more
than twice as deep as broad, the margins parallel, the angles well
rounded.
Length of body, ^, 47 mm., $, 59 mm.: antennae, ^, 20 mm., 9>
21.5 mm.; tegmina, t;^ , 38.5 mm., $, 5G mm.; hind femora, ^, 23
mm., $ , 33 mm.
2 (?, 10 9. Barber Co., Kans., Cragin (Bruner) ; Texas, Lincecum,
Belfrage; San Antonio, Tex., Newell (Bruner); Gulf coast of Texas,
Aaron; Montelovez, Coahuila, Mex., Sept. 20, Palmer.
SO. Schistocerca albolineata.
Acridlutn alboUneatum Tliom., Rep. U. S. Geol. Surv. West 100 Mer., V. 897, pi. 43,
fig. 1 (1875).
Specimens at hand come from Ames, Iowa (Bruner) ; Camas Pt.,
Idaho (Bruner); Mesilla, June 30, Morse, and Las Cruces, N. Mex.,
Aug. 19 (Bruner); and Grand Canon, Ariz., July 5 (Bruner). It was
originally described as probably from Arizona.
SCUDDER. — THE GENUS SCHISTOCERCA. 467
31. Schistocerca venusta sp. nov.
Of fully mediura size, moderately slender, olivaceous marked with fla-
vous and more or less int'uscated. Head not very large, flavo-olivaceous,
with a greenish fuscous suborbital genal streak, and a pair of similar
divergent stripes on the vertex, darker in color in front of than behind
the eyes, bordering a broad median flavous stripe; frontal costa subequal,
moderately sulcate below the ocellus; eyes somewhat prominent, espe-
cially in the male, distinctly longer than the itifraocular portion of the
gense ; antenna) more than half as long again as tlie head and pronotuin
in the male, flavo-luteous. Pronotum feebly subtectate, olivaceous,
sometimes punctate with flavous, the lateral lobes sometimes clouded with
dull flavous and always ligliter than the disk, which is more or less though
never strongly infuscated or of a deeper green, leaving however an ordi-
narily broad flavous median stripe; median carina slight, percurrent;
prozona considerably produced and strongly rounded in front, about as
long as the metazoua, which expands only a little so as to be only as
broad as ( ^^ ) or but little broader than ( 9 ) the width at the eyes, the
disk more or less tumid, at least in the female, posteriorly obtusangulate,
the angle generally broadly rounded. Prosternal spine slender, com-
pressed conical, subacuminate, erect. Tegmina extending considerably
beyond the abdomen, moderately slender, olivaceous, immaculate, the
edge of the anal area more or less flavous; wings hyaline with green veins.
Fore and middle femora slightly enlarged in the male; hind femora not
very stout, about reaching the tip of the abdomen, olivaceous, the outer
face more or less hoary and bisally flavescent, the genicular lobe flavous ;
hind tibiae red, the spines luteous with black tips. Male cerci fully half
as long ag-ain as broad, tapering but little, inbent at middle, apically
truncate with rounded angles and mesially emarginate, the lower lobe
projecting the most; subgenital plate very short scaphiform, upturned,
apically emarginate half way to base, forming a V-shaped incision, gener-
ally much deeper than broad, the angles hardly rounded.
Length of body, J*, 45 mm., 9, 56 mm.; antennae, ^, 20 mm., 9)
19 ram.; tegmina, (J, 41 mm., 9,51.5 mm.; hind femora, (J , 23 mm., 9,
30 mm.
22 (^ , 16 9 . Grant's Pass, Oregon, Sept. 8, Morse; Gazelle, Sept.
5, JMorse, Tulare, Aug. 5, Morse, Palm Springs, July 12-13, Morse, and
Indio, Cal., July 9, Morse ; Eeno, Nev., Aug. 16; Wasatch Mts. near
Beaver, July 12-18, Palmer, and Spring Lake Villa, Utah Co., Utah,
Aug. 1-4, Palmer; Ft. Buchanan, south of Tucson, Ariz., Palmer;
468 PROCEEDINGS OP THE AMERICAN ACADEMY.
Texas, June 25-26, Morse; San Luis Potosi, Mex., Palmer, and Sierra
de San Miguelito, Mex., Palmer.
The specimens, one female each, from the last two localities, differ from
the others in their greater robustness and the brevity of the tegmina,
which scarcely surpass the abdomen; the hind tibi?e also are not so bril-
liant red.
The specimens collected by Mr. Morse at Indio, Cal., were " common
on fruit and shade trees about ranches; they were shy and very active,
taking alarm easily and flying fast and far." Those taken in the north
at Grant's Pass, Or., and Gazelle, Cal., were found on willows.
32. Schistocerca mexicana sp. nov.
Of rather small size, not stout, testaceous, somewhat infuscated. Head
light testaceous, the carinse heavily infuscated, with a strong suborbital
geual fuscous stripe, and on the vertex a pair of deep fuscous diverging
and arcuate stripes, bordering the median testaceous stripe; frontal costa
subequal, faintly narrowed above and at the ocellus, sulcate below the
latter ; eyes not very prominent, ranch longer than the iufraocular por-
tion of the gense ; antennae fully half as long again as the head and
pronotum together, flavous. Pronotum feebly tectate with faint median
carina in a narrow flavo-testaceous median stripe, the rest of the disk
ferruofineo-testaceous somewhat infuscated, especially next the median
stripe, the lateral lobes ferrugineo-testaceous with a conspicuous oblong-
quadrate submedian luteous patch, edged with fuscous ; prozona produced
anteriorly and angulato-rotundate, about as long as the metazona, which
enlarges a little but is not broader than at the eyes, and behind is obtus-
angulate, the angle narrowly rounded. Prosternal spine small, erect,
conical, bluntly pointed. Tegmina extending a little beyond the abdomen,
not very slender, ferrugineo-testaceous, immaculate, the anal margin dull
flavous; wings very faintly infumate in the apical half, the veins luteous.
