PROCEEDINGS

OF THE

AMERICAN ACADEMY

OF

ARTS AND SCIENCES.

Vol. XXXV.

FROM MAY, 1899, TO MAY, 1900.

BOSTON, MASS.:
JOHN WILSON AND SON.

Santbcrsttg ^rcss.
1900.

^.51^

CONTENTS.

Page
I. The Development and Application of a General Equation for Free
Energy and Physico-Cheviical Equilibrium. By Gilbert
Newton Lewis 1

II. Short Studies of North American Tryxalinae. By Samuel H.

SCUDDER 39

III. A Revision of the Atomic Weight of Cobalt. Third Paper. —

The Analysis of Cobaltous Chloride and Oxide. By Theo-
dore William Richards and Gregory Paul Baxter . 59

IV. On the Thermal Conductivity of Vulcanite. By B. O. Peirce . 73

V. Tvo Genera of North American Decticinae. By Samuel H.

Scudder 81

VI. Note on the Finite Continuous Groups of the Plane. By F. B.

Williams 95

VII. The Echelon Spectroscope. By A. A. Michelson 109

VIII. The Electrochemical Equivalents of Copper and Silver. By
Theodore W. Richards, Edward Collins, and George
W. Heimrod 121

IX. Preliminary Diagnoses of Neiv Species of Laboidbeniaceae. — /.

By Roland Thaxter 151

X. Ferrous Iodide. By C Loring Jackson and I. H. Derby . 211

XI. Note on the Constitution of Diparabrombenzylcyanamide. By C.

Loring Jackson and R. W. Fuller 229

IV CONTENTS.

Page
XII. Note on the Chief Theorem of Lie's Theory of Continuous

Groups. By Stephen Elmek Slocum 237

XIII. A Rerixion of the Atomic Weight of Iron. Preliminary Paper.

By Theodore William Richards and Gregory Paul
Baxter 251

XIV. On Certain Colored Substances Derived from. Nitro Compounds.

Third Paper. By C. Loring Jackson and F. H. Gazzolo 261

XV. The Metamei-ism of the Hirudinea. By W. E. Castle . . . 283

XVI. 1. New Species and Varieties of Mexican Plants. By J. M.
Greenman.

2. Synopses of the Genera Jaegeria and Russelia. By B. L.

Robinson.

3. New Dioscoreas from Mexico. By E. B. Uline.

4. New Phaenogams, chiefly Gamopetalae, from Mexico and

Central America. By B. L. Robinson 305

XVII. The Freshwater Tertiary Formations of the Rocky Mountain

Region. By W. M. Davis 343

XVIII. On the Determination of Sulphuric Acid in the Presence of Iron ;
a Note on Solid Solutions. By Theodore William
Richards 375

XIX. The Species of the Orthopteran genus Derolmema. By Samuel

II. SCUDDER 385

XX. Vieio of the Carboniferous Fauna of the Narraganselt Basin.

By Alpheus S. Packard 397

XXI. Preliminary Diagnoses of New Species of Laboulbeniaceae. — //.

By Roland Thaxter 407

XXII. Historical Notes relating to Musical Pitch in the United States.

By Charles R. Cross 451

XXIII. The Driving Energy of Physico- Chemical Reaction, and its

Temperature Coefficient. By Theodore William Richards 469

XXIV. Supjilementary Note on the Chief Theorem of Lie's Theory of

Finite Continuous Groups. By Stephen Elmer Slocum . 481

CONTENTS. V

Page
XXV. 1. A Si/nopsis of the Mexican and Central American Species
of Salvia.

2. A Revision of the Mexican and Central American Solanums

of the Subsection Torvaria.

3. Some undescribed Mexican Phanerogams, chiefly Labiatae and

Solanaceae. By M. L. Feknald 487

XXVI. On the Singular Transformations of Groups Generated hy In-
finitesimal Transformations. By Henry Taber .... 575

Proceedings 601

A Table of Atomic Weights. By Theodore William Richards . 621

liEroRT OF the Council 623

Biographical Notices . 625

Epes Sargent Dixwell 625

John Cummings 628

John Codman Kopes 029

John Lowell 634

Officers and Committees for 1899-1900 643

List of the Fellows and Foreign Honorary Members . . . 644

Statutes and Standing Votes 653

Index 665

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 1. — July, 1899.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF

HARVARD COLLEGE.

THE DEVELOPMENT ANV APPLICATION OF A GEN-
ERAL EQUATION FOR FREE ENERGY AND
PHTSICO- CHEMICAL EQ UILIBRL UM.

By Gilbert Newton Lewis.

VOL. XXXV. — 1

MARINE BIOLOGICAL LABORATORY.

Received ^^^^-^^f^^^ru^^^^r^.^^
Accession No. /./'c/ v^.
Given by '--^'^^i>,..U<k^.?3^.,/^^
Place,

***No book OP pamphlet is to be

oratopy without the permission of the Trustees.

pemoved fpom the Iiab-

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF

HARVARD COLLEGE.

THE DEVELOPMENT AND APPLICATION OF A GEN-
ERAL EQUATION FOR FREE ENERGY AND
PHYSICO-CHEMICAL EQUILIBRIUM.

By Gilbert Newton Lewis.

Received June 5, 1809. Presented by T. W. Richards, June 14, 1899.

Introduction.

The advance of modern physical chemistry has been largely due to
the application to physico-chemical problems of the first and second laws
of thermodynamics and the gas law, — the latter both directly and by
analogy. Upon this basis the whole theoretical treatment of chemical
equilibrium rests at present. For this reason it may not be without
interest to attempt to express the relations deduced from these laws in a
single equation, the most convenient and the most general possible, which
may serve to systematize a part of our jDresent knowledge, and perhaps
point out new laws. Such an attempt will be made in this paper.

I. General Equations of Free Energy and Equilibrium.

The simplest expression embodying the first and second laws of ther-
modynamics is,

^=^^+^' (1)

where A is the diminution in free energy (Helmholtz) of a system in any

isothermal process ; (7 is the diminution in internal energy in the same

d A
process ; T is the absolute temperature ; -yy^ represents the change in

A as the temperature of the process varies, — the system changing in
every case from the same initial to the same final volume.

A, the diminution of free energy, is a quantity which denotes the max-
imum amount of work obtainable in an isothermal process, and which is
determined definitely by the initial and final states of a system. It may,

4 ' PROCEEDINGS OF THE AMERICAN ACADEMY.

therefore, be expressed as the difference between two quantities, one
representing wliat may be called the total free energy of the initial state;
the other, the total free energy of the final state. Representing these
quantities by Sli and ^2 ? ^e have,

Similarly, if we represent by ^1 the total internal energy of the system
in the initial state, and by Wi« the same in the final state,

equation (1) may be written,

dT dT

Si

d%

dT

and

d'E,
dT

a^ =T^ - T^ + 51, _ ^2 ,

beino; taken at constant volume.

0

This equation may be separated into two equations,

^1= 2'^ + m, + il/, (2 a)

and

d^
^,^ T^, + Wi, + M, (2b)

where 31 is an undeterminable quantity which can have only arbitrary
j)hysical significance, since we are in practice only concerned with changes
of free energy at constant temperature, and in such changes J/ always
disappears.

In two special cases, viz. the ideal gas and the dilute solution, the
expression for the change in free energy has been found to assume a very
simple form,

A = nET\n'^,

where n denotes the number of gram-molecules of the gas or solute ; R,
the gas constant ; T, the absolute temperature ; In, a natural logarithm;
Vy and Vn respectively, the initial and final molecular volumes. Al-
though the above expression gives a complete statement for the change
in free energy only in the two special cases mentioned, still we are led
by many considerations to believe that it forms an important factor in
many other, if not all, free energy changes. Evidence on this point is
ofi^ired by tlie fact that the above term is present in the general equations
of equilibrium which have found experimental verification in the most

LEWIS. — FREE ENERGY AND EQUILIBRIUM. 5

diverse systems. Moreover, we are led to the above expression for free
energy change by the kinetic theory, the extended application of which
to much more complicated cases than that of a perfect gas has already
produced important results.

The term n R T\n — may be written in the form

— nRT\x\i\ + nRT\nv.^.

Then we may consider that the quantity — nRT \ni\ represents an
essential part of the free energy of n gram-molecules of a simple sub-
stance in a given state, and similarly, nRT\\\ v.^ , a similar part of tlie
free energy in another state, and we may write for the free energy of
one gram-molecule of a simple substance the entirely arbitrary equation,

where Xi represents that part of the free energy not contained in the term
— R Tin t'l ; differentiating this equation,

^^^-Rlnv.+y,, (3)

dxi
where y^ = — .

Now substituting in equation (2 a),

(^^ = - RTlnv,+y,T+ IX, f M. (4)

Differentiating this at constant volume,

Comparing this with equation (3),

d^ ^l^,dM_ di/i _ 1 dm, 1 dM

dr'^ dT'^dT~ ' dT~ TdT~l'dT'

Since 5sli in this case represents the internal energy of one gram-

molecule, at constant volume will be equal to the molecular heat

capacity, at constant volume, which may be designated by c^^ . There-
fore,

6 PROCEEDINGS OP THE AMERICAN ACADEMY.

Integrating,

where fL^j is the integration constant whose value depends on the value
chosen for JJ, as the lower limit of integration. Substituting in equa-
tion (4), we obtain for '^i, the free energy, per gram-molecule, of a
simple substance in any given state,

^1

r.T\nv,- TT'-^dT+^T+m. + N, (5)

where

-^x

and ^is, like 3f, a quantity -which will vanish in any expression for the
difference in free energy at constant temperature.

From this we may obtain an equation for the change of free energy in
any process. In the most general case, a system composed of any num-
ber, ?/?! , of different molecular constituents,* and any amount, ni, n'l,
etc. gram-molecules of each constituent, undergoes any change, phys-
ical or chemical, arriving at a condition in which it is composed of
m^ molecular constituents, and rio , w'o , etc. gram-molecules of each
constituent.

If 3i, ^'i , etc. represent the free energies per gram-molecule of the
various simple molecular constituents, a total change in free energy will
be given by the equation,

A = (n, % + n\ Sl'i + . . . ) _ (n„ %. + n'. Sl'o + ...).

If we substitute for ^i , '^'i , etc., their values as in equation (5), the N
terms will all vanish and the equation will become,

A = — 11 T(7ii In Vi + n'l ]n v\ + . . . — rio In Vo — n'^ In v'.. — . . . )

^'"- c^, + n\ c' -\- . . . —n.c„ — n\ c' —

dT

T

+ n'l m'l + . . . - n. 5Ho — n\^ W ^ -...).

* Molecular constituent is used to mean a single molecular species in a single
phase. Thus two phases of tlie same molecular species are regarded as two con-
stituents, as are two different molecular species in the same phase.

LEWIS. — FREE ENERGY AND EQUILIBRIUM. 7

If we represent by C^^ and C^^ the total heat capacity, when each
constituent remains at constant volume, in the initial and final states
respectively ; by H the total change in the various functions denoted by
«if^i, n2^2 7 etc.; by U the total change in the internal energy of
the svstem, then the equation may be written,

,./ "'.

V 2 ?/„ 2 n^ o — G

A = RT\u^ ""^'-'-ri -^^—^dT+HT+a (6)

We have in this equation a perfectly general expression for the change
of free energy in any isothermal change, chemical or physical, in any
system, whether homogeneous or heterogeneous. The quantities con-
tained in the equation are all capable of direct experimental determination
with the exception of the quantity H, of which we only know that it is
not a function of the temperature, since it enters as the difference between
a number of integration constants each of which is indejiendent of the
temperature. The value of this function and the simple form that it
assumes in many important cases of equilibrium will be considered later.

The form which equation (G) takes in the simple cases where all the
molecular species participating in the reaction are gases and dilute solu-
tions may be shown as follows.

Equation (5), applying to the free energy of one of the simple con-
stituents of a system, is,

(Si = - E Tluv - T r ^ d T + "^ T-^Wi+ ^,
differentiated with respect to volume, at constant temperature,

J Tr. dv 1 dv dv ^ ^

dv V

Since d ^ represents the work accomplished in a reversible change,
we may write

d^ =z — p dv , V

where p denotes the gas pressure or the osmotic pressure, as the case
may be. Then

_, = _^_^ri,^,r+^a + ^^, (3)

V J T^T dv dv dv ^ ^

Now we know that in the case of perfect gases and dilute solutions the
heat of free expansion and the heat of dilution respectively are zero.

8 PROCEEDINGS OF THE AMERICAN ACADEMY.

dWi dc

Therefore — — = 0 ; moreover, -~ = 0. This is experimentally

proved in the case of gases, and may be shown in the case of solutions

as follows.

,dWi dm

dc^ ^ ^,/r ^ d'-m ^ " dv
dv do dvdT~ dT

(9)

now, ^'^^

dm ^ ~d^ _ , c?c, -

— — = 0 ; — ™- = 0 ; and -^ =: 0.
dv dT dv

Equation (8) becomes

V dv

R T

But since we know that p = is the characteristic equation of

V

d'^

the perfect gas and the dilute solution, -^ = 0.

In these two simple cases, therefore, we find that m, c^, P^, are all
independent of the volume. If then, equation (6) is applied to a reaction
between gases and dilute solutions, the quantities 5Ei, c,.j, P^i ; Wi,2i c^^,
^-2, etc. will not at any given temperature change with changing volume,
and the quantities U, C^^ — O^^, H, will be constant, however the ini-
tial and final volume conditions of the system are changed. For any one
temperature equation (6) may be written,

n J n\

A = RT\xi V^^,>"' ' ' + G (a constant). (10)

Vi 'v J 'i . . .

This equation, moreover, obviously applies to any system which con-
tains, besides gases and dilute solutions, any constituents participating in
the reaction, whose molecular volume is not changed appreciably by a
change . in the conditions of volume or pressure in the system. Thus
any solid phase of definite constitution in a heterogeneous system may be
considered constant in its molecular volume, as well as in its functions
5E, C„, f^, when the pressure of the system is varied through limits not
too wide.

Since equation (6) gives an expression for the change of free energy
in any isothermal process, we may derive immediately a general equation
for equilibrium in any system. Let us consider a system such as is ordi-
narily studied, upon which the only external force is a uniform pressure,

LEWIS. FREE ENERGY AND EQUILIBRIUM. \)

normal to the surface, and in which the effects of gravity, surface tension,
etc. may be neglected.

A necessary and sufficient condition for equilibrium is, that any change
in a system in equilibrium is reversible. In other words, the change in
the free energy of the system must be equal to the external work.*

In the case under consideration the external work is the product of the
external pressure, P, by the change, F, in the volume of the system.
Therefore in equilibrium,

A = PV. (11)

Let us consider a system, of any degree of complexity, which is capa-
ble of change. In general this change will consist in a loss by some
constituents of the system, accompanied by a corresponding gain by
others. Then according to equation (6),

A = PT\n^ \ " ~ T\ ^" ,,, 'VZr+ HT-\- U,

where quantities with subscript 1 refer to the constituents which suffer
loss; those with subscript 2, to those which gain, and n^, n\, etc., and
Ho 1 n\ , etc. represent the number of gram-molecules of each constitu-
ent lost and gained respectively.

Combining equations (6) and (11) we have as the general equation of
equilibrium,

« "2?/ "'2 r'^C — C

PV =RT\xx " ' \ • '■ - Tl " ^ "^-dT+ JIT+ U, (11 a)

where P is the external pressure and V is the total change in volume.

V — (jin Vn 4- n'o r'2 + • • • ) — («i ih + n\ v\ + ...).

This equation (11 a) expresses the equilibrium of the system in regard
to the particular change in question. When a system is in perfect equi-
librium there will be an equation of the above form for every change or
reaction which can take place independently. These equations, however,
will not all be independent. For example, if both liquid and gaseous
acetic acid are composed of two kinds of molecules, namely, CH3COOM
and (CHoC00H)2, then when acetic acid and its vapor are in equilib-

* Of course the change must not be great enough to disturb the condition of
equilibrium. Tlie following demonstration would be somewhat more rigorous if
the infinitesimal notation were used.

10 PROCEEDINGS OF THE AMERICAN ACADEMY.

rium, we have four equations of the type of (Ho); one referring to
the equilibrium iu the gaseous state of the reaction, 2 CH3COOH =:
(CH3COOH)2 ; another, referring to the same reaction in the liquid
phase ; a third, referring to the liquefaction or vaporization of the double
molecules ; the fourth, to the liquefaction or vaporization of the single
molecules. Of these four equations three are independent.

Returning to the discussion of the general equation of equilibrium,
equation (Ha), it is interesting first to determine what form it will take
when we limit the system considered to the conditions under which
equation (10) was deduced, namely, that the reacting system shall in-
clude, besides gases and dilute solutions, only " condensed " phases of
definite composition. Combining equations (10) and (11),

Now V, the change of volume, is due in this case to the change in vol-
ume of the gaseous constituents, and will therefore, at constant temper-
ature, be inversely proportional to P. Therefore P Fis a constant, and
at constant temperature,

In '' \ •' •= C\ and " ", ' "^ /v (a constant). (12)

This equation is the familiar mass law of Guldberg and "Waage, but
it is also, since it is not restricted to homogeneous systems, the law of
the constancy of the ratio of distribution among different phases. This
includes the law of Henry.

That equation (12) does not represent a universally accurate law of
nature is shown by comparison with equation (Ho) ; for it is only when

IT, H. C„ —Q. , and F Fare constant that —, — ,"„, is constant.

' - Vi ly'i 1 . . .

If this fraction for convenience is denoted by /i, which may be called,
instead of the equilibrium constant, the equilibrium ratio, then K is a
function, not only of the temperature, but also of U, H, 0^^ , C^„, P V.
The nature of this function may be shown from equation (11 a),

BTlnK=FV- U+ T C^ ^^'^ ~ ^'^dT-JIT; (13)

LEWIS. — FREE ENERGY AND EQUILIBRIUM.

11

A'=e

1 (P V V PT

'^C\',__ ^^''^dT-II

T

)

(15)

ill which e is the base of the natural system of logarithms.

From this general equation of equilibrium may be easily deduced the
expression for the change of the conditions of equilibrium with change of
temperature. In equation (14) let, for convenience,

L

TC — C

H=.F,

(16)

equation (14) differentiated with respect to tem2:)erature then becomes

'PV

) \ ^ TT TJ d f1

(17)

rfln K

dT

1

rd

dT

_}^dU U dF

]

Now 7^ is a function of T and also of In K, which is itself a function of
T. According to the laws of partial derivatives,

d_F_hF hF d\nK

d2'~ IT^ Ih^<:~dV

where 8 signifies a partial differential. Since // is independent of the
temperature, and

PT (J — C

then,

SF^a^^C^^ldJT
8T T TdT'

Therefore,

for 0..-C., = ^.

\ dU ^F d\uK

^ +

dT T dT^ hlnK dT '
and equation (17) becomes

d\uK _ 1 [ \ T J 1

dT i?L dT
and by transposition

dn U IdU 8F dl
Td T'^ T- Td T^ SluK d

'T J'

dT \ Ii8\nKj~Ji\_ TT^'^T^j

(18)

12 PROCEEDINGS OF THE AMERICAN ACADEMY.

If we consider the special case of a system whose volume is unchanged
by the reaction to which the equilibrium equation refers, then V = 0,
and

dlnKf 1 8F \ _ U

/>T O (J
"' „ — - d T — H, was shown to be con-

stant under the conditions which led to equations (10) and (12), when

the conditions of equilibrium changed at constant temperature. In such

?>F
cases, therefore, —-j — p = 0, and the equation becomes
o In A

d\nK_ U
dT ~R^'' ^'^^

This equation applies to both homogeneous and heterogeneous equilib-
rium. When applied to the former it is identical with the well known
equation of van't HofF, sometimes called the equation "isochore." This
equation, however, has been used to express the change of equilibrium
with the temperature, not merely in those systems in which the reaction
causes no change in volume, but in general. That this use is justifi-
able in the cases for which equation (20) was developed may be readily
shown. For, in systems subject to moderate pressure, the only consider-
able isothermal changes in volume are those of the gaseous phase. The
volume of the gases is, at a given pressure and temperature, proportional
to the total number of gram-molecules of the various gases present. If,
duriug the reaction to which the equation refers, there is a change of n
gram-molecules in the gaseous phase, then the total change in volume is,

,, nRT PV

V= p , or —^ = nR.

IF
In the case under discussion, where ^ — v,=: 0, equation (18) may be

written,

d\x\K _
dT

P V

Since — =- =. n R, a constant,

dh^K _ U

~d~T~ ~ Wr-'

LEWIS. — FREE ENERGY AND EQUILIBRIUM. 13

Similarly, for any system in general in which the pressure is not very
great, </ ( — =- j will be negligible, and, for a very close approximation,
equation (18) becomes

d\nAy 1 ^F\_ U

(IT \ lihla KJ ~ K T'

Comparing equations (20) and (21), we see that the conditions under
which the above law of van't Hoff holds true are practically the same
as those under which the equation " isotherm " of the mass law, equation
(12), is true; namely, that F, and therefore C^[~ C^.,^ and H are,
at constant temperature, independent of the volume conditions of the
system.

The various equations which have been here deduced from the general
equations of free energy and equilibrium can be best studied further by
their application to special cases of equilibrium, which we will proceed to
discuss somewhat systematically.

II. Application to Monomolecular Systems.

{A.) Homogeneous Systems,

1. Gases. — The simplest conceivable case of physico-chemical equi-
librium is offered by a single molecular species in a single phase in
equilibrium with the external pressure. For this case the equation of
condition has already been found, — equation (8), namely,

V J T^I dv dv dv

In the case of a perfect gas it has been shown that

dc„ ^ d"^ ^ dW^ RT

:7-^=0; -P = ^\ ^ = 0; and V = ^-^ . (22)

dv dv d V ■ ^ V ^-^

The next case that deserves attention is that of a compressed gas.
Here, also, the specific heat at constant volume is independent of the
volume. This has been shown to be true up to pressures of several
thousand atmospheres.* A recent work has questioned the absolute
accuracy of this law. We will return to this point later. Meanwhile

we may consider —~ = 0, and equation (8) then simplifies to
* Mallard and Le Cliatelier, Wied.-Beibl., XIV. 364.

IJ: PEOCEEDINGS OF THE AMERICAN ACADEMY.

V dv dv ^ ''

'.'dm'
d

From equation (9), since -^-^ = 0, — , r„ ^ = 0, and therefore the
^ ^ d V d 1

term — — is independent of the temperature when c„ is independent

of the volume. This term represents the heat developed in the free
expansion of a gas, and shows that when the internal energy of a gas
increases with increasing volume, — that is, when there is a cooling
effect on free exjjansion, — the gas will have lower than normal
pressure.

Since —p^ is a function of the volume alone, we may write equation
(23) in the form,

^ = (I- -('■))-- ^- (-)

And since -^ — is not a function of the temperature, the equation of
a gas at constant volume, the " isochore," is,

p ^^ A T — B, where A and B are constants. (25)

This equation has been proved experimentally by Ramsay and Young.*
The values of ~— for a few gases may be found from the porous

plug experiments of Joule and Thomson. These experimenters deter-

• ^ ^- 1 dm , dm T ^ 1 T . „ , .

mmed directly, not --. — , but — — , and lound that in all their experi-
ments with a variety of gases and mixtures of gases that the latter
quantity was independent of the actual pressure of the experiment.

dm

That is, at constant temperature — — is a constant. We may write

from the gas law, neglecting the errors introduced by the deviations from
this law, which are of much smaller order than the errors of the experi-
mental results that we are using. Then,

* Zeit. Phys. Chem., I. 433.

LEWIS. FKEE ENERGY AND EQUILIBRIUM. 15

dv v'^ dp

We may now write for equation (24),

/i? „, A ^ RTdU

(2G)

v'^ dp

Now, for reasons that will be obvious immediately, we will write with

P 7?

perfect generality in place of -^('Oj _ ^ . » where F(v)

and f(v) are different functions of v. Then,

ET ETdm

^ V — fiv) v dip

Now since the last term is independent of the temperature, — ; — R T

dm. .
must be independent of the temperature. But smce — — is constant

dm,
at constant temperature, R T —— is also independent of the volume.

This quantity, therefore, is constant under all conditions to which the

system is subjected. Let us write, — R T —^ — a, and we obtain,

P = 77-^ - -2 ' (28)

^ V —f{v) v^

This equation, which is identical with the familiar formula that van
der Waals has developed from purely kinetic considerations, is here
shown to be directly deducible from a general thermodynamic equation,
with the aid of two simple empirical observations, namely, the constancy
of the specific heat of gases at constant temperature, and the proportion-
ality between the cooling effect and the fall in pressure in the free
expansion of gases.

Equation (28) does not define the nature of /(v), and in this respect
is less explicit than the corresponding term in the equation of van der
Waals, which is a constant, h. Nevertheless it must be borne in mind
that b, the volume correction in the formula of van der Waals, may only
be regarded as constant when the volume is larce, and that it also is in
reality an undetermined function of the volume.

16 PROCEEDINGS OF THE AMERICAN ACADEMY.

7 my

The constancy of the quantity, R T— — or a, we have only proved

in the cases of the gases which were experimented ujion by Joule and
Thomson, and through limited range of j^ressure. However, since the
constancy of this term was established for all the gases tried, we may
assume that the same result would have been obtained with any other
gases through the same limits of pressure. It cannot be taken for
granted that this constancy will hold at all pressures. In fact, the inter-
pretation given by van der Waals for the quantity a, as the interraolec-
ular attraction, would lead to the idea that it must be a function of the
volume. Otherwise the attraction between two molecules would be
independent of their distance apart. The variation of a with the vol-
ume will be mentioned later.

Van der Waals considered a priori that a would not depend upon the
temperature. But it is evident from page 14 that it is a constant only if

-~ = 0. Although the change of specific heat with moderate change of

volume is negligible, the change through wide limits of volume is proba-
bly always a measurable quantity, as will be shown later.

Since the quantity a has been defined as equal to — M T—-~ , we

dp

should be able to determine its value from the data of Joule and

Thomson. A calculation which is in effect the same as this has already

been made by van der Waals,* who calculated from the values of a and

b, given by experiment, the cooling effect that should be observed in free

expansion. This cooling effect, in Centigrade degrees, for a change of

one atmosphere in pressure is given below, as calculated by van der

Waals and as determined by the averages of Joule and Thomson's

results : —

Temp. Cels.

Calculated.

Observed.

Air

17

0.265

0.259

ii

90

0.18

0.206

CO,

18

0.9

1.15

11

91.5

0.64

0.703

The agreement, although not close, is very satisfactory considering the
uncertainty of the experimental results.

2. Liquids. — The experimental proof of the continuity between the
liquid and gaseous states shows that we must regard them as identical,

* Kontinuitat der F. und G. Zustand, p. 116.

LEWIS. — FREE ENERGY AND EQUILIBRIUM. 17

and any formula that is universally applicable to one state must also be
applicable to the other. Thus all the equations already developed for
gases apply without change to simple liquids (having a constant molecu-
lar weight) under the restrictions already mentioned, namely, that in the

equations

RT rfm , RT a

■2 '

v—f{i-) dv ^ v — b V

—z — and — ; are independent of the temperature only when -~ = 0,

that a has not been proved to be a constant in all cases but may be a
function of the volume, and that b is not necessarily a constant with
changing volume, although it is in all cases independent of the temper-
ature.

Regarding the question of the variability of c^ with the volume in the
case of liquids we have no direct experimental evidence, but indirectly it
can be shown that c„ is practically constant, for Ramsay and Young
(page 14) found that equation (25), p = A T — B, applies to liquids

dc
as well as to gases, and this equation can only be true when — ^ = 0.*

dv

dc
If —r^ is not equal to zero, then the general equation (8) must be

used,

RT d^ dm^^ r-^i dc;

this, differentiated with respect to T, v constant, gives

dp _R d^ d'wi , ri dc, ,^ , dc.

+ / -— vr-f

dT V dv dv dT ' J T,T dv ^ dv

„. d-Wi. dc^ „ , ^^^

oince -z 777,= T — from equation (9),

dv dr dv ^ ^

^^^R_m+ rL,p,r=c+rp'-i, (29)

d 1 V dv J T^T dv J T^dv T ^ ^

where C is a constant; while from equation (25) ~rL= ^•

Notwithstanding this evidence for the approximate constancy of the
specific heat, the experimental work of Joly f on this subject seems to

* Compare Nernst, Theor. Chem., p. 202,

t Phil. Trans. Roy. Soc, 182A, 73; Proc. Eoy. Soc , XLVII. 218; LV. 390.

VOL. XXXV. 2

18 PROCEEDINGS OF THE AMERICAN ACADEMY,

show that the value of c„ in a gas does change considerably through wide
limits of volume. He has been the first to succeed in measuring directly
the specific heat of gases at constant volume. The values were deter-
mined by means of his differential steam calorimeter, a method which
seems to give very accurate and consistent results. The results showed
that the specific heat at constant volume could be expressed in the
following formulae,

For Air, c„ = .17151 + .02788 p,
ForCOa, c„=.1650 + .2125 p + .3400 p^

where p is the density in grams per cubic centimeter. According to
these formulae the specific heat at constant volume at atmospheric pres-
sure differs from that at infinite volume by only two hundredths of one
per cent in the case of air, and by three tenths of one per cent in the case
of carbon dioxide. Between the specific heats of the gases at atmos-
pheric pressure and in a highly compressed or liquid condition the
change is much greater. For example, the value given by the formula
for f„ in the case of carbon dioxide is about twice as great at the critical
volume and about three and one half times as great in the liquid at 0° C.
as the value for the gas at ordinary pressure. Further evidence of the
change of c„ between the liquid and gaseous condition will be given later.

In these variations in the specific heat we find the probable cause of
many of the deviations from the equation of van der Waals that have been
noticed. It may be found necessary, therefore, in order to obtain a more
exact equation of condition, to return to the more general equation (8),

V J 'v-T do

do dv

7 ^t*

in which the value of —^ contains not only the function for volume

dv ■^

correction, but also a term depending upon the value chosen for the lower
limit of integration, 7{) . If we write

^^-p^Fiv), then ^-F{v)T=—^-—T-r{o)T,
dv V V —f{v)

where y(y) denotes the same function of v that has been used in equations

(27) and (28), namely, the quantity h in the van der Waals formula, and

dc^,
F' (o), another function of v. Now — — ^' is practically independent of

the temperature and the equation may be written

LEWIS. — FREE ENERGY AND EQUILIBRIUM. 19

^ -,._/(,.) ^ T, dv ^ ^ dv

It is probable from the derivation of F' {v) that this is a function of the
same general nature as the preceding term, and will vary with the volume

d c p d c^, ,

in the same way that -— ^ does. Then, since the coefficient of -7— is

■' dv u I'

not a volume function, we may write, not as a complete expression, but
as an approximation one step nearer the truth than the equations pre-
viously obtained, the following isothermal equation of condition,

« = _-^^-^ + .^, (30a)

P V -/{!■) do ^ dv' ^ '

or, after the type of the van der Waals equation,

^==J^_4 + c^, (305)

^ V — b v' d V

where c is a third constant to be determined by experiment, and which
differs with different temperatures.

Equation (29) gives the corresponding equation isochore, which may
be written for moderate ranges of temperature,

IL^ C -^^\nT. (31)

dT dv ^ ^

These equations, (30) and (31), should furnish a more exact statement

of the behavior of gases and liquids than the equation of van der Waals

or the corresponding ones developed in this paper. Since the degree of

dc^,
accuracv is not known in the values of -r-^ at present determined, I

d 0

have not yet attempted an application of these equations.

Returning to the consideration of f(v) or h, it may be stated at once
that this quantity is always a function of the volume and decreases with
decreasing volume, for usually tlie total volume in the liquid state is less
than b calculated for the gaseous state, and if b were constant then
V — b would have a negative value, which would be meaningless, v — b
must always have a positive value.

The way in which the quantity a varies with the volume cannot be
predicted. If it represents the attraction between the molecules, then it
will vary inversely with the volume if the attraction between the molecules
varies inversely with the distance. But since the attraction observed is
probably the resultant of forces attractive and repulsive acting between

20 PROCEEDINGS OF THE AMERICAN ACADEMY,

the molecules, we cannot say a priori how it will change with the dis-
tance between the molecules. In fact it is not necessary to suppose that

a or — r — must always have a positive value. Joule and Thomson

found in the case of hydrogen a rise of temperature instead of the usual
cooling effect on free expansion. This would indicate a small negative
value of a, corresponding to a preponderance of repulsive force between
the molecules. Tlie unreliability of the experimental data, however,
precludes certainty on this point.
In place of the equation

RT dm. RT

a

.■2 '

-f{v) dv ' ^ v — b

the equation of van der Waals, can be applied to liquids with the under-
standing that a and b are not constants but volume functions to be deter-
mined. In all liquids p is small compared with the other two terms.
When p = 0, if we represent the volume by Vq,

^ = ~,; (32)

but since the volume of liquids is only slightly changed by changing the
external pressure, Vq — b will not differ materially from v — b at
atmospheric pressure. We may write, then, as the equation for liquids at
atmospheric pressure,

(oo)

V — b

,2

From this equation may be found the values of a and b when the vol-
ume of a liquid is known at two different temperatures. From the
values thus found it should be theoretically possible to calculate the com-
pressibility of the liquid at constant temperature. Thus by differentiating
the van der Waals equation we obtain the reciprocal of the compressi-

bi%' dp RT ^a

dv {c-by^ v^'

practically this method fails on account of the fact that the difference
between the last two terms is very small compared with their total val-
ues, and therefore any error in either of these terms is multiplied euor-

d 7)
mouslv ill the determination of ~— .

•' do

The values of a and h obtained from equation (33) will be of service

LEWIS. — FREE ENERGY AND EQUILIBRIUM. 21

later in discussing certain relations between the liquid and gaseous state.
The way in which a .and b change with varying conditions of a liquid is
illustrated by the following values, calculated for fluor-benzol, from the
data given by Young.* For the values at high pressures/* is not negli-
gible comiDared with — , equation (33) will not give exact results,

and recourse must be had to the original equation of van der Waals.
V represents molecular volume in litres, p is expressed in atmospheres,
and b and a in units corresponding.

V

b

a

.097

.00075

11.6

.111

.00078

12.8

.127

.00081

13.5

.157

.00088

15.0

.225

.00110

17.9

.232

.00114

18.2

lese val-

270 the

By extrapolation from these val-
ues we find for v =
critical volume 270 .00126 19.7

From the critical data a and b are

determined to be 270 .00128 19.9

It is evident that the values of a and b obtained from the data for the
thermal expansion of a liquid are entirely consistent with those deter-
mined from the critical data. The change of b with the volume is well
shown by the figures, and is typical. Eegarding the variation in «, how-
ever, it is not possible to say whether it is due to change in volume or
change in temperature, or both.

It must be emphasized that in all the preceding work on gases and
liquids we are dealing with substances composed of a single molecular
species. Gases and liquids in which dissociation or association occur are
not within the scope of this section.

(B.) Heterogeneous Systems.

1. Gases and Liquids. — The simplest case of heterogeneous equilib-
rium in a system composed of a single molecular species is that between
a simple liquid and its vapor. We may apply to this case the general
equation of equilibrium (11 a). If we consider the specific beat at con-

* Phil. Mag., XXXIII. 153.

22 PROCEEDINGS OF THE AMERICAN ACADEMY.

stant volume the same in the liquid and its vapor, the equation assumes
the simpler form

RT\n^^ =PV- U- HT, (34)

where Vy and v<, represent the molecular volumes in the liquid and gase-
ous states respectively. From the previous section we know that the
term H will enter simply as a volume correction. The exact manner in
which it so enters may be best shown by finding the free energy of the
process of liquefaction from the work that might be done if the vapor
were compressed isothermally and continuously until it reached the liquid
condition. This work would be equal to J^j) dv^ and ■p can be found in
terms of v from the equation of condition which holds good throughout
the process. From equation (27)

RT rfm

■f{v) dv

Therefore

= / pdv= I — dv- I dm

R T

The integration of -rj-z dv is only possible when the form of f{v\

v —f{v) ^ ^ '' ^ '

is known ; but since y(y) does not change greatly, and since it is only an

important part of the expression when v is small, it may be regarded as

a constant, and equal to the value oi f(v) in the liquid state, or b^. This

value may be found from equation (33). Then the above equation

becomes

■Vi — Oi
and for equilibrium from equation (11), A = P V.

R Tin "^ ~ f^ = PV- U. (35)

^1 — bi

Since bi is but small compared with v^, we may replace v„ — bi by Vo.
Since U represents the change of internal energy in vaporization, and
P Fthe external work, — U + P V will be equal to the ordinary heat of
vaporization per gram-molecule, including the external work. This whole
quantity may be designated by L. Then

RT\u—^=L. (36)

Vi — Ol

LEWIS. — FREE ENERGY AND EQUILIBRIUM. ^23

This equation of equilibrium, which is at the same time a simple for-
mula for the heat of vaporization, has been deduced for all cases in which
there is no change in the molecular weight or in the specific heat at con-
stant volume during vaporization. Let us, in the following table, com-
pare for a few substances the experimentally determined values of L with
those calculated from equation (36).

R is in such units that L will be in gram-calories, and equals 1.98.
The values of h^ are obtained from the coefficient of thermal expansion of
the liquid by equation (33). For the value of Vg, the molecular volume
of the vapor, in the cases in which they have not been experimentally
determined, I have used the volume which a gram-molecule of a perfect
gas would have under the same conditions. The error thus introduced
will be negligible, for a large error in v^ or in fi — \ will affect only

slightly the value of L.

Temp. C.

i-i

i-i

v\

L (ealc.)

L (obs.)

Ether .... 35

.1062*

.080

24.82

4,200

6,600 §

Benzol .... 80

.096 1

.065

28.87

4,800

7,20011

Methyl formiate . 40

.06271

.047

25.02 %

4,500

6,9001

Methyl propionate 80

.10471:

.077

27.50$

4,800

7,400**

The results are very interesting, the calculated values being in each
case about two thirds of the observed. The explanation of the lack
of agreement must lie in the failure of the substances to comply with
the conditions named above. We have no reason to suspect, in the
case of all these substances, the polymerization of the molecules in the
liquid state. We are therefore again led to the belief that the specific
heat at constant volume is, in general, different in the liquid and gaseous
states. Moreover, I think that it is possible to show this in another,
entirely independent way. The specific heat at constant pressure of a
substance is the sum of two quantities, one representing the internal
change in energy and the other the external, that is,

'dW\ ^fdv

c„ =

dT jp ' \dTyp

P[tiA^ (37)

* Kopp: .Tahresber. der Cliem., 1860.
t Pisati, Paterno : Jaliresber., 1874.
\ Young, Thomas : Phil. Mag., XXXIV. 508.
§ Brix : Pogg. Ann., LV. 841.
II Wirtz: Wied. Ann., XL. 438.
IF For 32.9^ Andrews: Pogg. Ann., LXXV. 501.
** Scliiff: Lieb. Ann., CCXXXIV. 338.

24 PROCEEDINGS OF THE AMERICAN ACADEMY.

where f -^ 1 and f -7-^ j represent the change of internal energy

and volume respectively, with a change of temperature at constant pres-
sure, P.

\dTJ^-\dTj, • \dv)XdT);

from (37),

^''~ \dT Jp

(

dm\ f dv

dv J x\d T I p

^^\ -c -p(^

dv
~dTjp

(dm\ _p(d_v\
\dv)^ \dT);

dv\ r„ , (d'^\
a T

(38)

C„ — Cp

We have found for a gas or a liquid, -y = — . Substituting in (38),

Since ( y— , j is the measure of thermal expansion, equation (39)

contains only quantities capable of experimental determination, and we
are able to calculate r,, from the experimentally found value of Cp. The
calculation is further simplified by the fact that in the case of most va-
pors — is negligible compared with P, and in liquids P is negligible

compared with — . For close approximations, therefore, we may write

for liquids,

a ( dv\ ,, s

and for vapors,

c^ = cp-p(^) =Cp-R, (406)

dT,p

that is, the equation for the specific heats of perfect gases applies to

LEWIS. — FREE ENP^RGY AND EQUILIBRTUM. 25

vapors in which -j is negligible compared with the atmospheric pres-
sure.

The molecular specific heats at constant pressure of ether in the liquid
and gaseous states are 40.48 and 31.67 respectively. At 30 degrees, the

coefficient of thermal expansion for the liquid is .000163; v = .10o4

a
litres; a, calculated by means of equation (33), in such units that — ^ is

in atmospheres, is equal to 10.84. From these data — [ -TTp] — .1590

in litre-atmospheres, or 3.843 in gram-calories. Subtracting the latter
value from 40.48, we obtain 36.64 as the molecular specific heat at con-
stant volume of liquid ether. From the value 31.67, subtracting the
value of i?, 1.98, we obtain 29.69 as the molecular specific heat at con-
stant volume of ether vapor. The difference between these values is far
greater than could be explained by experimental errors.

It is interesting to see whether an explanation of such variations in
specific heat at constant volume can be found from the kinetic point of
view. We must believe that the energy imparted to a substance for an
increase in the progressive motion of its molecules corresponding to a
definite rise in temperature must be independent of the conditions of the
substance. If, however, the heat energy of a body is due not only to
the progressive motion of its molecules, but also to some additional mo-
tion such as the vibration or rotation of molecules, then the energy given
to the body in raising its temperature would be in part used in increas-
ing the progressive motion and in part in increasing the secondary mo-
tions of vibration, rotation, etc. The quantity of energy required for the
latter would not necessarily be independent of the volume of the body,
but might depend upon the proximity or rate of collision of the molecules.
If, however, a body were composed of molecules incapable of any except
progressive motion, we should predict absolute constancy in the specific
heat at constant volume. Mercury vapor is believed to be such a body,
and if liquid mercury, as seems probable, also is composed of monatomic
molecules, the value of c^. in the two states should be identical. Unfor-
tunately the specific heat of mercury vapor is unknown, and we cannot
test the correctness of this supposition directly. But if it be true, equa-
tion (36) should give a correct result for the heat of vaporization of
mercury. A calculation similar to those whose results appear on page
23 gives, for T= 623, v^ = .01566, b^ = .01419, v., := 58.7, L = 13,070.
The value experimentally observed by Person (1846) was 12,400.*

* Comptes Eendus, XXIII. 162 et seq.

26 PROCEEDINGS OP THE AMERICAN ACADEMY.

The agreement is satisfactory and within the limits of experimental
error.

2. The Vapor Pressitre Curve. — The change with changing temper-
ature of the equilibrium between a vapor and its liquid or solid phase
may be obtained from equation (20) ; for if we consider the vapor to
obey with sufficient exactness the laws of perfect gases, the system under
consideration is one for which equation (20) was developed, since it con-
tains only a gaseous phase and a condensed phase of definite composition.

From equation (20), — ^- = -^^ .

In K = In V — In ?/,

where v is the molecular volume of the vapor, v' that of the solid or

,. ., din v' . ^ . ^ d\n v

liquid. „ is entirely negligible compared with , and equa-

tion (20) becomes d\nv U

~dT ^ 'RT^ '

hince V =^ , where p is the vapor pressure,

In y = — In JO + In ^ + In T,

and dlnv = — d\np + — d T.

^ , . . d\np U 1 -U+ RT
Substituting, -j-^' ^ — -^-^ + 7^ ^ iej-^ — •

As on page 22, let — U + R T = L,

din p L

then __ = __,

which is the familiar expression for the vapor pressure curve.

A complete expression for the vapor pressure in the case of liquids
to which equation (36) applies, if the vapor approximates a perfect gas,
may be derived as follows :

1 '^' — ^
V^^'~R~T'

R T

Substitute for v, and

V

RT L RT (±\

In :^ . or z;: aXR T/

^"^(y'-Z,') i?r °^ p{v'-b') ^

RT
whence p

(y' - b') e(^)

LEWIS. FREE ENERGY AND EQUILIBRIUM. 27

III. DiMOLECULAR SYSTEMS.

Sohitio7is. — The simplest case of a system composed of two molecular
species is one in which the two species are chemically neutral, only acting
on each other by the process of solution.

Before taking up such systems it will be convenient to consider briefly,
in the light of the results of the last section, a question concerning the
general equation of equilibrium, namely, the nature of the function If.
In equation (5), which expresses the free energy of a simple constituent
of a system, the quantity P^i entered as an integration constant, and
nothing was known regarding its nature except that it must be indepen-
dent of the temperature. After differentiating equation (5), it was found

that , was a function which could represent in the later equations

dvi

the volume correction corresponding to the quantity b in the equation of
van der Waals. Since this was true in the widely differing states of gas
and liquid, it is probable that in any state in which the molecule itself is

d ?^

not changed — — may be expressed as such a volume correction, rep-
d i'l •

resenting the diminution of the space available for the free motion of the
molecules, due to the space actually occupied by the molecules themselves.
If, therefore, we subtract from ^^ the term representing the volume
correction, there will remain a quantity which will be constant under all
conditions when the molecule itself does not change, and whose value will
dejjeud only on the nature of the substance considered. We will use |^
hereafter to denote this quantity. The volume correction enters in the
most general way in the consideration of a phase containing a number of
molecular species. When we consider each species, the volume must be
corrected for the volume actually occupied by its own molecules and also
for that occupied by the other molecules present. The volume with the
first correction may be expressed as in the van der Waals equation by
V — b, where b is the correction due to the space occupied by its own
molecules. The volume may be corrected for the space occupied by the
other kinds of molecules by multiplying the actual volume by a factor, r,
representing the fraction of any volume of the mixture which is left
available for the free motion of the molecules of the particular species
under consideration. The nature of this species should have no effect on
the quantity r. We have, therefore, for the corrected volume of a sub-
stance dissolved in any mixture the value r (v — b), where r depends
solely on the nature of the solvent, b on that of the solute.

28 PROCEEDINGS OF THE AMERICAN ACADEMY,

In taking up the subject of solutions it will be necessary to consider
their probable nature. Although there is some evidence that the phe-
nomenon of solution is accompanied by a molecular change, the prepon-
derance of evidence seems to be in favor of the theory that the molecule
of a substance in solution is free, and not combined in any chemical vpay
with the molecules of the solvent. If this is the case, certain properties
of the solute should remain unchanged reirardless of the nature of the
solvent. In the following work we will assume that, for a dissolved
substance, the quantity f) as defined above and the quantity c„ will be the
same in any solvent. If this supposition, which seems eminently proba-
ble, proves to be not entirely correct, then the equations developed below
will only be approximations to the truth. It is to be hoped that in any
case their application will conversely give us information concerning the
nature of solution.

Osmotic Pressure. — The simplest phenomenon of a homogeneous so-
lution is that of osmotic pressure, which may be determined in the fol-
lowing way. The change of free energy on addition of an infinitesimal
amount of solvent to a solution containing one gram-molecule of solute is
equal to the sum of the changes of free energy in the two constituents of

the system. From equation (5), if —-^ — 0, and -~ becomes the

volume correction, which, as is shown by Nernst,* is for each constituent,
in the case of osmotic pressure, only the correction for the space occupied
by the molecules of the constituent in question,

— dvx — dWii -\ r

t\ — 6>i Co — O2

where the subscript 1 refers to the solute, subscript 2 to the solvent.

dv-^ , the change in molecular volume of the solute, is also the change in

volume of the system, and d A = p dvi, where p is the osmotic pressure.

Therefore

RT dm. , nRT do^ .^^.

7'i — Oi dvi fo — 0.2 dvi

dW.

where --, — represents the heat produced by the addition of d v^ of the

solvent. Except in cases of solution of such great concentration that the
molecular volume of the solute and that of the solvent are not greatly

different, -^ is entirely negligible, and the equation for osmotic pres-
sure becomes »

* Theor. Chem., p. 209. (References to tliis book are to the first edition.)

dA = -^^^ dv, - dUr + r- dv^ - dW^^ ,

LEWIS. — FREE ENERGY AND EQUILIBRIUM. 29

^ V — 0 dv

Except that - — represents the heat of dilution instead of the heat of
^ dv

free expansion, equation (42) is identical with the equation obtained for
gases. The comparison of this formula with experiment is not possible
with the experimental data at present available, since, as far as I know,
the heat of dilution has been determined only in the case of electrolytes,
and in these cases a complication is introduced, due to the heat of disso-
ciation. An abstract * has just come to hand of a recent work by
Kistiakovskijt in which he shows that the lowering of the freezing point,
which is proportional to the osmotic pressure, is, in a concentrated solu-
tion, equal to the lowering calculated for an ideal solution plus a tei'm
that is proportional to the heat of dilution. This is the result that would
be given by equation (42). Kistiakovski shows that there is perfect
agreement between the lowering of the freezing point calculated in this
way and that found by experiment. It seems questionable, however,
whether his application of the formula to electrolytes, neglecting the heat
of dissociation, is justifiable. Unfortunately I have not access to his
original paper.

The osmotic work obtainable from the change of a gram-molecule of a
solute from a solution of one concentration to one of another concentra-
tion may be found by direct application of equation (6), or by integration
of jO dv, where p may be expressed in terms of v by means of equation
(42). Then

^ = i?nn^^?^-f U. (43)

For all except very concentrated solutions b may be neglected, and

A = E T\n^^+ U. (43a)

V

1

Cady % has recently shown that in a galvanic cell in which the total
result of the current is the jiassage of a certain amount of a metal from
an amalgam of one concentration to one of another,

IT = In \- q ,

n eg Wi

* Chem. Centr. Blat., 1899, I. 89.

t Jour. Russ. Phys. Chem. Ges., XXX. 576,

t Jour, of Phys. Chem., II. 551, 1898.

30 PROCEEDINGS OF THE AMERICAN ACADEMY,

where tt is the electromotive force, Vi and v^ are the moleciiLar volumes

of the metal iu tlie two amalgams, and q is the heat of the process in

electrical units. From this,

v
neoTT = B Tin ~ + U,

but n Pq -t is the electrical work per gram-molecule, which is equal to
the change of free energy, since the cell is a reversible one. Therefore
yl =r n Co TT, or

^ = ^rin^+ u,

which is identical with (43 ci).

When we consider solutions of all concentrations, varying from the
state where one of the constituents of the phase is in great excess to the
state where the other constituent is in great excess, as, for example,
when water is added continuously to a definite amount of alcohol, then
the form which the osmotic pressure curve assumes is very complicated.

Here equation (41 ) must be used, and -— ^ and — — will both be com-

d i\ a Vi

plex functions of i\. may be looked upon as the sum of two quanti-

ties, one due to the attraction of unlike, the other to the attraction of like
molecules. Concerning the manner in which the former 'will change
we are ignorant. The latter, however, according to reasoning exactly

similar to that which led van der Waals to the term — , may be
shown to be inversely proportional to the square of the volume, or equal

to -^ — 7, — ^ . We see from this that equation (41) is at least of the

'11-

third degree in v\. Bredig * and Noyes f have each proposed a general
formula for osmotic pressure based upon kinetic reasoning. Both these
equations are of the third degree in v. The osmotic pressure curve
represented by equation (41) is not necessarily, therefore, single valued.
There may be more than one volume corresponding to one osmotic pres-
sure. This is a further analogy between solutions and gases. In fact,
a luimber of cases are known in which the osmotic pressure may be
shown to be the same at two different concentrations, namely, the cases
of liquids that are mutually soluble to a limited extent, thus forming two
phases in equilibrium with each other. It is evident that in order to

* Zeit. Phys. Chem., IV. 444. t Zeit. Pliys. Chem., V. 53.

LEWIS. — FREE ENERGY AND EQUILIBRIUM. 31

preserve the equilibrium the osmotic pressure, not only of one but of each
of the constituents, must be the same in the two phases. No similar phe-
nomena have ever been observed in the case of solids dissolved in liquids,
but it seems not impossible that such may be found. Then a solid might
have two different solubilities in a solvent at one temperature, correspond-
ing to two concentrations in which the osmotic pressure would be equal
to the solution pressure.

Distribution of a Solute between two Solvents. — The equation of equi-
librium when a substance is distributed between two solvents may be
found directly from equation (11a), simplified by the considerations
advanced on page 27, namely,

P V= E T\n'''['''~^2 + ^. (44)

n (vi — b)

In all cases of this sort P Fis entirely negligible, and

r,{ih-b)^ ^

where ^2 and Vi are the molecular volumes of the solute in the two sol-
vents ; b is the volume correction for the solute molecules ; r^ is the
volume correction for the first solvent, and 1\ that for the second. U is
the heat given off when one gram-molecule of the solute passes from one
solvent to the other. It equals the difference between the heats of
solution of the solute in the two solvents. In all ordinary solutions b is
negligible, and the equation becomes

i?rin^-'-j- U=0. (46)

Since this is the general equation of distribution of a substance between
two solvents, it will hold true in the special case in which the solutions
are in equilibrium with the solute in the solid form. If we represent by
Si and 8-2 the solubilities in gram-molecules of the solute in one litre of

each of two solvents, then s^ = — and s, — — , and equation (44)
may be written

i^rin'^'-f ?7=0; (47)

or if TJ' is the heat of solution in the first solvent, TJ" in the second,

R T\u '^ + U' - U" = Q,

32 PROCEEDINGS OF THE AMERICAN ACADEMY.

or RTlu^-^^ U' ^RTln'-^ + U". (48)

This equation permits the calculation of the solubility of a substance
in any solvent if the solubility in any other solvent is known and the heat
of solution in each solvent, or the difference between these heats of solu-
tion, and if the ratio of the volume corrections for the two solvents is
known.

The heat of solution may be found from the change of solubility with
change of temperature by the well known equation

dJ-l-J^^ .49)

which may be obtained by direct application of equation (20), since the
conditions for which (20) was obtained are all fulfilled in this case, and
since the molecular volume, t"o, of the solid may be considered constant.

K = VqS ; In K = In Vq + In s ; d In K = d\n s.
Uniting equations (-18) and (49),

RT\u'^ + RT^'^ = RT\n'^+ RT^"^-^,

Ti d I I'n dT

a In —
or 1d?-1 4- 2^^^_ln!i = 0. (50)

Si d 1 To

When the values of r are known for the various solvents, if we know the
solubility of a substance in two solvents and the temperature coefficient
of the solubility in one, we may find the corresponding coefficient in the
other solvent. I hope soon to verify these formulas experimentally, and
to determine the values of r for some common solvents. It will be inter-
esting to see how r compares with the value of v — h found for the
solvents by equation (33).

IV. Polymolecular Systems.

(^.) Homogeneous Systems.

In accordance with the considerations advanced on page 27, the general
equation of equilibrium (11 a) may be put into the form

[r (I'l — 6i)]"i ... J To T

LEWIS. — FREE ENERGY AND EQUILIBRIUM. 33

where h is the sum of all the fj terms, and ( (7„i — C^.^ and h have the
same value regardless of the nature of the solvent. If the total change
in the number of molecules be n, then the quantity r will occur to the
power of n, and

(I'l - b^yh . . . J T, 2

Only in the case of equilibrium between gases is P K considerable, for
other cases P Vmaj be neglected. For equilibrium in liquid phases, if
we represent by k the equilibrium ratio with the volume corrections

/>T Q (J
"' „ — - dT by y)

then {R Tlnr'^k) + U + fT= 0, (52)

or (R Tin k)=-R Tin r" - U-fT. (52a)

We are now iu a position to answer the question how equilibrium is
influenced by the nature of the solvent. If we write for equilibrium in
two solvents two equations of the form of (52 a),

(R T\a k') =-RT\n /" - U' -fT,

(R Tin k") =-RTln /'" - U" -fT,

and subtract, we obtain, since _/ is the same in the two solvents, according
to the assumption made on page 28,

In V77 = — « lu ^ —
fc r

( u' - n"\

and we find that the condition of equilibrium depends on the values of
r and the heat of the reaction in the two solvents. Ordinarily when tlie
solvent does not enter into the reaction, the values of {i\ — b^), etc. may
be replaced by v^, etc., and k, the corrected equilibrium ratio, may then
be replaced by K, the ordinary equilibrium ratio. If we are dealing
with reactions in which the original and final number of molecules is the

/

same, or with any case where n In -j-, is negligible, the equation becomes

lnK'-lnK-= ^ ^ ^^ . (54)

In such cases the difference between the logarithms of the equilibrium
ratios in any two solvents at a given temperature is equal to the differ-
ence in the heats of the reaction divided by the gas constant and by the
absolute temperature. I know of no case in which the experimental

VOL. XXXV. — 3

34 PROCEEDINGS OF THE AMERICAN ACADEMY.

data are at present sufficiently complete to permit the testing of equa-
tion (54).

V. Application to Electro-Chemistrt.

Since the electrical work of a reversible cell is equal to the change of
free energy of the process taking place in the cell, the calculation of
electromotive force in many cases is possible by means of the general
equation (6).

Tiie general formula for the electromotive force of a concentration cell
has already been given on page 29.

Let us consider next the subject of a single potential difference be-
tween a metal and an electrolyte containing the ions of that metal. If
^1 be the free energy of one gram-molecule of the metal, and '^o that of
one gram-molecule of its ions, then the electrical work in the electrolytic
solution of one gram-molecule is

Now for '^2 we obtain from equation (5), modified according to page 27,

^, = -R T\nr(v, - h) - T r^-^dT+ l),T + Wi,,

where r is the correction for the particular solvent in which the ions are
dissolved, b^ the correction for the ions themselves, which will always be
negligible, c^.^ and |^2 are independent of the nature of the particular
solvent. Win is the internal energy of the ions, and if we represent by
5Ei the internal energy in the electrode, and by Uihe change of internal
energy in electrolytic solution, U= Wii — Wi2, and — SIo ™ay be re-
placed by U ■}- W.I- Then since ^i, 2Ei, c^^, fj2j are constants at con-
stant temperature,

A= C-{- (RTlnrv.) + U,

where C is a constant including the various terms mentioned above. If
we replace Vz by , where p is the osmotic pressure of the ions, we

may write

A= C-^ (RT\nBT)-RT\nl+ U=c-RT\n^+ U.
^ ^ r r

If for convenience we write c = R Tin P,

A = -RTln~^+ U,

LEWIS. — FREE ENERGY AND EQUILIBRIUM. 35

or, since ^ = v Cq tt, where v is the valence of the ion, Cq the electricity
carried by a gram-molecule of a univalent ion, and tt the electromotive

force, then RT^p U

17 = In-^-f — , (5o)

in which P differs from the electrolytic solution pressure of Nernst in
that it is at constant temperature the same, no matter what the solvent
may be in which the ions are dissolved, while the value of the Nernst
solution pressure holds good only for water solutions, r represents the
particular volume correction of the solvent.

An interesting type of cell is one in which two similar electrodes are
in contact with solutions of an electrolyte containing the electrode ion in
two different solvents ; as, for example, zinc, zinc sulphate in water, zinc
sulphate in alcohol, zinc. When a current passes through this cell, the
total change consists in the transfer of zinc sulphate from one solvent to
another. The free energy change in a cell of this type may be found
from equation (6), modified as in equation (46),

A = RT\n'^^,+ U, (56)

where i\ and v/ are the molecular volumes of the positive and negative
ions respectively, in the first solvent ; v,, and v^', in the second solvent.
Now v/ =^ g Vi , where ^ is a whole number or a simple fraction ; also,
V2 = g V2- Therefore,

^ = i?7^1n^V^+ U=2RTln'^^+ U. (57)

If ra represents the number of gram-molecules transferred from one
solvent to the other when the quantity of electricity, v^o, passes through
the cell, then

veoTT = 2 ;» i? T'ln-^' -f m U.

Let m U ^= q, then

. = ^i?rin^^ + ^. (58)

From the equation of Helmholtz,

and comparing (58) and (59) it is evident that

ll = l^E\n'-^\ (60)

a I V Co riVi

36 PROCEEDINGS OF THE AMERICAN ACADEMY.

In an investigation which I am now making on cells of the above type
I have attempted a verification of these equations in the following way.

7*

Choosing two solvents in which the value of In — may be neglected,
namely, water and a mixture of alcohol and water, then if t'g be made

d TT

equal to Vy, -pj^ should equal zero. Unfortunately the dissociation in

alcohol-water mixtures of the salts that are available for our purpose
has not hitherto been determined. If the two solutions are made up
with equivalent amounts of the original salt, then the concentration of
the ions in the alcohol-water solution will be less than that in the water
solution on account of the greater dissociative power of water. In the
following cells, made up in this way, we should expect, therefore, a small
temperature coefficient, and moreover, since the electrode in contact with
the water solution is found to be negative, this temperature coefficient
should be negative.

The following table gives the results obtained for the cells : —
(1) Zinc; zinc sulphate, tenth normal, in water and fifty per 'cent ethyl
alcohol. (2) Zinc ; zinc sulphate, tenth normal, in water and fifty per
cent methyl alcohol. (3) Cadmium; cadmium sulphate, tenth normal,
in water and sixty per cent ethyl alcohol. (4) Thallium; thalliiia
sulphate, hundredth normal, in water and twenty-seven per cent ethyl
alcohol. (5) Thallium ; thallium sulphate, hundredth normal, in water
and forty per cent methyl alcohol.

(1)

Temp. C

30
0

JT

.043
.047

dn
d T

—.00013

(2)

23

0

.0475
.0480

-.00002

(3)

28
0

.046
.049

—.00011

(4) A

23
0

.0212
.0212

.00000

B

24
0

.0216
.0218

-.00001

C

30
0

.0242
.0246

-.00001

(5)

24
10

.0380
.0380

.00000

LEWIS. — FREE ENERGY AND EQUILIBRIUM. 37

In order to determine the actual dissociation in the above cases con-
ductivity determinations were made. I found in the case of zinc and
cadmium sulphates that the degree of the dissociation could not be found
from conductivity determinations on account of peculiarities which will
be discussed in a later paper. It was found, however, that in the case
of cadmium sulphate the dissociation is at least five or six times as
great in water as in fifty per cent methyl alcohol. In the first three cases,

therefore, the value of — is undoubtedly great enough to account for

the values of -77=, found.
a 1

On the other hand, in the case of thallium sulphate it was found possi-
ble to determine the degree of dissociation from the conductivity data.
In dilute solutions the dissociation in water and in fifty per cent methyl
alcohol was found to be practically the same. This is in complete agree-
ment with the result in case (5), where the temperature coefficient was
zero. In case (5), then, the only one in which all the data are available,
equation (60) is thoroughly verified. I hope to publish soon more com-
plete results on this subject.

I wish to express my deep obligation to Professor Theodore TV.
Richards for his encouragement and friendly criticism of this work.

Summary.

I. (a) A general equation for change of free energy is developed.

(b) From this is derived a general expression for physico-chemical
equilibrium in homogeneous or heterogeneous systems, which includes as
special cases the law of isothermal mass-action and the laws of constancy
of distribution coefficients among several phases.

(c) For change of equilibrium with change of temperature a formula
is derived of which the equation of van't Hoff" is a specialized form.

II. (a) The application of the general equations to gases yields an
equation of condition which with the aid of two familiar empirical obser-
vations is shown to be identical with the equation of van der Waals.

(h) This equation of condition is applied to liquids in detail and special
cases are discussed.

(c) A more complete equation is proposed, recognizing the variability
of specific heat with changing volume.

(d) From the general equation a formula is obtained for equilibrium

38 PROCEEDINGS OF THE AMERICAN ACADEMY.

between a liquid and its vapor. Heats of vaporization are calculated
from this formula.

(e) The formula is inapplicable in cases where the specific heat at
constant volume differs in the liquid and gaseous state. A method of
calculating these specific heats is given.

(y ) The formula applies perfectly in the case of mercury.

{g) The vapor pressure curve is discussed.

III. (a) Application of the general equations to solutions leads to
simple expressions for osmotic pressure and osmotic work in concentrated
solutions.

(h) Equations are given for the distribution of a solute between two
solvents, and for the relation between the solubility curves of a substance
in different solvents.

IV. The influence of the nature of the solvent upon general homoge-
neous equilibrium is determined and formulas are given.

V. («) The general equation for free energy is applied to electro-
chemistry. For the single potential difference between a metal and an
electrolyte an equation is proposed which is an amplification of the equa-
tion of Nernst.

(6) Galvanic cells in which two solvents take part are discussed.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. Xo. 2. — August, 1899.

SHORT STUDIES OF NORTH AMERICAN TRTXALIN^.

By Samuel H. Scudder.

SHORT STUDIES OF NORTH AMERICAN TRYXALIN^.

By Samuel H. Scudder.

Received June 6, 1899. Presented June 14, 1899.

The review of a large series of Tryxalinaj collected on the Pacific
coast iu 1897 by Mr. A. P. Morse, and kindly placed unreservedly in
my hands, has provoked a re-examinatiou of the sj^ecies in a number of
genera scattered through the group. I have published elsewhere (Can.
Ent., XXXI. 177) a review of our species of Orphulella, and here gather
together other miscellaneous studies, all referring to the Tryxalinoe.

1. The United States Species of Mermiria.

A recent study of our Mermiria^ has brought to light a couple of new
species of Mermiria, and some slight extension of the known range of
some of the other species, so that I venture to publish the followiiio'
notes and descriptions, with a new table of the species, based primarily
on that published by McNeill.

Table of the United States Sjyecies of Mermiria.

fli. Head shorter than pronotum, or, if (rarely) as long, then the greatest width of
the fastigium is greater than its length beyond narrowest part of vertex ; last
ventral segment of male bluntly acuminate.
ii. Fastigium less prominent and blunter, its greatest breadth being consider-
ably greater than its length beyond narrowest part of vertex, especially in the
female.

c^. Stouter, the hind femora shorter, not reaching the tips of the tegmina in
the female ; disk of pronotum, in female, hardly or not more than twice as
long as greatest breadtii ; head with a broad occipital fuscous band.

texana Brun.
C.2 Slenderer, the hind femora longer, reaching the tips of the tegmina in the
female; disk of pronotum distinctly, generally much, more than twice as long
a- greatest breadth ; head with a narrow occipital band or none.

hivittata Serv.

b-. Fastigium of vertex more prominent and angulate, its greatest breadth

being scarcely greater, even in Uie female, than its length beyond narrowest

• part of vertex ; disk of pronotum considerably more than twice as long as

greatest breadth intertexta sp. nov.

42 PROCEEDINGS OF THE AMERICAN ACADEMY.

a-. Head as long as pronotura ; fastigium at least as long beyond narrowest part
of vertex as its extreme breadth ; last ventral segment of male more produced
and somewhat acutely acuminate.

b^. Head with a relatively narrow or no occipital median fuscous stripe, which

never nearly equals the width of the fastigium.

c^. Fastigium triangular, the sides converging in a nearly straight or onlj'

slightly curved line, the tip narrowly rounded alarris Scudd.

C-. Fastigium semielliptical, the sides converging with a well rounded curve,
the tip very broadly and bluntly rounded neomexicana Thorn.

h-. Head with a broad fuscous occipital stripe, almost or quite as broad as the

greatest width of tlie fastigium.

ci. Fastigium semielliptical and strongly rounded apically, the sides well

curved • . vir/ilans sp. nov.

C-. Fastigium acutely triangular, with the sides straight and the tip hardly
blunt rosti-atu McNeill.

Mermiria texana Brun.

I have seen no male of this species. My specimens come from Col-
orado, 7UU0', Morrison, and Coahuila, Mex., Palmer. It was originally
described from Texas and the State of Durango, Mex. McNeill also
credits it to Arizona.

Mermiria bivittata (Serv.).

I am inclined to think that Bruner's 31. macullpennis must be regarded
as at most only a variety of this species ; I have seen it from Texas and
Colorado only, and all are females. McNeill accepted it with doubt, and
my inaculate specimens, including one named by Bruiier, vary in the
breadth and arcuation of the fastigium to the same extent as do those
undoubtedly to be referred to bivittata.

This is our commonest species, and is wide spread. From east of the
Mississippi I have specimens from Georgia only, but numerous speci-
mens from Nebraska, Kansas, Colorado, Utah, Arizona, New Mexico,
and Texas, and some immature specimens, which may belong here, from
Iowa, collected by Allen. McNeill also credits it to Virginia, Illinois,
and Nevada. Morse found it common in New Mexico on bunch grass,
but it was rather shy and flew freely.

Mermiria intertexta sp. nov.

In color, markings, and size this species exactly resembles the preced-
ing ; in all specimens seen, however, there is a slender mediodorsal
fuscous stripe on head and pronotum. The head is distinctly shorter
than the pronotum, and the fastigium in both sexes is scarcely broader

SCUDDER. — NORTH AMERICAN TRYXALTN.E. 43

than its length beyond the narrowest part of the vertex, being triangular
( (^ ) or suhtriangular ( 9 )> ^'hh straight (^) or arcuate (9) sides and
blunt apex, the margins slightly ascending, and with faint or no median
carina. The median carina of the pronotum is pronounced throughout,
the lateral cariuce feeble, and the disk of the pronotum considerably more
than twice as long as broad, especially in the male, the transverse sulci
rather feebly impressed. The tegmina reach about to the tip of the
abdomen and are immaculate, with the base of the median area in-
fuscated and bordered by a submarginal costal flavous streak, as fre-
quently in M. bifittata. The hind femora are slender, and reach as far
back as the tegmina in both sexes. Last ventral segment of male a
little more elongate and gradually acuminate than in M. bivittata.

Length of body, $, 38 mm., 9 56 mm. ; pronotum, ^, 6 mm., 9,
8 5 mm. ; tegmina, ^, 27 mm., 9, 39 mm. ; hind femora, ^ , 2L5 mm,,
9,31 mm.

2 $ , 2 9 • Georgia, Morrison ; P^agle Pass, Tex., Schott.

Mermiria alacris Scudd.

I have seen specimens only from Georgia, Morrison ; Sandford, Fla.,
P'razer ; and Dallas, Tex., Boll.

Mermiria neomexicana (Thom.).

My specimens come only from Pueblo, Col., Aug. 30, 31 ; Dallas,
Boll, and Bosque Co., Tex., Belfrage. McNeill says it occurs from
Wyoming to New Mexico and eastward to Georgia. I suspect his
Georgia reference may be due to my remarks in describing M. alacris,
where I speak of the present species as occurring in Georgia, but I now
think that was a mistake.

Mermiria vigilans sp. nov.
Dull olivaceous with purplish and flavous markings, particularly the
former. The head has a broad mediodorsal purplish fuscous stripe,
broadening a little in passing backward, and as broad, at least poste-
riorly, as the fastigium ; this is separated by a narrow flavous stripe
from a broad and equal light purplish postocular stripe which continues
over the upper part of the lateral lobes of the pronotum ; below this the
genfe are olivaceous clouded with flavous, while the face is infuscated.
The pronotum is dull flavo-olivaceous, the disk often with a faint pur-
plish median stripe, and the lateral lobes narrowly margined below with

44 PROCEEDINGS OF THE AMERICAN ACADEMY.

purplish, the latter color in the male often suffusing the wliole ; tegraina
green, the median area, especially near base, more or less ferruginous ;
hind femora flavous, infuscated above ; hind tibife dull red.

Fastigium distinctly longer beyond the narrowest part of the vertex
than the extreme breadth, semielliptical, the sides straighter in the male
than in the female, very bluntly rounded at tip, especially in the female,
the margins scarcely ascending but plane, while the centre is rotundate
with no sign of median carina. Disk of pronotum about twice as long
as broad, the median carina sharp throughout, the lateral carin^e distinct
but not elevated, the posterior margin very broadly rounded. Tegmina
7'eaching the tip of the abdomen. Hind femora ver}' slender, reaching
the tip of tlie tegmina. Last ventral segment of male short but unusually
acuminate.

Length of body, ^ , 36 mm., 9, 48 mm. ; pronotum, ^ , 5 mm., 9,
7.25 mm. ; tegmina, ^, 2o mm., 9) 36 mm. ; hind femora, J* 20 mm.,
9, 26 mm.

2 (? , 4 9 . Smithville, N. C, Nov. 22.

In markings this species seems to bear a close resemblance to M. ros-
trata, which I have not seen, but differs from it as it does from M. alacris
in the form of the fastigium. It is most closely allied structurally to
M. neomexicann, but has a longer pronotum and a more pointed genital
segment in the male, while it differs to a considerable decree in color
and markings ; these, however, are variable in both species.

Mermiria rostrata McNeill.
This species, known only from Indian Territory, I have not seen.

2. ACENTETUS AND ITS SpECIES.

This genus was founded by McNeill (Proc. Dav. Acad. Nat. Sc, VI. 225)
on Acentefus unicolor, a species described by him in the same paper from
a single male without antennie, taken in Colorado. I have a single male
of the same species, taken by me at Lakin, Kansas, on Sept. 1, which
agrees with McNeill's figures and description except that the whole upper
surface of the head and pronotum is blackish fuscous, the genoe are nar-
rowly striped with pale flavous or luteous and pale fuscous, and the
lateral lobes have similar luteous stripes on a pale fuscous ground ; the
contrast of the dark disk and lighter lateral lobes does not well suit
the name unicolor. The anteiinaj, the description of which had of course
to be omitted from the generic characters, are depressed subfiliform, not

SCUDDER. — NORTH AMERICAN TRYXALIX^. 45

tapering, and distinctly longer than head and pronotum together. The
scai)ular area of the tegmina, as represented in McNeill's figure, is too
narrow ; at its widest, opposite the nexus of veins in the median area, it
is nearly one third the total width of the tegmina at this point.

I have a second species of Acentetus, also represented by a single male
and also without antennae, taken by me at Florissant, Col., Aug. 17-22.
It is testaceous, marked with griseous and fuscous ; the head is testa-
ceous, with a pair of narrow, arcuate, diverging, fuscous occipital stripes,
and on each side a pair of similar but straight postocular stripes ; the
disk of the pronotum is griseous and the lateral lobes testaceous below,
passing rather rapidly into blackish fuscous above; the hind femora are
testaceous, more or less infuscated but not at all banded, and rufous
beneath, the hind tibiiB dull red. Length of body, 16.5 mm.; tegmina,
10.5 mm. ; hind femora, 10.5 mm.

It differs from A. imicolor not only in color and markings, but also in
several structural peculiarities : The median carina of the fastigium is
less pronounced and indeed rather feeble ; the lateral oarinae of the pro-
notum, though having much the same divergence, are continuous and
eqsal throughout, thus requiring a modification of the generic definition
as given by McNeill ; the metazona is much more coarsely and deeply
punctate ; and the tegmina (in the male of course) have a very different
form, the costa being very strongly arched in the distal half and the
distal portion of the tegmina being much abbreviated, so that the tegmina
as a whole are less than three times longer than broad, instead of five
times as long as broail, as in A. unicolor ; the relative breadth of the
scapular area is even greater than in that species. It may be called
Acentetus carinatus.

3. A Second Species of Opeia.

Opeia was founded by McNeill in 1S97 (Proc. Dav. Acad. Nat. Sc,
VI. 214) upon a single species, Oxycoryphus obscurus Thom. I have seen
numerous specimens of this species coming from the Yellowstone valley
in Montana, Nebraska, Lakin, Kans., Sept. 1 (Scudder), Colorado, 5500',
7000' (Morrison), Ft. Collins, Col., Aug. 12, 25, "on Bouteloria oligo-
stachya" (Baker), Garden of the Gods, Col. (Scudder), Silver City,
N. Mex. (Bruner), and Bosque Co., Tex., "on prairies" (Belfrage), as
well as from Ft. Whipple, Arizona (Palmer). According to McNeill it
is "a species peculiar to the Great Plains."

In 1897, Mr. A. P. Morse brought a second species from California.

46 PROCEEDINGS OF THE AMERICAN ACADEMY.

— Lancaster, Aug. 1, Kern City, Aug. 4, Tulare, Aug. 5, and Lathrop,
Aug. 17 (2/) J , 179)5 which may be called Opeia tesiacea.

It is a slenderer insect with longer tegmina, which equal ( 9 ) or sur-
pass ( 1^ ) the abdomen and reach the base of the geniculation of the hind
femora, and with a paler, nearly uniform coloring with scarcely any green
in it, and immaculate tegmina in the female, where at most they merely
have an obscure unbroken fuscous streak in the proximal half of the
median area, while in 0. obscura the female tegmina have the median
area nearly always distinctly marked with fuscous, broken into quad-
rangular spots ; in the latter species the upper half of the lateral lobes
is generally marked with a broad or narrow fuscous (rarely greenish)
stripe, in both sexes, extending in extreme cases upon the head as a
postocular band ; this is extremely rare in 0. testacea and obscure at
best, though the lateral lobes are occasionally infuscated as a whole;
generally the whole pronotum is uniform pale testaceous ; the hind tibial
spurs are more slender in the new species, and the face a little more
oblique in both sexes. Measurements of average specimens are as fol-
lows. Length (if body, cf, 14 mm., 9, 25.5 mm. ; antennjB, ^, 5.75,
9, 8.9 mm.; tegmina, ^, 9.5 mm,, 9} 15.5 mm.; hind femora, ^,
9.4 mm., 9, 15.6 mm.

Other species occur in Northern Mexico, which appear to bo
undescribed.

4. A New Genus of Orphul^.

Among the Orthoptera brought by Mr. Morse from the Pacific coast
is a new form of Orphula3 nearly allied to Chloealtis. Our genera of
Orphula3 may be thus separated.

Table of the United States Genera of Orphidce.

(ji. Antenna; relatively short, at most but little longer than head and pronotum
together ; scapular area of tegmina not specially dilated.

^1. Foveola; of vertex more or less evident ; prozona not much longer than meta-
zona ; lateral lobes of pronotum transverse, that is, deeper than long; upper
ulnar vein of tegmina, at least in male, apically joining the lower ulnar vein at
a long distance beyond the end of the basodiscoidal field . . . Orphulella Stal.
Ifl. Foveolffi of vertex wanting ; prozona very much longer than metazona ;
lateral lobes of pronotum longitudinal, longer than or fully as long as deep ;
upper ulnar vein of tegmina, at least in male, apically strongly arched, joining
the lower ulnar vein not far beyond the end of the basodiscoidal field.

c^. Lateral lobes of pronotum plane above, meeting the disk at nearly right
angles ; lateral carina parallel throughout ; tegmina usually much shorter
than the abdomen in both sexes Dichromorpha Morse.

SCUDDER. — NORTH AMERICAN TRYXALTN.E. 47

c^. Lateral lobes of pronotum convex above, except for the carinae passing
rather gradually into the disk ; lateral carina} divergent on metazona ; tegmina
nearly or quite as long as tlie abdomen, at least in the male.

Clinocephalus Morse.
a-. Antenna) long, about or more than half as long again as head and pronotum
together ; fastigium of vertex with a median carina ; scapular area of tegmina dis-
tinctly dilated, at least in the male.

61. Antennae basally depressed but not expanded, subcqual to near the tip ; face
moderately oblique, the frontal costa subobsolete below the ocellus; lateral
foveola? of vertex obsolete ; pronotum posieriorly truncate, the lateral lobes as

deep as long Chloealtis Harr.

6-. Antennae basally depressed and expanded, tapering in the proximal half; face
strongly oblique, the frontal costa percurrent and sulcate below the ocellus;
lateral foveolae of vertex distinct, linear; pronotum posteriorly obtusangulate, the
lateral lobes longer than deep (Eonomus gen. nov.

CEonomus (oi'ocoVos) gen. nov.

Of slender foi-m. Head somewhat prominent, subconical, the face
strongly oblique ; fastigium of vertex triangular with rounded subrec-
tangulate apex, plane above with feebly raised blunt margins and a
median carina, the lateral foveola^ distinct, linear, invisible from above ;
frontal costa percurrent or almost percurrent, sulcate except in the
uppermost siibvertical portion ; eyes long-oval, oblique, not distant above ;
antenna? half as long again as head and pronotum together in the male,
nearly as long as that in the female, the proximal half beyond the second
joint depressed, expanded and tapering, in the male broader, at broadest,
than the interspace between the eyes, at extreme tip again tapering
slightly, at least in the male. Pronotum compressed, the disk nearly
plane, faintly tectate, with parallel sides, the lateral and median carina?
similar and parallel, the front margin gently convex, the hind margin
broadly obtusangulate, the lateral lobes vertical but gently rounded,
longer than deep. Tegmina shorter than the body, apically subangulate,
the scapular area expanded and scalariform in the male ; wings aborted.
Hind legs slender, the femora surpassing the abdomen, the inner spurs of
hind tibia? equal.

CEonomus altus sp. nov.

Testaceous with a lateral blackish fuscous stripe of variable width, but
generally broad and deepest in color above, extending from behind the
eyes across the pronotum, limited above by the lateral carina and con-
tinued upon the closed tegmina ; otherwise devoid of markings except
usually for a pair of feeble and obscure diverging fuscous stripes on the

48 PROCEEDINGS OF THE AMERICAN ACADEMY.

vertex, and that the antennae are generally much infuscated. Vertex
well rounded, slightly ascending; frontal costa sparsely and irregularly
punctate; eyes a little shorter, especially in the female, than the infra-
ocular portion of the genas, Pronotum with sharp and distinct carinas,
the i^rozona generally impuuctate except anteriorly and delicately, the
transverse sulci feeble, the principal sulcus angulate, especially in the
female, and situated distinctly behind the middle, the metazona densely
and rather finely punctate. The fuscous portion of the tegmina involves
the front margin of the discoidal area, but in the male leaves the distal
expanded half of the scapular area untouched, and in one female the dis-
coidal area is sparsely punctate with fuscous ; they are usually about as
long as the hind femora, but in the female are sometimes no longer than
head and pronotum together. Hind femora very slender, at least equal-
ling ( 9 ) or considerably surpassing ((^) the abdomen ; hind tibise red,
occasionally infuscated apically.

Length of body, ^, 18 mm., 9i 24 mm.; antenna;, ^, 9 mm., 9?
8.5 mm. ; tegmina, (J, 9.5 mm., 9» ^ mm.; hind femora, S, H mm.,
9 , 14.5 mm.

15 (^, 9 9 . ]\It. Wilson, Altadena, Cal., 2400', July 27, A. P. Morse.

5, A New Genus of Stenobothri.

The North American Stenobothri may be sejiarated as follows : —

a'. Antennae not ajjically clavate.

b^. Face considerably oblique, straight or little rounded ; lateral foveola of vertex
slender; lateral lobes of pronotum longer than or fully as long as deep.

ci. Fastigium with a distinct percurrent median carina ; antennae, at least in
female, depressed and more or less expanded basally.

Ji. Antennae much expanded basally, tapering, in the male as long as the
liind femora ; lateral carinse of pronotum subparallel, the disk subrectangu-
lar ; prosternum tuberculate, especially in the male ; tegmina shorter than

the abdomen Najmia McNeill.

d~. Antennas feebly expanded basally, subfiliform, much shorter than the
hind femora ; lateral carinie of pronotum strongly sinuate, the disk clepsydral ;
prosternum not tuberculate ; tegmina longer than the abdomen.

Iloresidotes gen. nov.
c^. Fastigium with no median carina, but at most a colored line, except some-
times in extreme anterior portion; antennfe filiform, the basal joints neither
expanded nor greatly depressed in either sex ; disk of pronotum clepsydral.

Stenobothrus Fiscli.
6'^. Face little oljlique, strongly rounded ; lateral foveolse of vertex moderately
broad, never more than twice as long as broad ; lateral lobes of pronotum deeper
than long.

SCUDDER. — NORTH AMERICAN TRYXALIN.E. 49

c^. Hind margin of pronotiim more angulate than front margin ; posterior
margin of lateral lobes straight ; tegmina and wings fully developed.

Plutybothrus Scudd.
c'-. Fore and liind margins of pronotum equally (and slightly) angulate; pos-
terior margin of lateral lobes sinuate ; tegmina abbreviate and wings aborted.

Bruneria McNeill.
a2. Antenna; apically clavate Gomphocerus Thunb.

Horesidotes (ipeo-iSc5T7js) gen. nov.

Allied to Napaia McNeill (which I have not seen) and separable from
it in the points mentioned in the above table. Head subpyramidal, the
face considerably oblique and straight; occiput with a median carina
extending to and invading the fastigium of the vertex and througliout
accompanied proximately by a pair of similar supjilemeutary carina; ;
foveolaj visible from above, elongate, shallow; eyes rather elongate;
antennse subfiliform, a little depressed but only feebly expanded basally,
moderately slender, a little longer than head and pronotum together in
both sexes but especially in the male, and much shorter than the hind
femora. Pronotum rather small, the disk markedly clepsydral, the
lateral carinas being strongly arcuate and as distinct as the median
carina ; prozona and metazona of sube<iual length, the hind margin
rounded obtusangulate ; lateral lobes slightly longer than deep; proster-
num not tuberculate. Tegmina extending beyond the abdomen, without
intercalary vein, the apical portion of the scapular field expanded in the
male. Inner calcaria of hind tibiae subequal.

Horesidotes cinereus sp. nov.

Varying greatly from light testaceous with a slight olivaceous tinge
and very feeble markings to dark cinereous with heavy markings, which
in some females includes a broad median testaceous stripe on head and
pronotum, bordered on the latter by a velvety black stripe cut by the
luteous lateral carina ; but in others these markings are wholly wanting,
the disk and tegmina are dark cinereous flecked with griseous, and the
lateral lobes are marked with a broad postocular blackish fuscous stripe
extending to the eyes and separated sharply and angularly from the clay-
yellow of the lower portion of the lobes. Similar differences occur in
the males, and there are also some of each sex in which all markings are
but faintly indicated. The antenna? are light castaneous, the wings are
feebly infumate apically with black veins, and the hind femora are of the
color of the upper surface, but where this is light, the upper outer carina
is often marked with black ; hind tibiae glauco-luteous.

VOL. XXXV. 4

50 PROCEEDINGS OF THE AMERICAN ACADEMY.

Length of body, ^, 14.5 mm., 9, 24.5 mm.; antennae, ^, 6.25 mm.,
9 , 8.75 mm. ; tegraina, ,$, 12 mm., 9 , 19 mm. ; hind femora, J, 10 mm.,
9 , 15.5 mm.

19 S, 119. Palm Springs, Cal., July 13, A. P. Morse. It occm-red
on grasses iu dry places in Palm Canon and West Canon.

6. The American Species of Stenobothrus.

I have numerous specimens of Stenobothrus from west of the Missis-
sippi and east of the Sierra Nevadas, including Dallas Co., Iowa (Allen),
Nebraska (Dodge), Colorado, 5500' and 8000' (Morrison). Ft. Collins,
Col. (Baker), Morris Ranch, Larimer Co., Col., 8500' (Baker), Garland
Col., Aug. 28-29 (Scudder), Salt Lake Valley, Utah, Aug. 1-4 (Scudder),
Spring Lake Villa, Utah Co., Utah (Palmer), and Yellowstone Park,
Sept. 6-12 (Scudder). These agree perfectly with eastern examples, and
do not have in any case all of the characters on which McNeill separates
S. colo?-adensis, on the basis of a single female specimen from Ft. Collins,
Col. All the points on which he separates this species vary to a certain
extent, except perhaps the length of the antenna? (in which he may have
been mistaken if the antennse were curled), and I am therefore inclined
to think that S. coluradensis must be placed as a synonym of S. ciirti-
pennis. My Yellowstone Park specimens, all males, are of a smaller
size than is usual ; while specimens collected by R. Thaxter iu sphagnum
bogs at Salmonier, Newfoundland, Aug. 1-15, are unusually green and
have a very peculiar aspect, but I have not discovered any good specific
distinction from S. curtijjenms.

On the other hand, Mr. A. P. Morse brought home from Oregon
(Corvallis, Cordley, Portland, Sept. 18, Divide- Cottage Grove, Sept. 12,
Drain, Sept. 11, Roseburg, Sept. 10, Glendale, Sept. 9, Grant's Pass,
Sept. 8, and Ashland, Sept. 7), and California (Gazelle, Sept. 5, Sisson,
Aug. 29, Baden, Aug. 24, Berkeley, Aug. 21, and San Francisco, Aug. 19),
numerous specimens of a closely allied but more heavily marked species,
which seems to be distinct and may bear the name Stenobothrus oregonensis.
As compared with S. curtipennis the antennae of the male are shorter
than, instead of as long as, the hind femora, the middle joints narrower
than the narrowest part of the frontal costa, instead of being at least as
broad as it; the fastigium of the vertex has a distinct median carina in
the anterior portion wanting or hardly discernible in S. curtipennis ; the
disk of the metazona is marked distinctly and generally broadly at the
sides with black, instead of being generally immaculate or narrowly mar-

SCUDDER. — NORTH AMERICAN TRYXALIN^. 51

giiied with black laterally ; the male tegmina are shorter, generally much
shorter than the abdomen, with rare exceptions as long as the abdomen,
instead of being at least as long as, in macropterous forms considerably
longer than the abdomen, the vena plicata joining the vena dividens
before the middle or fading at some distance before the middle, instead of
running free past the middle of the tegmina. In general it is more
heavily marked, has shorter tegmina and slenderer antennse. lu the
female the tegmina are but little longer than the head and pronotum
together, sometimes no longer. It should at least be distinguished as a
7-ace ; future collections in the intermediate regions will probably show
more clearly whether it should be regarded as wholly distinct.

7. PSOLOESSA AND SxiRAPLEURA.

In his Revision of the North American Tryxalinje (Proc. Dav. Acad.
Sc, VI.), McNeill placed these two genera side by side at the end of his
series. Later, in my Preliminary Classification of the same subfamily
(Psyche, VIII.), I placed them at some distance apart, Psoloessa among
the Phlibostromse and Stirapleura in the Scyllinai. This change of mine
was wrong and came from incorrect observation of the foveolae of the
vertex (a distinction on which I placed a wider reliance than McNeill),
for in Psoloessa they ai'e partially visible from above, their plane beino^
twisted feebly in relation to that of the margin of the vertex, so that while
they are visible from above on their inner half, they are not so on their
outer half. The other features of Psoloessa show that it belongs to the
Scyllinai, and I would restore it to the immediate vicinity of Stirapleura,
to which it is very closely allied.