Fore and middle femora very slightly enlarged; hind femora flavo-testa-
ceous, the outer face hoary on the lower half, heavily trifasciate with
blackish fuscous, the fasciations completely blended on the inner face and
partly blended on the upper half of the outer face; hind tibi* purplish
fuscous, the spines hardly lighter with black tips. Male cerci heavy, sub-
equal, nearly twice as long as broad, apically truncate and slightly emar-
ginate ; subgenital plate short haustrate, not compressed, apically fissate
about half way to the base, the lobes rounded and overlapping so as to
close the fissure.
SCUDDER. — THE GENUS SCHISTOCERCA. 4G9
Length of body, 34 mm.; antennae, 16+ mm. ; tegmina, 31 mm. ; hind
femora, 20.5 mm.
1 ^ . Sinaloa, Mex., Keels, Behrens.
33. Schistocerca separata sp. nov.
Of average size and stoutness, dark olivaceous. Head rather large,
flavous, much mottled and streaked with olivaceous, and with a pair of
dark olivaceous diverging stripes on the vertex, bordering a median flavo-
olivaceous stripe; frontal costa subequal, shallowly sulcate below the
ocellus; eyes somewhat prominent in the male, much longer than the
infraocular portion of the genre ; antennae rufous. Pronotum well arched,
in no way tectate, with a slight median carina, dark olivaceous, darker
on disk than on the lateral lobes, where it is faintly mottled with flavous,
with a rather broad and equal median stripe of very faint olivaceous,
sometimes wholly obscured and always confined to the prozona ; the latter
angularly produced in front, the angle rounded, almost as long as the
metazona, which is but slightly expanded so as to be slightly narrower
( (^ ) or slightly broader ( 9 ) than at the eyes, posteriorly only a little
obtusangulate, the angle narrowly rounded. Prosternal spine moderate,
conical, blunt, suberect. Tegmina reaching considerably beyond the
abdomen, not very slender, olivaceous, in the female the median area
more or less ferruginous and with the distal half showing the faintest
possible signs of small scattered maculations ; wings faintly luteo-infumate.
Fore and middle femora not enlarged in the male ; hind femora olivaceous,
the outer fiice lighter than the rest with a flavous tinge ; hind tibiae oli-
vaceo-testaceous, the spines luteous with black tips. Male cerci fully
twice as long as basal breadth, the lower margin straight, the upper
slightly convex and sloping, the upper posterior angle well rounded, the
apical margin roundly and a little obliquely truncate ; subgenital plate
slender scaphiform, much compressed, apically acuminate as seen laterally
and fissate to the base, the lateral lobes rounded.
Length of body, ^ , 32 mm., 9 , 55 mm. ; tegmina, ^ , 34.5 mm., 9 ,
55 mm.; hind femora, ^, 19.5 mm., 9, 33 mm.
1 (? ? 2 9- Chontales, Nicaragua; Costa Rica, Underwood (Bruner).
34. Schistocerca shoshone.
Acridlum shoshone Thorn., Proc. Acad. Nat. Sc. Philad., 1873, 295 (1873); Rep
U. S. Geol. Surv. West 100 Mer., V. 895, pi. 43, fig. 2 (1875).
Schistocerca shoshone Brun., Rep. Neb. St. Hort. Soc, 1894, 163 (1894).
470 PROCEEDINGS OP THE AMERICAN ACADEMY.
This is a strictly western species, I have seen specimens from Spring
Lake Villa, Utaii Co., Utah, Aug. 1-4, Palmer ; Truckee Valley, Nev.,
Eidgway ; Colorado desert, Cal. (Stanford Univ.) ; Ft. Whipple, Ariz.,
Palmer ; Mesilla, June 30, Morse, July 4, Cockerell (Morse) and Las
Cruces, N. Mex., Cockerell (Bruner) ; Pecos River, Pope, and El Paso,
Tex., Morse ; Juarez, Mex., July 3, Morse ; and Sonora, Schott. It was
originally described from Nevada and Utah, and besides the States men-
tioned above it has been reported from Idaho, Hunter, and Colorado,
Cockerell.
This species was occasionally met with by Mr. Morse at El Paso, Tex.
on shrubbery along streams and ditches.
35. Schistocerca obliquata sp. nov.
Of rather large size and moderate stoutness, testaceous (or flavo-testa-
ceous ?) somewhat iufuscated dorsally. Head not large, luteo-testaceous,
sometimes with a pair of diverging fuscous stripes ou' the vertex ; frontal
costa equal, sulcate, especially below the ocellns ; eyes not very promi-
nent in the male, somewhat ( J" ) or hardly (9) longer than the infra-
ocular portion of the gen?e; antenntc half as long again as the head and
pronotum together in the male, luteous. Pronotum feebly tectate, espe-
cially in the female, where the prozona is a little tumid, with a slender
median carina, testaceous, the disk niore or less infuscated, leaving some-
times a faint paler median line along the prozona ; prozona anteriorly
produced and strongly rounded, as long as the metazona, which broadens
a little so as to be a little narrower (J ) or a little wider ( 9 ) than at the
eyes, posteriorly obtusangulate, the angle a little rounded. Prosternal
spine slender, subconical, bluntly pointed, erect. Tegmina extending
somewhat beyond the abdomen, not very slender, testaceous, immaculate,
distally hyaline; wings hyaline, basally with a very faint flavous tinge.