The table given by me for the separation of the genera of Scyllinfe may
be altered by substituting the following for the final paragraph relatino-
to Stirapleura (Psyche, VIII. 231) : —

Pronotum constricted in tlie middle, the prozona slightly the shorter ; lateral
carin<e percurrent, more or less divergent in front and strongly divergent behind.
FoveoliB of vertex visible from above only on the inner half; lateral carinaj of
pronotum anteriorly but little or at least not strong!}' divergent, being gently
arcuate on the prozona ; lateral lobes of prozona feebly or not marked above tlie
middle with obliquely disposed short lunate carinules . . . Psoloessa Scudd.
Foveolffi of vertex visible from above throughout their length ; lateral carina of
pronotum very strongly divergent in front as well as behind, being strongly bent-
arcuate on the prozona; lateral lobes of prozona more or less conspicuously
rugose-carinate obliquely above the middle Stirapleura Scudd.

52 PROCEEDINGS OP THE AMERICAN ACADEMY.

This definition will leave P. texana Scudd. in Psoloessa, instead of
transferring it to Stirapleura, as was done by McNeill, doubtless on
account of the slightly more marked oblique carinaj of the lateral lobes of
the pronotum. The face is generally a little more oblique in Psoloessa
than in Stirapleura, but the distinctions drawn by McNeill from the
frontal costa hold in Psoloessa only for P. huddiana Brun.

As to the species of Stirapleura, I am inclined to look on the form
from southern California, heretofore regarded as identical with S. delica-
tula (Scudd.) of Colorado to be distinct from though closely allied to it.
I have before me a considerable series (more than a hundred) of each,
and I find the Californiau species to have a slenderer form, longer teg-
mina and wings, and the upper inner angle of the lateral foveolns of the
vertex distinctly more rounded so as to make them less distinctly rhom-
boidal than in S. delicatida. I describe it herewith, together with
another new species from Texas, remarkable for the delicacy of the
lateral carinas of the pronotum and approaching Psoloessa in the feeble-
ness of the oblique carinas of the lateral lobes.

Stirapleura pusilla sp. nov.

Head moderately prominent, subasceuding, the fastigium of the vertex
rather deeply sulcate with elevated, anteriorly acutangulate margins ;
lateral foveoliB nearly or quite half as long again as greatest breadth, sub-
rhomboidal, but narrower interiorly than exteriorly, with the inner upper
angle distinctly rounded; frontal costa much contracted at summit, more
or less gradually broadening, sulcate throughout but only feebly at base,
punctate within the raised and smooth margins ; antennsB distinctly but
not greatly longer than head and pronotum together, especially in the
male. Pronotum considerably constricted mesially, the posterior margin
obtusangulate, the median carina moderately prominent, e({ual, cut barely
in advance of the middle, the lateral carinae equally prominent, bent-
arcuate and strongly divergent, especially behind, so that the disk of the
pronotum is about twice as broad posteriorly as near the middle, the
lateral lobes more or less corrugated at the shoulder just below the lateral
carinas. Color cinereous, generally much marked with fuscous and black,
paler beneath than above, but very variable ; face generally testaceous,
the frontal costa more or less infuscated, the gense generally dotted with
fuscous or infuscated, sometimes with the exception of a broad arcuate
oblique subocular stripe ; the occiput may or may not be striped with
testaceous and fuscous, but there is usually a broad postocular fuscous
stripe extending across the lateral lobes, where it is often followed below

SCUDDER, — NORTH AMERICAN TRYXALIN/E. 53

by a testaceous stripe, below which the lateral lobes are again iufuscated,
but ofteu enlivened below the middle jDOsteriorly by a more or less con-
spicuous oblique Havo-testaceous bar, sometimes merged in the lighter
color of the lowest portion ; disk of pronotum testaceous more or less
infuscated, the carinas usually flavous and the metazona with a triangular
black patch on either side. Tegmina surpassing the hind femora in both
sexes, more or less heavily flecked with fuscous, ranging from a nearly
uniform sprinkling throughout with fuscous dots to a regular series of six
or more quadrate fuscous blocks in the median area ; wings hyaline with
black veins. Hind femora considerably surpassing the abdomen, cinere-
ous or cinereo-testaceous, generally marked above with a median trian-
gular black-edged brown spot and often also with less conspicuous basal
and postmedian fuscous patches, the geniculation more or less infuscated ;
hind tibiae pallid with a glaucous tinge and generally flecked more or less
conspicuously with fuscous, the base with a postgenicular infumated
annulus, the spines black-tipped.

Length of body, $, 10.5 mm., 9, 18 mm.; antenna?, S, 5.2 mm., 9,
i^.b mm. ; tegmina, $^ 10.5 mm., 9, 15 mm. ; hind femora, S , 8.5 mm.,
9 , 12.75 mm.

89 $, 80 9 . Mesilla, N. Max., July 1 (Morse) ; and the following
from California : San Diego, July 22 (Morse) ; Coronado, July 24
(Morse); San Bernardino, July 16 (Morse); Cahon Pass, July 18
(Morse) ; Los Angeles, July 21 (Morse), Oct. 24 (Scudder) ; Altadena,
July 29, and Mt. Wilson, 2400', July 27 (Morse) ; Pasadena, Oct. 23
(Scudder) ; Santa Barbara, Oct. 21 (Scudder) ; Lancaster, July 31,
Aug. 1 (Morse) ; Tehachapi, Aug. 2 (Morse) ; Kern City, Aug. 4 (Morse) ;
Tulare, Aug. 5 (Morse) ; Monterey, Oct. 18 (Scudder) ; Raymond, Aug. 16
(Morse), and Ahwanee, Aug. 15 (Morse).

At Mesilla, Mr. Morse found this insect common on sand and gravel
on the mesa, scarce on weeds and tall grass along ditches.

Stirapleura tenuioarina sp. nov.

Head not very prominent, the fastigium of the vertex moderately
narrow, deeply sulcate with elevated margins acutangulate in front ;
lateral foveolae almost exactly quadrate, barely longer than broad, dis-
tinctly impressed ; frontal costa pinched above, gradually enlarging
throughout and sulcate, though scarcely so at base and directly above the
ocellus, very sparsely punctate ; antennas a little shorter than head and
pronotum together. Pronotum not greatly constricted mesially, the
posterior margin broadly obtusangulate, a little rounded, the median

54 PROCEEDINGS OF THE AMERICAN ACADEMY.

carina ouly slightly prominent, equal, cut distinctly in advance of the
middle, the lateral carinte delicate, feebly elevated, briefly parallel just in
front of the middle, but widely divergent in advance, as behind; lateral
lobes with subdued corrugations at the shoulder just below the lateral
carinte. Color fusco-cinereous, the head testaceous with pallid gense,
jDlumbeo-fuscous above ; antennae testaceous interrupted with fuscous ;
pronotum nearly uniform obscure fusco-testaceous, the carina^ concolor-
ous ; tegmina cinereous much flecked with fuscous especially in quadrate
patches along the median area, one beyond the middle larger than the
rest and divided by an oblique pallid bar ; wings hyaline with black
veins ; hind femora cinereo-testaceous with a ferruginous tinge and
blotched with fuscous, above rather obscurely trimaculate with fuscous ;
hind tibice pale ferrugineo-luteous, deepening in tint distally.

Length of body, 22.5 mm.; antennse, 5.75 mm.; tegmina, 19.5 mm.;
hind femora, 13.75 mm.

1 9. Sierra Blanca, El Paso Co., Tex., June 26, A. P. Morse.

8. The Species of Aulocara.

Aulocara Scudd. has as synonyms Qildocara Scudd. and Coloradella
Brunn. (See Can. Ent., XXIX. 75^ and Psychej VII. 71.) No species
have been referred to the last named, but to the others five nominal spe-
cies have been referred; elliotti Tliom., ccBi'uleipes Scudd., decens Scudd.,
strangrdatiun Scudd., and scudderi Brun. Scudderi, as has been shown
by McNeill, belongs to Ageneotettix (Eremnus). The other four all
represent a single species, which must bear the oldest name, eJUotti.
Nevertheless we possess four species which may be separated by the fol-
lowing table : —

Tahle of the Species of Aulocara.

u^. Fastigium of vertex broader than long, its front margin obtusangulate ; meta-
zona feebly tumescent, its hind margin not very broadly obtusangulate ; hind

tibiae red rufum sp. nov.

fl'2. Fastigium of vertex at least as long as broad, its front margin rectangulate or
acutangulate ; nietazona plane or nearly plane, its hind margin broadly obtus-
angulate; hind tibiaj purple,* or glaucous.

61. Pronotum strongly constricted mesially, the disk with more or less conspicu-
ously decussate markings, the lateral carinae strongly divergent in front and
behind.

c^. Male antennae as long as thorax and abdomen combined ; lower margm of
lateral foveola; of vertex obsolete or obsolescent ; tegmina generally immacu-
late, much shorter than the long hind femora femoratiim sp. nov.

* The hind tibial oi A. parallelum are not known, but are presumably purple.

SCUDDER. — NORTH AMERICAN TRYXALIN^. 55

c^. Male antenna? shorter than thorax and abdomen combined ; lower margin
of lateral foveolte of vertex distinct ; tegmina generally maculate, generally

fully as long as the relatively shorter hind femora elliotti Thorn.

h-. Pronotum but feebly constricted mcsially, the disk of subequal width and uni-
colorous, the lateral carinse divergent only, and but little, behind.

parallelam sp. nov.

Aulocara rufum sp. nov.

Head well rounded, rather large, ferrugineo-cinereous, paler on face,
Avith a pair of obscure fuscous stripes on summit and more or less flecked
with fuscous or ferruginous on upper part of genai ; summit tumid, the
fastiorium much broader than long, with slightly raised, parallel lateral
margins, the front margin distinctly obtusaugulate ; lateral foveolte obso-
lescent, scarcely impressed, subtriangular, longer than broad ; frontal
costa of moderate breadth, subequal but feebly narrowed above, obsoles-
cent below with slightly and narrowly elevated margins; antenntie fer-
ruginous, more or less infuscated, in the male as long as the hind femora.
Pronotum nearly uniform rufous, more or less infuscated on disk, espe-
cially on metazona, not greatly constricted mesially, the metazona feebly
tumescent, the hind margin a little obtusangulate, the angle sometimes
much rounded, the median carina slight and confined to the metazona.
Teo-mina broad and well rounded, rufous, minutely sprinkled with fuscous,
hardly surpassing the hind femora; wings hyaline, the veins glaucous,
sometimes infuscated. Hiud femora cinereo-testaceous, often more or less
ferruginous, generally very obscurely (but occasionally in male distinctly)
bifasciate with fuscous ; hind tibije light red, pallescent basally.

Length of body, $, 14 mm., 9, 19 mm. ; antennae, $, 8 mm., 9,
8.75 mm.; tegmina, $, 11.5 mm., 9, 15.5 mm.; hind femora, $,
8 mm., 9? 12 mm.

5 (?, 4 9 . Pueblo, Col., Jitly 8-9, Aug. 30-31.

This species is very distinct from all the others, not only in the color-
vnff of the body, tegmina, and hind tibiag, but in the breadth of the fastigium
of the vertex, the obscurity of the foveolte, the absence of a median carina
on the prozona, and the lesser obtuseness of the hind margin of the
metazona.

Aulocara femoratum sp. nov.

Of minor size, the head well rounded and rather prominent, cinereo-
testaceous, a little infuscated above; summit tumid, the fastigium deeply
excavate, considerably longer than broad, the margins rather sharply
elevated, the lateral parallel, the front acutangulate ; lateral foveola? tri-
ano-ular, of moderate size, distinctly impressed, but with obsolescent or no

56 PROCEEDINGS OF THE AMERICAN ACADEMY.

inferior margin ; frontal costa rather narrow, subequal but slightly com-
pressed above, obsolete below, more or less but generally feebly sulcate ;
antennae testaceous, infuscated except near base, of unusual length though
not quite so long as the long hind femora. Pronotum testaceous, the
disk somewhat infuscated, considerably constricted mesially, the lateral
lobes with a large and distinct, quadrate, subcentral but superior, fuscous
patch, the metazona plane, the hind margin subtruncate but feebly angu-
late, the median carina distinct, equal, percurrent, the lateral carinas blunt
and obscure except where marked with pale testaceous, strongly diver-
gent in front and behind. Tegmiua short and well rounded, when closed
not nearly covering the abdomen, testaceous, sometimes feebly infuscated
or minutely and obscurely flecked with fuscous ; wings hyaline with
glaucous veins. Hind femora of unusual length, much surpassing the
abdomen, testaceous, obliquely biannulate with blackish fuscous, the
geniculation broadly black; hind tibia? glaucous, basally pallescent.

Length of body, 15 mm.; antenuEe, 9.5 mm.; tegmina, 7.5 mm. ; hind
femora, 10.75 mm.

5 (J. Lakin, Kans., Sept. 1 ; Colorado, 5500', Morrison ; Pueblo, Col.,
Aug. 30-31 ; Prove, Utah, Aug. 23-24.

I have based this species upon the male specimens just quoted, but I
have also three females from Lakin and Pueblo, with short wings and
long hind femora, which I think belong here, but which resemble
A. eUiotti more than do the males. The species is smaller than any of
its congeners, with shorter wings, longer hind femora, and subtruncate
posterior margin of the pronotum.

Aulocara elliotti.

Stanronohis elliotti Thorn., Proc. Acad. Nat. Sc, Philad., 1870, 82 (1870).

(Edocara elliotti Sauss., Prodr. CEdip., 79 (1884).

Aulocara elliotti Brun., Can. Ent., XVII. 10 (1885).

Aulocara ccEuileipes Scudd. !, Bull. Ha3'd. Surv., II. 266 (1876).

Aulocara decens Scudd. !, Ibid., II. 2G7 (1876).

(Edocara strangulatina Scudd. !, Ann. Rep. Wheel. Surv., 1876, 289 (1876).

I have specimens, mostly collected by myself, from Kansas (Lakin,
Sept. 1) ; Colorado (southern Colorado, June 11-20, Carpenter; Ani-
mas, July 8-9 ; Granada, July 8-9 ; Pueblo, July 8-9 ; Garden of the
Gods, July 13, Packard; Florissant. Aug. 17-22); Texas (Fort Worth,
July 4) ; New Mexico (Johnson's Basin, June 22, Townsend) ; Arizona
(Fort Whipple, Palmer) ; Utah (Castle Gate, Aug. 22 ; Lake Point,
Salt Lake, July 26, Packard ; Salt Lake valley, Aug. 1-4) ; and Cali-

SCUDDER. — NORTH AMERICAN TRYXALIN^. 57

fornia (Tehachapi, Aug. 2, A. P. Morse, the last of an uuusually large
size) ; also mountains 12 leagues east of San Luis, Mexico, E. Palmer.

It has also been credited to Nebraska (Bruner), Wyoming (Thomas),
the Yellowstone region (Bruner), Montana (Thomas), and Washington
(Bruner).

The species varies much in coloring, from nearly immaculate to
markedly maculate, and also in the length of the tegmiua, which, though
rarely as short as the hind femora, may sometimes far surpass them in
length.

Aulocara parallelum sp. nov.

Head well rounded, rather large and prominent, cinereo-testaceous,
feebly infuscated above, and with a broad and broadening postocular,
blackish fuscous stripe, extending over the prozona ; summit tumid, the
fastigium depressed, with well elevated margins, rather longer than
broad, at least in the male, the lateral margins parallel, the front acutan-
gulate, particularly in the male ; lateral foveolse triangular, elongate,
tolerably well impressed, with coarse raised margins ; frontal costa not
very broad, in the male much constricted above and gradually broaden-
ing, percurrent and rather deeply sulcate, in the female subequal, obsolete
below, faintly sulcate at and below the ocellus. Pronotum dull testaceous,
with the postocular stripe mentioned, very broad in front and rapidly
narrowing, sometimes extended over the metazona on the upper margin
of the lateral lobes, feebly constricted mesially, the metazona plane, the
hind margin very broadly obtusangulate, the median carina distinct and
percurrent, but feeble or obsolete between the sulci, the lateral carinte
diverging, and not greatly, only behind, so that the disk is subequal in
width. Tegmina extending to tip of abdomen, well rounded, testaceous,
more or less infuscated along the median area and especially in the
proximal half ; wings hyaline, the veins glaucous except apically in
anterior area where they are fuscous. Hind legs lost in only specimens
seen.

Length of body, ^, 19 mm., 9, 28 mm.; tegmina, ^, 13.5 mm.,
9 , 20 mm.

1 ^, 19. Salt Lake Valley, Utah, Aug. 1-4.

This species is remarkable for the parallelism of the lateral carina} on
the prozona, giving a disk of subequal width, and for the lack of decussate
markings, combined with the presence of a broad postocular stripe.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 3. — August, 1899.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE

A REVISION OF THE ATOMIC WEIGHT OF COBALT

THIRD PAPER. — THE ANALYSIS OF COBALTOUS CHLORIDE

AND OXIDE.

By Theodore William Richards and Gregory Paul Baxter.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF
HARVARD COLLEGE.

A REVISION OF THE ATOMIC WEIGHT OP^ COBALT.

THIRD PAPER: — THE ANALYSIS OF COBALTOUS
CHLORIDE AND OXIDE.

By Theodore William Richards and Gregory Paul Baxter.

Received June 24, 1899.

The two preceding papers,* describing the earlier part of this investi-
gation, were concerned with the complete analysis of cobaltous bromide.
Both the cobalt and the bromine in this compound, as well as the trace
of inevitable impurity, were determined with care, and the consistent
verdict of all the results pointed toward a value between 58.99 and 59.00
for the number soujjht. The danger of the existence of an unknown
error in a result thus obtained from a single compound is too well known
to need emphasis ; and with this danger in mind we sought to obtain
support from wholly different data. The present paper describes two
series of determinations, which, while less satisfactory than the work with
the bromide, are nevertheless of corroborative value, and moreover serve
to point out some of the pitfalls which ensnared the earlier experimenters.

I. The Reduction of Cobaltous Chloride.

Cobaltous chloride was chosen first as the basis of the confirmatory
work, both on account of the ease with which it can be prepared, and
because it can be heated to a much higher temperature than the bromide
without danger of vaporization. These circumstances allow greater
surety both in drying and in reducing.

Owing to the comparatively slight volatility of cobaltous chloride and
to the dilHcult}' of obtaining chlorine in a sufficiently pure state, we made
no attempt to purify the salt by sublimation. The preparation of
cobaltous chloride in aqueous solution was inadmissible on account of the
certain presence of oxide in the dried and ignited substance under these

* These Proceedings, XXXIIL 113 (1897), XXXIV. 351 (1898).

62 PROCEEDINGS OF THE AMERICAN ACADEMY.

conditions ; hence the method finally adopted was the ignition of care-
fully purified purpureo-cobaltic chloride, which may be dried before it
decomposes.

A considerable quantity of this compound was prepared by passing air
through a strongly ammoniacal solution of cobaltous chloride and subse-
quently acidifying the mixture with hydrochloric acid. The precipitate
was then recrystallized six times by solution in pure redistilled ammonia
in a platinum dish and precipitation with redistilled hydrochloric acid in
a Jena glass flask. This latter operation could not be performed in
platinum, because of the liberation of small quantities of chlorine. The
final precipitate was collected on a pure washed filter and dried. In
order to convert the amine into cobaltous chloride it was heated in an air
bath to about 200°, until it was entirely decomposed without fusing.
Owing to the fact that the dried salt melts at a temperature very slightly
above that at which it breaks up, the heat was applied very gradually.
Here again the liberation of traces of chlorine prevented the use of
platinum, a porcelain crucible being used to contain the salt. The cor-
responding purpureo-bromide of cobalt melts at a temperature too near
its decomposing point to have allowed the use of a similar method for
preparing cobaltous bromide in the earlier part of this research.

The cobaltous chloride thus prepared, still containing a large amount
of amnionic chloride, was now heated in the platinum boat in a rapid
current of dry nitrogen and hydrochloric acid gas until ammonic chloride
ceased to be expelled. The drying apparatus used in the work upon the
bromide was refilled, and easily altered for this purpose by substituting a
strong solution of hydrochloric acid for the bromine and red phosphorus.
The hydrochloric acid gas was displaced by pure nitrogen, and this in
turn by pure air ; and finally the di-y cobaltous chloride was transferred to
the weighing bottle by means of the usual bottling apparatus.

After having been weighed, the chloride was reduced in hydrogen in
the manner already described under the analysis of cobaltous bromide.

During the reduction no trace of cobaltous chloride sublimed from the
boat, but a slight sublimate of ammonic chloride appeared in the tube.
The amount of this sublimate was determined by washing out the tube
with pure water and Nesslerizing the solution, and this amount was sub-
tracted from the original weight of cobaltous chloride. Probably the
correction was somewhat too great, for traces of ammonic bromide
appeared in every reduction of cobaltous bromide, when no ammonia
could have been present in the salt ; but the error arising from this
source must have been small.

RICHARDS AND BAXTER. — ATOMIC WEIGHT OF COBALT.

63

A small quantity of alkaline chloride was found in the residual cobalt, and
was determined in the usual way by lixiviating the metal and evaporating
the solution. This correction was of course subtracted from the original
weight of the cobaltous chloride, as well as from that of the metal.

In the following table of results the specific gravity of cobaltous chlo-
ride is assumed to be 2.94, the value given by Joule and Playfair.* This
necessitates a correction to the vacuum standard of +.000262 gram per
gram of chloride. Cobalt itself has a specific gravity too near that of
brass to need any correction of this kind. CI = 35.455.

Observed

AVeiiiht of

Cobaltous

Chloride

Observed

Weight of

Cobalt
iu Vacuum.

Weight of
Residue.

Weight of
Amnionic
Chloride.

Corrected
Weight of
Cobaltous
Chloride.

Corrected

Weight of

Cobalt.

Atomic
Weight

of
Cobalt.

in Vacuum.

4.16702

1.89304

.00061

.00158

4.16483

1.89243

59.053

2.30588

1.04771

.00048
Averajre .

.00028

2.30512

1.04723

59.035
59.044

As the alkali had not been completely removed by the process of
crystallization to which the cobaltamine had been subjected, it is highly
probable that some silica also remained which was not removed from the
metal by leaching. This probability, together with the fact that the
ammonic chloride correction is undoubtedly somewhat too large, seems
sufficient to account for the fact that these results are slightly higher
than those obtained in the bromide series (58.995). At any rate the
evidence of this method is that the atomic weight of cobalt cannot be
greater than 59,05, for none of the probable errors would tend to make
this result too low.

IT. The Reduction of Cobaltous Oxide.

At this point experimental difficulties in the preparation of the chlo-
ride, together with the belief that silica could not be completely removed
from the salt without introducing platinum, led to a search for a more
satisfactory method. In the hands of the early experimenters upon the
atomic weights of nickel and cobalt, reduction of the monoxide yielded
good results, — results less satisfactory in the case of cobalt than in that
of nickel, however, on account of the greater difficulty in preparing the
monoxide in a pure state.

Kusselljt the first investigator to work with cobaltous oxide, found that
the higher oxides of cobalt were reduced to the monoxide by heating to

* Landolt and Burnstein, Tabellen (1894), p. 134.
t Jour. Chem. Soc. (2.), I. 51 (1863).

6i PROCEEDINGS OF THE AMERICAN ACADEMY.

a higli temperature in air, but that reoxidatiou of the mouoxide took
place while it was cooling. He showed this by igniting in air a crucible
containing the black oxide and cooling the crucible suddenly by plunging
it into cold water. Under these conditions a film of black oxide was
formed on the surface only, while the lower portions remained in the
form of brown monoxide. Evidently, to prevent reoxidation it was only
necessary to keep the oxide away from contact with oxygen. The method
adopted by Russell was to beat the oxide in a current of an inert gas until
constant weight was established, and then to reduce the compound iu
hydrogen, thus obtaining an indirect ratio between oxygen and cobalt.

This method of procedure, which was adopted by all subsequent experi-
menters on the oxide, seemed open to two serious disadvantages. The
first, the difficulty of preparing nitrogen or carbon dioxide free from
traces of oxygen, is one which can be surmounted, although not easily.
The possibility that small quantities of liberated oxygen may become
mechanically entangled in the oxide, in such a way as not to be speedily
removed by the current of gas, is a danger even less easily avoided.
This oxygen would of course reunite with the monoxide upon cooling,
thus making its weight too great.

As an alternative to heating in a current of inert gas, ignition in a
vacuum suggested itself as offering distinct advantages, since neither of
the dangers just mentioned would be incurred. For this purpose a
glazed porcelain tube, of twenty-eight millimeters internal diameter,
similar to that used for fusing silver in a vacuum, was employed.* The
tube was closed as usual with Hempel hollow brass stoppers and heated
in a perforated Fletcher furnace by means of a large blast lamp, while
an ordinary Sprengel air-pump served to exhaust the gases from the tube.

Very pure cobalt, which had been prepared by various methods during
the previous work, was converted into the oxide as follows. From the
solution of the metal iu pure nitric acid the pink granular cobaltous
hydrate was precipitated by adding an excess of pure freshly redistilled
ammonia, and subsequently digesting the mixture upon the steam bath
in a platinum bowl. This precipitate was collected upon a Gooch cru-
cible in which a circular piece of hardened filter paper was used instead
of an asbestos mat. After drying and converting the mass into the black
oxide by ignition with an alcohol lamp this material was used directly
for analysis. The metal which served as the source of this oxide was
known to be free from alkalis and silica,t and as neither the nitric acid

* Richards and Parker, These Proceedings, XXXII. G3 (1896)..
t These Proceedings, XXXIII. 120 (1897).

RICHARDS AND BAXTER. — ATOMIC WEIGHT OF COBALT. 65

nor the ammonia could have introduced either of these impurities, the
oxide must have been as pure as possible. It has already been shown
that specimens of cobalt prepared by any of the methods described in the
preceding parts of this paper are essentially identical.*

A weighed platinum boat, containing several grams of cobaltic oxide,
was placed in the porcelain tube, and, after the air had been exhausted,
the portion of the tube containing the boat was heated to full redness.
The expelled oxygen was removed with the air-pump ; and the tension
of the gases in the tube was finally reduced to a small fraction of a
millimeter of mercury. The tube was then cooled while the vacuum was
still maintained, air was admitted, and the boat was transferred to the
weighing bottle and weighed. Successive ignition for periods of about
two hours each reduced the weight of the cobaltous oxide slowly by
amounts varying from two to five tenths of a milligram, until constant
weight within one teoth of a milligram was obtained. During this
process the cobaltous oxide, which was light brown at first, gradually
assumed a darker tint.

Since complete reduction of this oxide by hydrogen at a temperature
which could be borne by hard glass is very slow, the rednction was con-
ducted in a porcelain tube heated by the furnace. The metallic cobalt,
which sintered together in such a fashion as to occupy not more than one
fifth the original volume of the oxide, was cooled in a vacuum, and then
transferred to the weighincr bottle and weighed. Constant weight was
obtained without difficulty. f

In the following table are given the results obtained by this method of
procedure. The correction to a vacuum standard of +-000069 gram per
gram of cobaltous oxide was applied, the specific gravity of the oxide
being taken as 5.68 (Joule and PIayfair).$ 0 = 16.000.

Number of
Analysis.

Weight of

Cobaltous Oxide

in Vacuum.

Weight of

Cobalt
in Vacuum.

Atomic

Weia;ht of

Cobalt.

1

grams.

7.04053

grams.

5.53779

58.962

2

6.69104

5.26312

58.974

3

7.83211

6.15963

58.927
58.954

* These Proceedings, XXXIII. 120 (1897).

t Sintered cobalt does not occlude important amounts of hydrogen, even when
cooled in this gas. A paper showing this fact will soon be published.
I Landolt and Bi'.rnstein, Tabellen (189i), p. 13i.
vor-. XXXV. — 5

66 PROCEEDINGS OF THE AMERICAN ACADEMY.

From the fact that the above results agree neither among themselves
nor with the results obtained from the bromide (58.995), it seemed highly
probable that all excess of oxygen had not been removed from the
cobaltous oxide. It has been shown that oxides made by ignition of
nitrates give off their extra " included " oxygen only upon long continued
heating, the gas gradually working its way out by a process of dissocia-
tion and recombination.* Here also the oxygen combined with the
innermost parts of the monoxide, too tightly enclosed to escape easily,
must have required time to force its way out by the same process.

Accordingly, in the next analysis the oxide was heated for much longer
periods in the highest vacuum obtainable. Under these conditions the
loss of weight between each two ignitions increased very materially,
amounting to more than a milligram in each case, and culminated in a
loss of three milligrams accompanied by a decided darkening of the
oxide. This called to mind an experiment made by Dr. Cushman in his
work upon the atomic weight of nickel. f In this experiment he ignited
nickelous oxide to bright redness in a current of nitrogen. Upon
removal from the ignition tube the oxide was found to have been partially
converted into metallic nickel, owing doubtless to the dissociation of the
oxide into metal and oxygen, and the subsequent carrying away of the
oxygen by the current of nitrogen. Cupric oxide, when heated only to
700°-800°, is known to dissociate perceptibly into cuprous oxide and
oxygen. $ As a matter of fact, when the cobaltous oxide was removed
from the boat, it was found to have been reduced to such an extent that
enough metallic cobalt had been formed to alloy with the platinum ; thus
cobaltous oxide also, at about 800°, dissociates very considerably into
cobalt and oxygen. In the zeal to obviate oxidation, the opposite error
had been encountered. Obviously the only method of obtaiuing the
monoxide free from the metal was to determine the tension of the oxygen
evolved from it at the temperature to which the oxide was heated, and
always to keep the tension above this point during the ignition.

In the earlier experiments, an exceedingly slight known leakage of
water through the brass stoppers into the tube may explain why the
reduction of the monoxide did not occur, for metallic cobalt decomposes
water at a high temperature.

Some of the brown monoxide which had suffered partial reduction in
the experiment described above, and which consequently must have been

* T. W. Ricliards, These Proceedings, XXXIII. 399.
t These Proceedings, XXXIII. 424.
J Ibid., 421.

RICHARDS AND BAXTER. — ATOMIC WEIGHT OF COBALT. 67

free from higher oxides, was uow heated in a hard glass tube sealed at
one end ; and the tube was exhausted to within two tenths of a milli-
meter of mercury by means of the Sprengel pump. When the pump was
stopped the tension of the gas slowly increased from four tenths of a
millimeter at 400° to one and five tenths millimeters at full redness, the
highest temperature obtainable with a Buusen lamp. The pressure of
the oxygen resulting from the dissociation of the black oxide of cobalt
was determined to be in the neighborhood of six hundred and thirty
millimeters at the same temperature. Evidently, if the tension of the
oxygen remains below six hundred and thirty millimeters of mercury, the
higher oxides of cobalt cannot exist (except in an imprisoned state) ;
and, on the other hand, if the tension of the oxygen does not fall below
one and five tenths millimeters there can be no reduction of the mon-
oxide. In the next analysis the oxygen was removed from the tube as
fast as it was evolved until the tension of the gas remainins: was about
one and five tenths millimeters. The pressure inside the tube was kept
at this point until heating ceased, when the tube was exhausted as com-
pletely as possible. In subsequent ignitions of this boat-load the tube
necessarily contained air, not oxygen, and allowance was made for this
fact by increasing the pressure five times. In this case also constant
weight was obtained only after many ignitions, although the differences in
the weight of the oxide were small. 7.74242 grams (in vacuum) of
cobaltous oxide yielded 6.09219 grams of cobalt, corresponding to an
atomic weight 59.068.

Evidently even here traces of metal had been found during the heating
of the monoxide. This uudesired reduction could have taken place only
during the cooling, for at all other times the tension of the oxygen was
greater than the dissociation pressure of the monoxide. Measures were
then taken to make this error impossible.

On account of the size of both tube and furnace, the cooling was
necessarily very slow ; this fact is indeed an advantage, because other-
wise the life of a tube would be short. Assuming that the tube cooled
always at the same rate, the tension of the oxygen from the dissociated
oxide after any given interval of time from the cessation of ignition
should always be the same. A table embodying such tensions was pre-
pared by repeatedly exhausting the tube to a tension slightly below the
supposed point, and allowing the pressure inside the tube to reach a
maximum. This table, which was obtained by averaging three different
series of observations, is given below. The readings were obtained by
means of a McLeod gauge.

68 PROCEEDINGS OF THE AMERICAN ACADEMY.

Time from Ces.aation
of Ignition.

Tension of Oxygen
Resulting from Dissociation

Minutes.

Millimeters.

15

.80

30

.56

45

.28

60

.21

75

.11

90

.07

These values have no meaning except for the particular process by
which they were obtained, but they show plainly how great a chance for
error existed when the tube was exhausted completely as soon as ignition
ceased.

In the next analysis the procedure was exactly similar to that in
Analysis 4, except that during cooling the values indicated in the table
were maintained. No essential loss of weight took place after the third
heating, although several additional ignitions were carried out for the
sake of certainty. 10.58678 grams (in vacuum) of cobaltous oxide yielded
8.32611 grams of cobalt, corresponding to an atomic weight 58.929.

In the face of this unsatisfactory result, there seemed nothing further
to do except to attempt the detection of the presence of impurities.
Three possibilities were open : the oxide might contain moisture, enough
nitrate might have been carried down by the cobaltous hydrate to cause
a small amount of nitrogen and oxygen to be occluded by the oxide, or
else some higher oxide of cobalt might be present.

It was easily shown that cobaltous oxide is as a matter of fact slightly
hygroscopic. Several grams exposed for twenty-four hours to an atmos-
phere saturated with moisture gained five milligrams in weight, most of
which was lost in a short time iti a desiccator. It was satisfactorily
proved, however, by suitable experiments which need not be recounted,
that not enough moisture could have crept in during any of the processes
used to cause any important error.

In order to prove the absence of included gases, a specimen of oxide
which had been treated in the same manner as the material used in
Analysis 5 was dissolved in hydrochloric acid in an apparatus for measur-
ing the gases evolved during the solution.* No gas was given off; but
upon treating the solution with potassic iodide and starch, an unmistak-
able liberation of iodine was detected. This showed conclusively that
some higher oxide of cobalt had remained undecomposed, for pure cobalt-

* These Proceedings, XXXIII. 403, and Jour. Chem. Soc. Trans., LXXI. 559.

EICHARDS AND BAXTER. ATOMIC WEIGHT OF COBALT. 69

ous chloride lias no immediate effect on hydriodic acid. The presence
of the imprisoned higher oxide is undoubtedly the reason i'or the low
value for the atomic weight of cobalt obtained in this way.

The cobalt itself obtained from the pure oxide by reduction at a high
temperature is very permanent in the air, and occludes only traces of
hydrogen.

With o-reat reluctance the work upon the monoxide was abandoned at
this point; for a substance whose composition varies so widely under
conditions which vary so sligiitly is obviously unfitted for woi-k of the
hio-hest accuracy. As in the case of cobaltous bromide, the preparation
of this substance even in a reasonably pure state appears to be an impos-
sibility. Unfortunately the accurate estimation of the worst impurity in
this case, cobakic oxide, is a very difficult matter. An attempt to use
pure cobaltic oxide itself as the basis of the determination proved to be
equally unsatisfactory.

The results of this part of the investigation are nevertheless of con-
siderable import in the light they throw upon previous analyses of the
monoxides of both nickel and cobalt as well as upon the accuracy of the
results obtained in the analyses of cobaltous halides described in this
paper. The material which gave the lowest result contained oxides of
cobalt higher than the monoxide, hence these results are too low. On
the other hand, the material which gave the highest result was heated
under conditions which theoretically should have caused, and evidently
did cause, partial reduction of the oxide to metal. The true value of the
atomic weight of cobalt must lie between these two extremes, 58.93 and
59.07 ; and until trustworthy evidence to the contrary has been produced,
the most probable value seems to be the mean 58.995, obtained from the
analysis of the bromide.

Critical Review of Earlier "Work.

In the light of the information gained during the protracted investiga-
tion which has just been described, it; is possible to discover the reasons
for many of the unusual discrepancies in the older work upon the atomic
weight of cobalt. Following is a chronological list of this work, with
references to the original articles. The atomic weight of oxygen, upon
which the appended values are based, is taken as 16.000.

Pogg. Ann., VIII. 184. 59

Pogg. Ann., CI. 387. 60.0

Bibl. Univ. fie Geneve. I. 372. 58.8

Ann. Cheni. Pharm., CXIII. 25. 59.1

RotlihoEE

1826

Schneider

1857

Marignac

1858

Dumas

1860

70

PROCEEDINGS OF THE AMERICAN ACADEMY.

Russell

1863

Sonimaruga

1866

Winkler

1867

Russell

1867

Weselsky

1868

Lee

1871

Ziuimerniiinn

1886

Kriiss and Schmidt

1889

Winkler

1889

Remmler

1891

Schiitzenberger

1892

Winkler

1893

Winkler

1894

Winkler

1895

Winkler

1895

Henipel and Tliiele

1896

Ricliarils and Baxter

1897

Winkler

1898

Riciiards and Baxter

1899

Richards and Baxter

1899

Jour. Chem. Soe. (2), I. 51. 58.7

Sitz. Wien. Akad., LIV. (11.), 50. 60.

Zeit. Anal. Chem., VI. 18. 59.4

Jour. Chem. Soc. (2.), VII. 294. 59.1 ?

Ber. d. d. Chem. Gesell , II. 592. 59.1

Am. Jour. Sci. & Arts, (3.), II. 44. 59.2

Ann. der Chem., CCXXXII. 321. 58.9

Ber. d. d. Chem. Gesell., XXII. 11.

XXII. 891.

Zeit. Anorg. Chem., II. 221. 58.8

Compt. Rend., CXIV. 1149. 60.1

Zeit. Anorg. Chem., IV. 10. 59.8

" IV. 462. 59.5

" VIII. 1.

" VIII. 291.

" XL 73. 58.9

" XVI. 362. 59.0

" XVIL 236.

Proc. Amer. Acad., XXXIV. 351. 59.0

The present paper. 59.0

All except four of these investigations were acconnpanied by similar
ones upon the atomic weight of nickel, and in general the criticisms made
by Richards and Cushman in a recent paper ui)on this subject may be
applied unchanged to the jiresent argument. It is only necessary to call
attention here to the features in which diftereuces existed.

The favorite method for the determination of both constants has been
the reduction of the oxides. It is evident from our work upon this
method that Russell's cobaltous oxide (1863) must have been contami-
nated with a higher oxide; that Zimmerraann's material (1886), treated
with greater care in more perfect apparatus, was more nearly normal,
although still impure ; and that Schiitzenberger (1892) must have driven
off some of the oxygen, which should have been weighed, by his employ-
ment of a high temperature. It was reserved for Herapel and Thiele
(1895) to discover by varying the conditions of work that cobaltous
oxide was incapable of giving constant results, although they did not
attempt to find the reason of the inconstancy. Their irregular observa-
tions were really a proof of the thoroughness and care of their work,
while the wonderful constancy of Zimmermanu's results simply shows a
constancy of imperfect conditions.

The far greater dissociation tension of the higher oxide of nickel makes
the method much less dangerous in this case ; hence the similar results
for the atomic weight of nickel are both more concordant and more nearly
correct than those for the atomic weiirht of cobalt.

RICHARDS AND BAXTER. ATOMIC WEIGHT OF COBALT. 71

Tlie four investigations which had no parallel in the case of nickel
were those of Sommaruga, Weselsky, Reoimler, and Hempel and Thiele.
The first of these need not claim serious attention, for there can be no
doubt that the result (60.) is far too high. Weselsky's contribution was
very similar to Lee's, which has been treated in Richard and Cushman's
criticism. Since the work was obviously crude and incomplete, we must
ascribe its accurate average result (58.96) to a compensation of errors.

Two more investigations remain to be considered. In 1891, when the
" Gnomium" dispute was at its height, Remmler entered the fray. He
fractionated cobaltous hydrate by treating a large quantity of this precipi-
tate with strong ammonia. Each fraction was subjected to a protracted
purification, which, although efTective in removing heavy metals, intro-
duced more impurity than it removed. His final solution of cobaltous
nitrate, from which he formed the oxide by ignition, contained the alkali
and silica taken up by at least six successive evaporations to dryness in
glass or porcelain, in three of which the solution was ammoniacal. With
such hopelessly contaminated material, it is no wonder that the results
from twenty-four fractions varied between limits which were nearly one
per cent apart. As his method of igniting the monoxide to constant
weight in an inert gas was subject to the same errors which have been
discussed in the case of Russell and Zimmermann, no further comment is
necessary. The average of twenty-four determinations was 58.80.

Hempel and Thiele, in 1895, at first attempted to find the ratio
CoO : Co from analyses of the monoxide. This part of their research
has been already discussed.

The remainder of their work consisted in the quantitative preparation
and analysis of cobaltous chloride. They were unable to prepare it in a
wholly anhydrous condition, and realized this fact; hence the part of
their calculation depending upon the weight of the chloride was unsatis-
factory. The final step of their investigation was the weighing of the
argentic chloride obtainable from the chloride combined with a known
weight of cobalt. Although not wholly satisfactory, this part of their
research far excelled the earlier work of a similar kind (Dumas) and
yielded one of the most trustworthy values for the atomic weight of
cobalt (59.91 to 59.94) which had been obtained at that time.

It is clear, from a careful consideration of the details just referred to,
that all the wide variations in the older results are explicable. Hence
no serious argument exists against the atomic weight of cobalt (58.995,
if 0 = 16) determined by our analyses of cobaltous bromide. Since
moreover the still more recent results with the chloride (less than 59.05)

72 PROCEEDINGS OP THE AMERICAN ACADEMY.

and oxide (58.93 to 59.07) confirm this value, one may assume that it is
not greatly iu error. In any case, the Harvard analyses of the bromides
of nickel and cobalt must stand and fall together, for they were made by
methods essentially similar ; and they must at least give the ratio of the
atomic weights (58.706 : 58.995) with some approach to accuracy. Each
of these numbers has found quite independent support for its actual as
well as the relative value, so that for the present the study of these
atomic weights will not be pursued further in this Laboratory.

Cambridge, June 23, 1899.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. Xo. 4. — October, 1899.

ON THE THERMAL CONDUCTIVITY OF VULCANITE.

Bv B. O. Pkikck.

ON THE THERMAL CONDUCTIVITY OF VULCANITE.

By B. O. Peirce.

Presented May 10, 1899. Received August 19, 1899.

Last year Dr. R. W. Willson aud I published in these Proceedings
an account of some determinations of the thermal conductivities of differ-
ent kinds of marble, made by the so called " Wall Method.''

A relatively thin, square, plane-faced slab of the material to be exam-
ined, enclosed between two other slabs of the same material, formed
a rectangular parallelopiped or prism, which was clamped, aud left for
many hours, between the steam chest, A, and the ice box, Z, of the
apparatus * represented by Figure 1. The final temperatures at the

Figure 1.

centres of the faces of the slab to be tested were determined by the aid
of thermal elements, and the flux of heat through a definite central por-
tion of the colder base of the prism was measured. Figure 2 represents
the revolving ice holder, which was kept in motion by the motor, M;
and Figure 3 shows the ice pot used in measuring the heat flux.

* This was partly constructed witli the help of an appropriation from the
Rumford Funl of the Academy.

76

PROCEEDINGS OP THE AMERICAN ACADEMY.

The present paper describes a series of experiments on the thermal
conductivities of different specimens of hard rubber, * or " vulcanite,"
made by the method whicli we described at length last year.

I had at my disposal fourteen different pieces of hard rubber thick
enough to be used conveniently in the apparatus, and much of this was
in sheets about 120 cm. by 50 cm. So much of it as was bought for the
purposes of this investigation came from three well known makers.
Besides this new rubber, however, there were a number of pieces from

Figure 2.

unknown sources, whicli had been lying in tlie laboratory for indefinite
periods. The specific gravities of the different specimens examined
differed slightly from each other, but by not nearly so much as one per
cent in the extreme cases. The specific gravity of the rubber used as a
standard is about 1.202, and its average specific heat between 25° C. and
100° C. is 0.339. The specific heat of sheet rubber usually increases
with the temperature, as may be inferred from the subjoined table, in

* Stefan, Sitzungsbericlite der Wiener Akademie der Wissenschaften, 1876;
Lees, Pliifosophical Transactions, 1893.

PEIRCE. — THERMAL CONDUCTIVITY OP VULCANITE. 77

the second column of which is given the number of calories required to
raise one gram of a certain kind of thin sheet rubber from 25° C. to
the temperatures given in the first column. The third column gives the
average specific heat between 25° C. and these tempei'atures.

Temperatures.

Calories.

Average Specific Heat

35°

3.17

0.317

45°

6.46

0.323

55°

9.91

0.330

65°

13.47

0.337

75°

17.20

0.345

85°

21.22

0.354

95°

25.48

0.364

100°

27.90

0.372

The average specific heat of another specimen of thin hard rubber be-
tween 25° C. and 100° C. was 0.370.

Since hard rubber is an extremely poor conductor of heat, a long time
was required in the case of any prism for the final state to be reached
approximately. After about seven or eight hours, if the temperatures of
the hot and cold boxes were kept quite constant, the temperatures at all
points on the axis could generally be assumed to be within one twentieth
of a degree of their final values ; and, in view of the variable composi-
tion of what is called "hard rubber," "ebonite," or "vulcanite" in the
market, and of the extreme difficulty of keeping the temperature of the
hot box constant within less than one tenth of a degree for any great
number of hours, it did not seem desirable to extend the experiments
further. In what follows, the conductivity is given in every case to
three decimal figures, but it is evident that the third figure is not quite
determined.

Experiment (a). — A compound slab, made up of two plates (A and B),
of a certain kind of hard rubber which I chose as a standard, with their
thermal elements, was placed between two other pieces of hard rubber
to form a prism. Each plate was about 60 era. long by 50 cm. broad.
The average thickness of A was 1.270 cm., and that of B 1.260 cm. In
the final state the thermal elements on the warmer side of A, between
A and B, and on the cooler side of B, indicated 74°. 9, 45°. 2, and 15°. 7
respectively. The rate of melting of the ice in the box was 102
grams in almost exactly 8,760 seconds. Assuming the area of the bot-
tom of the ice pot to be 126.7 square centimeters, and the latent heat

78

PROCEEDINGS OF THE AMERICAN ACADEMY.

of melting ice to be 79.25, this corresponds to a conductivity for each
slab of 0.000311.

Experiment (b). — Plate A with two thermal elements, enclosed by-
two other sheets of hard rubber, was made into a prism with three plates
of glass. In the final state the temperatures of the elements on the
faces of A were 60°. 1 and 24°. 9 respectively. In 11,220 seconds 154.8
grams of ice were melted. This again corresponds to a conductivity
between 60° and 25° of 0.000311.

Two other experiments in which the
final gradients on the axes of prisms built
lip of disks about 20 cm. in diameter
made of this rubber were determined,
failed to show any sensible variation of
the conductivity with the temperature
between 65" and 16°.

Experiment (c). — A compound slab
made of A and a plate, C, of the same
dimensions as A, but purporting to come
from another maker, with their thermal
elements and " guard plates " of rubber,
were used to form a prism for the large
apparatus. In the final state the indica-
tions of the elements on the warm side
of A, between A and C, and on the cool
side of C, were respectively 69°. 8, 41°. 1,
and 13°. 1, so that the conductivity of
C between 41° and 13° appears to be
0.000319.

After this a number of disks 20 cm.
in diameter, which seemed alike in their
physical i:»roperties, were cut from the
standard rubber and used with other
disks to form prisms for the smaller
apparatus.

Experiments (d) to (p). — By the aid
of these standard disks the conductivities of twelve other specimens were
easily obtained. I give in the next table the results, not in the order
in which the experiments were performed, but, for convenience; in the
order of the conductivities of the specimens.

Figure 3.

PEIRCE. — THERMAL

CONDU(

JTl

VITY OF -V

ULCANITE.

cimei

IS. Thickness in Centimeters.

Tempera

ture Limits.

Conductivity.

D

1.584

44° .4

and 16° .5

0.000317

E

1.265

40M

15°. 8

0.000313

F

1.297

40°. 1

14°.9

0.000309

G

1.254

40°. 5

16°. 1

0.000307

H

0.979

37°.0

16°.4

0.000305

I

0.499

50°. 0

23°. 9

0.000282

J

1.003

37°. 0

12M

0.000256

K

0.945

36°. 1

12°.6

0.000254

L

0.797

35°.2

12°. 5

0.000237

U

0.643

33°. 1

13°.2

0.000217

N

0.644

32°. 8

12°. 2

0.000214

0

0.498

30°. 6

12°. 8

0.000200

79

In view of the suspicious decrease of the thermal conductivity with
the thickness of the prisna experimented on, it is well to say that some
manufiicturers apparently use rubbers of different compositions for thick
and for thin plates. Much cheap thin rubber is to be had in the market,
and the specific electrical resistance of some of this is so low as to make
the material useless for insulating purposes. Of the specimens mentioned
in the table, G and H are pieces of expensive rubber made especially for
use in induction coils of high grade. K, which is nearly of the same
thickness as H, is an excellent rubber of high electrical resistance, but is
not quite equal in this respect to H, which has a very different thermal
conductivity. Extended experimentation seemed to show that the indica-
tions of the thermal elements might be trusted whether the plates were
thick or thin. A redetermination of the conductivities of N and O with
different arrangements of the prisms gave 0.000215 and 0.000199 respec-
tively. When a disk of the standard rubber was turned down thin, and
its specific conductivity redetermined, it was found to be almost the same
as before ; but when disks F and I were made slightly thinner by turning
off the polished surface on each side, the conductivity of the thinner disks
was increased by five per cent and three per cent respectively. This
seems to show that in the cases of some sheets of hard rubber the thermal
conductivity of the skin near the polished surfaces is somewhat less than
that of the mass of the material in the sheet. This difference, however,
is too slight to account for the differences of conductivity shown by the
table. It is well to emphasize the fact that the specific heat between
25° C. and 100° C. of the standard rubber is less (0.339) than that of
the thinnest sheet rubber (0.372) that I have experimented upon.

80 PROCEEDINGS OP THfc AMERICAN ACADEMY.

Stefan found for the thermal conductivity of a certain sheet of rubber,
0.787 cm. thick, the value 0.00026; this rubber had a somewhat greater
density (1.22) and a much lower specific heat (0.23) than any of my
specimens, so that the ratio of the conductivity to the specific heat per
unit volume was somewhat greater. Experimenting upon a thin disk of
"ebonite," 1.93 cm, in diameter and 0.0414 cm. thick, Lees determined
the conductivity to be 0.000403.

The Jefferson Physical Laboratory, Harvard University,
Cambridge, Mass., July, 1899.

Proceedings of the American Academy of Arts and Sciences.
You XXXV. No. 5. — October, 1899.

TJVO GENERA OF NORTH AMERICAN DECTICIN^.

By Samuel H. Scudder.

VOL. XXXV. — 6

TWO GENERA OF NORTH AMERICAN DECTICIN^E.
By Samuel II. Scudder.

Received September 4, 1899.

I. Tropizaspis.

Tropizaspis was founded by Brunuer in 1893 (Rev. Syst. Orth., 187),
but without specification of species, except tliat the author knew of but
one, coming from Washington Territory. As Arytropterh steindachneri
from Puget Sound, described by Herman in 1874 from specimens in
Brunner's collection, belongs here, as indicated by me in 1894 (Can,
Ent., XXVI. 180), and is the only species hitherto described, it should
doubtless be regarded as the type. Others are now described, but appar-
ently none of them are as common as that described by Herman, which
is found over a great extent of country, even beyond the Pacific coast,
as it has occurred in Texas and possibly in Nebraska, as will be seen
below. The species may be separated by the following table : —

Table of the Species of Tropizaspis.

A^. Hind femora scarcely or not more than half as long again as the pronotum, at
least in the male.

ii. Lateral carinas of pronotum diverging continuously from the very base ;
supraanal plate of male triangular, apically rectangulate ; anal cerci scarcely
longer than basal breadth ; supragenital plates very large, overlapping each
other and the lateral margins of the subgenital plate, together apically rounded ;

subgenital plate with distinct though small styles ovafa.

b'. Lateral carina; converging on the prozona and then diverging ; supraanal
plate of male quadrangular, apically transverse ; anal cerci nearly twice as long
as basal breadth ; supragenital plates relatively small, attingent, together apically
slightly acutangulate, not reaching lateral margins of subgenital plate ; the latter

with minute, scarcely noticeable styles castanea.

.!-. Hind femora twice or almost twice as long as pronotum ; lateral carina; of pro-
notum converging on the prozona and then diverging.

61. Pronotum broadest far behind the middle of the metazona in both sexes ;
ovipositor rather gently curved.

ci. Disk of pronotum irregularly mottled ; ovipositor slightly narrowed in the
middle (where it is narrower than the frontal fastigium) and faintly enlarged
beyond picturata.

84 PROCEEDINGS OF THE AMERICAN ACADEMY.

(•2. Disk of pronotum generally unicolorous, sometimes longitudinally striped,
rarely irregularly and rather obscurely mottled ; ovipositor not narrower in
the middle than beyond, and broader than the frontal fastigium.

sieindachneri^
b^. Prozona broadest scarcely behind the middle of the metazona, at least in the
female; ovipositor rather strongly curved, not narrowed in the middle; hind
femora basally stouter than in the other species diaholica.

Tropizaspis ovata sp. nov.

Head rather large, testaceous, the fastigium narrower than the de-
pressed basal joint of the antenna?, anteriorly truncate, sulcate. Pro-
notum nearly half as long again as broad, broadest posteriorly, ovate,
anteriorly truncate, the prozona plane, the metazona gently tumid, the
whole disk glistening testaceous longitudinally streaked with black, espe-
cially down the middle and at the sides ; lateral carina? rather sharp and
somewhat prominent laterally but scarcely elevated, diverging continu-
ously from the very base, the hind border strongly rounded and margined ;
median carina sharp but feeble, equal, percurrent ; lateral lobes uniform
testaceous. Legs testaceous, feebly clouded with fuscouS; the hind femora
less than half as long again as the pronotum. Abdomen uniform testa-
ceous ; supraanal plate of male triangular, apically rectangulate but
rounded, the sides with a slight basal tooth ; anal cerci stout, brief, coni-
cal, scarcely longer than basal breadth ; supragenital plates large, tumid,
superiorly carinate, overlapping each other and surpassing laterally the
subgenital plate, together apically rounded ; subgenital plate with dis-
tinct styles.

Length of body, 21.5 mm.; pronotum 11.5 mm.; breadth of same,
8 mm. ; length of hind femora, 16.5 mm.

1 ^ . California, D. A. Saunders.

Tropizaspis castanea sp nov.

Head rather large, testaceous, irregularly marked with fuscous in
slender stripes, the fastigium dotted with luteous, a little narrower than
the first antennal joint, apically truncate, plane ; antennae long enough
to reach a little beyond the hind femora, testaceous broadly banded with
fuscous. Pronotum with asymmetrically clepsydral disk, about half as
long again as broad, broadest posteriorly, very gently tumid, faintly pec-
tinate transversely and posteriorly faintly rugulose, castaneous, the lateral
carinae distinct but rather blunt, feeble and converging on the prozona,
on the metazona diverging and with the bluntly margined hind border
forming a rather short ovoid; median carina faint, equal, percurrent; lat-

SCUDDER. — NORTH AMERICAN DECTICIN^. 85

eral lobes castaneous, passing below into luteous. Legs castaneous, the
femora flecked above with fuscous, the hind pair only half as long again
as the prouotum. Abdomen castaneous, faintly clouded and obscurely
flecked irregularly with fuscous ; supraanal plate of male quadrangular,
apically squarely truncate, the lateral angles slightly prominent; cerci
rather short, but very much longer than basal breadth, subconical ; supra-
genital plates attingent, together triangular, apically acutangulate, lying
well within the margins of the subgenital plate ; the latter slightly com-
pressed, so as to be V-shaped as seen apically, the styles minute, scarcely
noticeable.

Length of body, 24 mm.; pronotum, 12 mm.; breadth of same,
7.5 mm. ; length of hind femora, 18 mm.

1 $. Los Angeles Co., California, Coquillett. Received from the
U. S. National Museum, where are other specimens.

Tropizaspis picturata sp. nov.

Head not very large, testaceous mottled with brown, the fastigium
narrower than the bat^al joint of the antennae ; antennae testaceous, broadly
interrupted with fuscous. Pronotum with clepsydral disk, which is more
than a third as long again as broad and broadest posteriorly, the bent-
arcuate lateral carinas sharp but little elevated, the posterior border deli-
cately margined, broadly rounded, the median carina sharp, not very
elevated, equal and percurrent, the whole disk testaceous, irregularly but
somewhat symmetrically pictured with brownish fuscous, the lateral lobes
fusco-iufumate except the luteo-testaceous lower margin. Legs testa-
ceous, the fore and middle femora flecked and banded more or less con-
spicuously with fuscous, the hind femora slender, more than twice as
long as the pronotum, testaceous dotted and clouded with fuscous,
assuming more or less the form of brief oblique fascise externally, just
above the outer inferior carina. Abdomen testaceous, the sides especially
on basal half heavily banded with fuscous, leaving only a dorsal testaceous
band about as wide as the narrowest part of the pronotal disk, rapidly
broadening on the fourth segment ; genitalia of the male much as in
T. steindachneri, but in the single male at hand distorted by contraction,
so as not to be described accurately ; ovipositor castaneous, basally lutes-
cent, rather gently curved, about three quarters the length of the hind
femora, sliijhtlv narrowed in the middle (where it is narrower than the
frontal fastigium) and faintly enlarged beyond, the denticulations acute,
erect, separated by their own height.

Length of body, $, 14.5 mm., 9? 18-5 mm.; pronotum, ^, 7 mm..

86 PROCEEDINGS OF THE AMERICAN ACADEMY.

9 , 8 mm. ; breadth of pronotum, ^ , 5 mm., 9, 5.5 mm. ; length of hind
femora, (?, 17 mm., 9, 20.5 mm. ; ovipositor, 15.5 mm.
1 ^, 19- Northwestern boundary survey, Kennedy.

Tropizaspis steindachneri.

Arytropteris steindachneri Herm., Verb. Zool.-Bot. Ges. Wien, XXIV. 204-205,
PI. 6, Figs. 98-102 (1874).

Tropizaspis steindachneri Scudd., Can. Ent., XXVI. 180, 18.3 (1894).

This is altogether the commonest form of Tropizaspis, to judge from
the specimens I have seen, and is somewhat variable in coloration and
markings, though usually rather uniformly testaceous or fusco-testaceous,
with the upper part of the lateral lobes of the pronotum infuscated.
Herman states that the median carina of the pronotum stops before the
bind border, but it is percurrent in all the specimens I have seen, 46 in
number.

It was originally described from Puget Sound. I have seen specimens
made on the northwestern boundary survey by Dr. Keunerly ; from Oi'e-
gon (Oregon City, Harford; Mary's Peak, Benton Co., Sept. 16, Divide,
Sept. 12, Drain, Sept. 11, Gleudale, Sept. 9, and Siskiyou, Sept. 6,
Morse); California (Mendocino; Napa Co., Edwards ; Napa or Sonoma
Co., Osten Sacken) ; and El Paso, Texas, Uhler. I have also specimens
fi'om Uhler's collection marked as taken by Suckley in Nebraska, but I
am inclined to think this an error, as Suckley collected also in the north-
west, and Bruner has not given it in his Nebraska lists.

Tropizaspis diabolica sp. nov.

Head moderately large, castaneo-testaceous, the fastigium slightly
broader than the basal joints of the antennte. Pronotum with rather
feebly clepsydral disk, which is considerably more than half as long again
as broad, broadest but little behind the middle of the metazona, and of
nearly equal breadth throughout the posterior half of the pronotum,
the lateral carinre not elevated on the prozona, on the metazona a little
elevated but blunt; hind border margined, broadly rounded ; median
carina feeble, subequal, percurrent ; the whole disk uniform castaneous,
transversely pectinate, feebly rugose posteriorly, the lateral lobes fusco-
fuliginous on upper half, castaneo-testaceous below. Legs luteo-testaceous,
the fore and middle femora twice banded with blackish fuscous, the hind
femora coarsely and a little transversely spotted above witli blackish
fuscous, much more than twice as long as the procotum, the basal portion

SCUDDER. — NORTH AMERICAN DECTICIN.E, 87

unusually stout. Abdomen dull luteo-testaceous, faintly iufuscated in
blotches laterally ; ovipositor rather strongly curved, not narrowed in
the middle more than beyoud, luteo-testaceous, a little infuscated in the
apical half, nearly two thirds as long as the hind femora, the denticula-
tions blunt, oblique, separated by more than their own height.

Length of body, 25 mm.; pronotum, 10.5 mm.; breadth of same,
6 mm.; length of hind femora, 25.5 mm. ; ovipositor, 16 mm.

1 9. Monte Diablo, California, August, 1872.

I have seen another species of Tropizaspis, closely allied to T. stein-
dachneri, but easily distinguished by the stout apical spine on the inner
margin of the narrower supragenital plates of the male, collected by
A. P. Morse on Mt. Wilson, Altadena, California, July 27. The hind
legs are long as in T. steindachneri, and the coloring similar. Unfortu-
nately the single specimen was badly damaged before opportunity occurred
to describe it.

II. Cacopteris.

Cacopteris was proposed by me in 1894 (Can. Ent., XXVI. 178, 181;
see also Guide N. A. Orth., 1897, 56) as a new genus of Decticiufe for
several Pacific coast species, all of which were said to be new and none
were at that time described. It was peculiar in that the upper surface
of the fore tibiae was sometimes spined and sometimes unarmed on the
inner side, this being sometimes the case even within the range of a single
species. The species known to me are herewith described, and a table
given for their separation. Probably others will be found, as many of
these are known by only one or two examples.

The genus may be characterized as having an unarmed prosternum ;
the pronotum smooth, similarly arched in front and behind, with rarely
any lateral or median carinas (and when present obscure and generally
partial) ; the legs long, the hind femora generally extending well beyond
the abdomen ; the tegmina mere pads mostly or quite concealed by the
pronotum in the female, small, apically rounded, and scarcely longer than
broad in the male; the supraanal plate of male bifid, the cerci generally
subcylindrical, often more or less incurved, with an inner median hook
generally of considerable size, the subgenital plate apically emarginate
with a pair of small and slender styles ; and the ovipositor long and
straight or very faintly upcurved. It has somewhat the aspect of the
European Antaxius.

88 PROCEEDINGS OP THE AMERICAN ACADEMY.

The only species known are confined to the Pacific coast region of
North America, though some occur east of the Sierra Nevadas. Some
of the species originally referred to it do not belong here, so that there
are inaccuracies in my brief remarks in the Canadian Entomologist, loc.
cit., p. 181. G. (xqualis may be regarded as the type.

Table op the Species of Cacopteris.

a^. Male with no spines on inner margin of fore tibiae above, and rarely with any
black spot at outer apex of tegmina. Female also with no spines on inner margin
of fore tibias above, or rarely with a single one.

b^. Pronotum relatively short, being but little longer than broad; apical denta-
tions of supraaiial plate of male blunt, at least as broad as long.

ci. Tegmina of male f usco-testaeeous, unicolorous ; supraanal plate of male

broadly and roundly emarginate at apex hiermis.

c". Tegmina of male liglit testaceous, witii a distinct black apical spot;
supraanal plate of male deeply and angularly emarginate apically.

Jliscopunctaia.
b^. Pronotum relatively long, being about half as long again as broad ; apical
dentations of supraanal plate of male acuminate, at least twice as long as
broad.

ci. Lateral lobes of pronotum with a slight inferior flange anteriorly ; fore
femora as long as pronotum ; cerci of male prolonged apically beyond tlie
inner median tooth to a bluntly acuminate spine reaching well beyond the

subgenital plate simialu.

C-. Lateral lobes of pronotum with a pronounced inferior flange anteriorly ;
fore femora shorter than pronotum ; cerci of male with a short blunt conical
extension beyond the inner median tooth, not reaching the tip of the sub-
genital plate e/ihipjiiata.

a^. Male with one or two, generally two, spines on inner margin of fore tibiae
above, and a distinct black spot at outer apex of tegmina.* Female generally
spined on fore tibise as much as male, rarely without spines.

b^. Ovipositor almost or quite as long as hind femora, which are little if any
longer than the body.

ci. Ovipositor mucii longer than hind femora ; apical dentations of supraanal
plate of male triangular, somewhat brief, separated by a V-sha[)ed cleft ;
apical hook of cerci comparatively stout, apically bent .... nevudensis.
c^. Ovipositor about as long as hind femora ; apical dentations of supraanal
plate of male slender, acuminate, separated by a U-shaped sinus ; apical
hook of cerci comparatively slender, feebly incurved at tip .... cequalis.
6'^. Ovipositor only tiiree fourths as long as hind femora, which are much longer
than the body Jemoruln.

* The male of C . femorala is unknown, but as the female has two or three spines
on the inner margin of the fore tibiae above, it evidently belongs in this category.

SCUDDER. — NORTH AMERICAN DECTICIN^. 89

Cacopteris inermis sp. nov.

Blackish fuscous, the whole dorsum of the female testaceous, ofteu
flecked or clouded with fuscous, the sides of the abdomeu of same
sex often sprinkled with testaceous. Head testaceous below the lower
level of the eyes, often more or less flecked inconspicuously with fuscous;
antennas rufo-fuscous, rufous basally, rather longer than the body. Pio-
notum with the lower and posterior margin edged with luteous, thu
lateral lobes deep, with scarcely perceptible posterior sinus, strongly
bent-arcuate below, the whole but little longer than broad. Tegmina of
male fusco-testaceous, unicolorous, of female testaceous. Legs rather
short, fusco-testaceous, the lower half of hind femora testaceous ; fore
femora much shorter than the pronotum ; fore tibiae normally with no
spines on inner margin above (a single female has one) ; hind femora
very much shorter than the body. Supraanal plate of male strongly
transverse, the apical lobes broad, short, and rounded, the intervening
sinus broad and arcuate ; cerci stout, strongly incurved, both the stout in-
ner median tooth and the depressed apical projection triangular, strongly
incurved and decurved, and sharply pointed ; subgenital plate with
acutangulate apical emargination, the styles as long as apical breadth of
plate ; ovipositor castaneous, marked with black at base, straight, much
longer than hind femora, though shorter than body.

Length of body, (J, 17 mm., 9 > 21 mm. ; pronotum, 5 mm., 9 > 6 mm. ;
fore femora, c? , 3.75 mm., 9, 4mm. ; hind femora,,^, 14 mm., 9 , 14.5 mm.;
ovipositor, 18.5 mm.

2 (?, 3 9- Vicinity of Lake Tahoe, Nevada, Sept., H. W. Henshaw,
Lt. Wheeler's exploration of 1876.

The supraanal plate and cerci of the male' are quite diifereut from
those of any other species. It also has shorter legs.

Cacopteris fuscopunctata sp. nov.

Dark fusco-testaceous, the face blackish fuscous and behind the eye a
narrow piceous stripe crossing the pronotum but broadly interrupted on
the prozona; the legs dark fusco-testaceous, the tegmina light testaceous
with a distinct black apical spot, the antennas half as long again as the
body, rufo-testaceous, narrowly ringed with fuscous at the apex of the
joints. Pronotum but little longer than broad, the lateral lobes deep
with no posterior sinus, bent-arcuate below, the disk with a faint median
carina. Legs of moderate length, the fore femora slightly shorter than
the pronotum, the fore tibite without spines on the inner margin above,

90 PROCEEDINGS OP THE AMERICAN ACADEMY.

the hiud femora about as long as the body. Supraanal plate broad tri-
angular with a very deep apical sinus, separating two dentiform triangular
lobes not so long as broad and apically acute, not reaching so far back as
the subgenital plate ; cerci fairly stout with similar and subequal, conical,
indirected, median and apical spines of no great size, blunt tipped ; sub-
genital plate subtruncate apically, with a feeble median emargination,
the styles slight, less than half as long as the apical breadth of the plate.
Length of body, 14 mm.; pronotum, 4.5 mm.; fore femora, 4 mm. ;
hind femora, 14 mm.

1 ^. Tehachapi, Cal., Aug. 2, A. P. Morse.

This species bears a close general resemblance to C. cBqualis, from
which it is distinguished by its unarmed fore tibiae, slightly shorter hind
femora, and the male genitalia.

Cacopteris sinuata sp. nov.

Light olivaceo-testaceous, the inferior margin of the lateral lobes of
pronotum broadly bordered with luteous, and faint luteous stripes in
place of lateral carinte. Face yellow testaceous, narrowly edged below
next clypeus with rufous. Pronotum about half as long again as broad,
the lower margin of lateral lobes sinuate, the lobes being but little deeper
on anterior than on posterior half of pronotum, the sinus rather slight.
Tegmina dark fuscous, with luteous veins. Legs long, the fore femora
slightly longer than the pronotum, the fore tibiae with no spines on
the inner margin above. Abdomen with a pair of narrow, luteous,
latero-dorsal stripes fading basally ; supraaual plate developing an
apical pair of long, triangular, uniformly tapering and acuminate
lobes, separated by a deep and narrow fissure, the lobes reaching as
far as the subgenital plate ; cerci cylindrical, long, sinuate, with a
brief postmedian, interior, conical, pointed tooth, beyond which the
cerci taper to a blunt point, this whole apical portion surpassing the
subgenital plate ; the latter with a subrectangulate apical emargination,
the styles slender, less than half as long as the apical breadth of the
plate.

Length of body, 20.5 mm. ; pronotum, 6 mm. ; fore femora, 6.5 mm.

2 $ . Fort Whipple, Arizona, E. Palmer.

The colors are probably affected by immersion in alcohol. Unfortu-
nately neither specimen has hiud legs preserved.

SCUDDER. — NORTH AMERICAN DECTICIN^. 91

Cacopteris ephippiata sp. nov.

Luteo-testaceous, the lower anterior margin of lateral lobes bordered
with luteous, the outer face of the stouter part of the hind femora with a
slender median and broader and longer inferior fuscous streaks, the teg-
mina blackish, with the main lateral veins luteous. Pronotum about
half as long again as broad, the lateral lobes twice as deep in the anterior
as in the posterior half, the sinus deep, the lower margin of deeper por-
tion of pronotum strongly arcuate. Fore femora considerably shorter
than pronotum, the fore tibite with no spines above on inner margin, the
hind femora long, as long as the body. Supraanal plate with a pair of
subattingent, long and slender, tapering, depressed and pointed dentiform
lobes, nearly reaching the tip of the subgenital plate, separated by a deep
and narrow fissure ; cerci similar to those of C. sinuata but stouter and
shorter, the tip not reaching so far back as the subgenital plate ; the lat-
ter rather narrow apically with rectangulate emargination.

Length of body, 20 mm. ; pronotum, 7.1 mm. ; fore femora, 6 mm. ;
hind femora, 20 mm.

1 $ . Sonora, Schott, Emory's Mexican boundary survey.

Cacopteris nevadensis sp. nov.

Fusco-griseous, the lateral lobes of pronotum more or less marked
with piceous especially above, with a luteous patch bordering the infero-
posterior margin ; head somewhat lighter below the lower level of the
eyes, the legs concolorous, the tegmina of the male dull testaceous, with
an apical black spot outwardly ; antennse much longer than the body,
rufo-testaceous. Pronotum but little longer than broad, the lateral lobes
deep, the lower margin strongly bent-arcuate with no sinus posteriorly.
Legs rather long, the fore femora as long as {^) ov shorter than ( 9 ) the
pronotum, the fore tibiae with one or two {^) or no (9) spines on the
inner margin above, the hind femora as long as ( (^ ) or shorter than ( 9 )
the body. Supraanal plate of male triangular, somewhat brief, fissured
apically by a deep V-shaped cleft, forming on either side a short, trian-
gular, pointed lobe not reaching so far as the subgenital plate ; cerci with
the basal half cylindrical, rather stout, at its end furnished with an inte-
rior hook, beyond which the cerci form an incurved, tapering, acuminate
spine, bent before the infuscated tip ; subgenital plate apically narrowed
by compression, rectangularly emarginate, the styles moderate, more
than half as long as the apical vv'idth of the plate ; ovipositor straight,
longer than the hind femora, but shorter than the body.

92 PROCEEDINGS OP THE AMERICAN ACADEMY.

Length of bod}', ^ , 17 mm., 9, 22.5 mm.; pronotum, J', 4.75 mm.,
9 , 5.5 mm. ; fore femora, J*, 4.75 mm., 9? 4.5 mm.; hiud femora, (J,
17 mm.,, 9 5 18 mm. ; ovipositor, 20 mm.

2 (^ , 19- Ruby Valley, Nevada, Ridgway ; mountains near Lake
Tahoe, Nevada, Oct. 14, Henshavv, Lt. Wheeler's explorations of 1876.

The female, which alone comes from the last named source, may possi-
bly belong to a distinct species ; it differs in having a darker and more
uniform coloring, with nearly uniformly fuscous lateral lobes (except the
luteous patch) as well as in the points brought out in the desci'iptiou.

Cacopteris sequalis sp. nov.

Fusco-testaceous, often clouded with fuscous, generally but not always
with a fuscous patch on the lateral lobes of the pronotum, posteriorly
overhanging a crescentic luteous bordering to the lower margin ; face gen-
erally a little lighter colored below the lower level of the eyes ; antennae
nearly twice as long as the body, luteo-testaceous ; tegmina of male testa-
ceous, with a distinct black spot at the apex ; hind femora testaceous,
generally more or less infuscated, occasionally streaked longitudinally
with blackish fuscous externally. Pronotum scarcely longer than inferior
breadth, with deep lateral lobes, the lower margin strongly bent-arcuate,
with a slight posterior sinus. Legs long, the fore femora nearly or quite
as long as the pronotum, the fore tibiae with a pair of spines on the
upper inner margin, the hind femora slightly longer than the body.
Supraanal plate of male with an apical pair of rather widely separated,
slender, acuminate denticulations, as long as the body of the plate, reach-
ing nearly as far as the tip of the subgeuital plate, and separated by a
wide U-shaped sinus ; cerci stout and cylindrical in basal half, the apical
half forming an inbent, apically incurved, slender, tapering hook, spring-
ing from the inner apical part of basal portion ; subgenital plate apically
rectangularly emarginate, the styles moderate, neai-ly as long as the apical
breadth of the plate ; ovipositor faintly upcurved, about as long as the
hind femora.

Length of body, r? , 19.75 mm., 9, 21 mm.: pronotum, $, 5.25 rnm.,
9, 5.5 mm.; fore femora, $, 5 mm., 9, 5.75 mm.; hind femora, J*,
20.5 mm., 9 , 22 mm. ; ovipositor, 22 mm.

9 (J , 5 9. Los Angeles, Cal., July 26, A. P. Morse, D. W. Coquil-
lett ; Cahon Pass, Cal., July 19, A. P. Morse. Two immature 9 from
Mt. Wilson, Altadena, Cal., July 27, A. P. Morse, may also belong
here. Other specimens are in the U. S. National Museum.

SCUDDER, — NORTH AMERICAN DECTIClNiE. 93

Cacopteris femorata sp. nov.

Nearly uniform very dark fusco-testaceous, the hind femora lighter
testaceous except for a black median stripe on the outer and inner sides,
larger without, the head fusco-castaneous ; antennae nearly twice as long
as the body, rufo-testaceous. Pronotum but little longer than inferior
breadth, the lateral lobes angularly separated from the disk posteriorly,
forming slight blunt lateral carina?, the lobes deep, a little lighter colored
below, the inferior margin strongly bent-arcuate, with a slight posterior
sinus. Legs unusually long, the fore femora slightly longer than the
pronotum, the fore tibije with two or three spines on the inner margin
above, the hind femora much longer than the body. Ovipositor concol-
orous, feebly upturned, almost as long as the body.

Length of body, 20.5 mm. ; pronotum, 5.5 mm. ; fore femora, 6 mm. ;
hind femora, 25 mm.; ovipositor, 19 mm.

1 9 . South Santa Monica, Cal., July 30, J. J. Rivers through A. P.
Morse.

Easily distinguished by the carinje of the pronotum, the long hind
legs, and relatively short ovipositor.

Proceeding of the American Academy of Arts and Sciences.
Vol. XXXV. No. 6. — November, 1899.

h

NOTE ON THE FINITE CONTINUOUS GROUPS

OF THE PLANE.

By F. B. Williams.

NOTE ON THE FINITE CONTINUOUS GROUPS OF

THE PLANE.

By F. B. Williams, Clark University, Mass.

Presented by Henry Taber, October 11, 1899,

Since Professor Study* made the important discovery that the special
linear homogeneous group contains singular transformations, i. e. transfor-
mations that cannot be generated by an infinitesimal transformation of
this group (in consequence of which the group is not continuous except
in the neighborhood of the identical transformation), such singular trans-
formations have been found by Professor Taber t and others, in many
other sub-groups of the general projective group. Thus, e.g., Mr. Eettger
has shown that of the 76 two and three term sub-groups of the projective
group in two variables, and of the general linear homogeneous group in
three variables, 21 contain singular transformations.! It was therefore
to be expected that, for example, among the groups of the plane given
by Lie on pages 360 and 361 of his Continmerliche Gruppen, some, not
sub-groups of the projective group, would be found to contain singular
transformations. This I find to be the case, as the second group consid-
ered below will show. The first group considered is projective for the
value of r taken ; and, in connection with the consideration of this group,
there is given a method by means of which we are able to ascertain
whether a group contains singular transformations or not.

Throughout this paper p = ^r— and q = ;r— .

Example I.
If ia the case of the group

q, xg, x-q . . . x'-^q, p, xjp^ayq,

(r>3)

* Leipziger Berichte, 1892.

t Bull. N. Y. Math. Soc, July, 1894; Math. Ann., Vol. XLVI. p. 561; Math.
Review, Vol. I. p. 154. See also Newson, Kansas Univ. Quart., 1896.
X See These Proceedings, Vol. XXXIII.

VOL. XXXV. — 7

98 PROCEEDINGS OF THE AMERICAN ACADEMY.

we put r = i, we have the group

q, xq, p, xp + ayq,

which is a sub-group of the general projective group.

The symbol of the general infinitesimal transformation is

U= (ai -\- a.x -{- a^ai/)q + {a-^ + a^x)p.
Hence,

Ux = a^ -{- a^ X,

U"x = a^Ui -\- a^ a;,

where ZZ^ re denotes U{Ux), etc. Similarly,
Uy = a^ + a2X + a^ay,

U'^y = «! «4 « + aoa^ax •\- a^^ u"y -\- 02 a^ x -\- a2 a^ ,
U^y = aiai^u" + a2a^a'^x-^ a^ v? y ^ a2a^ {a -}- l)a;+ a2a^a^{a-\- 1),

U'y = aia4»-^«"-^ + a2ai''-'a''-^x + a^Vy + a2a4""^(«""'+ «""'+ • • • « + l)a;

+ 02«3«4""^(«""^ + «"^ + ... + « + 1).

Therefore,* the transformation of the group generated by the general
infinitesimal transformation of the group is defined by the equations

X = xe -\ (e — I),

(1)

a^x

«4 (« — 1)

a., ^3 / « «4 «4 -1 , ^ , ^1 / " ^4 1 \

« a4^ (a — 1) a a4

Let this transformation be denoted by Ta- It transforms the point P
with coordinates (ar, ?/) into the point P' with coordinates (x', y). Let
the transformation 7^ of our group (generated by the infinitesimal trans-
formation (bi + b2X -\- b^ixy) q + (bs -\- b^ x)p) transform P' into P" with
coordinates (x" , y"). Ti, is then defined by the equations

* Lie, Differentialgleichungen, chap. 3, § 3.

WILLIAMS. — FINITE CONTINUOUS GROUPS, 99

x= X e -\- — (e — 1),

(2) ■

« 64 (m — 1) 0(04

Substituting in (2) the values of x' and y' from (1), we get the transfor-
mation 7), 2\ which carries the point P into the point P" . The equa-
tions defining Tf, T^ are then

x" = x/^ + '^ + ^ (e"^-*-'^ - e'O + f^ (/^ - 1),

(3) "^ **

«" = «/^"* + '^^ + ^ i "^ (,«(«. + *.) _ /. + «^^) + ^(,«^ _ /4) I
ce — 1 (^ a4 O4 )

-j — (e —ae —e + «e ) H (e — e )

aa4"(a — 1) tt«4

+ ~Tu n ^^ -ae -\ + u) + —r{e - 1).

a O4 (m — 1 ) u 0^

If now 7J, Ta, which is also a transformation of our group, is equiva-
lent to a transformation T^, generated by the infinitesimal transformation
(ci + C2 X + C4 ay) 7 + (C3 + Ci x)p, — that is, if T^ T^— T^, we have
also

X = xe -\ (e — 1),

(4)

0^4

Whence, first,

/* = /^ + \

o C4 a (ai + 64)

e = e ;

and therefore,

C4 r= a4 4- &4 + 2 Kiri,
aCi = a (04 + ^4) -f 2 k' TT t,

100 PROCEEDINGS OF THE AMERICAN ACADEMY.

where k and k are integers. From these equations it follows that a k
is an integer. Therefore, if « is irrational, k = 0. On the other hand,

if a is rational and equal to - , where fx and v are integers relatively

prime, k = A. j/, where X is an arbitrary integer.
We also derive from (3) and (4)

o'i + ^i + 2 /< TT ^ C «3 . , , 6,^ K , T. . )

e"i + ^* — I C «4 04 )

«4 + ^4 + 2 K TT Z ( «2 . « ('^'^ + 64) O4 + a 64 Jo . a 64 J4. )

= (^ (a, &),

«4 + ^4 + 2 K TT Z f 4> \p a{at + bi)

Co —

Ci =

^a (04 + fti)

■.TTl { <ji \p a

"F^ («4 + ^4 + 2K^ir(a-l) ^'

''l + ^4 -II \ 1 Cf2 ^3 / « (O4 + 64) (74 + a 64 a 6,

— «e — 14-«;H — ^- -r (e — ae — e

04- (« —.1)

a 64. «l,a(a4 + 64) 064, ^2^3 /»64 A

+ ae)H (e — e)-|- ^-r-r rr (e — ae

04 O4" (« — 1)

-1 + „) + *; (.-'-!)}

If for finite values of the «'s and J's, while some of the c's may remain
finite, one (or more) becomes infinite in all branches, there is no
infinitesimal transformation of the group that will generate T;, T^, i. e.
T^ T^ is a singular transformation.

Let n be irrational. Then k = 0 ; and for all finite values of the a's
and S's, Cj^ is finite. But, if 04 + (^4 = 2 m it i for some integer m r}i 0,
Ca is infinite, provided

«4 ^ O4 \«4 O4/

Similarly, if {n — 1) (^4 + ^4) = 2 m tt « 4^ 0, c^ is in general infinite;
and, if a (a^ + b^) = 2 m tt ^ =|= 0, Cj is in general infinite.

Let now « = ~ ; then c^ is, as before, finite. In this case, as stated

V

above, k = X v, where X is an arbitrary integer ; and if a4 + J4 = 2 m tti
^ 0, Cg, and therefore c^, are in general infinite unless

2 TT ^ {in -\- \v) = «4 + J,j + 2 K TT ^' = 0) —

WILLIAMS. — FINITE CONTINUOUS GROUPS. 101

that is, unless m contains v. Therefore, if a^ -{- b^ =■ 2 m tt i :i^ 0 and

V 4: 1, we can always so choose m that c^ shall, in general, be infinite
and Tf^ T^^ singular. On the other hand, if 04 + J^ = 2 m tt i, and if

V = 1 (i. 6. if a is an integer), one branch of c^ is always finite, and the
same is true for c^ and Co : so in this case 7], T^ can be generated by an
infinitesimal transformation of our group.

When a is rational there are, however, always singular transforma-
tions of the group. For let

, 2 m TT ^ 2mvTri
«4 + 04 =

— 1

p.

Then in general (i. e. provided the function of the a's and 5's found in
the second factor in the expression for c^ is not zero), Co, is infinite unless

h A. ) = « . + ^4 + 2 K TT ^■ = 0 ;

IX — V J

which is impossible if m is so chosen that it shall not contain [x — v.

, , . . , . . . .r. 7 2m7rz'^

Therefore, whether a is rational or irrational, 11 a^ + 6^ = -^ 0

a — 1

(where m is an integer which if « is rational and equal to - does not con-
tain fx — v), e.j is in general* infinite, and consequently Tf, T^ cannot be
generated by an infinitesimal transformation of our group; i.e. 7], T^^ is
then singular.

Among the singular transformations of our group obtained by putting

a 4- (^4 = 4^ 0 (where if « is rational and equal to — , the integer

« — 1 V

m does not contain /a — i/), let us consider those for which, further,
«g = ^3 = 0. These singular transformations are defined by the
equations

Irmri

(5)

y' =y e ''-1 + 31 X + ^V {M^ 0).

The singular transformations T'defined by equations (5) leave invariant,
as a whole, the system of lines x := const., but change each line into

* I. e. provided ^ (e""^*** + **) - e"* + «*4) + ^ (e^^* _ A), which in this case

«4 °4

becomes (-"e"* — ^) (e** - e"*«), is not

\"4 ^i^

zero.

102 PROCEEDINGS OF THE AMERICAN ACADEMY.

some other line of the system. Associated with T is a one-term group
whose path curves, jc = c, are as a whole unchanged by T. The path
curves of our group are given by the equation

dx dy

az + a^x «! + a^x -\- a^ay'

the solution of which gives

y = a/(i-«) ^"^ + "^^ ^ ^<— + ^(^^ + «*^)'''

where y is the constant of integration. If in the symbol of the general
infinitesimal transformation U we put a^ — 0, as — 0, and a^ and a™ finite,
we get the one-term group, whose symbol of infinitesimal transformation
is Ui = (oi + 02 x) q, and whose path curves are x — const. ; which is
then the one-term group associated with the singular transformations T.

Example II.

e g, xe q, .

. . x'^e q, p

K := 1, 2, .

. . m, «« =: const..

2pk -\- tn r — \,

r > 2.

Put r = 3, K = m = 1, and p^ — 1. We then have the group

e"'^^, xe-"^ q, p.

(«to)

The symbol of the general infinitesimal transformation is

7-7 / a.x \ aX\ I _ «-.

U = («i e -j- a^x e ) q + a^ j).
Hence,

Ux =^ Oj,

U-'x = 0,
and therefore x' = x + a^. Further,

TT aX \ (XX

U y = 016 -\- a<txe ,

TT-'* a X I a X I _ aX

(J-y =^ aia.,ae -\- ana-^e -\- a.^a^axe ,

T7"a 'i ^ o-X I Ck Q a X I 09 a3?

T-TW + 1 'iwaaji 'in. — lax, n n ax

LI ^ y ^ a^a^ a e ■\- na^a^ a. e -{- o^a^ a x e .

WILLIAMS. — FINITE CONTINUOUS GROUPS.

103

Therefore,

« flJo

(e""^ — 1)

r

+ 0= -1 -^ + "S
I"

'M ^IJ^. ^

»3 ' J J

Hence the oo^ of uon-singular transformatioas Ta have the form
x' = X + as,

(1)

= « + a:e"^,^(«) +e»^>/'(«)-

^ =y

where

e'^Of' _ 1

<}>(a) = a2 ,

e""3 _ 1

«/'(«) = Cti h «2 < ^ + «3

a a.,

\«3

<9a3

j-"-0+-.+°-f-)[

V «2 0!3 ^

« a. a ( \ M 03 / M «3 z )

(2)

Similarly, the transformation Tb may be defined by the equations

x" = x + ^3,

And, therefore, T'j Ta is defined by the equations

x" — X -\- a^ -{- Jg,

(3)

y" =y-r xe'^^\<^{a) + e""'^ (5)) + ^^ i^\, {a) + a^e^^^c^Ci)

If now Tft T'a = 7*^ , we have also

a; := a; + C3,

(4)

y" ^y\ xe<^^</.(c)4-e«^^(c).

Therefore,

C3 = a„ + Sg,

104 PROCEEDINGS OF THE AMERICAN ACADEMY,

Whence we derive
Cz — as + K

= n {a, b),

ea(a3 + 63)_l|_ „ I \ a(a, + b,)J

« (flSg + 63) 2 5 J

From these equations it follows that c^ is finite for finite values of tlie

a's and 6's ; but if 03 + ^3 = =[= 0, where k is an integer not zero

(« being either rational or irrational), then Ci and Co are, in general, both
infinite in all branches (and, indeed, Ci is infinite to the second order),
that is, unless

(5) -^ (e««3 _ 1) 4_ ir_ (ea(<'3 + &3) _ e-%) = 0,

a O3 « 63

and

— C

(6) -

L V «(«3 + 63)y «(«3 + ^3) 2 J

+ ,/. (a) + a, e'^''3 ^ (J) + g^^ ^ (5) ^ Q.
If (5) is satisfied and (6) is not, we have Co finite and c^ infinite to the

^ K ^ 2 ^ K 71 7/

first order for «. + ^3 = ^ — =1= 0. Therefore, if a, + b„ = =^ 0,

a , " a

where k is an integer, Tj T^ is, in general, singular.

Among the singular transformations of our group obtained by putting

^ K 7T t

ciz + ^3 = 4^ 0 (k an integer), let us consider those for which,

a

further, a.o, = bo = 0. Equation (5) is then satisfied ; and these singular

transformations are defined by the equations

, 2 K ni . . , ^-

X ^^ X -] {k an integer df 0),

a

(7)

/ = y + i¥e"^ (J/i 0).

WILLIAMS. — FINITE CONTINUOUS GROUPS.

105

The singular transformations T defined by equations (7) leave invariant,
as a whole, the system of lines x =: const., but change each line into some
other line of the system. Associated with Z" is a one-term group whose
path curves, a: = c, are as a whole unchanged by T. The path curves
generated by the general infinitesimal transformation of this group are
defined by the equation

dx dy

the solution of which gives

e"'^ ipi tt -{■ cu a X — a.2) = as a" y + c.

(c = const.)

If, now, in the symbol of the general infinitesimal transformation U, we
put ttg = 0, cTo = 0, and Oj finite, we get the one-term group, whose symbol
of infinitesimal transformation is l/i = aie'^^q, and whose path curves
are x = const. ; which is then the one-term group associated with the

singular transformation T.

The following groups do not contain singular transformations, and are
properly continuous groups.

Example III.

<1^

(^0 ix)q, . . . ^^_^ {x) q,
r> 2

y<i

Put r =: 3 ; we then have the group

7, <A(x)(7, yq-;

and
Therefore,

U z=^a^q ^ a.,(f> (x) q -f- a.yq.

Ux = 0',

Uy = ai + a.2 <j!) (x) + a.y,

IPy — a^ «3 + «2 "z 4> (^) + «3^y,

Vy^ «i ffs"- ^ + «2 «3" - 1 </> (x) + a,-y.
Hence the transformation T^ of this group is defined by

106 PROCEEDINGS OP THE AMERICAN ACADEMY.

(I)

and the transformation T^ by
(2)

Therefore, if T^ T, ^ 2\,

C3 = «3 + *3 + 2 K 71 ^■.

For finite values of the a's and 5's, every branch of Cg is finite, and at
least one branch both of Ci and of Cn is finite. For Ci and C2 can only be
infinite for a^-\- bz = 2 7nni {m an integer) ; but if «3 + Jg = 2 m tt /,
the branches of q and c.^ corresponding to k = — m are finite, being equal
respectively to

L*3 (e«3 - 1) ^ + (e*» - 1)1 = ^'' (^"' - 1) (^' - 1 \

03 + 63 = 2 m TT I
and

03 + 63 = 2nvi

The groap

is likewise continuous for values of r > 3 ; i. e. for values of r > 3, it
does not contain singular transformations. For, if r = p > 3 the trans-
formations Ta are defined by

/ai-\-a2^o(x)-ir as4>p,(^) + • ■ 4- flp-ic^p-i (a;)V "p_i\

WILLIAMS. — FINITE CONTINUOUS GROUPS.

107

Example IV.

a X ax par

e" q, xe" q, . . . , x^e" q, ij q, p
K — 1, 2, . . . , W2, «K = const., 2pK + m = r — 2, r>3

Put p^ =: 1, K := m = 1, and r = 4. We then have the group

and

Therefore,

e 'q, xe'^^q, y q, p ;
V = (flj e'^' + a^x e"^ + a^i/) q + 04^3 .

Ux = 04, ZPa: = 0, . . . ?7"a: = 0,
a;' = X + a^ ;

so that
and

fT'^y = aiOge"^ + a^agxe'"' + a^y + e*^(aia4a + Oo^O + xe"'*a2a4«,

f7^y = a^a^e'^ -\-a<2,a^xe'^ -{■aiy-{-e'^ {axai^aar^-{-a^a^a^-\-xe'^ a^a^aa^
+ e"^ (ai a/ «- + 2 as 04^ a) + a; e»^ Og a^ a%

U^y — a^aie'^^a^aixe'^-{-a^^y-\-e'^{axa^aa^-^aia^a^)^xe'^a^a^ua^"
+ a;e'"a2a4^a^a3,

Hence

y' =ye''^ ■\- X e

Cfn

(e««* _ e%)

+ ea2

(^ a 04 — «3 J

}fll(a3-aa4)+«2fl4l(e"8-l)-ai(a3-aff4)(e°"«-l)-q2fl'«e°^*(Q3-g"4 + ^]

(a3-aa4)'-2

For the values of r, p^ , and k, chosen, this group is continuous.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 7. — November, 1899.

THE ECHELON SPECTROSCOPE.

By a. a. Michelson.

Investigations on Light and Heat made and published wholly or in part ttith Appropriations

prom the rumfobd fuku-

THE ECHELON SPECTROSCOPE.
By a. a. Michelson.

Keceived and Presented October 11, 1899.

The important discovery of Zeeman of the influence of a magnetic
field upon the radiations of an approximately homogeneous source shows
more clearly than any other fact the great advantage of the highest at-
tainable dispersion -and resolving power in the spectroscopes employed
in such observations.

If we consider that in the great majority of cases the separation of
the component lines produced by the magnetic field is of the order of a
twentieth to a fiftieth of the distance between the sodium lines, it will be
readily admitted that, if the structure of the components themselves is
more or less complex, such structure would not be revealed by the most
powerful spectroscopes of the ordinary type.

In the case of the grating spectroscope, besides the difficulty of obtain-
ing sufficient resolving power, the intensity is so feeble that only the
brighter spectral lines can be observed, and even these must be aug-
mented by using powerful discharges, — which usually have the effect of
masking the structure to be investigated.

Some years ago I published a paper describing a method of analysis
of approximately homogeneous radiations which depends upon the
observation of the clearness of interference fringes produced by these
radiations. A curve was drawn showing the change in clearness with
increase in the difference of path of the two interfering pencils of light;
and it was shown that there is a fixed relation between such a " visi-
bility curve " and the distribution of light in the corresponding spectrum,
— at least in the case of symmetrical lines.*

It is precisely in the examinations of such minute variations as are
observed in the Zeeman effect that the advantages of this method
appear, for the observations are entirely free from instrumental errors ;
there is practically no limit to the resolving power, and there is plenty
of light.

* In the case of asymmetrical lines anotlier relation is necessary, and such is
furnished by what may be called the "phase curve."

112 PROCEEDINGS OF THE AMERICAN ACADEMY.

There is, however, the rather serious inconvenience that the examina-
tion of a single line requires a considerable time, often several minutes,
and during this time the character of the radiations themselves may be
changino;.

Besides this, nothing can be determined regarding the nature of these
radiations until the " visibility curve " is complete, and analyzed either
by calculation or by an equivalent mechanical operation.

Notwithstanding these difficulties, it was possible to obtain a number
of rather interesting results, such as the doubling or the tripling of the
central line of Zeeman's triplet, and the resolution of the lateral lines
into multiple lines ; also the resolution of the majority of the spectral
lines examined into more or less complex groups ; the observation of
the effects of temperature and pressure on the width of the lines, etc.

It is none the less evident that the inconveniences of this process are
so serious that a return to the spectroscopic methods would be desirable
if it were possible, 1st, to increase the resolving power of our gratings ;
2d, to concentrate all the light in one spectrum.

It is well known that the resolving power of a grating is measured by
the product of the number of lines by the order of ihe spectrum. Atten-
tion has hitherto been confined almost exclusively to the first of these
factors, and in the large six-inch grating of Professor Rowland there are
about one hundred thousand lines. It is possible that the limit in this
direction has already been reached; for it appears that gratings ruled
on the same engine with but half as many lines have almost the same
resolving power as the larger ones. This must be due to the errors in
spacing of the lines ; and if this error could be overcome, the resolving
power could be augmented indefinitely.

In the hope of accomplishing something in this direction, together with
Mr. S. W. Stratton, I constructed a ruling engine in which I make use
of the principle of the interferometer in order to correct the screw by
means of light-waves from a homogeneous source. This instrument
(only a small model of a larger one now under construction) has already
furnished rather good gratings of two inches ruled surface, and it seems
not unreasonable to hope for a twelve-inch grating with almost theoreti-
cally accurate rulings.

As regards the second factor, the order of the spectrum observed, but
little use is made of orders higher than the fourth, chiefly on account
of the faintness of the light. It is true that occasionally a grating is
ruled which gives exceptionally bright spectra of the. second, or third
order, and such gratings are as valuable as they are rare ; for it appears

MICHELSON. — THE ECHELON SPECTROSCOPE.

113

that this quality of throwing an excess of light in a particular spectrum
is due to the character of the ruling diamond, which canuot be deter-
mined except by the unsatisfactory process of trial and error.

If it were desired to produce rulings which should throw the greater
part of the incident light in a given spectrum, we should try to give the
rulings the form shown in section in Figure 1.

/ /

/

"X/ x/

FiGDKE 1.

I am aware of the difficulties to be encountered in the attempt to put
this idea into practical shape, and it may well be that they are in fact
insurmountable ; but in any case it seems to be well worth the attempt.

Meanwhile the idea suggested itself of avoiding the difficulty in the
following way.

/

/

Figure 2.

Plates of glass (p^, Fig. 2) accurately plane-parallel and of the same
thickness, are placed in contact, as shown in the figure. If the thicknesses
were exactly the same, and were it not for variations in the thickness
of the air-films between the plates, the retardations of the pencils
reflected by the successive surfaces would be exactly the same, the
reflected waves would be in the same conditions as in the case of a
reflecting grating, — except that the common retardation is enormously
greater.

VOL. XXXV. — 9

114

PROCEEDINGS OP THE AMERICAN ACADEMY.

The first condition is not very difficult to fulfil ; but, in consequence
of dust jiarticles which invariably deposit on the glass surfaces, — in
spite of the greatest possible precaution, ■ — it is practically impossible to
insure a perfect contact, or even constancy, iu the distances between
surfaces.*

If now instead of the retardation by reflection we make use of the
retardation by transmission thi-ough the glass, the difficulty disappears

almost completely. In particular the
air-films are compensated by equivalent
thicknesses of air outside, so that it is
no longer necessary that their thickness
should be constant. Besides, the ac-
curacy of parallelism and of thickness
of the glass plates necessary to insure
good results is now only one fourth
of that required of the reflection ar-
rangement.

In Figure 3 let ab =^ s, the breadth
of each pencil of rays ; bd = t, the
thickness of each element of the eche-
lon ; 0, the angle of diffi'action ; «, the
angle adb; m, the number of waves
of length A corresponding to the com-
mon difference of path of the successive
elemeuts. The difference of path is

cos (a + 0) ; or, since 0 is always very

1

1

1 »

1

r

1

1

1

C/--^

/

\d

' *i

/ /'

' , '

/ei

1

1

Figure 3.

mX

H-

t — ad.

ac ^

cos «

small,

t

ac

(cos a — ^ sin tt) = < (1 — 0 tan a),

cos a ^

I.

and m\— ijx— \)t ^ sQ

To find the angle corresponding to a given value d\, differentiate for

dp.

A, and we find -— = -
a A s

m — t

dX

Putting in this expression the approximate value of m
we have

(.-»^,

* Nevertheless I have succeeded with ten such plates, silvered on their front
surfaces, in obtaining spectra which, though somewhat confused, were still pure
enough to show phenomena such as the Zeeman effect, the broadening of lines
by pressure, etc. ; but evidently the limit had been nearly reached.

MICHELSON. — THE ECHELON SPECTROSCOPE. 115

dO

d\/\

|_^' ^ dXjs s

For the majority of optical glasses b varies between 0.5 and 1.0.

The expression II. measures the dispersion of the echelon. To obtain

tlie resolving power, put e = dk/k for the limit. For this limiting value

the ano'le 6 will be k/n s, where n is the number of elements ; hence

n s p= the effective diameter of the observing telescope. Substituting

these values, we find

A.

e =

biit'

III.

To obtain the angular distance between the spectra, differentiate I. for
m ; we find

dm s '

or, putting d m = unity,

de^=^. IV.

The quantity dk/k = E corresponding to this is found by substitut-
ing this value of c^^ in II., whence

Hence the limit of resolution is the 7iih part of the distance between
the spectra.

This fact is evidently a rather serious objection to this form of spectro-
scope. Thus, in observing the effect of increasing density on the breadth
of the sodium lines, if the broadening be of the order of k/bt the two
contiguous spectra (of the same line) will overlap. As a particular case,
let us take t — 7 mm., JiJ = 17 Joo- It will be impossible to examine lines
whose breadth is greater than the fourteenth part of the distance between
the D Hues. It is evidently advantageous to make t as small as possible.

Now the resolving power, which may be defined by - is proportional

to the product 71 1. Consequently in order to increase it as much as
possible it is necessary to use thick plates, or to increase their number.
But in consequence of the losses hj the successive reflections, experience
shows that this number is limited to from 20 to 35 plates, any excess not
contributing in any important degree to the efficiency.

I have constructed three echelons, the thickness of the plates being

116

PROCEEDINGS OP THE AMERICAN ACADEMY.

7 mm., 18 mm., and 30 mm. respectively, each containing the maximum
number of elements, — that is, 20 to 35, and whose theoretical resolving
powers are therefore of the order of 210000, 540000, and 900000,
respectively. In other words, they can resolve lines whose distances
apart is the two-hundredth, the five-hundredth, and the nine-hundredth
of the distance between the D lines.

Consequently the smallest of these echelons surpasses the resolving
power of the best gratings, and what is even more important, it concen-
trates all the light in a single spectrum.

The law of the distribution of intensities in the successive spectra is
readily deduced from the integral

Hence

rs,,

I cosp
J s/2

27r^

X ax. where p = — - 6.

A

I=A^-

sin- TT-a

A

-x'^"

This expression vanishes for 0 = t X / s, which is also the value of dOi,
the distance between the spectra.

Hence in general there are two spectra visible as indicated in Figure 4.

Figure i.

By slightly inclining the echelon, one of the sj^ectra is readily brought
to the centre of the field, while the adjacent ones are at the minima, and
disappear. The remaining spectra are practically invisible, except for
very bright lines.

As has just been indicated, the proximity of the successive spectra of
one and the same line is a serious objection, and as this proximity depends

MICHELSON. — THE ECHELON SPECTROSCOPE. 117

on the thickness of the plates — which for mechanical reasons cannot
well be reduced below 5 or 6 mm. — it is desirable to look to other
means for obviating the difficulty, among which may be mentioned tlie
use of a liquid instead of air.

In this case Formula II. becomes

dd t_

s

dXj\

1 ^ -, ^ dQj. — /xi)~j _ t

- (M — Ml) — '^ JT = C

.1^1

d\ s

and Formula IV. becomes

dm. /ji s

Repeating the same operations as in the former case, we find :

A
net

and E —— .

ct

The limit of resolution is still the nW). part of the distance between
the spectra, but both are increased in the ratio bjc.

Suppose for instance the liquid is water. Neglecting dispersion the
factor would be 3.55. Hence the distance between the spectra will be
increased in this proportion, but the limit of resolution will also be
multiplied by this factor. But as there is now a surfiice water-glass
which reflects the light, the loss due to this reflection will be very much
less, so that it will be possible to employ a greater number of elements,
thus restoring the resolving power. At the same time the degree of
accuracy necessary in working the plates is 3.55 times less than before.

For many radiations the absorption due to thicknesses of the order
of 50 cm. of glass would be a very serious objection to the employ-
ment of the transmission echelon. I have attempted, therefore, to
carry out the original idea of a reflecting echelon, and it may be of
interest to indicate in a general way how it is hoped the problem may be
solved.

Among the various processes which have suggested themselves the
following appear the most promising.

In the first a number of plates (20 to 30), of equal thickness, are
fastened together as in Figure 5, and the surfaces A and B are ground
and polished plane and parallel. They are then separated and i^laced on
an inclined plane surface, as indicated in Figure G.

118

PROCEEDINGS OF THE AMERICAN ACADEMY.

If there are differences in thickness of the air-films, the resulting
differences in the height of the plates will be less in the ratio tan «.

B

Figure 5.

An error of \jn may be admitted for each plate, — even in the most
unfavorable case in which the errors all add ; and consequently the
admissible errors in the thickness of the air-films may be of the order
A/^^«. For instance, for 20 plates the average error may be a whole
wave-length if the inclination a is o'^ . As there is always a more or

FiGUKE 6.

less perfect compensation of the errors, the number of plates or the
inclination may be correspondingly greater. Accordingly, it may be
possible to make use of 50 elements, and the plane may be inclined at
an angle of 20° to 30°. It would be necessary in this case, however,
to use a rather large objective. Possibly this may be avoided by cutting
the surface ^ to a spherical curvature, thus forming a sort of concave
echelon.

MICHELSON. — THE ECHELON SPECTROSCOPE.

119

Figure 7.

The second process differs from the first only in that each plate is
cut independently to the necessary height to give the required retardation.
The first approximation being made,
the plates are placed on a plane surface,
as in Figure 7.

The projections a and h (Fig. 8) are
then ground and polished until the up-
per surfaces are all parallel, and the
successive retardations equal. The
parallelism as well as the height is
verified by means of the interferom-
eter.

These processes are, it is freely
conceded, rather delicate, but prelim-
inary experiments have shown that with patience they may be suc-
cessful.

Figure 8.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 8. — December, 1899.

THE ELECTROCHEMICAL EQUIVALENTS OF COPPER

AND SILVER.

By Theodore W. Richards, Edward Collins, axd
George W. IIeimrod.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF
HARVARD COLLEGE.

THE ELECTROCHEMICAL EQUIVALENTS OF
COPPER AND SILVER.

By Theodore William Richards, Edward Collins, and George W.

Heimrod.

Presented June 14, 1899. Received October 30, 1899.

Four years ago we began an investigation which had as its object the
careful study of the well known irregularities of the "copper voltameter"
from the point of view of the modern theories of electrochemistry. One
reason for this study was the fact that the atomic weight of copper com-
puted from the relation of its electrochemical equivalent to that of silver
was noticeably smaller than the atomic weight obtained in this Labora-
tory by purely chemical methods.* Owing to circumstances beyond our
control the completion of the work was postponed far beyond its proper
time ; but at last it is in a condition suitable for partial publication. The
matter was complicated by the discovery of an error in the " silver volta-
meter," as well as in the " copper voltameter." In the mean time an
extremely interesting paper by Foerster and Seidel, and several other
valuable contributions by Kahle, Patterson and Guthe, and others, upon
allied subjects, have appeared.j These cover a part of the ground occu-
pied by our work, and thereby shorten the present account. The work
naturally falls under two headings, because each metal needed its special
investigation ; and this natural division will be retained in the following
description.

I. The Copper Voltameter.

The work of Gore, Lord Rayleigh and Mrs. Sidgwick, Gray, Shaw,
Vanni, and Gannon, | showed long ago that a side reaction, varying in

* T. W. Richards, These Proceedings, 26, 240 (1891).

t Foerster and Seidel, Zeitschr. anorg. Chem., 14, 106-140 ; K. Knlile, Wied.
Ann., 67, 1 (1899) ; Patterson and Guthe, Physical Review, 7, 257 (1808); E. H.
Griffiths, Nature, 56, 258 (1897), etc.

t Gore, Nature, 25, 473 (1882) ; Lord Rayleigh and Mrs. Sidgwick, Pliil. Trans.,
175, 458 (1884) ; Gray, Piiil. Mag., (5), 22, -389 (1886). also 25, 179(1888) ; Shaw,
Brit. Ass. Rep. 1886, p. .318 [Phil. Mag., (5), 23, 138]; Vanni, Wied. Ann., 44,
214 (1891) ; Gannon, Proc. Roy. Soc, 55, 66 (1894).

124 PROCEEDINGS OP THE AMERICAN ACADEMY.

prominence with circumstances, vitiates the electrochemical equivalent
of copper as it is ordinarily determined. Even a cursory study of these
careful investigations shows that the complication arises at the point of
contact of the copper cathode with the solution, and that the side reaction
has the effect of lessening the amount of copper deposited by the current.
With this in mind, we made a systematic investigation of the behavior
of copper plates in cupric solutions, which led us to precisely the same
conclusions as those attained by Foerster and Seidel in the paper already
mentioned. Since these gentlemen have described their work in great
detail, an abbreviated statement of our results will suffice.

1. Metallic copper slowly dissolves in an acid solution of cupric sulphate,
even when the solution has been freed from air and has been protected by
an atmosphere of hydrogen. This conclusion was tested many times, and
the losses of weight of the plates were found to be roughly proportional
to the respective areas of the plates, if the volume of solution was large.

2. A strongly acid solution does not differ materially in its action from
a weakly acid solution ; hence neither hydrogen nor SO4 ions can be
responsible for the phenomenon.

3. Other things being equal, the action is proportional to the concen-
tration of the cupric sulphate, being very slight when this is absent, even
if much sulphuric acid is present. Hence the cupric ions must be the
active agents, and they can only dissolve the copper according to the
reaction Cu + Cu++[+ SO/'] t.2 Cu+[+ SO"]. In a word, cuisrous
sulphate must be formed.

4. In contact with the air this action naturally takes place more
rapidly than in the absence of oxygen. Instead of losing only about
0.004 milligram per square centimeter per hour in a normal solution
of cupric sulphate at 20", as before, the loss was nearly doubled. Evi-
dently the cuprous becomes cupric sulphate in the oxidizing environment,
and thus opportunity for further reduction is ottered.

5. On the otlier hand, plates of copper immersed in neutral solutions
of cupric sulphate always gain in weight, becoming coated with a film
of cupi'ous oxide. This is a wholly separate phenomenon, due to tlie
hydrolysis of the cuprous sulphate. Of course this hydrolysis cannot
happen in an acid solution ; hence copper deposited electrolytically from
an acid solution is free from cuprous oxide. Obviously, too, the method
of Vaniii, which consists in adjusting the amount of acid so that the
plates neither gain nor lose, is a device for replacing dissolved coi)per
by an equal weight of cuprous oxide, and hence is very faulty from a
scientific point of view.

RICHARDS, COLLINS, AND HEIMROD, COPPER AND SILVER. 125

6. As usual, the speed of these reactions is increased by increase of
temperature; and moreover, the amount of cuprous sulphate which de-
termines equilibrium is greater iu a hot solution than in a cold one.

7. The addition of such substances as cane sugar, even in large
amounts, produces no important effect, for obvious reasons.

8. The addition of much sodic
sulphate diminishes the action, evi-
dently by diminishing the number
of cupric ions present, according to
the law of mass action. In one
case the loss was diminished to half
its normal amount.

Many of these conclusions are to
be found, either implicitly or expli-
citly, in the older accounts, but the
interpretation of the facts is now
more illuminating because of the
assistance of the dissociation hypoth-
esis. The only authority who denies
the solubility of copper in cupric
sulphate is Schuster,* whose experi-
ments involved only a very small
amount of solution, which evidently
required only unimportant amounts
of copper for its saturation.

Foerster and Seidel have shown
that it is possible to dissolve copper
iu allot solution, and crystallize the
metal by cooling. With the same
idea in mind, we constructed an ap-
paratus somewhat different from
theirs, arranged so that this opera-
tion may take place continuously,
with the production of very pure
copper in beautiful minute crystals.
A ring-shaped tube was fixed in a j^^^^j^
vertical plane, one side of the ring

being surrounded by a steam-jacket, and the other by cool water. This
arrangement naturally kept the liquid filling the tube in constant circula-

FiGURE 1. — Apparatus for recrys-
TALLizisG Metallic Copper.

A, steam-jacket for heating copper.
B, copper wires to be dissolved. C, re-
ceptacle for crystallized copper. D, cold

* Proc. Roy. See, 55, 84 (1894).

126 PROCEEDINGS OP THE AMERICAN ACADEMY.

tion ; and if an acid solution of cupric sulphate is thus made to circulate
over copper placed in the hot side, considerable amounts of copper may
be dissolved with ease. The crystals deposited upon cooling in the cold
limb cling loosely to the glass. A more advantageous form of this device
is shown in the foregoing illustration.

From these facts concerning the action of copper on a solution of
cupric sulphate, the precautions necessary for obtaining the weight of
deposit which most nearly represents the true electrochemical equivalent
are evidently the following : —

1. The solution must be as cold as possible.

2. The solution must be acidified, to avoid hydrolysis.

3. The solution must be as dilute as is compatible with the transpor-
tation of the required strength of current to a given electrode without
risk of deionizing hydrogen. The smaller the electrode, the stronger
is the necessary solution.

4. Air must be excluded.

Since even at the freezing point of the saturated solution the action
still takes place, although to a greatly diminished degree, the complete
fulfilment of these conditions will still yield too small a deposit of copper.
In order to overcome the difficulty, Foerster and Seidel * have recom-
mended the saturation of the solution with cuprous ions in the first place ;
but this plan oversteps its mark, and forms almost the only jsiece of faulty
logic to be found in their paper. Obviously if cuprous ions were the only
ones present, twice as much copper would be deposited with a given
current as with pure cupric ions. It is well known that when a mixture
of metals is present, that substance is deposited which gives the lowest
contact-potential with the solution, provided the current is weak and the
solution strong. It is easy, however, to deposit brass from a suitable
solution of zinc and copper sulphates with a strong current. The present
case presents some analogies to this example ; but to it is added the pro-
viso that the ions are convertible into one another, and that the cupric
and cuprous salts must be present in a ratio definite for each concentra-
tion. If some of the cupric salt is removed, for example, less cuprous
salt is capable of existing in the solution, and extra copper must neces-
sarily be deposited in order to establish the equilibrmra, according to the
equation 2 Cu+l;? Cu"*""^ + Cu. Experimental verification of this rea-
soning will be given later.

We are then between Scylla and Charybdis, — from a cupric solution

* Zeitschrift fur anorg. Chem., 14, 137 (1897).

RICHARDS, COLLINS, AND HEIMROD. — COPPER AND SILVER. 127

the deposit must be too small, while from a cuprous solution the deposit
must be too large. Both errors decrease as the temperature decreases ;
but it is evidently impossible to obtain results directly with the copper
voltameter which will exactly correspond with the atomic weight of
copper, except by a chance combination of circumstances. This com-
bination would be attained when the amount dissolved from the cathode
jy the side reaction Cu+ + + Cu ±=; 2 Cu+ just equalled the extra amount
of copper deposited from the trace of cuprous salt in the solution. Such
a fortuitous condition could not be maintained for any length of time,
because the amount of cuprous salt present must steadily increase by the
action at the cathode as well as at the anode. At first sight, it might be
supposed possible to increase the current density to such an extent that
all the copper in the neighborhood of the cathode should be completely
deionized ; but it will be shown that hydrogen begins to be deposited
instead of copper before this point can be reached.

Besides these sources of error another may exist, to which Foerster
and Seidel have called especial attention. This fourth eri'or consists in
the possible semi-deionizing (or partial reduction) of the cupric salt by
the electric current without the deposition of metal. Thus positive
electricity may make its exit from the solution without a corresponding
increase in the deposit, and the observed atomic weight of copper will be
too low. Evidently, however, this action cannot result in the formation
of more cuprous salt than the small proportion corresponding to its
equilibrium with the cupric remaining in the neighborhood. Hence this
fourth cause of error is essentially identical with the first, the simple
dissolving of copper in cupric sulphate. The question then arises as to
whether or not the equilibrium is established more quickly under the
influence of the current than without the assistance of outside electro-
motive force. Both Gore and Gray, in the papers already cited, give
evidence showing that reasonably small current-densities tend to increase
the apparent corrosion of the copper rather than to diminish it. This
might have been predicted, for it is easy to imagine that the equilibrium
is established more quickly with the help of nascent cuprous ions than
with the resting metallic surface of the copper plate. The authors just
cited give no evidence to prove, however, that a strong current causes a
less proportionate gain in weight than a weak one ; indeed. Gray contends
that a strong current may even j^rotect the electrode from corrosion.*
The phenomena which led him to this conclusion will be shown later to
be due at least in part to another cause.

* Phil. Mag., (5), 25, 182 (1888).

128

PROCEEDINGS OP THE AMERICAN ACADEMY.

The only possible method of determining the amounts of these various
modifying eifects seems to be to vary the size of the exposed surface of
cathode; for upon tlie extent of this exposed surface depend the most
serious causes of error. This mode of experimentation was adopted by
Shaw, and especially by Gray in the paper just cited. Gray's results
are so interesting that it seems worth while to repeat a part of the plate
which depicts his observations. The abscissoe register the number of
square centimeters area per ampere, while the ordinates give the values
proportional to the amount of copper deposited by a given current in a
given time. The solution of cupric sulphate had a specific gravity vary-
ing from 1.15 to 1.18, or contained perhaps 250 grams of crystallized
salt to the litre.

Area per ampere

Figure 2. — Diagram drawn from the Numerical Table given by Gray.

(Phil. Mag., [5], 25, 182 [1888].)

It is evident that with large areas of cathode (small current densities)
the fairly straight lines tend to converge at a point somewhat above the
top of the table. On the other hand, when the current density becomes
great, the curves fail of their mark and bend sharply to the left. Gray
dismisses this remarkable tendency with the assumption of the "pro-
tective action " of strong currents already mentioned, and takes as the
true value the results obtained at 2° Centigrade with current densities
higher than 0.01 ampere per square centimeter.

RICHARDS, COLLINS, AND HEIMROD. — COPPER AND SILVER. 129

Since it is precisely this variation of the highest values for copper
which interests us, — for we know that the lower values are certainly
too low, — it seemed desirable to obtain further light upon this question.
We therefore carried out a similar series to Gray's, using weaker solu-
tions, with which the point of inflection in the curve would naturally
come at a lower current density. Cathodes and anodes of copper of
different sizes were placed in separate portions of a solution of cupric
sulphate containing 100 grams to the litre, and the same current was
run through the series of cells. The amounts of copper thus deposited
are given in the following table.

No. of
Analysis.

Strength of
Current.

Area of Cathode
60 sq. cm.

Area of Cathode
25 sq. cm.

Area of Cathode
12| .sq. cm.

Ampere.

Grams.

Grams.

Grams.

n

0.12

0.07325

0.07341

0.07345

b

0.19

0.66856

0.66909

0.66921

c

0.21

0.44460

0.444C9

0.44462

d

0.22

0.31212

0.31225

0.31204

e

0.25

0.45869

0.45918

0.45907

The first fact which is at once apparent is the interesting phenomenon
of a decrease in the deposit on the smallest plate in experiments c. d, and e.
This corresponds to a downward turn in Gray's curve ; and if these
results are plotted upon his scale, the following diagram is obtained.
(See Figure 3, on the next page.)

Such a curve as this is only to be produced by the appearance with
high current densities of another cause, independent of the cuprous-cupric
reaction, but also tending to lower the observed weight of copper. It is
not difficult to suggest what this cause may be. When the current
increases in strength, the copper ions in the neighborhood of the cathode
are no longer able to carry all of the current, and a portion of the load
is taken by the only other positive ion present. But the hydrogen which
thus plays the electrical role of the copper weighs far less than the equiv-
alent weight of the red metal. A small amount of hydrogen, capable
of being occluded by the copper without change in the appearance of the
surface, would cause an immense deficiency in the gain in weight. When
the current strength increases only a little above 0.02 ampere per square
centimeter the deposit of the copper becomes powdery, probably because

VOL. XXXV. — 9

130

PROCEEDINGS OF THE AMERICAN ACADEMY.

1 .000

0.999 '

o
Q

Area per ampere

0.998

Figure 3. — Diagram showing Decrease in Copper with small Cathodes

(due to Hydrogen).

The table above is here jjlotted by taking the higliest weight of copper in each
determination as unity. Great exactness could not be expected from this method,
but it is sufBuiently definite to show a marked deflection of the line when the 100
sq. cm. line is approached, and a sudden downward tendency when the current
density approaclies 0.02 ampere per square centimeter.

of the action of the ocduded hydrogen, and the deficiency assumes
gigantic proportions. Indeed it has been well known for a long time
that this current density is the limit if one makes any pretence to
accuracy ; our results merely show that the action of the hydrogen
begins to affect the deposit loug before it becomes manifest to crude
observation. Thus it seems probable that the left-handed tendency of
Gray's curves, instead of being due to the approach to the true value and
a lessening of the corrosion, was really due to the beginning of a new
source of inaccuracy. From a theoretical standpoint such a phenome-
non appears only natural ; with more dilute solutions the deficiency would
of course appear at lower current densities. Probably the fact that in
analytical work the amount of the deposited metal never corresponds
to the quantity of electricity may often be due to the same cause, in
addition to the oxidizing action pointed out by Danneel.*

Returning now to weaker currents in fairly strong solutions (Exi^eri-
ments a and b), it is evident both from Gray's results and from our own
that the weight of the deposit in this case is approximately a linear
function of the area of the cathode ; in other words, the dissolving of
copper in cupric sulphate is approximately proportional to the surface
of copper exposed, even while the current is running. Since this is the
case, the obvious method of obtaining the true weight of copper would
be to extrapolate for a surface of zero area, — or to add to the weight
deposited upon a medium-sized area the difference between its gain and

* Zeitschr. Electrochem., 4, 153 (1898).

RICHARDS, COLLINS, AND HEIMROD. — COPPER AND SILVER. 131

the gain on an area of twice the size. This method of correction would
certainly give a correction which is too large, for during the electrolysis
the solution is steadily approaching saturation with cuprous ions, which
are deposited but not accounted for by the correction. Nevertheless it
was decided to carry out a series of observations with plates of two
sizes, calculating the results according to this method of correction, and
comparing both corrected and uncorrected values with the results of an
included silver voltameter. In this way an upper and a lower limit for
the electrochemical equivalent of copper might be fixed ; and by using a
low temperature, it seemed probable that the two limits might be brought
near together. Agitation would of course increase the amount of corro-
sion * and diminish the error due to hydrogen ; but we preferred to use
low current densities and rely upon the natural convection of the elec-
trolyte, which is then sufficient if the error is kept low by cold.

Experimental Details.

The cupric sulphate used for the following experiments, after having
been purified in usual ways, was allowed to stand for some time, first with
flocculent cupric hydroxide and then with pure copper, in order to free
it from possible traces of bismuth, silver, etc. Subsequently it was
recrystallized from the purest water. Metallic copper for the anodes
was prepared by electrolysis from this material, after strongly acidifying
its solution with •pnve sulphuric and nitric acids.

The silver used for the anodes was prepared according to the well
known chloride-milk-sugar method of Stas, and was amply pure enough
for its purpose. The argentic nitrate used as an electrolyte was made
from this silver by the action of pure nitric acid. This salt was recrys-
tallized, dried, and fused in order to insure its neutrality. For particulars
concerning these and other precautions, the reader is referred to recent
papers upon the atomic weight of copper and other atomic weights.f

In any given experiment, the galvanic current was sent tin-ough two
silver and at least two copper cells, all being arranged in series in such
a way as to avoid liability to leakage. Every point of contact with the
desk was amply insulated by glass plates, and the wires were arranged
as much as possible in air-lines.

The copper cells were contained in wide-mouthed bottles with a ca-
pacity of 500 cubic centimeters each, provided with well cleaned rubber

* Compare Foerster's results.

t Tl.ese Proceedings, 22, 3i2 (1887), 23, 177 (1887), 25, 95 (1890), 26, 240
(1891), etc.

1^9

PROCEEDINGS OP THE AMERICAN ACADEMY.

Stoppers. Through these stoppers ran the tubes for supplying the
hydrogen used to excIudB oxygen from the process, as well as the sup-

FlGURE 4.

ports for the electrodes. Two anodes (B and C) were used in each cell,
and in all except the earliest experiments the cathodes (A) were of
platinum. A diagram, Figure 4, while illustrating a later modification,

RICHARDS, COLLINS, AND HEIMROD. — COPPER AND SILVER. 133

sufficiently explains the arrangement. At the close of an experiment
the cathodes were immediately immersed in water, and after a thorough
washing in pure water and alcohol were dried as usual. The electrolyte
contained 100 grams of crystallized cupric sulphate in a litre.

The silver cells were essen-
tially similar to those used by
Lord Rayleigh, except that the
cathodes consisted of large light
platinum crucibles instead of
bowls. These crucibles weighed
only 60 grams, although they
were capable of holding 120
cubic centimeters ; they were
provided with lips. A crucible
exposes a smaller sui'face to the
impurities of the atmosphere
and gave in our experiments a
more evenly distributed deposit
than a bowl. The anode con-
sisted of a lump of pure silver,
wrapped in pure filter paper,
and the electrolyte contained
ten per cent of argentic nitrate.
When the current was stopped
the argentic nitrate was decant-
ed through a weighed Gooch
crucible, and, after standing
many hours with several fillings
of pure water, the crucibles
were washed with pure water and alcohol, and dried at 130°. The
decantates were filtered, in order to be sure that no silver had been
lost ; if the Gooch crucible was found to contain any of the deposit, its
weight was added to that of the greater part clinging to the cathode.
In one or two of the earlier determinations, plates were used instead of
crucibles as cathodes ; but since the results thus obtained showed no sys-
tematic variation from those with the crucibles, the table is not compli-
cated by calling attention to this inessential fact. The danirer of losino-
silver mechanically from the corners of a plate is so great that this form
of cathode was soon abandoned.

Figure 5. — Comjion Silver Volta-
meter (f natural size; section).

A, silver anode. B, filter paper covering.
C, lip of platinum vessel forming cathode.

134 PROCEEDINGS OF THE AMERICAN ACADEMY.

lu the table on page 135, each horizontal line reports a complete
experiment, which usually required a day for preparation and another
for accomplishment. At first the copper cells were multiplied ; but later
when the silver seemed to be quite as uncertain as the copper, the nobler
metal received its full share of representation. Some of the earlier
experiments failed for one reason or another, and these failures are not
included in the table. The last six determinations were consecutive ;
and in these the conditions of experiment were much more satisfactory
than in the earlier ones ; for much practice was naturally acquired in the
course of the work. In all the experiments excepting Nos. 1 and 2 the
cupric electrolyte was boiled and cooled to zero in an inert atmosphere
before the electrolysis; and in all excepting Nos. 1, 2, 3, and 6 the cop-
per electrolysis was conducted in an atmosphere of hydrogen.

Attention is again called to the fact that each corrected value for the
copper deposit is computed from two deposits, one on a large and one on
a small electrode. The exposed area of the large electrode was always
about 50 square centimeters ; and in all cases, except those indicated
by the signs ft below, the area of the smaller plate was half that of
the larger. The current densities varied slightly, but were always
less than 0.01 ampere per square centimeter on the smaller plate,
the total current strength usually being about 0.12 ampere for the
system. When the total current strength was kept below this limit,
no systematic variation due to its changes coidd be detected in the re-
sults; hence the individual figures are omitted in this published account.
In order to prevent saturation with cuprous salt, the volume of the
electrolyte used in each cell was fairly large, being usually about 200
cubic centimeters.

The corrected average value from the table for the atomic weight of
copper, G3.5G3, has a "probable error " of 0.004 as far as the accidental
inaccuracies of manipulation are concerned. It has already been shown,
however, that a value obtained in this way must correspond to a de-
posit of copper slightly too great ; for the mode of correction does not
take account of the growing, although slight, presence of cuprous salts.
On the other hand, the values computed from the directly observed
quantities of copper must be too small. The average observed values,
calculated from the sums at the bottoms of the respective columns, are
Cu = 63.525 from the larger plates, and Cu = G3.547 from the smaller
plates. It is evident then, that the true electrochemical atomic weight
of copper, when compared with the silver deposited in the ordinary
silver voltameter, must lie between the limits 63.563 and 63.547. Jt

RICHARDS, COLLINS, AND HEIMROD. — COPPER AND SILVER. 135

Comparison of the Silver with the Copper Voltameter.

Temperature of Silver Voltameter = 15° to 25^. Temperature of Copper Voltameter =: —2- to 0°.

Large

Copper

50 sq. cm.

Small

Copper

usually

25 sq. cm.

First

Silver

(in air).

Second

Silver

(in air).

Corrected
Copper.

Average

Silver.

(cor. to vac.)

Atomic

Weight

of Copper,

Ag=lU7.93.

1

2
3

0.44460

0.81555*

0.24965*

0.44469

0.31623*t

0.24967*1

1.51039
1.07476
0.84780

1.51098

0.84808

0.44478
0.31645
0.24968

1.51064
1.07473
0.84792

63.56
63.56
63.56

4
5
6

1.02182*
0.G6090*
0.62992*

1.02189

0.66128*

0.63018t

3.47066
2.24545
2.14008

2.14105

1.02186
0.66166
0.63027

3.47056
2.24538
2.14050

63.56
63.61
63.56

7
8
9

0.45869
0.39140
1.11054

0.45907t
0.39165t
1.11042

1.56000
1.33075
3.76996

3.77003

0.45919
0.39177
1.11030

1.55995
1.33071
3.76990

63.54
63.55
63.57

10
11

12

0.67592
0.48238
0.83036

0.67578
0.48235
0.83064

2.29667
1.63731

2.82200

2.29655
1.63816
2 82222

0.67564
0.48232
0.83092

2.29655
1.63768

2.82203

63.51
63.58
63.56

13
14
15

0.63407
0.69956
0.84341

0.63449
0.70029
0.84375

2.15780
2.37916
2.86580

2.15701
2.37840
2.86654

0.63491
0.70102
0.84409

2.15735

2.37868
2.86608

63.53
63.61
63.57

16
17

087448
0.69379

0.87455
0.69392

2.97121
2.35652

2.97126
2.35729

0.87462
0.69405

2.97114

2.35683

63.54
63.57

11.01704

11.02085

11.02.363

37.43593

63.563

* These determinations were made in duplicate. Tlius, in Experiments 2, 3,
and 5, four copper cells were in the current at once. The pairs agreed within
themselves remarkably well, the average deviation from the mean being only 0.03
milligram.

t In these experiments, the smaller copper was J as large as the larger one.

t In this determination the smaller copper was ^ as large as the larger one.

136 PROCEEDINGS OP THE AMERICAN ACADEMY,

is to be noticed that the value (Go. 54) obtained from Gray's observa-
tions by our method of correction is near these figures.*

If the absolute accuracy of Faraday's law is assumed, these results
show either that the silver voltameter must yield nearly 0.1 per cent too
much silver, or else that the value of the atomic weight of copper found
chemically in this Laboratory (63.604) must be 0.1 per cent too high.

Convincing as these conclusions seemed, the attempt was made to ob-
tain yet further light upon the copper voltameter by varying the conditions.
For example, several experiments were made with a mercury cathode,
in the hope that the amalgam of copper might be less easily influenced
by side reactions than the metal itself; but the difficulties in the way of
determining accurately the weight of the large volume of metal which
was needed to contain the copper, soon led to the abandonment of this
attempt. Moreover other chemical difficulties, due to the possible dis-
solving of mercury, added complications.

It seemed now worth while to make a few determinations of the
amount of copper deposited from a solution saturated with cuprous salt,
in order to fix the highest possible limit for the electrochemical equiva-
lent in a fashion wholly free from any correction, as well as to test
experimentally our criticism of Foerster and Seidel's remarks about
such a solution. In order to saturate the liquid with cuprous ions, the
weighed platinum cathode was I'aised above the solution by means of its
sliding support, and the current was run backward and forward for an
hour or more between the two coils of copper (wrapped in filter jjaper,
B and C, Fig. 4) which were to serve as anodes. Instead of the large
bottle of the earlier experiments, a test tube just large enough to contain
the electrodes was used, so that the small amount of necessary solution
could be more easily saturated. In due time the cathode was lowered,
and the quantitative electrolysis commenced. Of course the solution had
been boiled, and was protected by a current of hydrogen throughout
the experiment. The silver voltameter used was an improved arrange-
ment, but its results are reduced to the standard used in the earlier
experiments for the sake of ready comparison. This matter will be
fully explained in the sequel.

This value of the atomic weight, 63.573, is 0.06 per cent higher

* Gray in liis first paper reports that with plates of 240 sq. cm. area he obtained
0.29303 gram of copper for every gram of silver, while with plates of 50 sq. cm. area
he obtained 0.29407 gram. Hence with plates of zero area he would have obtained
0.29434, corresponding to the atomic weight given above. (Phil. Mag., [5], 22, 407
[1886]. Last three determinations in table.) ^

RICHARDS, COLLINS, AND HEIMROD, — COPPER AND SILVER. 137

Comparison of the ordinary Silver Voltameter with the Copper
Voltameter saturated with Cuprous Salt at 0°.

Weight of Copper
(ia Vacuum).

Weight of Silver
(ia Vacuum).

Atomic Weight
of Copper.

2i

0.718^7

2.43935

63.578

25

0.71861

2.43940

63.585

26

0.72019

2.44603

63.556

27 a

0.97193

3.30100

63.558

28

0.50916

1.72859

63.582

29 «

0.76188

2 58664

63.580

Average .

. 63.573

than the average of the forty observed results in cupric solutions, and
0.016 per cent higher than these results after correction. As far as the
copper deposit is concerned, this result is certainly the upper limit of the
electrochemical value, although it is still below the chemical atomic
weight. No experimental proof has yet been given that too much copper
was deposited, however.

If the cuprous salt really carries a part of the current, it is obvious
that higher temperatures, involving the presence of more cuprous salt,
should raise the observed value still higher. This matter is tested in the
results which follow.

Comparison of Silver Voltameter with Copper Voltameter saturated
WITH Cuprous Salt at higher Temperatures.

No of
Experiment.

Temperature.

Weight of Copper
(Vacuum).

Weight of Silver
(Vacuum).

Atomic Weight
of Copper.

27 5
296

or

55°

0.97295
0.70214

3.30100
2.58064

63.62
63.61

This gain in the apparent atomic weight, from 63.573 to 63.615, is
conclusive. Even without further knowledge of the silver voltameter
one is able to predict with reasonable certainty that higher temperatures

138 PROCEEDINGS OF THE AMERICAN ACADEMY.

should yield yet higher values, and hence that the presence of cuprous
salt really causes the deposit to be too heavy.

The results of all these varied operations assign a wide range of
apparent values to the atomic weight of copper. In order to present the
matter in a nutshell, a table is given below.

Uncorrected results of Rayleigh, Gray, Shaw, Vanni, etc. .
Large plates in cupric solutions, 20° (present research,)

ki U ii (.1. Q° i( ( (

Small " " " " " "

Medium plates, in cuprous solutions, 0° " " .

a t< tt ii gQO u a

Corrected results from cupric solutions, Go. 5 63

Atomic weight of copper by chemical processes, 63.604

Discrepancy . . 0.041

It has already been stated that the corrected value is probably a little
too high, although it is still nearly 0.07 of one per cent below the chem-
ical value. This discrepancy, taken in connection with the inexpli-
cable variations in the results of parallel decerminations with the silver
voltameter (see page 135), led us to suspect this instrument long before
the results given above were completed. Accordingly, many of the
electrolyses which we have been discussing were also used as a means
of testing the accuracy of the silver voltameter. The overlapping inves-
tigations have been separated in the descrijjtion only in order to make a
very complex matter less obscure to the reader. The outcome of the
experiments upon the silver voltameter is explained below.

II. The Silver Voltameter.

The inference to be drawn fi-om the preceding work is that the amount
of silver deposited in the silver voltameter, as described by Lord Rayleigh
and Mrs. Sidgwick * in their classical paper, may be too great by nearly
0.1 per cent. Jn spite of the great care taken by these investigators,
such a deviation from the exact value is not impossible ; and indeed
Lord Rayleigh does not claim for this work an accuracy over 0.1 per
cent.

The first possible cause of error to be investigated was the possible
occlusion (or rather inclusion) of electrolyte by the crystalline deposit.

* Phil. Trans., 2, 411 (1884).

RICHARDS, COLLINS, AND HEIMROD. — COPPER AND SILVER. 139

Lord Rayleigh did not neglect this source of error, and by some very
ingenious qualitative experiments showed that it could not be large. It
seemed to us w^orth while, however, to analyze quantitatively the precip-
itated silver, and the following table recounts this series of trials, in which
the silver was dissolved, and precipitated by the purest hydrobromic
acid with great care.*

No.

Silver taken
(Vacuum).

Argentic

Bromide

obtained

(Vacuum).

Silver
calculated
from AgBr.

Deficiency in
Silver.

Percentage
Deticiency.

30
31
32
33
34
35
36

0.71585
5.43807
3.76993
2.29649
2.15701
2.37893
2.97120

1.245G7
9.46557
6.5G21G
3.99820
3.75473
4.14187
5.17218

0.71573
5.43747
3.76961
2.29G74
2.15G89
2.37923
2.97114

+0.00012
+0.00060
+0.00032
-0.00025
+0.00012
+0.00035
+0.00008

+0.017
+0.011
+0.008
-0.011
+0.006
+0.015
+0.003

+0.007

The result of this table somewhat surprised us; for few crystalline
precipitates contain so little included mother liquor as 0.0] per cent.
While the analytical work is not perfect, (for the result of Experiment 33
can only be ascribed to accident,) one is forced to conclude that the pre-
cipitate is very nearly pure silver. Evidently the electrical method of
precipitation insures a more compact structure than is possible when the
precipitation does not involve outside electromotive force. One of us
found three times as much water in crystallized silver prepared at 0° in
another way.f The source of error in the voltameter is apparently more
subtle than such a merely mechanical cause as inclusion.

Further experimentation upon such silver deposits showed that, after
having been properly washed, they neither gained nor lost in weight
upon continued digestion with water or with neutral solutions of ar-

* For the method, consult recent papers on atomic vveiglits from this Labo-
ratory.

t These Proceedings, 23, 177 (1887).

140 PROCEEDINGS OP THE AMERICAN ACADEMY.

gentic nitrate, hence any possible error must have crept in during the
electrolysis.

The existence of flaws in the working of the silver voltameter is no
new idea, and a brief resume of earlier anomalies is necessary in order
to indicate our own train of thought.

Lord Rayleigh and Mrs. Sidgwick found that large cathodes yielded
higher results than small ones, and warm solutions yielded higher results
than cold ones. In some cases the deviations amounted to 0.1 per cent.

Schuster and Crossley * state that deposits made in a vacuum are
slightly heavier than those produced by the same current in air, and
these in turn are heavier than those produced in an atmosphere of
oxygen. Myers f verified these statements. Schuster and Crossley
showed also that with great current density argentic peroxide may form
at the anode, and in some unexplained fashion the result is a diminution
of the weight of the silver deposited ; moreover, they pointed out the
fact that the discrepancies observed by Lord Rayleigh and Mrs. Sidgwick
between large and small bowls disappear when the anodes are of the
same size. Apparently "//?e anode gives rise to secondary reactions^

Rodger and Watson % observed that on continued use of the electrolyte
the deposits grow heavier ; and they also found that, when a very strong
slightly acidified solution of argentic nitrate was electrolyzed by a
powerful current, the acid was removed, and the deposit was much too
heavy. They venture to say that a subsalt of silver is formed " having a
silver ion heavier than the arseutic salt."

Again, Kahle § has found that after boiling the electrolyte with oxide
of silver, the deposit is increased 5 parts in 10,000. In a later and very
important paper || he calls attention for the first time to the fact that in
a dilute electrolyte an acid is formed during the electrolysis. Further-
more, he shows again that old solutions give too high results, an error
which was sometimes removed by treatment with argentic oxide.
Colored spots sometimes appeared upon the silver in old acid solutions,
apparently caused by the licjuid descending from the anode ; these did
not form in a neutralized solution, or in one which had been allowed to
remain in contact with silver. Kahle's hypothetical explanation of these
phenomena essentially agrees with Rodger and Watson's.

* Proc. Roy. Soc, 50, 344 (1892).

t Wied. Ann., 55, 288 (1895).

t Phil. Trans., 186 A, 631 (1895).

§ Brit. Ass. Adv. Sc. Edinb. (1892), p. 148.

II Wied. Ann., 67, 1 (1899), or Zeitsch. f. Instrkunde, 18, 229, 267 (1898).

RICHARDS, COLLINS, AND HEIMROD. COPPER AND SILVER. 141

The latest work which has corae to our notice is that of Patterson and
Guthe.* In this paper the authors give the lion's share of their attention
to an admirably worked out treatment of the physical methods necessary
for fixing the electrical units, and pay little heed to the chemical side of
the question. Since they obtained constant results by treating their
electrolyte with argentic oxide, they recommend this method for general
use, in spite of the fact that both Rodger and Watson and Kahle have
shown that such treatment probably yields too high results.

The essence of all these investigations, as far as they concern the
irregularities of the silver voltameter, may be summed up in the follow-
ing words.

A substance which causes the deposition of too much silver seems to
be formed around a large anode. Oxygen tends to eliminate this sub-
stance, hence the substance must be a reducing agent. All the phenom-
ena agree with this interpretation. A very small anode (which causes
another irregularity, namely, the formation of argentic peroxide) may
result in the appearance of a lower weight of silver; perhaps argentic
peroxide may oxidize the reducing substance and thus remove it. The
formation of the reducing substance is accompanied by the liberation of
hydrogen ions in dilute solutions of argentic nitrate; but in stronger
solutions small amounts of acid may be neutralized. In warm solutions
the reducing agent is somewhat more active than in cold.

An important point is left in doubt : — Does the acid appear at the
anode or the cathode ? Nothing could be easier than to answer this
question ; and a large number of trials in which the anode was enclosed
in a porous cup showed conclusively that the acid was always formed at
the positive pole. The solution around the cathode remained wholly
neutral to methyl orange, while the liquid within the cup gave a strongly
acid reaction.

In marked contrast with the case of copper, the chief disturbing re-
actions seem then to exist at the point at which the positive electricity
enters the solution ; and the obvious remedy for the irregularity is to
enclose this positive electrode in a cell which will permit conductivity
but will shut out convection.

Experimental Details.

Small cylinders of Pukal's porous ware (Berlin), such as are used for
osmotic pressure experiments, were found to serve admirably as the

* Physical Review, 7, 251 (1898).

142

PROCEEDINGS OF THE AMERICAN ACADEMY.

permeable septum desired. These vessels were 50 millimeters high
and 20 in diameter ; their walls were not much over one millimeter in
thickness. Their impurities were removed by boiling with nitric acid
and thorough washing with water. They were suspended in the solution
by means of a platinum wire hung upon a glass hook, which insulated the

wire from the electric con-
nections. In four experi-
ments two concentric porous
cups were used, the annular
space between them being
charged with argentic oxide,
since Kahle found this sub-
stance to diminish the re-
ducing action in old solutions.
This last arrangement was
soon abandoned, however.
By means of a siphon the
liquid within the cup was
always kept at a lower level
than that without, so as to
prevent outward filtration.

The cathodes consisted of
the large lipped crucibles al-
ready described ; the anodes
were bars 5x1x1 centi-
meters of the purest silver,
supported by silver wires and
not enclosed in filter paper ;
and the electrolyte usually
contained 10 per cent of pure
freshly prepared argentic
nitrate.

The temperature of the
cell under investigation was
kept constant by means of
suitable baths. For a low temperature a mixture of pure ice and
a solution of aramonic nitrate answered well, and for the higher tem-
peratures a large beaker containing pure water was warmed from the
outside by enough encircling leaden steam-pipe to cause the desired
intensity of heat.

^^^^

Figure 6. — Porous Cup Voltameter
(I actual size).

A, glass hook for supporting anode. B, glass
ring for supporting porous cup. C, silver anode.
D, porous cup. E, platinum cathode.

RICHARDS, COLLINS, AND HEIMROD. — COPPER AND SILVER. 143

The manipulation was simple. The platinum crucibles were cleaned,
dried at 160°, and weighed after two or three hours' cooling in a large
desiccator kept in the balance-room. In order to prevent leakage during
the electrolysis, the several stands were insulated by separate glass plates,
and all the connections were air lines. The apparatus was protected, as
in the earlier experiments with copper, by a miniature house with walls
of fine cotton cloth, which helped to exclude dust. When the current
was broken, the electrolyte was removed, the silver was rinsed twice with
water, a third filling with water was allowed to stand in the crucible for
two or three hours, and a fourth one remained in it over night. The
wash-waters were collected and filtered if the silver showed any tendency
to break off. In such cases a Gooch crucible was employed to collect
the particles ; and a very small filter, afterwards burned, served to catch
the minute flakes of asbestos detached from the mat. On the next morn-
ing, the crucibles were washed once more, rinsed twice with pure alcohol,
and finally dried and weighed as before. This method of treatment gave
opportunity for the diffusion of mother liquor from the intricate recesses
of the crystallized mass, while it did not run the risk of dissolving silver
which may attend the use of boiling water for washing. It is probable
that traces of water remained even after drying at 160"^, but the experi-
ments on page 139 show that these traces were less than one part in ten
thousand.

Weighings were made upon the balance which served for the weigh-
ings in the earlier work upon copper, — one which has served also for
many determinations of atomic weights. Its results with small objects
may be depended upon to within ^ milligram. All weighings were
made by double substitution, a similar vessel being used as a tare, and
the weights were of course carefully standardized. Since the question
concerned merely the comparison of silver with silver, the results were
not corrected to the vacuum standard.

The object of the work described below was to determine the relative
values given by the silver voltameter under difi^erent conditions, and to
interpret the results in such a way as to point out the true value. For
such a purpose it is convenient to select some one method as a standard
of reference ; and because the porous cup method, conducted at 20°, soon
showed itself to be capable of giving the most constant results of all the
various modes of procedure, it was chosen as this standard. For
example, in one case a given current caused depositions of 2.4o744 and
2.43749 grams of silver in two cells placed in series, — a difference of
only 2 parts in 100,000.

144

PROCEEDINGS OF THE AMERICAN ACADEMY.

The Silver Voltajieter under Varting Conditions.

Number

of
Experi-
ment.

Weight of

silver

(stanaard) .

Cell at 20°-25°.

AVeiglit of

Silver iu

Voltameter

to be tested.

Difference.

Ratio of stan-
dard Silver
to heavier
Silver.

Method used in
Voltameter
to be tested.

37
38
39

Grams.
1.94124

1.76283

3.04996

Grams.

1.94267
1.76425
3.05270

Milligrams,
1.43

1.42

2.74

1.00074
1.00080
1.00090

Cathode a lipped cru-
cible ; anode wrapped
iu filter paper ; pre-
cisely like the cell
used in preceding
copper work.

Average = 1.00081

40
41

3.09629
2.53256

3.09918
2.53439

2.89
1.83

1.00094
1.00072

Lord Rayleigh's meth-
od. Cathode a howl ;
anode a plate in filter
paper; at 20° J;.

Average = 1.00083

42

2.5.3256

2.53357

1.01

1.00040

The same at 0°.

[1.0012]

The same warm.*

43
44

2.55012

1.89800

2.55460
1.90238

4.48
4.38

1.00176
1.00230

Patterson and Guthe's
metliod ; old electro-
lyte saturated with
AgoO.

Average = 1.00203

45

46

2.44412

3.29842

2.44599
3.30036

1.87
1.94

1.00077
1.00059

Cathode a crucible.
Anode enclo.sed in
porous cup like stan-
dard,but temperature
60-' (instead of 20-).

Average = 1.00008

47
48
49
50

2.63668
1.72724
2.58463
3.09629

2.63822
1.72724

2.58663
3.09988

1.54

0

2.00

3.59

1.00058
1.00000
1.00078
1.00116

i Like the above, ex-
cept that two porous
cups pficked with
Ag.,0 isolated the
anode (60'-).

Average = 1.00063

51

52

2.55012
1.89800

2.55029

1.89845

0.17
0.45

1.00007
1.00024

Cathode a crucible.
Anode enclosed in
porous cup like the
standard. Tempera-
ture 0°.

Average = 1.00016

* From Lord Rayleigh's figures.

EICHARDS, COLLINS, AND HEIMROD. — COPPER AND SILVER. 145

With this cell were compared the following modifications of the silver
voltameter: first, the large crucible voltameter at 20", with the anode
wrapped in filter paper, such as was used in the preceding experiments
upon copper ; secondly, a voltameter at 20° prepared according to Lord
Rayleigh's standard prescription (a platinum bowl, 10 centimeters in
diameter, containing a 15 per cent solution of argentic nitrate and a large
silver plate enclosed in filter paper) ; thirdly, the same arrangement at
0° ; fourthly, an exact imitation of Patterson and Guthe's method, for
which the electrolytes used in Lord Rayleigh's method were digested with
argentic oxide and filtered ; fifthly, the " porous cup method " at 60°
instead of at 20° ; sixthly, the same with a double porous cup enclosing
argentic oxide in the annular space ; and finally, the " porous cup method "
at 0°. The table opposite explains itself.

Discussion of Results.

The comparison of these results is highly instructive. The most
noticeable fact is that our "standard porous cup method" gives the low-
est result of any of the methods tried, since all the figures in the fifth
column are greater than unity. In the next place, we find that Patter-
son and Guthe's method gives a result 0.12 per cent higher than Lord
Rayleigh's ; a figure which corresponds closely with their own estimate,
0.11 per cent, while Lord Rayleigh's method gives results 0.082 per
cent higher than ours. Evidently the method used in the first three ex-
periments is essentially similar to Lord Rayleigh's, and may be averaged
with it. Again, it is apparent that the interposition of argentic oxide in
the hot determinations had a varying and unsatisfactory effect, as if it
introduced a new source of error. Finally, it appears that change of
temperature causes a somewhat smaller variation in the results from the
" porous cup method " than in those from Lord Rayleigh's old standard : —

Lord Rayleigh's Voltameter.

Porous Cup Voltameter

60°

1.0012 (?)

1.00068

20°

1.00083

1.00000

0°

1.00040

1.00016

The obvious inference from these facts is that the porous cup is really
effectual in protecting the cathode from a part at least of the disturbing
influences under consideration, especially at low temperatures. The
apparent gain of 0.016 per cent exhibited by the method at 0° was un-
doubtedly due to the difficulty in washing the very finely divided spongy
metal which is deposited from a freezing solution ; but Lord Rayleigh's

VOL. XXXV. — 10

146 PROCEEDINGS OF THE AMERICAN ACADEMY.

method at 0° is affected by the same source of error, so that the com-
parison is a fair one.

The only alternative to this inference, — the supposition that the po-
rous cup introduces a new complication causing too low a weight, — is
difficult to disprove ; but, on the other hand, the possible nature of such a
complication is not easily conceivable.

Recurring to the first inference, an hypothesis supported by many
facts, we admit it to be probable that the porous cup does not wholly
prevent the admixture of the hypothetical complex ions with the liquid
around the cathode ; for even if diffusion were wholly prevented, some of
these ions might migrate with the current. Both diffusion and migration
are increased in rapidity by a rise in temperature, hence even with the
porous cup the hot solution gives a somewhat higher value.

It is a matter of great interest to study the effect of the low value
obtained by the new method upon the electrical atomic weight of copper,
upon the electrolytic value assigned to the ampere, and upon the elec-
trical result for Joule's equivalent.

We concluded, at the close of the section devoted to copper, that the
atomic weight of this metal must lie between the limits 63.547 and
63.563, if the common silver voltameter is correct in its verdicts. We
have now shown that the results of this voltameter are in all probability
too high by 0.081 per cent; hence the conclusion is that the true atomic
weight of copper must lie between 68.598 and 63.615. Since the
elaborate chemical investigation of copper carried out some years ago in
this Laboratory yielded the value 63.604, we have here a remarkable con-
firmation of the results of the " porous cup voltameter."

In this connection, it is worth while to call attention to the reasons
why copper placed in an ice-cold solution of argentic nitrate causes the
deposition of the true amount of silver.* Of this automatic voltaic cell
the anode is copper ; hence the complications which arise around a silver
anode do not exist. On the other hand, the cathode is silver; hence the
complications which take place around a copper cathode do not exist.
The trace of cuprous salt undoubtedly formed at the anode does not de-
compose the very cold ions of nitric acid, but sooner or later deposits
its equivalent of silver. The action is so slow at the low temperature
that the current density where the silver is deposited is very small, hence
there is no danger of the deposition of a weighable amount of copper

* Richards, These Proceedings, 22, 342, 23, 177 (1887).

RICHARDS, COLLINS, AND HEIMROD. — COPPER AND SILVER. 147

with the silver; * and the concentration effect prevents the deposition of
silver in any place where the silver has become nearly exhausted from
the solution.

It is possible to apply a correction to the various values which have
been assigned to the electrochemical equivalents of copper and silver with
the help of the tables given above.

Tliis computation is made below ; it does not pretend to be precise,
but will show iu a general way the effect of the complications under
consideration.

(Grams per
Ampere-second.)

Lord Rayleigh and Mrs. Sidgwick . . . ^;^!!"i^,^ = 0.0011170 f

Fr. and W. Kohlrausch l'o008^>nO ^ ^■^^'^'^^^^

Kahle (fresh solutions) ^'^^^^^^^^ = 0.0011173

Patterson and Guthe -^-tttt^tt^ = 0.0011174

The average of these results is about 0.0011173, but perhaps 0.001 1172
(the mean between the two extremes) is a safer value to choose.

From this value of the electrochemical number for silver, that for
copper may be calculated simply by multiplication with the ratio of the
chemical equivalents ; for we have shown that its true value corresponds
to Faraday's law within one part in five thousand. Hence it may be

taken as 0.0011172 X ttt^^t-^ = 0.0003292 grams per ampere-second. |

* Lord Rayleigh and Mrs. Sidgwick, Phil. Trans., 175, 470 (1884).

t The probable reason for the lowness of this result is the fact that Lord
Rayleigh waslied and ignited the silver at high temperatures. All the otlier
values, including our own, may be a trifle high because of traces of occluded
mother liquor. No attempt is made at present to correct the results for this pos-
sible error. The results of Masquart (J. de Phys., [2], 3, 283 [1884]), Pellat and
Potier (Ibid , [2], 9, 381 [1890]), Kopsel, Heydweiller, and others are omitted from
this table, since they throw no further light chemically. Some are too low and
others too higli, hence they would not have much effect on the average. Compare
Patterson and Guthe's paper.

I The value 0.000o294 found by F. E. Beach from a solution containing chlorine
is probably affected by the presence of cuprous salts. (Am. J. Sc. [3], 46, 81
[1893].) All the other directly determined values for this quantity are lower than
our value.

148 PROCEEDINGS OF THE AMERICAN ACADEMY.

Thus 96,610 coulombs correspond to one gram equivalent of an
electrolyte.

These numbers are of interest from a theoretical standpoint, but their
practical value as means of determining current strength must depend
upon tlie ease of their application. The value for copper is a mean be-
tween two limits, and is not easy to reproduce with exactness ; but for
most purposes the weight deposited by current densities between 0.008
and 0.012 amperes per square centimeter, from ten per cent solutions of
cupric sulphate cooled below zero and protected by an inert atmosphere,
is sufficiently close to the real value. Nevertheless, empirical tables
like those of Gray will probably remain the most convenient method of
computing current strength from the deposition of copper.

On the other hand, the intricacy added by the porous cup in the silver
determination is not serious, and the results obtained seem trustworthy.
Obviously any method capable of being repeated with constant results, when
standardized with accuracy by means of known currents, would serve all
practical purposes ; and a method which is constant because the sources
of eri-or have been at least partially removed is certainly worthy of
attention.

This point suggests a discussion which has recently ariseu concern-
ing the electrical determination of Joule's equivalent, which depends upon
measuring the heat corresponding to a known amount of electrical
energy. Griffiths, in an appendix* to an elaborate paper f upon this
subject, calls attention to the fact that an error in the electrochemical
equivalent of silver of 0.1 per cent would explain the difference between
his value of the mechanical equivalent of heat and Rowland's. The
equation of Griffiths for the calculation of his value is

in which i? signifies the constant potential difference, R:= the resistance
converting electricity into heat, M the heat-capacity which is heated, 8t
the interval of time, and SO the change of temperature. The electro-
chemical equivalent of silver enters into this expression in the determina-
tion of the value of £ in reference to H. Griffiths's value for J is larger
than Rowland's, hence his value for E is greater than that which Row-
land would have observed if he had used an electrical method; or, in
other words, too small an amount of silver may have been taken as the

* Nature, 56, 258 (1897). i Proc. Roy. Soc, 53, 6 (1S03J.

RICHARDS, COLLINS, AND HEIMROD. COPPER AND SILVER. 149

amount corresponding to the ampere in computing Griffiths's E. Such
an inaccuracy may have come about iu two ways : either the electro-
magnetic methods for determining the energy of an ampere may have
caused the original investigators of the ampere equivalent to overestimate
the strength of their current, or else they may have used a form of volta-
meter which gave a smaller deposit of silver than the form used by
Glazebrook in standardizing Griffiths's cells. This last supposition is the
only one with which the voltameter is concerned, for it is clear that
a constant error in the amount of silver, occurring both in the original
comparison of the ampere value, as well as in the standardizing of
Griffiths's cells, could have no effect, — the weight of silver would be
eliminated from the result. The only safe mode of comparing results of
this kind is to reduce all figures to the standard of some one form of
silver voltameter, capable of giving constant values, as has been done
above. ■Patterson and Guthe, in their otherwise valuable paper, do not
seem to realize this fact, and only add to the confusion by appearing to
suppose that the results with their plethoric voltameter are directly appli-
cable to Griffiths's equivalent. (See line 4, p. 281 of their paper in the
Physical Review, Vol. 7.)

This paper is by no means a final statement of the matters which it
concerns. It is in the nature of a preliminary contribution, and we in-
tend now to test more rigidly the accuracy of the "porous cup volta-
meter," as well as to study from a physico-chemical standpoint the cause
of the anomalies which are rendei'ed harmless by its use. Some light
has already been obtained, and more seems not far off.

Summary.

In conclusion, a brief summary of the chief points treated in the
paper may serve as an index to the more intricate details.

1. Metallic copper dissolves in cupric sulphate with the formation of
cuprous sulphate, and unless acid is present cuprous oxide or hydroxide
is formed by hydrolysis ; in short, most of Foerster and Seidel's results
are confirmed.

2. On the other hand, the saturation of the solution with cuprous
salts is shown to cause too high results in the copper voltameter.

3. Vanni's method of correcting the copper voltameter by regulating
the amount of acid present is shown to be unsound, although its results
are not very erroneous.

4. Before the cupric solution is made so dilute that its solvent action

150 PROCEEDINGS OF THE AMERICAN ACADEMY.

becomes negligible with a given current density, the deposition of
hydrogen interferes with the exact determination of the electrochemical
equivalent, even before hydrogen is actually evolved in bubbles. Hence
the use of a small cathode is dangerous beyond a limit which depends
upon the accuracy required. With small current density an approximate
correction may be made for the area of the cathode.

5. Some possible causes of error in the silver voltameter are obviated
by a simple device ; and with the help of this apparatus the older
methods are compared with one another. The danger of a fallacy in
such a comparison is pointed out.

6. The electrochemical equivalent of silver appears to be about
0.0011172 gram per ampere per second.

7. The electrochemical equivalent of copper appears to be between
0.00032915 and 0.00032925 gram per ampere per second.

8. The discrepancy between the chemical atomic weights of copper
and silver and their electrochemical equivalents is explained ; hence
Faraday's law is verified with two cathions more precisely than before.
Conversely, assuming Faraday's law to hold rigidly, and the value 63. GO
to represent the true atomic weight of copper, the agreement furnishes
evidence of the accuracy of the new method.

Harvard University, Cambridge, Mass.,
October 14, 1899.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 9. — December, 1899.

CONTRIBUTIONS FROM THE CRYPTOGAMIC LABORATORY OF
HARVARD UNIVERSITY. — XLL

PRELIMINARY DIAGNOSES OF NEW SPECIES OF
LABOULBENIA CE^. — I.

By Roland Thaxter.

»

CONTRIBUTIONS FROM THE CRYPTOGAMIC LABORATORY OF
HARVARD UNIVERSITY. — XLI.

PRELIMINARY DIAGNOSES OF NEW SPECIES OF
LABOULBENIACE^. — I.

By Roland Thaxter.

Keceived November 3, 1899. Presented November 8, 1899.

Since the publication by the Academy of the writer's " Contribution
toward a Monograph of the Laboulbeniacete," * a large amount of
material of the group has been accumulated, the greater portion of which
has been derived from an examination of the entomological collections
at the Jardin des Plantes in Paris, the South Kensington Museum of
Natural History in London, the Hope Museum in Oxford, the collection
of Italian Carabidce in the Museo di Storia Naturale at Florence, and the
National Museum at Washington. It is the writer's intention to publish
as soon as practicable a Supplement to the Monograph just mentioned,
including figures of all the species ; but although a portion of the illustra-
tions are already finished, it will be impossible to complete them without
considerable delay, and it has therefore seemed advisable to publish pre-
liminary diagnoses of the new species of the genus Laboulbenia, which
will be followed shortly by a second paper, including such new forms as
belong to other genera.

Tlie writer desires in this connection to express his great obligations to
the gentlemen in charge of the collections mentioned for tlie courtesies
extended to him, due acknowledgment of which will be made in connec-
tion with the Supplement already referred to. It should be mentioned
also that a set of duplicate preparations has been prepared and will be
deposited at Paris, London, and Oxford, so that a majority of the new
forms, as well as many others, will be accessible to European students
of the group.

In the followinf^ descriptions no comparison of the forms has been
attempted in the absence of figures which might render such comparison

* Memoirs, Vol. XH. No. 3.

154 PROCEEDINGS OF THE AMERICAN ACADEMY.

iutelligible. The very large mass of material obtained naturally contains
many of the species previously reported, illustrating more fully thau was
formerly possible the geographical distribution, variation, and hosts of
many of the species, notes concerning which are likewise reserved for a
more extended paper. The numbers referred to apply to the specimens
examined, which in the Hope and British Museum collections were
labelled if found to be infested. A portion of the specimens examined
at Paris were also labelled, but unfortunately this was not done in all
cases.

Laboulbenia Acrogenis nov. sp.

Perithecium rather slender, free except at its base, pale yellowish or
straw-colored, becoming yellowish brown, tapering to the neck -like rather
slender often almost clavate apex, the lip-cells hyaline, well distinguished,
the pore obliquely external : below more or less deeply suffused with
smoky brown, especially the outer half, the posterior (outer) sub-
terminal wall-cells very often becoming deeply sutfused with smoky
brown. Receptacle rather short, becoming slightly suffused with yellow-
ish brown, normal in form, sometimes slightly inflated between cells I
and II. Outer appendage consisting of a large rounded basal cell
bearing four to eight branches arranged an tero -posteriorly in a more or
less definitely double row, their insertions forming a continuous deeply
blackened area, their cells inflated, with blackened septa, successively
once to three times branched antero-posteriorly. The inner append-
age consisting of a much smaller basal cell producing from one to two
branches similar to those of the outer, one on either side. The antheridia
solitary or from two to four, borne rather regularly on short branchlets.
Perithecium 90-175x28-35^. Antheridia 14^ long. Total length
to tip of the perithecium 190-360 /x. Appendages 85-100 ^w. Spores
about 40 |U.

On Acrogenys Mrsuta MacLeary, Brit. Mus. Nos. G68 and 528,
" Australia " and Union Reefs, Australia, Occurring on the inferior
posterior margin of the prothorax and the adjacent portions of the
thorax.

Laboulbenia adunca nov. sp.

Perithecium long and slender, straight or nearly so, the outer half
clear dark translucent brown, the inner pale olivaceous, wholly free, a
very short narrow hyaline neck formed by the basal wall-cells ; the tip
well distinguished, wholly blackish below or especially on the inner

THAXTEK. — NEW LABOULBENIACE^. 155

side, the inner lips black-margined, curved outward over the outer lips in
a characteristic hook-like fashion. Receptacle uniform dirty olivaceous,
cell I paler, the cells about it faintly punctate. Insertion-cell black,
broad, indistinguishable from the blackened basal portions of the basal
cells of the appendages. Outer appendage consisting of a large nearly
triano-ular basal cell, becoming wholly blackened except its upper margin
and surmounted by a series of usually six superposed hyaline cells
curved toward the perithecium, each cell producing externally a single
simple banch consisting of a basal portion made up of two roundish
short cells constricted at the dark septa and a distal portion six or more
times as long, tapering, hyaline or tinged with brownish ; the inner
appendage consisting of a basal cell wholly or almost wholly black, giving
rise on either side to a short series of cells, usually three or four, similar
to that of the outer appendage and similarly branched. Perithecia
225-245 X iO/x (including neck, which is about 18 ^i long). Total
length to tip of perithecium, average 450 /t; to insertion-cell 200^/,
greatest width 50 _w. Appendages 150 /<.

On Galerita unicolor Dej., Brit. Mus. No. 516, Amazon River. In-
ferior surface of prothorax.

Laboulbenia Aerogenidii nov. sp.

Form short and stout. Perithecium suffused with smoky brown, trans-
lucent, becoming distally rather abruptly subhyaline, contrasting with the
abruptly differentiated tip ; the latter black-brown, opaque or nearly so
below, the lip-cells usually symmetrical, rounded, spreading broadly
and rather abruptly, the lip-edges translucent ; sometimes asymmetrical
with the inner lip-cells only prominent and the pore external : the wall-
cells with a slight spiral twist. Receptacle hyaline below, becoming
more or less tinged with smoky brown above, especially cells III and IV
and the region immediately below the perithecium ; the suffused parts
indistinctly punctate. Appendages consisting of two basal cells ; the
inner producing two short branches on either side, which usually bear a
rather compact cluster of antheridia ; the outer giving rise to from two
to four branches placed antero-posteriorly with more or less regularity,
the outer for the most part soon broken and deeply blackened, the rest
especially the outermost more or less suffused with brown, commonly
twice branched, above the basal and subbasal cells, elongate though for
the most part broken. Total length to tip of perithecium 200-220 ^.
Appendages 200-250 ^w. Perithecium 110-120 X 35-45//, the apex
reaching a width of 35 ^u. »

156 PROCEEDINGS OF THE AMERICAN ACADEMY.

On Aerogenidion Bedeli Tsch., Mon-Pin (China ?), Paris, No. 179.
Along the margins of the elytra.

Laboulbenia Anaplogenii nov. sp.

Perithecium nearly hyaline or pale yellowisli, becoming tinged with
pale amber-brown, stout, free except at its base, the outer margin mostly
straight, the inner convex ; the tip rather small and abruptly distinguished,
blackish except the hyaline lip-margins. Receptacle concolorous with
the perithecium, rather short and stout, cell V relatively large, cell IV
divided by transverse septa into from two to several superposed cells,
usually extending upward beyond the insertion-cell, thus forming a blunt
outgrowth external to it. Basal cells of the appendages nearly equal,
each producing as a rule two similar branches antero-posteriorly, once or
twice branched in the same plane, the branchlets rather elongate and
slender, concolorous with the receptacle. Spores 50 X 4.5 /<. Total
length to tip of perithecium 300-350 ju, to insertion-cell 200-240^/,
greatest width 50-70 ^u. Appendages, longest 600 /i, average 300 ^.

On Anaplogenius circumcindus Moh., Brit. Mus. Nos. 622 and 665,
China. Also immature material of apparently the same species on an
undetermined Carabid, Paris Museum, No. 4, from Madagascar. On the
elytra.

Laboulbenia Anchonoderi nov. sp.

Pale amber-yellow, perithecium slightly darker, about two thirds free,
the inner lip-cells suffused below with blackish brown, an external
prominence involves the upper portion of the subbasal and lower portions
of the subterminal wall-cells, forming, in the specimens from A. suh-
cBneus, an abruptly defined hunch. Receptacle normal, the basal and
subbasal cells usually slender and somewhat darker, the subbasal more
distinctly marked with fine transverse striations which are less distinct
on the cells above it. Appendages concolorous, the outer simple
(always?), the inner consisting of a basal cell similar to that of the outer,
about twice as long as broad, giving rise on either side to a single branch
which may be once branched above its basal cell; all the branches some-
what flexed. Insertion-cell red-brown or purplish, more or less oblique
through the upgrowth of cell V. Spores about 45-48 /n. Perithecia
(larger) 185 X 48/^, average 125 X 48 /x. Total length to tip of peri-
thecium 276-500^ (longest 535 /m). Appendages about 200 fi.

On Anchonoderus suhceneus Reiohe, San Felix, Panama, and A.
binotatus Reiche, Guatemala City, Brit. Mus. (Biologia coll.), Nos. 706
and 707. On. the elytra.

THAXTER — NEW LABOULBENIACE^. 157

Laboulbenia angularis no v. sp.

Perithecium wholly free, symmetrically inflated, straight, evenly suf-
fused with smoky brown except the very short narrow hyaline neck-like
base ; the tip well differentiated, black, distally hyaline, the whole peri-
thecium inserted nearly at right angles to the axis of the receptacle.
Receptacle amber-yellow or amber-brown, the basal cell tinged with
smoky brown, cell II abruptly broader and long, all the cells except cell
I more or less conspicuously marked by short transverse striae, cell V
bulging on the inner side and carrying the black insertion-cell out free
from the neck of the perithecium. Outer appendage consisting of an erect
series of about six obliquely superposed cells, the lower becoming nearly
opaque, the basal one larger and darker, opaque externally and below;
each cell bearing externally a single simple branch, the branches consist-
ing of a basal portion of three short cells prominently constricted at the
blackish septa, and a terminal portion (broken in the types) : the inner
appendage consisting of a basal cell which gives rise to a series of super-
posed cells on either side, like that of the outer appendage and similarly
branched, one of the series in the types much shorter than the other (two
and four celled), the branches all erect, closely apposed, more or less suf-
fused with brownish. Perithecia 280 X 45-55 /x. Total length to tip
of perithecium 680 /x ; to insertion-cell 420 /x ; greatest width 75 fi. Ap-
pendages (broken) 55 fx.

On Galerita unicolor Dej., Brit. Mus. No. 516, Amazon River. In-
ferior surface of prothorax.

Laboulbenia anomala nov. sp.

Perithecia nearly symmetrical, free except at the base, tapering dis-
tally, the tip somewhat lighter, subtruncate with one (or more ?) of the
lip-edges modified to form a prominent large somewhat irregular blackish
brown median projection which causes the tip to appear notched on the
inner side. Receptacle dirty olive-brown, finely punctuate, the whole
curved almost to a horseshoe form : cells I and II about equal in length
and diameter, the base of the perithecium opposite cell III, insertion-cell
unmodified external subtriangular, cell V similar to it, about twice as large
and occupying a corresponding position on the inner side ; the two to-
gether with the terminal portion of cell IV, on either side of which they
lie, form the free broadly clavate terminal portion of the receptacle above
the narrower and also wholly free basal half or more of cell IV. Basal
cells of appendages not distinguishable, giving rise to a number of promi-
nences (about twelve) each forming the base of a branch, the branches

158 PROCEEDINGS OF THE AMERICAN ACADEMY.

simple or rarely branched, hyaline, of two kinds, the one with long
slightly inflated basal cells, the other closely septate, intermediate forms
occurring in which the septa are more remote ; all the septa brownish
purple. Perithecia 140 X 40 /x. Total length to tip of perithecium
about 450 |u; to tip of free end of receptacle 450 /a. Appendages 140-
160 fi.

On Orectogyrus suturalis Reg., Paris, No. 102, Zambesi River, Africa;
on 0. glaucus Klug., Brit. Mus. No. 465, Coast Castle, Egypt. On
elytra.

Laboulbenia aquatica nov. sp.

Perithecium free nearly to its base, rather short and stout becoming
dark olive-brown, the wall-cells very slightly twisted, the blackened tip
well differentiated ; nearly symmetrical, subtruncate, the lip-edges hyaline,
the lip " valves " prominent. Receptacle rather short and distally broad,
olivaceous, lighter below. Insertion-cell unmodified, external, about as
large as and nearly symmetrical with cell V, the two lying side by side
above cell IV. The appendages consisting of an outer basal cell from
which is formed a primary dark brown terminal prominence bearino-
several small branches distally, while about its base externally and often
on only one side several secondary unmodified prominences arise, each
producing usually a single terminal branch. The inner basal cell mod-
erately distinct, at first simple, later apparently divided or lobed and bear-
ing several branches. All the branches hyaline, constricted at the lower
(four to about six) dark, mostly oblique septa ; the distal portion slender,
elongate, subrigid", tapering, without constrictions or colored septa, the
basal segments bearing numerous typical flask-shaped antheridia. Spores
70 X 4.5 /x. Perithecia, average 100 X 38 /x. Total length to tip of
perithecium average 220 /x ; to insertion-cell 150 /x; greatest width 30 /x.
Appendages, longest 175 /x.

On Gyretes? sp., Paris Museum, No. 106, Venezuela. On elytra.

Laboulbenia aristata nov. sp.

Perithecium less than one half free from the receptacle, pale amber-
yellow, straight and symmetrical or nearly so, the tip well distinguished,
symmetrical, blackish, the lip-cells symmetrically rounded and protruding
slightly on either side, their distal margins straight and horizontal or
nearly so. Receptacle concolorous with the perithecium, cell I tinged
with brown, a brownish shade below the base of the perithecium, cell IV
bulging distally where it is rather faintly suffused with blackish. Outer

THAXTER. — NEW LABOULBENIACE^. 159

appendage simple, the basal cell more than twice as long as it is broad,
the two cells above it abruptly narrower and equal ; two or three of the
cells above these broader with thicker walls, the rest of the appendage
taperin<T ; the whole rigid, straight, erect or bent toward the tip of the peri-
thecium, a part of which it may overlap; inner appendage consisting of
a small basal cell bearing a very short one or two celled branch on either
side. Perithecia 95-110 X 30 /a. Total length to tip of perithecium
240-260 /a; to insertion-cell 185-200 /a; greatest width 60 //,. Append-
ages 260-330 fji.

On Carabid (? near Pericalliis), Hope Collection, No. 322, Bouro,
East Indies (A. R. Wallace). On superior margin of prothorax.

Laboulbenia Asiatica no v. sp.

Perithecium and receptacle as in typical forms of L. elongata, more or
less suffused with dirty brownish yellow, the perithecium somewhat smoky
brown above the basal wall-cells. Outer appendage consisting of a large
subtriangular hyaline basal cell, which gives rise from a deeply blackened
area of insertion to an antero-posterior series of short stout cells, them-
selves outwardly blackened and giving rise to from one to three branch-
lets also placed antero-posteriorly and themselves mostly once branched
in a similar plane, the ultimate branchlets rather long and remotely
septate, the basal septa only blackened. The inner appendage consists
of a smaller basal cell which gives rise on either side to a series of from
two to three branches arranged antero-posteriorly on a blackened inser-
tion as in the outer appendage, the branches once or twice branched, the
branchlets shorter than those of the outer appendage. Perithecium 140
X 45 |U,. Total length to tip of perithecium 400-500 /x. Appendages
(longest) 400^.

On Casnonia sp., Asia, Paris, No. 139. Elytra.

Laboulbenia Assamensis nov. sp.

' Perithecium mostly long and slender, free except at its very base,
evenly suffused with dark slightly reddish brown, paler distally, tapering
very slightly and gradually to the not abruptly differentiated tip which is
somewhat rounded and slightly bent inward, suffused with blackish brown ;
the left lateral lip-cell prolonged to form a blackish blunt-tipped promi-
nent outgrowth which is either erect or bent slightly outward or inward.
Receptacle dirty brownish yellow with fine rather indistinct transverse
striations, cell V very small, often hardly visible against the perithecium.

160 PROCEEDINGS OF THE AMERICAN ACADEMY.

Appendages concolorous with the receptacle, the outer simple, straight,
rigid, mostly somewhat divergent; the inner consisting of a basal cell
often as large as that of the outer, and giviog rise on either side
distally to a branch similar to the outer appendage which bears several
lateral antheridia near the base. Perithecium (including outgrowth 18 /x)
185-190 X 33-37 /x. Total length to tip of perithecium 375-450 /x;
to insertion-cell 185-300 fx ; greatest width about 48 /t. Appendages,
150 /x.

On Catascopus? sp., Brit. Mus. No. 663, Assam, India. Inferior
surface.

Laboulbenia barbata nov. sp.

Perithecium large, pale yellowish or brownish becoming slightly tinged
with amber-brown, often symmetrically inflated and tapering gradually
to the tip, which is brownish below becoming black on the inner side, the
lips subhyaline, turned slightly outward. Receptacle concolorous with
the perithecium, normal except that cell V is greatly enlarged and
extends upward covering nearly one half of the inner margin of the
perithecium, bulging strongly outward and carrying out the insertion-
cell which is thus made lateral in position. Insertion-cell externally con-
cave or indented. The outer appendage consisting of a small roundish
basal cell giving rise to two, rarely more, branches antero-posteriorly,
which are usually once branched near the base, the branchlets very
elongate, slender, attenuated, curved outward and downward : the inner
appendage consisting of a similar basal cell from which arise usually two
branches, one on either side, similar to those of the outer appendage, the
whole forming a hanging beard-like tuft. Perithecia 190-200 X 55-60 /t.
Total length to tip of perithecium, average 400 ^ ; to insertion-cell, aver-
age 1 90-200 jM. Appendages, longest 450-650 |U. Spores 100 X 6 ^.

On 3I(irio Georgii Pal., Brit. Mus. No. 690 (Biologia Collection),
El Zambador, Mexico ; on M. simplex, Dej., Brit. Mus. No. 581, Cayenne;
on M. mom'licorms, Latr., Hope Collection, No. 289, " North America."
On the elytra.

Laboulbenia bicornis nov. sp.

Perithecium wholly free, dark brown tinged with olive becoming
opaque, lighter at the base ; very elongate, nearly straight, slightly and
gradually inflated from the neck-like base to about the middle, thence
tapering very slightly to the tip, which is distinctly though not abruptly
differentiated : the two inner lip-cells symmetrical each terminating in a
small rounded prominence which bears a second smaller rounded terminal

THAXTER. NEW LABOULBENIACE^. 161

promiuence ; from the base of the lower prominence a long slender
brown or olive-brown horn-like process grows downward, inward, and
upward, the two symmetrical and similar and resembling the horns of an
ox ; though formed from the inner lip-cells, a slight twist in the wall-
cells usually makes them appear lateral or even external. The two
outer lip-cells grow beyond the inner and are closely united forming a
large bluntly pointed nose-like projection, its inner margin slightly con-
vex while externally it is nearly straight and slightly oblique. Recep-
tacle short and stout, evenly dark olive-brown ; cell I short, slender, cell
II abruptly larger, broad and short. The basal cells of the appendages
opaque and indistinguishable, giving rise as in allied aquatic species to
cladophorous prominences, the branches once or twice branched, hyaline,
the lower three or four septa dark, the cells between them slightly
inflated, the distal portion elongate, cylindrical, thin-walled, blunt-tipped,
without dark septa, the whole forming a dense tuft. Spores 125-140
X 7-8 1.1. Perithecia 340-750 X 60-75 ^. Total length to tip of perithe-
cium, longest 1150 ^w; to insertion-cell 340-400 ^u ; greatest width 120-
130//.