Fore and middle femora slightly enlarged in the male ; hind femora about
reaching the tip of the abdomen, luteo-testaceous, sometimes obscured
with fuscous above, basally ; hind tibise luteous or ferrugineo-luteous,
the spines with black tips. Male cerci stout, scarcely tapering, incurved
a little, less than half as long again as broad, apically truncate and feebly
emarginate ; subgenital plate rather short haustrate, a little upturned,
not compressed, apically fissate half way to the base, the angles rounded.
Length of body, ^, 41 mm., 9, 55 mm.; antennae, ^, 18 mm., 9,
30 mm. ; tegmina, (^ , 38.5 mm., 9 , 51.5 mm. ; hind femora, ^' , 23 mm.,
9 , 29 mm.
SCUDDER. — THE GENUS SCHISTOCERCA. 471
1 (^ , 2 9 , San Jose del Cabo, Mex., Cal. Acad. Sc. (Bruner) ; Chi-
huahua, Mex. (Morse).
All the specimens have been immersed in alcohol.
36. Schistocerca perturbans sp. nov.
Of rather large size, moderately slender, testaceous. Head not very
large, testaceous, with a pair of obscure, diverging, fuscous stripes on the
vertex ; frontal costa contracted feebly at summit and just below the
ocelhis, gradually broadening below, sulcate at and below the ocellus ;
eyes longer than the infraocular portion of the gense ; antennae luteo-
testaceous. Pronotum rather feebly tectate, testaceous, obscurely infus-
cated on the sides of the disk, especially posteriorly, and also below the
middle of the lateral lobes, where the fuscous cloud is abruptly and
obliquely delimited below ; median carina rather coarse, percurrent ;
prozona anteriorly produced and angulato-rotundate, scarcely shorter than
the metazona, wliich is laterally shouldered and posteriorly only a little
obtusangulate, the angle generally narrowly rounded. Prosternal S2)ine
strong, cylindrical, blunt, erect. Tegraina very much surpassing the
abdomen, rather slender, testaceous, the median area proximally obscured
feebly with fuscous, distally faintly and rather profusely maculate with
fuscous in obliquely transverse stripes ; wings hyaline, feebly infumate
in apical half, basally very feintly lutescent, the veins luteo-ferruginous.
Hind femora testaceous, the outer face hoary, the genicular arc fuscous ;
hind tibite luteo-testaceous, the spines with black tips.
Length of body, 43 mm. ; antennae, 19 mm. ; tegmina, 51 mm. ; hind
femora, 2G.5 mm.
5 9 • Paramaribo, Dutch Guiana, Richardson (Mus. Comp. Zool.) ;
Para and Rio de Janeiro, Brazil, Thayer Exp. (Mus. Comp. Zool.).
All the specimens have been long immersed in alcohol.
37. Schistocerca bivittata.
*
Grjillus flavkornis var. 5, € Thunb., Mem. Acad. St. Petersb., V. 226 (1815) t. Stal.
Acridium {Schistocerca) biviltatum Stal, Rec. Orth., I. 66 (1873).
I have not seen this species, the provenance of which is unknown. It
comes pretty certainly from South America.
472 PROCEEDINGS OP THE AMERICAN ACADEMY.
38. Schistocerca peregrina.
Acridium peregrimim Oliv., Voy. Emp. Ott., II. 424 (1807).
Acridium (Schistocerca) pereyrinum Stfil, Rec. Ortli., I. 65 (1873).
Acridium (Schistocerca) peregrimim (pars) Berg., Aual. Soc. Cient. Argent., IX. 275
(1880).
Schistocerca peregrina Brunn., Prodr. Eur. Ortli., 215 (1882).
Gryllus migratoriits (pars) Thunb., Mem. Acad. St. Pe'tersb., V. 243 (1315) t. Stal.
Gryllus rufescens Thunb., Loc. cit., V. 245 (1815) t. Stal.
Acridium flaviventre Burm., Handb. Ent., II. 631 (1838).
This is the migratory species which occurs on both sides of the Atlantic.
I have seen specimens from Brazil, Janson, and Panama, Hassler Exp.
(Mas. Comp. Zool.), in this comitrj ; and in the Old World from Spain,
Bolivar, Cairo, Egypt, Upper Egypt, and the Cape of Good Hope, the
last from Schaum's collection, XnheWeA Jlaviventre. Its distribution in the
Old World is given by Brunner as northern Africa, from Senegal to the
Red Sea, but not farther south, Syria, the Balearic Islands, and Portugal.
The published references to its occurrence in America have been obscured
by other species being confounded with it, so that no further details can
be given than are found above. Stal, however, states that the Stockholm
museum has specimens from Argentina, Montevideo, and Bahia in
America, and from j\Iadeira, Teneriffe, Algeria, Egypt, Nubia, and the
East Indies.
As this is the only species of Schistocerca known in the Old World,
and as it occurs in both hemispheres, there can be little doubt that it
originated in America, — the home of the genus. Notwithstanding its
known powers of extended flight one would hesitate to affirm that it crossed
the Atlantic Ocean on the wing, were it not that it has actually been taken
on vessels in midocean,* viz. in Lat. 25° 28' N., Long. 41° 33' W., which
is about as near one continent as the other, but slightly nearer Africa.
"The clouds and ship's sails were full of them for two days" early in
November.
39. Schistocerca paranensis.
Acridium paranense Burm., Reise La Plata, I. 491 (1861).