On Dineutes aereus Klug., Brit. Mus. No. 461, Hadramaut, Arabia;
on Dineutes sp., Brit. Mus. No. 463, Ambaca, Angola, W. Africa. On
abdomen, elytra, thorax, and head.

Laboulbenia bidentata nov. sp.

Wholly suffused with pale dirty brown. Perithecium rather darker,
free except at its base, somewhat inflated below, tapering symmetrically
from the extremity of the basal wall-cells to the rather narrow apex ; the
two inner lip-cells suffused with blackish brown and projecting beyond
the others to form two blunt prominences, their tips hyaline, contrasting.
Receptacle rather short and stout, uniform pale dirty brown, incon-
spicuously flecked with minute somewhat darker points, the lower part
of the small basal cell mostly hyaline. Outer appendage consisting of a
basal cell several times as large as that of the inner, bearing a single
branch, the basal cell of which is somewhat rounded and gives rise dis-
tally to an outer and an inner branchlet, each commonly once branched,
the outer deeply blackened at the base (usually broken) : the inner
appendage consisting of a basal cell which bears a branch on either side
which may be twice branched, bearing solitary antheridia laterally.
Spores 50 X 4^. Perithecia 115-130 X 45 //. Total length to tip of
perithecium 220-270 [i ; to insertion-cell 135-150 //. Greatest breadth
70^.

VOL. XXXV. — 11

162 PEOCEEDINGS OF THE AMERICAN ACADEMY.

On Homothis sp., St. George's Sound, Australia, Hope Coll., No. 309.
On elytra.

Laboulbenia Brachionychi nov. sp.

Perithecium wholly free, rather deeply suflfiised with smoky or reddish
brown, subclavate or almost symmetrically inflated, tapering from the
middle to the nearly symmetrical, or externally slightly oblique blunt
tip, the basal wall-cells contracted to form a long slender neck-like sub-
hyaline stalk. Receptacle abnormally developed, cells I and II nearly
equal, cells III and V forming a stalk which is deeply suffused with
blackish brown especially externally, and rather coarsely punctate in
the darker areas, of equal diameter throughout and quite free, except at
its very base, from the rest of the receptacle : cell IV larger than cell
III, cell V small and separated by an oblique septum ; cell VII and the
basal cells of the perithecium small. Outer appendage consisting of a
basal cell deeply blackened externally, and bearing a long slender simple
branch, distally hyaline, more or less suffused with smoky brown towards
the base : the inner appendage consisting of a basal cell about half as
long as that of the outer, producing a single branch on either side which
may be once or twice branched, the antheridia borne usually in twos or
threes on short branchlets near the base. Spore 110 X 7jM. Perithecia
exclusive of stalk, 270-340 X 65 ju ; including stalk 340-430 ^u. Total
length to tip of perithecium 650-800^; to insertion-cell 400-500^.
Stalk-like portion of receptacle 138 X 38 /<. Appendages 400-470 ju.

On Brncldonychus sp., Nos. 99 and 822, Cochin China ; on Epis-
cosoma laticollis, No. 85, Cochin China; on Episcosoma sp., No. 86,
Java. All in Paris Coll. Usually on inferior surface of thorax.

Laboulbenia Cafii nov. sp.

Perithecium almost wholly free, pale amber-colored or straw-yellow,
transparent, stout, the tip blunt, with blackish basal suffusions, well dis-
tinguished, especially on the inner side. Receptacle short and stout, pale
amber-colored, normal in form. Insertion-cell broad, often not deeply
blackened. The appendages consisting of an outer and an inner series
of from four to six superposed cells which, through a twist of the insertion-
cell, become lateral instead of antero-posterior in position ; each cell of
these series produces externally a single simple short branch usually
three-celled, the two lower cells short and inflated, the upper longer
tapering to a blunt apex. Perithecia 140-165 X 60-70 /x. Total length
to tip of perithecium 310-350 /x, to insertion-cell about 170 /x. Append-
ages, longest 85 /x.

THAXTER. — NEW LABOULBENIACE^. 163

On Cafius seminitens Horn, and C. canescens Mann, U. S. National
Museum, Los Angeles, California. On C. sericeus Holme, Brit. Mus.
No. 437, Great Britain. On Ccifius sp., Brit. Mus. No. 425, " Europe,"
No. 395, Hong Kong ; C. Msulcatus Sol., Chili. On elytra and legs.

Laboulbenia celestialis nov. sp.

Perithecia almost wholly free, rather deeply suffused with dark red-
dish brown except the almost hyaline basal wall-cells ; tapering slightly
to the well distinguished rather large tip, which is turned slightly out-
ward, the lip-cells blackened except around the pore. Receptacle
uniformly pale dirty brown, rather short and stout, somewhat abruptly
expanded below the perithecium ; cell IV broader than insertion-cell.
Outer appendage consisting of a somewhat rounded basal cell deep brown
externally and bearing usually two brauches placed antero-posteriorly,
the basal cell of the outer nearly round, deep brown externally and
bearing two branches placed antero-posteriorly which are very long and
slender, remotely septate and more or less suffused with dirty brown :
the inner appendage consisting of a smaller basal cell producing a branch
on either side, the basal cells of which are short with suffused septa and
bear solitary antheridia or short sterile branches which are blunt and
shorter than the perithecium. Perithecia 110 X 35 /x. Total length to
tip of perithecium 280 /x ; to insertion-cell 156 /x. Greatest width 55-
60 IX. Appendages (longest) 400 /x.

On Drypta lineola Dej., Brit. Mus. No. 507, China. Elytra.

Laboulbenia ceratophora nov. sp.

Perithecium wholly free, borne on a short nearly hyaline stalk, tinged
not deeply with brown, for the most part nearly straight, rather slender,
hardly inflated, the outer margin usually slightly concave; tapering
slightly and rather abruptly to the peculiarly modified ti]?; the tip black
except externally immediately about the pore, a large blunt tooth-like
prominence i^rojecting inward the upper surface of which is nearly
horizontal while it is continued upward and outward into a longer
more slender horn-like bluntly tipped outgrowth, the lower or external
margin of which may be partly hyaline. Receptacle yellowish tinged
with brown subdistally, rather narrow dis tally, cell IV small and
hardly longer thqn cell V. Insertion-cell opposite the distal extremity
of the short perithecial stalk. The basal cell of the outer appendage
longer and narrower than that of the inner, becoming concolorous with

164 PROCEEDINGS OF THE AMERICAN ACADEMY.

the insertion-cell bearing a single branch of usually three cells more or
less deeply tinged with brown, each of which may give rise distally on
the inner side to a short simple branchlet ; the inner appendage consists
of a basal cell shorter and broader, which usually bears distally on either
side a branch, the basal cell of which usually bears distally two erect
simple branchlets, one of them sometimes replaced by a single large
long-necked brown antheridium. Perithecia, including the stalk {17 fi),
150-155 X 30 fi. Total length to tip of perithecium 290-310 fx ; to inser-
tion-cell 150^; greatest width 35 /x. Appendages, longer 200 /x.

On Serrimargo guttiger Schaum, Hope Collection, No. 300, Sumatra ;
on Miscelus Javanus King., Hope Collection, No. 303, Java; on Miscelus
sp., New Guinea, Paris Museum, No. 114. On elytra and inferior sur-
face of the prothorax.

Laboulbenia Ceylonensis nov. sp.

Perithecium wholly free, suffused with smoky brown, relatively short
and small, straight, slightly inflated, somewhat abruptly contracted distally
to form the unusually large broad tip, the distal half of which is hyaline,
distinguished from the opaque black lower half by a clean cut line of
demarcation : the basal cells forming a short stout hyaline neck narrower
than the body of the perithecium. Receptacle elongate, of nearly the
same diameter throughout, the basal cell nearly hyaline, the cells above
it more or less deeply suffused with smoky brown, cell V and the upjier
portion of cells IV and VII hyaline, the distal suffused portion obscurely
punctate. Outer appendage consisting of a basal cell deeply blackened
externally (the blackened area continuous with the black insertion-cell
and involving also the external walls of the two cells immediately above
it) producing from one to three branches arranged antero-posteriorly,
which may be once or twice branched in a similar fashion, the branchlets
long, slender, drooping, hyaline, some of the lower cells suffused with red-
dish brown : the inner appendage consisting of a basal cell about half as
large as that of the outer, producing in the types a single branch which
may be once branched as in the outer. Perithecium 105 X 32 ^w. Total
length to tip of perithecium 445 ^ci ; to insertion-cell 320 ^m. Greatest
breadth 42^0. Appendages 340 /<.

On Hexagonia? , Ceylon, Hope Coll. No. 288. On elytra.

Laboulbenia Chiriquensis nov. sp.
Perithecium yellowish brown, straight, the inner margin convex, the
outer slightly concave, tapering distally to the broad short blackened tip,

THAXTEB, — NEW LABOULBENIACEiE. 165

which IS bent abruptly inward almost at right angles. Receptacle yellow-
ish brown, deeper in the region of cell III, the basal cell slightly curved,
its upper half rather abruptly expanded ; a more or less distinct bulge
below the perithecium. Outer appendage usually simple, stout, the basal
cell of the inner usually much smaller bearing a branch on either side
usually once branched above the basal cell ; all the branches stout and
constricted at the lower septa. Spores 60 X 4: jn. Perithecium 100-125 X
37 jM. Total length to tip of perithecium 200-270 /<; to insertion-cell
135-1 GO ,(/. Greatest width 55 ^.

On Calleida scintillans Bates, Brit. Mus. (Biologia Coll.), No. 735,
Vale de Chiriqui, Panama. On margin of elytron.

Laboulbenia Clivinalis nov. sp.

Usually rather slender. Perithecium three fourths or more free, more
or less deeply tinged with blackish olive-brown, distally curved slightly
outward, the tip black with hyaline lip-edges. Receptacle wholly hya-
line, or cells III and IV becoming more or less suffused with blackish
brown, the suffused areas punctate : bulging distally below the perithe-
cium. Insertiou-cell well differentiated, black. Appendages consisting
of an inner and an outer basal cell, which may remain simple or become
longitudinally divided, sometimes also transversely or even obliquely :
the outer basal cell hyaline, often several times as long as broad, its dis-
tal septum blackened ; when simple, bearing a single branch, if divided,
several ; the basal cell of the outer branch usually rather short and some-
what suffused, commonly bearing two branchlets. The basal cell of the
inner appendage usually smaller than that of the outer, sometimes equal,
and like it simple or divided ; when simple, bearing a branch on either
side, or more if it is divided. All the branches of both appendages hya-
line or nearly so, mostly once branched above their basal cells. Spores
about 55 X 4.5 /i. Perithecia 120-150 X 35-50 jM. Total length to tip
of perithecium 275-400^; to insertion-cell 200-340^. Appendages,
longest 300-400^.

On CUvina collaris Herbst, Hope Coll. No. 348, and Brit. Mus, No.
456, both from England. On GUvina fossor Linn., Hope Coll. Nos. 353,
275 and 475, England; No. 295, "Europe"; Florence Mus., Italy.
Usually on elytra and superior prothorax, but occurring elsewhere.

Laboulbenia coarctata nov. sp.

Perithecium short and stout, dark brown and opaque, its upper half
free, its contour evenly rounded, the small papillate translucent tip turned

166 PROCEEDINGS OP THE AMERICAN ACADEMY.

slightly inward and abruptly differentiated, the lips well defined. Recejv
tacle dirty brown finely and obscurely punctate, of the typical form, lighter
below, short, stout, cell I much narrowed below, cell II very broad, slightly
inflated. Basal cells of appendages well developed bearing about eight
erect short stout simple branches with broad basal constricted blackened
septa, their rounded tips slightly exceeding the tip of the perithecium.
Spores 70x7;u. Perithecia 140-170 X 62^11. Total length to tip of
perithecium 325-400 ^u; to insertion-cell 250-300^; greatest width
100 /i.

On Orectochilus'^, Hope ColL Nos. 233 and 234; Brit. Mus. No. 4G6,
Bengal, India. Along median depression of elytra.

Laboulbenia Colpodis nov. sp.

Perithecium wholly free, very large, uniform clear translucent brown,
slightly inflated or the margins straight, the basal wall-cells forming a
well marked hyaline neck as broad as the base of the ascigerous portion
from which it is not abruptly distinguished ; the tip rather narrow and
well distinguished, darker, the distinct lip-edges hyaline or translucent.
Receptacle very small, cell I hyaline or yellowish, the rest more or less
deeply suffused with blackish brown becoming nearly opaque, especially
cells III and IV, which lie side by side and are about equal in size, cell
III forming a rounded jirominence just below the outer edge of the inser-
tion-cell, cell V triangular, about as large as cells III and V combined.
Insertion-cell not at first deeply blackened, but becoming indistinguishable
in the general blackish brown suffusion of the adjacent cells. Outer ap-
pendage consisting of a nearly erect series of obliquely superposed small
cells (sometimes as many as thirteen) at first hyaline, the basal cell and
sometimes several of the cells above it becoming deeply suffused, which
produce externally a single simple branch curved upward, its basal por-
tion consisting of two cells, sometimes three, longer than broad and
more or less deeply suffused with brown, constricted at the dark
septa ; its distal portion elongate, reaching nearly to the tip of the peri-
thecium, strongly tapering hyaline except at its base where it is in-
volved by the brown suffusion of the basal part ; the inner appendage
consisting of a basal cell surmounted on either side by a series of cells
similar to that of the outer appendage, but fewer in number, similarly
branched except that the two or three lowest cells of each series bear
single large stout straight brown antheridia, the basal cells of the series,
as in the outer appendage, becoming suffused with blackish brown. Peri-
thecia 190-220x40-45^. Total length to tip of perithecium 300-

1

THAXTER. — NEW LABOULBENIACE^. 167

375 fi ; to insertiou-cell 75-80 fi ; greatest width 55 /.i. Appendages,
longest 220 ^.

On Colpodes Chiriqiiimis Bates, Brit. Mus. No. 735 (Biologia Coll.),
Vale de Cbiriqui, Panama. Ou elytra.

Laboulbenia constricta nov. sp.

Perithecium more than one half free, short, stout, inflated, somewhat
olive-brown, the tip not differentiated, one of the lip-edges becoming modi-
fied to form a flattish or roundish brown prominence which lies usually
in a median position between two rather ill defined prominences on either
side of it. Receptacle usually with a double curvature, its basal cell very
large, somewhat inflated distally, the base and posterior margin paler, cell
II shorter, sufi^used below, strongly constricted distally, the constricted
portion paler or hyaline : the distal portion of the receptacle short,
deeply suffiised, bulging strongly anteriorly below the insertion of the
appendages. Insertion-cell not blackened, the appendages arising much
as in L. Orectochili and L. strangulata, the branches hyaline, the basal
portion closely septate, simple, without persistent blackened basal por-
tions. Spores 70 X 6^. Perithecia, average 130 X GO^u. Total length
to tip of perithecium, average 400 ^ ; to insertion-cell 340 ^. Appendages
70 (i (or more ?).

On Orectogyrus glauciis King., Brit. Mus. No. 465, Coast Castle,
Egypt. On elytra.

Laboulbenia Copteae nov. sp.

Very slender. Perithecium free except the base, nearly straight,
curved outward slightly at the tip, which is blackish with hyaline lip-
edges ; dull translucent olive-brown, concolorous with the receptacle,' or
with the basal wall-cells nearly hyaline. Receptacle slender, bulging
slightly at the base of the perithecium, cell II narrower distally than cell
I, the base of which is hyaline. Outer appendage consisting of a basal
cell several times as long as broad, hyaline, usually curved strongly
outward, constricted distally where the septum is suffused with dark
brown below the single straight branch, which is more slender, elongate,
tapering, hyaline, simple or once branched above its basal cell. Inner
appendage consisting of a small basal cell bearing a short branch on
either side, simple or once branched, usually not longer than the perithe-
cium. Perithecium 120-140 X 35 ^<. Total length to tip of perithecium
340-470 ;u; to insertion-cell 220-300 ^m. Appendages (longest) 400 /i.

168 PROCEEDINGS OF THE AMERICAN ACADEMY.

On Copfea armata Lnp., Brit. Mus. No. 595, Santarem, Amazon
River, Brazil. On elytra and superior prothorax.

Laboulbenia corethropsis nov. sp.

Perithecium relatively large, translucent brown or yellov(?ish brown,
generally larger distally through a subterminal external bulge, the tip
moderately well differentiated, often bent rather abruptly inward, wholly
blackish brown or hyaline in the median line, the inner lip-cells smaller
and usually more prominent than the outer, sometimes overlapping them,
the whole perithecium free from the receptacle. Receptacle short and
rather stout, cells II-IV becoming externally blackish brown, the rest
concolorous with the perithecium ; cells III and IV rather prominent.
Outer appendage wholly opaque, its successive cells indistinguishable as
are the basal and insertion-cells, consisting of a blackened axis erect or
bent outward, from the inner side of which arise several more or less
blackened branches, the basal portion of the appendage only, as a rule,
remaining. The inner appendage, consisting of a basal cell which is in-
distinguishable from the insertion-cell, being quite opaque, producing
a branch on either side, the axis of the branches erect, becoming
blackened and opaque, except the inner margins of the distal cells, bear-
ing externally from six to ten or even more branchlets which are more
or less deeply suffused. The general habit suggesting a species of
Gorethromyces. Perithecium 130-150 X 30-33 /x. Total length to tip
of perithecium 190-240 yu,; to insertion-cell 90-130 ;u.. Greatest width
40-50 jU. Appendages, longest (broken) 185 fj..

On Miscelus Javanus Klug., Hope Collection, No, 304, Java; on Mi-
scehts sp., Paris Museum, No. 114, New Guinea. On inferior surface of
abdomen, and on elytra.

Laboulbenia corrugata nov. sp.

Perithecium rather small, irregular, free from the receptacle except at
the base, blackish brown, darker and almost opaque below, a rounded
distal elevation extending completely round the perithecium, broader
externally, forming a nearly opaque broad collar above which the tip is
very abruptly distinguished ; the latter subhyaline basally, a dark median
external and internal suffusion, the lips translucent, faintly brownish,
the two outer rounded and curved inward between the two inner, which
form two free slightly unequal divergent blunt-pointed projections. Re-
ceptacle rather long and slender, pale brownish yellow with deeper

THAXTER. — NEW LABOULBENIACE.E. 169

brown suffusions about the distal region of cell I and the basal region of
cell II ; the basal cells of the perithecium opaque and cell IV and V
more deeply suffused ; a series of prominent blackish brown tuberculate
rido^es extend more than half way across the receptacle from the anterior
margin, beginning below the insertion cell and present as far as the base
of cell II; cell IV bulging distally outside the insertion-cell. Outer
appendage simple, erect, tapering, yellowish, perhaps once branched above
the basal cell ; inner appendage consisting of a smaller basal cell with a
similar and shorter branch on either side. Perithecia 120 X 37-40 /x.
Total length to tip of perithecium 340 [x ; to insertion-cell 200 fx. ; greatest
width 60 /x. Appendages \85 fx.

On Serrimargo guUlger Schaum., Hope Collection, No. 300, Sarawak,
Borneo. On base of elytra.

Laboulbenia Cubensis nov. sp.

Perithecium short and stout, free except ^at the base, slightly curved
toward the appendages, blackish olive, the lip-cells prominent but flat-
tened backward and outward. Receptacle elongate, cells I and II dirty
olive, cell III hyaline, long, contrasting with cells IV and V, which,
together with the basal cells of the perithecium, are deeply suffused with
blackish olive ; cell VI as long as cell III and lying beside it, becoming
tinged with dirty olive ; basal cells of appendages dark olive, indistin-
guishable, producing concolorous prominences which give rise to a
number of hyaline branches, the basal cells of which are large, swollen
distally, and bear numerous terminal and subterminal branchlets exter-
nally; the branchlets once or twice branched, the septa olive. Perithe-
cia 155-190 X 75-85 /x. Total length to tip of perithecium 480-800 [x.
Appendages, longest 140 /x.

On Dineutes longimanus Oliv., Paris Museum, No. 101, Cuba. On tip
of abdomen.

Laboulbenia dactylophora nov. sp.

Perithecium relatively small, its upper third only free from the recep-
tacle, translucent smoky brown, the tip prominent, not abruptly differen-
tiated, its upper half hyaline, black-tipped, symmetrical or irregularly
sulcate, the lower half blackened. Receptacle very elongate, concolorous
with the perithecium or paler, finely punctate, tapering below, its curved
base hyaline, the basal cell expanding distally to the very broad upper
septum, which forms the middle of a somewhat one-sided enlargement
involving the adjacent extremities of cells I and II, cell II more dis-

170 PROCEEDINGS OF THE AMERICAN ACADEMY.

tinctly punctate, very long, tapering very slightly nearly to its distal
extremity, cell V producing two outgrowths, one on either side, erect,
usually similar, brown, distally somewhat enlarged, the rounded tips
often bent sliglitly outward, extending to or beyond the tip of the peri-
thecium. Appendages consisting of two (or more?) basal cells from
which several protrusions arise bearing groups of branches which "are
several times branched, forming a dense tuft not as long as the outgrowths
from cell V, the lower septa brown or blackish, between the short cells.
Perithecia 170-190 X 48-50 /x,. Total length to tip of perithecium 750-
880 /i; to insertion -cell G80-810/i ; cell II about 425 /x long. Out-
growths from cell V, 58 /i. Appendages 50^.

On Orectogyrus specular is Aube, Paris Museum, No. 100, Gold
Coast, W. Africa. Margin of elytra.

Laboulbenia Darwinii nov. sp.

Perithecium hyaline becoming pale straw or amber-yellow, sometimes
with a shade of brown, relatively small, its upper third or less free from
the receptacle, the tip black, abruptly distinguished, the broad lip-edges
translucent purplish brown. Receptacle relatively stout, indistinctly
punctate with short lines or dots of darker yellowish color, cells II, III,
and IV often unusually broad. Insertion-cell well developed, deep
purplish brown or black. Outer ajipendage consisting of a short angular
brownish basal cell, which bears an outer and an inner branch distally,
the outer simple or once branched above its basal cell, the inner mostly
simple ; the outer branchlet mostly shorter, divergent, often deeply
suffused with reddish brown, the rest less deeply colored, rigid, erect.
The inner appendage consisting of a smaller basal cell which may pro-
duce'a single branch, or two placed laterally or antero-posteriorly, short,
simple with lateral antheridia or once branched. Perithecia 100 X
30-35/1. Total length to tip of perithecium 150-275 /x; to insertion-cell
135-250 /i. Appendages, longest 200-250 fi.

On Oezena parallela W., Brit. Mus. No. 572, Rio de Janeiro (legit C.
Darwin), on Pachyteles spp., Paris Mus. No. 137, South America, Hope
Collection, Nos. 284 and 285, Brazil. Occurring usually at the base of
the posterior legs.

Laboulbenia denticulata nov. sp.

Perithecium free, olivaceous with blackish shades below the paler tip,
rather narrow, straight ; the tip broad, one of the inner lip-cells forming
a short brown conical terminal prominence which is straight or bent

THAXTER. — NEW LABOULBENIACEiE. 171

toward the pore contrasting with the nearly colorless lip-edges below it.
Receptacle pale brown or dirty olivaceous, elongate, normal. The in-
sertion-cell nearly horizontal external to cell V, unmodified. The outer
basal cell of the appendages giving rise to a single subconical brown
prominence bearing branches terminally and externally and protruding
beyond the inner basal cell, which is indistinguishable from the very
nusnerous bi-anches arising from it in all directions ; all the branches
hyaline, their basal cells bearing distally several branchlets which may
again be branched, the four to eight lower septa dark. Perithecia,
average 175x44/^; the spine-like apex 10-12 /:i. Total length to tip
of perithecium 400-575 fi ; to insertion-cell 275-400 (j, ; greatest width
55-70 /i. Appendages about 70 jj..

On Dineutes ?, Brit. Mus. No. 482, Adelaide River, Australia.

Laboulbenia Dineutis nov. sp.

Perithecium free except at the base, dark brown, the inner margin
straight, the outer curved from the base to the tip, which is not well
differentiated, the lip-cells inconspicuously modified to form an inner
tooth-like brown prominence curved inward, and a median small brown
rounded elevation, which is covered by an outer hyaline externally
brown broad irregular elevation. Receptacle elongate or rather short,
dark brown inconspicuously punctate. Appendages much as in Z. hete-
rocheila, the branchlets closely sefitate with dark septa near the base
and (in unbroken specimens) with long hyaline aseptate or remotely
septate terminations three to four times as long as the basal part.
Measurements very variable. Perithecium, (Ceylon) 275X72/^,
(India) 140 X 40 /x. Total length to tip of perithecium, (Ceylon) 1 mm.,
(India) 400 /x, (Madagascar) 350 /i. Appendages, (Ceylon) 200 fx.
Si^ores, (Ceylon) 75 X 6 /i.

On Dineutes suhspinosns Klug., Paris, Nos. 33 and 34, Madagascar and
Isle de France; Hope Collection, No. 236, no locality. On Dineutes
spp.,- Hope Collection, Nos. 230, 231, 232, and 235, Bengal, " Pondich
Guera," Asia, Ceylon, Mauritius. On Dineutes, Brit. IVIus. No. 483,
Nilgiri Hills, India. On margin of elytra and tip of abdomen.

Laboulbenia Dercyli nov. sp.
Perithecia free except at the base, short, stout, becoming tinged with
brown, straight or the usually very broad tip turned slightly outward, the
latter black, contrasting with the hyaline lip-edges. Receptacle elongate.

172 PROCEEDINGS OF THE AMERICAN ACADEMY. i

faintly and uniformly tinged with smoky brown, normal except that cell I
V is pushed up even beyond the hyaliue unmodified insertion-cell which !
becomes thus external to it. Appendages recalling those of L. variabilis, j
typically consisting of an inner and an outer basal cell, the outer bearing ;
a single oblique or nearly vertical more or less irregular row of branches j
arising antero-posteriorly ; the inner basal cell bearing a similar row on |
either side, all the branches hyaline or yellowish brown, more or less I
copiously branched ; the lower cells somewhat inflated, the septa black- |
ened, often oblique, the basal cells of the branches bearing distally and ]
externally from one to three obliquely superposed branchlets with
blackened septa, which may branch again ; the ultimate branchlets taper-
ing slightly, the septa transverse and hyaliue, usually cohering in an erect I
mass. A third group of branches similar to the rest sometimes arises j
between these and the perithecium, apparently from the distal portion of !
cell V. Perithecia 140-200 X 55-60 //. Total length to tip of perithe- I
cium 475-875 ft; to insertion-cell 400-700 ^. Appendages, longest about ]
175 ^.

On Dercylus tenehriosus Laf. (= Eiirysoma tenebrioides?), Hope Coll. -,

No. 328, Para; Brit. Mus. No. 586, " S. America." Margin of right i

elytron. |

Laboulbenia distincta nov. sp. I

Perithecium short and broad, wholly united to the receptacle except 'i
the tip, clear dark brown, darker distally, the tip large, blackish, somewhat |
compressed, the lips translucent smoky brown, not prominently distin- i
guished. Receptacle short and stout, the distal jwrtion larger than the I
basal ; cells I, II, and VI transparent yellowish, the rest smoky brown, !
marked by closely set fine transverse lines ; cells III and IV relatively j
very large, subequal. Insertion-cell two thirds as broad as cells IV- V. j
Outer appendage consisting of three superposed flat cells, hyaline becom- j
ing dark brown, the two lower larger and nearly equal, the middle cell
producing a single branch curving upward from its inner side, the third ,
cell producing a similar branch from its inner side and a terminal branch i
which is deep brown curved outward and upward, slender, simple ; the '
inner branches of the three cells all at first hyaline, mostly once branched, !
later developing a dark brown contrasting suffusion above their basal i
cells : the inner appendage consisting of a small basal cell producing a |
branch on either side the basal cell of which is larger than that of the 1
appendage, and bears two branchlets basally suffused with brown like j
those of the outer appendage. Perithecia 130 X 50 fi. Total length to i

THAXTER. — NEW LABOULBENIACE^. 173

tip of perithecium 275 (j, ; to msertion-cell 250 fi ; greatest width 95 fx.
Appendages, lougest 235 fx.

Oil Pericallus cceruleovirens, Tat., Brit. Mus. No. 570, Singapore. On
margin of elytra.

Laboulbenia drepanalis nov. sp.

Perithecium smoky olive, the inner half or less usually much paler,
the upper three fourths free, falcate ; the inner margin concave, the tip
undifferentiated, the lip-edges forming a small hyaline rounded abruptly
distinguished papilla. Receptacle rather short, concolorous with peri-
thecium ; cell 1 paler or hyaline with a basal blackish sufEusion ; cell V
larc^e, growing upward above the oblique insertion cell which is thus
pushed out free from the perithecium together with the basal cells of the
appendages. Insertion-cell small, unmodified. Basal cells of the appen-
dages closely united, finally iudistini^uishable from one another, forming
a prominent rounded outgrowth which gives rise to about six or eight
branches, their dark basal septa of variable diameter only remaining, as a
rule ; the basal cells of the branches are distally inflated, and bear several
branchlets externally and terminally, the basal cells of the branchlets
mostly similar to those of the primary branches and similarly branched,
the ultimate branchlets closely septate, the septa dark. Perithecia,
Mexican specimens 100 X 'iO /.i, Amazon 140 X 35^. Total length to
tip of perithecium, Mexican 210^, Amazon 275^; to insertion-cell,
Mexican 130 ^u, Amazon lAO fz.

On Gyretes acutangidus, Sharp, Brit. Mus. No. 771 (Biologia Coll.),
Bugaba, Panama; on Gyretes sp., Brit. Mus. No. 477, Amazon. On
mid-elytron.

Laboulbenia EgaB nov. sp.

Perithecium free or nearly so, slender, usually somewhat curved in-
ward, becoming evenly suffused with pale olive-brown, the broad tip not
differentiated from the body of the perithecium, the lip-cells more or less
suffused with darker brown. Receptacle pale yellowish, often elongate,
the basal cell short ; cell II several times as long, becoming amber-brovv'n
with transverse striations ; cells III and VI about equal, elongate. In-
sertion-cell not deeply blackened. Outer appendage consisting of a
usually somewhat inflated basal cell with thick outer wall, bearing one or
two branches placed close together antero-posteriorly, the branches simple
or once branched, subhyaliue ; the inner appendage consisting of a basal
cell half as large as that of the outer, bearing usually a single simple or

174 PROCEEDINGS OP THE AMERICAN ACADEMY.

once divided branch on either side : the branches of both appendages
subhyaline, mostly thin-walled, erect in a compact small tuft, tapering
slightly, hardly exceeding the tip of the perithecium. Perithecium 190-
225 X 34 fi. Total length to tip of perithecium 540-680 X Gofi; to
insertion-cell 375-470 |U. Appendages 175-200 ^u.

On Uga sp., Paris Mus. No. 151, Acapulco, Mexico. On Ega Sallei
Chev., Brit. Mus. No. 705, Biologia Coll., Paso Antonio and Champerico,
Guatemala. On elytra.

Laboulbenia equatorialis nov. sp.

Perithecium free, hyaline becoming tinged with brown, slender and
elongate ; the basal wall-cells forming a well marked though not clearly
differentiated hyaline stalk ; a median and subtermiual external promi-
nence; the rather broad tip more or less deeply tinged with blackish
brown, rather abruptly differentiated and bent outward, the lip-edges
hyaline, the inner lip-cells prominent, more deeply suffused. Eeceptacle
very long and slender, cell I short, cell II greatly elongated, cells III
and VI about equal. Insertion-cell broad, deeply blackened. Outer
appendage consisting of a somewhat rounded basal cell, which is hyaline
externally, distally suffused with blackish brown, and bears two branches
antero-posteriorly ; the outer of which consists of a nearly isodiaraetric
basal cell, opaque except its upper inner angle, from which arise two
branches, the inner simple, becoming red-brown, its two lower cells in-
flated inward, the outer consisting of a small basal cell, opaque, except
its inner upper hyaline angle and bearing two branches, an inner dark
red-brown and slender, an outer curved outward and upward, more slen-
der, deep red-brown, opaque toward the base, usually broken : the inner
branch from the basal cell of the outer appendage consists of a basal cell
like that of the outer branch, which bears distally two bi-anches red-brown,
about equal, the lower two cells inflated inward. The inner appendage
consists of a slightly smaller basal cell bearing a branch on either side, the
basal cell of each branch shorter, and giving rise typically to two
branchlets from the basal cells of which the large, long, slender deep
red-brown antheridia are produced in pairs. Perithecia 400-760 (x.
Total length to tip of perithecium, average 550 ^ ; longest over 1 mm. ;
average breadth 50^. Appendages, longest ^Ibfi.

On Casnonia sp., Brit. Mus. No. 502, Amazon River. On upper
surface of prothorax, at base of elytra and on legs.

THAXTER. — NEW LABOULBENIACE^. 175

Laboulbenia erecta nov. sp.

Perithecium pale yellowish brown, slightly and evenly rounded, united
to the receptacle as far as its subterminal cells, the nearly symmetrical
tip rather broad, often flattened distally, becoming deeply suffused with
blackish, except the narrow translucent margins of the lips. Receptacle
pale yellowish, cell I short and stout, distally suffused with blackish
brown ; cell VI very small, cell V free from the perithecium, sometimes
lateral as in L. paupercxda. Insertion-cell thick and black, sometimes
carried forward opposite the apex of the perithecium through the elonga-
tion of cells IV and V. Outer appendage consisting of a large basal cell
several times longer than broad, which may bear terminally one or two
branches, the outer sometimes once branched, the branchlets very long,
slender, tapering, hyaline or pale yellowish, more or less flexuous : the
inner appendage sometimes laterally placed, usually single, simple, con-
sisting of two or three short cells with one or two lateral antheridia and
sometimes producing longer branchlets, the basal cell much smaller than
that of the outer appendage. Perithecia 110-120 X 35-40 /z. Total
length to tip of perithecium 200-275 /x ; to insertion-cell 200-250 /x.
Appendages, longest 675 ^.

On " Colpodes agil'is Chd.," Jalapa, Mexico, Brit. Mus. (Biologia
Coll.), No. 696; on Q. evanescens Bates, U. S. National Museum, Bio-
logia Coll., Mexico. Elytra.

Laboulbenia falcata nov. sp.

Perithecium free or nearly so, mostly very large, pale yellowish, the
inner half or more suffused with smoky brown, darker toward the margin,
the base more or less strongly curved so that the perithecium is directed
outward nearly at right angles to the axis of the receptacle or even re-
curved, basally inflated along the inner margin, tapering gradually from
about the lower third to the apex; the tip not differentiated from the body
of the perithecium, the lip-cells with darker longitudinal brown shades.
Receptacle short, straight or nearly so, the basal cell broad, sometimes
slightly inflated, a slight constriction often present between cells II and
III, all the remaining cells unusually small in proportion. Outer append-
age consisting of a small rounded basal cell bearing a single branch sepa-
rated from it by a blackened septum and consisting of a hyaline externally
blackened basal cell bearing two branchlets ; an outer (usually broken)
externally blackened and bearing several vertical branchlets ; an inner
usually simple, hyaline or yellowish. Inner appendage consisting of a

176 PROCEEDINGS OF THE AMERICAN ACADEMY.

basal cell similar to that of the outer or slightly smaller, producing a
branch on either side, each usually ouce branched, all the branches pale
yellowish with occasional brown suffusions, the longest not greatly ex-
ceeding the tip of the perithecium. Spores 35 X 3 /x. Perithecium 150-
200 X 35-55 yx. Total length to tip of perithecium 275-380 fi ; to inser-
tion-cell 140-190^. Width 34-40 /t. Appendages 175-275 /i.

On Gasnonia sp., Paris Mus. No. 116 bis, Bahia, Brazil. At base of
elytra and on superior prothorax.

Laboulbenia fallax nov. sp.

Perithecium becoming dark dirty olive-brown, the tip blackened, bent
outward, the lips hyaline. Receptacle becoming concolorous with the
perithecium except the hyaline slender basal cell, the remaining cells
usually short and stout except cell V, which extends up along the inner
margin of the perithecium nearly to its tij), its inner margin continuing
the curvature of the tip down to the insertion of the appendages, so that
the perithecium seems at first sight twice its actual size. Insertion-cell
unmodified, forming a slight rounded external prominence within which
the basal cells of the appendages form an evenly curved base from which
arises a single antero-posterior row of branches about twelve in number,
their lower cells slightly inflated, hyaline with dark septa, usually twice
branched ; the ultimate branchlets above the third or fourth septum slen-
der without dark septa, scarcely exceeding the tip of the perithecium.
Perithecium 100-120 X 35-40 /x. Total length to tip of perithecium
190-325 /Li; to insertion-cell 120-250 /x. Greatest width 85 /x. Appendages
50 /x. (The larger measurements are from the Amazon specimens.)

On Gyretes acutangulus Sharp, Brit. Mus. No. 771 (Biologia Coll.),
Bugaba, Panama; on Gyretes sp., Brit. Mus. No. 477, Amazon River; on
Gyretes sp., Hope Coll. No. 229, Rio de Janeiro. At tips of elytra.

Laboulbenia finitima nov. sp.

Perithecium one half to two thirds or more free, olivaceous brown,
lighter distally, becoming wholly dark brown, straight or curved outward
rarely inward ; the tip broad, rounded, generally not well differentiated,
blackish. Receptacle relatively small, the basal cell hyaline or yellowish,
the rest concolorous with the perithecium; cells III-V lighter, cell VI
extending down almost to cell I. Appendages brownish or pale oliva-
ceous, the outer sinaple, its basal cell twice as long as broad, externally
more deeply tinged with brown ; the basal cell of the inner appendage

THAXTER. NEW LABOULBENIACE^. 177

half as large, bearing a branch like the outer appendage on either side :
all the branches erect, closely associated, and often bent terminally across
the tip of the perithecium which they scarcely exceed, Perithecia, aver-
age 125 X 45 jM. Total length to tip of perithecium, average 240 /^t ; to
insertion-cell 145-150^; greatest width 48-50 yit. Appendages 30 fx.

On Pei'icallus guttatus Chev., Paris Museum, No. 78, Brit. Mus. No.
571, Java; on P. cceruleovirens Tat., Brit. Mus. No. 570, Singapore.
On the legs.

Laboulbenia fissa nov. sp.

Perithecium three fourths or more free, short and stout, slightly bent
toward the appendages, dirty olive-brown becoming deeply suffused with
blackish brown except distally just below the tip, which is abruptly dis-
tinguished, long, narrow, hyaline below its black distal portion, furcate,
the inner fork formed by the upgrowth of one of the inner (right) lip-
cells wliich grows outward and abruptly upward beside the deep black
blunt-tipped projection formed by the other three which it may equal in
length, though narrower and somewhat paler. Receptacle short, the
basal cell largest, pale yellowish or hyaline, of about the same diameter
throughout, broader than cell II, which is short, narrow, of equal diameter
throughout, hyaline or yellowish at the very base, the rest opaque and
indistinguishable from the remainder of the receptacle, which expands
abruptly above, becoming opaque except the upper part of cell IV and
cell V. Insertion-cell much narrower than cells IV-V. Outer append-
age consisting of several superposed cells, which form a black opaque
axis, usually broken off, curved outward, each cell producing a short hya-
line or brown edged brauchlet distally on the inner side ; the inner ap-
pendage consists of a smaller basal cell, which gives rise on either side to
a branch much like the outer appendage, its main axis less deeply black-
ened, curving outward on either side of the perithecium, the hyaline
branchlets arising from its convex side mostly once branched. Perithe-
cia, average 150 X 48 ;u, including the tip, which is about 45 X 18-20 /^t.
Total length to tip of perithecium 290-300 /n; to insertion-cell 185/*;
greatest width 70 /^. Appendages about 110 /x.

On Pericallus giitattiis Chev., Paris Museum, No. 78; Brit. Mus.
No. 571; Hope Collection, No. 301, Java. On P.favogutiulus Dej.,
E. Indies. On elytra.

TOL. XXXV. — 12

178 PROCEEDINGS OP THE AMERICAN ACADEMY.

Laboulbenia forficulata nov. sp.

Perithecium free, straight or somewhat curved, slightly iDflated, brown
except the basal wall-cells which form a mostly hyaline well developed
narrow neck-like stalk less than one third as long as the ascigerous portion,
the latter tapering rather abruptly at the tip, two of the lip-cells prolonged
to form a pair of erect pointed hyaline symmetrical apposed outgrowths
which resemble the tips of a pair of shears. Receptacle short, stout, sub-
triangular, cell I hyaline below, forming a short slender usually curved
pedicel ; the body of the receptacle suffused deeply below with blackish
brown, the more deeply suffused portions coarsely punctate with darker
spots. Appendages arising much as in L. pahnella, the outer consisting
of a basal cell from the blackened upper and outer margin of which arise
usually three branches in an antero-posterior series, which are curved
slightly outward and give rise from their convex side to secondary
branches which in turn may bear branchlets in a similar fashion : of the
primary branches the inmost is more copiously branched, the outmost
being small, usually broken ; all the branches black externally and brown
on the inner margins, or wholly opaque, usually constricted on the inner
side at the blackened septa, the terminal cells of some of the ultimate
branchlets abruptly inflated at the base : the basal cell of the inner ap-
pendage gives rise to a branch on either side, the two divergent and very
similar in character and mode of branching to those of the outer append-
age. Perithecium exclusive of neck, 150-200 X 28-38 /x ; the neck 35-
50y:iloug. Total length to tip of perithecium 300-450 /x; to insertion-
cell 100-150 /Lt; greatest width 50-70;;*. Appendages 175-200 /x.

On Thyreopferus striatus Guer., Hope Collection, No. 302, Madagas-
car. On elytra.

Laboulbenia geniculata nov. sp.

Perithecium free, long and narrow, translucent olive-brown except the
short somewhat constricted hyaline neck, the tip well distinguished
blackish brown bent outward, the lips coarse subhyaline oblique outward.
Receptacle nearly hyaline, except cell II and sometimes the upper part
of cell I, long and slender geniculate above cell II. Insertion-cell
broad, black, close to base of perithecial neck. Outer appendage con-
sisting of a larger basal subtriangular cell becoming more or less suffused
with olive-brown, surmounted by a series of five or six obliquely super-
posed hyaline cells which curves inward toward the perithecium ; each
cell of the series bearing externally a simple branch consisting of two
short faintly brownish basal cells constricted at the blackish septa, and a

THAXTER. — NEW LABOULBENIACE^, 179

terminal hyaline tapering portion about twice as long : the inner append-
age consisting; of a basal cell blackened below, from which arises on
either side a series of superposed cells like that of the outer appendage
and similarly branched excej^t that one or two of the lower cells of the
series bear antheridia, single, sessile or on a one-celled stalk. Perithecia
200-285 X 37 fi (the neck, 18-20 /x, included). Total length to tip of
perithecium 500-670 /x ; to insertion-cell 275-370 /x ; greatest width 55 /x.
Spores 65-70 X 5 fj.. Appendages 150-175^.

On GaJerita sp., Paris Museum, No. 160, Rosario, Argentine Republic.
On left side of inferior prothorax.

Laboulbenia gibbifera nov. sp.

Perithecium free or nearly so, somewhat narrower than in L. Der-
cyli, the apex narrower and more abruptly distinguished, the lips distinct,
turned slightly inward ; an external outgrowth just below the blackened
tip, which it may exceed in length, forming a free protuberance straight
or bent sidewise or inward, its apex evenly rounded or slightly inflated,
its outer margin continuous with the nearly straight margin of the peri-
thecium. Receptacle much as in L. Dercyli, more slender and shorter, the
appendages pushed outward by the enlargement of cell V. Appendages
much as in L. Dercyli, the basal cells projecting upward more prominently,
a group of branches in almost all cases arising apparently from cell V or
from a small cell separated from it, the lower segments of the branches
clearly differentiated and bearing externally three to six external and ter-
minal brauchlets. Perithecium 150-167 X 50^. Total length to tip
of perithecium 400-450 /a; to insertion-cell 275-340 /x. Appendages
150 /x.

On Dercylus tenebriosus Laf. {Eurysoma tenebrioides?), Hope Coll,
No. 328, Para; Brit. Mus. No. 586, " S. America." Inferior surface of
thorax and prothorax near base of two anterior pairs of legs. Possibly
a variety of L. Dercyli.

Laboulbenia heteroclieila nov. sp.

Perithecium olive-brown united to cell IV nearly to its base, rather
short, inflated below, tapering distally, the tip not abruptly differentiated,
blackened below, the four lip-cells all differently modified : of the two inner
lips one produces an erect rather slender brown finger-like terminal out-
growth, the other a shorter much broader paler outgrowth turned inward
nearly at right angles : of the two outer lips one forms merely a rounded

180 PROCEEDINGS OF THE AMERICAN ACADEMY.

prominence while the other grows out into a large prominent brown blunt-
tipped tooth-like projection which becomes bent slightly outward and
sideways. Receptacle rather long and slender, suffused with brown,
the darker distal portion marked by fine transverse indistinct striations.
The basal cells of the appendages more or less indistinguishable and
giving rise to indistinct prominences bearing numerous branches, the
basal cells of which are short, inflated distally, and bear a terminal and
external series of closely septate branchlets ; the latter once or twice
branched, the septa dark, the whole forming a dense tuft about half as
long as the perithecium. Spores 60 X 6 /x. Perithecium 200 X 70 /x.
Total length to tip of perithecium 550 n ; to insertion-cell 400 p.. Longest
lip-prominence 35 fi.

On Dineutes? sp., Brit. Mus. No. 486, Timor, E. Indies. Elytra.

Laboulbenia imitans nov. sp.

Perithecium free, long, rather slender, curved slightly outward, suffused
with dark smoky brown, tapering rather abruptly to a somewhat trun-
cate tip, the inner lip-cells darker, the basal cells forming a hyaline
well developed neck. Receptacle short, stout, subtriangular, usually
abruptly bent above the basal cell, becoming deeply suffused with
blackish brown, coarsely and conspicuously punctate, except where quite
opaque, the opacity first involving the anterior and lower portions above
the hyaline basal cell : cells IV and V nearly equal, cell IV bulging
outward more than half its upper surface, being free from and external
to the black insertion-cell. Appendages not reaching the tip of the
perithecium ; the outer consisting of a hyaline basal cell which gives rise
to an antero-posterior series of rigid rather slender slightly incurved
branches about six in number, arising from a deeply blackened ridge of
insertion ; the branches becoming deeply suffused with blackish brown,
except along their inner margins, closely septate, the lower cells giving
ri?e distally and inwardly to secondary branches similar to the primary
ones and often developing short hyaline branchlets in a similar fashion.
The inner appendage consisting of a basal cell which gives rise on either
side to a series of from two to three branches similar in character to
those of the outer appendage : the whole forming a somewhat crest-like
dense tuft. Spores 42^ long. Perithecium (exclusive of neck) 150-
155 X 28/x; the neck 20-30 X 20 fi. Total length to tip of perithecium
270-340 ;x; to insertion-cell 100 /i ; greatest width 50^. Appendages,
average 45 fi long.

THAXTER. — NEW LABOULBENIACE^. 181

On Nj/cteis sp., Paris Museum, No. 29, Madagascar. On legs, elytra,
and abdomen. Resembling a sjiecies of Corethromyces in general appear-
ance.

Laboulbenia insularis nov. sp.

Perithecium one third or more free, inflated, dark brown, becoming
almost opa(iue, tapering to the rather pointed apex, the lip-edges hyaline.
Peceptacle elongate, the distal portion concolorous with the perithe-
cium ; cell I suffused with brown above and below, cell II hyaline
except for a brown suffusion at its base and distal end. Outer appendage
consisting of a large subconical basal cell becoming dark brown and
bearing terminally a single short erect slender branch, the two basal cells
of which are dark blackish brown, the septa usually oblique ; the distal
cell somewhat longer than the rest of the appendage, hyaline and soon
broken. The inner appendage consisting of a very small basal cell,
usually producing a single short hyaline branch about as long as that of
the outer appendage. Perithecium 85-95 x 40 yx. Total length to tip
of perithecium 215-275 /:i; to insertion-cell 200-250 /;i. Appendages,
longest 100 /x.

On Bemhldium suhlimatum Woll., and B. Grayanum Woll., Brit. Mus.
No. 409, Island of St. Helena.

Laboulbenia intermedia nov. sp.

Perithecium about three fourths free, suffused with yellowish brown
deeper below the apex, rather stout, slightly inflated, the apex very broad,
rounded, often almost flat, short, wholly blackened, the lips indistinguish-
able, the wall-cells with a slight spiral twist. Receptacle short and
stout, at first hyaline below, becoming concolorous with the perithecium.
Outer appendage consisting of a rather large subisodiametric cell bearing
distally two or three antero-posterior branches, once or twice branched,
the external branchlets brown, basally deeply blackened. Inner append-
age consisting of a basal cell similar to that of the outer and producing
usually two simple or once branched branches placed antero-posteriorly,
the ultimate branchlets in both appendages elongate, tapering slightly
distally, hyaline. Perithecia 95-105 x 35-40 fi. Total length to tip of
perithecium, average 190 /t ; to insertion-cell 120 fx. Appendages, longest
300^.

On Anisodactyliis tricuspidatus A. Mor., Paris Museum, No. 199,
Mon-Pin, (China?). Margin of the elytra.

182 PROCEEDINGS OF THE AMEEICAN ACADEMY.

Laboulbenia Italica nov. sp.

Peritheciiim free except at its base, rather short and stout, the upper
half or third curved strongly outward, the tip large, sulcate, blackened, the
lips coarse, nearly equal, subhyaline. Receptacle concolorous with the
perithecium, the base nearly hyaline, usually bent between cells I and
II, short, abruptly expanded above cell II, the anterior margin straight
above cell I. Appendages arranged very much as in L. orientalis, the
basal cells subtriangular, the outer producing externally an oblique row
of about four superposed branches from a blackened area of insertion, the
branches erect mostly twice subdichotomously branched, all the lower
septa blackened and constricted, the inner appendage similar to the outer :
the insertion-cell normally placed, broad, subhyaline, close to base of
perithecium. Antheridia brown, the vente much inflated, the neck
becoming pointed, 23 X 8 //. Perithecia 100 X 42 jU. Total length to
tip of perithecium 275 ^i ; to insertion-cell 175 fi. Appendages 140jM.

On Brachinus explodens Duft., Florence Museum, Florence, Italy.

Laboulbenia Javana nov. sp.

Perithecium about two thirds or less free, sessile, tinged with brown,
paler in the middle, the distal portion hyaline and tapei'ing considerably to
the greatly modified tip, which is deeply suffused with blackish brown and
bears two often symmetrically placed divergent outgrowths, the inner
broader at the base and much shorter than the outer which is finger-like,
with a hyaline area above, close beside the subterminal pore : the
outer of these two projections appears, through a twist in the perithe-
cium, to lie on the inner side. Receptacle pale yellowish with brownish
shades, the deeply suffused base of the perithecium opposite the upper
half of cell III or lower ; the distal portion usually so twisted and bent
in conjunction with the perithecium that it crosses the latter and the
appendages at a considerable angle. Insertion-cell higher than the middle
of the perithecium. Appendages concolorous with the receptacle, consist-
ing of an outer basal cell which bears a single simple branch, slightly
nodulose above and below the septa, the basal cell of the inner appendage
smaller, bearing a single branch on either side which may be once
branched, Perithecia about 100 ^u long exclusive of outgrowth, 27/*
broad. Total length to tip of perithecium 190 /* ; to insertion-cell 140 /i.
Appendages 130 /i. Distance from tip to tip of perithecial outgrowths

45-48 IX.

On Pericallus cicinddoides MacLeary, Paris Museum, No 143, Ton-
gou, Java. On inferior surface of thorax.

THAXTER. — iNEW LABOULBENIACE^. 183

Laboulbenia leucophaea nov. sp.

Perithecium dark brown, almost opaijue, rather small, hardly more
thau the tip free from the receptacle ; the tip relatively large and long,
bent slightly outward, not abruptly differentiated, black except around
the pore, the right inner lip forming a hyaline nearly median blunt out-
growth which is bent slightly outward. Receptacle sometimes twisted at
the distal end of and above cell II j cell I hyaline; cell II suffused with
brown, in some cases with deeper brown tranverse elevations on one
side ; the lower half of cell III and cell VI hyaline, the rest of the
receptacle concolorous with the perithecium. The outer appendage
simple, its basal cell four or more times as long as broad, curved toward
the perithecium ; the basal cell of the inner appendage very small,
bearing in general a single short branch, both appendages pale yellowish.
Perithecium to tip of outgrowth 130-150 X 35-40 ju. Total length to
tip of perithecium 325-375 /x ; to insertion-cell 250-290 /a ; greatest
width 55-65 ^. Appendages about 200 jjl.

On Serrimargo guttiger Scbaum., Hope Collection, No. 300, Sumatra.
Mid-elytron and base of legs.

Laboulbenia Loxandri nov. sp.

Perithecium about three fourths free, suffused with brownish, translu-
cent, the distal half narrow and strongly curved inward, especially at the
tip, the latter externally and distally blackened, the lip-edges hyaline.
Receptacle rather stout, pale dirty brownish ; cell II basally and distally
and cell VI externally more deeply suffused. Cells IV and V elongated
so that they become parallel and carry the insertion-cell upward and out-
ward free from the perithecium. Outer appendage consisting of a
rounded basal cell bearing a single terminal branch, the basal cell and
one or two cells above it rounded, constricted at the mostly blackened
septa, simple or each of the lower cells producing distally on the inner
side a branchlet, the branchlets and the terminal portion of the main
branch hyaline, slender, thin-walled, tapering. Inner appendage con-
sisting of a basal cell like the outer and like it producing a branch on
either side. Spores about 45 X 4 ^. Perithecium 140 X 40 /i. Total
length to tip of perithecium 340 /x ; to insertion-cell 275 fi. Appendages,
longest 120^.

On Loxnndrus unistigma Bates, Brit. Mus. No. 659 (Biologia Coll.),
Paso Antonio, Guatemala. Elytra.

184 PROCEEDINGS OP THE AMERICAN ACADEMY.

Laboulbenia maculata nov. sp.

Perithecium free, dark brown becoming nearly opaque, the outer
margin more convex than the inner ; somewhat constricted at the base,
the tip rather abruptly distinguished externally, the margins nearly
straight, the inner lips small and prominent, the outer broad, straight,
oblique. Receptacle abnormal, cell I short, slender, curved, opaque ; cell

II nearly hyaline in the middle, brownish above, coarsely spotted with
blackish brown below, becoming darker and indistinguishable from cell
I at its base ; cell VI distally nearly hyaline and narrow, extending
down beside cell II nearly if not quite to cell I, its base spotted as in cell
II ; cell VII (the " secondary stalk-cell ") external to it, the margin
blackish brown especially distally, extending down beside cell VI to
within a short distance of its base where it is similarly punctate
towards its base or throughout ; cell III narrow, external to the upper
two thirds of cell II, punctate below, its distal end close beside the cor-
responding termination of cell II ; the base of cell IV overlapping cell

III so that a cross section in this region would cut cells II, III, IV,
VI, and VII : distal portion of the receptacle concolorous with the
perithecium or somewhat paler. The perithecium bent toward and
partly or wholly overlapping the insertion-cell. Appendages directed
across the lower half of the perithecium sometimes at right angles ;
consisting of a large outer basal cell sometimes slightly inflated, bearing
distally one or usually two antero-jjosterior simple branches which are
slender, rather rigid, their diameter much less than that of the basal cell :
the inner appendage consisting of a smaller basal cell which may pro-
duce one or two branches similar to those of the outer appendage ; all
the branches slender, rather rigid and straight, parallel and closely ap-
proximated, tapering but slightly. Perithecia 225 x 60 /x. Total length
to tip of perithecium 560 /x ; to insertion-cell 375 /x; greatest width 140/x.
Appendages about 200-250 /x.

On Serrimargo guttiger Schaum., Brit. Mus. No. 559, Penang, East
Indies. On anterior legs.

Laboulbenia Madagascarensis nov. sp.

Perithecium free, mostly straight, the inner margin more convex, uni-
formly clear dark brown or blackish except just below the black tip,
abruptly distinguished from and contrasting with the receptacle ; the tip
rather abruptly distinguished, straight or bent slightly inward, with
hyaline lip-margins, the wall-cells with a slight spiral turn. Receptacle

THAXTER. — NEW LABOULBENIACE^. 185

hyaliue or finally yellowish, cells HI and VI about equal. Insertion
cell opposite base of peritheciuni. Outer appendage consisting of a
rather small basal cell, its outer wall blackened, the blackening con-
tinuous with the insertion cell, producing distally usually two branches,
an outer blackened externally or suffused with brown at its base and
once branched, and an inner usually simple and hyaline. The inner
appendage consists of a basal cell like that of the outer, and produces a
single branch on either side which may be once branched, all the branch-
lets of both appendages rather stout and stiff, tapering, slightly curved
outward, hyaline or becoming dirty yellowish. Perithecia 100-120 X
40-45 iJL. Total length to tip of perithecium 240-270 /x; to insertion-cell
140-155 fi. Appendages, longest 250 /x.

On a Carabid allied to Harpalus, Paris Museum, No. 3, Madagascar.
On margins of both elytra.

Laboulbenia Madeirse nov. sp.

Perithecium united to receptacle for about two thirds of its length,
pale straw-yellow becoming brownish yellow, the whole tip clear con-
trasting black or blackish brown, the hyaline lip-edges turned outward.
Receptacle concolorous with perithecium, rather short, normal. Cell V
relatively large, its upper margin free between the perithecium and the
insertion-cell, the latter oblique, clear black, contrasting. Outer append-
age often simple, elongate, sometimes once branched above its subbasal
cell ; the branches divergent : inner appendage consisting of a basal cell
smaller than that of the outer, and bearing one or two short branches
commonly three-celled. Spores 75 X 6 /x. Perithecia 100-130 x 35-
40 /x. Totallength to tip of perithecium 225-250 /x; to insertion-cell
175-210 /i. Appendages, longer 350 /^i.

On Calathus complanatus Dej., Paris Museum, No. 211, Madeira. On
elytra.

Laboulbenia Malayensis nov. sp.
Perithecium clear translucent brown with a slight olive tinge, becom-
ing almost opaque ; united to the receptacle except the abruptly dis-
tinguished tip which is hyaline, except the blackened lips ; the latter
turned abruptly usually to the right, forming a lateral somewhat irregu-
larly four-lobed papilla in which the hyaline pore is central. Cells I
and II of the receptacle about equal in length, nearly hyaline, often dis-
tally olivaceous; cells III and IV relatively large, translucent olive-
brown, cell IV bulging distally so that the dark but not opaque inser-

186 PROCEEDINGS OP THE AMERICAN ACADEMY.

tion-cell is turned obliquely toward the tip of the perithecium : cells
VI and the basal cells of the receptacle more or less tinged with olive-
brown forming an elevation so that the perithecium appears indented
below, all the cells except cell T marked by fine transverse striations.
Outer appendage simple, the basal cell rather large, often externally
indented near the base, the second and third cells nearly equal, nari'ower
than the basal cell and the cells immediately above them; the rest of the
appendage tapering to the hyaline attenuated elongate distal portion ; the
inner appendage consisting of a basal cell one third as large as that of
the outer and bearing a single branch on either side, one or both of
which may be elongate much like the outer appendage, bearing one or two
short slender antheridial branches near the base which are bent rather
abrujitly upward from their point of origin ; the branches all distally
hyaline and attenuated ; the basal cells faintly reddish. Ferithecia,
average 110 X 37 /t. Total length to tip of perithecium 260-280 /j.; to
insertion-cell 250-275 fi ; greatest width 75 fj.. Appendages, longest
375 ,x.

On PericaUus cceruleovirens Tat., Brit. Mus. No. 570, Singapore. At
base of posterior legs.

Laboulbenia melanaria nov. sp.

Perithecium nearly free, uniformly suffused with clear blackish brown,
straight or bent slightly outward ; the tip more deeply colored, the lip-
edges hyaline, contrasting, externally oblique. Receptacle hyaline
becoming yellowish, often suffused with blackish brown except the lower
portions of cells I, III, and VI, and usually cell V. Outer appendage
consisting of a basal cell mostly free, bearing terminally a single branch
typically once branched above its basal cell, the branchlets elongate,
thick-walled, rigid, more or less tinged with brown. The inner append-
age consisting of a much smaller basal cell, producing either a short
two-celled branch with one or two terminal antheridia, or two longer
branches which may be once branched ; the branches like those of the
outer appendage but shorter. Perithecium 120 X 35 \i. Total length
to tip of perithecium 275 /x ; to insertion-cell 150 /x. Apj^endages,
longest, 550 /u.

On Diachrormis germanus Linn., Florence Museum, Florence, Hope
Coll. No. 344i, 319, France, Portugal ; on Anisodactylus militaris, No.
315, Sardinia; on A. heros Fabr., No. 316, "Europe."

THAXTER. — NEW LABOULBENIACE^. 187

Labovilbenia melanopus nov. sp.

Perithecium free except at the base, large, rather deeply suffused with
smoky brown, translucent, not contrasting ; the subdistal wall-cells
lighter, tapering abruptly to the narrow somewhat incurved tip, the
inner lip-cells only deeply blackened. Receptacle somewhat curved,
tapering below to the short slender basal cell which is smoky black
except at its base, the distal cells gradually suffused with yellowish
brown. Insertion-cell opaque only externally, the basal cells of the
appendages becoming apparently divided into several cells which are
opaque or nearly so and indistinguishable in the mature plant, giving
rise to numerous branches the basal cells of which are distally inflated
and bear terminally numerous branchlets (about six to ten), the latter
very slender, flexuous, not as long as the perithecium. Spores about
60 X 4.5 /I. Perithecium 290 x 70 ^. Total length to tip of perithe-
cium QlOfi; to insertion-cell -tOO/i. Appendages, longer 140 /i.

On Carabid (allied to Harpalusf), Paris Mus. No. 115, Africa. On
tip of abdomen.

Laboulbenia microscopica nov. sp.

Perithecium one half or wholly free, pale olivaceous, somewhat inflated,
tapering to the relatively long narrow subtruncate blackened tip which
is bent slightly inward. Lower half of receptacle greatly reduced in size,
the basal cell hyaline or nearly so, the rest suffused with dark brown,
cell III paler, cell II broader than long, cells III to V relatively large,
bulging prominently outward beyond and below the'^insertion-cell. Outer
appendage consisting of a basal cell which becomes sub-triangular through
the protrusion of its upper outer angle which renders its distal margin
twice as broad as the basal cell of the single branch which rises from its
upper inner half. The inner appendage arising from a much smaller
basal cell which produces two branches. Spores about 35 X 3 fx. Peri-
thecia 75-93 X 27-34 fi. Total length to tip of perithecium 120-140 fj. ;
to insertion-cell 75-90 n. Greatest width 45-60 ^. Appendages
about 70 /x.

On Pelmatellus nitescens Bates, Brit. Mus. (Biologia Coll.), No. 683,
Vera Paz, Guatemala. On elytra.

Laboulbenia microsoma nov. sp.

Perithecium free, several times as large as the receptacle, smoky brown
darker basally and distally, the longitudinal septa subhyaline, the outer
margin concave, the inner convex ; a subterminal external small rounded

188 PROCEEDINGS OF THE AMERICAN ACADEMY.

elevation ; the tip very broad, short, almost flat-topped, the outer angle
almost a right angle, the inner rounded. Receptacle consisting of a basal
cell which is nearly hyaline, above which cells II, III, and VI form an
almost transverse row ; cell II median, triangular, lying between the
other two, the receptacle abruptly expanded in this region; cells III and
IV small and flattened; cell V hardly distinguishable. Insertion-cell
and basal cells of the appendages nearly opaque and indistinguishable
from one another, the outer basal cell apparently producing two branches
antero-posteriorly ; the inner a branch on either side, all the branches
(broken) brown, stiflf, erect or slightly divergent. Perithecium 185 X
6C /x. Total length to tip of perithecium 295 /x ; to insertion-cell 90 /x;
greatest width 65 ft.

On Serrimargo guttiger Schaura., Brit. Mus. No. 560, Penang, East
Indies. At base of posterior legs.

Laboulbenia minimalis nov. sp.

Perithecium free, becoming olivaceous brown, mostly straight, the
basal wall-cells forming a very short stalk, the tip rather abruptly distin-
guished, mostly straight symmetrical black, distally hyaline. Receptacle
olivaceous yellow with brown suffusions ; cell I slightly suffused with
brown, somewhat longer than cell II, both rather narrow ; the receptacle
expanding rather abruptly above cell II ; cells III, IV, and V nearly
equal becoming rather deeply suffused with brown. Insertion-cell broad,
blackened. Outer appendage consisting of a large triangular basal cell
becoming deep blacrfsh brown, above which four to six small nearly
hyaline cells obliquely superposed, or with their long axes nearly vertical,
form a series which runs obliquely toward the perithecium, each cell pro-
ducing externally a single branch ; the branches either simple and mostly
three-celfed or branched above their basal cells ; the branchlets two in
number, mostly four-celled, their basal and terminal cells very small, all
the septa somewhat dark, slightly constricted : the inner appendage con-
sisting: of a basal cell bearing on either side a short series of cells like
that of the outer appendage and similarly branched, except that the three
or four lower branches consist of a single cell bearing terminally a pair
of rather stout long-necked antheridia ; the three series closely apposed
or united. Perithecia 100 X 30^. Total length to tip of perithecium
200-235 /x ; to insertion-cell 110 yu.; greatest width 45 yu,. Appendages
60-75 f,.

On Galerita sp., Paris Museum, No. 74, Venezuela. On mid-elytron.

THAXTER. — NEW LABOULBENIACE^. 189

Laboulbenia Misceii nov. sp.

Peritbecium free, loug aud sleuder, trauslucent, pale brownish olive,
narrowed at the base to form a short paler stalk which lies opposite the
insertion-cell; the tip long, not very abruptly distinguished, paler below,
straight or turned slightly outward, distally blackened on the inner side ;
the lips variable, rather prominent. Receptacle rather short and stout,
darker olive-brown ; the basal cell pale yellowish. Insertion-cell nearly
as broad as cells IV- V. Basal cell of outer appendage blackish brown
externally, bearing a single terminal branch of less diameter externally
suffused with blackish brown, slightly curved outward and bearing two
or three branchlets from the inner side which are hyaline, the basal cells
somewhat suffused with brown ; the basal cell of the inner appendage
smaller than that of the outer, nearly hyaline aud bearing a branch on
either side similar to the outer appendage. Perithecia including base
145-180 [X. Total length to tip of peritbecium 240-300 [x ; to insertion-
cell 90-130 /x ; greatest width 35-40 fi. Appendages, longer 150 fx.

On Miscelus sp., Paris Museum, No. 114, Isles des Moluques. At
base of posterior legs.

Laboulbenia obtusa nov. sp.

Peritbecium about three fourths free, becoming opaque, black-brown,
very stout, the outer margin slightly and more or less symmetrically
convex, the inner bulging prominently distally and curved abruptly to
the brown blunt rounded hardly differentiated apex, the pore external.
Receptacle short, cell VI together with the basal cells of the peritbecium
concolorous with the latter, becoming indistinguishable ; cell VI ex-
tending to or towards the base of cell 1 1, which is mostly suffused above,
hyaline and contrasting below as is cell I; cells III and IV with
median brown shades. Insertion-cell black-brown. Appendages hyaline
becoming tinged with brown, the outer basal cell twice as long as the
inner, each bearing one to two branches which form a compact group
curved toward aud against the peritbecium. Peritbecium 120 X G5 /x.
Total length to tip of peritbecium, average 260 /x ; to insertion-cell
175 jx. Appendages, broken, 35 fx.

On Aerogenidion Bedeli Tsch., Paris Museum, No. 198, Mon-Pin,
(China?). On left inferior margin of prothorax.

Laboulbenia CEdodactyli nov. sp.
Peritbecium free except at the base, pale transparent amber-yellow,
somewhat inflated at the base aud tapering gradually thence to the

190 PROCEEDINGS OF THE AMERICAN ACADEMY.

sleuder tip, a blackish shade below the nearly hyaline lips which are
turned slightly outward. Receptacle amber-colored, deeper anteriorly,
cell II sometimes elongate, cell VI very short, so that the base of the
perithecium comes opposite cell III. Insertion-cell and the inner margin
of cell V usually free from the perithecium. The outer appendage
simple divergent, the basal cell very large ; the basal cell of the inner
appendage much smaller, bearing one to two short branchlets. Spores
o5-40 X 41 jw. Perithecia 120 X 35 /j. Total length to tip of perithe-
cium 175-380 fi (longest) ; to insertion-cell 115-275 //.

On " (Edodactijlus fuscobrunneus," Brit. Mus. No. 397, Chili. On
elytra.

Laboulbenia Oopteri nov. sp.

Perithecium three fourths or more free, translucent blackish brown,
the inner margin evenly curved outward, the outer margin with slight
elevations at the septa and curved abruptly inward to form the well
differentiated tip which is pale brownish, with dark inferior suffusions.
Receptacle concolorous with jDerithecium, except that cells I and II are
usually hyaline, the suffused portions sparsely and rather coarsely and
distinctly punctate. Insertion-cell broad and black. Basal cell of the
outer appendage for the most part very long, bearing distally an outer
and an inner branch, the former with blackened basal septum, simple,
or once branched in which case the basal septum of the outer branchlet
is also blackened. Inner appendage consisting of a very small basal cell
bearing a short branch on either side. Perithecia 95-110 X 30-35 (x.
Total length to tip of perithecium 175-275 fx ; to insei'tion-cell 85-160 ^.
Appendage broken, 200 jw, probably much longer.

On Oopterus rotundicollis White, Brit. Mus. No. 613, New Zealand.
On elytra.

Laboulbenia Ophoni nov. sp.

Perithecium free except at the base, short and stout, pale straw-colored
or nearly hyaline, somewhat inflated ; the inner margin more convex, the
black tip abruptly differentiated on its inner side, black, contrasting, the
lip-edges hyaline turned slightly outward. Receptacle short, stout, nor-
mal, concolorous with the perithecium. Insertion-cell black, contrasting.
The onter appendage divergent, simple or once to three times branched,
the ultimate branchlets distally attenuated ; the inner appendage consist-
ins of a basal cell half as larcre as that of the outer, bearing; a short branch
on either side which may be several times branched, the antheridia borne
in small groups. Spores 28x3//. Perithecia 70 X 30-34/1. Total

THAXTER. — NEW LABOULBENIACEiE. 191

length to tip of peritheciuai IGo/i; to insertion-cell 85-100 ;U ; width 40 ^t/.
Appendages, longest 200 /i.

On Ophonus ohscurus Fabr., 0. brevicolUs Dej., 0. azurens Fabr.,
Harpnlus neglectus Dej., H. sei-ripes Quensel, H. sulpluiripes Germ.,
// tardus Panz., in Florence Museum collection of Italian Coleoptera.
On Ophonus sp.?, Interlaken, Switzerland. On Ophonus sp., Paris Mu-
seum, No. 37, Algeria. On elytra, inferior thorax and prothorax, and
abdomen.

Laboulbenia Orectochili nov. sp.

Perithecium free except at the base, more or less evenly suffused with
smoky brown, with a subtermiual external blackish patch, nearly sym-
metrical and straight, slightly inflated, tapering gradually to the hyaline tip
which is surmounted by a median straight pointed purplish tooth-like pro-
jection formed by the outgrowth of one of the lip-cells; the inner lip-cells
forming a small hyaline or partly purplish lateral papilla. Receptacle elon-
gate, cells I and II stout, the latter slightly if at all narrower distally, cells
IV and V and the basal cells of the perithecium darker brown, the rest
very pale yellowish or purplish brown, finely punctate, the dots scarcely
visible except in the more deeply suffused areas. The insertion-cell
broad, blackened, extending completely across the distal margins of cells
IV and V. Appendages consisting of an inner and outer basal cell,
giving rise in all to from five to ten erect subconical prominences, each
of which becomes separated as the basal cell of a very short two-celled
branch of which only a blackened basal portion remains in mature speci-
mens, the rounded purplish slightly inflated terminal portion of the
upper cell usually breaking off above its blackened slightly constricted
basal half. Of the branches that borne by the protuberance first formed
from the outer basal cell is always somewhat larger and more prominent
than the rest. Perithecia 190 x 59 ,«. Total length to tip of perithecium
475-680 fi ; to insertion-cell 400-550 fi.

On Orectochilus cordatus Reg., Paris No. 99, " Asia." On elytra.

Laboulbenia orientalis nov. sp.
Perithecium straight, its base free from and higher than the insertion
of the appendages, straight to strongly recurved, becoming suffused with
pale brownish ; the tip blackish brown in normal specimens, well dis-
tinguished, with prominent lips (when curved, not abruptly distinguished,
somewhat pointed, with ill defined lips), the translucent edges dirty brown.
Receptacle hyaline or concolorous with the perithecium, sometimes be-

192 PROCEEDINGS OP THE AMERICAN ACADEMY.

coming dark smoky brown ; cell V often as large as cell IV, pushing the
small subtriaugular unmodified insertion-cell outward so that it may
become lateral, with its transverse diameter vertical, cell VII unusually
large. Appendages consisting of an outer and an inner basal cell, the
two free from one another except at the base, mostly several times as
long as broad and overlapping slightly ; the outer bearing an external
row of superposed branches, usually seven or eight in number, formed by
the successive proliferation of the tip of the basal cell, and separated from
it by broadly blackened septa ; the branches successively subdichoto-
mously branched several to eight or more times, the basal and sometimes
the subbasal cell often producing more than two branchlets (two to four)
superposed in a single row. The inner appendage like the outer, the
basal cell producing a single similar row of branches fewer (usually two
to four) in number, overlapping those of the outer appendage and bear-
ing antheridia in gi'oups of from one to eight not characteristically grouped,
the venter rather abruptly distinguished from the straight cylindrical
purplish neck : the branches of both appendages directed outward, hyaline
or distally reddish or purplish, constricted at the lower purplish septa.
Perithecia (largest) 230 X 55 /j,; average 170 X 40 ^u. Total length to tip
of perithecium very variable, from 275 /^ to 1 mm. Appendages 200-
350 |U. Antheridia 16 X 4ju.

On Brachimis Chinensis Chaud., Paris Museum, Nos. 58, 59, Manila,
Philippine Islands, and Macao, China. Brit. Mus. Nos. 536 (bis), China.
Hope Coll. No. 244, China. On Brachinus spp., Brit. Mus. Nos. 537,
539, 540, China and Philippine Islands. Usually on inferior surface of
thorax and prothorax.

Laboulbenia Ortliomi nov. sp.

Perithecium free, long and straight, slender, sometimes slightly inflated
distally, deep clear brown ; the tip broad, prominent, not abruptly differ-
entiated ; the lips rather large and prominent, the lower wall-cell as a
rule elongated to form a hyaline neck, usually well marked and contrasting
with the body of the perithecium. Receptacle shorter than the perithe-
cium, olive-brown, except the hyaline or slightly yellowish basal cell.
Insertion-cell not as broad as cell IV. The outer appendage consisting of
a basal cell longer than broad, blackened externally, producing usually
a single simple terminal branch, the two lower cells of which are black-
ened externally and sometimes give rise to erect simple branchlets. The
inner appendage consisting of a basal cell similar to that of the outer, and
producing on either side a straight hyaline erect branch. Spores 50 X

THAXTER. — NEW LABOULBENIACE^E. 193

ifi. Perithecia 130-140 X 30-35 |U exclusive of the variably developed
neck, which may be 18/x long. Total length to tip of perithecium 2G0ju ;
to insertioa-cell 100 ^tt; width 40^. Longer appendages 200-270^.

On Orthomus aquilus Coquer, Algeria ?, Paris Museum, No. 41. On
margin of elytra.

Laboulbenia pallida nov. sp.

Perithecium almost wholly free from the receptacle, colorless becoming
faintly yellowish, bent outward ; the prominent tip abruptly distinguished,
coarse lipped, hyaline except for an inner blackish patch. Receptacle
concolorous with the perithecium, the basal cell large and broad, longer
than cell II, the cells of the distal portion relatively small, cell III
roundish, about as large as cells IV and V together. Insertion-cell thick,
contrasting purplish black. Outer appendage consisting of a basal cell,
rectangular or distally enlarged and producing usually two, sometimes
but one branch, the branches once or even twice branched, the ultimate
branchlets sometimes very elongate and attenuated : the inner appendage
consistincr of a basal cell much smaller than that of the outer and some-
times lateral in position, bearing one or two branches which may be
short or elongate like those of the outer appendage. Perithecia 70 X
2b (X. Total length to tip of perithecium 110-175 /x; to insertion-cell
85-120 ju. Appendages, longest 285 ^a.

On a Carabid allied to Harpalus, Paris Museum, No. 94, Java. On
elytra.

Laboulbenia Papuana nov. sp.

Perithecium nearly two thirds free, straight or curved slightly outward,
the inner margin convex ; pale brownish yellow, the tip rather well dif-
ferentiated, blackish ; the hyaline irregularly prominent lip-edges turned
inward. Receptacle elongate, cell II, and cell I except at its base, con-
spicuously tinged with blackish and faintly marked by fine transverse
striations, the rest of the receptacle concolorous with the perithecium ;
cell lY externally concave, the whole receptacle more or less prominently
bent anteriorly in the region of cells III and VI. Insertion-cell broad,
horizontal, black, narrower than cells IV-V. The basal cells of the
appendages simple and distinct, the outer producing a single branch
which may branch once ; the branchlets short ; the basal cell of the inner
appendage producing two small branches which may be once branched.
Spores about 65 X 5.5 n. Perithecium 160-200 X 55-65 ^. Total length
to tip of perithecium 650-880 /a ; to insertion-cell 544-700 ^. Appendages,
longest seen, 140 ^u.

VOL. XXXV. — 13

194 PROCEEDINGS OF THE AMERICAN ACADEMY.

On Morio sp., Paris Museum, No. 112, New Guinea. On anterior
inferior surface of thorax on right side.

Laboulbenia Pericalli nov. sp.

Peritheciura becoming almost opaque, its upper fourth, sometimes only
the tip, free from the receptacle, relatively small, the tip more or less
prominent, sometimes subconical, short, and wholly black (often more
prominent, abruptly distinguished, the rounded lips well defined with
hyaline edges). Receptacle normal, cells III and IV large and broad,
concolorous with the peritliecium ; cells I and II together with the lower
end of cell VI pale yellowish. Outer appendage mostly simple, stout,
the lower cells slightly inflated ; inner appendage consisting of a much
smaller basal cell, bearing a usually simple branch on either side very
similar to the outer appendage ; all the branches yellowish or becoming
tinged with brown, especially toward the base. Perithecia 110-130 X
37-45 1.1. Total length to tip of perithecium 200-300 /x ; greatest breadth
about 75 fi.

On Pericallus guttatus Chev., Paris Museum, No. 78, Java ; on Misce-
lus sp., Paris Museum, Nos. 113, 114, 115, New Guinea.

Laboulbenia platystoma nov. sp.

Perithecium free except at the very base, straight, rather long and
narrow, pale amber-yellow becoming slightly tinged with brown, slightly
and symmetrically inflated ; the tip black, opaque, the lip-cells forming
an abruptly spreading almost flat symmetrical termination with a slight
median indentation, two of the lips forming a small median hyaline
truncate cone, the other two arching over them from the outer and inner
side, wholly opaque except their inner margins, the blackened part not
quite meeting in the median line, the whole resembling the end of a
pair of horizontal cut plyers. Receptacle medium, pale amber-yellow
becoming tinged with brown distally. Outer appendage mostly simple,
the basal cell about twice as long as broad, often slightly inflated, the
rest of the appendage much narrower, straight, rigid, tinged with brown
tapering somewhat distally ; the inner appendage consisting of a basal
cell about half as long as that of the outer, producing a branch on either
side ; the branch usually bearing an antheridial branchlet near its base,
the branches and the outer appendage similar, often curved slightly out-
ward, becoming tinged with brown. Perithecia 175 X 40/i; the tip 37 /x
broad. Total length to tip of perithecium 325-400 /i; to insertion-cell
185-230 /Lt; greatest width 55-63 /^t. Appendages about llofi (longest).

THAXTER. — NEW LABOULBENIACE^. 195

On Catoscopus sp., Paris Museum, No. 119, New Guinea. On in-
ferior surface.

Laboulbenia PolyhirmsB no v. sp.

Perithecium rather slender, almost wholly free, nearly hyaline or
faintly brownish yellow ; tapering slightly toward the moderately well
differentiated tip which is usually bent slightly outward, more or less
blackened on the inner side, sometimes wholly black, the inner lips often
prominently rounded and terminal. Receptacle concolorous with the
{lerithecium, marked by faint transverse striations, long and slender;
cell I [ usually greatly elongated; the distal portion small and normal.
Outer appendage consisting of a small basal cell more or less rounded
and producing distally from two to four branches, usually four ; an
outer and an inner, the two others placed between them one on either
side, the branches simple or usually not more than once branched above
the basal cell ; the branchlets slender, often flexed, rather closely septate,
hardly tapering: the inner appendage consisting of a smaller rounded ba-
sal cell which produces on either side a single branch, which may branch
several times and bears hyaline lateral or terminal autheridia singly
or in pairs; all the branches nearly hyaline. Perithecia 130-190 X
30-40 fi. Total length to tip of perithecium 400-700 ^, average 500 ^,
to insertion-cell 275-450 fi. Appendages, longest 350-400 ^. Greatest
width 40-50 /f.

On Polyhirma sp., Paris Museum, Nos. 5, 6, and 168, Tangar, Algeria.
On inferior surface of abdomen and thorax especially in depressions
at base of posterior legs.

Laboulbenia prominens nov. sp.

Perithecium short and stout, less than half free from the receptacle,
nearly opaque blackish brown lighter distally, the short broad blunt
black tip rather abruptly distinguished on the inner side. Receptacle
rather stout, cells I and II dirty yellowish or subhyaline, the rest more
or less deeply suffused with blackish brown; all the cells except cell I
marked by fine transverse striations more distinct on the suffused por-
tions ; cells III and IV large and prominent, the insertion-cell broad but
narrower than cells IV- V. Outer appendage consisting of a short
irregular cell abruptly narrowed distally and bearing two branches an-
tero-posteriorly, the outer externally deep blackish brown, once branched ;
the outer branchlet also branched, the inner consisting of a short basal
cell which bears an inner and an outer branchlet, the branchlets pale

196 PROCEEDINGS OF THE AMERICAN ACADEMY.

brown, curved outward : the inner appendage consisting of a basal cell
smaller than that of the outer, irregular and bearing a brancli on either
side ; each once branched, the branchlets like those of the outer append-
age. Perithecia lo0x52|U. Total length to tip of perithecium 310-
330 /x; to insertion-cell 21b ji. Greatest width 95-100 ^t.

On Pericallus guttatus Chev., Brit. Mus. No. 571, Java. On legs.

Laboulbenia protrudens nov. sp.

Perithecium smoky brown with a tinge of olive, the outer margin con-
cave, relatively small ; the tip only free from the receptacle, short, rounded,
not abruptly differentiated, bent slightly outward, black except around
the pore. Receptacle dirty olivaceous ; cell I, except at the base, con-
colorous with j^erithecium ; cells IV and V forming a somewhat angular
protrusion which carries the insertion-cell out free from and beyond the
tip of the perithecium. Insertion-cell less than half as broad as the ad-
jacent distal margins of cells IV and V, which form a flat surface in which
the insertion-cell is mostly central. Outer appendage arising from a
small roundish basal cell, simple or once branched, the branches short,
tapering nearly hyaline ; the basal cell of the inner appendage very small
bearing one or two short tapering hyaline branches. Perithecia 95-
110 X 34 ju. Total length to tip of perithecium 280 ^u; to insertion-cell
about the same measurement; greatest width 66 /i.

On Pericallus cicindeJoides MacLeary, Paris Museum, No. 144:, Ton-
gou, Java. On mid-elytron.

Laboulbenia Pseudomasci nov. sp.

Perithecia dark rich brown, slightly and rather evenly inflated, diver-
gent, free from the receptacle except at the very base, the broad opaque
tip hardly differentiated ; the lip-edges nearly hyaline, not prominent,
turned slightly outward. Receptacle colorless or yellowish below, dis-
tally brownish, slender, its axis coincident with that of the appendages;
cell I usually larger and longer than cell II, the rest of the receptacle
relatively unusually small, the inner margin of cell V partly free from
the perithecium. Insertion-cell free, black. Outer appendage con-
sisting of a basal cell several times longer than it is broad, which may
branch above its basal or subbasal cell one to three times successively,
the branchlets divergent ; the inner appendage consisting of a much
shorter basal cell bearing one or rarely two brandies, sometimes simple,
mostly one to three times branched, all the branches becoming more or

THAXTER. NEW LABOULBENIACE^. 197

less sufTused with brown. Spores about 65 X 45 /x. Perithecia 120-165
X QO fi. Total length to tip of perithecium 225-300 //, ; to insertion-cell
170-240 ^w. Appendages, longest 100 /a.

On Pseudomascus nigrita, Fab., Paris Museum, No. 201, Mongolia.
Near upper inferior margin of prothorax on left side.

Laboulbenia punctata nov. sp.

Perithecium free, straight, translucent brown becoming almost opaque,
except the broad short neck formed by the basal wall-cells, which is
nearly hyaline and as broad as the ascigerous portion ; the lower half or
more of the suffused body of the perithecium covered with irregular
more or less rounded dark spots, irregularly distributed, the lower
larger; the tip rather abruptly distinguished, narrow, black, distally
translucent. Receptacle rather short and stout, the basal cell rather
narrow and hyaline or yellowish, contrasting; the rest of the recep-
tacle subtriangular and deeply suffused ; cell VI paler, cells III and IV
side by side, nearly vertical, almost opaque, except the upper edge ;
cell V very large, subhemispherical, becoming opaque ; all the suffused
cells where not opaque, more or less conspicuously and rather coarsely
punctate. Insertion-cell very broad, black, close beside the base of the
perithecial stalk. Outer appendage consisting of a large triangular basal
cell externally blackish brown, forming the base of a series of (eight
or less) much smaller cells obliquely superposed, which curves toward
the perithecium ; each cell producing externally a single simple erect
branch, rather closely septate, the (usually six) septa dark, constric-ted,
the terminal cell short with rounded apex. The inner appendage con-
sisting of a basal cell giving rise to a series of cells on either side like
that of the outer appendage, but shorter, one to three of the lower
branches consisting of a single cell bearing terminally a long slender
flask-shaped brown antheridium. Perithecia 200-220 X 40 ^u, ; smaller
130 X 48 ^ (including the neck about 35 /x). Total length to tip of
perithecium, average 350 /x ; to insertion-cell 145 ^u; greatest width 75 /x.
Appendages 1 10-1 30 /x.

On Galerita sp., Paris Museum, No. 74, Venezuela, No. 136, " South
America." On head.

Laboulbenia punctulata nov. sp.
Perithecium about three fourths free, dark brown translucent, curved
toward the appendages which cross it obliquely, the broad short fiat-

198 PROCEEDINGS OP THE AMERICAN ACADEMY.

topped snout-like tip slightly upcurved. Receptacle short and stout,
the basal cell small, short, hyaline, contrasting, the rest concolorous
with the perithecium, but darker and distinctly punctate with dark
brown spots. Outer appendage consisting of a series of from three to
about six successively smaller superposed cells, from each of which
a simple tapering browQ branch arises, blackened about its subbasal
septum, the successive branches superposed in a vertical external row
as in L. Pacliytelis, the basal cell of the inner appendage producing
usually a short one-celled antheridial branch. Perithecia 120 X 45 ^u.
Total length to tip of perithecium 200-220 /;x; to insertion-cell 125 /a.
Appendages 100-120 /x,.

On Pacliyteles parallelus, Chaud., Brit. Mus. No. 575, Para: on P. por-
rectus Chaud., Brit. Mus. No. 670 (Biologia Coll.), Pantaleon, Guate-
mala. On legs.

Laboulbenia pygmeea nov. sp.

Perithecium dark brown becoming almost opaque, coarsely punctate
throughout or only toward the base, the basal wall-cells forming a well
defined hyaline contrasting short neck slightly narrower than the body
of the perithecium, the tip usually not very abruptly distinguished and
bent very slightly outward, or straight, rather blunt, the lip-edges trans-
lucent, the lip-cells blackened below, especially on the inner side.
Receptacle very short and subtriangular, cell I short, slender, curved,
hyaline at the base, distally becoming opaque blackish and indistinguish-
able from cell II, which is wholly opaque. Cells III and IV elongated
and lying obliquely side by side, cell III forming a more or less prominent
rounded projection a little below the insertion-cell, both cells becoming
opaque ; cell V rather large, at first hyaline, becoming later suffused
with brown ; all the other suffused parts rather coarsely punctate. In-
sertion-cell black, very broad, often becoming indistinguishable from the
basal cells of the appendage. Outer appendage consisting of a sub-
triangular basal cell distally rounded, becoming deeply suffused with
blackish brown, prominent externally ; surmounted by a series of
obliquely superposed cells close set, their long (transverse) axes some-
times almost perpendicular, each bearing externally a single simple
branch, the two lower cells of which are longer than broad, tinged with
brown, the septa dark and often oblique ; the distal portion hyaline, twice
as long, blunt-tipped: the inner appendage consisting of a smaller basal
cell also becoming almost entirely suffused, surmounted on either side by
a series of cells like that of the outer appendage and similarly branched,

THAXTER. — NEW LABOULBENIACE/E. 199

except that the two or three lowest cells of the series bear a short oue-
celled branch termiuated by usually three slightly curved brown anthe-
ridia. Spores 52 X 4 /x. Perithecia 110 X 22-150 X 33 yu,, exclusive
of neck which is 20-30 fx. Total length to tip of perithecium 175-300 /a;
to insertion-cell about 90-110 /x ; greatest width 40-55 ^. Appendages
90-130 IX.