Schistocerca paranensis Brun., Inf. Com. Inv. Lang., I. 1, pi., figs 1-3, 6 (1898).
Acridium (Schistocerca) peregrimim (pars) Berg, Anal. Soc. Cient. Argent., IX.
275 (1880).
This is the destructive locust of Argentina, which has sometimes been
confounded with the preceding. The only specimens I have seen are
* See Psyche, II. 24.
BCUDDER. — THE GENUS SCHISTOCERCA. 473
those sent to me by Professor Bruner from Argentina, while engaged in
studying its ravages there. His report states that it occurs not only in
Argentina, but in " the adjoining portions of Uruguay, Paraguay, Brazil,
Bolivia, and Chile." Burmeister received it from the valley of the Pa-
rana, whence it derives its name. Its home, according to Bruner, appears
to be to the north of the settled district of Santa Fe, Cordoba, and
Entre Rios.
40. Schistocerca exsul.
Schistocerca exsul Scudd.!, Bull. Mus. Comp. ZoiJl., XXV. 4 (1893).
The single female I have seen was taken " two hundred and fifty miles
off the west coast of South America."
41. Schistocerca pallens.
Gri/Uus pallens Thunb., Mem. Acad. St. Pe'tersb., V. 237 (1815).
Acridium {Schistocerca) pallens Stal, Rec. Orth., I. 66 (1873).
Schistocerca pallens Brunn.-Redt., Proc. Zool. Soc. Lond., 1892, 210 (1892).
1 Acridium cubense Sauss., Orth. Nova Amer., II. 14 (1861).
? Cyrtacanthacris pectoralis Walk., Cat. Derm. Salt. Brit. Mus., III. 551 (1870).
Specimens are before me from Mexico, Packard (Mus. Comp. Zool.) ;
Vera Cruz, Mex., Heyde (Bruner); Motzoronga, Mex. (Bruner);
Meriden, Yucatan ; Realejo, Nicaragua, April, McNeil ; San Domingo,
Frazer ; Surinam, Schaum ; Pernambuco, May (Bruner), Chapada,
Aug. (Mus. Comp. Zool.), and Victoria, Brazil, May (Bruner) ; and
Uruguay River, Wyman (Mus. Comp. Zool.). It is also reported from
Cuba by Stal.
42. Schistocerca cancellata.
Acridium cancellatum Serv., Orth., 664 (1839) ; Blanch., Gay, Hist. Fis. Chile,
Zool., VI. 71 (1849).
I have specimens from Chile which agree perfectly with the descriptions
of the original specimens from the same country. There is also a speci-
men in the Cambridge Museum from Para, July. It differs but little
from S. americana, and may perhaps be regarded as a mere race of that
species.
Blanchard in Gay's Chile refers to a figure of this species in the atlas
of that work (Plate 2, Figure 7), but I cannot discover that more than
one plate of Orthoptera was ever published. Four copies have been
examined in a vain search for more.
474 PROCEEDINGS OP THE AMERICAN ACADEMY.
43. Scliistocerca americana.
Gryllus americanus Drury, 111. Nat. Hist., I. 128, pi. 49, fig. 2 (1770).
Acridium americantim Scudd., Bost. Journ. Nat. Hist., VII. 4'6G (1868).
Cyrtacanthacris americana Walk., Cat. Derm. Salt. Brit. Mus., III. 550 (1870).
Acridium (Schistocerca) americanum Stai, Rec. Orth., I. GO (1873).
Schistocerca americana Brun., Bull. Washb. Coll., I. 136 (1885).
Gryllus succinctus (pars) Linn., Syst. Nat., ed. xii. I. 099 (1707) t. Stal.
Gryllus rusticus Fabr., Syst. Ent., 292 (1775).
Acridium rusticum Burm.!, Handb. Ent., II. 633 (1838).
Gryllus tartaricus Panz., Urury, pi. 49, fig. 2 (1785-88).
Gryllus serialis Thunb.!, Mem. Acad. St. Pe'tersb., V. 241 (1815),
Acridium vittatum Palis., Ins. Afr. Amer., 140, pi. 4, fig. 5 (1817).
Acridium ambiguum Thorn., Ann. Rep. U. S. Geol. Surv. Terr., V. 447 (1872).
Acridium obscurum Bol. (nee Fabr.), Me'ni. Soc. Zool. France, 1888, 144 (]
Excepting S. peregrina, which has crossed the ocean and colonized
another world, this species of Schistocerca is the most widely distributed
of the genus, and merits its name, ranging as it does from North America
east of the Great Plains and south of about latitude 40°, through the
West Indies, Mexico, and Central America to South America, where it
occurs as far as Colombia in the west and Argentina in the east, though
the records of its occurrence in South America are few. North of north
latitude 40° or thereabouts, sporadic cases of its appearance are recorded,
notably in Massachusetts and southern Ontario; these are doubtless
accidental visitants, flying from their proper home farther south.
I have before me specimens from Pennsylvania, Uhler; Maryland,
Sept. 19, 30, Oct. 6, Uhler; North Carolina, Holder, Uhler, Shute ;
Dingo Bluff, Nov. 15, Parker, Maynard, and Smithville, N. C, Nov.
22 ; Georgia, Morrison ; Florida, Norton, Wurderaann ; Jacksonville,
April, Maynard, Dummet's Grove, Feb., Maynard, Green Cove Spring,
Boardman, F'ort Reed, Comstock, Tallahassee, Glover, Indian River,
Palmer, Key West, Maynard, Palmer, Morrison, Burgess, and Logger-
head Key, Fla., Mayer; Illinois, Uhler; southern Illinois, Kennicott,
Uhler, and Moline, 111., McNeill ; Indiana, Oct. 15, Blatchley (Morse);
New Madrid and St. Louis, Mo., Engelmann ; Lincoln, Nebr., May 3
(Bruner); Texas, Lincecum, Belfrage * ; Goliad, Dec. 3, Palmer,
Dallas, Feb. 20, Mar. 12, May 26, Boll, Bosque Co., July 27, Belfrage,
and Carrizo Springs, Tex., Wadgymar (Bruner) ; Gulf coast of Texas,
Aaron ; Mexico, Packard (Mus. Comp. Zool.) ; Matamoras, Tamaulipas,
Couch, Uhler; Cuernavaca, May 12, Barrett (Bruner), Tepic, Calif.
* Where it is said to be not rare and " found only in woods, particularly in pine
barrens, being never seen in open fields or in suuuner."
SCUDDER. THE GENUS SCHISTOCERCA. 475
Acad. Sc. (Bruner), and Vera Cruz, Mex., Heyde (Bruner), Salle;
Yucatan, Schott; Chontales, Nicaragua; Cauca River near Medelin,
Colombia, Denton (Morse) ; Inagua, Bahamas ; Cuba, Gundlach, "Wright,
Uhler ; Isle of Pines, Scudder ; San Domingo, Frazer ; Jamaica,
Cockerell (Bruner); South America ; and Carcaraiia, Argentina, Jan.,
Feb. (Bruner).
It has also been reported from Ontario, Moffatt ; Toronto, Walker ;
eastern Massachusetts, Sprague, Morse ; New York, Thomas, Beuten-
miiller; Staten Island, Davis; New Jersey, Smith; Michigan and Ohio,
Bruner; Minnesota, Lugger; Iowa, Bessey, Ball; S. Dakota, Bruner;
Kansas, Bruner ; Kentucky, Garman ; Tennessee, Thomas ; Mississippi,
Weed ; Virginia, Howard ; District of Columbia, Thomas ; and St.
Bartholomew, Stal.
Specimens from the West Indies are almost always of a small size, and
it was to these that Thuuberg applied the name serialis.
44. Schistocerca damnifica.
Acridium damnijicum. Sauss., Orth. Nova Amer., II. 14 (1861).
Cyrtacanthacris unilitmita Walk., Cat. Derm. Salt. Brit. Mus., IV. 611 (1870).
Acridium rubicjinosum Thorn, (nee Harr.), Rep. U. S. Geol. Surv. Terr., V. 170
(1873).
Acridium rugosum Prov.!, Nat. Canad., VIII. Ill (1876).
Acridium appendiculatum Uhl.!, MS., Scudd., Proc. Bost. Soc. Nat. Hist., XIX. 86
(1877).
This is a common southern species differing greatly from all other
species of Schistocerca, occurring in the United States east of the Great
Plains from Pennsylvania, Indiana, Illinois, and Arkansas to the Gulf.
It is not known to extend into Mexico. It is our smallest species.
I have seen specimens from Pennsylvania, Schaum ; Maryland, April
26, Sept. 30, Oct. 8, 18, Uhler; Middle States, Osten Sacken, early in
spring ; Virginia, May (Bruner) ; North Carolina, Shute ; Dingo Bluff,
Nov, 15, Parker, Maynard, and Smithville, N. C, Mar. 22, Maynard ;
Georgia, Morrison, Oemler; Florida, Mar. 10, Uhler, Norton, Archer,
March (Bruner) ; northeast Florida ; east Florida, Ashmead (Morse) ;
Jacksonville, April, Maynard, Palmer, Key West, Palmer, and Appa-
lachicola, Fla., Thaxter; Alabama, Morrison (Bruner), Baker; Texas,
Belfrage; Tiger Mills, Schaupp (Bruner), and Dallas, Tex., July 26,
Boll ; and Johnson, Aug. 18, Ashmead (Morse) and Fayetteville, Ark.,
McNeill (Morse).
476 PROCEEDINGS OF THE AMERICAN ACADEMY.
It has also been reported from Tennessee, Saussure ; Indiana, "Walker ;
southern Illinois, Thomas ; and even from Quebec, Canada, Provancher.
I have myself seen Provancher's specimen, but must confess to some
doubt whether it was actually taken in that part of Canada.
The following described species I have not been able to determine.
Acridium carneipes Serv., Orth., G65 (1839). Berg regards this as a synonym of
<S. peregrina.
Cyrtacanthacris concolor Walk., Cat. Derm. Salt. Brit. Mus., IV. 610 (1870).
Cyrtacanihacris diversifera Walk., Log. cit., IV. 611 (1870). Probably a Melano-
plus.
Acridium emortuale Sauss., Orth. Nov. Amer., II. 13 (1861). This appears to ap-
proach S. desiliens Scudd., but the eyes of the latter species are not " maxime
appropinquati."
Acridium luridescens Walk., Loc. cit., III. 583 (1870).
Acridium panthennum Walk., Loc. cit., IV. 623 (1870). Pretty certainly not a
Schistocerca.
Acridium picei/rons Walk., Loc. cit.. III. 578 (1870).
Acridium proprium Walk., Loc. cit., IV. 621 (1870).
Acridium strenuum Walk., Loc. cit.. III. 580 (1870).
Acridium toltecum Sauss., Loc. cit., II. 14 (1861). Probably not a Schistocerca.
Acridium varipes Walk., Loc. cit.. III. 681 (1870). Very doubtful if it be a Schis-
tocerca.
Acridium vicarium Walk., Loc. cit, III. 580 (1870).