On Trichognathus sp., Paris Museum, No. 72, Venezuela. On " T.
margivdtns" Brit. Mus. No. 526, Brazil; on T. marginipetinis hatr.,
Brit. Mus. No. 525, Tamaz, S. America; on Galerlta occidentalis Oliv.,
Brit. Mus. No. 515, Bolivia; on Galerita sp., Hope Coll., No. 258,
Bahia, Brazil. On all parts of host.

Laboulbenia rliinophora nov. sp.

Perithecium large and stout, dirty smoky brown, the lower half much
deeper and united to the receptacle, the base nearly opaque, one of the
subtermiual wall-cells forming a terminal blunt finger-like brown out-
growth, close beside the rather small internally blackened tip, which it
exceeds in length. Receptacle stout, the basal cell small hyaline ; a
blackish brown suifusion becoming opaque involves the upper part of
cell II, cells III, YI, and VII, as well as the basal cells of the perithe-
cium; cells IV and V very large and nearly parallel, translucent ; the
suffused parts, when not opaque, marked by darker transverse dots and
strise. Insertion-cell very large, triangular, quite unmodified. Append-
ages consisting of two basal cells concolorous with insertion-cell, the
outer usually somewhat larger, both protruding upward and slightly over-
lapping, producing directly numerous branches (four to eight from each
cell) which arise in more than one row from their outer surfaces; all the
branches once to twice branched, the lower segments deeply constricted
at the purplish septa, the distal cells without constrictions at the hyaline
septa. Spores 75 x 5 /x. Perithecia to tip of protuberance 275-300 X
85 ju; to insertion-cell 300-340 jx. Appendages about 200 ^/.

On Brach'mus sp., Hope Coll. No. 252, Madagascar. On legs.

Laboulbenia rostellata nov. sp.

Perithecium about one half free, becoming more or less deeply suffused
with blackish brown, relatively small, narrow and curved toward the
appendages, the tip monstrously developed, bulging terminally and
externally to form a rounded prominence from the inner side of which
a blunt blackened outgrowth is developed, the hyaline contrasting tip of

200 PROCEEDINGS OF THE AMERICAN ACADEMY.

which is bent slightly upward. Receptacle more or less suffused, some-
times deeply colored with blackish brown, cells I and II paler, slender,
of nearly equal width, cell II longer ; cells IV and V very broad, carrying
out the insertion-cell free from the receptacle so that it becomes oblique
or even vertical and lateral. Outer and inner appendages similar, bent
away from the perithecium, their bases overlapping ; consisting of a series
of superposed cells which are sucessively smaller from below up, each
bearing distally and externally a short tapering branch ; all the branches
brown, the subbasal septa blackened, all simple except the lower branch
of the outer appendage which bears two to three short branchlets ; the
basal cell of the inner appendage bears a short antheridial branch from
its inner side ; the outer appendage somewhat longer than the inner, the
superposed cells usually eight in number. Perithecia 140-100 X 40-45 ^.
Total length to tip of perithecium 400-550 /a ; to insertion-cell 270-
450 fx. Appendages, 140-170 fx. Tip of perithecium, including out-
growth, 50-60 /A.

On Brachinus lateralis Dej., Hope Coll. No. 246, "North America";
on Brachinus sp., Eustis, Florida, October. At base of anterior legs.

Laboulbenia separata nov. sp.

Perithecium pale olivaceous, the inner margin convex, the outer nearly
straight; the tip rather abruptly distinguished, blackened, but not uni-
formly, below the inner lip-edges, prominent olivaceous translucent, the
right inner lip prolonged to form a slender nearly hyaline projection, the
tip of which is blunt and somewhat swollen. Receptacle dull olivaceous,
cells II, III, and IV sometimes becoming blackish brown externally,
cells III and IV rather large ; the insertion-cell close to the base of the
perithecium, half as broad as cells IV and V. Appendages much as in
L.jissa^ the outer curved strongly outward, opaque and indistinguishable
from the insertion-eel], bearing three or four branches from its convex
side which are mostly once branched (the outer appendage usually
broken) ; the inner appendage consisting of a small basal cell, bearing a
branch on either side externally blackened, somewhat curved outward,
and bearing three or four branchlets which are curved toward the peri-
thecium, externally or wholly brownish toward the base. Perithecia
100-110 X 25-30 iJL. Total length to tip of perithecium 220-260/^ ; to
insertion-cell 130-165 fx. ; greatest width 55 /x. The prolongation of the
lip-cell extending about 20 [i or more beyond the tip of the perithecium.

On Pericallus guttatus Chev., Brit. Mus. No. 571, Java. Margins of
elytra.

THAXTER. — NEW LABOULBENIACE^. 201

Laboulbenia Serrimarginis nov. sp.

Perithecium about three fourths free, dull olive-brown, the outer
murtfiu nearly straight, the inner bent rather abruptly below the tip ;
the latter not abruptly distinguished, broad, blunt, the lips not well de-
fined, dull blackish. Receptacle dull olive-brown, except the almost hya-
line rather slender basal cell ; the rest of the receptacle short and stout,
concolorous with the perithecium, cell VI broad and extending down to
cell I beside cell II, cell VII extending half way down cell VI exter-
nally, the septa of cells III and IV oblique; cell III extending up to the
base of cell V. Appendages as in L. maculata, stouter, brown, darker or
blackish toward the base. Perithecia 136 X 50 /x. Total length to tip
of perithecium about 300 ^i; to insertion-cell 190 /x; greatest width 85 //.
Appendages 260^.

On Serrimargo giittiger Schaum., Brit. Mus. No. 558, Penang, East
Indies. At base of anterior legs.

Laboulbenia speciosa nov. sp.

Perithecium free, long and narrow, the inner half or less hyaline, the
outer dark clear blackish brown ; the basal wall-cells forming a short
hyaline stalk narrower than the ascigerous portion ; the tip hardly dis-
tinguished, bluntly rounded, sliglitly oblique outwardly, black, hyaline
about the pore. Receptacle very elongate, hyaline except cells IV and
V which are tinged with amber-color and indistinctly punctate, as is cell
III; cell II very elongate. Insertion-cell black, just free from the stalk
of the perithecium through the slight enlargement of cell V. Outer
appendage consisting of a series (in the unique type) of eight cells super-
posed not very obliquely, the basal one larger tinged with brown below,
the rest hyaline ; all the cells producing externally a single simple
branch, the lower three cells of which are mostly not longer than broad,
constricted at the black septa, distally faintly brownish yellow ; the
inner appendage consisting of a basal cell which bears on either side a
series of four cells like that of the outer appendage, and bearing similar
branches in a similar fashion, the three series quite distinct from one
another. Perithecium, including its short neck, 280 X 55 ^. Total
length to tip of perithecium 925 ji<; to insertion-cell 650^. Append-
ages about 200 (-250) ju.

On Galerila unicolor Dej., Brit. Mus. No. 517, Brazil. On inferior
surface of the prothorax.

202 PROCEEDINGS OF THE AMERICAN ACADEMY.

Laboulbenia spiralis nov. sp.

Perithecium one half to one third (sometimes only the tip), free
from the receptacle, dark dull amber-brown with dirty brown suffusions,
rather stout, the tip moderately well distiui>uished, rather short and
stout, deep black-brown, except the distal hyaline lip-edges. Distal por-
tion of the receptacle concolorous with the perithecium; cells I and II
much paler ; cell I longer than cell II, the two forming a rather slender
stalk of about the same diameter throughout, above which the distal
portion of the receptacle is somewhat abruptly distinguished ; the lighter
portions marked by fine transverse striations not visible in the deeply
colored often opaque distal region. Outer appendage consisting of a
main straight divergent axis formed by usually three nearly equal cells,
deeply blackened externally, bearing distally and from each of their
upper inner angles a very long slender ei-ect simple branch which is
reddish brown, paler and spirally twisted distally. The inner append-
age consisting of a basal cell about as large as that of the outer
appendage and bearing on either side a branch which may give rise to
one or two erect simple branchlets similar to the branches of the outer
appendage. Perithecia about 150 X 55 ^i. Total length to tip of peri-
thecium 300-390 fj, ; to insertion-cell 275-325 fi. Appendages, longest
480-GlO fi.

On Hexagonia sp. ?, Hope Coll., No. 288, Ceylon (Thwaites).

Laboulbenia strangulata nov. sp.

Perithecium one third to one half free, dark brown, concolorous below
with the distally almost opaque recejitacle ; symmetrical, straight ; the
tip undifferentiated, bluntly rounded except for a hyaline flattish ter-
minal 2^apilla formed by the projection of one of the external lip-cells.
Receptacle slender, cell I usually basally curved, broader at the distal
end where it is rather deeply suffused with brown ; cell II much longer,
its lower two thirds often distinctly inflated, deep brown distally, rather
abruptly constricted to less than half its greatest diameter, the constricted
portion hyaline, the short remaining portion above the constriction
becoming deep brown, concolorous with the upper portion of the recep-
tacle. Insertion-cell normal as in L. Orectochili^ oblique, often concave
above, the appendages consisting of an outer and an inner basal cell, the
two producing in all from three to six outgrowths as in L. Orectochili,
somewhat narrower, hyaline except the first one formed from the outer
basal cell which is always external to those subsequently formed and is

THAXTER. — NEW LABOULBENIACEiE. 203

often divergent, deep brown, the suffusion involving the basal cell which
bears it : each branch consists of a single simple cylindrical cell the distal
portion of which is usually broken off leaving the deep brown contrast-
ing base. Spores 60 X 4.5 /x.. Periihecium 120-140 X 40-45 ^u ; to
insertion-cell, average 275^. Appendages 100 /x.

On Orectochilus ?, Brit. Mus. Nos. 480 and 484, Timor, East Indies.
Margin of elytra.

Labotilbenia subconstricta nov. sp.

Perithecium less than one half free, rather small, curved strongly and
evenly toward the appendages, evenly suffused with dark brown or lighter
distally; the tip black, not abruptly distinguished, the lips distinct with a
broad hyaline margin about the pore. Receptacle dull amber-yellow,
cells IV and V, sometimes cell III, less deeply suffused with brown;
cell I narrow, straight, but slightly enlarged distally; cell II abruptly
much larger, symmetrically and prominently constricted in the middle ;
cells III and IV large and broad. Outer appendage simple, the basal
cell moderately large ; externally or wholly blackened ; the rest of the
appendage straight, rigid, directed across the tip of the perithecium :
inner appendage consisting of a smaller basal cell which bears one or two
branches similar to the outer appendage. Perithecia 85 X 25 ^. Total
length to tip of perithecium about 200 /a ; to insertion-cell 165 /x ; greatest
width 50 jtt. Appendages, broken, about 150 ^u or a little more.

On Catoscopus sp., Paris Museum, No. 116, New Guinea. On anterior
inferior margin of thorax on the rioht side.

Laboulbenia Sumatrae nov. sp.

Perithecium small, pale amber-colored, darker and somewhat inflated
below, tapering distally ; the tip only free from the receptacle not distin-
guished from the body of the perithecium, long, nearly hyaline below,
the lip-cells abruptly spreading, contrasting, black except about the pore
and peculiarly modified, the two outer broadly rounded, the two inner
developing outgrowths which grow inward and upward ; that on the
right side longer than the left, narrower and indented near the base on
the inner side. Receptacle ooncolorous with the perithecium, paler
below, marked by faint fine tranverse striations. Appendages concolor-
ous with the receptacle, the outer simple, stiff, curved slightly outward,
the inner consisting of a basal cell smaller than that of the outer and
bearing ou either side a branch which may be once branched at the base.

204 PEOCEEDINGS OF THE AMERICAN ACADEMY.

Perithecium 92 X 22 ^u. Total length to tip of perithecium 200 //, ; to
insertion-cell 166 jjl. Appendages about 185 jx.

On '' Catoacopus cupi-ipennis Thorn.," Plope Collection, No 291,
Borneo No. 299, Sumatra. At base of anterior legs.

Laboulbenia Taenodemse nov. sp.

Perithecium nearly free, slightly inflated, clear translucent brown;
the tip abruptly differentiated, extei'nally black with an inner inferior
hyaline patch, the lips well distinguished hyaline, the inner prominent,
rounded, the pore external. Receptacle long and slender through the
elongation of cell II, pale dirty yellowish brown or nearly hyaline.
Insertion-cell large, free, blackish, hardly narrower than cells IV and V.
Appendages concolorous with the receptacle, the outer consisting of a
large basal cell which bears distally as a rule two antero-posterior
branches ; the inner simple, the outer furcate above its basal cell ; the
branches stout, slightly tapering, elongate, the inner erect, the outer
divergent ; its branclilets curving upward : the inner appendage con-
sistins: of a much smaller basal cell, rounded and bearing; one or two short
branchlets. Spores about 70 X b jx. Perithecia 12.5-155 X 35-50 /x.
Total length to tip of perithecium 400-650 jx ; to insertion-cell 300-
400 /x. Appendages, longest about 600 p..

On Tcenodema sp., Brit. Mus. No. 391. Ega, Amazon. On elytra and
superior prothorax.

Laboulbenia tenuis nov. sp.

Perithecium relatively small and narrow, dull amber-yellow to brown,
the upper half or more free from the receptacle, tapering somewhat
distally, curved toward and partly across the appendages ; the tip not
well distinguished, broad, its distal margin often concave, the lips
projecting slightly on either side, the lip-cells black except about the
pore. Receptacle slender amber-yellow, becoming tinged with brown
especially distally, strongly curved throughout, the concave side anterior.
Insertion-cell black and thick, narrower than cells IV- V. Appendages
as in L. plat>/stoma, the basal cell of the outer appendage blackened
externally, all the branches erect and lying across the tip of the perithe-
cium which is bent toward them. The material on Cdtoscopus much
larger and darker than tliat on Mlscelus. Perithecia 90-130 X 22-26 u.
Total length to tip of perithecium 250-500 /^ ; to insertion-cell 425-185^.
Greatest width 37-55 /a. Appendages more than 200 ^ (broken).

THAXTER. NEW LABOULBENIACE^. 205

On Miseelus Javaniis King., Hope Collection, No. 308, Java ; on
Miscelus sp., Paris Museum, Nos. 114 and 115, New Guinea; on
Catoscopus ? sp. Brit. Mus. No. 663, Assam, India. On the elytra and
inferior surface.

Laboulbenia Thyreopteri nov. sp.

Perithecium nearly free, proportionately large, dull amber-brown,
straiulit, narrower at the base, the inner margin slightly convex, the outer
concave through the presence of a prominent subterminal hump, which
is suffused with blackish brown, the suffusion often involving a fainter
discoloration of the subbasal wall-cell below it ; the tip small, prominent,
and abruptly differentiated, blackish with broad hyaline lips. Receptacle
slender, the basal cell black, opaque, mostly curved below, very slender ;
the subbasal cell broader, suffused with blackish, mostly verrucose or
coarsely punctate, the remaining cells normal and concolorous with the
perithecium. Insertion-cell thick and narrow. The outer appendage
simple, its basal cell long, undifferentiated ; the basal cell of the inner
appendage shorter bearing a branch distally on either side, all the
branches similar crowded, concolorous with the perithecium, erect, straight
or bent slightly toward the perithecium, the inmost in contact with it.
Spores 55 X 4|U. Perithecia 140-175 X 14-25 ^u. Total length to tip
of perithecium 340-400 yu, ; to insertion-cell 250-270 /x. Appendages
120-140/..

On Tlujreopterus flavosignatus Dej., Brit. Mus. No. 561, Port Natal,
Africa. On Thyreopterus sp., Paris Museum, No. 125, Africa. On
elytra.

LaboTilbenia tibialis nov. sp.

Perithecium deeply suffused with blackish brown, somewhat inflated,
the small tip rather abruptly distinguished. Receptacle stout, deeply
suffused with blackish brown except cell I and the lower part of cell II,
which are hyaline or nearly so, abruptly contrasting, and cell V which
is yellowish ; cells IV and V broad, the insertion-cell broad and in con-
tact with the base of the i^erithecium. Appendages as in L. rostellata
except that the iimer is larger and longer than the outer without over-
lapping it at the base and the lower branch of the outer is simple.
Perithecia 150-175 x 60-70//. Total length to tip of perithecium
300-325 fx ; to insertion-cell 200-2-25 ^. Appendages exclusive of the
branches, inner 100-120//, outer 85 /i.

On Brachinus sp., Eustis, Florida, October. On the legs.

206 PROCEEDINGS OF THE AMERICAN ACADEMY.

Laboulbenia tortuosa nov. sp.

Peritbecium with hardly more than the tip free, relatively small, ex-
ternally suffused with smoky brown and concave through the presence of
a well defined snbterraiiial hunch, above which the somewhat pointed
well defined outwardly oblique blackish-brown tip is abruptly differen-
tiated, the lip-edges pale brown translucent. Receptacle very pale red-
dish or yellowish, variously bent, sometimes at right angles or at an
angle of forty-five degrees above cell II ; cells I and II straight or more
frequently cell II curved strongly, while there is often a less pronounced
curvature in the opposite direction immediately above it. Insertion-cell
horizontal, about opposite the external hunch of the perithecium. Outer
appendage consisting of a very large basal and somewhat broader sub-
basal cell, the two commonly as broad as or broader than any portion of
the receptacle with which they are concolorous ; the upper outer angle
of both cells marked by the black insertion of a short simple branch, usu-
ally broken off, the subbasal cell surmounted by a small flattish cell which
bears a simple terminal branch with blackened base like those developed
laterally below it : the inner appendage consists of a very small basal
cell which usually produces directly a pair of relatively large antheridia
with inflated venters and brown necks. Perithecia 85 X 27 ja. Total
length to tip of perithecium about 275 /x. Appendage, to upper blackened
septum 50 /n, by 28 /x wide.

On Pa chyteUs festaceus Horn, U. S. National Museum, Arizona. Along
the adjacent inferior margins of the thorax and jDrothorax, on left side.

Laboulbenia Trichognathi nov. sp.

Perithecium free, generally straight, long, narrow and of nearly equal
diameter throughout to the base of the rather abruptly differentiated tip,
(sometimes however shorter, stout and slightly inflated,) pale yellowish
or becoming rather deep, evenly translucent smoky brown ; the basal wall-
cells forming a very short scarcely noticeable stalk ; the tip obliquely black
below the rather coarse and prominent hyaline lips. Receptacle gener-
ally very long and slender, pale yellowish, the basal cell tinged with
brown ; cells IV and V amber or often becoming wholly amber-brown or
smoky brown ; cells IV and V large, prominently marked by short
transverse lines or patches which are less numerous and distinct on the
other cells. Insertion-cell carried out free from the perithecium by the
enlargement of cell V, well differentiated, black. Outer appendage con-
sisting of a series of obliquely superposed cells three to ten in number.

THAXTER. — NEW LAB0ULBENIACEJ3. 207

the basal one subtriangular and blackish, the rest hyaline or yellowish,
each producing distally and externally a single simple straight branch,
slightly constricted at the three to four lower black septa ; the distal por-
tion without black septa, sometimes shorty sometimes elongate and tapering.
The inner appendage consists of a basal cell usually giving rise on either
side to a series of superposed cells similar to those of the outer append-
age, but mostly shorter and similarly branched ; the three series in general
partly united at least at the base, the basal cell in some cases bearing
more than two such series so that a very dense tuft of branchlets results.
Peritliecia IGo X 55-350 X 90 /li, average 275 X QO fi. Receptacle very
variable. Total length to tip of perithecium 425-1480 /x ; to insertion-cell
275-1260^. Appendages, longest 600 /:i, average about 400 /x.

On Trichognnthus marginipennis Latr., Hope Collection, No. 267,
Columbia, Brit. Mus. No. 525, "Tamaz," S. America ; on " T. margi//atus
Latr." Brit. Mus. No. 526, Brazil, Hope Collection, No. 266, " S.
America"; on THchognathns sp. indet., Paris Museum, Nos. 70, 71, and
\'6b, Venezuela and " S. America." On all parts of host.

Laboulbenia triordinata nov. sp.

Perithecium usually wholly free from the receptacle, very variable,
amber-brown or usually becoming almost black, generally elongate, often
with the wall-cells showing a distinct spiral twist, the basal wall-cells
forming a hyaline or less deeply suffused neck ; the tip wholly black
except the lip-edges, which may be translucent and more or less distinctly
differentiated. Receptacle elongate or rather short, pale amber-brown or
amber-yellow, gradually tapering from the base to the very broad distal
portion, or more commonly cells I and H forming a slender stalk above
which the rest of the receptacle expands abruptly, the unmodified tri-
angular insertion-cell carried up and out free from the receptacle through
the enlargement of cell V, the inner margin of which is mostly free.
Outer appendage consisting of a series of about seven to ten obliquely
superposed cells, each bearing distally and externally a single simple
branch ; the branches constricted at the first, second, and third septa,
which are deeply blackened ; the second less broadly, the first and second
cells variably suffused with clear brown especially near the septa, the
distal portion of the branch hyaline, tapering, its cells longer, the lower
swollen below the septum : inner appendage consisting of a basal cell
which gives rise on either side to a branch resembling the outer append-
age often with fewer cells (sometimes only 3-celled) similarly branched,

208 PROCEEDINGS OP THE AMERICAN ACADEMY,

except that the lower branchlet of each branch bears a single large
terminal antheridium which becomes brown ; the outer appendage and
the two branches of the inner erect and close together or more or less
strongly and irregularly divergent. Perithecia, average 200 X 52/i
(110-260 X 45-60 /i) including stalk. Total length to tip of perithe-
cium 260-600 /i; to insertion-cell 185-370 /* ; greatest width 65-100 /i.
Appendages 220-330 fi.

On Calojjhoena hifasciata Oliv., Brit. Mus. No. 509, South America ;
on CalojilKjena sp., Brit. Mus. No. 512, Nanta, Amazon; on Cordistes
bicinctus Dej., Hope Collection, No. 268, Columbia; on Cordistes? sp.,
U. S. National Museum, Central America ; on Helluomorpha melanarla
Reich., Brit. Mus. No. 527, Ega, Amazon.

Laboulbenia tuberculifera nov. sp.

Perithecium deeply suffused with smoky brown, free except the lower
fifth ; the tip not abruptly distinguished, nearly black, the distal margin
somewliat oblique, mostly straight with an outer more or less ill defined
tooth-like prominence ; in general variable, the lip-cells not prominent.
Receptacle leather long and slender, cell I quite hyaline below, distally
tinged with deep brown and coarsely punctate ; cell II tinged with
brown and punctate below, otherwise nearly hyaline, except for the
presence of a series of deep brown short tubercular transverse ridges on
one side of cell IV and the basal cells of the perithecium which are deep
brown, the rest subhyaline, cell IV bulging. Outer appendage simple,
slender, the basal cell sometimes long and somewhat inflated : the inner
appendage consisting of a small basal cell with a short branch on either
side. Perithecia about 140 X 50 /x. Total length to tip of perithe-
cium, average 340 fx. ; to insertion-cell, average 225 fi ; greatest width 60/*.
Appendages 225^ (longest).

On Serrimargo guttiger Schaum., Brit. Mus. No. 558, Penang, East
Indies. On base of elytra.

Laboulbenia uncinata nov. sp.

Pale yellowish, becoming tinged with pale reddish yellow. Perithe-
cium more than one half free, stout, nearly oblong, the blackened lip-cells
recurved externally through the rounded upgrowth of the inner distal
portion of the perithecium which makes them almost lateral in position,
the outer lip-cell often twice as long as the others and more prominently
recurved, its projecting portion translucent, rounded. Receptacle

THAXTER. — NEW LABOULBENIACE^. 209

medium, normal. Appendages normal, rather short, several times
brauched, the cells rather short and somewhat rounded. Spores
50 X 3.0 fi. Perithecium (not including lip-cells) 120 X 62 fi. Total
length to tip of perithecium 100-120 /a; to insertion-cell 72-92 /x.
Appendages (longest) 40 fx.

On Harpahis ceneus Fabr., Selenga, Siberia, Paris IMus. No. 12. At
base of anterior legs.

Laboulbenia verrucosa nov. sp.

Perithecium becoming deeply suffused with smoky brown, straight, the
line of demarcatiou between the subterminal and subbasal wall-cells indi-
cated by a more or less well defined ridge forming a rather prominent
external hunch in this region, above which the perithecium is abruptly
contracted, almost at right angles iu the type, below the rather narrow
nearly erect tip, the lip-cells black below, with the broadly hyaline edges
turned obliquely outward. Receptacle dirty yellow-brown, becoming
more or less suffused with smoky brown, especially the two basal cells,
and covered with irregular w^art-like prominences which are more or less
definitely arranged in transverse rows. Appendages of the '' L.Jlagellata"
type, the outer once to twice branched, the inner consisting of a smaller
basal cell giving rise on either side to single branches which may be from
once to three times branched ; all the branches pale dirty yellowish with
brown shades above the lower septa. Perithecia 150-170 X 45-50 /li.
Total length to tip of perithecium 550-610 /x; to insertion-cell 430-480 ji.
Appendages (longest) 400 fx.

On a Carabid allied to Platt/nus, Hope Coll. No. 342 (without label)
and U. S. Nat. Museum, No. 7, Mt. Coffee, Liberia, Africa. Elytra.

VOL. XXXV. — 14

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 10. — December, 1899.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF
HARVARD COLLEGE.

FERROUS IODIDE.

By C. Loring Jackson and 1. H. Derby.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.

FERROUS IODIDE.

By C. Loring Jackson and I. H. Derby.

Presented May 10, 1899. Received November 24, 1899.

The accounts of anhydrous ferrous iodide which we have been able to
find in the chemical literature are very contradictory. Serullas * seems
to have been the first to obtain it by the rather strange method of passing
iodine and steam through a red-hot iron tube filled with charcoal. He
describes it as shining yellow crystals looking like gold filings. Gay
Lussac and Davy * say it is a brown mass which fuses at a red heat and
sublimes at a higher temperature. A. T. Thomson f treated one part
of iron with two of iodine suspended in water, and after evaporating off
the water in a flask containing some free iron, obtained a steel gray lam-
inated mass melting at 177° and possessing a sharp puckery taste. When
heated in air it left a residue of ferric oxide. In 1861 Carius and Wank-
lyn X prepared it as a gray laminated mass by heating iron filings and
iodine in a porcelain crucible. In 1863 De Luca § states that Faville
had made it under his direction, and that it is a pure white amorphous
powder. He gives no statement of the method of preparation in any of
the journals which are accessible to us. Finally Erdmann || in his Lehr-
buch (1898) describes it as a reddish brown body, also without giving
the method of preparation. The color of ferrous iodide, therefore, is not
by any means settled, the balance of evidence in the chemical literature
being in favor of gray or white ; but as no analyses of the anhydrous
salt have been published, so far as we can find, none of the statements
quoted above can be considered to rest on a solid foundation. Under

* Gmelin-Kraut, Handbuch, III. 350, 6th edition.

t Pharmaceut. .Journ., I. 44 (1842).

t Ann. Chem., CXX. 69.

§ Comptes-Rendus, LV. 615.

II Lehrbuch der anorganischen Chemie, 62.

214 PROCEEDINGS OF THE AMERICAN ACADEMY.

these circumstances it seemed worth while to take up the study of the fer-
rous iodide, and the results of our work in this field are contained in the
following paper.

Preparation op Ferrous Iodide.

The method adopted consisted in passing the vapor of iodine over
heated iron. In the first experiments a combustion tube was used with
a stream of carbonic dioxide to carry the iodine, but as we found a tem-
perature above the softening point of hard glass was necessary, we re-
placed the glass tube with one of iron, which we heated with a Fletcher's
tube furnace. At the high temperature thus obtained the iron was oxi-
dized by the carbonic dioxide, and the film of oxide thus formed protected
the metal from the action of the iodine ; accordingly the carbonic dioxide
was replaced by nitrogen, after which good results were obtained.

The apparatus in its final form consisted of two large bottles, which
served as gas holders for the nitrogen, and were united by glass tubes and
rubber connectors in such a way that, by letting in water from the water
works, the gas could be forced out of one of them at a regulated rate
through the purifiers and the iron tube to the other bottle, in which it
was collected. When this was full of gas, by a simple rearrangement
of the pinch-cocks the nitrogen could be forced back through the appara-
tus again in the same direction as before into the first bottle. By means
of this device the same nitrogen could be used over and over again, thus
causing a great saving in time and labor. The nitrogen was made by
the excellent method of Dr. Gibbs,* — equal parts of sodic nitrite, am-
monic nitrate, and potassic dichromate were pulverized, and mixed with
three parts of water in a flask with a wide delivery tube. The gas came
off when the mixture was gently warmed. Twenty-four grams of the
mixture gave about 1200 cc. of nitrogen. The nitrogen, after leaving
the gas holder, was purified by passing through a solution of sodic hydrate,
sulphuric acid, and a long column of calcic chloride, after which the oxy-
gen was removed by means of a layer of finely divided copper heated in
a tube of hard glass by means of a kerosene stove. This precaution was
necessary, as the nitrogen was contaminated with air introduced when
the tubes were opened, or from the water in the gas holders. We tried
at first an alkaline solution of pyrogallol for the removal of the oxygen,
but as we found this was not efficient, we resorted to the layer of finely
divided copper, which was prepared by reducing cupric oxide.

* Ber. d. chem. Ges., 1877, p. 1387.

JACKSON AND DERBY. — FERROUS IODIDE. 215

The reaction took place in three steel bicycle tubes, the largest of
which was one inch (2.54 cm.) in outside diameter and about one six-
teenth of an inch (1.59 mm.) thick. The other tubes were somewhat
thinner and small enough to slip tightly into the larger tube. The larger
tube contained the iron, which was used in the form of wire clippings
(card teeth), as it was found by experiment that these gave a better yield
of iodide than turnings, filings, or iron by hydrogen. The iron was freed
from organic matter by heating it in a stream of hydrogen before using
it. The layer of iron in the tube was about 23 cm. long, and this por-
tion of the tube was heated directly by the Fletcher furnace to a bright
red heat. One of the smaller tubes was slipped into the larger one at its
end nearest the gas holder containing the nitrogen, in this was placed the
iodine, which was slowly volatilized by the heat conducted along the iron
tube, and the vapor thus formed was swept over the hot iron in the fur-
nace by the stream of nitrogen ; this tube was connected with the purifiers
by a perforated cork and glass tube far enough from the heat to avoid
danger of burning. It was found wise to protect this cork from the
iodine vapor, some of which crept back even against the current, by an
asbestos or iron diaphragm. The second smaller iron tube was slipped
into the further end of the larger tube, and reached the heated portion of
it, so that the vapor of ferrous iodide formed in the larger tube condensed
in the smaller one, from which it could be easily collected by withdraw-
ing this smaller tube after it had cooled in the atmosphere of nitrogen.
The joints of the iron tubes were rendered air tight by means of large
rubber connectors, which were kept cool by winding them with lamp-
wicking that dipped into a beaker of water. The receiving tube just
described was connected with some bottles of strong sulphuric acid to
prevent the creeping back of any aqueous vapor into the tube. The
delivery tube from these bottles conveyed the nitrogen to one of the gas
holders already described.

In carrying on the process it was observed that a considerable amount
of iodine was taken up by the iron before any iodide distilled over ; but
after it began to appear, it continued to collect for several hours, so that
the yield was a good one, and when the apparatus has been once set up
the ferrous iodide can be obtained in quantity with little trouble.

As the product of the reaction just described had a deep red color in
thin plates, and looked almost black when the mass was even moderately
thick, we supposed at first that it could not be a ferrous salt, and accord-
ingly applied the following tests. A little of it treated with water, in
which it dissolved completely and with great ease, gave no color with a

216 PROCEEDINGS OF THE AMERICAN ACADEMY.

solution of potassic sulphocyanate, but a strong test for a ferrous salt
with potassic ferricyanide. As it was possible that it might be a ferric
iodide which was decomposed by water into the ferrous salt and free
iodine, we next tested for iodine both with starch paste and with carbonic
disulphide, but with negative results. These qualitative experiments
indicated that the substance, in spite of its red color, was ferrous iodide ;
and that this is the case was proved by the following analyses. The
substances used in I.-V. and in IX. and X. were the direct product of
the reaction. In VI., VII., and VIII. this product had been purified
by sublimation before it was analyzed.

I. 0.6717 gram of substance gave 1.0202 gram of argentic iodide
and 0.1720 gram of oxide of iron.
II. 1.0400 gram of substance gave 1.5736 gram of argentic iodide.

III. 0.9658 gram of substance gave 1.4601 gram of argentic iodide.

IV. 0.7493 gram of substance gave 1.1471 gram of argentic iodide

and 0.1934 gram of ferric oxide.
V. 0.8327 gram of substance gave 1.2586 gram of argentic iodide
and 0.2209 gram of ferric oxide.
VI. 1.2675 gram of substance gave 1.9232 gram of argentic iodide
and 0.3261 gram of ferric oxide.
VII. 0.7396 gram of substance gave 1.1192 gram of argentic iodide

and 0.1932 gram of ferric oxide.
VIII. 0.5944 gram of substance gave 0.8978 gram of argentic iodide.
IX. 0.3831 gram of substance gave by ignition 0.0982 gram of ferric

oxide.
X. 1.7526 gram of substance gave by ignition 0.4658 gram of ferric
oxide.

Percentages.

Iodine.

Iron.

Calculated for Fel2

81.92

18.08

Found I.

82.10

17.92

II.

81.77

III.

81.74

IV.

82.71

18.07

V.

81.66

18.56

VI.

82.00

18.01

VII.

81.76

18.28

VIII.

81.64

IX.

17-94

X.

18.60

JACKSON AND DERBY. — FERROUS IODIDE. 217

Of these analyses I.-VITI. were made by passing sulphurous anhy-
dride into the aqueous solution to convert any free iodine (formed by the
action of air on the solution) into hydriodic acid, and to dissolve the ferric
hydrate, if any, formed at the same time as the free iodine ; then, after
driving off the excess of the sulphurous anhydride, the iodine was precip-
itated as argentic iodide and the iron as ferric hydrate in the usual way.
Analyses IX. and X. were made by simple ignition of the ferrous iodide
in a porcelain crucible, when it loses all its iodine and is converted into
ferric oxide.

Properties of Ferrous Iodide.

The anhydrous salt prepared as described above forms plates of a deep
full red color with a slight brownish tint. This color is a much more pro-
nounced red than any appearing among the ferric salts with the exception
of ferric sulphocyanate. In thicker masses it is nearly or quite black ;
when melted and allowed to solidify, it forms a blackish red crystalline
mass, which looks brown on a surface of fracture.* When the plates are
examined under the polarizing microscope, they give evidence that they
belong to one of the uniaxial systems, but the outlines of these crystals
are so ragged that they furnish no means of deciding to which system
they belong. Accordingly some of the plates immersed in ether, in which
they dissolve very slowly, were observed under the microscope, when the
formation of regular hexagonal holes was observed ; there is no doubt,
therefore, that the crystals belong to the hexagonal system. If the salt
is resublimed slowly, it can be obtained in broad thin plates with a bril-
liant vitreous lustre ; but on one occasion very small yellow plates were
formed which called to mind the " gold filings " of Serullas. The melt-
ing point of the substance is given by Thomson as 177° ; we have not
attempted to determine this point anew, as it seemed to us of little prac-
tical value on account of the very deliquescent nature of the salt. If
heated to about the fusing point of soft glass in an inert gas it sublimes
essentially unaltered. "When sublimed in a stream of nitrogen, part of the
iodide is deposited in the red plates already described, the rest is carried
on further in the tube as a yellowish cloud, which deposits a dark reddish
brown amorphous powder ; this is found to be much more susceptible to
the action of atmospheric moisture than the red plates, but we think this
difference in behavior is due only to the difference in the size of the par-

* The dark color of the ferrous iodide is not without analogy, as G. P. Baxter
has found in this Laboratory that ferrous bromide has a dark yellow color with a
greenisli brown tinge.

218 PROCEEDINGS OF THE AMERICAN ACADEMY,

tides in the two specimens. The sublimation may take place in nitrogen,
carbonic dioxide, or even hydrogen, but oxygen must be carefully ex-
cluded, as this decomposes the iodide rapidly at not very high temperatures,
so that on gently heating a sample of it in a porcelain crucible in a few
minutes all the iodine has passed off in violet fumes, and there is left a
reddish brown residue of ferric oxide, which retains the shape of the
crystals of the ferrous iodide, and is magnetic. That this latter property
was not due to the presence of some magnetic oxide or metallic iron in
the residue was proved by dissolving it in hydrochloric acid, when a strong
test for ferric chloride, but none whatever for ferrous chloride, was ob-
tained. The iodine is so completely removed by this ignition in air that
the process was used as a method of analyzing the ferrous iodide (Anal-
yses IX. and X.). De Luca observed a similar decomposition of his
(white) ferrous iodide. The same decomposition of the ferrous iodide by
the oxygen of the air takes place slowly at ordinary temperatures ; a
specimen of it kept in a good desiccator gives off" enough iodine vapor in
time to color the air within the desiccator a distinct purple. A sample
of the salt was kept in a desiccator over potassic hydrate, to absorb the
iodine, for thirteen months, and the decomposition followed by occasional
weighings, when it was found that about seven eighths of the total loss
of iodine was given off in the first six months, but that even at the end
of thirteen months the decomposition was not complete, as the residue
still contained a little iodine, and gave tests for both ferrous and ferric
compounds. It follows from these observations that the salt should not
be allowed to stand even in a desiccator before analysis, but should be
analyzed as soon as possible after its preparation.

If the salt is exposed to the atmosphere, it deliquesces quickly to a
brown solution. The thinner plates, when kept in a corked flask or a
desiccator, behaved in a very striking way ; the red plate showed at first
a white border around its edge, which gradually spread inward, until the
whole plate became white ; this change was soon followed by a liberation
of iodine, wliich imparted a gray or even black appearance to the mass.
The appearance of the iodine is undoubtedly due to the decomposition by
air just described, but the cause of the whitening of the crystals was not
so evident. The most obvious theories were, that the white substance was
a second modification of the ferrous iodide, or that it was produced by the
absorption of water, which might either form enough of the green ferrous
iodide containing four molecules of water of crystallization to neutralize
optically the red color of the anhydrous salt, or might give a definite
compound containing less water of crystallization than the green salt.

JACKSON AND DERBY. — FERROUS IODIDE. 219

The idea that the white substance was a second modification gained some
support from the observations of Thomson, Carius and Wanklyn, and
De Luca, who describe ferrous iodide as white or gray, and also from the
fact that the white compound was formed in atmospheres containing very
little water, although in none of these cases had moisture been rigorously
excluded. We, therefore, tried a number of experiments to settle this
point. In all of these it was necessary to select specimens of the salt
which show this change from red to white, such as rather thin red plates
or the brown amorphous powder obtained by sublimation, since thicker
plates or masses of the iodide show the phenomenon with difficulty or not
at all. We see no reason, however, to ascribe these differences in behav-
ior to anything except the size of the crystals. Two similar portions of
the ferrous iodide were exposed, the one to dry carbonic dioxide, the
other to moist carbonic dioxide, the experiments being carried on at the
same time and under parallel conditions. The specimen in the moist
gas turned white quickly, while that in the dry gas remained unaltered.
Other similar experiments showed that the change of color took place
much more slowly in a gas containing little moisture than in one contain-
ing more. In all these experiments the amount of moisture must be
kept small, as, if too much aqueous vapor is used, the salt deliquesces so
rapidly that the formation of the white body cannot be observed. A sam-
ple of ferrous iodide was immersed in absolute ether (dried with sodium),
and underwent no change of color even after long standing ; when, how-
ever, a minute fraction of a drop of water was added, it rapidly turned
white. Upon heating a specimen of the white body in a tube filled with
dry carbonic dioxide a cloud appeared which looked like moisture and
condensed in the cold part of the tube, the substance turning red at the
same time. As the tube cooled this moisture was quickly reabsorbed,
and the white body was formed again. This experiment was repeated
several times with the same specimen. These experiments leave no
doubt that the white substance is not a second isomeric modification of
the ferrous iodide, but is formed by the action of water on the red anhy-
drous salt.

The question next arises. Is the white body a mixture of the red anhy-
drous and the green hydrous ferrous iodide, or is it a crystallized salt
with a definite amount of water ? Our experiments are not sufficiently
conclusive to answer this question with certainty, but we would present
the following argument in favor of the second view with all necessary
reserve. A sample of ferrous iodide was exposed at ordinary temper-
atures (in this case about 22°) to an atmosphere of hydrogen kept moist

220 PROCEEDINGS OF THE AMERICAN ACADEMY.

by contact with dilute sulphuric acid containing 33 per cent of acid, and
the absorption of water followed by successive weighings at intervals of
about a day. Upon plotting these results a curve was obtained, which
showed a change in tlie rate of absorption at a certain point, and a calcu-
lation of the amount of water absorbed at this point gave the following
result:

0.4289 gram of ferrous iodide absorbed 0.0491 gram of water.

Calculated for Fel22H20. Found.

Water 10.40 10.28

The coincidence of these numbers is certainly striking, and, so far as
it goes, tells in favor of a definite composition for the white body ; but
little weight can be given to a single determination, and in this case the
result must be regarded with the more suspicion, because the ferrous
iodide, after it had turned from red to white in this experiment, became
gray from decomposition with 'liberation of iodine.

When one of the red plates was viewed under the polarizing micro-
scope in ordinary air, it appeared at first granular and then white, at
the same time assuming a lower order of symmetry, as extinction was
observed at certain angles when the stage was rotated, an effect which
could not have been produced by the original crystal, as it was a hexa-
gonal plate bounded by two basal planes. This observation shows little
more than that the white substance was crystalline.

The anhydrous ferrous iodide dissolves rapidly and freely in water
with a considerable evolution of heat. The dilute solutions are colorless,
the stronger ones green. The hot saturated solutions deposit green
crystals of Fel24H20. Our analysis of this salt gave results which agreed
only approximately with those required by the theory, but this is not
strange, as it is very deliquescent and rapidly decomposed by the oxygen
of the air with liberation of iodine. The water cannot be determined by
loss, as the salt gives off hydriodic acid when it is heated. Fortunately
it was not necessary for us to spend our time in attempting to obtain
better analytical results, as Volkmann * has made a careful study of
crystallized ferrous iodide, and has isolated it with four, six, or nine mol-
ecules of water of crystallization. From his description of the prepara-
tion of these salts there can be no question that the salt obtained by us
was Fel24H20.

The solution of ferrous iodide, when exposed to the air, is decomposed,

* Ber. d. chem. Ges. 1894, R. 618.

JACKSON AND DERBY. — FERROUS IODIDE. 221

turning at first brown and then throwing down a yellowish brown precip-
itate of ferric hydrate or a basic ferric salt ; iodine is set free at the same
time, and apparently some hydriodic acid is also formed.

The ferrous iodide is very susceptible to the action of oxydizing agents,
as appears from the account given above of the action of air on it at or-
dinary temperatures, but it shows a very striking resistance to the action
of reducing agents. Its behavior in this respect was studied with some
care in the hope of perhaps encountering a subiodide. As has been al-
ready stated, ferrous iodide can be sublimed essentially unaltered in an
atmosphere of dry hydrogen. This is not perhaps altogether unexpected,
as it has been found in this Laboratory that cobaltous bromide is only
partially reduced when heated in dry hydrogen, the remainder subliming
unaltered, while nickelous bromide was reduced under the same condi-
tions essentially without sublimation.* When hydrogen was passed
throufifh melted ferrous iodide, no more reduction was obtained. Ferrous
iodide sublimed through a layer of iron by hydrogen remained unaltered.
Mercury alone or when emulsified by grinding with diatomaceous earth
produced no effect on ferrous iodide even when distilled with it. When
melted with molecular silver, on the other hand, the reduction to metallic
iron was complete. Reductions in the moist way with hydriodic acid and
iron by hydrogen or zinc dust also seemed to give negative results.

C. Liebermann and H. Sachse,t and also P. Bruck,| have recommended
ferrous iodide as a useful carrier in adding iodine to derivatives of acety-
lene, such as phenylpropiolic acid, acetyleuedicarboxylic acid, or propar-
gylic acid. We have also tried some experiments on the behavior of
ferrous iodide with organic compounds, but as the results are of little
interest, we shall summarize them as briefly as possible. Ferrous iodide,
when heated to redness in ethylene, was reduced, leaving a magnetic
residue containing carbon. With acetylene a very gentle heat was suffi-
cient to start a reaction, which afterward proceeded by itself accompanied
by a dull red glow. The products were a volatile yellow liquid contain-
ing much iodine and a voluminous residue consisting of carbon and very
little iron. The fact that the ferrous iodide is reduced by these hydro-
carbons, when it is almost unaffected by hydrogen, is remarkable. Iso-
propyl alcohol heated with ferrous iodide gave a little of an insoluble oil,
probably the iodide. When treated with an ethereal solution of aniline,
the ferrous iodide formed a white compound much more voluminous than

* Richards and Baxter, These Proceedings, XXXIV, 352.
t Ber. d. chem. Ges., XXIV. 4112.
t Ibid. 4118.

222 PROCEEDINGS OF THE AMERICAN ACADEMY.

the original salt; this in all probability is an aniline addition product
analogous to the ammonia compound described later. Benzol, benzyl-
chloride, ether, phenol, or hydroquiuone gave no perceptible action with
ferrous iodide.

Other Methods of Preparing Ferrous Iodide.

In this section of the paper we shall describe a revision of the methods
used by our jiredecessors, and also some new methods of our own.

The revision of the older methods was necessary, because in three
cases the ferrous iodide formed was described as gray or white ; and
although we have proved that the white samples formed from the red
salt owed their change of color to absorption of water, it did not follow
that those described by our predecessors were not anhydrous ferrous
iodide in a second isomeric form. De Luca,* who states that ferrous
iodide when pure is white, gives no method of preparation in any of the
journals to which we have access, so that it was impossible for us to
repeat his work. A. T. Thomson f prepared his gray ferrous iodide by
evaporating an aqueous solution in a flask containing some free iron.
On repeating this experiment we had so much difficulty from oxidation
that we carried on our later attempts in an atmosphere of carbonic diox-
ide. Under these conditions, when the water of crystallization began to
come off, the liquid became black and opaque-looking as if it contained
a suspended black precipitate. Later the surface of the boiling mass
became grayish white with a silvery lustre, but when all the water had
been driven off, the residue was brownish red like the ferrous iodide
prepared by our method. This experiment makes it probable that the
gray ferrous iodide of Thomson owed its color to the presence of water,
for in the absence of analyses he would undoubtedly have ascribed the
brownish red color of the final residue to the formation of a ferric salt by
oxidation.

As the method of Carius and Wanklyn | consisted in heating iron
and iodine in a porcelain crucible, it seemed at first sight that the gray
color of the product could hardly be due to the presence of water, and
therefore we repeated their work, following their directions as carefully
as possible. One gram of rather coarse iron filiugs,§ mixed with a little

* Coniptes-Rendus, LV. 615.
t Gmelin-Kraut, Handbuch, III. 350, 6th edition.
t Ann. Chem., CXX. 69.

§ The success of the process depends on using filings of the right degree of
coarseness. With too fine or too coarse filings unsatisfactory results are obtained.

JACKSON AND DERBY. — FERROUS IODIDE. 223

iodine to drive out the air, was heated in a covered porcelain crucible to
redness as quickly as possible ; four grams of iodine were then added in
two portions, and the heating continued until comparatively little iodine
vapor escaped around the edges of the cover. When cool, the product
was found to be a fused mass, consisting, after it had been broken out of
the crucible, of red scales exactly like those described earlier in this paper.
These scales upon exposure to the air soon turned grayish white, and
then deliquesced to a brown solution. We have repeated this process of
Carius and Wanklyn many times, and in every case the ferrous iodide
formed has appeared in red scales, before it had been exposed to the air
long enough to absorb moisture. We, therefore, are compelled to ascribe
the gray color mentioned by Carius and Wanklyn to the absorption of
water from the air, since they, like Thomson, would in the absence of
analyses have supposed that the red color was due to oxidation. Our
present knowledge, therefore, gives no reason for supposing that an-
hydrous ferrous iodide has any other color than red.

Another method of preparation may be described here, as it was tried
to see whether a white form might not be obtained at comparatively low
temperatures. A small quantity of iron by hydrogen was placed in v,he
middle of a small glass tube, and secured by two plugs of asbestos ; an
excess of iodine was inserted in one end of the tube, which was then
filled with nitrogen and sealed at both ends. The end containing the
iodine and iron was then heated to 120° by means of an air bath, and
after most of the iodine had sublimed into the cold upper part of the tube,
it was reversed so that this end and the iron were heated to 120°. After
repeating this distillation of the iodine through the iron several times, all
the metal was converted into the red ferrous iodide. This is a good
method for preparing a specimen to be kept to show the color. Ferrous
iodide was also made by passing hydrogen and iodine vapor over heated
ferric oxide.

Of these methods of preparation, that of Carius and Wanklyn is the
easiest, but it gives a product apt to be contaminated with metallic iron
and oxide of iron. Thomson's method is tedious, and does not give a
pure product, as we observed evolution of hydriodic acid toward the end
of the evaporation even in an atmosphere of carbonic dioxide. The
sealed tube method is also slow and imperfect, so that the best method,
if the pure salt is desired, is the heating of iron and iodine in nitrogen
given in the first part of this paper.

224 PROCEEDINGS OF THE AMERICAN ACADEMY.

Action of Ferrous Iodide with Ammonia.

When anhydrous ferrous iodide was exposed to a stream of ammonia
gas at any temjjerature between 0° and 100°, a large amount of the gas
was absorbed, and the iodide was converted into a voluminous white pow-
der. The composition of this substance was determined by finding the
increase in weight during the treatment with ammonia.

I. 3.5507 grams of ferrous iodide gained 1.1701 grams of ammonia.
II. 2.5411 grams gained 0.8449 gram of ammonia.

Calculated for Found.

FeUGNHa. I. II.

Ammonia 24.76 24.78 24.95

The substance therefore corresponds in formula to the compound FeCl2-
6NH3 made from ferrous chloride under the same conditions.

Properties of the Hexammonia Ferrous Iodide.

It forms a pure white amorphous powder, which occupies many times
the volume of the ferrous iodide from which it is made, and has a very
low specific gravity. Water decomposes it at once with a considerable
rise of temperature ; the products are ferrous hydrate, amraonic iodide,
and ammonia. We did not succeed in finding an organic solvent for
it. When exposed to the air it gives off ammonia rapidly and turns
brown ; in an atmosphere of ammonia it remains unaltered even at 100°;
heated in an indifferent gas it loses ammonia, and is converted into a dark
brown powder, which seems to be stable in the air. Two analyses indi-
cated that this brown substance contained between two and three mole-
cules of ammonia to each molecule of ferrous iodide. Its stability in air
leads us to think that it is a new compound, but evidently it was not pure,
as the analyses differed by three per cent. Unfortunately we could find
no way of purifying it.

Action of Bromine on Hexammonia Ferrous Iodide.

When the compound FeT26NH3 was exposed to the vapor of bromine, it
absorbed a large quantity, so that the product in one case contained 71.03
and in another 70.41 per cent of bromine. This apparent constancy in
the amount of bromine absorbed led us to examine the product more care-
fully in the hope that it might be a definite compound, but our exper-
iments have convinced us that it is only a mixture essentially made up
of ferric bromide, ammonic bromide, and an ammonic bromiodobromide

JACKSON AND DERBY, — FERROUS IODIDE. 225

NH^BrlBr. The product had a reddish brown color, and gave off bro-
mine when exposed to the air. It was investigated by treating it with
ether, in which a portion dissolved with considerable evolution of heat.
The residue insoluble in ether had a yellowish color, and seemed to be a
mixture of ferric bromide and ammouic bromide in varying proportions.
These two substances were recognized by the usual tests. The ether
extract on evaporation left a residue which was green with a brilliant
metallic lustre by reflected light, deep blood red by transmitted light ; it
resembled rosaniline, therefore, but the red color was less purplish than
in that body. Examined with the microscope, crystals which looked like
ammonic bromide were observed. It had a strong odor similar to that
of bromide of iodine, and, although not really stable, could be kept for
some time without much decomposition. Upon ti-eating it again with
ether a fresh quantity of the insoluble yellowish residue was left behind.
The residue obtained by evaporating one of our first ether extracts was
analyzed with the following results :

0.7149 gram of the substance gave 0.0216 gram of ferric oxide, 0.0405
gram of ammonia, 1.4289 grams of mixed silver salts which yielded
0.7652 gram of silver.

In calculating these results, as the iron from its small amount was
evidently an impurity, the amount of ferric bromide corresponding to it
has been subtracted from the amount of substance taken, and the bromine
contained in this ferric bromide from the bromine found. As the sub-
stance gave a test for a ferrous salt, it would have perhaps been more
correct to consider the iron present as fei-roferric bromide, and this would
have given numbers agreeing better with those calculated from the form-
ula ; but we think it wiser to give the results on the assumption that ferric
bromide was present, as we cannot believe that the amount of ferrous
bromide was more than a trace, since the substance had been exposed to
an excess of bromine.

Calculated for NH^BrlBr. Found*

Ammonium 5.90 6.38

Iodine 41.64 40.96

Bromine 52.46 53.32

These numbers would seem conclusive if it were not for the presence
of the small amount of ferric bromide (considered in calculating them)

* If the ferric bromide was not subtracted, the numbers were, NH4 5.67, iodine
36.-3!), bromine 56.41.

VOL. XXXV. — 15

226 PROCEEDINGS OF THE AMERICAN ACADEMY.

and the appearauce under the microscope of crystals resembling ammonic
bromide ; but with these sources of error, we thought it necessary to
support our analysis with evidence of some other sort, and therefore
undertook the preparation of NH4BrIBr direct in order to compare its
properties with those of our compound. Upon treating ammonic bromide
with au ethereal solution of bromide of iodine, the salt dissolved, giving
a red solution, and on evaporation of the ether a residue was obtained
the properties of which were identical with those of the ammonic brom-
iodobromide described above. Further, upon exposing ammonic iodide
to the vapor of bromine an amount was absorbed corresponding very
neai'ly to that required by the formula.

1.6487 grams of ammonic iodide absorbed 1.7804 grams of bromine.

Calculated for Per cent of Bromine

NHjBrlBr. in the Product.

Bromine 52.46 51.92

During the action of the bromine the ammonic iodide at first turned
black, but as the absorption went on it finally became scarlet, and this
product proved to be identical with the ammonic bromiodobromide. A
curve constructed from the increase of weight of the ammonic iodide
showed that there was a marked diminution in the speed of the absorption
after the first atom of bromine had been added. This decrease in the
speed of absorption is undoubtedly connected with the conversion of the
black substance into the red NH4BrIBr, but we are unable to decide
whether this black intermediate product is another compound NH4BrI,
or only a mixture of ammonic bromide and free iodine. It is a curious
fact, in view of the absorption of free bromine by ammonic iodide, that,
so far as we could find, an ethereal solution of bromine had no action on
this salt. We have not continued our experiments in this direction for
fear of intruding on the field of work occupied by Wheeler, who with
Pratt and Barnes has prepared the ammonic triiodide.*

We have adopted provisionally the name ammonic bromiodobromide
and the formula NH^BrlBr for this substance, because it is made by the
action of bromide of iodine on ammonic bromide, and when it decomposes
bromide of iodine is given off and ammonic bromide left behind ; but we
do not feel that these observations settle its constitution finally, although
they certainly are strong arguments in favor of the formula adopted. If
this is correct, the formation from ammonic iodide and bromine must be

* Am. Chem. Journ., XIX. 675 (1897).

JACKSON AND DERBY. — FERROUS IODIDE. 227

preceded by the replacement of the iodiue in the salt by bromine, and this
may account for the fact noted above, that an ethereal solution of bromine
does not act on ammonic iodide.

Properties of Ammonic Bromiodobromide NH.BrlBr. '

The salt has a blood-red color in transmitted light, but is a brilliant ,

green with a metallic lustre by reflected light. The specimens examined j

by us contained very elaborately twinned crystals with the branches at J

right angles, but we are inclined to think that these consisted of ammonic |

bromide, which is easily formed by the decomposition of our salt ; on j

this account it smells of bromide of iodine, and on long standing leaves a |

white residue of ammonic bromide ; the same decomposition takes place j

to a limited extent when it is dissolved in ether ; the ethereal solution is j

dark red. When treated with water, bromine and iodine are set free, ]

1
and a solution is formed. Hydrochloric acid sets free iodine. Sodic '

hydrate or ammonic hydrate gives a black precipitate of iodide of nitro-
gen; upon adding hydrochloric acid to the filtrate from this iodide of
nitrogen, bromine is set free. Sulphurous acid gives hydrobromic acid
and hydriodic acid. ;

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 11. — December, 1899.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.

NOTE ON THE CONSTITUTION OF DIPARABROM-

BENZYL GTANAMIDE.

By C. Lorixg Jackson and R. W. Fuller.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.

JsOTE ON THE CONSTITUTION OF DIPARABROMBEN-

ZYLCYANAMIDE.

By C. Lorixg Jackson and R. W. Fuller.

Received Norember 24, 1899. Presented December 13, 1899.

The work described in this paper consists of the conversion of the
silver salt of cyanamide into a dialkylcyanamide, and the determination
of the coustitutiou of this body. Last summer (after this work was
finished) a paper a|ipeared in the " Berichte der deutschen chemischen
Gesellschaft," in which Wallach * described a number of substituted
cyanamides obtained by the action of bromide of cyanogen on secondary
amines. For fear of approaching too near the field thus reserved by
Wallach we shall abandon the further study of dialkylcyanamides, but
the study of alkyl compounds of dicyandiamide and dicyanimide will be
taken up in this Laboratory ; in fact work on this latter substance is
already in progress.

Theoretically a dialkylcyanamide derived from the silver salt of
cyanamide might have either of the following formulas R-N=C=N-R or
R.rN-CN, and it is easy to determine by experiment which of these two
formulas is correct. So far as we can find, but a single experiment of
this sort has been tried ; this was published some years ago by Fileti and
Robert Schiff,t who prepai'ed diethylcyanaraide by the action of ethyl-
iodide on argentic cyanamide at 100° for some hours. The product was
extracted with ether, and divided into two portions ; one was distilled,
and gave a boiling point of 186°, whereas Cloez and Cannizzaro, J who
prepared it by the decomposition of ethyl cyanamide, found a boiling
j)oint of 190°. Fileti and R. Schiff analyzed their distillate, and ob-
tained carbon 60.66 instead of 61.22 and hydrogen 10.11 instead of 10.30.

* Ber. d. chem. Ges., XXXIL 1872.
t Id., X. 425 (1877).
t Ann. Chem., XC. 95.

232 PROCEEDINGS OF THE AMERICAN ACADEMY.

The other portion of their product (which had not been distilled) was
decomposed by means of hydrochloric acid on the water bath ; the
chlorides obtained by evaporating the hydrochloric acid solution were
converted into chlorplatinates and crystallized fractionally, when they
obtained two end fractions in which the platinum was determined with
the following results : —

Found.

(NHi),PtCl6

I.

II.

((CoH5)3NH2)2PtCl6

inum

44.04*

42.51

36.3

35.30 *

(C2H5NH3)2PtCl6 requires 39.24 * per cent of platinum.

In considering these results of Fileti and R. Schiff, it is to be observed
that the diethylcyanamide was not purified, and that no very sharp cri-
terion of purity was given (Wallach states that these substances are
decomposed by distillation under ordinary pressure, so that the boiling
point is not of much value in this respect), further that their analyses of
the platinum salts did not give numbers very near to those calculated.
It seemed to us, therefore, worth while to try similar experiments with,
if possible, a crystalline disubstituted cyanamide, which could therefore
be obtained in a state of undoubted purity, and also with one which
would yield amines more easily separated than ammonia and diethyl-
amine. We selected for this purpose the diparabrombenzylcyanamide,
since the parabrorabenzyl compounds show a great tendency to crystallize,
and the diparabrombenzylamine, if formed, could be recognized by its
melting point, 50° (dibenzylamine is a liquid), while the parabromben-
zylamine, if that were the product, gives a carbonate with a definite
melting point, and both these amines could be separated without difficulty
from ammonia.

The diparabrombenzylcyanamide proved to be a well crystallized solid,
melting at 133°. On decomposition with dilute sulphuric acid it gave
diparabrombenzylamine, ammonia, and carbonic dioxide by the following
reaction : —

(C6H4BrCH2)2NCN + 2HoO = (C6H4BrCH2)2NH + NH3 + COo .

Our results, therefore, confirm those of Fileti and R. Schiff, and leave
no doubt that the dialkyl derivatives from argentic cyanamide are cyan-
amides and not carbodiimides. If they are formed by direct replace-
ment of the silver in argentic cyanamide by the alkyl radicals, the same

* These are the numbers given by Fileti and R. Schiflf. Tliey would be some-
what altered by using modern atomic weights.

JACKSON AND PULLER. — DIPARABROMBENZYLCYANAMIDE. '2o3

constitution (AgaNCN) must be ascribed to this substance and to cyan-
amide. If, on the other hand, these compounds are formed by successive
additions of the alkyl bromide with splitting off of argentic bromide, the
disubstituted cyanamides could be iormed from a silver salt with a
carbodiimide formula, as is shown by the following reactions : —

R

NAg Br NAg NR

^ \ / //

C +RBr=C =AgBr+C

^ ^ ^

NAg NAg NAg

R
NR Br NR NRj

// \ / /

C + RBr = C = C + AgBr.

NAg NAg N

Our results, therefore, prove nothing in regard to the true formula of
cyanamide.

Preparation op Diparabrombenzylcyanamide,
(C6H4BrCH2)2NCN.

The yellow silver salt of cyanamide, AgaNCN, prepared according to
Walther,* was mixed with a benzol solution of parabrombenzylbromide
in the proportion of two molecules of the bromide to one of the salt,
which should be finely powdered. The mixture was heated in a flask
with a return-condenser on the steam-bath, until after four or five hours
the full yellow color of the argentic cyanamide had been completely
replaced by the yellowish white color of argentic bromide. The precipi-
tate was then filtered out and washed thoroughly with hot benzol, and
the filtrate and washings evaporated to dryness, when a thick reddish
yellow oil was left. To purify this it was dissolved in hot alcohol, and
the strong solution allowed to cool slowly ; a yellow oil was deposited
at first followed by a white crystalline substance, which was obtained by
pouring the solution off from the oil as soon as the crystals began to
appear. By repeated recrystallizations of this sort the melting-point of
the substance was raised to 133°, where it remained constant. It was
dried in vacuo and analyzed with the following results : —

* Journ. prakt. Chem., 1896, 510.

234 PROCEEDINGS OF THE AMERICAN ACADEMY.

I. 0.1508 gram of the substance gave by the method of Carius 0.1494
gram of argentic bromide.
II. 0.3041 gram of the substance gave 20.2 c.c. of nitrogen at a tem-
perature of 18°. 6 and a pressure of 760.4 mm.

Calculated for Found.

(C7U5Br)2NCN I. II.

Bromine 42.11 42.18

Nitrogen 7.37 7.64

In view of the great tendency of cyanamides to polymerize, it was
thouglit safer to determine the molecular weight of the body * by the
metiiod of freezing a benzol solution, which gave the following results : — ■

0.375 gram of the substance dissolved in 16.25 grams of benzol produced
a depression of 0°.28 in the freezing point.

Calculated for (C7HeBr)3NCN. Found.

Molecular weight 380 337

There can be no doubt, therefore, that the substance is really
(C6H,BrCH2)oNCN.

Properties of Diparabrombenzylcyanamide.

The substance crystallizes from benzol in sheaves of white crystals
shaped like the blade of a lancet, sometimes united laterally into groups
wiih comb ends. It melts at 133°. It is freely soluble in benzol, chlo-
roform, acetone, or acetic ester ; soluble in toluol ; slightly soluble in cold
ethyl or methyl alcohol or glacial acetic acid, freely soluble in these
solvents when they are hot; slightly soluble in ether, carbonic disulphide,
or in hot or cold water ; essentially insoluble in ligroine. It is slowly
decomposed by cold strong sulphuric acid; apparently unaffected by
hydrochloric acid or nitric acid in the cold. The best solvent for it is
hot alcohol.

In order to see if it could form a chloride, a portion of the dipara-
brombenzylcyanamide was dissolved in anhydrous benzol and saturated
with dry hydrochloric acid gas. No precipitate was formed even after
the mixture had stood for two weeks, and on evaporating off the benzol
the original substance was recovered unaltered. It would seem from
this experiment that the diparabrombenzylcyanamide cannot unite with
hydrochloric acid.

* This work was done before the appearance of Wallach's statement that di-
substituted cyanamides show no tendency to pol^^merize.

JACKSON AND FULLER. — DIPARABROMBEXZYLCYANAMIDE, 235

A number of experiments were tried in the hope of obtaining polymeres
of the diparubrombenzylcyanamide. A dilute solution of sodic or potas-
sic hydrate produced no effect on the substance, either by long standing
in the cold or by boiling the mixture. The substance was boiled for two
weeks with a solution of ammoiiic hydrate, care being taken to replace the
ammonia which escaped, but the only change observed was that the color
turned from white to pale brown, evidently due to a slight decomposition,
since the melting point of the substituted cyanamide was essentially unal-
tered. Water alone was also tried at the boiling point, but produced no
chano-e. Upon heating the substance above its melting point it remained
unaltered to 160° ; above this point it turned first yellow, and at higher
temperatures red, and on cooling gave an oily product, which we could
not bring into a fit state for analysis. A similar viscous product was ob-
tained when the substance was heated with sodic acetate. In both these
cases it seemed evident that a decomposition had taken place rather than
a polymerization, and our experiments, therefore, confirm the statement of
Wallach that these dialkylcyanamides show no tendency to polymerize.

DECOMPOSITIOlSr OF DiPARABROMBENZYLCYANAMIDE.

As some preliminary experiments showed that the substituted cyan-
amide was decomposed with difficulty by hydrochloric acid in open vessels,
we adopted a dilute sulphuric acid having a specific gravity of 1.44,
which has frequently given good results in this Laboratory. Several
grams of the diparabrombenzylcyanamide were boiled with a large excess
of this acid in a flask with a return-condenser. Soon after the substance
melted in the hot acid an effervescence was observed, and upon testing
the gas given off with baryta water, it proved to be carbonic dioxide.
As the oily drops did not disappear, the heating was continued for ten
hours, which reduced the amount of oil, but did not entirely remove it.
On cooling, the whole liquid became filled with a voluminous white crys-
talline precipitate ; we determined, therefore, to stop the process at this
point and isolate this crystalline sul)Stance, which could be easily separated
from the portion undissolved in the hot sulphuric acid. This latter sub-
stance which solidified on cooling, seemed to consist of undecomposed
diparabrombenzylcyanamide, as it gave a fresh quantity of tlie crystalline
product on boiling again with sulphuric acid ; its amount was insignifi-
cant. To obtain the crystalline pro<luct it was filtered from the acid
liquid, washed with a little cold water to free it from adhering acid, and
then dissolved in hot water, which left behind the few black lumps of
undecomposed cyanamide. The solution was then filtered, and the sul-

236 PROCEEDINGS OF THE AMERICAN ACADEMY.

phate decomposed by the addition of a strong solution of sodic hydrate,
which set free the base as a pasty mass solidifying on cooling. This
base, after purification by crystallization from alcohol, showed the con-
stant melting point 50°, which is that of the diparabrombenzylamine*
(C6H4BrCH2)2NH. For greater security the chlorplatinate was pre-
pared and analyzed. Chlorplatinic acid added to an alcoholic solution of
the base gave a yellow precipitate, which was purified by washing with
alcohol, and dried in vacuo.

0.1783 gram of the salt gave on ignition 0.0309 gram of platinum.

Calculated for [(C7H,;Br),NH2]2PtCl|i, Found.

Platinum 17.58 17.34

The sulphuric acid filtrate from the crystals of the diparabrombenzyl-
amine sulphate was treated at first with sodic carbonate, and finally with
a large excess of sodic hydrate, and distilled with steam, the distillate
being collected in a series of flasks containing hydrochloric acid. After
the distillation was finished, the contents of the flasks were evaporated to
dryness on the steam bath, and the white residue dissolved in a little water
and converted into the chlorplatinate ; this was a yellow precipitate
crystallizing in octaiiedia, which was washed with water and alcohol,
dried in vacuo, and analyzed with the following results: —

0.2061 gram of the chlorplatinate gave on ignition 0.0903 gram of
platinum.

Calculated for (NH4)2PtCl8. Found.

Platinum 43.91 ' 43.81

It is evident from the experiment described above that the products of
the decomposition of diparabrombenzylcyanamide, when boiled with di-
lute sulphuric acid, are carbonic dioxide, diparabrombenzylamine, and
ammonia. The reaction, therefore, I'uns as follows: —

(C6H4BrCHo)2NCN -f 2H.,0 = (CJl4BrCH,)oNH -f NHg + COj.

* Jackson, These Proceedings, XVI. 254.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 12. — January, 1900.

NOTE ON THE CHIEF THEOREM OF LIE'S THEORY
OF CONTINUOUS GROUPS.

By Stephen Elmer Slocum.

NOTE ON THE CHIEF THEOREM OF LIE'S THEORY OF

CONTINUOUS GROUPS.

By Stephen Elmer Slocum, Clark University, Worcester, Mass.

Presented by Henry Taber, November 8, 1899. Received December 1, 1899.

The chief (Jiaupt) theorem of Lie's theory of finite continuous groups
is that a system of r independent infinitesimal transformations

X-l, Xn, . . . X,.,

such that

(Xj, X,) = i c,,, X* U, h^\,2...r),

for constant coefficients Cj^.^, generates a continuous group with r param-
eters, that is, a group with r parameters in which each transformation
can be generated by an infinitesimal transformation of the group.f

Professor Study, however, has shown that, notwithstanding the infini-
tesimal transformations of the special linear homogeneous group satisfy
Lie's criterion, nevertheless not every transformation can be generated
by an infinitesimal transformation of this group. % Consequently Lie's
theorem is subject to certain limitations. The precise nature of the error
in the demonstration of Lie's theorem has, so far as I know, not been
pointed out ; and to show wherein it consists is the object of this paper.
For this purpose I shall carry out, for the case of a particular group, the
successive steps (as given in the " Continuierliche Gruppen," pp. 368-
377) in Lie's demonstration of the first fundamental theorem of his the-
ory, upon which the chief theorem (the second fundamental theorem)
depends. At a certain point in this demonstration an assumption is
made in which Lie's error consists.

* Xs denotes i^-^^ix^. . . x„)~ + . . . + |s„(^i . . .Xn)-

3 Xi 3 •'■n

t Lie : Continuierliche Gruppen, pp. 211, 390.

X Engel : Leipziger Berichte, 1892, p. 279. See also Taber : Am. Jour. Maths.,
XVI.; Bull. N. Y. Math. Soc, July, 1894, April, 1896, Jan., 1897; Math. Ann.,
XLVI. ; Rettger: Proc. Am. Acad., XXXIII. ; Newson : Kansas Univ. Quarterly,
1896 ; and These Proceedings, p. 97.

240 PROCEEDINGS OF THE AMERICAN ACADEMY.

I shall number the equations the same, and shall use the same nota- j
tion for the special group considered as employed by Lie in his general
demonstration.

Lie first shows that if the n equations • i

(1) ^/ =/i {^1^ • • • ^n^ «i . . . a,) (i — 1, 2, . . . n) ]

represent a group with r parameters, the x""s as functions of the x's and |
a's satisfy certain differential equations of the form !

9x' »■ '

(9) ■— = 1j xPj^ (ai . . . a,.) ^ji (x/ . . . x^) j

{i=\,2,...n;k=\,2,...r), ■ • \

in which the determinant of the i(/j/. ^ 0 ; — that, consequently, these i
equations may be written in the form

^ 9x' '

(10) 4(x/ . . . x,^) = Ij. aj^ («!... «^) -^ ;

1 do^ I

(« = 1,2, . . .71; J = 1,2, . . . r), ■

where the determinant of th'e a^t i 0 ; and that, further, no linear rela- '■
tion of the form ;

e,^,,{x') + ... + e,^,,{x') = 0, .

with constant coefficients e, persists, simultaneously, for ^' = 1, 2, . . . w. '

We shall consider a case for which both n and r are equal to two. The |
equations

Xi —Xi-\- 02 =/i (x, a), I

Xo = e'^'iX.2 + Oi r=/2 (^j «)> I

define 00" of transformations T^^ which constitute a group. For, by the 1
elimination of xl, xJ from (1) and

(2)

we derive

(3)
where

(4)

^' = x^ + b, =fi(x',h),

X

Xo!' — e''2x/ + 61 =/o ix', h).

x{' = .ri + Co =/i {x, c),
x^' = e''2 iCa + <^i = f-i (x, c) ,
Ci = «! e''2 -}- bi = (fii (a, b),

C2 = 02 + ^2 = ^2 (Sh ^)*

SLOCUM. — FINITE CONTINUOUS GROUPS.

241

That is, functional equations persist of the form

(5) / {f{x, a), h) =f, (x, cj. (a, b)) (^ = 1, 2).

Therefore, the composition of two transformations JJ, and Tf, of the fam-
ily is equivalent to a single transformation of the family. It is to be
observed, as noted by Lie, § 1 in the demonstration of the general case,
that the functions ^i and ^2 ^.re independent of each other with respect
to ^1 and ^2- l^or

9 b,

1, «i eh

0,

is not identically zero.

We may, therefore, regard x^, Xo, a^, a^, c^, Cn, as independent variables,
and x/, x/, x(' , x<^', b^, ^2, as dependent variables. Then the differen-
tiation of the functional equations (5), or of

(5')

i. e., of
(p'a)

fKi^',b)=A{x,c) (A =.1,2),

Xl -{- b.2 = Xy -\- C2,,

bi + e*2 X:^ = Ci + 6*^2 X2,
with respect to the a's gives

^"'' + 1^ = 0,

c^fli 5«i

(7)

clXx , 9 b.

9a2

+ '^ = 0,

J 9x/ , 9bi , ^9bo

9ai 9 ai

^ 9x2 , 9 bi

€"2 1- \- x.{ eh

9a2 9 a^

9ai

9h

9ao

0,

0.

In order to obtain expressions for ^^ , we differentiate (4) with

9a^

respect to ctj and agj and thus obtain

9b^

0 =

9b2
9ai'

C^02

VOL. XXXV. — 16

242 PROCEEDINGS OF THE AMERICAN ACADEMY.

cJClo 0 0.2

These equations give

0— = — e*=, ;^ = «i eS n~ = ^' n~ = " ^

cyOi c'«2 cyfli c'a2

and, inserting these values, equations (7) become

0,

-1,

5«i
9a<2,

e^s- — = e*2(a;2 - «i).

5 02

If in (8o) we insert tlie values of a;/ and a^g' derived from

3C-^ — 37j ~r~ (z^j

^2 = e^s X2 + Oi,

they become equations between the independent quantities x^, X2, «i, 02,
bi, b.2 ; and must, therefore, be satisfied identical)}'. Hence, equations
(8 a) will still persist identically, in virtue of equations (1), if we assign
definite values to the ^'s.

For this purpose let e^ = 1. We then have

(9a)^r— = 0, -^ — = 1, -^ — = 1, ^ — =zx2—ai.

dtti da^ dtti dao

Therefore, if we define functions i of the x"s and functions 1// of the a's
as follows,

$ii(x)=0, ^,„(x') = -l, ^,,(x') = -l, i,„(x')=-X2',

xpn («) = — !, i/'i2 («) = «i. «A2i («) = 0, >/'22 ( '0 = — J 5

we have

SLOCUM. — FINITE CONTINUOUS GROUPS.

243

(9)

3x[
9 a.

= ^n («) in (^') + 'A2t («) 4i (^')

({- 1, 2; /t = 1, 2).
The determinant of the if/'s, namely,

'/'2U "/'22

- 1,

«!

0, -1

not being identically zero, equations (9) may be solved for the |'s, giving,
(10)

4 (x/, X,') = a,, (a) ^ + a,, (a) ^

where

(i =l,2;j = l, 2),

cm («H «2) = — Ij «12 («1> «2) = 0)

agi (ai, ao) = — Gi, 022 («i> ^2) = — 1-

It is to be observed that

ei ill {x') + e, ^21 (a^O = — ^2,

^1 112 («') + ^2 I22 (a;') = — ei — ^2 ^1''

Therefore, these expressions linear in the |'s cannot both be simulta-
neously zero (for all values of the a;"s) if e^ and e^ are constants other
than zero. That is to say, no two constants ei and eo, not zero, can be
found for which

ei in {x') + Co 4.- {x') = 0,

for i = 1 and i = 2, simultaneously.

We come next to the demonstration of the second part of Lie's first
fundamental theorem. Starting with a system of equations which define
a family with r essential parameters (r = 2 in the case considered), and
satisfying differential equations of the form (9), we proceed to follow the
steps in Lie's demonstration that this family (provided certain other
conditions are also satisfied) constitutes a group.

As shown above, the family of 00- of transformations T^^ defined by

(1)

Xi =Xi-\- On =fi (x, a),

xJ =:r e"2a: + «i =.fi (x, a),

satisfy differential equations of the form (9) in which the determinant of
the il/jk is not identically zero ; also the differential equations

244

PROCEEDINGS OF THE AMERICAN ACADEMY.

iji {^1 ,Xi) = 0.J1 (a) -^^ + aj. (a) ^r—

where the ^'s and u's are defined as follows :

^n (^') = 0, i,, (xO = - 1, io, (x') =-1, ^22 (x') = - x,\

«ii (a) = — 1, ax2 («) = 0, a2i (a) = — «!, Ojo (a) = — 1.

Moreover, if we put Oi'"' = Oo'"* = 0, then Oj = aj"", ag = ^''2"'* gives the
identical transformation ; and the determinant of the aj^. (a'"'), namely,

0

„ (0) „ (0)

an , ai2

"21 ) "22

1,

-a, -1

is neither zero nor infinite.

In order to prove that this family constitutes a group, we proceed to
integrate equations (10). For this purpose, introduce a new auxiliary
variable t by means of the equations

(11)

dt
dt

= Ai an (cii) 0,0) + A2 ooi (cti, CI2) ^^ — \\ — Cli X^,
= Xi a.\i\Cti, 0,2) + ^^2 0.22 (^1) ^2) =■ — ^2

^2>

where Xi and X2 are any arbitrary but definite constants. To determine
the constants of integration, we assume that ai, a^ take the values «!, a^
for < = F. The integrals of equations (11) are then

a2 = a2 — X2 (t — t),

r(Xi + fliXs)] -

Xi (; — ?) = /x-i, X2 (< — ?)= ^2 ;

the integral equations then become

Let now

Ml , «1 Ml ^ / -x

«i = — ITT- + — TT- - = $1 (^, a),

(12)

a^ = ao — Ma = *2 (Mj <^)'
It is to be observed that the a's are independent functions of the /a's : for

SLOCUM. — FINITE CONTINUOUS GROUPS. 245

9 a.

9fij

_ 1 /eM2- 1\

is not identically zero.

By means of equations (12) we can introduce in equations (1) the
new parameters fii, fi^ instead of Ui, a^. Solving (12) for /^i, jMo) ^e have

(«2-«2)(«i- -^je"^-"^
H = i-e^.-cH = ^^ (''' «)'

fxo = 02 — a2 =. Mg (a, a).

If we introduce these new parameters fxi, /^.o in equations (1), the ar"s
become definite functions of Xi, x.^, fii, /j-z, «i5 «2 5 3,nd, since the fi's contain
t, the a;"s will also be functions of the auxiliary variable t. Multiplying
equations (10) by Ai, X^, and adding, we get

4 ^ e / ' 'A % 9x{.daf,

2,A,4.(x, .,) = >. 5^^ (^ = h2),

or

dx'

—jj' ^^ '^1 ^11 (•'^l J ^2 ) + ^^2 ^21 (3^1 J a^2 ) = — ■^2>

(14)

-^ — ^1 ^12 (a:*/, x.{) + A2 ^22 (a?/, xl) = — Ai — A2 a^a'^

which define the a;"s as functions of t. Since «!, Oo reduce to Ox, 02 for
< = ^, we see that x{^ x^'-, for < = 7, take the initial values

Xi = a?! + «„ =/i (a-, o),
(15) _ _ _ _

a^o' = e^-'-Xo + «i ==^2 (^1 «)•

Further, the integrals of equations (14) contain the A's only in the
combination /^i, /^o ; and, therefore, can be expressed in the form

Xx =: -Tl {X-x , X<2, , fXi, ^2) = a*! — /X2,

(16)

, XT / — r ~ r - A''! f 111 '^''

a:2 = i'o (a:^i , a^2 » /^i, /^2) = — ( — 1 + —•

Inserting therein the values

^/ =/• (^» «)? '< =/• (^' «)j ^k = <^4 (ft, «)

(i = l, 2; ^-=1, 2),

246 PROCEEDINGS OF THE AMERICAN ACADEMY

the totality of equations (1) to (12), we obtain the functional equations

(17) F, if(x, a), /.) =f, (x, ^ (ix, a))

(i = 1, 2).

We have already denoted by Ta the transformation defined by (1), in
which the parameters are a^, a^. Consequently, equations (15) define
the transformation Ta. We may now denote the transformation defined
by (16) in which the parameters are yui, /xo by ^^. Then the functional
equations (17) may be expressed in the single formula

(18) TaE^=Ta.

For Ta transforms x^ into x- ^=f(x, a), and E^ transforms x^ into
Fi {x' , yu), while, in virtue of (12), T^ can also be written in the form

^l "^fi (^j * (m, «)) («■ =1,2).

The relation (18) persists, therefore, provided the three parameter
systems (aj, a.,), (/xi, [x.^, («^, Oo) fire connected by relations (12).
Therein a^, «o denote definitely chosen general values of (7i, a^.

We now make use of the assumption that the transformation T^., de-
fined by (1), shall become the identical transformation for Oi = Oi"'*,
«2 = a.^'^\ Namely, for the moment, let a^ = a/'', a„ = Og'*" ; whereupon
Ta becomes the identical transformation J[,(0). Then, because the deter-
minant of the a^j (a'"') is, according to assumption, neither zero nor in-
finite, and, therefore, the former considerations are also valid for a = a'"',
— the Oj, Oo in virtue of (12) assume the values

(19) a, = ^, (/xi, /X2, a A «./') ('^■=1,2);

that is to say, since we take aj"" = aJ^^ = 0, Oi and a^ assume the values

(19, «)

ai

/Xi /Xi
/W2 e"2 f^2 '

Hence, (18) becomes
and, therefore.

do

(20)

E, = t;.

Whence it follows that each transformation E^ belongs to the family of
transformations T„, defined hy equations (1). If now, conversely, we
could establish that each transformation T^ belonged to the family of

SLOCUM. — FINITE CONTINUOUS GROUPS. 247

transformations ^^, the first fundamental theorem would then be proved.
For taking the a's arbitrarily, we could then find a system of parameters
/x such that -Efj, = Ta; and, the «'s being also an arbitrarily chosen sys-
tem of values of the parameters a of equations (1), we should have,
symbolically,

J a J a = J a J^fi. = -^ cii

or

where

fi if{x, a), a) = Fi(J{x, a), fx) =/, {x, a)
0'=1, 2),

a, = ^, (/x, a) = *, (31 (a, a""), ") ;

that is .to say, the composition of two arbitrary transformations Ta and
7'a of the family gives again a transformation Ta of the family. This is
precisely the step taken by Lie, who assumes that because, — as men-
tioned above, page 244, — the ^g^'s are independent functions of the ;u,'s,
each transformation of (1) belongs to the family ^^. But, although the
functions aj, a^, as defined by (19 a), are independent of the /a's, since
the Jacobian

1. H"^ f^2\ e^V

is not identically zero, nevertheless, for certain values of the a's, the /x's
are infinite. Infinite values of the /x's, however, are expressly excluded
from consideration. For //.^ = Aj (< — 7), and since t and } cannot be
infinite, if /x^ is infinite A^ is infinite ; and, by supposition, the A's are
arbitrary but definite constants in the integration on -page 244. So we
cannot assume that every transformation Ta of the family (1) belongs
to the family JS^. Thus, solving equations (19, a), we have

(19, b)

For ai rj: 0, and ao an even multiple of tt/, [j.i becomes infinite. More-
over, this transformation of the family (1) is distinct from any transfor-
mation of this family Ta for which the /x's are finite.

On page 375 of the " Continuierliche Gruppen " Lie points out that
every transformation of the family E^ is generated by an infinitesimal
transformation. The infinitesimal transformation in question is repre-
sented by the symbol

248 PROCEEDINGS OP THE AMERICAN ACADEMY.

r r n £j f

^j ^j ^jf = 2^- ^i h ^ji (^1 • • • a^«) -^ >
and adds to an arbitrary function y(rri' . . . a-/) the increment

and, therefore, to a:^ adds the increment

8xi = iji(xi . . . x„) • St.
This shows its relation to the simultaneous system on page 245, namely,
dxf r

dt

r — '^3 \- ^a (^/ • • • ^n) (i = 1,2 . . . n).

1

If the theorem stated by Lie, jjage 375, " Continuierliche Gruppen,"
was true without exception, namely, that every transformation of the
family T„ belonged to the fiimily E^, it would then follow that every
transformation of the family T„ could be generated by an infinitesimal
transformation ; for then taking the a's arbitrarily, we should have

^ ^M = To..

But, for a system of values of the a's for which one or both of the func-
tions Ml (a, a*"'), M2 (a, a*°') are infinite, there is no equivalent transforma-
tion of the family £J^ ; and, consequently, such a transformation cannot
be generated by an infinitesimal transformation of the group. E. g., the
transformation T,, considered above, for which oi iji 0 and a^ is an even
multiple of tt i, cannot be generated by an infinitesimal transformation of
the group.

1
In demonstrating the second fundamental theorem (the chief theorem)
Lie assumes the results of the first fundamental theorem. He shows
that a system of r independent infinitesimal transformations*

Xif= 5, $,, (xi... x,,)^ (^• = 1, 2 . . . r)

1 dx^

generate a family of transformations K„, with r essential parameters,
which contains the identical transformation, and is defined by the
equations

* Lie terms the infinitesimal transformations or symbols of infinitesimal
transformations A'^, X2, . . . Xr independent if they satisfy no linear relation
eiXif+ . . . + Cr Xrf^ 0, with constant coefficients e.

SLOCUM. — FINITE CONTINUOUS GROUPS, 2i9

r r r

(A) x! = Xi+ 2fc ofj X, Xi + i 2, :S, a^ Oi A;. X; x, + ... = (i (x, a)
1 11

(i = 1, 2 . . . r) ;
further, that, if aud only if

r

will this family satisfy differential equations of the form required by the
first fundamental theorem. Consequently, only if this criterion is satis-
fied by the infinitesimal transformations can they generate a group.

Proceeding now, as in the. demonstration of the first fundamental
theorem, we introduce certain new parameters [x, and, finally, obtain the
equation

where a^ = 0^ (fj., a) (^-=1,2,... r). As before, since the family of
transformations 9Ea, defined now by equations (A), contains the identical
transformation, we have

where a^ =: </>* (jx, a""), and fij. = ^j. (a, a'"') (k = 1, 2, . . . r)* and thus

Ea Ea = Ea-

In the former case we saw, page 247, that, if the a's were chosen arbitra-
rily, one or more of the fx's might be infinite. In the present case the
/x's are numerical multiples of the a's f ; aud, consequently, the /a's are
finite whenever the a's are finite. E.g. (n = r = 2),

fi (x, a) — Xi -\- tto,

12 {x, «) == e«2 Xo -f- Oi

02«'"= (e°2 — M2 — 1) /^2^^- (^"^~ ''= — 1)

/^l («2 — /^2)

/A2(e°2-M2 — 1)'
a, ^ f/'2 (ilAj «) = «2 /^2J

which give ai = ^/Ji (//, a*"') = — /^i,

Oo = 02 (z^, «"") = — /U.2-

* In the present case the values of the a's giving the identical transformation
are a^^) = Oo'"' = 0.

t Cf. Lie : Transformationsgruppen, III. G07 et seq.

250 PROCEEDINGS OF THE AMERICAN ACADEMY.

Therefore, the a's and a's being taken arbitrarily, we have

where

. «, z= 0, [|H (a, a<"'), «] (k=l,2... r).

But, whereas the parameters a with which we dealt in that part of this
paper, pages 240-248, relating to the first fundamental theorem were not
restricted in range, the present parameters a, involved in the transforma-
tions JE-a defined by equations (A), must be finite, at least if the trans-
formation QTa is to be generated by an infinitesimal transformation. And
if a and a are chosen arbitrarily,

«* = ** (a, «) (k^l,2...r)
may be infinite.

Thus, in the example chosen, if Oi 4^ 0, ai 4^ 0, and a^ + a^ = an
even multiple of 7^^, the transformation

Xi =z fi (a;, a) — a^i -f- a

*^2 —

= tz {x,

a) = e"

is

finite.

while

«i

= dh

(a, a)

-_ «2

+ «2
°2_ 1

/e«2 — 1 \

is infinite ; and the transformation SEa cannot be generated by an infin-
itesimal transformation.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 13. — January, 1900.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.

A REVISION OF THE ATOMIC WEIGHT OF IRON.

PRELIMINARY PAPER.

By Theodore William Richards and (tregory Paul Baxter.

]

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.

A REVISION OF THE ATOMIC WEIGHT OF IRON. —

PRELIMINARY PAPER.