Acridium vitticeps Walk., Loc. cit., III. 579 (1870).
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. No. IS. — April, 1899.
CONTRIBUTIONS FROM THE HARVARD MINERALOGICAL
MUSEUJVL
VII. — O.V HARDYSTONITE, A NEW CALCIUM-ZING
SILICATE FROM FRANKLIN FURNACE,
NEW JERSEY.
By JoHx E. Wolff.
COMMUNICATIONS FROM THE HARVARD MINERALOGICAL
MUSEUM.
VIL — ON HARDYSTONITE, A NEW CALCIUM-ZINC SIL-
ICATE FROM FRANKLIN FURNACE, NEW JERSEY.
By John E. Wolff.
Presented March 8, 1899. Received March 13, 1899.
In the fall of 1898, while studying the ore deposit in the new work-
ings at North Mine Hill, Franklin Furnace, in connection with the Frank-
lin Folio of the United States Geological Survey, the writer took down
from the wall of a cross-cut at the extreme north workings, and about
900 feet below the surface and near the limestone foot-wall, a specimen
of ore composed of small irregularly interlocking grains of green and
reddish willemite, lilac-brown rhodonite, franklinite in abundance, and
a white mineral which is the subject of this paper. The ore is banded,
and the grains average about a millimeter in diameter.
Chemical Composition. — Part of the specimen was pulverized and
passed through a 90-mesh sieve, the franklinite and rhodonite taken out
by the electro-magnet and portions of the white mineral obtained by
careful hand-picking, which were then purified from a trace of calcite
by the Thoulet solution, but still contained a few specks of franklinite
and an occasional grain of willemite. The following analyses (I. to IV.)
were made on separately picked portions : —
I. II. III. IV.
1.59
SiOa
38.00 38.10 38.10
ZnO
24.00 . .
24.30
MnO
3.48
1.50
CaO
.33.70 . .
33.85
MgO
1.43 . ,
1.62
Fe.303
0.58 . ,
0.57
Ignition
0.58 . .
0.52
Total
101.77
100.46
0.86
V.
VI. VII.
VIII.
IX.
38.34
.639 .639
38.34
88.66
24.45
1.50
•302}. 323
.021 )
25.88
24.47
1.43
34.07
1.62
.040 5
35.78
33 83
1.61
100.00
100.00
100.00
For the iron and manganese determination of IV. over a gram of
material was used which had been separated from the willemite, etc., by
480 PROCEEDINGS OP THE AMERICAN ACADEMY.
the barium-mercury-iodide solution, for tlie others about f of a gram. In
I. the mineral was decomposed by HCl alone, in the others after pre-
vious fusion with Na.jCOs. Iron was first precipitated as basic acetate,
then the zinc as sulphide, dissolved and re-precipitated as carbonate, the
manganese as sulphide and carbonate, calcium and magnesium, as usual.
The manganese carbonate in I. was accidentally contaminated and is in-
accurate. Taking III. for computation, in V. it is recalculated to 100,
omitting iron and ignition as non-essential. VI. and VII. give the mo-
lecular ratios from which we deduce (ZnMn)O : (CaMg)O : SiOs as
1:2:2 with Mn to Zn and Mg to Ca as 1 to 15 respectively. The
formula is hence (ZnMn)O, 2 (CaMg)O, 2 SiO. or ZnO, 2 CaO, 2 SiO.,.
VIII. gives tlie theoretical composition of the latter, and IX. of the
former. The mineral gives off a little chlorine with liCl indicating a
higher oxidation for some of the manganese unless due to traces of
franklinite, and the ferric iron may also be due in part to that impurity.
Crystal System and Physical Characters. — The grains have no dis-
tinct crystalline boundaries, but show several cleavages. Thin sections,
prepared by scattering the grains in balsam and grinding thin with em-
ery, give in polarized light numerous basal sections which show a distinct
uniaxial cross (without any perceptible opening into hyperbolas) and neg-
ative optical sign. These basal sections show two distinct sets of rec-
tangular cleavages apparently normal to the base, and at 45° to each
other, one a little better than the other. Prismatic sections show a good
basal cleavage (i. e. normal to the negative optical direction) and par-
allel extinction, with a strong bi-refringence. The mineral is therefore
tetragonal with a good basal cleavage and secondary cleavages parallel to
the prisms of the first and second orders. The specific gravity determined
with the pycnometer on over a gram of the material used in Analyses 1.
and IV. was respectively 3.397 and 3.395. Hardness between 3 and 4,
color white to transparent, with a glassy lustre. Pyrognostics: unaltered
in the open and closed tubes. It fuses in the forceps with difficulty to a
cloudy glass, giving an intense red calcium flame, especially when moist-
ened with HCl. On charcoal the powdered mineral glows intensely, and
gives a heavy zinc coating, which is intensified by using NaoCOg, the
centre of the assay is then colored bluish green (Mn). A manganese
reaction with borax ; gelatinizes easily with HCl. Aside from the crystal
system it may be separated from willemite by the hardness, negative
optical character, and intense calcium flame.
The mineral is named Hardystonite from Hardyston township, in which
the Franklin mines are situated.
WOLFF. — HARDYSTONITE. 481
Its systematic position is perhaps not clear without crystallographic
material. Having an oxygen ratio of 4:3 it would belong among the
intermediate silicates (Dana's System, 6th ed.), and its tetragonal system
and cleavages would place it near ganomalite (Pb3Si20,). Some genetic
connection with clinohedrite * (HoCaZnSiOg) may be surmised, as both
minerals come from the same workings, and the writer is informed that
a white mineral, probably hardystonite, has been found there in con-
siderable quantity. It is hoped in a future visit to the mines to obtain
further light on these questions.