By Theodore William Richards and Gregory Paul Baxter.

Received and presented December 13, 1899.

It is rather surprising that the atomic weight of a metal so important
as iron should have received so little recent attention as it has. Berze-
lius* was among the first to determine the quantity under discussion ; and
in order to accomplish the determination he converted Swedish piano-
wire into ferric oxide. For many years his very low result, 54.3, was
almost unchallenged, when in 1843 Wackenroder called attention to the
fact that in some early analyses of ferric oxide Stromeyer and he had
found the substance to contain only 30.1 per cent of oxygen, correspond-
ing to an atomic weight of nearly 56. This led immediately to several
investigations by Svanberg and Norlin, Berzelius, and Erdmann and
Marchaud, whose individual results upon the same ratio varied from 55.81
to 50.23, showing without a doubt that the low value was incorrect, and
settling the question with all the accuracy necessary in that day. Rivot
in 1850 published two poor analyses which shed no further light upon
the subject; and the more recent work of Dumas, who analyzed the
chlorides of iron in 1859, was no more accurate than his similar work
on other metals. Since that time no work on the atomic weight
worthy of notice has appeared, so that our present knowledge depends
upon data obtained over fifty years ago. In the intervening time Stas
has taught the world accuracy in chemical analysis, and countless com-
plications and sources of inaccuracy have been discovered which then
were unsuspected. The variations of nearly 0.8 per cent in the best de-
terminations (all depending upon the composition of ferric oxide), point
with emphasis toward the necessity of a modern determination. Such

* The references to all the sources of information are to be found iu the critical
discussion at the close of this paper.

254 PROCEEDINGS OF THE AMERICAN ACADEMY.

questions as the possible presence of gases in the oxide, the possibility
of incomplete reduction, and the danger of contamination from the con-
taining vessels, should obviously receive attention.

The balance and weights used in the present study vrere the same as
those used in other recent determinations of atomic weitrhts.* It is need-
less to state that the weights were carefully justified, and that all the
usual precautions necessary to procure accuracy in such work were ob-
served. All the weighings were reduced to the vacuum standard, and the
values of the atomic weights refer to the universally accepted standard
0 = 16.000.

One of the most important steps in such an investigation is clearly the
choice of the substance to be analyzed, and the choice usually requires a
careful preliminary qualitative and quantitative study of many compounds
of the element under consideration. This work occupied a number of
weeks, and in the course of it we found that the preparation of either
ferrous or ferric halides in a pure anhydrous state offered so many diffi-
culties that as yet no satisfactory method for the purpose has been de-
vised. Ferric oxide, on the other hand, can be obtained in a variety of
ways, hence the choice of the early experimenters seems to have been
a wise one. The first step in a systematic treatment of the subject is
obviously a repetition of the discordant work of half a century ago, with
especial reference to the possible errors noted above.

Owing to the fact that most oxides formed from nitrates are known to
include considerable quantities of oxygen and nitrogen, f ignition of ferric
nitrate as a means of obtaining ferric oxide was at first rejected, and ferric
hydroxide Was used instead. Very pure iron ribbon was dissolved in
sulphuric acid, and the metal was then deposited electrolytically from an
ammouic oxalate solution. After the film had been dissolved in the pur-
est nitric acid, the carbon was removed by filtration, and the iron \vas
precipitated as ferric hydrate by the addition of an excess of ammonia
which had been distilled into pure water contained in a platinum dish.
The precipitate was washed thoroughly, collected upon a pure washed
filter paper, separated from the filter paper while moist, and finally dried
in a platinum dish upon the steam bath. The lumps of dried hydrate
were crushed in an agate mortar and finally ignited for several hours in
a current of pure dry air. During the latter operation the weighed
platinum boat with the oxide was contained in a large porcelain tube

* These Proceedings, 26, 242 (IWH).

t Ricliards, These Proceedings, 26, 281 (1891). Richards and Rogers, ibid., 28,
200.

RICHARDS AND BAXTER. — ATOMIC WEIGHT OF IRON.

255

heated by a Fletcher furnace to about 900°. Coiistaut weight within
a few hundredths of a milligram was obtained without difficulty. The
weighed oxide was then reduced in a current of electrolytic hydrogen
which had been dried by means of fused potassic hydroxide. Ferric
oxide is only very slowly reduced at the highest temperature which can
be used with glass tubes, so that this reduction also was of necessity
carried on in porcelain tubes. Even at about 900°,* long continued
heating wms needed to complete the reduction, the total period of ignition
amounting sometimes to twenty hours. The progress of the reaction was
sliglitly accelerated by alternate oxidation and reduction.

Two analyses, carried out in the fashion described, are tabulated be-
low. Corrections to a vacuum standard have been applied by adding
0.00009 gram to every apparent gram of ferric oxide, and O.OOOOl
gram to every apparent gram of metallic ii'on, the specific gravities of
these two substances being assumed to be 5.2 and 7.9 respectively, f

SERIES I.

Number of
Analysis.

Weight of Ferric
Oxide in Vacuum.

Weiglit of Iron
iu Vacuum.

Atomic Weight
of Iron.

1

2

3.17485
3.61235

2.22096
2.52750

55.885
55.916

Average . . . .

. 55.900

The fact that both of these results are considerably lower than the
accepted value of the atomic weight of iron (56.0) seemed to confirm a
suspicion which had already arisen that ferric hydrate might not be
completely converted into the oxide by ignition at a high temperature.
That the oxide upon ignition easily reached constant weight is, however,
evidence in the other direction. Since no direct method of proving the
presence or absence of retained water was at hand, it seemed best to
make use of another method of preparing ferric oxide. Although, as
has been already stated, oxides formed from nitrates occlude nitrogen
and oxygen, yet these gases are evolved upon solution of the oxides and

* At the temperature to which the porcelain tubes were heated pure silver did
not melt, while sodic chloride was easily fused.

t Landolt und Br.rnstein Tabellen, 118, 133 (1894).
VOL. XXXV. — 17

256 PROCEEDINGS OP THE AMERICAN ACADEMY.

can be determined. Ignition of ferric nitrate was therefore chosen as
the next method of obtaining ferric oxide.

The material used in the following work was purified with great care.
In the first place a solution of a so-called " chemically pure " chloride of
iron was treated with an excess of sulphuretted hydrogen and the result-
ing sulphides and sulphur were removed by filtration. From the filtrate,
after oxidation with nitric acid, the iron was precipitated with an excess
of amraonic hydrate and the precipitate was thoroughly washed by de-
cautation. In this process traces of a number of metals which might
have been present must have been removed. The precipitate was dis-
solved in sulphuric acid, and by cautious addition of ammonic hydrate
basic ferric sulphate was thrown down ; a precipitate which is much
more easily washed than ferric hydrate, but which was redissolved in
sulphuric acid with considerable difiiculty. In order to reduce to tlie
ferrous state the ferric sulphate thus formed, the solution was next sub-
jected to the action of a galvanic current of several amperes. The solu-
tion was contained in a large platinum dish which served as the negative
electrode, the positive electrode being a flat spiral of platinum wire.
Since the solution was very concentrated and contained a considerable
excess of sulphuric acid, the greater part of the ferrous sulphate crystal-
lized out when the solution was allowed to cool. By alternately elec-
trolyzing and cooling, almost all of the iron was eventually obtained as
ferrous sulphate.

The next process in the purification was to deposit the iron electrolyti-
cally from a solution of its oxalate. By means of this separation, the
aluminum and manganese could be eliminated, since the first, together
with alkaline and other impurities, is not deposited by electrolysis, while
the second is deposited upon the anode. The ammonic oxalate needed
was prepared with great care. Oxalic acid which had been repeatedly
crystallized with hydrochloric acid, and then until free from chlorine, was
saturated with pure redistilled ammonia, and the resulting ammonic oxa-
late was subjected to several crystallizations. To a hot concentrated so-
lution of this ammonic oxalate, the ferrous sulphate was added as long as
no permanent precipitate resulted, and the solution was then electrolyzed
until the greater part of the iron had been deposited. A slight deposit
of hydrated manganese peroxide appeared upon the positive pole, which,
however, presented enough surface to prevent the deposit from becoming
detached. The film of iron was carefully washed and dissolved in pure
dilute nitric acid. When all of the ferrous sulphate had been thus con-
verted into ferric nitrate, the solution of the nitrate was filtered and

RICHARDS AND BAXTER. ATOMIC WEIGHT OF IRON. 257

evaporated, and the salt was three times recry stall ized from a large
amount of nitric acid. This acid had been distilled in a platinum still
and condenser and collected in a platinum receiver. The mother liquor
of the first crystallization vpas analyzed for impurities as follows : After
precipitation of the iron as ferric hydrate by the addition of an excess of
pure ammonia and filtration, the filtrate was evaporated to dryness and
the residue was ignited. Since only a negligible amount of alkaline
solid remained, it can fairly be assumed that the final preparation of ferric
nitrate must have been free from such impurity. During the latter part
of this purification platinum vessels only were employed, and the water
and reagents were the purest obtainable.

The next stejD, that of converting the nitrate into oxide, was carried
out in part by gentle heating in an electric oven at about 150°.* The
resulting cake of basic nitrate, after removal from the platinum dish in
which it was contained during the above operation, was crushed in an
agate mortar, and the reaction was completed by ignition in an open
porcelain tube at a temperature of about 900° for several hours. The
oxide, which was now contained in a large platinum boat, was carefully
weighed and again heated. The loss of weight during this second igni-
tion amounted to one tenth of a milligram, while a third prolonged
heating caused a further loss of three-tenths of a milligram. This differ-
ence must have been caused by the escape of included gases, and was sO'
small in comparison with the weight of the oxide (seventeen grams), that
further heating was deemed unnecessary.

The determination of the gases contained in this oxide proved to be a
matter of considerable difficulty. At first the method used was that of
dissolving the oxide in a Sprengel vacuum in fused acid potassic sulphate.
Owing to the fact that the sulphuric acid, which was evolved from the
fused sulphate, attacked both the rubber connections used for joining the
apparatus, and also the stick potash used as a drying agent, it was uncei--
tain whether the small quantity of gas which was collected in the air
pump was originally included in the oxide. A more successful method
was eventually found by dissolving the oxide in hot hydrochloric acid con-
taining stannous chloride. The operation was carried on in an apparatus
for determining the gases' evolved during the solution of a substance,
which has been already described in papers from this laboratory. f In
two experiments half a gram of ferric oxide yielded a mere trace of gas.
Cupric oxide treated in the same way evolved considerable gas, less, how-

* Richards, Am. Cliem. Jour., 22, 45 (1899).
t These Proceedings, 28, 200.

258

PROCEEDINGS OF THE AMERICAN ACADEMY.

ever, than that evolved when the same cupric oxide was dissolved in sul-
phuric acid. It was shown, however, that the gas which resulted upon
solution in hydrochloric acid and stannous chloride consisted wholly of
nitrogen, and that the volume of this nitrogen corresponded to the nitro-
gen in the gases set free by solution of the same oxide in sulphuric acid.
Although it was evident that the oxygen was absorbed by the stannous
chloride, yet, since in the case of all oxides heretofore examined the in-
cluded gases consisted chiefly of nitrogen, it is fair to conclude that this
was so in the case of ferric oxide. Only a trace even of nitrogen was
evolved from the ferric oxide, hence this preparation must have been
practically free from included gases.

In order to avoid any error due to hygroscopicity, the oxide used for
each analysis was bottled in dry air in the bottling apparatus, which has
already been described elsewhere.* The platinum boat containing the
oxide was heated in the hard glass tube while the air was exhausted from
the apparatus by means of a Sprengel air pump. Dry air was then
admitted, and the boat was pushed into the weighing bottle which was
then stoppered and weighed with the usual precautions. Since in two
instances repetition of the bottling failed to change the weight of the
oxide, subsequently this precaution was omitted.

The oxide was then reduced in the manner already described, constant
weight of the metal being attained much more easily than in the case of
■oxide prepared from ferric hydrate. Five closely agreeing results were
obtained from analyses of material prepared as above.

SERIES II.

Number of
Analysis.

Weight of Ferric
Oxide iu Vacuum.

Weight of Iron
in Vacuum.

Atomic Weiglat
of Irou.

3

3.97557

2.78115

55.883

4

4.89655

3.42558

55.891

6

4.35955

3.04990

55.891

6

7.14115

4.99533

55.870

7

6.42021

4.49130

55.882

Average . . . .

. 55.883

* These Proceedings, 32, 59 (1896).

RICHARDS AND BAXTP:R. ATOMIC WEIGHT OP IRON. 259

In order to determine if the reduction had been complete in the hist
two analyses of this series, the resulting metal was dissolved in cold dilute
sulphuric acid. In each case a few tenths of a milligram of black insolu-
ble matter remained, which contained platinum. This impurity probably
came from the alloying of the boat with the iron during the reduction,
and hence introduced no error. A trace of ferric oxide also was found.
Even if the platinum had been originally present, the combined weight
of the two impurities could not have raised the final result more than
0.01 per cent; hence no correction was applied.

One of us* has recently shown that iron when ignited at a high tem-
perature, has practically no tendency to occlude hydrogen, so that in the
present case this possible source of error is excluded. On comparing
these results with the two determinations of the first series, one is forced
to conclude that the oxide made from the nitrate has essentially the same
composition as the oxide made from the hydroxide. The higher value
indicated by Experiment 2, was probably caused by incomplete reduction,
and the first experiment agrees with the average of the second series.

It is interesting to review the older determinations f in the light of the
experience gained in these analyses. Evidently the occlusion (or better,

* Baxter, Am. Chem. Journ., 22, 363 (1899).

t A complete list of these determinations is given below : — Fe =

1814. WoUaston, from Thenard's Results, Phil. Trans., 104, 21 (1814) . . 55.2
1826. Stromeyer, Pogg. Ann , 6, 471 (1826). Reduction Tea O3 : Fe . . . 55.&
1826. Berzelius, Pogg. Ann., 8, 185 (1826). Oxidation Fe^ : Fe., 0;3 . . . 54.3
1830±. Wackenrodcr, Archive Pharm., 35, 279 ; 36, 22 (1843). Reduction

Feo Og : Fe. 55.9

1844. Svanberg and Norlin, Liebig's Ann. Chem. Pharm., 50, 432 (1844),
also Berzelius Jahresbericht, 24, 121 (1845), and 25, 42 (1846).

Oxidation Fe., : Fe. O3 . . 55.87

Reduction Fe, O3 : Fe. 56.09

1844. Berzelius (the same references). Oxidation Fe, : Fe. O3 .... 56.02
1814. Erdmann and Marchand, J. Pr. Ch., 33, 1; Lieb. Ann., 52, 212

(1845). Reduction of oxide from oxalate Fe2 O3 : Fe., 56.02

1846. Maumene, Ann. Chim. Phys. [3], 30, 380 (1850). Oxidation Fe, :

Fe., O3 ' . 56.00

1850. Rivot, Ann. Chim. Phys. [3], 30, 192, (1850). Reduction Fe.^ O3 : Fe 54.2
18.59. Dumas, Ann. Chim. Phys. [3], 55, 157 (18-59). Lieb. Ann., 113, 26

(1860). Analysis of Chlorides 56 2

[The results of Magnus, Dobereiner, Capitaine (1841), and Torrey (1888), alluded
to by Clarke, are not worthy of serious consideration. See Smithson. Misc. Coll. ;
Constants of Nature, 5, 287 (1897) ; also 4, 65 (Becker, 1880)].

Frr)m some of these values Clarke (1897) computed the value 56.021, while
Messrs. Landolt, Ostwald, and Seubert [1898] chose the number 56.0.

260 PROCEEDINGS OF THE AMERICAN ACADEMY.

inclusion) of gases by the oxide is a less serious cause of error in the case
of iron than in most other cases, if the temperature of ignition is reason-
ably high. This is the chief probable error which would tend to make
the results too low. The three most probable errors having the opposite
effect are : — first, the possible presence of magnetic oxide in the ferric
oxide ; second, the possibility of incomplete reduction during the analysis
of this substance; and third, the possible presence of alkaline, siliceous,
or other non-reducible material. Wackenroder seems to have been the
only experimenter who thought seriously of the first two of these errors;
and none of the early experimenters paid any attention to the last, for
until recently glass was supposed by most chemists to be wholly insoluble.
Wackenroder's work was perhaps the most intelligently carried out of all
the older determinations, although his individual analyses did not agree
among themselves quite so well as some of the others. His greatest
omission was the recognized lack of purity in his hydrogen, although he
could not observe an error due to this cause. It is interesting to note
that the average of his six results gives the value 55.82 for the atomic
weight in question, while a corrected average obtained by rejecting the
two most discordant values gives the value 55.92. The result of the
present work indicates a value midway between the two averages.
Wackenroder's valuable work does not enter into Clarke's average for the
atomic weight of iron, for Clarke was unable to find the original paper.

In order to prepare pure iron for his later work, Berzelius fused his
metal with ferrous oxide, but gave no proof of the effectiveness of his
treatment. It is not impossible that traces of the oxide may have been
held by the iron, as copper dissolves cuprous oxide. Erdmann and Mar-
chand made the ferric oxide, which they reduced quantitatively, by the
ignition of ferrous oxalate , — a method which invites the presence of
magnetic oxide. Maumene's method was the process of filtration used
in common analysis, in which the possible errors in both directions are
plentiful.

These detailed criticisms, taken in connection with the general lack of
accuracy which is to be observed in the quantitative work of half a cen-
tury ago, seem to show that there is nothing improbable in the present
result, 55.88. Of course the analysis of a single compound is never con-
clusive, so that this result is announced only as a preliminary one, which
we shall hope to support or disprove in the near future.

Harvard University, Cambridge, Mass.,
October, 1899.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 11. — January, 1900.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF
HARVARD COLLEGE.

ON CERTAIN COLORED SUBSTANCES DERIVED FROM

NITRO COMPOUNDS.

THIRD PAPER.

By C. Loring Jackson and F. H. Gazzolo.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.

ON CERTAIN COLORED SUBSTANCES DERIVED FROM
NITRO COMPOUNDS. — THIRD PAPER.

By C. Loring Jackson and F. H. Gazzolo.

Presented December 13, 1899. Received December 30, 1899.

The colored substances formed by the action of sodic alcoholates and
certain nitro compounds have been studied by Victor Meyer,* Lobry de
Bruyn,t and in this Laboratory, t but as yet no satisfactory constitutional
formula has been assigned to them.

In continuing this investigation, we tried first to replace the sodic
alcoholates by other similar reagents, and succeeded in obtaining colored
products from trinitranisol or trinitrobenzol by the action of sodic malo-
nic ester, sodic acetacetic ester, sodic phenylate, the sodium compound of
benzylcyanide, and perhaps the sodium compound of phloroglucine, al-
though in this last case the action was not well marked. As it has been
shown already that similar compounds are formed with various sodic
alcoholates, § and even with sodic hydrate, || it appears that this behavior
with nitro compounds is a very general reaction of alkaline substances.

Of these new colored products only those with sodic malonic ester or
sodic acetacetic ester were stable enough to be prepared for analysis, but
they were unusually stable for bodies of this class. All four of the sub-
stances formed from trinitranisol or trinitrobenzol and these two sodium
esters were analyzed, and were proved to consist of three molecules of
the sodium ester combined with one of the trinitro compound ; for in-
stance, the malonic ester trinitrobenzol compound has this formula,

* Ber. d. chem. Ges., XXVII. 3153; XXIX. 848.

t Rec. Trav. Chim. Pays-Bas, XIV. 89, 150; XV. 848.

t Jackson and Ittner, Am. Chem. Journ., XIX. 109, where a historical account
of the previous work is given ; Jackson and Boos, These Proceedings, XXXIII.
173.

§ These Proceedings, XXXIII. 173.

II Hepp, Ann. Chem. (Liebig), CCXV. 359.

264 PROCEEDINGS OF THE AMERICAN ACADEMY.

C6H3(N02)n[CHNa(COOC2H5)o]3. The formation of compounds with
three molecules of the sodium constituent is noteworthy, since all the
compounds analyzed heretofore have contained the two constituents in
the proportion of one molecule of each. Similar experiments with sodic
methylate, ethylate, or amylate and trinitrobenzol also led to products
apparently containing three molecules of the alcoholate to each molecule
of the nitro compound * — a surprising result, since Lobry de Bruyn and
Van Leent f obtained from trinitrobenzol a substance with the following
formula, C6H3(N02)3KOCH3|^H20. The difference in the results is un-
questionably due to differences in the method of preparation. Lobry de
Bruyn and Van Leent's compound was obtained by crystallization,
whereas all our products with three molecules of the alkaline material
were precipitated from an alcoholic solution with benzol. Experiments
are now in progress to test this explanation of the phenomena.

The discovery of these sodic malonic or acetacetic comjiounds would
furnish a strong argument, if that were needed, against the only theory
for these colored substances as yet published, — that of Victor Meyer, t
who supposed they were formed b}' the replacement of atoms of hydro-
gen on the benzol ring by atoms of sodium. This theory has been dis-
proved by the observations of Lobry de Bruyn, § supported by those
made in this Laboratory ; || and among other arguments the point was
made that Victor Meyer's theory necessitated the assumption of alcohol
of crystallization in every compound of this class which had been ana-
lyzed. In these malonic and acetacetic compounds the presence of malonic
ester or acetacetic ester of crystallization must be assumed, if this theory
is adopted ; and, further, the number of molecules of " ester of crystalli-
zation " corresponds in each case to the number of atoms of sodium ; the
view, therefoie, that the colored bodies are addition and not substitution
compounds is confirmed by these observations.

Other experiments were tried to study the effect on the formation of
the colors of increasing or diminishing the negative nature of the aromatic
constituent. That there is some effect of this sort has been shown already,

* The ethyl and methyl compounds seemed to contain alcohol of crystallization,
to judge from the percentages of sodium obtained. The publication of these results
will, therefore, be postponed until further analytical data have been collected. The
amyl compound, on the other hand, gave a percentage of sodium corresponding

to C6H3(N02)3(Na005Hu)8-

t Rec. Trav. Chira. Pay.s-Bas, XIV. 150.
t Ber. d. chem. Ges., XXVII. 3153.
§ Rec. Trav. Chim. Pays-Bas, XIV. 89.
II These Proceedings, XXXIII. 174.

JACKSON AND GAZZOLO, — NITRO COMPOUNDS. 265

since certain substituted toluols give less stable colored derivatives than
the corresponding benzoic acids.* Picramide, the first substance selected
for this work, gave colored compounds with sodic methylate or sodic
malonic ester, but too unstable to analyze, whereas trinitranisol or tri-
uitrobenzol, in which the negative character of the nitro groups is not
weakened by the presence of a positive radical like NHo, gave stable
well marked colors. Dinitroxylol ((CH3)., 1. 3. (NOo)^ 4. 6.) also gave a
slight and evanescent coloration with sodic methylate, and no reaction
with sodic malonic ester, whereas trinitroxylol ((CIi3)2l.3. (N02)3 2. 4. 6.)
gave colored compounds with both these reagents, which, although much
more stable, could not be prepared for analysis. These results, there-
fore, as far as they go, show that an increase in the negative nature of the
aromatic constituent increases the tendency to form colored compounds.

The next subject considered by us was the effect of the presence of
methyl groups attached to the benzol ring on the formation of colors.
Dinitrotoluol ((^03)2 2. 4.) gave colored compounds with sodic methy-
late or sodic malonic ester ; dinitroxylol ((€113)2 1. 3. (NO-2)2 4. 6.) gave
only a passing coloration f with sodic methylate, none at all with sodic
malonic ester ; and dinitromesitylene gave no color with either reagent.
Trinitroxylol gave strong color reactions with both reagents, trinitrome-
sitylene none whatever. It is evident, therefore, that the presence of
methyl groups on the benzol ring diminishes the tendency to form these
colored compounds. Whether this effect is due to a specific action t of
the methyl group, or to the fact that these groups stand in the ortho
position to the nitro groups, or to both these causes, cannot be deter-
mined from the facts at present at our disposal.

In consideration of the complete absence of a color reaction with
trinitromesitylene and sodic methylate, it is interesting to note that M.
Konowalow § obtained red salts from nitromesitylenes in which one of
the nitro groups stands in the side chain. We cannot find that he ana-
lyzed these salts to determine whether they were true salts or addition
products with sodic hydrate. If the latter, they would have a strong
bearing on the discussion given above.

Another series of experiments was tried with aromatic bodies rich in

* Am. Chem. Journ. (Remsen), XIX. 201.

t This may have been due to a small quantity of a thiophene compound. As a
rule we have not considered that a colored product belonged to the class we are
studying unless we could obtain a copious precipitate of it with benzol.

I Lobry de Bruyn, Rec. Trav. Chim. Pays-Bas, XIV. 95.

§ Ber. d. chem. Ges., 1896, 2204.

266 PROCEEDINGS OF THE AMERICAN ACADEMY.

negative radicals but containing no nitro groups ; for, if colors of the
same class could be obtained from these, it would 2:)rove that the addition
of the alkaline substance took place on the benzol ring and not on the
nitro group. We were encouraged to undertake these experiments by
the striking resemblance in properties * between our colored products
and the green bodies made by Astre f from the action of sodic alcoholates
on quinone. In the quinones, however, the formation of hemiacetals t is
possible, and it may be that the green bodies belonged to this class ; we
accordingly used for our new experiments substances in which the forma-
tion of hemiacetals could not occur, such as trimesic triethylester, which
is especially fit for these experiments, since it has a still stronger re-
semblance to trinitrobenzol than quinone has, because it contains three
negative radicals symmetrically disposed. "We have not succeeded in
obtaining any colored or other addition-products from this substance, or
from the free trimesic acid, although the attempts have been repeated
often and under varying conditions. Nor did we have better success
with other bodies free from nitro groups, such as phloroglucine, or resor-
cine ; pyrocatechine, it is true, gave a temporary coloration with sodic
methylate, but we think this reaction does not belong to the series under
discussion. These experiments, as they have given negative results,
throw no light on the constitution of our colored compounds.

It has been shown earlier in this paper that the only theory as yet
proposed for these colored compounds (that of Victor Meyer) is inadmiss-
ible, because they are addition-, not substitution-products. The facts
now at our disposal are not sufficient to furnish an absolute proof of the
structure of these compounds, but it is possible to show that certain
constitutional formulas explain these facts better than others, and it seems
to us that the work has arrived at a point where such a discussion of the
possible formulas will be useful. In this discussion the following proper-
ties must be considered, as they seem to be characteristic of all the
members of this group. (1) The very marked color. (2) The ease
with which they are decomposed even by dilute acids, giving the aromatic
constituent unaltered. (3) Their behavior with alcohols, which we
describe here in some detail, because the principal observations are new.
When the methyl compound C6Ho(NOo)30CH3NaOCH3 is allowed to
stand for some time with benzyl alcohol, both the methyls are replaced
by benzyls, and the compound C6H2(N02)30C7H7NaOC7H7 is formed.

* These Proceedings, XXXIII. 175.
t Comptes Rendus, CXXI. 530 (1895).
I These Proceedings, XXX. 41 1 .

JACKSON AND GAZZOLO. — NITRO COMPOUNDS. 267

Conversely this benzyl compound is converted into the corresponding
methyl compound if boiled with methyl alcohol. In the same way * the
methyl is converted into the ethyl compound by crystallization from
common alcohol. f

There are three possible ways in which these compounds can be
formed : — First, the addition of the sodic methylate (or other alkaline
substance) may take place upon the carbon atoms of the benzol ring.
Second^ it may take place on the nitro group alone. Third, it may take
place partly on the nitro group and partly on the carbon of the benzol
riucr.

The first method of addition, that on the carbon alone, seems to us
much less probable than the second or third, in which a nitro group takes
part, especially since the work of Nef and others has shown that the
sodium is attached to the nitro group in the sodium salt of nitromethane.
As we have succeeded in finding no analogous case in which an alkaline
substance is added to carbon atoms with the formation of a strongly col-
ored product, we think that this first hypothesis is not worthy of a
detailed discussion.

Turning to the formulas in which the nitro group is affected, we have
the second method of addition, in which the sodic methylate is attached
to the nitro group only ; this would give rise to a structure such as the
following : | —

I.

OCH3

C6H20CH3(NO.,)o— N

ll\

O ONa;

while the third hypothesis, according to which both the nitro group
and the carbon of the benzol ring take part in the addition, would be

* These Proceedings, XXXIII. 177.

t Some experiments of less importance may be mentioned here, witli the
remark that they are not incompatible with tlie formula adopted later as
giving the best explanation of the observed facts. Bromine decomposes
CsH3(NO.,)3[CHNa(COOC.,H5)2]3, giving trinitrobenzol as one of the prod-
ucts of the reaction. No salts with other basic radicals could be obtained
from C,.H.,(NOo)30CH3XaOCH3. No sodic iodide was formed by heating
CsH3(N02)3[CHNa(COOCoH5)._,]3 with ethyl iodide to 140°. On tlie other hand,
it looked as if benzoyl chloride acted on these bodies, but the end of the college
year prevented us from studying this reaction.

t This is analogous to that given by Hantzsch and Rinckenberger, Ber. d. chem.
Ges., XXXII. 628, for their dinitroethanester acid.

268

PROCEEDINGS OF THE AMERICAN ACADEMY.

represented by the formulas given below, in which it is supposed that an
isonitro * compound is formed with the development of a quinoid structure
in the benzol ring. The difference between the two formulas is that in
II. the quiuoid structure is developed in the para position, in III. in the
ortho position.

II. III.

0=N-0-Na

0=N-0-Na

We have used the formula of the addition product from sodic methylate
and trinitranisol, as it is the simplest that will serve in the argument
which follows. In applying these formulas to the malonic ester com-
pounds it must be assumed that the malonic ester radical which is added
to the benzol ring has the constitution -OC(OC2H5)=CIiCOOC2H5, as,
if it is assumed to be -CH(COOC.,H5)2, we should have an attachment of
carbon to carbon incompatible with the instability of these compounds.

In applying Formulas I., II., and III. to the explanation of the observed
properties of these compounds, we consider first the strong color, their
most marked characteristic ; this is explained by the quinoid structure in
Formulas II. or III., but is not accounted for by Formula I., since, accord-
ing to Hantzsch and Riuckenberger,t their substance

CH.3

\ /

CH-NO

OC2H5

NO,

/

\

OK,

which contains the group characteristic of Formula I., has only a pale
yellow color.

The easy decomposition of the colored body by hydrochloric acid with
regeneration of the trinitranisol, from which it was formed, is accounted
for by either of the three formulas I., II., or III., but the preference should
be given to II. or III., since Hantzsch and Rinckenberger t state that
their

* Compare Hantzsch, Ber. d. chem. Ges., XXXII. 575-641, and also the ortho-
benzoldioxime of Zincke and Schwarz, Ann. Chem. (Liebig), CCCVII. 28.
t Ber. d. chem. Ges., XXXII. 628. t Ibid.

JACKSON AND GAZZOLO. — NITRO COMPOUNDS. 269

CH3

OC^Hfi

\
CH-

-NO

NO2

\
OH

is a true stable acid ten times as strong as acetic acid, and it is fair to '
suppose, therefore, that the substance

OCII3

C6H.,OCH3(N02)oNO

\
OH

(formed by hydrochloric acid on our sodium salt, if it has Formula I.)
would also be comparatively stable, and not drop at once into the trini-
tranisol, which, as a matter of fact, is formed immediately by the action
of hydrochloric acid on the colored compound. On the other hand, this
rapid decomposition by acid would be explained according to Formula
II. or III. by the strong tendency of quinoid bodies to pass into the hy-
droquinoid form, which might easily cause the splitting off of methyl
alcohol as soon as the atom of sodium was replaced by hydrogen. This
rapid decomposition with removal of methyl alcohol when the colored
bodies are treated with dilute hydrochloric acid recalls the similar behav-
ior of the dichlordimethoxyquinonedimethylhemiacetal ; * and the simi-
larity of these phenomena may tell in favor of classing the colored salts
with quinone derivatives, although the two reactions are not strictly
analogous.

The third point in favor of Formula II. or IH. is the replacement of
the two methyls in C6H^(N02);30CH3NaOCH3 by benzyls when the com-
pound is soaked in benzyl alcohol, and the reverse change when the
benzyl compound is boiled with methyl alcohol. As under the same
conditions benzyl alcohol has no action on methyl picrate, or methyl
alcohol on benzyl picrate, it is obvious that the complete replacement of
one radical by the other here depends on the structure of the addition
product ; and, whereas Formula I. gives no reason why the change should
proceed beyond the methyl attached to the nitro group, it is easy to
see that in a substance constituted like Formula II. or III. any reagent
which affected one methyl would act in a similar way on the other,
so that the methyl compound would be completely converted into
C6H2(NO,)30C,H,NaOC7H7.

* These Proceedings, XXX. 431.

270 PROCEEDINGS OP THE AMERICAN ACADEMY.

The inferences drawn in the foregoing discussion may be briefly reca-
pitulated as follows : It is improbable that the sodic methylate is added
to the carbon of the benzol ring only. A quinoid formula (II. or III.)
explains the observed facts better than one in which the sodic methylate
is added to the nitro group alone (I.), but this latter structure is not
definitely excluded. Under these circumstances we think it would be
premature to contrive names for these colored bodies, or to give struc-
tural formulas in the experimental part of this paper. We hope that the
continuation of the work, now in progress in this Laboratory, will
definitely settle the constitution of these colored substances.

Experimental Part.

Preparation of Picryl Chloride.

As the method of making picryl chloride used by us in this work was
an improvement on that given by Pisani,* we describe it. 25 grams of
dry picric acid were mixed with 50 grams of phosplioric pentachloride in
a large Erlenmeyer flask provided with an air condenser, and heated on
the water bath until the violent reaction had ceased, and the contents
had assumed a very dark brown color. When cold, the flask was sur-
rounded with ice, and its contents treated with ice water, care being
taken to avoid any considerable rise of temperature. The precipitate
formed in this way was filtered out, dried, washed with ether, and crys-
tallized from a mixture of benzol and alcohol to purify it. The advan-
tages in our method are that there is a considerable saving of time, and
there is much less danger that the substance will be converted into a
tarry decomposition product, as happens in Pisani's method if the heat
runs too high in either the preparation or the removal of the phosphoric
oxychloride by distillation.

Action of Sodic Acetacetic Ester with Trinitranisol.

In our first experiment in this direction we prepared our sodic acet-
acetic ester with sodic methylate, and obtained a red precipitate which
gave the following result on analysis : —

0.2438 gram of the substance gave 0.0576 gram of sodic sulphate.

Calculated for CeH,(N0o)30Cn3Na0CH3. Found.

Sodium 7.69 7.66

* Ann. Chem. (Liebig), XCII. 326.

JACKSON AND GAZZOLO. — NITRO COMPOUNDS, 271

It was evident, therefore, that we had only the color formed from sodic
methylate, and that the acetacetic ester took no part in the reaction. In
order, then, to obtain an acetacetic ester addition, it was obviously neces-
sary to exclude all alcohol and alcoholates ; we accordingly proceeded as
follows. To an excess of acetacetic ester mixed with beuzol a quantity
of sodium in the form of ribbon was added (iu our later preparations the
amount of sodium used provided three atoms of it to each molecule of
triuitranisol). After the sodium had disappeared, the liquid thus obtained
was added drop by drop to a benzol solution of triuitranisol. It is un-
necessary to say that absolute benzol was used iu all this work. The
first drop imparted a deep vermilion color to the solution, and this color
became more and more intense as the reaction proceeded. During the
process the mixture was kept cool by surrounding the beaker with ice.
After all the sodic acetacetic ester had been added, the liquid was mixed
with an excess of anhydrous benzol, which threw down a semi-gelatinous
or oily precipitate. This was filtered out, washed with beuzol, and
pressed upon a porous plate, all these operations being carried, on as
quickly as possible.' The dark colored dried product crumbled easily
into a red amorphous powder of a much darker color than the addition
product from sodic methylate. It was dried in vacuo, and analyzed with
the following results : — •

I. 0.2376 gram of the substance gave 0.0740 gram of sodic sulphate.

11. 0.2196 gram of the substance gave 0.0697 gram of sodic sulphate.

III. 0.1928 gram of the substance gave 0.0700 gram of sodic sulphate.

IV. 0.2596 gram of the substance gave on combustion 0.4028 gram of

carbonic dioxide and 0.1148 gram of water. In this combustion
the substance was mixed with chromic oxide to drive out car-
bonic dioxide from the carbonate formed, and was spread out in
a long copper boat, which was heated gently and gradually to
avoid explosions.

Calculated for
CeH,(NO,)30CH3(CH3COCHNaC00C2H5)3 I.

Sodium 9.87 10.10

Found.
II. III.

10.28 11.75

IV.

Carbon 42.91

42.32

Hydrogen 4.57

4.91

There can be no doubt, therefore, as each analysis is of the product of
a separate preparation, that the substance is a definite compound, and is
formed by the addition of three molecules of sodic acetacetic ester to
one of trinitranisol. The variation in the percentages of sodium in the

272 PROCEEDINGS OF THE AMERICAN ACADEMY.

different specimens is no more than would be expected, when it is remem-
bered that the product was purified only by washing with benzol.

Properties of the Addition Product of Trinitranisol and Sodic Acetacetic
Ester, C6H2(N02)30CH3(CH3COCHNaCOOC2H5)3.

This substance forms a deep crimson powder, which we have not suc-
ceeded in bringing into a crystalline state. It is decidedly stable for a
body of this class, keeping for even several days in a desiccator, but
finally decomposing into a black tar. When heated it is slightly explo-
sive. It dissolves completely in water without decomposition, to judge
from the color ; is soluble in common alcohol, but gives a turbid solu-
tion ; on the other hand, it dissolves, forming a clear solution, in methyl
alcohol; soluble in acetone; insoluble in benzol, ether, chloroform, car-
bonic disulphide, or ligroin. Acids decompose it instantly, as was shown
by the destruction of the color.

Action of Sodic Malonic Ester on Trinitranisol.

Two jrrams of trinitranisol dissolved in absolute benzol were mixed
with a benzol solution of 4.5 grams of sodic malonic ester prepared by
the direct action of sodium on the malonic ester, — that is, three mole-
cules of the sodium ester to each molecule of trinitranisol. As the two
solutions came together, an intense cherry red color appeared, with the
formation of a thick gelatinous precipitate of the same color, which
increased in volume and deepened in color as the reaction continued.
After the mixture had stood some time at ordinary temperatures, a large
enough quantity of benzol was added to produce complete precipitation,
the product was then filtered rapidly, washed with benzol till the filtrate
was colorless, pressed quickly on the porous plate, and dried in vacuo.
This reaction seemed to run more quickly and cleanly than the corre-
sponding one with sodic acetacetic ester, giving a purer product which
was very easily handled and washed.

I. 0.2028 gram of the substance gave 0.0.536 gram of sodic sulphate.
II. 0.3027 gram of the substance gave 0.0796 gram of sodic sulphate.

Caloulatpfl for Found.

CeH2(NO2)30CH,,[CUN'a(C00C2H5)2]3 I. II.

Sodium 8.74 " 8.56 8.52

JACKSON AND GAZZOLO. — NITRO COMPOUNDS. 273

Properties of the Addition Product of Trinitranisol and Sodic Malonic
Ester, C6H2(N02)30CH3[CPINa(COOC2U5)2]3.

This substance is an amorphous powder with a deep maroon color.
"We have not succeeded in crystallizing it. It is one of the most stable
bodies of its class, as when exposed to the air it usually remains unaltered
for nearly five days ; at the end of this time it begins to grow moist, then
turns black, and is finally converted into a black powder with a some-
what tarry consistency. When heated, it explodes with a slight puff, but
with little or no noise ; it is, however, apparently stable at as high a
temperature as 140°. It dissolves completely in water, forming a clear
cherry red solution ; soluble, although more slowly, in ethyl alcohol ;
completely and quickly soluble in methyl alcohol, but this solution seems
to be attended by some decomposition, as a fading of the color was ob
served ; soluble in acetone ; insoluble iu ether, benzol, chloroform, car
bonic disulphide, or ligroin. A few drops of hydrochloric acid added to
its aqueous solution changes the red color to yellow instantly, and causes
a precipitate which, on filtration, solution in alcohol, and evaporation of
the solvent, proves to be a reddish oil containing malonic ester, to judge
from the smell, and trinitranisol, since this substance crystallizes out on
standing.

As this substance was more stable than most others of its class, we
tried the action of ethyliodide upon it in the hope of replacing the atoms
of sodium with ethyl. For this purpose 0.5 gram of the addition prod-
uct was heated in a sealed tube with ethyl iodide, at first to 100°, but,
as this produced no apparent effect, later to. 140° for an hour and a half,
and then it was kept at 100° for two days. The contents of the tube
were treated with benzol, after the ethyl iodide had evaporated, which
gave a red solution and a black residue ; the residue was extracted with
water, and the extract gave no test for an iodide. It is obvious, there-
fore, that the ethyl iodide had not acted at all, but that the unmanage-
able black product was produced by the decomposition of the addition
compound.

Action of Sodic Malonic Ester with Trinitrohenzol.

The sodic malonic ester was prepared with sodium alone, benzol was
used as the solvent, and the proportions were three molecules of the
ester to one of the trinitrobenzol. As soon as the solutions were mixed,
a deep scarlet lumpy precipitate was formed ; it was found best, there-
fore, to add the solution of the sodic malonic ester in small portions at a

VOL. XXXV. — 18

274 PROCEEDINGS OF THE AMERICAN ACADEMY.

time with constant stirring. Tiie beaker was cooled by immersing it in
ice. The precipitate was washed with benzol until the filtrate was color-
less, and then dried on a porous plate and in vacuo. Analyses I. and II.
are of two different products prepared in this way. As in these prepara-
tions and the other similar ones described in this paper we had used three
molecules of the sodium compound to one of the nitro body, there seemed
some danger that our products might be not definite compounds, but
mixtures of an addition compound containing only one atom of sodium,
with the two additional molecules of the sodic malonic ester (or the
corresponding reagent) precipitated by the large excess of the benzol.
This objection to our results did not seem a very important one, because
they agreed better with the theoretical numbers than would be probable if
this theory were true, but we felt that it was necessary to test it by ex-
periment, and for this purpose repeated the preparation, using two mole-
cules of sodic malonic ester to each molecule of trinitrobenzol (1 gram of
trinitrobenzol and 1.7 grams of the sodic malonic ester). Analysis III.
was made with the specimen prepared in this way, and proves that our
substances are definite compounds and not mixtures, since it agrees with
those prepared with three molecules of the sodium ester.

I. 0.2038 gram of the substance gave 0.0544 gram of sodic sulphate.

II. 0.2154 gram of the substance gave 0.0590 gram of sodic sulphate.

III. 0.2630 gram of the substance gave 0.0756 gram of sodic sulphate.

Calculated for Found.

CcH3(N02)3[CHNIa(000C„H5)2]3. I. II. III.

Sodium 9.09 8.66 8.87 9.31

Properties of the Addition Product of Trinitrobenzol and Sodic Malonic'

Ester, CeH3(N02)3[CHNa(COOaH5)j3.

This body has a rich maroon color brighter than that of the corre-
sponding compound of trinitranisol and sodic malonic ester. It is stable
for some time if kept dry and cool, otherwise it gradually undergoes
decomposition, as shown by its change of color and becoming gummy.
In its other properties it resembles the corresponding trinitranisol com-
pound most closely. When treated with hydrochloric acid the color is
destroyed, and a thick oily brownish yellow precipitate is formed ; by
washing this with small quantities of alcohol to remove the malonic ester
the trinitrobenzol was recovered in quantity, and recognized by its
melting-point, 121°-122°, after recrystallization from benzol. As soon,
therefore, as the three atoms of sodium are replaced by hydrogen, the
addition product splits into its constituents.

JACKSON AND GAZZOLO. — NITRO COMPOUNDS. 275

Action of Bromine on the Addition Product of Trinitrobenzol and Sodic

Malonic Ester.

The addition product C6H3(NO,,)3[CHNa(COOCaH5)2]3 was added in
small successive portions to a chloroform solution of bromine cooled by
immersing the vessel in ice. The color of the solid changed instantly
from maroon to white. After the mixture had stood over night, the solid
was filtered out, and the filtrate allowed to evaporate spontaneously, when
it left a thick brownish i-ed oil, which after standing two days deposited
crystals identified as trinitrobenzol by their melting-point, 121°, the form
of the crystals, and the formation of the characteristic red color with
sodic alcoholates. The portion insoluble in chloroform, after thorough
washing with chloroform and boiling benzol, proved to be sodic bromide.
This experiment does not absolutely disprove the formation of some
bromtriuitrobenzol, since a small amount of it might have remained dis-
solved in the oil from which the trinitrobenzol was deposited, but it
shows that trinitrobenzol is one of the principal products of the reaction ;
and as this separated from the oil in a nearly pure state, it is very prob-
able at least that no bromtriuitrobenzol was formed.

Action of Sodic Acetacetic Ester on Trinitrobenzol.

Tlie product was prepared in the same way as the corresponding addi-
tion compound of trinitrobenzol and sodic malonic ester. In this case
the precipitate had a deeper red color than that produced with sodic
malonic ester, and the reaction ran less neatly. Analyses of three
different preparations dried in vacuo gave the following results : —

I. 0.1865 gram of the substance gave 0.0G23 gram of sodic sulphate.

II. 0.2017 gram of the substance gave 0.0658 gram of sodic sulphate.

III. 0.2104 gram of the substance gave 0.0680 gram of sodic sulphate.

Calculated for

round.

C6H3(N02)3[CH3COCHNaCOOC2n5]3,

I.

11,

III.

10.32

10.82

10.56

10.47

Sodium

The addition product of trinitrobenzol and sodic acetacetic ester is a
rich brownish red amorphous powder darker than the corresponding
product from trinitrobenzol and sodic malonic ester. It is fairly stable,
if kept dry. In its other properties it is exactly similar to the colored
substances already described in this paper.

276 PROCEEDINGS OF THE AMERICAN ACADEMY.

Preparation of the Tri Sodic Amylate Addition Product of

Trinitrobenzol.

To a benzol solution of one gram of trinitrobenzol 1.5 grams of sodic
amylate were added gradually, care being taken to keep the mixture
cool. The pro^jortions are three molecules of the amylate to each mole-
cule of trinitrobenzol- A heavy scarlet precipitate was formed as soon
as the substances came together; this was filtered quickly, thoroughly
washed with benzol, and dried on a porous plate, after which it was
analyzed, with the following results : —

I. 0.2596 gram of the substance gave 0.1070 gram of sodic sulphate.

II. 0.1700 gram of the substance gave 0.0680 gram of sodic sulphate.

III. 0.3190 gram of the substance gave 0.1288 gram of sodic sulphate.

Calculated for Found.

C6H3(N02)3(05UuONa)3 I. II. HI.

Sodium 12.71 13.36 12.96 13.07

Properties of the Tri Sodic Amylate Addition Compound of Trinitro-
benzol, C6H3(N02)3(C5HnONa)3.

The dry substance is a dark crimson amorphous powder. It is re-
markably stable for bodies of this class, since it did not change in color,
or show any tendency to become moist, even after standing for two weeks
in contact with the air. It is soluble in ethyl or methyl alcohol or ace-
tone; very soluble in water; insoluble in benzol, chloroform, carbonic
disulphide, or ligroin. The strong acids decompose it at once, giving
trinitrobenzol as one of the decomposition products.

Upon treating trinitrobenzol with sodic methylate or sodic ethylate
under the same conditions, products were obtained with the following
formulas, if we may judge from the sodium determinations :

C6H3(N02)3(CH30Na)3Cri30H and CcH3(N02)3(aH50Na)3C2H50H,

but as these seem an insufficient foundation for such formulas, we shall
postpone the description of these substances until we have collected suffi-
cient analytical data to establish tlieir composition. They are both red,
but decompose more rapidly than the amylate, becoming moist and dis-
colored after exposure to the air for a few hours. Heating also decom-
posed the methylate body, so that the presence of methyl alcohol of
crystallization could not be established in this way. The discussion of
the conditions under which these tri bodies are formed instead of the
mono compounds will also be postponed until it has been thoroughly
settled by further experiments.

JACKSON AND GAZZOLO. — NITRO COMPOUNDS. 277

Attempts to obtain Colored Compounds with other Reagents.

Sodic phenylate, made by adding sodium to an excess of phenol, gave
with trinitrobenzol a clear red color, but no precipitate. A similar result
was obtained when an alcoholic solution of sodic phenylate was added to
a benzol solution of triuitranisol ; but this latter coloration does not
necessarily proceed from the sodic phenylate, as part of it may have been
converted into sodic ethylate by the alcohol.

Sodic hydrate also gives a red color with trinitrobenzol, as was observed -
by Hepp,* but as there seemed little chance of isolating this in a state fit
for analysis, we did not attempt to study it.

The sodium salt of phloroglucine, made by treating an excess of it
with sodic hydrate, gave a light reddish color when treated with a ben-
zol solution of trinitrobenzol, and upon adding an excess of benzol a most
uninviting sticky precipitate was formed which it would have been foolish
to try to analyze. We doubt whether this colored substance was really
a phloroglucine compound, as it is very possible that it was formed from
a little sodic hydrate produced by the decomposition of the sodium salt of
the phloroglucine.

Benzvl cyanide treated with metallic sodium, after the slight action
with the sodium was finished, was mixed with trinitrobenzol. Upon
stirring for a few seconds a deep blood-red precipitate appeared in large
quantity; but it was so unstable that even the addition of benzol to wash
out the excess of benzylcyanide converted it into a black tarry mass, so
that we were obliged to give up all idea of analyzing it.

Attempts to obtain Colored Compounds from other Nitro Bodies.

Picramide, CgH2(N02)3NH2, treated with a mixture of sodic methylate,
methyl alcohol, and anhydrous benzol, gave at once a strongly colored
dark crimson solution, which deposited a brick red precipitate ; but in
collecting it for analysis the substance decomposed as soon as it dried on
the porous plate, forming a brownish mass which later became tarry.
We were unable, therefore, to make an analysis,

Trichlorbromdinitrobenzol (CI3 1. 3, 5. Br2(N"02)2 4. 6.) gives a strong
vermilion color with an alcoholic solution of sodic ethylate, as already
stated by us in a previous paper, f

Dinitrotoluol ((N02)2 2. 4.) melting at 70° .5 gave with sodic methylate

* Ann. Cheni. (Liebig), CCXV. 359.
t These Proceedings, XXXIV. 148.

278 PROCEEDINGS OF THE AMERICAN ACADEMY.

a deep vermilion colored solution, from which a precipitate was obtained
with an excess of benzol. A benzol solution of the diuitrotoluol save
with sodic malonic ester a crimson red solution and a colored precipitate,
but both this and the precipitate of the methylate compound decomposed
while drying on the porous plate.

Symmetrical dinitroxylol melting at 93° ((0113)2 1. 3, (N02)2 4- 6.) gave
with sodic methylate after a few seconds a faint greenish color, which
turned rapidly to a deep purple, and finally became brownish black. It
was evidently therefore very unstable. Neither sodic malonic ester nor
sodic acetacetic ester gave any trace of color,

Trinitroxylol ((CHo)^ 1.3. (N02)3 2. 4. 6.) gave a deep cherry red solu-
tion with either sodic methylate or a benzol solution of sodic malonic
ester or of sodic acetacetic ester. An excess of benzol precipitated from
each of these solutions a gummy reddish body which decomposed before
it could be prepared for analysis. In these cases the decomposition-
product had a pinkish white color.

Neither dinitromesitylene nor trinitromesitylene gave a trace of color
after standing with sodic methylate. At the m'oment the trinitromesity-
lene was mixed with the sodic methylate we thought in one or two cases
we perceived a very faint coloration, but it was so indistinct that we felt
doubtful of its existence, and at best it was very evanescent. Sodic ma-
lonic ester and sodic acetacetic ester also gave negative results with both
these bodies.

Dinitrophloroglucine triethylether, C6H(OC2H5)3(N04)2, gave no
color with sodic methylate, sodic malonic ester, or sodic acetacetic ester.

Attempts to obtain Colored Compounds from Bodies which contain no

Nitro Group.

Pyrocatechin gave no color with sodic malonic ester, but with sodic
methylate a bright green color was formed along the edges, which soon
darkened, and finally gave a black oil. This coloration is probably sim-
ilar to those observed by Kunz Krause * on treating various phenols with
sodium and alcohol, but we do not feel sure that these colors are related
to those obtained from nitro compounds.

Resorcine gave no color with either sodic methylate, sodic malonic
ester, or sodic acetacetic ester. The same negative results were obtained
with phloroglucine.

Neither trimesic acid ((C00H)3 1. 3. 5.) nor its ester C6H3(COOC2H5)3

* Arch. Pharm., CCXXXVI. 542.

JACKSON AND GAZZOLO. — NITRO COMPOUNDS. 279

gave any sign of color with sodic methylate, although the experiments
were tried with great care, and under conditions which gave colors even
with some of the less reactive nitro compounds.

Experiments on the Replacement of the ATkijl Radical in the Colored

Compounds.

Action of Mdhyl Alcohol on the Benzyl Compound. — The addition
product of benzyl picrate and sodic benzylate, discovered by W. F. Boos
and one of us* was heated with methyl alcohol for about half an hour,
and the methyl alcohol was then allowed to evaporate at ordinary tem-
peratures. The product consisted of glistening scarlet crystals, which
were at once decolorized by hydrochloric acid, yielding a substance melt-
ing at 64°, and crystallizing in yellow rhombic plates from benzol. It
was therefore trinitranisol, and the methyl alcohol had replaced the benzyl
groups in the original addition compound by two methyls.

Action of Methyl Alcohol on Benzyl Picrate. — Benzyl picrate was
prepared according to the method given by Boos and one of us.f The
melting point of this substance is 145°, not 115° as given in the paper
just cited; the number 115° was due to a mistake in copying the melting
point from the note-book. A quantity of the benzyl picrate was recrys-
tallized four times from boiling methyl alcohol, and after each crystal-
lization the melting point remained constant at 145°, thus showing that
the benzyl picrate is not converted into methyl picrate by methyl alcohol
at its boiling point.

Action of Benzyl Alcohol on the Addition Product of Trinitranisol and
Sodic Methylate. — The colored compound was dissolved in benzyl alcohol
with the aid of gentle heat, and the mixture was allowed to stand at
ordinary temperatures until crystals separated. The red substance ob-
tained in this way was decomposed with hydrochloric acid, when the
product after crystallization showed the constant melting point 145°, and
was therefore benzyl picrate. In this case, therefore, the benzyl alcohol
had converted the colored methyl compound into the corresponding benzyl
compound.

Action of Benzyl Alcohol on Trinitranisol. — A solution of trinitranisol
in benzyl alcohol was allowed to stand in a paraffin desiccator until all
the benzyl alcohol had evaporated ; the residue showed the melting point
of trinitranisol, 64°.

* These Proceedings, XXXIII. 177.
t Ibid., 180.

280 PROCEEDINGS OP THE AMERICAN ACADEMY.

Benzyl alcohol, therefore, does not affect trinitranisol under the condi-
tions used in the experiment described in the last paragraph.

Attempts to prepare Derivatives from the Addition Compound of
Trinitranisol and Sodic Methylate.

Salts. — The sodium salt C6H2(NOo)30CH3NaOCH3 was treated with
the salts of various metals in the hope of obtaining other salts. The
chlorides of calcium, barium, mercury, and zinc, in mixed methyl alco-
holic and aqueous solutions produced no change. Cupric chloride, on
the other hand, formed a brown precipitate, from which trinitranisol was
isolated, and tests were obtained for copper and picric acid. We decided,
therefore, that the cupric chloride had decomposed the colored compound,
and neither this nor any of the other experiments we tried seemed to
point to the formation of salts of the colored compounds by metathetical
reactions.

Treatment with Benzoyl Chloride. — The addition compound
CsH2(N02)aOCH3NaOCH3, if dissolved in methyl alcohol and treated
with benzoyl chloride, was at once decolorized, even when sodic methyl-
ate was also present. Upon treating the dry compound with benzoyl
chloride, and allowing the mixture to stand over night, the amorphous
powder had become converted into masses resembling cauliflowers, with
an even more intense scarlet color than at first. An attempt to introduce
the benzoyl group by the Baumann-Schotten method led to a similar result.
One gram of the addition product was added to 25 grams of an 18 per
cent solution of sodic hydrate, and then 5 grams of benzoyl chloride were
gradually poured into the mixture ; the granular red powder was gradu-
ally converted into masses resembling cauliflowers, most of which dis-
solved in the alkaline liquid with a distinct intensification of the red color.
On acidifying with hydrochloric acid, the color was discharged and a
white precipitate of benzoic acid was formed. The filtrate apparently
contained picric acid. Unfortunately we had not time to study this reac-
tion more carefully, but we hope it will be investigated in this Laboratory
during the coming year, and also that the behavior of this compound with
methyl iodide may be studied then.

Postscript. The manuscript of the foregoing paper was ready for the
press, when I received an article * on colored compounds of this class by
Hantzsch and Kissel, in which they ascribe to them formulas with the

* Ber. d. chem. Ges., XXXII. 3137 (1899).

JACKSON AND GAZZOLO. — NITRO COMPOUNDS. 281

sodic alcoholate added to the nitro group only (I.). I cannot find any
reason in their article for changing the conclusion to which I had already
come, that a quiuoid formula (II. or III.) explains all the observed facts
better than the formula (I.) adopted by them. Their most important
new facts are the isolation of the free acid from the addition product of
potassic methylate and trinitrotoluol, and the formation of the correspond-
ing acetyl compound, both of which are explained better by the quinoid
formula than by theirs. They also call attention to the fact that the
free acid is a weak one instead of being a strong one, as it should be, if
derived from a salt with their formula, and that the marked color of the
compounds would not be expected from this structure ; both of these
anomalies disappear if the quinoid formula is adopted. It seems, there-
fore, that their observations tend to confirm this quinoid formula.

The authors also claim to have disproved definitely the theory of
Victor Meyer that these bodies are substitution products, but neglect to
mention that Lobry de Bruyn * in 1895 proved the incorrectness of this
theory by treating trinitrobenzol in boiling xylol with sodium. Therefore
all subsequent arguments against Victor Meyer's theory (of which I have
furnished three) must be considered as only confirmatory of Lobry de
Bruyn's work.

It may not be out of place to repeat here that work on this subject is
still in active progress in this Laboratory.

C. LoRiNG Jackson.
December 27, 1899.

* Kec. Trav. Chim. Pays-Bas, XIV. 89.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 15. — February, IDOO.

CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF THE |

MUSEUM OF COMPARATIVE ZOr)LOGY AT HARVARD COLLEGE. '

E. L. MARK, DIRECTOR. — No. 108.

THE METAMERISM OF THE HIRUDINEA.

By W. E. Castle.

CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF THE
MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE.
E. L. MARK, DIRECTOR. — No. 108.

THE METAMERISM OF THE HIRUDINEA.
By W. E. Castle.

Received January 20, Presented by E. L. Mark, February 14, 1900.

CONTENTS.

I. Introduction 285

II. Limits of the Somite in the

RliynehobdellidcB .... 286

1. Neuromeres as a Criterion of

Somite Limits 280

a. Somite Reduction in GIos-

siphonia 287

2. Septa and other Metameric

Organs as Criteria of So-
mite Limits 290

III. Limits of the Somite in the

Gnathobdellidae 291

1. Relation of Five-ringed to

Three-ringed Type of So-
mite 291

2. Somite Reduction in Nephe-

lis 293

IV. Somite Growth 296

1. Rhynchobdellida; .... 297

2. Gnathobdellidie 299

V. Primitive Condition of tlie

Leecii Somite 300

Summary 302

Papers cited 303

I. Introduction.

As long ago as the year 1862 Gratiolet pointed out that in the
medicinal leech, Hirudo medicinalis L., the successive rings of the body
are not all alike, but that they fall into similar groups of five each, which
are repeated in regular succession throughout the greater part of the
body. He found in each such group a single nerve ganglion and a single
pair of nephropores ; in other words, he discovered that there was a
segmentation of internal organs (metamerism) which agreed not with
single external rings but with the groups of rings.

Gratiolet called the groups of rings " zoonites " (somites). He noticed
on the exterior of the body two natural landmarks, which indicated to

286 PROCEEDINGS OF THE AMERICAN ACADEMY.

him their limits. These were, first, some whitish flecks (sensillse) found
dorsally on the ring which contains the ganglion (sensory ring) ; secondly,
a pair of pores (nephropores) lying in the furrow just posterior to ring 5,
counting the sensory ring as ring 1.

Gratiolet seems to have assumed that these convenient landmarks
indicated respectively the anterior and posterior limits of the mor-
phological units or somites. In this assumption nearly all subsequent
investigators have followed him uuquestioningly. Vaillant ('70) alone
expressed the view that it would be more logical to look for septa, which
should, as in the Cha^topoda, mark the limits of the somites. In the
case of Pontobdella, Vaillant believed that he had succeeded in finding
such septa, but subsequent investigators do not agree with him in this
view.

Gratiolet further observed that toward the ends of the body the
number of rings in a somite becomes smaller, and the ganglia are
crowded closer together. He suggested that the terminal ganglion at
either end of the body probably had resulted from a fusion of originally
distinct ganglia, as is the case with the ganglia of certain Mollusca
{Helix, etc.).

Gratiolet thus opened the way for a study of the metamerism of the
leech in two directions : first, to determine the number of primitive
somites present in the body of the leech ; and secondly, to determine the
natural limits of these somites. The first problem indicated has received
a satisfactory solution through the splendid work of Whitman ('92) on
the Rhynchobdellidre, and of Bristol ('99) on the Gnathobdellidaj.

From studies of my own, the results of which are in process of pub-
lication, I can confirm the conclusion of these investigators, that the
body of the leech contains thirty-four distinct somites.

The second problem, namely, the determination of the true limits of
the somite, has, as already indicated, received practically no attention.
To a discussion of this problem and some considerations growing out of
it, I now invite attention.

II. Limits of the Somite in the RHTNCHOBDELLiD.iE.

1. Neuromeres as a Criterion of Somite Limits.

In determining the number of somites found in the body of the leech,
it has been found necessary to rely solely on a study of the nervous
system. It would seem natural, therefore, to look in the same direction
for a solution of the problem conceiniug the limits of the somite.

CASTLE, — METAMERISM OF THE HIRDDINEA. 287

Whitman ('92) has shown that in the genus Glossiphonia (Clepsiue)
each distinct ganglion gives oti' three pairs of nerves, wliich are dis-
tributed respectively to the ring in which the ganglion typically lies
(sensory ring), the ring before it, and the ring behind it (Fig. 1, VIII.).*

If we take neuromeres as criteria of somites, the three rings thus
innervated from a common souree should be regarded as constituting a
somite. According to the commonly accepted view, however, the last
two of these rings with the nest following ring of the body, which is
innervated from a different ganglion, represent a somite. See Figure 1,

VI r.

The only other view possible would be to regard the somite as com-
posed of a sensory ring and the two rings which precede it, but I am not
aware that such a view has ever been suggested, and I am unacquainted
with any facts which could be presented in its support. Accordingly
discussion may be limited to the two views before stated, which are
graphically presented at the right and left margins respectively of
Figure 1.

a. Somite Reduction in Glossiphonia.

What first suggested to me the possibility that neuromere limits may
coincide with somite limits, was a study of the method of somite reduc-
tion at the ends of the body of Glossiphonia. See Figure 1, I. -IV. ;
Figure 2, XXV.-XVII. In the regions indicated the metameric sense
organs occur, not as in the middle of the body on every third ring, but
on every second ring, or even on successive rings. This fact shows that
in those particular regions the somite contains but two rings, or, in
extreme cases, only one. If we regard such somites as having had
originally each its full quota of three rings, we must suppose that one or
two of the three primitive rings have subsequently disappeared. This
disappearance might be explained as due either to a complete suppression
of a ring or to its fusion with an adjacent ring. A comparative study of
the abbreviated somites in different species of Glossiphonia shows that
the latter process is invariably the first step in the abbreviation of a
somite. An examination of Figures 1 and 2 will make this clear. In
the case of somite IV. (Fig. 1) a fusion has taken jjlace between the
sensory ring and the ring which precedes it. A similar union has
occurred in somites III. (Fig. 1) and XXV. (Fig. 2). In the case of

* This leaves out of account a branch {d, Fig. 1) which arises from the posterior
of the three nerves and is distributed to sense organs on the dorsal surface of all
three rings innervated by the ganglion.

288

PROCEEDINGS OF THE AMERICAN ACADEMY.

VII.

VI.':'

Vlllj

Fig. 1. Dorsal view of the anterior part of tlie body of Glossiphonia para-
sitica Say (Clepsine hollensis Whitman), showing in the right lialf of the figure the
nervous system ; in the left half, the lymph system (stippled). The pear-shaped
dark bodies shown on certain rings are dorsal sense organs, d., stump of a nerve
branch which innervates dorsal sense organs on all three rings of a somite.
nph'po., nephropore. va.t., transverse lymph vessel. Somite limits are indicated
by roman numerals, at the left of the figure (I. -VII.'), according to Whitman's
view ; at the right (I.-VIII.), according to the view of the writer. After Whit-
man ('92), simplified.

CASTLE. METAMERISM OP THE HIRUDINEA.

289

somites I. and XXVII. a fusion has taken place between the sensory
ring and both tlie adjacent rings, the one which follows as well as the
one which precedes it. An intermediate condition is found in somites
II., XXV., and XXVI., where the ring following the sensory ring is

XXlll/

XXIV.' -;

XXIV.

XXV.'

XXV.

XXVI.

XXVil.

Fig. 2. Posterior end of the body of GlossipJionia star/nnlis L. (Clepsine biocu-
lata Sav.), dorsal view. Somite limits as commonly placed are shown at the left
of the figure (XXIII. '-XXVII.') ; as the writer would place them, at the right
(XXIV.-XXVIL).

considerably reduced in size, but still distinct from the sensory ring
anterior to it. The amount of shortening which the somite undergoes
is thus seen to be increasingly greater as one progresses from the middle
toward either end of the body. The successive steps of abbreviation
are : —

1. A fusion takes place between the sensory ring and the ring which
precedes it.

2. The ring which follows the sensory ring is reduced in size.

3. It fuses with the sensory ring preceding it.

Accordingly, at the conclusion of this process, we find united in a single
ring three primitively distinct rings, — a sensory ring and the two rings
adjacent to it. But these three rings are typically innervated from the
same ganglion (Fig. 1, VII., VIII.). This fact substantiates the view
already expressed, that neuromere limits coincide with somite limits.

If we do not admit that neuromere limits coincide with somite limits,

VOL. XXXV. — 19

290

PROCEEDINGS OF THE AMERICAN ACADEMY.

Whitmau's ('92) theory that all the somites of the body are of equal
morphological value becomes untenable. For, suppose the body to con-
tain thirty-four distinct ganglia, each innervating three separate rings
(the hypothetical primitive condition realized in all unabbreviated so-
mites). Now if the somite limits be marked off so as to include in a
somite a sensory ring and the two rings following it, as is the practice of
Whitman and others (Fig. 3, left half, compare P'ig. 1, left half), we
shall have the absurdity of a ring at the anterior end of the body belong-
ing to none of the thirty-four somites, and somite XXXIV. at the poste-
rior end of the body will contain only two rings. On the other hand, if

neuromere limits and somite limits
are regarded as coinciding, all somites
of the body are of equal morphologi-
cal value, each somite consisting of
three rings innervated from a single
ganglion (Fig. 3, right half).

2. Septa and other Metameric Organs
as Criteria of Somite Limits.

If we are unwilling to be guided
solely by the nervous system in
determininor somite limits, what oth-
er criterion can we find ? Vaillant
("70) suggested a search for septa.
But septa such as are found in Chte-
topoda are wanting in the leech,
unless we so regard the entire blocks
of mesodermal tissue which lie be-
tween the transverse lymph vessels.
The position of tlie lymph vessels is
indicated by stipfding in the left half
of Fi<rui"e 1 in the case of Glos-
siphon! a parasitica Say {Clepsine
hollensis Whitman). The transverse
lymph vessels are found regularly in
the sensory ring of each somite (Fig.
1, va. t.). It is generally admitted
that the lymph vessels represent a greatly reduced coelom. The blocks
of tissue between the transverse vessels are, then, morphologically equiv-
alent to the septa of Chaetopoda. But the middle of each such block falls

XXXI

XXXIt!

XXXIV/

XXXII.

-1 XXXIII.

XXXIV.

Fig. 3. Diagram showing the relation
between neuromeres and somites in the
body of a leecli with three-ringed so-
mites, — at the left of the diagram
(I '-XXXIV.'), according to the view of
Whitman; at the right (I.-XXXIV.),
according to the writer's view.

CASTLE. METAMERISM OF THE HIRUDINEA. 291

midway between the sensory rings, exactly the position indicated for the
somite limits by the study of the nervous system (Fig. 1, right half).

In his work on " The Embryology of Clepsine," Whitman ('78) states
that the testes appear in the septa between the mesodermal somites.
Now the position of the testis in the adult is regularly in the two non-
sensory rings, that is, exactly midway between successive sensory rings.
This is just the position in which the boundary between somites should
be on the grounds already examined.

Crop diverticula are repeated at metameric intervals throughout the
middle part of the body in most species of Glossiphonia. They inva-
riably arise, so far as my observations go, in the sensory ring of the
somite, a fact readily explained by the occurrence in that position of the
interseptal portions of the ccelom, represented by the transverse lymph
vessels.

There is one other set of organs which is metameric in its arrangement,
— the nephridia. A single nephridium, according to the account of
Oka ('94), lies principally in the ring which contains the nephropore
(sensory ring, Fig. 1, nplipo.) and in the two rings which precede it.
Accordingly the nephridium is not confined to a single somite, either as
somites are commonly limited or as I have suggested placing their limits.
This need not surprise us, since in the Choetopoda also the nephridium
lies partly in the somite preceding that in which the greater part of it is
found.

III. Limits of the Somite in the Gnathobdellid.e.

Let us next inquire whether the suggested criteria of somite limits are
applicable also in the case of the Gnathobdellidas.

1. Relation of Five-ringed to Three-ringed Type of Somite.

Whitman suggested and Bristol ('99) has demonstrated that the five-
ringed type of somite, found in the Gnathobdellidfe, may be derived from
the three-ringed type, found in the Rhynchobdellidaj, by supposing that
each of the two non-sensory rings of the latter type has divided, while
the sensory ring has remained unchanged.

A comparison of the* distribution of the nerves arising from a single
ganglion in Glossiphonia (Fig. 1) and in Nephelis (Fig. 4) shows the
correctness of the homology suggested. The anterior two of the three
metameric nerves of Glossiphonia are united in the case of Nephelis
(Fig. 4, a). They are distributed, as we should expect, to the sensory
ring and the two rings which precede it. The posterior nerve of Glossi-

292

PROCEEDINGS OF THE AMERICAN ACADEMY.

vii/.:

Figure 4.

CASTLE. — METAMERISM OP THE HIRUDINEA. 293

Fig. 4. The first seven somites and part of the eighth somite of Nephelis
lateralis, dorsal view, a., anterior nerve ; d., dorsal branch of posterior nerve,
which supplies nervous structures in all five rings of a somite ; cuin. n., nerve ring
found typically in each of the two annuli adjacent to a sensory annulus ; p, pos-
terior nerve ; .r, a branch of the anterior nerve which ramifies in annulus 1 of
its somite.

Roman numerals indicate the somite limits, at the left of the figure (I.'-VII.'),
according to the view of Bristol ; at tiie right of the figure (I.-VIII.), according lo
the writer's view. 2\fter Bristol ('99), simplified.

phonia (Fig. 1) retains its separateness in Nephelis, and is distributed
chiefly to the two rings which follow the sensory ring. In Glossiphonia
the posterior nerve gives off a branch (of, Fig. 1), which is connected
with sense organs on the dorsal surface of all three rings. In Nephelis
the posterior nerve gives rise to a homologous branch {d, Fig. 4), which
is distributed to sense organs on the dorsal surface of rings 1, 3, and 5,
and is also connected with the nerve rings {a7i7i. n., Fig. 4) discovered
by Bristol in rings 2 and 4.

Hence, if neuromere limits coincide with somite limits, the latter must
fall, in the case of somites VI.-VIII. of Nephelis, as indicated in the
right half of Figure 4.* Bristol, working on the basis of Gratiolet's
conception, that the sensory ring occurs at the anterior end of the somite,
places the somite limits as shown in the left half of Figure 4.

2. Somite Reduction in Nephelis.

An examination of the abbreviated somites at the ends of the body
should, as in the case of Glossiphonia, throw light on the qtiestion
whether somite limits and neuromere limits are identical.

Giving attention first to the anterior end of the body, somite VIII., as
I place the somite limits (Fig. 4, right half), is the most anterior
unabbreviated somite in the body of Nephelis. In the case of somite
VII., ring 1 has disappeared. For, the nerve branch (x., somite VIII.)
which typically ramifies in ring 1 is wanting in somite VII. Moreover
the most anterior annulus of somite VII. is shown to be homolotrous with
annulus 2 of somite YIII. by the fact that it contains the anterior nerve

* No entire unabbreviated somite is shown in Figure 4, but the structure of a
typical somite may be learned from a study of the last six rings of the figure,
which represent unabbreviated the sensory ring and the two following rings of
somite VII., and the sensory and two preceditvj rings of somite VIII.

294 PROCEEDINGS OP THE AMERICAN ACADEMY.

ring of the somite (ann. n.), a structure typically found in annulus 2
(see somite VIII,, Fig. 4).

Somite VI. clearly consists of three annuli, of which we can identify
3 and 2 by the dorsal seusillipe and the anterior nerve ring respectively.
In place of annuli 4 and 5 we have in somite VI. a single annulus, in
which, however, Bristol found no nerve ring.

Somite V. consists, apparently, of two annuli. But what the distri-
bution of nerve V.'' shows to be undoubtedly the anterior nerve ring of
this somite lies just within the limits of the next anterior annulus (IV.',
left half of Fig. 4). Our first impulse would be to include that annulus
in somite V.; but a careful study of the distribution of nerves I.^-IV."
shows that the annulus in question contains important structures (includ-
ing the eye and associated sense organs) belonging unquestionably to
somite IV. It is possible to suppose that in this case the most anterior
part of somite V. has fused with somite IV. It seems to me, however,
more reasonable to explain the condition as illustrating a general
tendency in the head region for nerves to be cari'ied forward of the
somites to which typically they would be limited. For example, nerve
v." (Fig. 4) is carried forward ventrally as far as the last annulus of
somite II.; nerve IV." runs into the anterior annulus of somite II. ;
nerves II." and III." are carried forward into somite I. In Glossiphonia
(Fig. 1) also a similar tendency can be recognized in the distribution of
nerves III."-V." The nerve ring found in the posterior portion of the
annulus assigned by Bristol to somite IV. , (Fig. 4, left half) cannot
reasonably be considered the posterior nerve ring of somite IV., because
it is connected exclusively with nerve V." (See Bristol's PI. Yll.,
Fig. IG.) Moreover the posterior nerve ring has disappeared in somites
V. and VI., we should therefore expect to find it wanting also in the
more abbreviated somite IV Finally, unless this nerve ring does
belong to somite V., that somite contains no nerve ring at all, and somite
IV., which is much more extensively al)breviated, still retains a nerve
ring. This seems very improbable, for we find that metameric structures
omitted from an abbreviated somite, do not appear in other somites still
more strongly abbreviated.

In determining the external limits of somites I. -IV., Bristol has been
guided by the position of the metameric sense organs and the distribution
of the metameric nerves I. "-IV.", which represent the nerves given off
from one side of primitive ganglia I.-IV. respectively, completely fused
into single trunks. The somite limits indicated by him in the case of
these four metameres are, accordingly, neuromere limits, and, so far as I

CASTLE. — METAMERISM OF THE HIRUDINEA.

295

can judge from second-haud knowledge of the subject, are entirely
accurate.

A comparison of Figures 1 and 4 shows that abbreviation has been
more extensive in the head end of Nephelis than in that of Glossiphonia.
The process also does not progress anteriorly with such even and regular
gradations as in the case of Glossiphonia. Nevertheless, abbreviation
takes place in both cases by practically the same steps. First the
anterior end of the somite is affected, then the posterior end.

XXIV.

XXV.

Fig. 5. Posterior end of tlie body of Nephelis lateralis, dorsal view. Somite
limits according to the view of Bristol are shown at the left of the figure
(XXIV.'-XXVII.') ; according to tiie writer's view, at the right (XXIV.-XXVII.).
After Bristol ('99).

Figure 5 gives a dorsal view of the abbreviated somites at the
posterior end of Nephelis. Somite XXIV. (riffht half of figure) is the
last unabbreviated somite and shows the typical condition, — a sensory
ring in the middle of the somite preceded as well as followed by two
non-sensory rings. In somite XXV. (rijrht side of figure), the sensillae
indicating the fixed third ring from which to reckon, rings 4 and 5 are
seen to be partially united. In somite XXVI., rings 1 and 2 are com-
pletely fused together and imperfectly united with the sensory rinjr, 3;
rings 4 and 5 are likewise fused together, but retain their distinctness

296 PROCEEDINGS OF THE AxMERICAN ACADEMY.

from the sensory ring, 3. In somite XXVII. a complete union has
taken place between rings 1—3 ; rings 4 and 5 are probably repre-
sented in the broad, curved part of the body which bears distally the
acetabulum. This part is imperfectly divided by transverse creases
(shown too distinctly in the figure), but is sharply marked ofE from the
preceding ring.

To recapitulate : — We find in the case of Nephelis the same rules gov-
erning the process of somite reduction as in the case of Glossiphonia.

(1). Rings fuse into groups innervated typically from a common
ganglion. P^xamples : Figure 5, right half, somites XXV.-XXVII. ;
Figure 4, somites III. and IV. The union, at the margins of the body,
of somite I. with the anterior ring of somite II. presents an apparent
exception, but here we have to do with an extreme case, in which all the
rings of one somite (I.) have fused with one another, and in addition a
fusion between successive somites (I. and II.) is foreshadowed.

(2). The first rings to disappear are those at the ends of the somite,
the anterior end being usually the one which is first affected. Examples:
in Figure 4, somite VII. (only the most anterior ring, 1, wanting) ; so-
mite VI. (a ring wanting at either end of the somite) ; somite V. (two
rings wanting at the anterior end of the somite, only one wanting at
the posterior end) ; in Figure 5, somite XXVI. (rings 1 and 2 fused
together and partially united with ring 3, rings 4 and 5 united) ; somite
XXVII. (rings 1-3 united, rings 4 and 5 united).

In somite XXV. alone we have a case where abbreviation affects the
posterior part of the somite sooner than the anterior part. But it will
be observed that the middle of somite XXV. marks the boundary
between the abbreviated and unabbreviated portions of the posterior
half of the body. This may explain why in this single instance rings 4
and 5 are reduced, while rings 1 and 2 are unaffected.

From the foregoing fads it appears that in the case of Nephelis as well
as of Glossiphonia there are natural somite limits, which coincide with
the limits of the neuromeres.

IV. Somite Growth.

Having now examined with some care the process of somite reduction
in the leech, it may be instructive to study also the reverse process,
namely that of increase in the number of rings in a somite. This can
best be done by an examination of typical somites from various genera
of the two commonly recognized families of leeches.

CASTLE. — METAMERISM OF THE HIRUDINEA.

297

1 . RhynchohdellidcE.

We are already sufficiently familiar with the typical three- ringed
somite of Glossiphonia (Fig. 6, C). In Hsementeria, a closely related
genus (Fig. 6, D), each of the non-sensory rings (1 and 3) is divided
ventrally but not dorsally.

Branchellion has a three-ringed somite like that of Glossiphonia, ex-
cept that the middle (sensory) ring bears a pulsating respiratory vesicle

D

B

m--

Fig. 6. Somites of Rhynchobdellidse.

A, a much abbreviated somite of Glossiphonia ;

B, a less abbreviated somite of Glossiplionia ;

C, a typical somite of Glossiphonia;

D, " " flEementeria ;

E, " " Pontobdella;

F, " " Trachelobdella;

G, " " Cystobranchus ;
The sensory ring and its derivatives are stippled.

H

I

K

M

1

2

3

4

5

'■--,

3J

Fig. 7. Typical somites of Gnathobdellidae. E, Nephelis ; /, Dina ; J-il,
Trocheta. Tlie sensory ring and its derivatives are stippled.

connected with the lymph system. In Ozobranchus, a related genus,
one of the non-sensory rings is divided in large individuals. Blanchard
('94), who has described this genus, states that the ring which divides is
the third ring of the somite, by which I understand him to mean the ring
yi\nc\\ precedes the sensory ring, — ring 1 of my enumeration.

In Pontobdella (Fig. 6, E), a typical somite consists sometimes of

298 PROCEEDINGS OF THE AMERICAN ACADEMY.

three rings, but oftener of four. The sensory ring, which contains a
res^iiratory vesicle, is always the broadest ring of the somite. The ring
which is usually present, but sometimes wanting, is a narrow one, which
has been added at one end of the somite. It is impossible to say, without
studying the nerve distribution, at which end it has been added. Appar-
ently, however, it is at the anterior end, for in those animals which I
have had an opportunity to examine, the new ring appears to be united
more closely with the ring which precedes than with that which follows
a sensory ring. Moreover, the ring which precedes the sensory ring is
usually not so broad as the one which follows it. This is an indica-
tion that it is the former rather than the latter which has undergone
division.*

In Trachelobdella (Fig. 6, F), according to Blanchard. the somite con-
sists sometimes of three, sometimes of six rings, each of the three prim-
itive rings having, in the latter case, divided. The sensillse remain on
the anterior derivative of the original sensory ring.

In Cystobranchus (Fig. 6, G) the somite consists of seven rings, two

of which bear the respiratory vesicles and doubtless have arisen, as in

Trachelobdella, from the division of the original sensory ring. The

remaining five rings have probably been derived from the two primitive

non-sensory rings, — three from one non-sensory ring, two from the

other. One of the non-sensory rings must, accordingly, have divided

only once, as in Trachelobdella ; the other, twice. The position of the

genital pores would seem to indicate that the posterior non-sensory ring

is the one which has divided a second time; but it is impossible to say

without study of internal structures whether this is really the case or

not.

A typical somite of Piscicola contains, according to Apathy ('88),

twelve annuli ; according to Blanchard ('94), fourteen. Not having

had an opportunity to examine this genus myself, I do not venture to

express an opinion as to the number or morphological value of the rings.

Blanchard, however, states that the respiratory vesicle is borne on two

rings of the somite, the anterior of which contains the nephropore. This

would seem to indicate that the sensory ring had divided only once, and

that the remaining twelve rings (or ten, Apathy) had arisen by repeated

division of the two primary non-sensory rings.

To recapitulate : — We find in the Rhynchobdellidte that, starting with

a somite of three rings of equal width, increase in the number of annuli

* Vaillant ('70), who was guided by tlie position of tlie "septa," also placed the
Bomite limits so as to make two rings precede the sensory ring and one follow it.

CASTLE. — METAMERISM OF THE HIRUDINEA. 299

takes place (1) most often by division of one or both of the non-sensory
rings ; (2) occasionally by division of the sensory ring also. In other
words, ring multiplication is most active at the ends of the somite.

2. GnathobdelUdcB.

The typical five-ringed somite of Nephelis, Hirudo, and other genera
of Gnathobdellidae (Fig. 7, H) has undoubtedly been derived from a
three-ringed somite through division of each of its non-sensory rings.
This idea, suggested by Whitman ('92), has been abundantly confirmed
by the careful work of Bristol ('99).

In the two genera Dina and Trocheta, as described by Blanchard ('94),
we find a further multiplication of annuli taking place. In the former
genus (Fig. 7, /), the last of the five rings which, according to my view,
constitute a somite (ring 5, Fig. 7, /) is broader than any of the others
and is partially divided by a transverse furrow (not shown in the figure).
In Tiocheta (Fig. 7, J) the fifth ring is completely but unequally di-
vided, the posterior of the two rings thus formed being narrower than its
mate. Sometimes no further evidences of ring multiplication are found
in the somite of Trocheta, but usually other divisions occur forming a
somite of seven (Fig. 7, K) or more frequently of eight rings (Fig. 7, L).
Next in order after the division of ring 5 comes that of ring 1 (Fig. 7, A'),
then that of ring 2 (Fig. 7, Z). Blanchard finds that individuals obtained
from the Crimea often show evidence in certain somites of still further
divisions affecting the sensory ring and the two broad rings which follow it.
This brings the number of rings in the somite up to eleven (Fig. 7, H).

It is interesting to note how these eleven rings are related to the three
primary rings of a typical somite of Glossiphonia. The primitive sensory
ring of Glossiphonia (2, Fig. 6, G) is represented in Trocheta by two
narrow rings (stippled in Fig. 7, M). It may, therefore, be regarded as
having divided once, the division being among the latest to occur in any
of the rings. In place of ring 1 of Glossiphonia (Fig. 6, C), there are
four rings in Trocheta (Fig. 7, M). Primitive ring 1 may accordingly
be said to have divided twice (compare Figs. 6, C; 7, H\ and 7, M).
In place of ring 3 of Glossiphonia (Fig. 6, C), we find in Trocheta (Fig.
7, il/) five rings ; in other words, the primitive posterior ring of tlie somite
has divided twice, and the most posterior of the four rings thus formed
has divided a thvdtime (compare Figs. 6, C ; 7, H\ 7, /; and 7, M).

To recapitulate: — In ring multiplication among the Gnathobdellidas
(1) all five rings occasionally divide ; (2) more often only the non-sensory
rings divide, the sensory ring remaining unaffected ; (3) division takes

800 PROCEEDINGS OF THE AMERICAN ACADEMY.

place ^rs< and most often in those rings which stand at the limits of the
somite (rings 1 and 5, Fig. 7).

If one marks off the somite limits in any other way than that which
I have followed, regarding the sensory ring either as the first or as the
last ring of the somite, the pi'ocess of ring multiplication becomes unin-
telligible, taking place now in the middle, now at one end of a somite.
The reader can test this for himself by reconstructing the diagrams given
in Figures 6 and 7 so as to make the stippled ring come at one end or
the other of the somite.

From the foregoing discussion the conclusion seems warranted that
neuromeric groups of rings, that is, rings innervated typically fi'om the
same ganglion, are natural groups behaving as units both in the process
of abbreviation and in that of elongation of jiarticular body regions; in
other words, that they are the true morphological units or somites.

V. Primitive Condition of the Leech Somite.

We have seen that the process of ring multiplication is very general
among the leeches, and that it takes place in quite an orderly manner,
new annuli being formed, in the great majority of cases, at the ends of
the somite. We have seen, further, that the complicated forms of somite
found in the Gnathobdellidaj and some of the Rhynchobdellida3 are all
derivable from a primitive three-ringed type of somite, like that of Glos-
siphonia. In view of these facts, does it not appear probable that the
three-ringed type of somite itself has been derived from a simpler primi-
tive condition? I am stronjjiy inclined to think so.

As to what this simpler condition was, we perhaps may get an idea
from an examination of the abbreviated somites of Glossiphonia (Fig. 6,
B, A). For we have found that multiplication of annuli is the reverse
process of somite reduction, both alike affecting the e7ids of the somite,
the sensory (middle) ring being the stable component of the somite in
changes of either sort.

The final result of somite reduction in Glossiphonia (Fig. &, A) points
to a primitive condition of the leech somite, in which it consisted, like the
somite of a choetopod, of a single ring. A tendency to increase the
number of annuli in the somite would, in harmony with what we know
of the process of ring multiplication in the leeches, have called for the
formation of a new and narrower ring at one end or the other of the so-
mite, or at both. If new rings were formed simultaneously at both ends
of the primitive one-ringed somite, we should arrive, by a single step, at
substantially the condition of somite found in Glossiphonia (Fig. 6, G).

CASTLE. — METAMERISM OF THE HIRUDINEA.

301

If, however, a new ring were formed first at only one end of the somite,
the probabilities are that it would be at the posterior end. For we have
found that ring multiplication occurs rather oftener there than at the
anterior end of the somite. Moreover, in somite reduction in Glossi-
phonia, we get a stage (Fig. 6, B) intermediate between the one-ringed
(Fig. fi, ^) and three-ringed (Fig. 6, C) condition of the somite, in
which rings 1 and 2 are united into a single broad anterior ring, while
ring three is entirely distinct.

llie intermediate condition just described not improbably represents a
true phylogenetic stage in the formation of the three-ringed type of somite,
for it co7-responds exactly with the condition found
in a typical body somite of Branchiobdella or of
Bdellodrilus (Fig. 8), those curious leech-like
annelids, which nevertheless possess certain chas-
topod characters. The ganglion {g.. Fig. 8) of a
somite of Bdellodrilus or Branchiobdella is situ-
ated, as we should expect if the views just ex-
pressed are correct, in the broader anterior ring
of the somite. This ring, according to the
homology suggested, corresponds with rings 1
and 2 of Glossiphonia (Fig. G, ^ and C).

If, as I believe, the common assumption is well
grounded, that leeches and chaetopods have been
derived from a common stock, can we discover
any reason why ring multiplication should take
place in one group and not in the other ? I think

we have a sufficient explanation in the fact that the leech body contains
always the same definite number of somites, no matter how large or how
small the animal may be. This number is thirty-four, both in the
Rhynchobdellidae and in the Gnathobdellidse.

In the chaetopod, on the other hand, the body contains at first a
relatively small number of somites, which is increased, as the animal
grows in size, through the formation of new somites in some limited
region of the body, usually at its posterior end. In some cases the newly
formed somites may separate themselves oflE as a distinct individual, in
other cases they serve merely for the elongation of the original indi-
vidual to an indefinite extent.

In the leech, however, there is no provision for increase in the
number of somites. A definite number of somites, thirty-four, is laid
down early in ontogeny, and never increased. Elongation of the body

Fig. 8. A single borly
somite of Bdellodrilus illii-
minatus, lateral view, a.,
anterior; d., dorsal; tj.,
ganglion; p., posterior;
v., ventral. After Moore
('95).

302 PROCEEDINGS OF THE AMERICAN ACADEMY.

takes place, therefore, only by elongation of the individual somites
which compose it. This method of growth is probably what has called
forth the phenomenon of ring multiplication. In support of this view
may be cited facts such as that stated by Blanchard ('94) in regard to
Ozobranchus, that one of the three rings of a typical somite is doubled
in the case of large individucds.

Summary.

1. The number of somites in the body of the leech has been deter-
mined correctly by Whitman ('92) for the Rhyuchobdellid;c, by Bristol
('99) for the Gnathobdellidae. In both cases the number is thirty-four.

2. The lindts of the leech somite have been placed incorrectly by all
students of leech metamerism, with the possible exception of Vaillant
('70)» in the case of a single genus, Pontobdella, from the time of
Gratiolet ('62) to the present.

3. The natural and true limits of the somite coincide with the limits
of the neuromere ; that is, a somite includes those annuli which typically
are innervated from the same nerve ganglion.

4. The foregoing statement is confirmed by an examination of meta-
merically repeated structures other than ganglia ; namely, septa, testes,
and crop diverticula.

5. Neuromeric groups of rings, that is, somites as defined under 3,
behave as structural units (a) in somite abbreviation (reduction in the
number of rings in a somite), (b) in somite elongation (increase in
the number of rings in a somite).

6. Both reduction and increase in the number of rings take place
chiefly at the ends of the somite. The sensory ring occupies the middle
of the somite and is least often and least extensively affected in the two
processes just named. It represents the stable component of the somite.

7. The five-ringed type of somite found in the Gnathobdellidae has
been derived from the three-ringed type found in the Rhynchobdellida?,
as suggested by Whitman and demonstrated by Bristol. This has been
brought about by division of the non-sensory ring at either end of
the somite.

8. The wide prevalence of ring multiplication among the Hirudinea
suggests the derivation of the three-ringed type of somite from a still
simpler type consisting, as in Chastopoda, of a single ring.

9. A phylogenetically intermediate stage between the one-ringed and
three-ringed types of somite is probably represented in a typical body
somite of Branchiobdella. The same type of somite structure appears

CASTLE. — META3IERISM OF THE HIRUDINEA. 303

also in the abbreviated somites of Glossiphonia, as a stage intermediate
between the three-ringed and one-ringed conditions of the somite.

10. The phenomenon of ring multiplication in the Hirudinea is
correlated with the restricted number of somites found in the body.
Increase in the number of somites does not take place in the body of
the adult leech. Without this, elongation of the body is possible only
through lensthenius of individual somites. Lenaftheninsr of the individual
somite has probably been the cause, phylogeuetically, of increase in the
number of rings in a somite.

Papers Cited.
Apathy. S.

88. Analyse der jiusseren Korperforiii der Hirudiueen. Mitth. Zool. Sta.
Neapel. Bd. 8, pp. 153-232. Taf. 8, 9.
Blanchard, R.

"94. Hirudiuees de I'ltalie coutiueutale et insulaire. Boll. Mus. Zool. e
Anat. comp. Torino. Vol. 9, No. 192, pp. 1-81. 30 Fig.
Bristol, C. L.

'99. Tlie Metamerism of Nephelis. Jour, of Morph. Vol. 15, pp. 17-72.
PI. 4-8.
Gratiolet, P.

'62. Recherches sur rorganisatiou du systeme vasculaire dans la sangsue
medicinale et I'Austome vorace. Ann. Sci. Nat., Zool. T. 17, pp. 175-225,
PI. 7.
Moore, J. P.

'95. The Anatomy of Bdellodriliis illurainatus, an American Discodrilid.
Jour, of Morph. Vol. 10, pp. 497-540. PI. 28-32.
Oka, A.

'94. Beitrage zur Anatomic der Clepsine. Zeit. f. wiss. Zool. Bd. 58, pp.
79-151. Taf. 4-6.
Vaillant, L.

'70. Contribution a I'etude anatomique du genre Pontobdelle. Ann. Sci.
Nat., Zool. T. 13, pp. 31-71. PL 8-10.
Whitman, CO.
'78. The Embryology of Clepsine. Quart. Jour. Micr. Sci. Vol. 18, pp.
215-315. PI. 12-15.
Whitman, C. O.
'92. The Metamerism of Clepsine. Pestschr. Leuckarts. pp. 385-395. PI.
39, 40.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 16. — March, 1900.

CONTRIBUnONS FROM THE GRAY HERBARIUM OF
HARVARD UNIVERSITY.

New Series. — No. XVIII.

I. New Species and Varieties of Mexican Plants. By J. M.
Greenman.

II. Synopses of the Genera Jaegeria and Eusselia. By B. L. Rob-
inson.

III. New Dioscoreas from Mexico. By E. B. Uline.

IV. New Phaenogams, chiefly Gamopetalae, from Mexico and Central

America. By B. L. Robinson.

VOL. XXXV. — 20

CONTPxTBUTIONS FROM THE GRAY HERBARimi OF HARVARD
UNIVERSITY, NEW SERIES, No. XVIII.

Presented by B. L. Robinson, January 10, 1900. Received January 26, 1900.

I. — NEW SPECIES AND VARIETIES OF MEXICAN

PLANTS.

By J. M. Greenman.

Spiranthes Nelsonii. Aphyllous : roots clustered, tuberous, densely
covered with long hairs : stems 4 to 5.5 dm. high, clothed below with
oblong-ovate short-acuminate white scarious overlapping glabrous nervate
bracts, these somewhat more remote toward the inflorescence, glabrous
below, above and throughout the inflorescence more or less glandular-
pubescent: spikes oblong, 6 to 10 cm. long, about 2.5 cm. broad, rather
closely flowered ; bracts ovate, acuminate, shorter than the flowers :
flowers sessile, about 2 cm. long, apparently white ; the perianth later be-
coming reflexed or more or less folded on itself ; the outer lateral
divisions of the perianth or lateral sepals oblong, slightly expanded above
the middle, 1.1 to 1.2 cm. long, 3.5 mm. broad, obtuse, the upper sepal
similar but slightly larger ; the inner latei'al divisions of the perianth or
lateral petals oblong-spatulate, about 1 cm. in length, obtuse, much nar-
rowed toward the base; lip oblong-ligulate, 1.3 to 1.4 cm. long, 4 mm.
broad, subtruncate or obtusish at the apex, cuneate at the base, callous-
tuberculate on each side just above the base, bearing in the lower half on
either side near the margins a thick or fleshy pubescent ridge ; clinan-
drium pubescent along the free surface ; anther broadly ovate, 3 mm.
long, obtuse : beak of the stigma short-oblong, minutely 2-toothed at the
apex ; gland linear-oblong. — Collected by E. W. Nelson, between Rio
Verde and Panixtlahuaca, Oaxaca, altitude 125 to 370 m., 25 February,
1895, no. 2384.

Spiranthes Prixglei, Watson, var. minor. Flowers green or
greenish white, one-half to two-thirds as large as in the species proper ;
in technical characters and in general habit corresponding well with the
type specimen of the species. — Collected by C. G. Pringle in grassy
fields near Jalapa, altitude 1,225 m., 17, 21 May, 1899, no. 8197.

308 PROCEEDINGS OF THE AMERICAN ACADEMY.

Spiranthes tenuiflora. Roots several, fleshy, tuberous : stems
3 to 6.0 dm. high, bearing at the base remnauts of the old radical leaves,
above clothed with alternate lanceolate-acuminate white scarious imbri-
cated bracts or sheaths, these somewhat more remote toward the
inflorescence : racemes 1 to 2 dm. long, many-flowered, somev/hat
glandular-pubescent ; bracts ovate, lanceolate, acuminate, 3-uerved, shorter
than the sessile slender flowers (2 cm. long) : outer lateral divisions of the
perianth lance-oblong, obtuse, 3-nerved, about 6 mm. long, 2 mm. broad,
the upper sepal somewhat longer and broader than the lateral one ; lateral
petals narrowly lance-oblong, obtuse, shorter than and more or less
connivent with the upper sepal; labellum in general outline oblong, about
7 mm. long, short-uuguiculate, slightly hastate by a rather prominent
tubercle on either side at the base, terminated by a broadly ovate-reni-
form acute somewhat turbinate conspicuous expansion, 5-7-nerved, bear-
ing just above the hastate base minute papillae. — Collected by C. G.
Priugle in lava fields near Cuernavaca, Morelos, altitude 1,525 m., 16
February, 1899, no. 6995. A very striking and characteristic species,
readily recognized by the white showy bracts sheathing the stem, and
by the elongated racemes with long slender more or less remote reddish
or reddish-yellow sparingly glandular-pubescent flowers.

HosACKiA REPKN3, Don, Gen. Syst. of Gard. & Bot., ii. 200;
Walp. Rep., ii. 855. Specimens collected by C. G. Pringle in ditches
at TIalpam, Valley of Mexico, 20 August, 1896, no. 7240, and by Dr.
J. W. Harshberger, no. 170, are referred with little doubt to the above
meagrely described and little understood species. As the specimens
cited accord well with the brief characterization of Don, it seems best to
take up the above name, and to present herewith the following amplified
description. — An herbaceous perennial: stems repent or ascending,
3 to 7 dm. long, more or less branched, glabrous below, pubescent above :
leaves including the petiole 2 to 6 cm. long, 5-9-foliolate ; stipules
triangular, acuminate, 2 to 5 mm. long ; leaflets obovate-oblong, 1 to

2 cm. long, 4 to 8 mm. broad, obtuse, mucronate, narrowed below
into a conspicuously blackened rather slender (1 mm. long) petiolule,
green and glabrous above, somewhat paler and pubescent beneath :
inflorescence corymbose on long slender peduncles, these 8 to 18 cm. in
length, much exceeding the leaves; umbel usually 5-flowered, and sub-
tended by a short-petioled 3-5-foliolate bract: flowers 1 to 1.4 cm.
long, yellow, on short blackish pubescent pedicels : calyx-teeth lanceolate,

3 mm. long, about equalling the more or less reddish-maculate tube, both
tube and calyx-teeth pubescent : standard obovate-oblong, slightly

GREENMAN. — MEXICAN PLANTS. 309

retuse, abruptly narrowed below into a rather long claw : legume linear,
about 3 cm. long, compressed, glabrous. In general appearance H.
repens, Don, resembles H. Torreyi, Gray, but in the repent habit, the
very long peduncles, and in the longer claw of the standard, as well as in
the conspicuously blackened petiolules, it is quite different.

Hosackia oaxacana. An herbaceous perennial : stems erect or
ascending, 1 to 'l.o dm. high, usually branched, glabrous : leaves includ-
ing the short petioles l.o to 3 cm. long, 5-9-foliolale ; stipules narrowly
ovate, 3 to 5 mm. long, acute ; leaflets somewhat rhombic-obovate or
obloug-obovate, 0.5 to 1 cm. long, 3 to 6 mm. broad, obtuse or sub-
mucronate, green above, pale and somewhat glaucous beneath, having a
few scattered hairs on the midrib beneath and on the margins, otherwise
glabrous: peduncles usually exceeding the leaves, 2.5 to 5 cm. in length,
glabrous ; umbels 3-6-flowered, subtended by a nearly sessile 1-3-foliolate
leaf-like bract: flowers purplish, 1.2 to 1.4 cm. long: calvx-teeth
slightly unequal, triangular-ovate, 1.5 to 2 mm. long, acutish, sparingly
pubescent, barely one half as long as the glabrous more or less purplish
and somewhat glaucous tube ; standard obovate, slightly retuse, narrowed
below into a claw, not auricled at the base or subcordate : mature fruit not
seen. — Collected by the late Rev. Lucius C. Smith, Sierra de Clavellinas,
Oaxaca, altitude 3,000 m., 26 June, 1894, no. 57. The affinity of this
species is perhaps with the Californiau H. gracilis, Benth., biit it is readily
distinguished from it by the shorter petiolate leaves, the much shorter-
petioled and less conspicuous leaf-like bract subtending the umbellate
floral cluster, by the shorter broader and more obtuse calyx-lobes, and
finally by the subcuneate rather than distinctly auricled or subcordate
standard.

Arbutus glandulosa, Mart. & Gal. Bull. Acad. Brux. ix. 533
(reprint p. 5). This species, although originally described as having
" sharply spinulose-denticulate " leaves, seems to vary considerably in
the character of the leaf margin. It not infrequently happens that even
on the same individual specimen one finds leaves from perfectly entire
to sharply spinulose-dentate. This variation is well shown by Pringle's
no. 8033, in herb. Gray. To this species may also be referred, as a
form with entire or subentire leaves, Pringle's no. 8002, collected on the
Sierra de Pachuca, 21 February, 1899.

Styrax Ramirezii, Greenm. Proc. Am. Acad, xxxiv. 568. This
species, described from flowering specimens, has been collected in the
fruiting stage by Mr. C. G. Pringle on the Sierra de Tepoxtlan,
Morelos, altitude 2,300 m., 7 February, 1898, no. 8023. The fruit,

310 PROCEEDINGS OP THE AMERICAN ACADEMY.

which emerges from the persistent cup-shaped calyx, is oblong, 1 to
1.0 cm. iu length, 6 to 7 mm. thick, pale green and closely stellate-
lepidote over the slightly corrugated surface.

Stenamadenia macrophylla. Stems covered vrith a grayish bark
and dotted with lenticels : leaves oblong-obovate to oblong-lanceolate,
including the petiole 1 to 2.5 dm. long, 4 to 7.5 cm. broad, acuminate,
entire, narrowed below into a short (5 to 15 mm. long) petiole, glabrous
upon each surface, dark green above, slightly paler beneath : inflores-
cence subracemose at the bifurcation of the branches, few-flowered;
peduncles 3 cm. long : calyx deeply 5-parted, segments unequal, the two
outer shorter, 4 to 5 mm. long, somewhat narrowed toward the apex,
the inner oblong or slightly obovate-oblong, about 7 mm. long, rounded
at the apex : corolla tubular-funnelform, about 6 cm. long, tube nar-
rowed below, ampliated above, lobes spreading : fruit not seen. — Col-
lected by H. von Tuerckheim at Pansamala, Depart. Alta Verapaz,
altitude 1,100 m., January, 1886, no. 981.

This number (no. 981) of Mr. John Donnell Smith's collection was ori-
ginally distributed as " Odontadenia ? " * and subsequently corrected to
Stemmadenia hignoniaejiora^ Miers, and under this name Donnell
Smith's no. 1800 was also distributed. The latter species is character-
ized as having a calyx 1.3 to 1.5 cm. long. Stemmadenia macrophylla
on the other hand has the calyx less than half this length ; thus by the
character of the calyx alone the two species may be readily distinguished.

Stemmadenia tom.entosa. Shrub, 3 to 5 m. high : stems some-
what dichotomously branched above, covered with a grayish bark and
dotted here and there with minute lenticels : leaves opposite, ovate-
oblong, or slightly obovate, short-acuminate, entire, narrowed at the
base into a short slender petiole, glabrous above, densely soft-tomentose
beneath ; petioles slightly connate below and forming with the persistent
leaf scar a narrow ridire about the stem in the axil of which is a continu-
ous row of minute glands similar to those of the calyx: rather close
floral clusters terminating the branches : flowers large : calyx deeply 5-
parted, lobes broadly ovate, acute, about 5 mm. long : corolla yellow,
tubular-funnelform, greatly ampliated above, narrowed below, usually
more or less twisted in the throat, and with large subrotund spreading
lobes : mature fruit not seen. — Collected by C. G. Pringle in lava beds
near Zapotlan, State of Jalisco, 19 May, 1893, no. 4370, and distributed
as " S. mollis, Benth. ? " The nearest ally of ^S". tomentosa is apparently

* See Donnell Smith, Enum. PI. Guat. i. 26.

GREENMAN, — MEXICAN PLANTS. 311

S. Pahneri, Rose, as shown by the calyx characters and by the general
habit, but the former differs conspicuously from the latter by the presence
of a dense tomentum over the entire lower surface of the leaves. The
examination of a considerable number of specimens of S. Palmeri shows
minute tufts of pubescence along the midrib on the lower surface of the
leaves, otherwise the foliage is perfectly glabrous. This character seems
to be fairly constant, and as no intergrading forms have yet been seen, it
seems best to the writer, for the present at least, to regard Mr. Priugle's
plant as a distinct species.

Pliysalis acuminata. Suffrutescent : stems dichotomously branched,
glabrous below, glandular-viscid above : leaves single or in pairs, ovate,
4 to 10 cm. long, 2 to 6 cm, broad, acuminate, acute, entire, narrowed
below into a slightly marginate glandular-viscid petiole, more or less
glandular-pubescent upon both surfaces especially upon the midrib and
veins beneath : pedicels 1 to 2 cm. long, later usually reflexed : calyx in
anthesis 1.3 cm. long, densely glandular-pubescent; lobes ovate, acumi-
nate : corolla large, 2.5 to 3 cm. in diameter when fully expanded, sinu-
ately 5-lobed with short acutish lobes, and very shallow sinuses, ciliate,
conspicuously maculate iu 5 areas with numerous dark purple spots,
densely pubescent at the base : filaments about equal, glabrous ; anthers
purplish : fruiting calyx ovate, about 4 cm. long, equally 10-angled ;
lobes convei-gent : berry 2 cm. in diameter, glabrous. — Collected by
C. G. Pringle on the Sierra de las Cruces, State of Mexico, 23 October,
1892, no. 5315. Mr. Pringle's plant was at first doubtfully referred to
P. glutinosa, Schl., but from this species it differs materially in the char-
acter of the foliage. The leaves are all entire and narrowed below into
a petiole, not cordate, and, moreover, the fruiting calyx is equall}' 10-
angled, not 5-angled as iu P. glutinosa. To P. Pringlei the plant under
consideration bears a strong superficial resemblance, but differs in having
the leaves all entire, and in the character of the fruiting calyx.

Physalis Pringlei. Suffrutescent : stems 1 to 1.5 m. long, more or
less bi'anched, below terete, glabrous or nearly so, above and throughout
the younger parts densely glandular-viscid : leaves single or in pairs,
ovate to ovate-lanceolate, 3 to 8 cm. long, two-thirds as broad, more or
less acuminate, acute, remotely sinuate-dentate or sublobate to subentire,_
narrowed at the more or less unequal base into a slightly winged petiole,
glandular-viscid upon both surfaces: pedicels 1 to 1.5 cm. in length, at
first erect or at least ascending, later reflexed : calyx during anthesis 7
to 9 mm. long ; lobes ovate, attenuate, about equalling the tube : corolla
about 2 cm. in diameter, repandly rotate, ciliate, conspicuously maculate

312 PROCEEDINGS OF THE AMERICAN ACADEMY.

in 5 distinct areas with numerous dark purple spots, internally at the
base densely pubescent ; lobes obtusish : filaments about equal, glabrous;
anthers purplisli : calyx in fruit oblong-ovate, about 3 cm. long, definitely
o-augled, with intermediate ribs : berry about 1 cm. in diameter, gla-
brous. — Collected by C. G. Pringle on the Sierra de Clavellinas, State
of Oaxaca, altitude 2,700 m., 18 October, 1894, no. 6001, distributed as
" P. foetens, Poir." ; also on the Sierra de Ajusco, Federal District, alti-
tude 2,600 m., 2 October, 1895, no. 6216, distributed as " P. glatinosa,
Schlecht." P. Pri)iglei is evidently allied to P. foetens, but is quite
different in the much less dentate-lobed leaves, larger flowers, longer
calyx, and in the marking of the corolla. From P. glutinosa on the
other hand P. Pringlei differs in the leaves being contracted and attenu-
ate at the base into a petiole, not cordate.

Lamourouxia Conzattii. Herbaceous, probably perennial ; stems
and branches dark red or purplish, minutely puberulent in decussating
lines : leaves undivided, spreading or somewhat reflexed, 1-3-nerved, linear
to linear-lanceolate, 1 to 2.5 cm. long, 1 to 4 mm. broad, acute, serrulate,
slightly narrowed into a subpetiolate base, glabrous upon both surfaces,
margins more or less revolute : inflorescence in terminal usually one-
sided racemes; pedicels 0.5 to 1 cm. in length, minutely puberulent:
flowers large, 5 to 6 cm. long, scarlet, dark red or reddish-yellow; calyx
rather prominently 4-ribbed with intermediate less prominent ribs, puber-
ulent to nearly glabrous, lobes more or less unequal, lanceolate, 4 to 9
mm. long, equalling or exceeding the tube, obtusish, entire: corolla
externally pubescent, rather deeply bilabiate ; tube below narrow, ampli-
ated just above the base but not conspicuously ventricose ; lower 3-lobed
lip about 1.5 cm. long, only slightly spreading: perfect stamens 2,
shorter than and included under the galea ; anthers calcarate at the base,
villous-pubescent ; the posterior pair of stamens much shorter and bear-
ing reduced villous anthers : style exceeding the anthers in length, pu-
bescent ; capsule ovate, narrowed above, glabrous. — Collected by V.
Gonzalez and C. Conzatti on the Cerro de Papalo, camino de Cheve,
Cuicatlan, Oaxaca, altitude 2,500 m., 16-22 June, 1898, no. 760, and
by C. Conzatti on the Cerro de Yalina, District of Villa Alta, Oaxaca,
altitude 1,500 m., June, 1899, no. 965. Tiie author takes pleasure in
dedicating this beautiful species to Seilor Prof. C. Conzatti, whose ener-
iretic efforts have contributed much to our knowledge of the flora of
Southern Mexico. As the species of this genus are treated in Am. Jour.
Sci. 1. 169, L. Conzattii would stand near L. tenuifolia, Mart. & Gal.,
hut it is readily distinguished by the larger, longer, and more deeply
colored flowers.

GREENMAN. — MEXICAN PLANTS. 313

Lajiourouxia tendifolia, var. micrantha. Flowers only 1.8 to
2 cm. long: calyx cleft (in anthesis) nearly to the base: otherwise like
the typical form. — Collected on the west side of Valley of Cuicatlan,
Oaxaca, altitude 2,000 to 2,100 m., by E. W. Nelson, 10 November,
1894, no. 1891. Types in herb. Gray and herb. U. S. Nat. Museum.

Viburnum microcarpuji, Cham. & Schl., var. evanescens. Leaves
conspicuously discolorous; green and, except for a few scattered stellate
hairs on the sunken midrib, glabrous above, white and densely stellate-
tomentose beneath : in all other characters corresponding well with
the type specimen of the species. — Collected by C. G. Pringle in hedges
and thickets at San Miguel del Soldado, Vei-a Cruz, altitude 1,850 m.,
20 April, 1899, no. 8172.

Galium trtfloruji, Michx. Fl. i. 80. This species, hitherto doubt-
fully attributed to Mexico, may now be definitely incorporated in the
Mexican flora. Specimens were collected by C. G. Pringle in swamps
near Jalapa, altitude 1,200 m., 17 April, 1899, no. 8155.

Vernonia serratuloides, HBK., Nov. Gen. & Sp. iv. 33, t. 316.
Specimens collected by Dr. Edward Palmer in southwestern Chihuahua,
August to November, 1885, no. 279, and described by Dr. Gray, Proc.
Am. Acad, xxi. 393, as Perezia panieulata, prove upon careful examina-
tion to be a Vei'nonia, and moreover, correspond in all essential details
with the original description and excellent illustration of V. serratuloides,
HBK. Here also may be referred Goldman's no. 156, collected in the
Sierra Madre, near Guasarachi on road from Parral to Batopilas, Chi-
huahua, altitude 2,000 to 2,100 m., 26 September, 1898.

Coreopsis rhyacophila. SufFrutescent : stems erect, terete, covered
below with a thin grayish brown more or less deciduous cortex : leaves
opposite, petiolate, 3 to 8 cm. long, nearly as broad, pinnately tri- or
quadri-partite, glabrous upon both surfaces, glandular-punctate, divisions
narrowly lanceolate, apiculate-acute ; petioles 1.5 to 3 cm. long: heads
1 to 1.3 cm. high, and including the rays 2.5 to 3 cm. in diameter, few
on long slender striate naked or 1-3-bracteate peduncles; involucral
scales bi-seriate, the outer shorter, somewhat spreading, linear, acutish,
the inner oblong-elliptic, 8 mm. long, half as broad, margins subsca-
rious, glabrous except at the apex: rays yellow, longitudinally striate
with brownish lines : achenes about 4 mm. long, the outer surface gla-
brous, the inner pubescent, margins conspicuously ciliate ; pappus of two
stoutish ciliated awns, equalling or slightly exceeding the achene. — Col-
lected by C. G. Pringle in lava fields near Cuernavaca, Morelos, altitude
2,500 m., 9 February, 1899, no. 7866. In foliar characters C. rhyaco-

314 PROCEEDINGS OF THE AMERICAN ACADEMY,

phlla most closely resembles C. petrophlla, Gray, but from this species it
is readily distiuguished by the inflorescence, and by the much larger
heads. In the nature of the inflorescence G. rhyacopldla is perhaps
more nearly allied to C. anthemoides, DC, but from this species again
our plant differs markedly in the coarser character of the leaves, and by
the erect ligneous stem.

Spilanthes fllipes. Herbaceous : stems erect, freely branching from
the base, 3 to 4 dm. high, rather sparsely pubescent with jointed hairs ;
leaves opposite, petiolate, 3-nerved from near the base, ovate to oblong-
lanceolate, including the petiole 2.5 to 6 cm. long, 0.8 to 2 cm. broad,
obtusish, submucronate, crenate-dentate, narrowed rather abruptly or
even attenuate at the base into a ciliated petiole, glabrous or with a few
scattered hairs on the upper surface and on the midrib and lateral nerves
beneath: heads radiate, 4 to 5 mm. high, nearly or quite as broad, borne
on long slender (5 to 10 cm. long) peduncles which either terminate the
stems and branches or spring from the upper axils ; involucral scales
oblong-ovate, 2 to 2.5 mm. long, slightly acuminate, acute, inconspicu-
ously ciliate ; ligules about 5, including the tubular portion 4 mm. in
length; achenes of ray-flowers triangular: flowers of the disk 2.5 mm.
long; achenes laterally compressed bearing two minute pappus-awns,
ciliated, glabrous or nearly so on each face, at maturity about 1 mm. in
length ; pales rounded or subtruncate at the apex. — Collected by Dr.
G. F. Gaumer in Yucatan, 1896, no. 1122. In general appearance
S.fiUpes most resembles ^S*. r-amosa, Hemsl., but it differs matei-ially by
having much larger leaves, smaller heads, and stricter habit.

Dysodia Seleri, Robinson & Greenman, n. sp. Stoutish herb:
branches terete, striate, puberulent, at length glabrate, glandless : leaves
simple, opposite or sub-opposite, oblong, shortly acuminate, rounded or
more often cordate at the sessile base, sharply and rather finely serrate,
6 to 13 cm. long, 2.4 to 4 cm. broad, dotted with numerous round pellu-
cid glands; the uppermost leaves reduced, elliptical, often alternate:
peduncles terminal, more or less clavate-thickened toward the summit :
heads radiate, lar^e, 3 cm. in diameter exclusive of the ligules: involucre
puberulent, glandless; outer scales free, about 2-seriate, unequal, lanceo-
late, acute, herbaceous at least near the apex ; inner scales united for
two-thirds their length, their tips attenuate: rays 12 to 20, brown in a
dried state, broadly elliptical, about 2 cm. long including the tapering
tubular base ; pales short, thickish, fimbriate : tubular flowers 2 cm. long
including the sericeous-villous achenes : pappus 2-seriate, of about 20
subequal setose-fimbriate scales, 1 cm. in length : style-branches passing

ROBINSON. — JAEGERIA AND RUSSELIA. 315

abruptly into a slender acuminate appendage. — Collected by C. & E.
Seler, near Xochicato, Cueruavaca, Mexico, December, 1887, no. 410,
also near the Hacienda S. Gaspar in the same region, 18 December,
1887, no. 317. The affinity of this species appears to be with D. gran-
dijlora, DC, D. serratifolia, DC, D. integrifolia, Gray, and D. squam-
mosa, Gray. From all these it is readily distinguished by its glandless
involucre with larger herbaceous outer scales. Type specimens are pre-
served in the Royal Botanical Museum, Berlin.

Lygodesmia ramosissima. Much branched from a perennial base ;
branches striate-angled, glabrous, somewhat junciform, not spinescent:
leaves linear to subulate-linear, 0.5 to 7 cm. long, entire or remotely den-
ticulate, occasionally with more conspicuously spreading teeth, glabrous :
heads o-6-flowered, terminating the ultimate branchlets on either short
or elongated peduncles : involucre 2 to 2.3 cm. long, the outer calyculate
bracts ciliate, the inner bordered by a hyaline margin and bearing near
the apex a keel-like appendage : mature achenes subterete, smooth and
glabrous, about 1.5 cm. long; pappus nearly or quite equalling the
achenes, tawny. — Lygodesmia juncea, Gray, PI. Wright, i. 129, not Don.
— Collected by Charles Wright on the expedition from western Texas to
El Paso, prairies of the Pecos River, August, 1849, no. 417; by C G.
Pringle on plains near the city of Chihuahua, Mexico, 18 August, 1885,
no. 578 and by E. W. Nelson, between Ramos and Inde, Durango, 11 to
14 August, 1898, no. 4710. In habit L. ramosissima resembles most
closely L. juncea, Don, but from this species it is readily distinguished
by the size of the heads. From L. aphylla, DC, and its variety, our
plant is at once separated by the copiously branched stems.

IL — SYNOPSES OF THE GENERA JAEGERIA AND

RUSSELIA.

By B. L. Eobinsox.

The Genus Jaegeria.

The small helianthoid genus Jaegeria, inhabiting muddy shores and
shallow pools of tropical America, is exceedingly well marked by its non-
imbricated involucral bracts. These are similar to each other in form
and are equal in number to the rays, in fact each stands just in front of

316 PROCEEDINGS OF THE AMERICAN ACADEMY.

a ray and envelops the ray-achene by its thin infolded usually ciliate alar
portions. The achenes are glabrous and crowned with a shallow callous
cup, but otherwise pappusless.

Although thus clearly marked, the species of this genus have from
habital similarities been much confused with those of Sabazia, Gallnsoga,
Melampodium, and Spilanthes. The following synopsis, drawn up during
the study of the two new species here proposed, may therefore be found
useful in showing the present status of the group.

» Heads axillary, pedunculate, racemose, relatively large (including the rays 1.6
to 2 cm. in diameter) : rays about 12, conspicuous, pale yellow with more or
less deep roseate tinge: scales of the involucre ciliolate, otherwise glabrous:
weak aquatic essentially glabrous perennials.

-t- Leaves slender-petioled.

1. J. petiolaris. Sabazia glabra, Wats. Proc. Am. Acad, xxiii. 277.
— Shallow water, pine plains, base of the Sierra Madre, Pringle, no.
1295. The involute involucral scales enveloping the ray achenes clearly
show this plant a Jaegeria. The transfer of Dr. Watson's name glabra
to Jaegeria is rendered undesirable by the varietal use of this designation
under J. hirta by Mr. Baker in the Flora Brasiliensis as mentioned

below.

1- ■<- Leaves sessile, amplexicaul.

2. J. purpurascens. Decumbent stem simple or sparingly branched,
purplish, glabrous, leafy : leaves oblong-lanceolate, acute, sessile by a
clasping base, entire, glabrous, 1.4 to 2 cm. long, 3 to 4 mm. wide: pe-
duncles terete, purplish, 3 to 4.2 cm. long, puberulent near the summit,
borne in the upper axils : disk 8 mm. in diameter, hemisiDherical, at
length somewhat conical: rays 11 to 13, linear or narrowly oblong,
purple, 6 mm. long : bracts of the open Hattish involucre uniseriate, ovate,
acute, strongly ciliate at the broadest involute part, green and glabrous
dorsally : disk-ilowers numerous ; corollas pale yellow, 2 mm. long, gla-
brous except for a tomentulose ring at the base of the tube : achenes
glabrous, 1.7 mm. long, crowned with a narrow callous ring otherwise
calvous : ligules pale-yellow, shading to deep purple at the tip, 5 mm.
lono-, 1.8 mm. broad. — Collected in shallow water near Durango, Mex-
ico, by Dr. Edward Palmer, November, 1896, no. 805 (distributed as
Sabazia glabra).

* * Heads solitary and axillary (.7. prorepens) or more often terminal in the forks
of the stem, or, when several, borne in leafy cymes : more or less pubescent
plants of muddy shores or drier habitat : rays yellow or white.

ROBINSON. — -JAEGERIA AND RUSSELIA. 317

■*- Heads relatively large, including the well-exserted conspicuous yellow rays, 1.2

to 1.5 cm. broad.

■w. Main stem prostrate, rooting at the nodes ; branches ascending, few-headed :

bracts foliar.

3. J. MACROCEPHALA, Less. Svu. Gen. Comp. 223, & Linnaea, ix.
270. — Jalapa, Mexico, Schiede & Deppe, C. L. Smith, nos. 1624, 1835;
brook-sides uear Patzcuaro, Michoacan, Pringle, no. 4166.

++ ++ Main stem erect from a short decumbent base : heads many : bracts reduced.

4. J. PKDUNCUi.ATA, Hook. & Am. Bot. Beech. 299. — Jalisco,
Beechey, Palmer, no. 427, ace. to Gray, Proc. Am. Acad. xxii. 424 ; Prin-
gle, no. 1772; also Palmer, no. 47 in part (Spilanthes sessilis, Gray, I.e.
428, not HemsL). The identity of Mr. Pringle's no. 1772 with the type
specimen at Kew has been kindly verified by Mr. TV. B. Hemsley.

-t- -i- Heads considerably smaller; rays inconspicuous, scarcely exserted, yellow or
white : pubescent or liirsute annuals.

*+ Dwarf but not creeping, very slender; pubescence scanty : leaves small, ovate,

not at all clasping at the bate.

o. J. MNioiDES, HBK. Nov. Gen. & Spec. iv. 278, t. 400. The only
plant in herb. Gray corresponding at all satisfactorily to the plate and
description of this Michoacan species is some rather poor material col-
lected in Costa Rica by Oersted. In it the stems are very slender, simple
or sub-simple, and erect, while the leaves are small and truly ovate, sub-
sessile, but in no sense clasping. Kunth describes the rays as yellow.
It is not unlikely that this original species of the genus was only a
starved condition of the common J. kirta, Less.

++ ++ Tall, inclining to be repent at the base : leaves ovate, acutish : peduncles
filiform, several times as long as the heads : pubescence usually copious and
spreading.

6. J. HiRTA, Less. Syn. Gen. Comp. 223. J. repens, DC. Prodr. v.
544. Acmella hirta, Lag. Nov. Gen. & Spec. 31 ; Less. Linnaea, v. 153.
Melampodium hradajglossiun, J. D. Smith, Bot. Gaz. xiii. 74. Spilan-
thes sessilifolia, Coulter in J. Donnell Smith, Enum. PI. Guat. i. 23, iii.
43. Jaegeria caloa, Wats., a binomial needlessly published by Rusby,
Mem. Torr. Club, iii. no. 3, 62.*

* Dr. Watson was quite right in regarding this plant as the characteristic cal-
vous part of Sehultz's mixed tj-pe of the unpublished Galinsorja cnlva, as the exam-
ination of Mandon's no. 80 (a portion of which has been kindly forwarded from
tlie Kew Herbarium for comparison) clearly shows. Dr. Rusby's publication of
aiiotlier Galinmtga calra, founded on a pappus-bearing plant (vvliich in tlie well
known confusion of the Mandon numbers may or may not have had any relation to

318 PROCEEDINGS OF THE AMERICAN ACADEMY.

This is evidently the commonest and most widely distributed species of
the genus extending from western Mexico to Tropical Brazil. It is
highly variable in stature, foliage, and degree of pubescence ; yet floral
or even vegetative characters for a satisfactory segregation appear to be
lacking. Var. glabra, Baker, in Mart. Fl. Bras. vi. pt. 3, 167, is, to
judge from its first mentioned type (Mandon's no. 80), merely a smoother
not an altogether glabrous form as its description would imply.
•H4- ■»-)- -w- Dwarf, not creeping, freely branched, smoothish : leaves, at least tlie

upper ones, obovate or oblong, sessile by a narrowed but still somewhat

clasping base : peduncles sliort or none.

7. J. DiscoiDEA, Klatt, Arbeit, des Hamb. Botan. Mus. 1893, p. 2
of reprint. Heads small (not discoid even in Klatt's type specimen!),
short-peduncled or sessile : rays small although slightly exserted, white
or pale yellow. — Pringle, no. 4279 from the Sierra de las Cruces, State
of Mexico (type), also Pringle, no. 7349, Bonrgcau, no. 1232, and
Schaffner, no. 286, all collected in or near the Valley of Mexico. Per-
haps too near J. hirta. Klatt's ill-choseu name must be retained with
regret.

++++++++ Low, creeping : leaves rounded at the base : Galapagos Islands.

8. J. PROREPENS, Hook. f. Trans. Linn. Soc. xx. 214. — James
Island, Darwin.

Schultz's type) appears to me a mistalcen fidelity to an indefinite and inappropriate
specific name, especially when Dr. Rusby explains so carefully that he himself and
not Schultz should stand as authority for the pappus-bearing G. calva. Jaegeria
calva, as applied by Dr. Watson, was merely an herbarium name, resting upon an
obvious clerical error. Far from being "astonishing," slips of this sort appear to
be tolerably frequent in manuscript work of even the most careful botanists. The
thoughtless publication of such mistaken names, however, merely adds to tlie mass
of an already burdensome synonymy. It may be worth while to note in passing
that Dr. Rusby's name G. calva is not only inappropriate and misleading, but that
according to the Rochester Code it has no right to stand. G. calva, Sch. Bip., is
not, as Dr. Rusby maintains, a nomen nudum. It is clearly given (as Dr. Rusby
admits) by Baker in the Fl. Bras. vi. pt. 3, 167 (1884) in the synonymy of Jaegeria
hirta, var. glabra. Its status is thereby established, and it is impossible to deny its
publication, since it has been used in print together with a description, defining
synonymy, and the citation of Mandon's no. 80, one of the types originally men-
tioned by Schultz. But as thus defined it was applied to a Jaegeria. Whether we
write Golinsoga calva, Sch. Bip., or G. caira, Baker, in our synonymy of Jaegeria.
hirta, the combination Galinsoga calva has been a published binomial since 1884,
and Dr. Rusby's G. calva, applied to a different plant, is thus a later homonym
which a rational application of the Rochester rules would discard. It is to be
hoped that if any one feels impelled, on account of the "doctrine of homon^-ms,"
to assign a new name to Dr. Rusby's species, the choice may be more felicitous.

ROBINSON. — JAEGERIA AND RUSSELIA. 319

* * * Heads small, discoid : braachcd pubescent annual : Galapagos Islands.

9. J. GRACILIS, Hook. f. I.e. 213. — Charles Island, Darwin.

Doubtful species.

J. BELLiuioiDES, Spreng. Syst. iii. 591, of Uruguay, is unrecognized
and [)robably, like nearly all of the other species referred by Sprengel
to Jaegeria, belongs to some other genus.

The Genus Russelia.

Considerable recently acquired material of Russelia seems to show
Bentham's extreme reduction of the species ill-advised. At least a
dozen may be recognized with advantage. The following provisional
key may prove of assistance iu identifying the members of this confused
genus.

* Stems and branches sharply 4-angled, the angles bearing ciliated wings : pe-

duncles opposite, axillary, solitary. S. America.

1. R. alata, Cham. & Schlecht. Linnaea, iii. 3 (1828) ; Schmidt in
Mart. Fl. Bras. viii. pt. 1, 269, t. 44, f. 11. — Tropical Brazil, Sellow
& Riedel.

* * Stems and branches sharply angled not winged; the angles prominent, often

thickened ; the intervening areas flat or concave.

f~ Juncoid, excessively branched : peduncles filiform, 1-2 (-3)-flowered, much ex-
ceeding the subtending bracts.

2. R. EQUiSETiFORMis, Schlecht. & Cham. Linnaea, vi. 377 (1831).
R. juncea, Zucc. Flora, 1832, II. Beibl. 99 (1832), is only a large
flowered form of this species. The type of R. equisetiformis has been
examined by Mr. J. M. Greenman in Berlin and proves to have corollas
which vary from 1 to 2 cm, iu length, thus invalidating the chief dis-
tinction between the species, a fact also noted by Dr. Loesener, Bull.
Boiss. ii. 562.

s- -t- Peduncles short, the primary' ones never equalling the subtending leaf-like
bracts, usually several-many-flowered.

++ Leaves entire, subcoriaceous, lucid.

3. R. SUBCORIACEA, Robiusou & Seaton, Proc. Am. Acad, xxviii.
113. — Tamasopo Canon, San Luis Potosi, Mexico, Pringle, no. 5086.

++ +* Leaves serrate : calyx-lobes oblong-lanceolate, gradually attenuate, not at
all subulate at the tip : flowers 2 to 2.4 cm. in length.

4. R. jaliscensis. Copiously branched : stems and branches sharply

320 PROCEEDINGS OF THE AMERICAN ACADEMY.

quadrangular, glabrous : leaves opposite, thin, ovate to ovate-lanceolate,
acute, attenuate at the base, covered on both surfaces with a sparse
more or less deciduous pubescence and resinous( ?) puuctation, ciliolate
at least when young, sharply serrate; petioles 2 mm. long; blade about
2 era. long, half as broad : floriferous branches leafy-bracted ; bracts
somewhat obovate, cuneate at the base ; peduncles 2-flowered ; pedicels
slender, 4 mm. in length : calyx herbaceous, 6 mm. long, divided nearly
to the base ; segments oblong-lanceolate, gradually and evenly narrowed
almost from the base to the flat acute but not at all subulate tip : corolla
scarlet, 2.4 cm. long, extei'nally glabrous, internally pubescent near the
base and bearded along a ventral line ; lobes orbicular, sj^reading : cap-
sule ovate, stramineous, 6 mm. long. — R. sarmentosa, Gray in Wats.
Proc. Am. Acad. xxi. 434, not Jacq. — Collected in Jalisco in barrancas
near Guadalajara by Dr. Edward Palmer, June, 1886, no. 126, and
by C. G. Priiigle, June, 1889, no. 2568. Types in herb. Gray. This
species differs from any other known to the author in the shape of its
calyx-lobes.

++ -M- ++ Leaves serrate : calyx-lobes broadly ovate, acuminate to subulate tips.

^= Stems and branches chiefly 4-angled, glabrous or glabrate.

a. Leaves not cordate.

5. R. SARMENTOSA, Jacq. Enum. PI. Carib. 25 (1762) nomen, Stirp.
Select. Am. 178 (1763), t. 113.

h. Leaves cordate.

6. R. FLORiBUNDA, HBK. Nov. Gen. & Spec. ii. 359 (1817).
Doubtless represented by Palmer's no. 252 from near Acapulco.

7. R. SYRiNGAEFOLiA, Cham. & Schlecht. Linnaea, vi. 376 (1831).

= = Stems and branches 6-many-angled.
a. Stems glabrous or soon glabrate.

8. R. VERTiciLLATA, HBK. Nov. Gen. & Spec. ii. 360 (1817).

h. Stems tomentulose or pubescent.

1. Leaves small.

9. R. POLTEDRA, Zucc. Abh. Akad. Muench. ii. 328 (1832-36).
R. retrorsa, Greene, Pittonia, i. 176, is hardly distinct.

2. Leaves large.

10. R. TERNiFOLiA, HBK. Nov. Gen. & Spec. ii. 359 (1817).

ROBINSON. — JAEGERIA AND RUSSELIA. 321

* * * Stems sub-terete, merely striate-angulate.

■^- Branches of the inflorescence pseutlo-racemose, elongated, loosely flowered:
leaves large, thickish, veiny, tomentulose beneath.

11. R. ROTUXDiFOLiA, Cav. Icoii. V. 9 (1799), t. 415. — Near Aca-
pulco, Mexico. Dr. Palmer's uo. 509 must be of this species.

-<- -(- Branches of the glomerate inflorescence cyraose, many-flowered : leaves
thin, acute or acutish : stem glabrous.

12. R. MULTiFLORA, Sims, Bot. Mag. t. 1528 (1813). — i?. pani-
cidata, Mart. & Gal. Bull. Acad. Brux. xii. pt. 2. 19 (1845), appears to
be merely a verticillate- leaved form of this species.

•i- M- J- Inflorescences cymose, very sliort, opposite on prolonged branches :
flowers very small : leaves buUate, obtuse : stem pubescent.

13. R. tepicensis. Probably suflfruticose ; stems subterete, striate-
angulate with many fine ribs, subappressed-pubescent, internodes con-
siderably exceeding the ternate broadly ovate cordate petiolate obtuse
crenate-dentate leaves ; these membranaceous, glabrous and with im-
pressed veins, strigillose on the veins and resinous-punctate on the
surface beneath, 3 cm. long, 2.4 cm. broad; petioles 3 mm. long, ap-
pressed-pubescent : floriferous branches (1 to 1.5 dm. long) mostly verti-
cillate toward the ends of the stems: bracts ovate, petiolate, crenulate,
small but exceeding the short peduncles (2 to 3 mm. in lengtli) ; pedicels
and subulate bractlets pubescent: calyx-lobes ovate, acuminate to slender
recurved tips, puberulent near the middle, thin at the margin, 2 mm.
long: corolla 6 to 7 mm. long, bearded near the base within, externally
glabrous ; lobes spreading : capsule ovoid, nearly spherical, glabrous,
lucid, small, 3 to 4 mm. in length, slightly grooved between the carpels ;
these slightly carinate dorsally. — Collected at Zopelote, Tepic, Mexico,
February, 1895, altitude 600 to 900 m., by Frank H. Lamb, no. 578.
Type in herb. Gray. This species is distinguished from B. sarmentosa,
Jacq., by its subterete pubescent stems, ternate subcordate leaves, and
very small flowers. R. rotundifoUa, Cav. and R. multijlora, Sims, the
only other species with subterete stems known to the writer, have much
larger inflorescences and also differ in the form and pubescence of the
leaves.

VOL. XXXV. — 21

322 PROCEEDINGS OF THE AMERICAN ACADEMY.

TIT. — NEW DIOSCOREAS FROM MEXICO.
ByE. B. Uline.

Dioscorea pumicicola. Only (^ plant seen. Stem slender, gla-
brous, dextrorsely twisted ; leaves cordate at base, deltoid or sub-trilobate,
lateral lobes rounded, median part ovate and blunt or abruptly acuminate
at apex, about 7-nerved, glabrous above, sparsely pubescent on the
nerves beneath ; flowers arranged iu 3-5-flowered cymules, which are
short-pedunculate, and situated at tlie angles of the slender rhachis ;
anthers subsessile, triangularly disposed at the centre of the oi)en cam-
panulate greenish-brown corolla.

Leaves 8 to 12 cm. broad, little longer than wide. Petiole 4 or 5 cm.
long, puberulent. Rhachis slender, angulate-flexuous, puberulent, 1 to 1.8
dm. long. Cymules 4 to 8 mm. apart, subsessile or on short peduncles.
Flowers sessile, 3 mm. wide, with orbicular corolla-lobes, the inner
whorl shorter. Anthers very minute. — This species is intimately re-
lated to D. macrostachya and D. splcuUJiora, its inflorescence and flowers
being essentially the same as the latter, but it differs radically from both
in its lobed leaves and jjubescence.

Morelos, lava beds near Cuernavaca; C. G. Pringle no. 7227 ^ , 189G.

D. oaxacensis, Uline in Engl. & Prantl, Nat. Pflanzenf. Nachtr. z.
II, 5, p. 8G, without characterization. — Glabrous, glaucous ; leaves
chartaceous, cordiform with wide sinus, 9-nerved, exterior nerves on
both sides more or less forked; J racemes elongate, simple; rhachises
angulate ; flowers shortly pedicellate, in 2~3-flowered rather remote fas-
cicles (rarely solitary), segments of the subrotate G-parted perianth
oblong, rather acute, the interior ones slightly narrower ; fertile stamens
3, subcentral, a little shorter than the segments of the corolla, strongly
incurved, anthers extrorse in the bud, sterile filaments 3, very short, flat,
entire, sometimes bearing effete anthers.

Stem compressed, angular, sinistrorsely twisted. Leaves scattered 2.5
to 4 cm. wide, acuminate. Petioles about 2 cm. long. Racemes as long
as 1.5 dm. Floral fascicles 4 to 12 mm. distant, shortly pedunculate.
Flowers 6 mm. wide, on short slender pedicels ; corolla-segments thin-
membranaceous, unequal, greenish-yellow.

Syn. D. macrostachya, Mart. & Gal. Bull. Acad. Brux. ix. pt. 2, p.
391 (1842), not Benth.

Mexico near Reyes, altitude 1,800 to 2,050 m., E. W. Nelson, no.
1786, in part, i. e., as to $ plant, in 1894. The only other specimen

ROBINSON. — MEXICAN PLANTS. 323

of this plant known to be in existence was collected by Galeotti (no.
5-456 J) and is in the herbarium at Brussels.

D. LOB AT A, Uline, var. morelosana. Middle lobe of leaf obtuse,
mucronate ; larger leaves measuring 2.5 dm. long and wide ; flowers
with deeper tube ; pedicels shorter and nearer together than in the
species.

Morelos, lava beds near Cuernavaca, altitude 1,540 m., C. G. Pringle,
no. 7341 ^.

IV. — NEW PHANEROGAMS, CHIEFLY GAMOPETALAE,
FROM MEXICO AND CENTRAL AMERICA.

By B. L. Robinson.

Hechtia sphaeroblasta. Only ^ plant known : leaves 7 dm. long,
4.5 cm. broad at the base, gradually attenuate to a slender pungent apex,
gray and scurfy on the lower (in dried state concave) surface, pale green
but nearly smooth upon the reverse face, armed on the margins at inter-
vals of about 2 cm. with very sharp incurved stramineous spines (2 mm.
in length): inflorescence mauy-branched ; axes smooth and glabrous, at
least 7 dm. high ; bracts ovate-oblong, acute, entire, 1 cm. or more long,
scarious, with many fine dark veins ; branches 1 to 3 dm. long, angled,
ascending, glabrous, 5-9-branched near the middle, bracts of the second
order, small, scarious, about 2 mm. long; branchlets 4 to 12 cm. long,
slender, ascending, loosely floriferous from the base to the tip, their
bractlets scarious, 1 mm. long ; buds broadly ovoid almost globose, 3
mm. long just before expanding: flowers 14 to 30 on each branchlet,
borne singly or subverticillate, nearly sessile, 6 mm. broad in anthesis :
calyx-divisions oval, 2 mm. long; petals (in dried state yellowish white)
suborbicular, distinct or nearly so, 3 mm. long, obtuse : rudimentary
ovary in (^ flowers glabrous. — Collected by E. W. Nelson at Tlapa,
Guerrero, Mexico, altitude 1,200 m., 3 December, 1894, no. 2044.
Types in herb. Gray and herb. U. S. Nat. Museum. This species pos-
sessing subsessile flowers, elongated flowering branchlets, and glabrous
ovary, evidently belongs in the group with H. rosea, Morr. (which has
much larger flowers), H. stenopetala, Kl. (which has a much denser
inflorescence and acuter petals), and H. gamopetala, ]\Iez (which has the
petals united).

Smilax domingensis, Willd., var. microscola. Stem rather thickly

324 PROCEEDINGS OF THE AMERICAN ACADEMY.

and branches sparingly covered with very short straight conical spines
(0.5 mm. in length) : otherwise corresponding closely to the typical West
Indian plant. — Collected by E. W. Nelson between Tumbala and El
Salto, Chiapas, Mexico, altitude 466 to 1,400 m., 29 October, 1895, no.
3392. Types in herb. Gray and herb. U. S. Nat. Museum.

Polygonum Meisnerianum, Cham. & Schlecht., var. jalapense.
Stems, ochreae, leaf-margins, and bracts covered with numerous stalked
glands ; prickles (developed in the typical form) obsolete : leaves inclin-
ing to hastate lobing at the base. — This is probably the P. Meisneri-
anum of Cham. & Schlecht. in Linnaea, v. 90, but much more glandular
than the typical form, characterized in Linnaea, iii. 40. — Collected by
C. G. Pringle in wet places near Jalapa, Mexico, altitude 1,225 m., 5
April, 1899, no. 8111. This species appears to have been overlooked by
Mr. Hemsley, as it is not recorded in the Biologia Ceutrali-Americana.

Telanthera mollis. Branches lignescent, terete becoming sub-
quadrangular near the slightly tumid nodes, covered when young with a
soft short spreading or even reflexed pubescence, at length quite
.glabrate: leaves lance-oblong, entire, acute at each end, finely appressed-
pubescent and (in the dried state) rugulose above, much paler and
soriceous-tomentose beneath, 4 to 6 or more cm. long, about half as
broad ; pubescent petioles about 1 cm. long : heads oval, obtuse, leaf-
less, 1.4 to 2.2 cm. long, 1.4 cm. in diameter, borne in 2-4-chotomous
spreading round-topped panicles ; ultimate tomentose pedicels 1 to 1.5
cm. long; bracts ovate, acute: sepals oblong, acute and pungent, sub-
equal, 7 mm. long, dorsally covered with long dense silky hair : stamineal
tube a little shorter than the ovary, the sterile segments 2.3 mm. long,
fimbriated above, about equalling the filaments and somewhat surpassed
by the anthers : style nearly as long as the ovary ; stigma globose,
unlobed. — Collected in Oaxaca in a canftn above Totolapam by C. & E.
Seler, 3 January, 1896, no. 1637. Near T. pubijiora, Moq. and (ex
char.) T. pycnantha, Moq. Differing from the former in its rounded
many-headed panicles and from both in its larger strongly villous flowers.
Type in herb. Gray.

Mimosa Deam.ii. Arborescent, 3 m. high: branches terete, unarmed,
fuscous, tomentulose and roughened by the enlarged persistent bases of
strigose hairs; stipules short, thickish, subulate, strigose : petioles 1.5 to
2.7 cm. long : pinnae a single pair ; their rhachises about 5 cm. long,
each bearing 5 to 6 pairs of leaflets, strigose ; leaflets elliptical, rounded
at each end, coriaceous, glabrous above, sparingly strigose (chiefly on
the excentric midnerve) beneath, 2.3 to 3.6 cm. long, half as broad, 3-4-

ROBINSON. — MEXICAN PLANTS. 325

nerved, strigose-ciliate : flowers roseate, in globose heads ; peduncles
2.5 cm. long, borne by 2's or 3's at the nodes of unbranched terminal
racemes : heads (including exserted stamens) 1.6 cm. in diameter :
bractlets small, spatulate or oblanceolate, with thickened pubescent acute
ends : calyx translucent, closely appressed, scarcely toothed but strongly
ciliate: corolla 4-lobed, 4-nerved, 2.7 mm. long, the lobes oblong, sub-
carinate, tomentulose toward the summit, about half the length of the
tube : stamens 4 : filaments 3 times as lon^ as the corolla, slightly en-
larged above : fruit green, 3-4-jointed, constricted between the oval
moderately flattened strigose segments, stipitate, 3 to 3. 4 cm. long, 8 mm.
broad. — Collected at Salina Cruz, State of Oaxaca, Mexico, by Charles
C. Deam, 21 December, 1898, no. 113.

This species (of § Eumimosa) is most nearly related to M. psilocarpa,
Robinson, but is readily distinguished by its strigose fruit and narrow
strigose not soft-pubescent leaflets.

M. psilocarpa. Branches ligneous, terete, unarmed, roughened by
ferrugineous lenticels : petioles 2.3 to 5.4 cm. long; leaves l-jugate;
leaflets on each division normally 3 pairs (the inner member of the low-
est pair often reduced), obovate, oblique or semicordate at the base, firm
and subcoriaceous, 2.8 to 3.6 cm. long, 1.5 to 2.5 cm. broad, when young
finely pubescent upon both surfaces, in age pale, lead-colored (in dried
specimens), slightly lucid, glabrate except for a few short appressed or
incurved stramineous bristles upon the slightly thickened margin and
upon the nerves beneath : peduncles about 3 cm. long, divergent, borne
singly or by twos at the nodes of short terminal racemes ; bracts subu-
late, ciliate ; flowers perfect (so far as observed), capitate : calyx 0.7 mm.
long, slightly toothed, ciliolate : corolla 4-toothed, 4 mm. long, 4-nerved ;
the teeth lanceolate, one-third the length of the tube : stamens 4 : ovary
slender-stiped ; pod 2-4-jointed, glabrous, wholly unarmed, 2 to 3 cm.
long exclusive of the stipe (nearly half as long), 1.3 cm. broad, con-
stricted between the segments, mucronate. — Mountain woods between
" S. Carl and S. Bartolo Yuahntepec," Oaxaca, collected by C. & E. Seler,
6 January, 1896, no. 1727. This species belongs to Bentham's series
Castae and may be placed near M. lactijiua, Delile, although in habit
and large leaflets it suggests some of the Sensitivae.

CuPHEA Reipublicae, Rob. & Sea. Proc. Am. Acad, xxviii. 106.
The type number of this species, unfortunately cited at the time of
publication as Pringle's no. 4112, is his no. 4142.

Vernonia liatroides, DC. Prodr. v. 34 (1836). Add syn. Eupa-
torium tulammi, Klatt, Abh. Naturf, Ges. Halle, xv. 323 (1882).

326 PROCEEDINGS OF THE AMERICAN ACADEMY.

Stevia arachnoidea. Stem, branches of inflorescence, &c., at first
white-lanate, soon arachnoid, and at length merely glandular-puberulent :
leaves opposite, lance-oblong, acute, cuneate at the base to a very short
petiole, pinnately veined, 8 cm. or more in length, dentate, puberuleut
above, much paler and tomeutose beneath : inflorescence a round-topped
dense compound corymb, its branches mostly alternate, spreading at a
considerable angle and often exceeded by the long narrow linear subtend-
ing bracts ; heads slender, crowded ; involucral bracts puberulent, acute,
6 mm. long: corollas pubescent, 5 mm. long, the tube purplish, the iJmb
white or at least pale : achenes black, glabrous : pappus of several
short unequal acute scales, without long awns. — *S'. cornpacta, Coult.
in J. D. Smith, Enum, PI. Guat. ii. 33, not Benth. — Collected
by Mr. J. Donnell Smith on the Volcano of Agua, Department
Zacatepequez, altitude 2,700 m., April, 1890, no. 2327. Type in herb.
Gray. S. compacta, Benth. PI. Hartw. 197, must from description
have been a very different plant with ovate leaves a tliird to half as long
and involucres also much smaller. Schultz Bipontinus (Linnaea, xxv. 279)
regarded Bentham's species as probably a form of S. rhombifolia, HBK.

S. AsCHENBORNiANA, Sch. Bip. Linnaea, xxv. 2G9 (1852). Mr.
J. M. Greenman informs me that his .S'. diffusa, Proc. Am. Acad, xxxii.
307 (1897), proves on comparison with authentic material at the Eoyal
Botanical Museum in Berlin to be identical wiih S. Jschenborniana.

S. bicrenata, Klatt, Engl. Jahrb. viii. 32, was founded upon a poorly
developed specimen of *S'. elatior^ HBK.

S. elatior, HBK., var. decumbens, Rob. & Greenman, Am. Jour. Sci.
1. 152., later published as S. decumbens, Greene, Pittonia, iii. 32, is S.
Liebmannii, Sch. Bip. in Klatt, Leopoldina, xx. 75.

S. MADRENSis, Gray, Proc. Am. Acad. xxi. 382, should, I believe, be
reduced to S. Plummerae, Gray. There seem to be no differences of
moment.

S. MONARDAEFOLiA, HBK., var. amblyolepis. Leaves large and
very thin, the blade somewhat deltoid-lanceolate or ovate, rather
abruptly contracted at the base : scales of the involucre obtuse or
rounded at the apex. — Collected by E. A. Goldman, at Pinal,
Chihuahua, Mexico, altitude about 1,800 m., 19 September, 1898, no.
116.

S. MONARDAEFOLIA, HBK., var. macrophylla. Leaves rhombic-
ovate, 1.2 dm. long, 5 to 6 cm. broad, cuneately decurrent upon the
petioles 3 cm. or more in length : involucral bracts acute. — Collected
by C. G. Pringle, woods of Eslava, Valley of Mexico, Federal District,

ROBINSON. — MEXICAN PLANTS. 827

altitude 2,350 m., 28 January, 1899, no. 8057, and earlier in woods at
base of mountains in the Valley of Mexico, Federal District, 6 November,

1896, no. 7328.

S. neurophylla, Robinson & Green jian, n. sp. Erect strict
slender somewhat rigid herb, 3 to 6 dm. high : stem terete, puberulent,
simple or sparingly branched : leaves strictly opposite, thick, rigid, ap-
pressed, linear-oblong, longitudinally ribbed with many close parallel
nerves, 2 to 5 cm. long, 3 to 5 mm. broad, sessile by a subclasping base,
coucolorous, green and nearly glabrous upon each surface ; the lowest
shorter, the upper also much reduced and distant : heads capitately
aggregated at the summit of the stem ; bracts of the involucre rather
rigid, glandular-puberulent, acute, 5 to 7 mm. long : corolla white,
nearly glabrous : achene dark nearly black, hispid both between and
upon the lighter colored angles ; pappus of 4 to 6 long almost plumose
awns and as many oblong intermediate scales. — Collected by E. W.
Nelson between Ayusinapa and Petatlan, Guerrero, Mexico, altitude
1,500 to 2,100 m., 14 December, 1894, no. 2133. Types in herb. Gray
and herb. U. S. Nat. Museum.

S. Rosei. Herbaceous smoothish perennial herb with tough dark
fibrous roots : stems single, erect or slightly decumbent, 3 to 5 dm. high,
usually branched from the middle, puberulent (under a lens) with
crisped white hairs or quite glabrate ; branches opposite, ascending:
leaves opposite, sessile, oblong or laiice-obloug, subcoriaceous, veiny,
crenate-serrate or nearly entire, usually obtuse, with 3 to 5 more or less
anastomosing nerves: heads in several small dense somewhat leafy-
involucrate capitate corymbs, these together forming a very loose leafy
corymb; involucral bracts green, linear, attenuate, very acute, 8 to 10
mm. long, almost glabrous, not much surpassed by the white or pale
purplish corollas : achenes black, glabrous ; pappus of 3 or more oblong
obtuse scales with or without 2 or 3 naked awns. — Collected by Dr.
J. N. Rose on the Sierra Madre, west of Balanos, Jalisco, 15 to 17
September, 1897, no. 2994; also in the state of Durango, 13 August,

1897, no. 2258. This species is evidently related to S. Pringlei, but
differs in its short-pedicelled glomerate heads, larger leaves, etc. It is
also related to »S. venosa, Gray, and S. Phimmerae, Gray, but differs
from both in inflorescence and character of its involucral bracts.

S. Seleriana. Stems stoutish, lignescent, terete, white-lanate upon
the younger internodes, soon arachnoid, glabrate, and at length verrucosa
with prominulous lenticels : leaves all opposite, decussate, petiolate,
oblong, crenate at the sides, subentire at the acutish apex, entire at the

328 PROCEEDINGS OF THE AMERICAN ACADEMY.

cuneate base, thickish, glabrous and covered on both surfaces (under a
lens) with very numerous dark glandular dots; blade 8 to 10 cm. long,
2.3 to 3 cm. broad, pinnate-veined, the midnerve prominent beneath,
grooved above; petiole 1.8 cm. long; the upper axils proliferous: in-
florescence relatively small, 4 cm. broad, a trichotomous short-^jeduncled
corymb ; involucral bracts 5 to 6 mm. long, viscid-pubescent upon the
back, acutish : flowers 5 ; corolla, apparently white, about twice the
length of the involucre, the tube viscid-puberulent : achenes smooth or
nearly so, crowned with several narrow acute or subulate scales ; more
elongated awns apparently lacking. — Collected by C. & E. Seler in
mountain woods between St. Carlos and St. Bartolo, Yantepec, Oaxaca,
Mexico, 6 January, 1896, no. 1G95. Well marked. Ty^^es in herb.
Gray and herb. Royal Botanical Museum, Berlin.

S. SKR^ATA, Cav., var. arguta. Leaves narrowly linear, ascending
or appressed, sharply serrate with incurved teeth : scales of the involucre
attenuate to a dark subulate tip. — Collected by C. G. Pringle on the
Cerro Ventoso above Pachuca, altitude 2,900 m., 18 August, 1898, no.
7652.

S. tephra. Ashy throughout with a close pulverulent pubescence
not glandular: stems herbaceous, terete, purplish: leaves scattered,
elliptical, obtuse, crenate-serrate, 3 cm. long, half as broad, 3-nerved,
finely pubescent upon both surfaces, slightly paler beneath, mostly con-
duplicate along the midrib, gradually narrowed to a petiole (1 cm. in
length) ; fascicles of smaller leaves on short or scarcely developed axil-
lary shoots: inflorescence a moderately dense compound corymb; its
branches chiefly alternate, ascending, the lower and longer about 8 cm.
in length, the heads densely clustered at their summits ; individual pedi-
cels 1 to 2 mm. long; involucral scales merely acutish, cinereous-
puberulent, without evident glands, green or purplish-tinged, 4 to 5 mm.
long: flowers 5, about twice the length of the involucre: corolla-tube
purplish, puberulent ; limb white or nearly so : achenes black, lucid,
4 mm. long, minutely roughened upon the angles ; pappus of three awns
(equalling the corolla) and 3 short blunt intermediate scales. — Col-
lected by C. G. Pringle, Sierra de Pachuca, Hidalgo, altitude 2,925 m.,
13 September, 1899, no. 8229.

Piptothrix Goldmanii. Glabrous : stems terete, striate, purple,
slightly lucid; internodes 6 to 8 cm. long: branches opposite: leaves
deltoid-ovate, acuminate, serrate, 3-nerved fz'om the shortly cuneate base,
green and glabrous upon both surfaces, 8 to 10 cm. long, 6 to 8 cm.
broad, thin but firm in texture, reticulated veins slightly prominent upon

ROBINSON. — MEXICAN PLANTS. 329

the scarcely paler under surface; petioles slender, 1.5 cm. long, purj^le :
heads small, crowded in round-topped corymbs at the ends of the
branches; pedicels and bracts filiform; scales of the involucre lineai-,
substriate, 2.5 mm. long, slightly ciliated near the acute tip : corollas
white, 3 mm. long, enlarged at the base of the tube, glabrous : achene
scarcely over 1 mm. long, black, glabrous ; pappus-bristles 5 to 8, often
5 one over each angle of the achene, unequal, white, caducous. — Col-
lected by E. A. Goldman near Batopilas, Chihuahua, Mexico, altitude
1,700 to 1,750 m., 4-5 October, 1898, no. 196. Readily distinguished
from the other species of the genus by its much broader and well-
petioled leaves. This may possibly prove identical with Eupatorlum
triangulatum, Alam., or E. rubrocaule^ HBK.

An effort to secure a more accurate classification of the tropical Ameri-
can Eupatoriums represented in the Gray Herbarium has suggested the
publication of the following new species and specific reductions. The
writer hopes later to publish a synopsis of the Mexican and Central
American species and thereby show more clearly the affinities of the
forms here proposed.

EuPATORiu.M ADENACHAENiUM, Sch. Bip. in Klatt, Leopoldina, xx.
75 (1884). Add syn. E. adenochaetum, Hemsl. Biol. Cent.-Am. Bot. ii.
91 (1881), nomen nudum.

E. ADSPERSUM, Klatt, Bull. Soc. Bot. Belg. xxxv. 279. Add syn.
E. polanlhum, Klatt, 1. c. 281. It is impossible to detect differences of
any moment whatever. I doubt whether either is really distinct from
Dr. Klatt's earlier species E. anisochromum, Bull. Soc. Bot. Belg. xxxi.
186 (1892).

E. aegirophyllum. Stem terete, clothed with close sordid or tawny
slightly scabrous tomentum ; internodes long: leaves opposite, ovate,
subrotund, shortly acuminate, shallowly and broadly cordate, 1.2 dm.
long, nearly or quite as broad, puberulent and slightly scabrous above,
much paler and at first tomentose but at length glabrate and veiny
beneath, serrulate (sometimes obsoletely so) ; nerves at base pinnate then
palmate somewhat above the base ; petioles tomentose, 5 to 7 cm. long :
inflorescence a large round-topped many-headed panicle ; branches oppo-
site, tomentose ; primary bracts petiolate, similar to the leaves ; the
secondary and ultimate ones linear or subulate ; heads 1 cm. long, 20-
30-flowered; involucral scales about 22, lance-linear, subequal, acute,
glandular-puberulent on the back, loosely imbricated: corollas 5 mm.
long, about equalling the whitish pappus ; the slender proper tube ex-

330 PROCEEDINGS OF THE AMERICAN ACADEMY.

ceeding the gradually ampliated throat : achenes slender, brown, attenu-
ate below, 3 mm. long. — Collected by E. W. Nelson on eastern slope
of Mt. Orizaba, altitude 1,500 to 2,100 m., 21 March, 1894, no. 304.
Near SJ. suhpenninervium, Klatt, but with smaller more numerous heads,
and with petioles 2 to 3 times as long, etc.

E. ALBiCAULE, Sch. Bip. in Klatt, Leopoldina, xs. 89 (1884). Add
syn. E. drepanophyllum^ Klatt, Ann. k. k. Naturh. Hofmus. Wieu, ix.
356 (1894). The type of the latter species stated by Dr. Klatt as col-
lected by Scniwne is Gaumer's no. 122 from Cozumel Isl.

E. ALBICAULE, Sch. Bip., var. laxius. Inflorescence somewhat looser
than in the type: scales of the invDlucre linear-attenuate, 5 mm. long. —
Collected by Dr. Edward Palmer in Ymala, Mexico, 16 to 25 August,
1891, no. 1474 (type) and by Dr. J. N. Rose at Rosario, Sinaloa, 7
July, 1897, no. 1566. Dr. Rose's no. 1839 from the same locality is
probably the same, although the inflorescence is still more open.

B. amblyolepis. Smoothish herb or undershrub, 3 to 6 dm. high;
stems slender, terete, at first puberulent, soon glabrate and lucid, copi-
ously branched ; branches opposite, ascending : leaves opposite, lanceo-
late, acute to attenuate at each end, serrulate to subentire, thin, green,
glabrous, and somewhat lucid, minutely pellucid-punctate, 5.5 cm. long,
1.3 cm. broad, 3-nerved from somewhat above the cuneate base; petioles
4 to 6 mm. long : heads small, numerous, in flat-topped corymbs ; pedi-
cels 5 to 8 mm. long, bracteolate near the summit ; involucral bracts
about 13, in 2 subequal series, obovate-oblong, pale green, striate,
rounded and ciliolate at the summit, 3 mm. long, at length spreading:
flowers about 15, small : corolla probably white, 3.5 mm. long, somewhat
exceeding the very delicate and rather sparse bright white pappus : achenes
1.5 mm. long, black, minutely roughened on the sharp angles. — Collected
by C. G. Pringle on cool ledges of the Sierra de Tepoxtlan, 24 km. north-
east of Cuernavaca, altitude 2,300 m., 7 February, 1899, no. 8034.

E. AREOLARE, DC, var. leiocarpum. Nearly glabrous throughout:
flowers somewhat larger than in the typical form and borne in condensed
cymes : achenes smooth. — Collected by E. W. Nelson, near Chilpan-
cingo, Guerrero, Mexico, altitude 2,700 to 3,100 m., 24 December, 1894,
no. 2221. Types in herb. Gray and herb. U. S. Nat. Museum. There
are no marked differences between the typical form of this species and the
later E. tiihiflorum, Benth. PI. Hartw. 76, so far as the tolerably detailed
descriptions indicate. Hemsley refers to Ghiesbreght's no. 530 under
both names. The vaguely characterized E. papantlense of Lessing may
perhaps be the same.

ROBINSON. — MEXICAN PLANTS. 331

E. BADiUM, Klatt, Bull. Soc. Bot. Belg. xxxi, 186 (1892), a species
based upon a Costa Rican type, is well matched by Mr. E. W. Nelson's
no. 2251 collected near Cliilpancingo, Guerrero, Mexico.

E. Bertholdii, Sch. Bip., var. stenophyllum. Leaves lance-
oblong, attenuate at each end, 9 cm. long, 2 to 2.4 cm. broad, entire. —
Collected by C. G. Pringle in canons near Guadalajara, Jalisco, 8 Decem-
ber, 1888, no. 1831. A shrub 2 m. high. The typical form of north-
western Mexico has ovate serrulate leaves.

E. hrickelloides^ Klatt, Jahrb. Hamb. Wissensch. Austalten, ix., pt. 2,
1892, 124, is a Trixis !

E. capnoresbium. Slirub with slender opposite terete glabrate
branches ; branchlets, petioles, and pedicels sordid glandular-tomentose :
leaves opposite, elliptic-oblong, acute at each end, impunctate, more or
less serrate from the middle to the apex, not revolute at the base, pubes-
cent above with short rather sparse hairs, paler and pubescent especially
on the miduerve and pinnate veins beneath, 3.5 cm. long, 1.6 cm. broad;
petiole 6 mm. long : heads (1 cm. in length) in trichotomous round-
topped fastigiate corymbs, slender-pedicelled, about 10-flowered; invo-
lucral scales lance-oblong, acutish, purple-tinged, glandular-puberulent
dorsally, one or two of the outer lance-linear and shorter than the rest :
corollas 6 mm. long, purplish : achenes prismatic, 3 mm. long, puberu-
lent. — Collected by C. G. Pringle in forests at the foot of tlie Sierra de
Ajusco in the Federal District, Mexico, 26 June, 1897, no. 7438, type
(in herb. Gray), and earlier by Schaffner on Mt. Popocatapetl, altitude
2,700 m., Se[)tember, 1855, also by Bourgeau in a forest of the Desert
Viego near City of Mexico, 1865-66, no. 844. This species has the
inflorescence and much the habit of E. micranthum, Less., which, how-
ever, has the leaves finely, but distinctly punctate on both surfaces and
revolute at tlie base. See remarks under E. glaucum, below.

E. eardiophyllum. Stout, tall herb, short-pubescent or tomentulose
and scabrous throughout ; stem terete, striate, fistulose : leaves opposite,
ovate-orbicular, short-acuminate, cordate with an open sinus, crenate, 1
to 1.6 dm. long, about as broad; petioles 2 or more cm. long, connate at
the base : rounded terminal corymb loose, ample, 2 to 3.4 dm. broad, its
branches subtended by bract-like much reduced leaves; bractlets linear
or filiform ; branches and pedicels dark purple, glandular-pubescent :
heads 8 mm. long, 24-flowered ; involucral scales about 15, linear, acute,
purple, puberulent, unequal, not strongly imbricated : corollas 5 to 6
mm. long, exceeding the nearly white pappus : achenes black, wliite-
puberulent under a lens. — Collected by C. G. Pringle on slopes of

332 PROCEEDINGS OF THE AMERICAN ACADEMY.

canons near Guadalajara, Jalisco, Mexico, 29 October, 1889, no. 2343,
and earlier on the Rio Blanco in the same neighborhood by Dr. Edw.
Palmer, October, 1886, no. 665. Both plants were distributed as E.
Benihamii. However, Klatt's species of that name (which is a clear
syuouym of the earlier E. Ehreiibergii, Hemsl.) has ovate serrate not
cordate leaves, more numerously flowered heads, etc.

E. CHAPALENSE, Wats., var. salicifolium. Leaves narrowly lance-
olate, merely serrulate, acute at both ends, 4 to 6 cm. long, 1.2 to 1.4
cm. broad : scales of the involucre oblanceolate, broader than in the
typical form. — Collected by C. G. Pringle on mountains near Lake
Chapala, Jalisco, Mexico, 18 October, 1895, no. 7071. The leaves of
the typical form are more than 4 cm. broad. The variety retains the
floral characters and unequal pappus of the type.

E. chiapense. Stem terete, clothed with a dense purplish brown
tomentum ; intex'nodes long (6 to 8 cm.) : leaves opposite, ovate,
acuminate, serrate-dentate except near the I'ounded base, dull green
and except on the subpinnate nerves nearly glabrate above, paler and
ferrugiueous-tomentose upon the nerves and veins beneath, 1 to 1.4 dm.
long, 6.5 to 8 cm. broad ; petioles densely tomentose, 4 to 5.5 cm. long:
heads numerous, 9 mm. high, about 24-flowered, borne in a compound
round-topped corymb (1.5 dm. broad) ; the opposite branches and
pedicels tomentose with purple gland-tipped hairs ; involucre subsimple
and scarcely imbricated ; scales about 10, narrowly linear, attenuate,
mostly 2-ribbed, 6 mm. long, pubescent : corollas glabrous, 5 mm. long,
exceeding the dull white pappus ; throat equalling the proper tube :•
achenes black, glabrous, tapering toward the base, 4 mm. long. —
Collected by E. W. Nelson, near Pinabete, Chiapas, Mexico, altitude
2,000 to 2,400 m., 8 February, 1896, no. 3786. This species is related
to E. Ehrenhei-gli, Hemsl., but has much narrower involucral scales,
smaller heads, glabrous corollas, etc. It is also and perhaps still more
closely related to E. Donnell-Smitlt'ii, Coulter, which, however, has
pale obtusish involucral bracts, and flowers much smaller, scarcely
more than half as long.

B. crassirameum. Large shrub or small tree, 3 to 5 m. high,
glabrous throughout : branches thick, terete, soft-woody to the inflores-
cence, covered with a pale gray cortex : leaves opposite, petiolate, large,
deltoid-ovate, thin, pellucid-punctate, glabrous, acuminate, repandly few-
toothed at the sides, pinnately veined, becoming 1.5 dm. long and
about equally broad ; petioles 5 cm. in length : heads numerous, about
15-flowered, pedicellate in dense rounded compound corymbs (9 cm. in

ROBINSON. — MEXICAN PLANTS. 333

breadth) ; involucres only 2 or 3 mm. long, little imbricated ; the scales
oblong, scarcely acute, unequal, ciliolate : flowers purple, 6 mm. in length,
— Collected by C. G. Pringle on lava fields near Cuernavaca, altitude
1,500 m., no. 8271. The flowers on the fleshy naked branches were
secured 16 February, and the leaves 29 September, 1899. Very
distinct in habit from any species of wliich I can find record.

E. crenaeum. Erect herb ; simple stems so grooved as to be-
come cross-shaped in transverse section, pulverulent-puberulent ; in-
ternodes mostly long, the lower more than 1 dm. in length : leaves
opposite, decussate, ovate, obtusish, creuate, obtuse at the subentire base,
finely and sparsely pubescent upon both surfaces, 9 to 11 cm. long,
7 to 8 cm. broad, 5-nerved from the base, slightly paler beneath ; petioles
relatively short, 1 to 1.3 cm. long, puberulent : heads about 50 in a
terminal flat-topped corymb ; pedicels filiform, flexuous, 5 mm. long ;
involucral bracts about 14, subequal, linear, acute, 1-3-nerved toward
the base, 6 mm. long; flowers about 16 ; corolla probably white or pale,
5 mm. long, equalled by the white pappus ; achenes glabrous, gray. —
Collected by C. G. Pringle in springy places near Guadalajara, Jalisco,
Mexico, 6 July, 1889, no. 2878.

E. desquamans. Copiously branched shrub ; branches spreading,
lignescent nearly to the tips; internodes mostl}^ very short, soon covered
by a grayish cortex : leaves opposite, petiolate, small, oval, obtuse or
merely acutish at each end, 1.8 cm. long, 8 to 11 mm. broad, pinnately
veined, finely serrate, at maturity green above and puberulent only
under a good lens, paler beneath, when young clothed especially beneath
with a close silvery web which soon breaks up into deciduous scalelike
remnants ; petioles 3 to 5 mm. long : inflorescences simple or nearly so,
terminal, 5-9-headed, racemiform ; pedicels opposite, 1 cm. long, bracteo-
late near the middle ; "involucre subsimple, of linear acute brownish
puberulent subequal bracts about 13 in number and 5 mm. long : flowers
about 15, 1 cm. long: corolla apparently white or pale, about equalled
by the copious nearly white pappus : achenes smooth, black, shining,
sharply angled. — Collected by C. G. Pringle on summit ledges of
the Sierra de San Felipe, Oaxaca, altitude 3,100 m., 22 May, 1894,
no. 5858. Type in herb. Gray.

E. GLAUCUii, Klatt, Leopoldina, xx. 89 (1884). The publication of
this species has been unfortunate in several ways. The name first
appears in print as " E. glaucum, Sch. Bip. ex A. Gray " in Hemsl. Biol.
Cent.-Am. Bot. ii. 95 (1881), where Parry & Palmer's no. 335 from
San Luis Potosi is referred to it. This mention cannot constitute

334 PROCEEDINGS OP THE AMERICAN ACADEMY.

publication, however, as it is not accompanied by any description or
synonymy whatever. The specimen cited is E. micranthum, Less.
(B. Ugustrinum, DC).

Dr. Gray evidently got his idea of Schultz's E. glauciim from a plant now
in herb. Gray collected by SchafFner on Popocatapetl and labelled in
Schultz's own band ^' E. glaucum, Sch. Bip. iu Ehrenb. pi. Mex. no. 397."
This plant is a species evidently near E. micranthum, Less., but differing
in its impunctate leaves, &c.

In 1884 Dr. Klatt published, 1. c., the first description of E. glaucum,
but it is quite evident from the characters given as well as from a good
drawing and some fragments iu his herbarium that Dr. Klatt had quite a
different plant before him from either E. glaucum, Gray (nomeu
nudum) or E. glaucum, Sch. Bip. in herb. (coll. Schaffner). Identical
with Dr. Klatt's E. glaucum (which, being the first species described
under that name, must stand) is E. Orizabae, Sch. Bip., described on
the subsequent page (Leopoldina, xx. 90) by Klatt himself. This species
is clearly shown by the excellent specimens distributed by Liebmann,
no. 80, and Nelson, no. 1737 (from near Reyes, Oaxaca). The leaves
are small, thickish, and subsessile and the branches numerous and
ascending.

The original E. glaucum, Sch. Bip. in herb., which is quite different
was also distributed by Schultz under a name approaching " E. popoca-
tapetlense," but with some differences in the spelling. Mr. Ilemsley
refers E. popocatapetlense (noraen nudum) to Schlechtendahl and cites
under it Ghiesbreght's no. 529 (which with its glandular-punctate leaves
i-evolute at the base is E. micranthum, Less.). Schultz's name in Dr.
Gray's hp,nd appears on the label of Ghiesbreght's specimen in herb.
Gray, and the substitution of Schlechtendahl as authority is certainly a
clerical error in the Biologia Cent. -Am. and Index Kewensis. While
Schultz's species seems to be a good one, it would seem unwise to
launch it under a different spelling of a name already used in another
significance. It has therefore seemed best to describe it as above under
a new name E. capnoreshium.

The synonymy of the related species here mentioned may be stated
thus : —

E. MICRANTHUM, Less. Linnaea, v. 138 (1830), not of Lag. (which
was an Ageratum).

E. Ugustrinum, DC. Prodr. v. 181 (1836).

E. semialatum, Benth. PI. Hartw. 76 (1841).

E. popocatapetlense, Hemsl. Biol. Ceut.-Am. Bot. ii. 99 (1881).

ROBINSON. — MEXICAN PLANTS. 335

E. fflaucum, Hemsl. 1. c. 95 (1881), not Sch. Bip. nor Klatt.

E. CAPNORESBIUM, n. sp.

£!. glaucum, Sch. Bip. in herb.

Also under another unedited name by Schultz.
E. GLAUCUM, Khitt, Leopoklina, xx. 89 (188-4), not Sch. Bip. (nomen

nudum) nor Ilemsl. (nomen nudum).

K Orizabae, Sch. Bip. in Klatt, I.e. 90 (1884).
E. heterolepis. Soft woody stems glabrous, lucid : leaves opposite,
approximate, decussate, lance-oblong, entire, acute at each end, coriaceous,
1.2 to 1.4 dm. long, 3.6 to 4.8 cm. broad, nigrescent, pinnately and rather
obscurely veined, pulverulent under a good lens; petioles 2 to 3 cm. long;
heads 8 mm. long, 1 to 1.2 cm. in diameter, about 18-flowered, borne oq
filiform flexuous divergent pedicels in an open pyramidal panicle ; scales
of the involucre of two kinds, the inner loag, linear, acute, about equal-
ling the flowers, stramineous, at length deciduous; the outer (about 6 in
number) ovate to lanceolate, acutish, much shorter, persistent, nigrescent:
corollas 5 to 6 mm. long, with slender proper tube and distinctly en-
larged throat : pappus of about equal length, somewhat sordid ; achenes
glabrous, 2 mm. long. — Collected by Dr. Sereno Watson in " eastern
portions of Vera Paz and Chiquimula," Guatemala, 1885, no 65. Type
in herb. Gray. A very distinct species with a calyculate involucre re-
calling a Senerio.

B. Heydeanum. Stems somewhat flattened, especially beneath the
nodes, striate, tomentulose above, soon glabrate and slightly scabrous :
leaves opposite, ovate to ovate-lanceolate, caudate-acuminate at the apex,
cuneate at the base, entire, 3-nerved from above the base, sordid-tomen-
tulose upon the nerves above, at first sordid-tomentose at length merely
jiubescent and slightly paler beneath, thin, 1.7 dm. long, 7 to 8 cm. broad ;
petioles sordid-tomentose, 4 to 9 cm. long : panicle opposite-branched,
rounded, terminal, 1 dm. broad ; heads about 22-fiowered, 9 mm. long,
short-pedicelled ; involucral scales linear, acute, about 16, loosely imbri-
cated, sordid-tomentulose on the outer surface, the outer shorter : corollas
5 to 6 mm. long, gradually enlarged from the base to the short-toothed
limb, without sharply differentiated throat: achenes slender, 5 mm. long,
hispidulous on the angles; pappus sordid, of nearly equal length. — E.
Ekrenberc/ii, Coulter in J. D. Smith, Enutn. PI. Guat. iv. 74, not Hemsl.
— Collected by Heyde & Lux on Rio de Las Canas, Depart. Santa Rosa,
Guatemala, altitude 800 m., March, 1892, no. 3427 of Mr. J. Donnell
Smith's sets. Type in herb. Gray, E. Ekrenhergii, Hemsl., has serrate
leaves, much larger beads, and broader involucral scales.

886 PROCEEDINGS OF THE AMERICAN ACADEMY,

E. lanicaule. Stems 1 to 3 m. high, terete, striate, covered at least
when youug with shaggy tawny wool : leaves opposite, thickish, short-
petioled, oblong, acuminate, sharply serrate, narrowed to a crisped base,
appressed-villous above especially upon the veins, soft-pubescent upon the
pinnate prominulous veins beneath, concolorous, slightly reticulate, 1.3 to
1.6 dm. long, 2 to 5 cm. broad; petioles lanate, 5 mm. long; inflores-
cence somewhat rigid, rather dense ; lower branches widely spreading,
thickish ; bracts lance-linear, nearly equalling the heads ; involucre sub-
turbinate; the scales pluriseriate, elliptical, regularly imbricated, at length
campanulately spreading, finely striate, ciliolate : flowers 1.2 cm, long:
corolla 7 mm, long, cylindrical with little enlargement at the throat :
pappus slightly tawny, of the same length : achenes columnar, puberulent
on the angles. — Collected by Dr. Sereno Watson in the " eastern por-
tions of Vera Paz and Chiquimula," 188-5, nos. 74 a, 123. Types in herb.
Gray. Nearly allied to the W. Indian E. placheoides, Griseb., but dif-
fering in its larger heads, broad smoothish obtuse involucral scales, etc.

E. Loesenerii. Shrub, 2 to 3 m, high ; branches stout, terete,
opposite, curved-ascending, covered with a smooth gray cortex with
longitudinal rifts, the lenticels not prominent ; branchlets stoutish,
elongated, covered with a fine ochraceous tomentum (sometimes reduced
to a mere pulverulence) : leaves opposite, petiolate, broadly ovate or
oval, shallowly cordate, rounded or very obtuse with slight mucro at the
apex, obsoletely serrate or entire, 5 to 6 cm. long, 4 to 5 cm. broad,
subpalmately veined, pale green and pulverulent above, tomentulose and
ochraceous beneath ; petioles 2 to 3.5 cm. long, strongly channelled above,
subconnate at the base : heads numerous, about 20-flowered, in terminal
round-topped compound corymbs (9 to 12 cm. broad) ; pedicels 7 to 9 mm.
long; involucres tomentulose, 3 mm. long, campanulate, the scales sub-
equal, obtuse or barely acute, ciliate, ochraceous or purplish, scarcely at
all herbaceous : corolla white or pale purple, 4 mm. long, glabrous,
resinous-punctate on the teeth : pappus-bristles about 20, white, unequal ;
angles of the achene upwardly hispid, — Collected in Oaxaca by C. G.
Pringle on calcareous hills, Las Sedas, altitude 1,800 m., 30 October, 1894,
no, 6022, type (in herb. Gray), This plant was unfortunately determined
as E. sordidum, Less., from the not altogether satisfactory description
of that species. A comparison by Mr. J. M. Greenman of Mr. Pringle's
plant with authoritative material of E. sordidum now shows that the latter
is quite distinct, being in fact of the Hebecliniiim group. E. Loesenerii
has also been collected by E. W. Nelson along the Cuicatlan road, Valley
of Oaxaca, altitude 2,100 to 2,400 in., 3 October, 1894, no. 1549; by

ROBINSON. MEXICAN PLANTS. 337

L. C. Smith, Rancho de Calderon, San Juan del Estado, altitude 1,700
m., 4 November, 1894, no. 275 ; and by C. & E. Seler at Cuauhtilla, 28
November, 1895, no. 1537, between Yanbuitlan and Teposcolula, Oaxaca,
3 December, 1895, no. 1430, and Canada Sta. Maria, 8 December, 1895,
no. 1596. I take pleasure in dedicating this species to Dr. Theodor
Loesener of the Royal Botanical Museum at Berlin.