* Penfield and Foote, Am. Journ. Sci., 1898, Vol. V. p. 289.
VOL. XXXIV. — 31
Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIV. No. 19. — April, 1899.
CONTRIBUTIONS FROM THE GRAY HERBARIUM OF
HARVARD UNIVERSITY.
New Series. — No. XV.
<
By M. L. Fernald.
I. Eleocharis ovata and its American Allies.
II. Scirpus Eriophorum and some Related Forms.
With a Plate.
CONTRIBUTIONS FROM THE GRAY HERBARIUM OF
HARVARD UNIVERSITY, NEW SERIES, No. XV.
By M. L. Fernald.
Presented by B. L. Robinson, March 8, 1899. Received March 15, 1899.
I. — ELEOCHARIS OVATA AND ITS AMERICAN ALLIES.
In attempting to place satisfactorily a number of strikingly different
American plants, which, according to the standard works upon that
group, must all be called Eleocharis ovata, a detailed study of the achenes
has shown that our present conception of the species — especially in
America — is remarkably indefinite. The commonest American plant
passing as Eleocharis ovata is an annnal species with many erect or
ascending comparatively stout culms from 1 to 5 dm. high, capped by
thick globose-ovoid or ovate-oblong obtuse densely flowered brown heads,
3 to 13 mm. long. The closely a^ipressed ascending obovate-oblong or
suborbicular scales are blunt, with scarious margins. The tubercle,
usually as broad as the cuneate-obovate achene, is depressed, somewhat
resembling in outline a high-crowned tam-o'-shanter cap; it is generally
one third as high as the body of the achene. This common American
plant (Figs. 1 to 7), now accepted as E. ovata, was described in 1 809 by
Willdenow as Scirpus obtnsus, and it was subsequently transferred by
Schultes to Eleocharis. For three fourths of a century the plant was
generally treated by Torrey, Gray, and other recognized authorities on
the group, as a distinct American species. In his monograph of the
Cyperacece, however, Burkeler reduced^ our common American Eleo-
charis {Heleocharis) ohtusa to the well known E. ovata. of central
Europe. This disposition of the plant was accepted by Mr. C. B. Clarke
in his study of the European species of Eleocharis,'^ and it has been
adopted by subsequent American students of the group, — Britton, Wat-
son, etc. Habitally the two plants are essentially alike, but a careful
examination of achenes from an abundance of American and European
specimens reveals certain differences which appear quite constant. The
1 BGckeler, Linnaea, (1869-70), XXXVI. 463. 2 Jour. Bot., XXV. 268.
486 PROCEEDINGS OP THE AMERICAN ACADEMY.
achene of the European E. ovata (Figs. 9, 10) is obovate or inverted-
pyriform in outline, and it is about three fourths as high and two thirds as
broad as the obovate or cuneate-obovate achene of the typical American
plant (Figs. 4 to 7) which commonly passes under that name. The
tubercle of true £J. ovata averages four sevenths as broad as the achene,
while that of the American plant equals the achene in breadth. Though,
as already stated, the European . E. ovata and the American plant
recently united with it are not readily distinguished by superficial
characters, the apparently constant differences in their achenes and
tubercles are sufficient to justify us in regarding our own plant as dis-
tinct from that of Europe, and in restoring to it the distinctive name
Eleocharis obtusa, under which it was so long known to American
botanists.
Although the common Americiin plant, which, for the last three de-
cades, has passed as Eleocharis ovata, proves on critical study to differ
from that species in certain well marked and constant characters, the true
jEJ. ovata of Europe is not entirely wanting in our American flora. The
American plant, however, which not only in habit but in the characters
of achene and tubercle closely matches the European specimens and
plates, is as yet known from only four northern stations, in New Bruns-
wick, Maine, Vermont, and Mioliigan. In these specimens, in habit and
achenes undoubtedly E. ovata, the oblong or ovate-oblong scales are
very dark chestnut-brown or purplish, distinctly darker than is usual in
E. obtusa.
In October, 1878, Mr. E. H. Hitchings collected in Dedham, Massa-
chusetts (presumably in Purgatory Swamp), an Eleocharis which has
proved unusually puzzling to those who have subsequently worked upon
the genus. Two sheets of the plant, showing large and small specimens,
are preserved in the Gray Flerbarium, where they have been frequently
shifted from one species cover to another. Originally Dr. Gray wrote
upon one of the sheets, a " remarkable form, I think, of Eleocharis inter-
media." Subsequently both sheets were referred by Dr. Watson to
E. obtusa ; but when studying the plants in the prejmration of his synop-
sis of "The Genus Eleocharis in North America," ^ Dr. N. L. Britton re-
ferred the two Dedham sheets to different species, the smaller specimens
to E. olivacea, the other to E. palustris. Why the two sheets should be
thus separated we cannot make out. They are, to be sure, hardly iden-
tical in size, but in general habit, scales, and achenes they are the same,
1 Jour. N. Y. Microsc. Soc, V. 95-111.
FERNALD. — ELEOCHARIS OVATA. 487
and they are both annuals with small tufts of merely fibrous roots. Both
Eleocharis palustris and E. oUvacea, on the other hand, are well known,
and are always described as perennials with definite root-stocks. Another
Massachusetts specimen, collected by C. E. Perkins at AVinchester, soon
after Mr. Hitchings found the puzzling Dedham plant, has likewise been
referred to both E. obtusa and E. oHvacea, and doubtfully to E. diandra.
In attempting, then, to place satisfactorily this anomalous plant, recent
botanists have associated it at different times with no less than five
species.
The plant is probably of general, though perhaps not of abundant,
distribution throughout eastern Massachusetts. In the middle of October,
1897, a small plant, which may well be a depauperate form of the Ded-
ham plant, was collected by E. F. Williams and J. M. Greenman at
Massapoag Pond in Sharon. P^xceptional individuals among these
autumnal specimens have short capillary stolons, but, except for this
unusual development, they can hardly be distinguished from the smallest
specimens collected by Mr, Hitchings. A little later, specimens identical
with the larger Dedham plant were collected by Mr. Williams in the bog
south of Annursnack Hill in Concord.
In northern Maine, on the upper waters of the St. John and Penobscot
Rivers, where Eleocharis palustris and E. intermedia are common species,
this Dedham plant is also abundant. There it has been carefully watched
in the field, where it forms dense tufts of generally slender and decidedly
flexuous culms, which are often quite prostrate upon the ground, giving
the plant a superficial resemblance to E. intermedia. From the gener-
ally common E. obtusa, whose place this slender flexuous plant (Figs. 15
to 22) seems to take in northern Maine, it is otherwise superficially distin-
guished by its dark chestnut or purple ovate or ovate-lanceolate acutish
scales, which are looser in the heads and more spreading than the paler
brown ascending closely appressed obovate obovate-oblong or suborbicu-
lar blunt scales of E. obtusa. The color of the scales, though fairly
constant, is not, however, so distinctive a character of the Dedham
and northern Maine plant as the size and shape of the tubercle. The
tubercle of this dark-scaled plant is deltoid-conical, slightly or not at all
constricted at the base, suggesting in outline a half-closed parasol with
incurved edge ; it is about three fifths as wide as the obovate or inverted-
pyriform achene which it caps, and usually about three sevenths as high
as the body of the achene. The tubercle of E. obtusa, on tlie other hand,
as already described, is usually as broad as the cuneate-obovate achene, and
it is depressed and generally one third as high as the body of the achene.
488 PROCEEDINGS OF THE AMERICAN ACADEMY.
The retrorsely barbed bristles of this species, too, are slightly coarser than
in the other plant, though for distinguishing the species this character is
less to be relied upon than those found in the scales and tubercles.
From Eleocharis pnlustris and E. olwacea this northern plant may
generally be quickly separated by its annual habit, though, as noted in
the Massapoag specimens, it very rarely produces late autumnal stolons.
Its flexuous densely-clustered slender culms and its comparatively short
ovate heads sufficiently distinguish it from the taller erect Ji!. palustris,
with its narrower elongated heads. In habit the plant strongly suggests
E. olivacea (Figs. 23, 24), but in this latter perennial species the tubercle
is narrower and lower, and of different outline : the sides, instead of being
essentially straight, have a strong concave curve ; and below, instead of
rounding gradually to a slightly constricted base, the tubercle flares some-
what like a saucer.
Like Eleocharis intermedia and E. diandra, to which the northeastern
plant has sometimes been referred, it is an annual. In habit it strongly
suggests the former species, but that (Figs. 25, 20) has narrower spikes,
and the more elongated achene is capped by a decidedly narrow deltoid
conical tubercle reminding one of a very tall fool's cap. Nor is the plant
satisfactorily referred to Charles Wright's obscure E. diandra. From
such specimens as we know (the original material) that species (Figs.
53 to 58) seems to be of erect habit, and the narrower scales are pale
brown with dark green midribs. The plant is unicpie in tliis group of
annual species (excepting forms of the very different Engelmanni sec-
tion) in its entire lack of bristles; and its smaller obovate achene is
capped by a depressed tubercle about as broad and half as high as that
of Mr. Ilitchings's plant, but in outline resembling a miniature tubercle
of E. obtusa. In short, the northeastern plant, which has been referred
at various times to the five species here discussed, is as distinct from all
of them as are they from one another, and the only other described plant
which seems to approach it is a form of E. ovata of continental Europe.
Though E. ovata is an erect plant, and has been so described by most
European botanists, a single sheet in the herbarium of Dr. Charles "\V.
Swan shows an extreme form collected by Seidel at Reichenbach in
Silesia, which is identical with the low flexuous-culmed plant first found
in America by Mr. Ilitchings. This depressed plant with flexuous culms
hardly suggests to the casual observer the familiar erect E. ovata, but it
is certainly difficult if not impossible to find in their achenes any satisfac-
tory distinctions ; and in northern Maine, at the single known station
for the erect E. ovata, there are puzzling specimens clearly intermediate
FERNALD. — ELEOCHARIS OVATA. 489
between the two. This low form, more common in New England than
the tj'pical erect E. ovata, is doubtless the Silesian variety Beuseri of
Uechtritz. From the description alone of Terracciano's var. hutnifusa^
our plant may be the same as that Italian form. No specimens of the
latter have been seen, and as the New England plant is clearly identi-
cal with the more northern var. Heuseri, Uechtritz, that name will here
be taken up.
This flexuous dark-headed plant is not the only anomalous form long
referred to Eleocharis ovata. A tall northwestern plant, 7 or 8 dm. high,
has been considered by Mr. C. B. Clarke as a variety of this species.
Aside from its unusual size, this plant is well characterized by the re-
markable broadly obcordate tubercle (Figs. 11, 12), which is not at all
compressed and fully half as high as the achene itself Other north-
western plants, however, connect this extreme form directly with the
typical E. obtusa, so that it seems undoubtedly an extreme variety of
that species. Another striking form which an abundance of material
shows to be in reality an extreme variatio 