E. Nelsonii. Slender shrub, l.G to 3 ra. high: stems terete, striate,
dark brown, minutely gray-pubescent, somewhat flexuous ; internodes
rather long : leaves opposite, ovate, caudate-acuminate, dentate, about
7-nerved from the deeply cordate and more or less hastate base, mem-
branaceous, dull-green and finely gray-pubescent upon both surfaces, 8 to
10 cm. long, 5 to 8 cm. broad ; petioles flexuous, 3 to 4.5 era. long,
puberulent: heads about 16-flowered, numerous, 8 mm. long, borne in
op230site axillary and terminal trichotomous nutant corymbs ; peduncles
and pedicels pubescent, filiform; involucral scales very unequal, light
green, striate, elliptical, obtuse, rounded or erose at the summit, imbri-
cated in 3 or 4 rows: corollas at anthesis 4 mm. long, of essentially
uniform diameter throughout, thus not clearly differentiated into tube
and throat, pale, said to be yellowish : styles dark, strongly clavate ;
acheues (of the genus) hispidulous, 3 mm. long, tapering almost from the
summit to the base, at length subglabrate and nearly black. — Collected
by E. W. Nelson between Ayusinapa and Petatlan, Guerrero, 14 Decem-
ber, 1894, no. 2144; by C. G. Pringle on Monte Alban, Oaxaca, altitude
1,700 m., 21 December, 1894, no. 5637 (leaves less deeply cordate) ; and
by C. & E. Seler, in mountain woods between Yanhuitlan and Teposcolula,
Oaxaca, Mexico, 3 December, 1895, no. 1447. Types in herb. U. S.
Nat. Museum, herb. Royal Bot. Museum, Berlin, and herb. Gray. This
species has the involucre and clavate style-branches of a BrickelUa. The
achene, however, is distinctly that of a Eupatorium.

E. Oerstedianum, Benth. in Oersted, Vidensk. Meddel. 1852, p. 74.
Add syn. E. vtmonioides, Coult. Bot. Gaz. xx. 45.

E. oresbium. Upper part of the stem, peduncles, and pedicels
loosely pubescent with a sparse more or less deciduous tawny wool :
leaves opposite, ovate, acuminate, rounded or subcordate below, very
shortly cuneate at the attachment of the petiole, cuspidate-dentic-
ulate, thin, very finely pellucid-punctate, green and glabrous above,
scarcely paler and conspicuously villous upon the nerves beneath, 5—7-
nerved from a point somewhat above the base, 1.2 to 1.4 dm. long, 1 dm.
broad ; petioles 8 to 10 cm. long, puberulent, also tomentose along the
grooved upper surface : corymb open, rounded or subpyramidal, tri-
voL. XXXV. — 22

338 PROCEEDINGS OP THE AMERICAN ACADEMY.

chotomous ; pedicels filiform, 5 mm. long ; iuvolucral bracts thin, pale
green, very unequal, the inner oblong, obtuse, ciliate and minutely
puberulent, 6 mm. long, the outer much shorter, ovate-lanceolate, acute,
glabrous: flowers about 16: corolla purple, 6 mm. long, considerably
exceeding the delicate bright white pappus: achenes glabrous, dark
gray, 2 mm. long. — Collected by C. G. Pringle in a wet mountain
canon above Cueruavaca, Mexico, altitude 2,000 m., 15 February, 1899,
no. 8030. Three to five meters hish.

E. PACACANUM, Klatt, Botanisches Beiblatt zur Leopoldina, 1895,
p. 3. Add syn. E. roseum, Klatt. Bull. Soc. Bot. Belg. xxxi. 194 (1892),
not Gard. Both of Dr. Klatt's species here mentioned rest upon
Pittier's no. 3324. The existence of an earlier valid E. roseum neces-
sitates the adoption of Dr. Klatt's second name.

E. phoenicolepis. Stems terete, probably herbaceous, purplish
brown, covered with dense very short glandular sordid tomentum : leaves
opposite, ovate, acuminate, cordate with a narrow sinus, serrate, dull green,
pulverulent and sometimes bullate above, paler, tomentose and reticulate-
veiny beneath, 3-nerved from near the base, 5 to 7 cm. lon^, 2.4 to 4.5
cm. broad; petioles 1.4 to 3 cm. long; bracts similar to the leaves ex-
cept in their smaller size, extending up into the many-headed opposite-
branched round-topped panicle; heads 1 cm. in height, about 18-
flowered; bracts of the involucre purple, flat, striate, acutish, very
unequal, regularly imbricated in 4 to 5 series, the outer very short,
ovate, dark and dull, the inner lighter and somewhat lucid, about equal-
ling the flowers: corollas apparently deep purple or crimson, 5 mm.
long, equalling the delicate and not very copious pappus : style-branches
clavate; achenes 2.2 mm. long, upwardly hispidulous under a lens. —
Collected by E. W. Nelson between San Cristobal and Teopisca, Chiapas,
Mexico, 4 December, 1895, no. 3475. Types in herb. Gray and herb.
U. S. Nat. Museum. A species with the styles and to some extent the
involucre of a Brickellia but the achenes of a Eupatorium. It is most
nearly related to E. Bigelovii, Gray, which, however, has much more
acute scales and smaller leaves not cordate at the base.

B. photinum. Glabrous up to the slightly pulverulent-puberulent
inflorescence, herbaceous, erect, 5 dm. high; stems terete, subsimple,
dark purple, with rather long internodes : leaves opposite, ovate or
ovate-lanceolate, 7 to 11 cm. long, 3 to 4 era. broad, caudate-attenuate,
serrate except near the subacute base, green, glabrous, and somewhat
lucid upon both surfaces, 3-nerved from near the base ; reticulate veins
slightly translucent; petioles purple, 1.5 to 2 cm. long: heads small,

ROBINSON. — MEXICAN PLANTS. 339

5 ram. high, about IG-flowered, borne in small round-topped corymbs at
the ends of the opposite branches, also forming a larger dense terminal
corymb; pedicels 4 mm. long; involucral bracts subequal, oblong,
obtuse, pale green, striate, 3 mm. long : corollas nearly white, 2.5 mm.
long, the well-marked throat nearly as long as the proper tube : pappus
(not copious) about equalling the corolla: achenes black, 1.2 mm. long,
minutely roughened on the angles. — Collected by C. G. Pringle on
shaded cliffs of the Sierra de Tepoxtlan, 24 km. northeast of Cuernavaca,
Mexico, altitude 2,300 m., 7 February, 1899, no. 8029.

E. platyphyllum. Shrubs with stems pithy, 6-angled, stoutish but
flexuous as if scandeut, at first sordid tomentulose-pulverulent, at length
glabrate ; internodes rather long : leaves opposite, 1.2 to 1.8 dm. long, the
largest 2.2 dm. broad, shallowly 3-lobed, acuminate, rounded below to a
very shortly cuneate base, thin, green and glabrous or nearly so on each
surface, 3-nerved from the base, coarsely reticulate-veined, the veins pale
and slightly prominulous beneath; lobes acuminate, cuspidate-denticulate ;
petioles slender, 3 cm. long: heads large, numerous, (i0-7o-fiowered,
aggregated at the ends of the slender ascending branches of a broad
somewhat pyramidal panicle; pedicels filiform, about 1 cm. long; in-
volucre campanulate, 1 cm. high, 1.3 cm. in diameter ; scales pluriseri-
ate, regularly but loosely imbricated, very unequal in length, stramineous,
striate, scarious and erose near the rounded tips: corollas 6 mm. long,
probably white, very slender, without distinctly enlarged throat, equalled
by the copious bright white pappus : achenes glabrous, 2 mm. long. —
Collected by E, "NV. Nelson, near Chicharras, Chiapas, Mexico, altitude
900 to 1,800 m., 6 February, 1896, no. 3765; also by E. A. Goldman
at Metlaltoyuca, Puebla, no. 78. Types iu herb. Gray and herb. U. S.
Nat Museum.

E. Saltivarti, Sch. Bip. in Rob. & Sea. Proc. Am. Acad, xxviii.
108 (1893). Add syn. E. abronium, Klatt, Ann. k. k. Naturh. Hofmus.
Wien, ix. .855 (1894).

E. scabrellum. Stems angled and furrowed above, finely grayish
pubescent or tomentulose: leaves subalternate (at least the upper ones),
petiolate, ovate, scarcely acute, subcordate or rounded at the base,
crenate-denticulate, 5 to 16 cm. long, 3 to 12 cm. broad, thin but of
harsh texture, green upon both sides not paler beneath, glabrous to the
naked eye, but somewhat scabrous and under a lens puberulent, palraately
3-nerved somewhat above the pinnately several-nerved base : inflores-
cence an alternate-branched many-headed racemose panicle, its lower
branches (subtended by leafy bracts) 1.2 dm. long, naked below,

840 PROCEEDINGS OF THE AMERICAN ACADEMY.

branched and somewhat thyrsoid from the middle, pubescent ; invohicral
bracts about 10, oblong, pubescent, green, subequal (only 2.5 mm.
long) except the 1 or 2 calyculate external ones: flowers about 10:
corollas white, 2.5 mm. long, the throat funnel-formed without sharp
demarcation from the relatively short projier tube : achenes at length
black, 1.5 mm. long: pappus white, equalling the corolla, not copious. —
Collected by E. A. Goldman near Batopilas, Chihuahua, Mexico, alti-
tude 1,700 to 2,000 m., 4 to 5 October, 1898, no. 197. Types in herb.
U. S. Nat. Museum and herb. Gray.

E. scoRODONioiDES, Gray, var. grossedentatum. Leaves deltoid,
hastate, rather deeply cordate, coarsely crenate-toothed : pubescence less
copious and shorter than in the typical form. — Collected by C. G. Prin-
gle on limestone hills near Tula, Hidalgo, 20 September, 1899, no 8244.

E. Selerianum. Branches opposite, weak, flexuous, terete, striate,
finely pubescent : leaves opposite, petiolate, ovate, abrupt or cordate at
the base, acuminate or narrowed to an obtusish tip, glabrate above, gray-
tomentulose beneath, 1.5 to 3 cm. long, two-thirds as broad, crenate ;
petioles 1 to 1.5 cm. long: inflorescence a lax flat-topped compound
corymb ; the branches filiform, opposite or alternate, bearing 1 to several
small nearly orbicular short-petioled bracts (4 to 5 mm. in diameter) ;
pedicels 4 to 7 mm. long, minutely bracteolate : scales of the involucre
about 13, oblong, pale green or stramineous, striate, obtuse or erose at
the summit, the inner about 4 mm. long, the two or three outer consider-
ably shorter; heads about 15-flowered; flowers apparently pale or white:
corolla 2.3 mm. long ; the throat exceeding the proper tube : pappus
white, not very copious : achenes black, glabrous. — Collected by C. &
E. Seler, at Hacienda Petapa, Tuxtla, Chiapas, Mexico, 1896, no. 1939.
Type in herb. Gray. This species is related to E. guadalupense^ but is
probably perennial and differs much in pubescence and in the presence
of the orbicular bracts.

E. Smithii. Shrub, 1 to 2 m. high, cinereous-puberulent or tomen-
tulose : branches opposite, spreading, covered with fissured buff cortex :
leaves opposite, ovate inclining to be deltoid, rounded at the apex, trun-
cate or subcordate at the base, pale green and merely pulverulent above,
cinereous-tomentulose and veiny beneath, crenate to entire, 2.3 to 6 cm.
long, 2 to 4.5 cm. broad; petioles 1 to 1.5 cm. long, channelled above:
inflorescences trichotomous round-topped corymbose panicles, seldom
over 5 cm. in breadth ; pedicels and pale green subherbaceous loosely
imbricated obtusish iuvolucral scales slightly woolly : flowers about 23 in
a head, 7 mm. in length ; corollas white, gradually enlarged from the

ROBINSON. — MEXICAN PLANTS. 341

base and without definitely marked tliroat ; segments ovate-oblong. —
Collected in Oaxaca by C. G. Pringle, on Monte Alban, altitude 1,800
m., 8 October, 189-1, no. 4959, type (in herb. Gray) ; by L. C. Smith
on mountains of San Juan del Estado, altitude 2,100 m., 4 November,

1894, no. 274 ; and by C. & E. Seler, above Porian, 27 November,

1895, no. 1547. This species stands very close to £J. Loesenerii and may
eventually prove a variety of it, but as yet, although each species has
been independently secured by three collectors, no intergradation has
been observed. The differences are as follows : E. Smithii has smaller
leaves of a more deltoid contour and with a cinereous rather than ochra-
ceous indumentum ; its inflorescences are considerably smaller, while the
involucral bracts are larger, more herbaceous, and less closely appressed ;
finally the teeth of the corolla are, on careful comparison, longer and
ovate-oblonw ratlier tlian deltoid as in E. Loesenerii.

E. Thieleanum, Klatt, Bull. Soc. Bot. Belg. xxxi. 191 (1892).
Add syn. E. myrianthum, Klatt, Botanisches Beiblatt zur Leopoldina,
1895, p. 3, founded on the same specimens and described in the same
words.

E. Valverdeanum, Klatt, Bull. Soc. Bot. Belg. xxxi. 188, (1892).
Add syn. E. chrysocephalam, Klatt, Botanisches Beiblatt zur Leopoldina,
1895, p. 2. Dr. Klatt's two species are founded upon the same speci-
mens and their descriptions are identical. The earlier name must, of
course, be adopted.

E. Vitalbae, DC. Prodr. v. 163 (1836). Add syn. E. Ecuadorae,
Klatt, Ann. k. k. Naturh. Hofmus. Wien, ix. 356 (1894).

MiKANiA ERIOPHOUA, Sch. Bip., var. chiapensis. Pubescence
more hirsute than in the type, the hairs spreading, not at all felted or
interplexed : leaves large, deeply cordate, becoming 2.2 dm. long, and
1.7 dm. broad: involucral scales early glabrate. — Collected by E, W,
Nelson at Chicharras, Chiapas, Mexico, altitude 900 to 1,800 m., 6 Feb-
ruary, 1896, no. 3769. This variety has much the habit of 3L pyrami-
data, J. D. Smith, but that has the leaves smaller, rounded at the base,
and the pubescence subappressed on the pedicels.

Porophyllum brachypodum. Shrub, scarcely or not at all glau-
cous ; stems somewhat rigid, branched toward the end, lucid : leaves
(only the upper ones shown) lanceolate to linear, narrowed to an obtus-
ish glanduliferous point, cuneate at the base, entire, 3.5 cm. long, 4 to 8
mm. broad ; petioles 4 mm. long ; glands on the blade all marginal, large,
subremote, on the narrower leaves often alternate : pedicels 5 to 8 mm.
long, slightly clavate, spreading ; involucral bracts oblong, rather broad,

342 PROCEEDINGS OF THE AMERICAN ACADEMY.

obtuse, tufted at the cuspidate apex, 8 to 9 mm. long, usually carinate,
pale green with broad diaphanous margins ; glands biseriate, approximate
to the keel, elliptical or oblong, not linear, amber colored: flowers about
16: corollas 7 mm. long, at anthesis about equalling the sordid pappus;
the slender proper tube about twice the length of the throat : achenes
slender, dull black, upwardly hispid, 7 to 8 mm. long. — Collected by
Dr. J. N. Rose at Guaymas, Sonora, Mexico, 5 to 11 June, 1897,
no. 1272. Near P. ohlus>'foUum, DC, which, however, has more acute
dark purple involucral scales marked with dark linear glands.

LiABUM Pkinglei, Rob. & Greenm. Proc. Am. Acad, xxxii. 49
(1896j. The type specimen of this species published by clerical error
as no. 6215 is 6214 of Pringle's Plantae Mexicanae.

Perezia Nelsonii. Tufted with rusty wool at the slightly decum-
bent base ; stem soon erect (probably 1 m. tall), subterete, striate,
branched, leafy about to the middle, glabrous and smooth, purplish:
leaves large (1.7 dm. long, 6 to 8 cm. broad), firm, reticulated, oblong-
oblanceolate, rounded, obtuse or acute at the apex, cordate-clasping at
the base, sharply dentate with small spreading teeth : branches ascending,
nearly naked : heads few, large, many-flowered, 2.5 to 3 cm. long and
broad; pedicels 2 to 5 cm. long, bearing several small lanceolate attenu-
ate appressed scale-like bracts passing gradually into the broader ovate
to obloufj or at length lance-oblono[ involucral scales, these s'^en with a
slight purple tinge, ciliolate, imbricated in about 7 rows, the middle and
inner ones obtuse or rounded and mucronate at the tips, the outer acute,
slightly spreading and pungent: purple strongly bilabiate corolla 1.5 cm.
long: columnar granulated achene 4 mm. long. — Collected by E, W.
Nelson on mountains near Talpa, Jalisco, Mexico, altitude 1,400 to
1,500 m., 7 March, 1897, no. 4037. A species near P. michoacana
and P. cuernavacana, but clearly distinct from either.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. Xo. 17. — March, 1900.

THE FRESHWATER TERTIARY FORMATIONS OF THE
ROCKY MOUNTAIN REGION.

By W. M. Davis.

THE FRESHWATER TERTIARY FORMATIONS
OF THE KOCKY MOUNTAIN REGION.

By W. M. Davis.

Receiyed January 29, 1900. Presented February 14, 1900.

CONTENTS.

1. Piedmont Depositories of Moun-

tain Waste 345

2. The Various Origins of Stratified

Deposits 347

3. Accounts of Tertiary Lakes in

tiie Rocky Mountain Region . 347

4. Characteristics of Lake Deposits 351

5. The Vermillion Creek Beds of

Wyoming 354

6. The Arapahoe and Denver For-

mations of Colorado .... 356

7. Lacustrine and Fluviatile Quater-

nary Deposits 359

8. Continental Deposits .... 360

9. Fluviatile Deposits .... 360

10. The Indo-Gangetic Fluviatile

Plain 362

11. Fluviatile Deposits of the Great

Plains 363

12. Fluviatile Basin Deposits . , 364

13. Rocky Mountain Basin De-

posits 365

14. Deposits in Arid Basins . . . 366

15. ./Eolian Deposits 369

16. Summary 370

1. Piedmont Depositories of Mountain Waste. — Extensive plains
stretch forward from the base of many mountain ranges. They are
formed of the waste brought from the valleys in the mountains by the
streams that have for ages drained them. Sometimes the plains are
open and slope gently to the seashore, as in northern India and northern
Italy ; sometitnes they occupy a basin drained through a gap in one of
its enclosing ranges, as in Hungary and California ; sometimes they fill
the floors of interior continental basins from which no river escapes to
the sea, as in Utah and Nevada, Persia and Central Asia. Although
Piedmont fluviatile plains are familiar geographical features, their genetic
association with mountains is seldom stated explicitly. It is seldom that
sufficient importance is given to the complimentary relation in which the
plains stand to the mountain valleys whence their materials have been
derived, even though some emphasis may be allowed to the fact that the

346 PROCEEDINGS OF THE AMERICAN ACADEMY.

form of all vigorous mountains is much more due to the carviug of val-
leys in their uplifted mass than to irregularities in their initial form. The
hahit of regarding rivers as destructive agents, except near their mouths
where they may build deltas in bodies of standing water, seems to have
retarded the recognition of the imiwrtance of rivers as constructive agents.
It is perhaps for this reason that the first suggestion made as to the
origio of piedmont plains generally ascribes them to deposition in the
sea or in lakes. It only after closer study has been given to their com-
position, structure, and fossils, as well as to the activities of aggrading
rivers, that the prevailing fluviatile origin of most piedmont plains has
been recognized. Instead of having been built up from the floor of the
sea or of a lake basin, fluviatile plains have frequently accumulated
where a slow depression of the land — perhaps an isostatic movement, or
a compensation for a slow upheaval of the adjacent mountains — has
given them appropriate location. The thickness of piedmont plains on
continental margins, as in northern India, may then be much greater than
the height of their surface above sea level. The laterally confluent fans
of which these plains are built, each fan heading at the mouth of a moun-
tain valley, is good evidence of their origin in the subaerial deposits of
aggrading rivers.

Many piedmont plains are to-day slightly dissected, especially near the
mountain base, by the rivers that formerly built them up, as in northern
India. Others are more strongly dissected, as in the group of great fans
at the base of the Pyrenees in southwestern France. Some are tilted
and deformed, being thus added as marginal strata to the older mountains
from which their materials were derived ; the marginal ridges to which
the Righi belongs at the northern base of the Swiss Alps, and the Siwalik
range at the base of the Himalayas in northern India are composed of
strata of this kind. Yet common as are existing piedmont fluviatile plains,
either of interior or exterior drainage, they have not been frequently
recognized among geological formations. Their prevailing absence from
the geological column may be objective or subjective ; due on the one hand
to their real absence in past geological ages ; due on the other hand to the
interpretation of fluviatile deposits as formations of non-fluviatile origin.

One reason why the fluviatile deposits of piedmont plains are not more
generally recognized in their true relations is that they have no common
English name. If their flatter parts are described simply as plains, they
are not sufficiently distinguished from plains of other kinds. If their
steeper parts are described as slopes, they may be confused with the body
of the mountains from which they are derived. The want of an appro-

DAVIS. — ROCKY MOUNTAIN TERTIARIES. 347

priate general name for the aggraded floors of interior continental basins
is also unfortunate ; for several of their features are peculiar to them-
selves and all are highly specialized and well correlated.

2. The Various Origins of Stratified Deposits. — Let us look now at the
problem of fluviatile deposits from the other side. A century ago, it was
the habit of geologists to regard all stratified deposits as of marine origin.
About half a century ago, certain stratified deposits in which no marine
fossils were found came to be regarded as of lacustrine origin. In still
more recent years, the importance of fluviatile, jKolian, and other strictly
subaerial agencies in the formation of extensive deposits has been recog-
nized ; yet it is still almost habitual to attribute extensive bodies of strati-
fied deposits to a source in bodies of standing water, without a sufilcient
consideration of other possible origins. It is therefore the intention of
this paper to suggest a revision of certain so-called lacustrine deposits,
in view of the possibility that they may have been at least in part
formed by great rivers, by small streams, or even by the wash and
creep of a subarid climate, and that lakes may have had but a very
subordinate part to play in their accumulation. With this object in
mind, attention may be given first to some extracts from the accounts of
certain Tertiary formations in the Rocky IMountain region which have
been repeatedly described as fresh-water " lake deposits " in the reports
of our western surveys ; and further on it may be considered whether
many of these deposits are not in reality ancient analogues of the fluvia-
tile deposits in modern piedmont plains.

3. Accounts of Tertiary Lakes in the Rocky Mountain Region. —
Among the earliest references to Tertiary lakes in the Rocky mountain
region are those by Hayden. He wrote : — "I would infer that this
great fresh-water lake [White river] must have spread over 150,000
square miles" (Geol. Surv. Terr., 1st Ann. Rep. (1867), 58). " With the
commencement of the Tertiary was ushered in the dawn of the great lake
period of the West. The evidence seems to point to the conclusion that
from the dawn of the Tertiary period, even up to the commencement of
the present, there was a continuous series of fresh-water lakes all over
the continent west of the Mississippi river. . . . The earliest of these
great lakes marked the commencement of the Tertiary period, and seems
to have covered a very large portion of the American continent west of
the Mississippi, from the Arctic sea to the Isthmus of Darien. . . . Every
year, as the limits of my explorations are extended in any direction, I
find evidences of what appear to be separate lake basins, covering greater
or less areas" (Ibid., 2nd Ann. Rep. (1868), 114, 115).

348 PROCEEDINGS OF THE AMERICAN ACADEMY,

In a chapter on " The ancient lakes of western America," Newherry
discussed " the history of this old lake country, as it is recorded in the
alternations of strata which accumulated at the bottoms of its water
basins" (Geol. Surv. Terr., 1870 (1871), 333).

Powell says that the bad-laud rocks of the Green river basin " are the
sediments of a dead lake, and are quite variable in lithologic character-
istics. We find thinly laminated shales, hard limestones, . . . crum-
bling bad-land rocks, and homogenous, heavily bedded sandstones "
(Colorado river of the west, 1875, 150). Again, the sedimentary beds,
north and south of the Uinta mountains, are described as having been
"deposited in waters whose shores were limited by the lower reaches of
the range; that is, they [the sedimentary beds] all gave evidence of lit-
toral origin " (Ibid., 1G6). Incidental reference is made to lakes by the
same author when describing certain strata of the Green river basin, in
such phrases as "the old shore line," " the old Bitter creek lake," "in
this great fresh-water basin," and " sands were accumulating in the bed of
the lake" (GeoL Uinta Mountains, 1876, 163-165).

Newton's Geology of the Black Hills, as edited by Gilbert, describes
the " bed of gravel " at the base of the White river Tertiary formation
as " piles of loose pebbles, having all the appearance of a gravel beach
on the sea-shore. Indeed, it is evidently a shore deposit, the remains of
the beach of the old fresh-water lake, formed before the waters attained
their greatest height and while they were sorting over material brought
by rivers from the neighboring Black hills." Further on, it is noted as
" worthy of remark that while the deposits of the Cretaceous, from the
demolition of which the White river Tei'tiaries were evidently largely
formed, are dark carbonaceous clays or shales, the latter are notable for
their light color. This is probably due to the making over of the sedi-
ment in shallow water, whereby the carbonaceous matter was oxidized,
leaving the accumulation* nearly white in color. A similar phenomenon
is to be seen in the work of the modern streams. . . . The rationale
appears to be that the finely divided silt derived from the Cretaceous
clays has by constant exposure to the air had its carbonaceous coloring
matter oxidized and burnt out " (Geol. of the Black Hills, 188, 189).

An essay by Cope on the "' Tertiary Formations of the Central Region
of the United States" (Amer. Nat., xvi, 1882, 177-195) is naturally
devoted for the most part to paleontological matters, but it contains such
statements as the following : — '■' The general history of the succession
of the Tertiary lakes of the interior of the North American continent and
their deposits has been developed by the labors of various geologists,

DAVIS, — ROCKY MOUNTAIN TERTIARIES. 349

prominent among whom must be mentioned Ilayden, Newberry, and
King" (177). " Extensive lakes were formed in the depressions of the
Laramie and older beds. . . . After an interval of time another series
of lakes was formed, which have left their deposits over a wider extent
of the continent than have those of any other epoch. These constitute
the beds of the Loup fork period" (179).

Many explicit quotations may be made from King's writings. In
those here following, the page numbers refer to the first volume of the
Fortieth Parallel Survey reports. The Vermillion creek beds in the
Green river basin are explained by accumulation in a depression " occu-
pied by an early Eocene lake, whose northern portion corresponded with
approximate accuracy to the present drainage-basin of Green river"
(359). After a description of localities, the author states that " from
the outcrops thus broadly sketched, it is clear that a single lake extended
from longitude 106° 30' to 112'^, stretching northward probably over
the greater part of the Green river basin and southward to an unknown
distance" (374). Reference is later made to the "great lake of the
Green river period" (393). King's recapitulation is as follows: —
" Tertiary time in the region of the Fortieth parallel is therefore repre-
sented by nine lakes : four Eocene lakes which occupied the middle
Cordilleras . . . ; two Miocene lakes, one in the province of the Plains,
the other in eastern Oregon and western Nevada ; and, lastly, the three
I'liocene lakes " (457).

The early pages of Marsh's monograph on the Dinocerata abounds in
references to Tertiary lakes. The Eocene lake of the Green river basin
"remained a lake so long that the deposits formed in it, during Eocene
time, reached a vertical thickness of more than a mile. ... As these
[Rocky and Wahsatch] mountain chains were elevated, the inclosed
Cretaceous sea, cut off from the ocean, gradually freshened, and formed
these extensive lakes, while the surrounding land was covered with a lux-
uriant tropical vegetation, and with many strange forms of animal life."
The Dinocerata " which nearly equalled the elephant in size, roamed in
great numbers about the borders of the ancient tropical lake in which
many of them were entombed" (U. S. G. 8. Monogr., x, 1886, 1, 6).

Button states that " The Tertiary system of the Plateau country [of
Utah] is lacustrine throughout, with the exception of a few layers near
the base of the series, which have yielded estuarine fossils. The widely
varying strata were accumulated upon the bottom of a lake of vast
dimensions" (Geol. High Plateaus of Utah, 158). The strata here
referred to are chiefly marls and hence may be largely lacustrine, but

350 PROCEEDINGS OF THE AMERICAN ACADEMY.

the detailed sections of the High Plateau Tertiaries by Gilbert and
Howell (Wheeler's Survey, iii, 157, 267) show the occurrence of numer-
ous alternations between shales, sandstones, marls, and limestones.

Referring to the Tei'tiaries of northwestern Colorado, C. A. White
says : — " They were deposited in great lakes, the existence, extent, and
elevation of which were respectively determined by the varying configu-
ration of the general land surface as elevation and degradation progressed
(9th Ann. Rep., U. S. G. S., 695).

Weed makes brief mention of " the lake whose sediments formed the
Livingston beds " in the lower Tertiary of Montana ; and of the gather-
ing of " lake waters in the subsidence that caused the deposition of the
great thickness of sandstones and clays that form the Crazy mountains "
(Bull. 105, U. S. G. S., 27).

In Clark's " Correlation of the Eocene Deposits of the United States,"
attention was not especially directed to conditions of deposition ; hence
his remarks on this subject may be taken rather as a reflection of general
opinion than as a result of direct and independent investigation. They
present only conclusions without argument, for example : — " From an
open sea of Cretaceous age, in which the life was marine, a gradual
change took place to great fresh-water lakes in which the typical Ter-
tiary deposits of the Interior were accumulated. . . . The great fresh-
water lakes continued with successively dismissed areas during the
remainder of the Eocene period. With the advent of the Neocene
an extensive area was again covered with fresh-water lakes which
finally became drained in the orographic movements accompanying the
elevaiiou of the Rocky Mountains" (Bull. 83, U. S. G. S., Ill, 132).

The Neocene correlation paper by Dall and Harris (Bull. 84, U. S.
G. S.) contains many extracts from accounts of the fresh-water Ter-
tiaries of the Rocky mountain region, in which frequent references to
great lakes seem to be approvingly quoted. As in the preceding paper,
the origin of deposits was not the chief subject of the authors' attention
in this publication.

The abundant details concerning the Tertiary formations of Nebraska
in Darton's recent report on the " Geology and Water Resources " of the
western part of that state (19th Ann. Rep., U. S. G. S.) include mention
of local and general unconformities, pebbly beds, cross-bedding, and
conglomerate-filled channels. Conditions of origin are hardly touched
upon in this report, but two phrases suggest deposition in lakes. A
certain soil bed is taken to " indicate that there was here a land sur-
face while the Gering formation was being deposited to the south"

DAVIS. ROCKY MOUNTAIN TERTIARIES. 351

(751) ; and volcanic ash is said to Lave been ''undoubtedly borne on
the winds and deposited in the waters which laid down the several Ter-
tiary formations" (761).

Barbour's account of the Devil's Corkscrew or Da^monelix (Bull.
Geol. Soc. Amer., viii, 305-314) reflects the generally accepted lacus-
trine origin of the Loup fork beds of Nebraska (Arikaree of Darton, the
formation that contains a number of old channels filled with conglom-
erate) in certain phrases, such as " on the sandbars of this lake " (307),
"plants which drifted into the lake" (312), "the inference from the
branching is that the seaweeds or rootlets . . . grew downward in the
sand, not upward in the water" (309).

4. Characteristics of Lake Deposits. — The extracts given above might
be greatly multiplied. A review of their sources will show that the
explajnation of fresh-water strata as lake deposits has been almost uni-
versal among the geologists of our western surveys. The deposits are
as a rule well stratified ; they contain no marine fossils ; they frequently
preserve an abundant land or fresh-water fauna and tlora. Without
explicit discussion of the various conditions under which such deposits
could be formed, the earlier observers seem to have taken it for granted
that all the fresh-water sediments were gathered in ancient lake basins ;
and the later observers have generally followed the belief of their prede-
cessors. In a word, stratification has been taken to mean deposition
under water and not merely by water. There does not seem to be a
single instance, in the accounts of the fresh-water Tertiary formations
above referred to, of a deliberate inquiry into the essential characteristics
of lacustrine deposits, followed by a comparison between the consequences
deduced from the assumed theoretical conditions and the facts deter-
mined by observation. In the absence of such an inquiry, it does not
seem too much to say that it became habitual to speak of strata bearing
fresh-water and land fossils as lake deposits, just as strata bearing marine
fossils are habitually spoken of as marine deposits. Under the guidance
of this habit of interpretation, it was no more worth while to enter into
a deliberate inquiry as to the origin of the so-called lake deposits of
Tertiary age than into an inquiry as to the origin of the accepted marine
deposits of earlier ages.

We may now turn to a theoretical consideration of lake deposits,
■without attempting to present anything more than a brief outline of
their inferred characteristics. It will not be necessary here to consider
small lakes, for those inferred to exist in the Rocky mountain region
were scores or hundred of miles in extent; and for the moment, shallow

352 PROCEEDINGS OF THE AMERICAN ACADEMY.

lakes need not be borne in mind, as their deposits will be discussed
further on. The present paragraph deals only with the deposits of
large and comparatively deep lakes. Such deposits are of two kinds,
marginal and central. The marginal deposits may be relatively coarse,
but as such they can extend but little distance from the shore line : the
marginal strata of subaerial origin should be more nearly horizontal
than those of subaqueous origin. The central deposits are shown by
studies of existing lakes to be of very fine texture, such as clays, marls,
or very fine sands. The stratification of these fine sediments must be
veiy even, with few variations in texture or composition. The move-
ments of the waters of large lakes, either in waves or currents, do not
suffice to sweep pebbles out to deep water ; hence conglomerates and
pebbly sandstones with inclined and cross-bedded layei's must be limited
to a narrow belt around the lake margin. It can hardly be imagined
that the sediments deposited on the floor of a large lake should contain
frequent alternations of finer and coarser beds, such as clays and sands ;
but it may be inferred that the gradual filling of such a lake would allow
the encroachment of the later marginal beds upon the earlier central
ones ; and thus a relatively thin cover of coarse and variable deposits
might come to overlie a heavy body of fine and uniform deposits.

If deformation or climatic change should cause repeated variations in
the area and depth of a large lake, a complicated series of lacustrine,
fluviatile, and subaerial deposits might result; but this will not be
further discussed for the present, as the reports referred to above sel-
dom explicitly recognize variations of area and depth, excepting such as
occur at the times of separation of successive formations to which differ-
ent names are given. The manner of mention of the water body in
which each formation was believed to have been deposited implies
clearly enough that it was thought to be a single, large, continuous
lake.

If we now turn to the deposits of large shallow lakes, they are found
to be more variable in composition, texture, and structure ; for the
waves may stir up the materials of the bottom and the currents may
shift the materials from place to place ; but conglomerates need not
be expected to occur among them except close to the shore line. Large
shallow lakes cannot, however, be of great importance geologically, for
they must be rare and short-lived : rare, because their production de-
pends on the accidental concurrence of unrelated conditions ; namely,
the crustal deformation of aflat region by a small and nearly uniform
amount over a large area ; short-lived, because their preservation de-

DAVIS. — ROCKY MOUNTAIN TERTIARIES. 353

pends on the accidental compensation of several imperfectly related
processes ; namely, marginal and central deposition, marginal and cen-
tral deformation, and lowering of level by erosion at the outlet, or in
interior basins by evaporation. It may be pointed out that the pres-
ervation of open fluviatile plains during the slow deformation of the floor
on which they are accumulating does not involve the balance of conditions
or processes so little related as in the case of extensive shallow lakes ;
for the rate of fluviatile deposition responds most delicately to any
deformation of the fluviatile plain. If slow local depression occurs, the
rate of aggradation increases and the swinging rivers build up the de-
pressed area about as fast as it sinks. If slow elevation occurs, the
swinging rivers tend either to degrade the uplifted area, or to con-
centrate' their deposits on the districts that are not uplifted and thus
maintain a generally even surface.

If a shallow lake of large extent should happen to be formed, it would
probably soon change by depression to a deep lake or to an arm of the
sea ; or by elevation, marginal deposition, climatic change, and erosion
of outlet to a land surface. Hence the deposits of shallow lakes are
likely to be followed either by deep lacustrine deposits or by marine
deposits, each recognizable as such ; or by fluviatile and other sub-
aerial deposits, particularly around the margin of their basin. The
converse of these statements does not seem to hold true; for an open
fluviatile plain is more likely to maintain itself as such than a shallow
lake is to persist as a broad and thin sheet of water. The deposits
of a fluviatile plain are therefore not likely to be associated with those
lakes, unless in interior basins.

If a review is now made of the descriptions of the so-called lacustrine
formations in the Rocky mountain region, frequent mention will be
found of strata that are strikingly unlike those which might be expected
fi'om the observation of existing large lakes and from the reasonable
extension of principles based on such observations ; but before giving
examples of this kind, it should be remarked on the other hand that
some of the deposits are entirely consistent with a lacustrine origin.
For example, King reports that there are even-bedded " paper shales "
of very fine texture and with numerous fossil fish in the Green river
basin, "giving general evidence of accumulation in still, rather deep
water" (40 Par. Surv., i , 447). Russell writes regarding the basin
of Lake John Day in the northwest : " Before the Columbia lava was
broken and tilted ... its surface over the whole of central Washington
and probably far into Idaho and Oregon, was covered by the waters of a
VOL. XXXV. — 23

354 PROCEEDINGS OF THE AMERICAN ACADEMY.

great lake" in which "the accumulation of fine sediment . . . went on
with but slight changes for a sufficient length of time for more than a
thousand feet of evenly bedded strata to be laid down one above an-
other " (Bull. 108, U. S. G. S., 22, 23).

It is evident from such examples as these that certain western Ter-
tiary deposits have a well characterized lacustrine facies : but it is also
evident that just in so far as Tertiary lakes are inferred from the occur-
rence of fine and uniform sediments, they are excluded by the occurrence
of coarse and variable sediments. It must be concluded that the lakes,
whose quiet waters permitted the accumulation of fine-textured, tliin-
bedded shales, had disappeared when the deposition of cross-bedded con-
glomerates began. It is possible that some variation of lacustrine con-
ditions may have been tacitly assumed by the writers of the survey
reports. It may have been taken for granted that subordinate varia-
tions of shore line and depth were produced by orographic movements
during the existence of each great Tertiary lake, sufficient to have
caused slight changes in the bottom deposits, but not sufficient to have
produced significant unconformities in the accumulated strata ; or it
may have been intended that some small share of the deposits were laid
down outside of the margin or above the level of the lake waters ; but
it would be going beyond the spirit as well as beyond the letter of the
reports above quoted to give a considerable value to deposits of subaerial
origin, resulting from alternations of land and lake conditions, in any of
the western Tertiary formations. "When a student reads these reports,
he will without question conclude that their authors regarded subaerial
deposits as of negligible volume because no mention is made of them,
and that the formations are essentially lacustrine because continual
mention is made of deposition in large lakes.

Reference will be made again to fine-textured deposits on a later page;
attention being now turned to special examples of coarse-textured and
variable strata such as are repeatedly described in the accounts of our
western " lake deposits."

5. The Vermillion Greek Beds of Wyoming. — Many authors might be
quoted to show how frequently various kinds of deposits in addition to
those of fine texture and even bedding occur in formations that are
described as lacustrine ; but in this section extracts will be made only
from King's account of the Vermillion creek (lower Eocene) beds in the
Green river basin, enclosed page numbers again referring to the first
volume of the Fortieth Parallel Survey reports. The strata that are
found at some distance from the inferred " shore-line of the lake," and

DAVIS. ROCKY MOUNTAIN TERTIARIES. 355

therefore within what may be called the central area of the basin are
"coarse red sandstones, iuterbedded with more or less clays and arenace-
ous marls of pinkish and creamy colors" (361), "easily eroded red clay
and sandy beds " (361), '' thin, reddish, flaggy sandstones . . . underlaid
by whitish clays" (364), "sandstones and calcareous shales, with slight
seams of lignite and several thin beds of a limestone " (365), "dark-drab
and gray gravelly sandstones" (367), " several bird-bones from a coarse,
gritty, buff sandstone " (367). Near the border of the basin, "con-
glomerates become more and more important, until directly north of the
upper canon of Weber river the mountain wall is composed of exces-
sively coarse conglomerate between 3,000 and 4,000 feet thick. It is
almost structureless, and lines of stratification can rarely be perceived.
The blocks of which the conglomerate is chiefly formed range from the
size of a pea to masses with a weight of several tons. . . . The rapidity
with which these conglomerates grow finer in advancing from the shore
along the Uinta is very conspicuous" (369). Sections here exposed in
certain canyons "display the graduation of the material from the coarse
conglomerate immediately in contact with the older rocks out toward
the north, until . . . they have become fine-grained, sandy beds, devoid
of pebbles " (370). Near Echo city are presented " 3,800 feet of strata,
chiefly of these Indian-red sandstones, containing toward the upper
limit gray shale-beds, with occasional sheets of fine conglomerate "
(371). "West of Concrete plateau there is an enormous development
of red sandstones and clays, with prominent belts of conglomerate, the
whole increasing in coarseness of sediment as it approaches the Uinta
on the south and the Wahsatch on the west. Here is an area about
sixty miles from north to south by fifty miles from east to west, which
is essentially a plateau of Vermillion creek beds, in general approxi-
mately horizontal, but in the vicinity of the Wahsatch rising to 14° "
(372). "The thickest exposures of the Vermillion creek series are
in the immediate vicinity of the Wahsatch • . . where is exposed not less
than 4,000 feet. The most characteristic section ... is made up of a
heavy, gritty series at the base, . . .characterized by. . . red sandstones
and clays. . . . The middle members are of finer material and are more
intercalated with clays, while the upper part of the series . . .is made
up of striped and banded sandstones" (375). "The entire Vermillion
creek series . . . was made up of sandstones and intercalated clays, with
more or less conglomerates near the old shores of the lake " (380).

It seems perfectly clear from the context that all this formation was
re"-arded as having been deposited within the lake waters, on the lake

356 PROCEEDINGS OF THE AMERICAN ACADEMY.

side and not on the land side of the shore line, below the water level
and not above it. Not even the coarse conglomerates are described as
the superaqueous or subaerial parts of marginal deltas ; no mention is
made of the gradual encroachment of marginal deposits upon the area
eailier occupied by lake waters. Whatever qualifying ideas the author
may have held on these points, they do not appear in his descriptions,
and no student would gather them from his text. Yet it is rather in
explicitness and detail of statement than in interpretation that this
report differs from many others.

6. The Arapahoe and Denver Formations of Colorado. — It was not
I^erhaps unnatural, at a time when little attention had been given to the
importance of subaerial deposits, that geologists should fall into the
habit of regarding all non-marine formations as lacustrine. To-day
there is more reason for critical discrimination, and several exti'acts
given further on will show that some geologists have been led to new
interpretations of the origin of certain western fresh water Tertiary
formations. It is not improbable that other than a lacustrine origin
would be attributed to many of these formations, or to many parts of
them, if they were now seen for the first time. There are indeed some
indications that an unpublished, perhaps unconscious change of opinion
has to some extent taken place on this subject, similar to that by which
many of the geologists of Great Britain have been transformed from sup-
porters of the theory of marine abrasion to that of subaerial degradation
in the production of peneplains. No one would suppose by reading
British geological essays of recent years that the British geologists of
to-day had very largely given up Ramsay's theory concerning plains of
marine denudation, and substitiited therefor the theory of subaerial
degradation, as advocated by Sir A. Geikie; yet conversation with a
number of them last year convinced me that such was commonly the
case. Similarly, it may be that our western Tertiary deposits are no
longer regarded as exclusively lacustrine by a certain number of Ameri-
can geologists who, although they have published nothing to indicate a
change of opinion, may have come by more or less unconscious revision
of theories to recognize the great accumulative work that various sub-
aerial processes can have accomplished in Tertiary time. But the
change of opinion cannot be universal, for some of the extracts given
above from the accounts of our western Tertiaries are from reports of
recent years, and one of the recent monographs of the U. S. Geological
Survey shows that some of our most experienced geologists still follow
the interpretation of earlier years in referring even coarse-textured

DAVIS. — ROCKY MOUNTAIN TERTIARIES. 357

fresh-water deposits to lakes, without any published discussion of other
possibilities.

The following extracts from the monograph on the " Geology of tlie
Denver Basin in Colorado," above referred to, gives a fair idea of the
emphasis that there is placed upon the lacustrine origin of the Arapahoe
and Denver formations ; all the page references here made being to
Monograph XXVII. of the U. S. Geological Survey. "After an
erosion of the Laramie beds ... a considerable fresh-water lake
[Arapahoe] was formed and sedimentation again set in. What the
exact area of this lake was it is not possible now to determine; its
extent was undoubtedly considerably larger than that covered by its
beds at the present day, especially to the northward" (31). "The
movement which caused the drainage of the [Arapahoe] lake . . .
was succeeded after a considerable lapse of time by a depression sufficient
to allow of the formation of a second [Denver] lake. . . . The nature
of the depression which produced such lakes without admitting marine
waters ... is not readily conceivable " (32). " The beds deposited
in the Denver lake reached a thickness of over 1,400 feet along the
flanks of the mountains, but were probably somewhat thinner toward
the middle of the basin" (33). Certain deposits are referred to as hav-
ing been laid down near the shore line, but always as if on the lake
side of it ; for example, " some exposures of Denver strata which clearly
show the immediate proximity of the old shore-line" (183); ". . .
materials might have been derived from the eastern shores" (201).
Several contemporaneous lava flows are said to have been " poured out
upon the surface of the sea-bottom " (34). One of these is " the basaltic
sheet of Table mountain which was poured out upon the floor of the
shallow Denver sea" (161 ; see also 291, 292) ; not that direct proof is
given of the presence of standing water into which the flows advanced,
but that the presence of standing water is involved in the theory of the
lacustrine origin of the underlying strata, even though some of them are
conglomerates. One of the most significant observations bearinor on the
conditions of deposition is likely to pass unnoticed by many readers, because
it is given only an inconspicuous place in the description of details : —
" The presence of considerable tree stumps in erect position with roots in
mud layers and broken trunks in sand or gravel, shows that the water
was shallow or even that low-land masses alternated with shallow seas.
Probably the latter was the case " (168). Nevertheless, the alternations
of low-land and shallow water here suggested are elsewhere unmentioned,
the usual terms for the area of deposition being " the sea," " the lake."

358 PROCEEDINGS OF THE AMERICAN ACADEMY.

The materials of the Arapahoe and Denver formations seldom suggest
typical lake deposits. "The lower 50 to 200 feet [of the Arapahoe
formation] were conglomerates, the upper 400 to 600 feet arenaceous
clays" (31). " In passing eastward . . . the conglomerates are gradually
replaced by sandstones " (153). " The lower 400 feet of the [Denver]
series are composed entirely of eruptive debris ; above this point Archean
and sedimentary debris are found in small but increasing proportion,
and above 900 feet the material derived from . . . Archean rocks is
largely predominant " (33). The debris here referred to is elsewhere
described as largely conglomeratic near the mountains; on advancing
over the Plains, the sediments become finer, but still contain ^^lentiful
coarse sands and occasional pebbles, with numerous alternations between
fine conglomerate, grits, sandstones, and clays (180, 193, 195).

Repeated instances are given of structures that are much more sug-
gestive of fluviatile than of lacustrine origin. " That the Denver beds
were deposited in shallow waters is shown by the frequent cross bedding
observable both in sandstone and conglomei'ate " (33). In the foot-
hills, " the sandy parts of the bed develop in places to wedge-shaped
masses exhibiting in their relations to each other and to the conglomerate
a very marked cross-bedding" (163). "The coarser-grained beds show
cross-bedding" (165). Describing a local clay deposit occurring as a
break in a conglomerate layer, it is remarked : " Probably the conglom-
erate succeeding it was dejjosited in turbulent waters" (177). "The
study of the conglomerate series made it evident that fine-grained beds
of local development might occur at almost any horizon" (177). On
the Plains at ten or twenty miles from the '• shore line," special mention
is made of " the irregular unconformable contact so frequently seen to
exist between a conglomerate or grit layer above and a clay or shale
below. . . . Often the unconformability is very marked. . . . The
changes in conditions of sedimentation which give rise to such strati-
graphical relations of consecutive beds were, however, common in both
Denver and Arapahoe epochs. Fine sediments were often disturbed
and locally removed at the beginning of periods of rapid deposition of
coarser materials' (180, 181).

Tlie fossils, both of plants and animals, give no clear suggestion of a
lacustrine origin. " Plant remains and standing tree stumps . . .
abound at certain horizons " (33). " The only animal remains yet
found in the Arapahoe beds are the bones of vertebrates of new and
remarkable types. These occur in the conglomerate along the foot-hills
and in the basal sandstones and overlying clays beneath the prairies. In

DAVIS. — ROCKY MOUNTAIN TERTI ARIES. 359

the conglomerate but few have been found, and these are more or less
worn ; iu the clays they are abundant and tlieir articulation, edges, and
muscular insertions are sharp and clearly defined. . . . They are found
at all horizons in the formation, and occur buried in the clays or sand-
stones or partially weathered out upon the surface " (lo-A). " Fossil
wood, leaves, and stems are abundant" (169) in certain strata of South
Table mountain. Within the limits of the city of Denver "there was
formerly a very good outcrop of Denver sandstones and clays, with cross-
bedding structure, and full of plant remains in certain layers. Here, too,
occurs a thin local seam of coal. ... In these same strata Mr. T. W.
Stanton found some molluscan remains, associated with plants, and a
small but perfect crocodile tooth" (193). The fossil moUusca seem to be
of Huviatile rather than of lacustrine types.

Nowhere in this report or elsewhere have I been able to find any
discussion of the share that fluviatile processes may have had in the
origin of the formations, otherwise so elaborately described. Hills has
suggested that some of the materials may have been brought from the
South park region which " then as now, drained into the Denver
basin," thus implying river action in the collection of land waste; but he
does not directly discuss the condition of deposition, although the context
indicates that he accepted the prevalent theory of a lacustrine origin
(Proc. Colorado Sci. Soc, iii, 1890, 393-394). Yet to my reading the
record of observations on nearly every page of the Monograph suggests
that a fluviatile origin is at least as probable if not more probable than a
lacustrine. Conglomerates near the mountains, pebbles and sands
alternating with clays on the plains, cross-bedding and local uncon-
formities, standing tree stumps and fossils of large land animals are
in my reading all witnesses to rivers rather than to lakes.

7. Lacustrine and Fluviatile Quaternary Deposits. — The body of
scientific opinion above quoted regarding the interpretation of our west-
ern fresh-water Tertiary formations as lake deposits stands in marked
contrast to another body of opinion that might be adduced regarding
the oriffin of the Quaternary basin deposits of the same region. The
Quaternary deposits tliat are interpreted as lacustrine are clays and
marls, with the addition of the strictly marginal gravels and sands near
the shore lines. Mere wedges of gravel between clay and marl beds
in the Bonneville basin, wedges that are trifling in volume when com-
pared to many Tertiary conglomerates that have been described as
lacustrine, are interpreted by Gilbert as indicating a reduction of Lake
Bonneville "so far as to bring subaerial agencies locally into play,"

360 PROCEEDINGS OP THE AMERICAN ACADEMY.

(Monogr. I., U. S. G. S., 193). Beds of cross-bedded gravel and sand,
associated with sandy loam, occur between a lower and a higher marl in
the Lahontan basin : of these Russell says : " the remarkable similarity
of the middle member of the Lahontan section, as exposed in certain
localities, to the . . . deposit formed by meandering streams, leads us to
refer its origin with considerable confidence to similar causes " (Monogr.
XL, U. S. G. S., 129). Indeed the whole theory of the variations of
Quaternary climate in the Great Basin depends on a subaerial origin
of certain gravel and sand deposits which are in many ways similar to
deposits that have been repeatedly described as lacustrine in accounts of
Tertiary formations.

8. Continental Deposits. — It was during a western excursion with
Professor Penck of Vienna in the summer of 1897 that a possible or
probable non-lacustrine origin of many of our western fresh-water
Tertiaries was first clearly presented to me. Since then, I have had
opportunity of seeing something of the great fluviatile plain of thePo, and
of recalling what I had long before seen of similar plains in California
and in northern India, as well as of reviewing several essays that bear
on the general problem here considered ; and the problem has thus come
to have an importance that warrants the present review and summary.

Penck's views on this subject may be found in his " Morphologie der
Erdoberfljiche " (ii, 24-36), where he discusses the occurrence of deposits
formed on subaerial plains in the older geological systems. Recognizing
that non-mai'ine formations may result under the action of various sub-
aerial agents as well as within lakes, he suggests the name, continental,
to include all such formations, leaving the discrimination of particular
deposits to further study. Penck's term deserves acceptance among
geologists, as an aid in the general consideration that it seems desirable
to give to the problem of our western Tertiaries ; they might be called
"continental" in order to avoid implication of either lacustrine or
fluviatile origin. Yet as far as the published descriptions of these
deposits afford evidence of their detailed structure, it appears to me
probable that streams and rivers have had more than lakes or winds to
do with their formation, and hence that "fluviatile" might often to
advantage replace "lacustrine" in describing them.

9. Fluviatile Deposits. — It is perhaps because so much has been
written regarding the erosive power of rivers that their constructive
powers have been too little considered; but their capacity to aggrade a
sinking area deserves as careful examination as their capacity to degrade
a rising area. When acting as aggrading agents, they spread out broad

DAVIS. — ROCKY MOUNTAIN TERTIARIES. 361

sheets of gravels, sands, or clays, the coarser sediments frequently show-
ing cross-bedding and local unconformities, the finer sediments generally
possessing an even stratification. The area over which such deposits may
accumulate is shown by existing fluviatile plains to rival that of the
western Tertiary deposits. In a region of considerable precipitation,
with a background of mountains from which abundant waste is shed to
lowlands in the foreground where the rivers have free discharge to the
sea, the activity of fluviatile aggradation is often equal to the average
activity of the deforming forces that tend to cause marine submergence
or to produce broad lake basins. A slight acceleration of littoral
depression might cause submergence, or a rapid local warping might pro-
duce a lake ; a pause iu these movements would allow the rivers to
convert the sea border or the lake into a fluviatile plain again. The
preponderance of one condition or the other might be determined by the
proportion of fine, evenly stratified layers (if such deposits are necessarily
marine or lacustrine, and out of the reach of river action) to variable
strata with cross-bedding and local unconformities in the resulting'
deposits, as well as by a study of the fossils that they contain.

The capacity of rivers to form extensive deposits of fine texture and
even stratification seems in particular to be underrated. It is true that
a torrential river, gathering coarse detritus and exposed to heavy floods
in its headwaters among lofty mountains, may carry cobbles and pebbles
many miles forward upon a piedmont fluviatile plain. The artificial
enclosure of its channel by dikgs to prevent overflow probably increases
the distance to which pebbles can be carried, as on the plain of the Po ;
but if the river is free to spread upon an aggrading surface, the pebbles
would be sooner laid down. In arid regions the coarse piedmont depos-
its assume great importance, as is more fully stated below. On the other
hand, rivers of moderate size, rising in uplands of moderate height, may
contribute chiefly very fine and well stratified sands and clays to the
plains that they aggrade. This will be especially true if their headwaters
drain regions of deep soils, such as occur on slightly elevated peneplains ;
or of weak strata, such as are found in basin deposits of earlier date. At
times of high water and overflow, rivers of this kind will spread layers
of fine silt far and wide over their plains, and the repetition of this pro-
cess must lead to the formation of thick deposits, fine in texture and even
in structure, with little admixture of coarser sands and pebbles. Ripple
marks, foot-prints, and raindrops may be preserved in the sediments of
shallow flood-plain lagoons, and mud-cracks may form as the lagoons are
dried up. The plains of the Po and of the Ganges, and the great fan of

362 PROCEEDINGS OF THE AMERICAN ACADEMY.

the Hwang-ho are very largely composed of fine sediments ; the propor-
tion of line to coarse materials in the extensive deposits of these rivers
seems to be greater than it is in many of the so-called lake beds of the
West.

The surface of many extensive fluviatile plains seems level as far as
the eye can reach. This level surface is the best obtainable index of
the evenness of structure that must prevail both in the strata already
laid down beneath the plain, and in those yet to be deposited upon the
plain. In the absence of special studies on the degree of continuity of
river deposits, it cannot now be said how far a single stratum or a group
of strata, marked by recognizable peculiarities of texture or color, may
extend ; but it may be urged that mere continuity of even bedded depos-
its, such as is reported in our western Tertiaries, even if occurring over
areas of many square miles, should not alone be taken as conclusive
evidence of lacustrine origin. Some other criterion than continuity is
needed to distinguish fluviatile from lacustrine deposits of fine texture.
No other feature seems so likely to serve this need as the filled channels
and lateral unconformities that must occur, albeit rarely, even in the finer
fiuviatile deposits. The occasional presence of these distinguishing struc-
tures might readily escape notice in beds whose continuity has been
traced only by observations of colored strata, such as are visible at a
distance on the barren slopes of arid regions.

10. The Indo-Gangetic Fluviatile Plain. — The alluvial deposits of
the Indo-Gangetic plain stretch over hundreds and hundreds of miles.
They are well described in the Manual of the Geology of India by
Medlicott and Blanford (or in the second edition by Oldham, 1893, 427-
458), where references are given to original articles for further details.
Gravels and conglomerates are abundant near the sloping borders of the
plain, while the prevailing material of the central area is some form of
clay, more or less sandy, with subordinate deposits of sand, gravel, and
conglomerates ; but pebbles are scarce at greater distances than twenty
or thirty miles from the enclosing hills. Borings show the deposit to be
hundreds of feet, and at one |)oint more than a thousand feet in depth,
with no trace of marine fossils ; and from this it is inferred that depres-
sion accompanied accumulation. Organic remains are not common, but
shells of river and marsh molluscs are occasionally found, and calcareous
material is not lacking ; the latter is frequently gathered in concretionary
nodules ; it sometimes forms compact beds of earthy limestone. The clays
bordering the Jumna, as well as the calcareous shoals of this river, have
yielded remains of a variety of vertebrates, including elephant, hippopota-

DAVIS. — ROCKY MOUNTAIN TERTIARIES. 363

mus, ox, horse, antelope, crocodile, and various fish. The borings yield
pebbles, sands, and clays, with peat and wood, and remains of terrestrial
mammals, fluviatile reptiles, and fresh-water molluscs.

Still more pertinent to the present discussion is the account given of
the inclined strata of the Siwalik (Tertiary) formation in the Himalayan
foot-hills (Ibid., 356-368). "The lower portion of the system is char-
acterized by a great thickness of fine grained grey, micaceous, pepper and
salt sandstone, iuterbedded with clay bands near its lower portion, while
the upper part of the system is composed of soft earthy clays, undistiu-
guishable from the alluvium of the plains, . . . and coarse conglomerates
of well rounded pebbles and boulders " (356). These strata are much
tilted; their thickness is estimated at 14,000 or 15,000 feet. The ver-
tebrate fauna of the Siwalik formation is well known to paleontologists.
" The earlier observers regarded this great series of beds as having been
deposited in a sea, a supposition which is sufficiently disproved by the
complete absence of any marine organisms, and by the occurrence of the
remains of fresh water molluscs, fishes, and tortoises. It is hardly possi-
ble that they could have been deposited in a fresh water lake, for it is
not conceivable that a fresh water lake extending the whole lengih of the
Himalayas could have existed. Moreover, the fresh water organisms
whose remains have been found are all such as inhabit streams, and not
lakes. The very close resemblance between the upper Siwalik beds and
the recent deposits of the Gangetic plain leaves little room for doubt that
thfe Siwalik beds were deposited subaerially by streams and rivers " (358).
While it does not seem necessary to deny the possibility of conceiving
the existence of a lake all along the base of the Himalayas, the necessity
for believing in such a lake seems to be removed by the striking re-
semblance between the upturned Siwalik strata and those of the Indo-
Gangetic plain.

11. Fluviatile Deposits of the Great Plains. — In view of these various
considerations and examples bearing on the competence of rivers to form
extensive stratified deposits of fine as well as of coarse texture, there
seems little room for doubt that some part of the fresh-water Tertiary
formations that stretch forward from the Rocky mountains across the
open slope of the Great Plains may be of fluviatile and not lacustrine
origin. The first observer to reach this conclusion was Gilbert, whose
views are to be found in a report on underground water in eastern Colo-
rado (17th Ann. Rep., U. S. G. S., 553-601). He ascribes the Tertiary
strata that unconformably overlie the Cretaceous of the Plains chiefly to
river action, but partly to transportation by wind and to deposition in

364 PROCEEDINGS OF THE AMERICAN ACADEMY.

lakes. The change from the erosion of the Cretaceous floor to the depo-
silioa of the Tertiary cover " was brought about by some modification of
conditions which is not yet clearly understood. Perhaps the plains region
was depressed at the west, and the slopes thus rendered so gentle that the
streams could no longer carry off the detritus which came from the moun-
tains, and it was deposited on the way. Perhaps a barrier was lifted at
the east, so that the base level stood higher. Whatever the cause, the
streams . . . filled their channels so that their beds lay higher than the
neighboring country . . . and they thus came to flow in succession over
all parts of the plains and to distribute their deposit widely, so that
the whole plain in the district here described was covered by sands and
gravels brought from the canyons and valleys, of the Rocky mountains.
The chief material is coarse sand ... in irregular beds with much oblique
lamination. In the sand are occasional pebbles and . . . beds of gravel.
... At the northeast, . . . clays, marls, and other fine-grained beds
alternate with the sand in the lower part of the formation, and these
are probably continuations of the lake deposits observed in Kansas"
(575, 576).

Haworth goes even further than Gilbert in excluding lacustrine condi-
tions in his discussion of the " Origin and Mode of P'ormation of the
Tertiary" in Kansas (Univ'y Geol. Surv. Kansas, ii. 1897, 281-284).
After quoting extracts from Gilbert's report, this author says : " The
relative positions of the gravel, sand, and clay of the Tertiary over the
whole of Kansas . . . correspond much better to river deposits than to
lake deposits. The irregularity of formation succession, the limited lat-
eral extent of the beds of gravel, clay and sand, the frequent steepness
of the cross-bedding planes, all correspond to river deposits, but are not
characteristic of lake deposits. ... It is quite possible that during
Tertiary time, in which there were so many changes in the velocity of
the water carrving the sediments, lesser local lakes and la^oous and
swamps and marshes may have existed in different places and for varying
lengths of time. But when we consider the Kansas Tertiarv as a whole
and yet in detail, it must be admitted that the materials themselves have
many indications of river deposits and a very few of lake deposits " (283).

12. Fluviatile Basin Deposits. — Broad plains frequently occupy
basin-like areas enclosed by mountains. Streams flow from the enclos-
ing slopes to the central depression, whence the united waters find escape
through a gorge in the bordering highlands. Such plains frequently
give the impression of having been once occupied by a lake. The plain
of Hungary is an admirable example of this kind, yet there is much

DAVIS. — ROCKY MOUNTAIN TERTIARIES. 365

probability that its strata have been deposited for tlie most part by
aggrading rivers. The sediments to-day laid down by the Theiss and
other rivers that wander over the central parts of the plain are of very
fine texture. Borings show that similar sediments underlie the surface
to depths of 100 to 200 meters. As described by Penck, these deposits
consist of a complex of fine sand and clay layers whose sections, dis-
closed in neiglil)oring bore holes, are so unlike that deposition in a lake
is held to be impossible. The plain of the middle Rhine is a longitudi-
nal graben, enclosed by uplands through which the river has cut its
narrow gorsje north of Bingen ; but here again the evidence of borings is
taken by Penck to be decidedly in favor of a fluviatile origin for the
deposits (Morph. der Erdoberfl., ii, 15). In both cases it must be con-
cluded tliat the deformation by which the basins were produced was so
slow that the production of lakes was prevented by deposition on the
depressed floor and by erosion on the rising rim. An older example of
this kind is offered by the fresh-water Molasse of Switzerland, flanking
the Alps on the north, and now uplifted, tilted, and eroded. It is often
referred to as a lacustrine formation, and its marls may well justify such
a reference ; but its heavy sandstones and conglomerates, such as are
now upturned in the marginal range of which the Righi is a member,
give strong suggestion of fluviatile origin. The resemblance of the Swiss
Molasse to the Siwalik beds of northern India has been pointed out by
Medlicott (Quart. Journ. Geol. Soc, 1868, 45, 46).

The subrecent deposits of Kashmir, Nepal, and Hundes, all basins
within the Himalayas, are of special interest in the present connection,
for they have been described as lake deposits, although now referred
chiefly to fluviatile agencies (Manual Geol. of India, 2nd edition, 422).
Their dimensions are comparable to those of some of our western Ter-
tiary basins. They frequently contain fine deposits in the more central
areas, and these may have been laid down in temporary lakes; but in
Kashmir the repeated occurrence of beds of shingle and sand alternating
with thin layers of lignite point to a subaerial origin ; and while the
central deposits of Ilundes are " a fine homogeneous clay with but little
gravel in it, . . . there is nothing to show that the whole [series of layers]
. . . might not be of subaerial origin, as it is almost certain that the
bulk of them might have been." The clays and gravels now dissected to
a depth of 3,000 feet contain mammalian remains, including the rhi-
noceros, ox, horse, hyena, sheep, and goat.

13. Rocky Mountain Basin Deposits. — The resemblance is so
strong between several of the basin deposits here described and the Ter-

366 PROCEEDINGS OF THE AMERICAN ACADEMY.

tiary formations of the Green river and other Rocky mountain basins,
that a fiuviatile origin for many of the latter becomes probable ; but re-
examination in the field with special attention to discriminating struc-
tures will be necessary before definite conclusions can be announced. It
is also probable that the basins produced by the Pliocene deformation of
the previously denuded Rocky mountains of Montana may have been at
least during part of their existence occupied by fluviatile plains as well
as by lakes. Hayden referred to them only as lakes : — " These . . .
broad valleys [of the Missouri headwaters in Montana] have all been lake-
basins during the last portion of the Tertiary period," and on another
page he says: — "The great valleys . . . during the latter Tertiary period
were the basins of fresh-water lakes, so that we have everywhere the
white and yellowish-white sands, marls, clays, sandstones, and pudding-
stones of the Pliocene lake deposits passing up into the Quaternary or
local drift" (Geol. Surv. Teir., 1871 (1872)", 147, 141).

The same comment may be made regarding deposits of the basins of
South park and San Luis valley, Colorado, which were described as
lacustrine by Stevenson (Wheeler's Survey, iii, 453, 461).

14. Deposits in Arid Basins. — In regions of drier climate, such as
interior continental basins, calcareous, saline, and alkaline matter may
slowly accumulate along with detritus of finer or coarser texture in the
central depression, while conglomerates, gravels, and sands would gather
to greater thickness in laterally confluent fans around the mountain bor-
ders. The importance of marginal deposits of this kind, both recent and
Tertiary, is attested by the following quotation from Powell, who says:
"I think that many geologists would ascribe this [Bishop mountain] con-
glomerate to the action of ice, but throughout all that portion of the
Rocky mountain region which I have studied, I have so frequently found
gravels and conglomerates of subaerial origin, and have in so many cases
found reason to change my opinion concerning them, often having attrib-
uted a drift-like deposit to glacial action, and afterward on further study
abandoned the theory, being able to demonstrate its subaerial origin, and
witnessing on every hand the accumulation of such gravels in valleys,
and over plains where mountains rise to higher altitudes on either side,
and having in many cases actually seen cliffs breaking down and the
gravels rolling out on the floods of a storm, I am not willing to disregard
explanations so obvious and so certain for an extraordiuary and more
violent hypothesis. . . . Nor need the thickness and extent of this
Bishop mountain conglomerate serve to weaken this explanation, for the
sub-aerial gravels in the valleys between the ranges in the Basin province

DAVIS. — ROCKY MOUNTAIN TERTIARIES. 367

are of equal and often of greater development. Whenever a low plain,
valley, or basin is for a comparatively long period but little elevated
above the base level of erosion, and during this time mountains and hills
stand about the lowlands, there must be a great accumulation of drift,
and where the highlands are areas of progressive elevation and the low-
lands areas of progressive subsidence this accumulation may continue
indefinitely" (Geol. Uinta Mountains, 170, 171).

Button's Report on the High Plateaus of Utah may be quoted to the
same end : " There is another class of conglomerates which claims our
special attention. These are of alluvial origin, formed, not beneath the
surface of the sea nor of lakes, but on the land itself. They do not seem
to have received from investigators all the attention and study which
they merit. . . . Throughout great portions of the Rocky mountain
region they are accumulating to-day upon a grand scale and have accu-
mulated very extensively in the past." After describing the transporting
action of torrents in mountains, and their depositing action at the moun-
tain base, the same author says: "The formation thus built up is an
'alluvial cone.' . . . The slopes near the circumference usually lie
between 1° and 2° ; those near the apex between 2° and 3h°. The
lengths of the radii of the bases often exceed 3 miles, sometimes exceed
4 miles, and seldom fall below 2 miles. ... So nearly together are the
gateways along the mountain and plateau flanks, each having its own
alluvial cone, that the cones are confluent laterally; giving rise to a con-
tinuous marginal belt . . . consisting of alluvial slopes which are sensibly
nearly uniform. . . . Our surprise is often great at finding the cone
wonderfully well stratified." Examples are then given of heavy Tertiary
conglomerates which are ascribed to an alluvial origin, although they are
" about as well stratified as the average of those which are attributed to
sub-aqueous deposition" (Geol. of the High Plateaus, 1880, 219-223).

The occurrence of calcareous materials other than pebbles in fluviatile
and subaerial deposits is seldom discussed. Marls and limestones, even
if impure, are taken as evidence of marine or lacustrine conditions.
Yet earthy limestones are explicitly recognized among the strata of
the fluviatile plain of northern India, as above noted, and calcareous
material may be deposited as a cement in the piedmont wash of an arid
reirion. Concerninsr the latter. Hill writes as follows: — "Throughout
the limestone regions of the hot climates of America a superficial crust
of white-lime material is found, called tepetate. Sometimes it is compar-
atively free from foreign material, or occurs as the matrix or cement
of conglomerates. This is a concentrate of the lime which has been

368 PROCEEDINGS OF THE AMERICAN ACADEMY.

dissolved from the surface, transported in solution by the torrential
streams, and redeposited through evaporation. . . . Tepetate is forming
great incrustations around the margins of the bolson plains of northern
Mexico " (18th Ann. Rep., U. S. G. S., 256).

Many other references might be made to the importance of the sub-
aerial deposits or " wash " at the base of sub-arid mountains, for such
deposits are well known in Utah, Nevada, and southern California, as
well as in more distant parts of the world (see Hilgard, Cienegas of
Southern California, Bull. Geol. Soc. Amer., iii, 127 ; Manual Geol.
of India, 2nd edition, 417, 418 ; Blanford, Superficial Deposits in the
Valleys and Deserts of Central Persia, Quart. Journ. Geol. Soc, xxix,
1873, 493), but a return to the reports of Fortieth Parallel Survey will
suffice. Here one may find abundant testimony to the competence of
subaerial processes to form extensive deposits flanking mountain ranges,
but attention is given almost exclusively to the coarse, unstratified de-
posits that are formed by storm floods near the mouths of mountain
valleys. For example: — "The interior valleys of the Cordilleras,
from California eastward to the "VVahsatch range, are all filled to a
varying depth with subaerial Quaternary accumulations. ... In each
one of these [Great Basin] depressions is a considerable covering of
angular and sub-rounded Quaternary gravel, always of an evidently
local character, directly to be traced to the flanking mountain ranges.
Its coarseness varies from large bowlders, weighing many tons, to fine
gravel, sands, and clay. Except where it has been rearranged in the
now extinct Quaternary lakes, it is altogether an unstratified deposit,
brought down by the rush of floods from the flanks and canons of the
mountains " (40th Par. Surv., i, 460). The sands and clays that are
gradually washed far forward from the piedmont fans of coarse gravels
and conglomerates can hardly have been in mind when describing these
subaerial deposits as " altogether unstratified." I cannot find that any
cross-reference was made from the account of these heavy unstratified
conglomerates to the description of the almost structureless conglom-
erates, between 3,000 and 4,000 feet thick, already quoted from the
description of the Vermillion creek lake deposits (380). It is possible
that a re-examination of certain "lacustrine" conglomerates in the Rocky
mountain Tertiaries might lead to their explanation as arid subaerial
deposits.

The central deposits of arid interior basins may be as fine as the
marginal deposits are coarse. The playas or mud plains of Nevada, as
described by Russell, and the plains around Lob Nor in the central

DAVIS. — ROCKY MOUNTAIN TERTIARIES. 369

basin of Asia, as described by Svea Hedin, are composed of exces-
sively fine materials, yet they are rather of fluviatile than of lacustrine
origin in the ordinary sense of these words. If a shallow lake occurs
in the lowest part of interior basins of this character, it occupies but a
small part of the entire depression ; it is variable in position, shifting to
a new site as it is driven about by growing deltas; it is still more vari-
able in volume, changing with the weather, the season, the century, and
the climatic cycle. Only in epochs of moist climate does a playa lake
reach dimensions comparable to those attributed to the Tertiary lakes
of the Rocky mountain region ; yet in dry as well as moist epochs,
fine sediments gather in the basin, aggrading its floor. The manner in
which the fine mud of playas is distributed when the very shallow water
is agitated by wind is said to prevent the production of well defined strati-
fication. Ancient playa deposits would therefore be prevailingly of a
massive structure, instead of being finely laminated like typical lake
■ beds ; and they would be associated with wedges of coarser deposits that
were washed forward on the basin floor by intermittent streams.

15. kalian Deposits. — Wind-borne dust and showers of volcanic ashes
are both of importance as possible contributors to subaerial deposits,
particularly in regions where the streams are aggrading the surface,
ana where running water is incompetent to remove the aerial sedi-
ments. The recoCTuitioa of wind-borne dust is still a matter of uncer-
tainty, or at least a subject of disagreement. Volcanic materials are
ill recent years generally detected by the aid of the microscope.
Matthew has recently called attention to the importance of wind de-
l)Osits in an article entitled, " Is the "White River Tertiary an JEolian
Formation?" (Amer. Nat. xxxiii, 1899, 403-408). This author begins
by saying that the Loup fork beds have been " shown to be largely a
flood-plain deposit " (403), and closes with a statement that the Rocky
mountain Eocene is " probably a mixture of lake and fluviatile sediment
— what proportion of each would not be easy to determine" (408). He
objects to the origin of the White river clays in a lake because of the
size of the lake required, and because of the absence of an eastern
barrier and of shore lines ; furthermore, he states that the clays are
not well stratified ; they contain land mammals in abundance, but they
preserve no plants, no fish, and no aquatic reptiles or invertebrates.
While the included sandstones are thought to be river-laid, the clavs
are said to be better accounted for by aeolian action, such as is now
going on in the production of loess on the open grassy surface of the
subarid Plains. This article is of especial interest, not only from its
VOL. XXXV. — 24

370 PROCEEDINGS OF THE AMERICAN ACADEMY.

critical quality, but also because it is concerned with a fine-textured
formation which has been referred to a lacustrine origin with more
confidence than almost any other in the Rocky mountain region.

16. Summary. — The object of this paper is to promote consideration
rather than to announce conclusions. The author's leaninsj towards the
explanation of parts of our western Tertiary formations as fluviatile
rather than as lacustrine deposits is not to be concealed, but he would
prefer to leave the decision of the points here mooted to those who have
a larger personal knowledge of the Rocky mountain region than he has.
Observations recorded by western explorers and geologists are in no
case disputed ; descriptions of observed structures are accepted as if
they were first-hand facts ; but the origin attributed to the observed
structures is not always accepted, because it is thought that other
interpretations than those offered by the observers are in some cases
to be preferred. This essay is therefore not concerned with the direct
observation of facts in the Rocky mountain region, but with the the-'
oretical discussion of recorded observations, a discussion in which any
one may reasonably take part, whether he has visited the region under
consideration or not, provided that the facts of structure are well observed
and described by those who have been there.

It is in no spirit of antagonism, but simply from interest in a
theoretical problem that expression is here given to an opinion in a
matter of interpretation different from that adopted by the authors
whose observations are quoted; an opinion which it seems advisable to
express freely in order to bring the discussion of the problem to the
attention of geologists interested in this aspect of their science.

The points in discussion seem to be in the main these : — Lake waters
may receive materials of various kinds from their shores and from the
rivers running into them ; and the materials thus gathered, coarse and
fine, will be deposited in stratified arrangement near the border and over
the floor of the lake. Their volume will depend on the rate, area, and
period of accumulation. River basins may also, under certain con-
ditions, receive in their lower portions more material than can be carried
forward and discharged into the sea ; or interior basins not containing
lakes may be gradually aggraded by the materials washed in by streams
from the enclosing highlands : here again the materials may be coarse
and fine ; they may have a stratified arrangement, the area of deposition
may be small or large ; and the volume of the deposits will depend on the
rate, area, and period of accumulation.

Recognizing these possibilities, assured by observation in various parts

DAVIS. — ROCKY MOUNTAIN TERTIARIES. 371

of the world to-day, how can the products of similar conditions in the
past be recognized ? In both cases, the deposits are stratified ; in both*
cases, the deposits may include fine as well as coarse materials ; in both
cases, the area of deposition may be large as well as small ; in both cases,
the thickness of deposits may be great as well as light; in both cases,
the strata may bear ripple-marks, mud-cracks, cross-bedding, and other
indications of small and variable water depth. With all these similarities,
it would not be remarkable if a lake deposit were sometimes called a
river deposit, or if a river deposit were mistaken for a lake deposit ; for
the safe discrimination of the two classes of deposits must depend on
their differences, not on their resemblances. While the marginal sedi-
ments of a lake may be coarse, the body of the central sediments must
be fine and uniform. The marginal parts of a fluviatile deposit may also
be coarser than the forward parts, but the latter may be characterized
by frequent variations of texture and structure, and occasionally by
filled channels and lateral uncomformities. The oi'igin of the western
fresh-water Tertiary formations should be considered with all these
items in mind, and with an equal hosj)itality to the fluviatile and to the
lacustrine theory.

In whatever way the discussion on this subject may end, it may be
noted a considerable body of geological opinion will follow its decision.
The lacustrine origin of the fresh-water Tertiary formations has had an
acceptance so general and undisputed that it has for some years held a
place in the geological history of the Eocky mountain region as an
established doctrine. Many examples of this might be given. Button
writes : — "I know of no more impressive and surprising fact in western
geology than the well attested observation that most of that [High
Plateau] area has been covered by fresh-water lakes. . . . The marvel
is not in the fact that here and there we find the vestiges of a great lake,
but that we find those vestiges everywhere. The whole region, with
the exception of the mountain platforms and pre-existing mainlands, has
passed through this lacustrine stage" (Monogr. II., U. S. G. S., 216).
The occurrence of numerous Tertiary lakes is made the basis of infer-
ences concerning Tertiary climate by the same author : — '•' We know
that the Miocene climate of the west was moist and subtropical. This is
indicated by the great extent of fresh-water lakes in some joortions of
the west, their abundant vegetable remains, and the exuberance of land
life " (Ibid., 223) ; but fluviatile plains do not imply a moist climate.
A low stand of the western part of the continent during part of Tertiary
time as inferred by the same geologist from the long endurance of

372 PROCEEDINGS OF THE AMERICAN ACADEMY.

lacustrine deposition : — "A considerable number of large lakes being
'formed, the next process was the desiccation of these lakes and
the evolution of river systems. So long as the region occupied a
low altitude this process, we may infer, would be very protracted.
Before a large lake can be drained its outlet must be cut down. But
several causes in the present instance would combine to render this
action very slow and feeble. The elevation being small, the declivity
and consequent corrasive power at the outlet must be correspondingly
small. Moreover, the waters issuing from a large lake contain little or
no sediment. . . . Corrasion by clear water is an exceedingly slow pro-
cess " (Ibid., 218). On the other hand, fluviatile and subaerial deposits
may accumulate at considerable altitudes above sea level in interior
basins.

The preservation of numerous vertebrate fossils was explained by
Marsh as " probably, without exception, due to their entombment be-
neath the waters of the great fresh-water lakes which existed in this
[Colorado] region during Mesozoic and Cenozoic time " (U. S. G. S.,
Monogr. XXVII., 525). From the time when Warren first called
attention to the inclined position of the Pliocene strata of the Plains,
their attitude has been taken to prove a jiost-Pliocene elevation of the
Rocky mountain system by all writers who have considered the subject.
If the Pliocene of the Plains is fluviatile instead of lacustrine, a much
smaller elevation may be demanded.

The Tertiary lakes of the Rocky mountain district have become
stock subjects of geological teaching, as the subject is represented by
the text-books generally in use, and it is here that my own interest in
the matter is especially aroused. Dana, Leconte, Scott, and Tarr all
assert the existence of Tertiary lakes without qualification ; they
give no indication that a large share of the so-called lacustrine
formations may really be of fluviatile or other subaerial origin. Similar
statements are naturally made by the standard European text-books,
such as those by Geikie, Lapparent, and Credner, who naturally adopt
the lacustrine origin of our western Tertiaries without demur. The
older generation of geologists, who had a first-hand acquaintance with
the facts, may have interpreted " lacustrine " very liberally, including
therein a considerable share of marginal subaerial deposits ; but the
brief and direct statements of the text-books leave the coming generation
of geologists no option in the matter ; they will accept the completely
lacustrine origin of all the deposits so-called. The probability or possi-
bility of fluviatile origin is not given a chance to gain a hold in the

DAVIS. — ROCKY MOUNTAIN TERTIARIES. 373

mind, for its place is taken by an opinion already established in favor of
lacustrine origin. In all efforts to visualize the Tertiary geography of
the Rocky mountain region, our students are now led to imagine broad
sheets of level water surface, scores or hundreds of miles across, and
well deserving the name of "seas" often given to them in various re-
ports ; and the long slopes of subaerial plains to which an important
place may come to be given now receive very scanty consideration.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. Xo. 18. - March, lUUO.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.

ON THE DETERMINATION OF SULPHURIC ACID IN

THE PRESENCE OF IRON; A NOTE

ON SOLID SOLUTIONS.

By Theodore William Richards.

CONTIUBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.

ON THE DETERMINATION OF SULPHURIC ACID IN

THE PRESENCE OF IRON; A NOTE

UPON SOLID SOLUTIONS.

By Thkodoue William Richards.
Keceived February 5, 1900. Presented February 14, 1900.

Within the past year have appeared three interesting articles by
Kiister and Thiel * concerning several methods of precipitating bai'ic
sulphate free from iron when that element has been present in the solu-
tion. One of the methods which they have proposed is an excellent
addition to our always imperfect analytical repertory, and the chemical
world owes them a debt for the ingenious although exceedingly simple
suggestion.

While not wishinor in the least to detract from the merit and interest
of this work, I feel that attention should be called to several omissions
in their paper, especially those which concern theoretical considerations
involved in previous work upon the subject.

In their first paper, where the admirable analytical device is explained,
Kiister and Thiel ascribe the '' occlusion" of iron to the presence of the
ferric ion. It is interesting that this idea, which they themselves have
since shown to be false, should have led them to the desired goal. The
circumstance might be taken as an argument in favor of the proposition
that even an incorrect idea is better than none. In a reference to this
paper, Ostwald t points out that the occlusion is in all probability due
not to the ferric ion, but rather to a molecular complex ; and in their last
paper Kiister and Thiel describe experiments which prove beyond much
doubt the correctness of Ostwald's suggestion. Oddly enough, however,
they still cling, in their summing up of the matter, to the statement that
the ferric ion is the essential agent.

* Kiister and Thiel, Z. anorg. Cliera., 19, 97 (1899); 21, 73 (1899); 22, 424
(1900).

t Ostwald, Z. pbys. Chem , 29, 340 (1899).

378 PROCEEDINGS OP THE AMERICAN ACADEMY.

It is curious also that they should have overlooked the following state-
ment in a paper published over ten years ago by Professor Jannasch and
myself. In speaking of the cause of the phenomenon under consideration,
we said : '' Es Itisst sich uumittelbar annehmen, dass das Eisen als eine
Doppelverbiudung mit Baryum als Baryum-Ferrisulphat gefiiUt wird,
welches wahrscheinlich eiu Molektil gebuudeues Wasser (sogen. Consti-
tutionswasser) enthalt."* Klister and Thiel have offered new arguments
which they consider as being in favor of the first part of this proposition,
without appearing to recognize that the proposition is not itself new.
They have moreover ascribed an arbitrary formula to the substancej
while we did not feel prepared either to assume its exact nature or to
explain the mechanism of its occlusion.

Three years after this publication Schneider described some interesting
experiments in which evidence was given that the occlusion of the double
or complex compound is a variety of solid solution. f The idea was
new at the time, and did not meet with general favor ; but a knowledge
of the work would have saved Kiister and Thiel trouble.

In 1894, 1 after the passing of two more years, Gladding published a
modification of Lunge's method of precipitating baric sulphate free from
iron, which is the most convenient of any which had appeared up to that
time. He precipitated the iron with ammonia, filtered without especial
washing, acidified the filtrate, added an excess of baric chloride, and then
collected on one filter both the main mass of the baric sulphate and the
small amount of the same substance which he obtained from the ferric
hydroxide. The chief difference between this method and the method of
Kiister and Thiel is the fact that the latter pointed out the uselessness
of filtering off the ferric hydroxide before adding the bai-ic chloride. It
is quite probable, however, that Kiister and Thiel did not know of Glad-
ding's work, and hence independently devised the whole process.

Let us consider for a moment the nature and cause of this kind of
occlusion. In the first place it must be distinguished sharply from the
mechanical retention of mother-liquor in minute cells which all crystalline
precipitates exhibit. Even large crystals rarely contain cells of included
mother-liquor plentiful enough to increase their weight by more than the
fraction of a percent; and this impurity is of course chiefly water. It is
easy to test the magnitude of this inclusion by precipitating such a sub-
stance as calcic carbonate from a solution containing a large quantity of

* Jannasch and Richards, J. prakt. Chem. [2], 39, 821 (1889).

t E. A. Schneider, Zeitschr. phys. Chem., 10, 425 (1892).

X T. S. Gladding, Jl. Am. Chem. Soc, 16, 398 (1894); 17, 347 (1897).

RICHARDS. — SOLID SOLUTIONS. ■. 379

some easily identified substance such as sodic chloride. Mr. R. P.
Gushing kindly made a number of such experiments for me, and found
that not even from a saturated solution of salt did more than 0.1 percent
of the precipitate consist of sodic chloride. Of course this value fixes
the maximum, for some of the substance may well have been occluded
instead of included.

Very different from the mechanical process of inclnsion is the behavior
of baric sulphate. Here the impurities may amount to several perceiit,
and the foreign material is disseminated equally throughout the mass.
The experiments of Schneider show that the amount of the occlusion is
greater as the concentration of the impurity in the solution increases,
although not in direct proportion. It is very clear that we are dealing
here with a special case of the distribution law ; but in this case the
distribution can take place only at the moment of precipitation, because
afterwards the rigidity of the solid prevents free motion. Kiister and
Thiel evidently thought of this possibility, but they did not amplify the
idea. In the present paper I have retained the name occlusion for this
phenomenon because no other term seems to be applicable. There is
indeed a certain analogy between the occlusion of hydrogen in palladium
and the dissemination of a solute out of a solution into a solid ; and since
the name gives rise to no misunderstanding it will answer its purpose.

Obviously the study of the distribution law in other cases should throw
light upon this one by analogy. "When hydrochloric acid is distributed
between its aqueous solution and the vapor phase, the concentration of
the undissociated part alone is concerned in the distribution, because ions
can exist only in the solution. In the same way, when a moderately
strong acid is distributed between water and ether, only the undissociated
part of the acid comes into play, because ether does not cause important
dissociation. From these analogies one might infer that the o-roup which
is concerned in cases of concomitant precipitation are not ions, but rather
electrically neutral complexes. The inference is not a perfectly safe one,
for little is known about the possible degree of dissociation in solids ;
but experimental evidence is at hand to support this view. In 1894
Richards and Parker made the observation that the occlusion of baric
chloride by baric sulphate is much increased by the previous addition of
hydrochloric acid to the solution of baric chloride.* For example, the
addition of twenty cubic centimeters of strong hydrochloric acid caused
almost twice as much occlusion as the addition of ten cubic centimeters

* Proc. Am. Acad,, 31, 74 (1894) ; Z. anorg. Chem., 8, 420 (1895).

380 PROCEEDINGS OF THE AMERICAN ACADEMY.

in another otherwise similar case. The conditions attendant upon mixing
two liquids are so uncertain that it is unsafe to attempt the quantitative
interpretation of the relationship ; but qualitatively this result is wholly
in accord with the law of mass action, provided that the undissociated
group is supposed to be the one concerned.

The nature of the electrically neutral group with which we have to do
in the present case is less easy to imagine. Kiister and Thiel assume
that it is Ba[Fe(S04)2]2 ; but this is evidently an unsafe assumption.
The analogy to the case of chromium which they quote is an excellent
point, but they do not seem to have made themselves conversant with
the literature upon this subject, and hence apply their useful analogy
unfortunately.

They seem to consider that chromium exists in the green solution as
part of an acid ion, instead of in the basic semi-colloidal form ascribed to
it upon excellent grounds by both Recoura and Whitney.* One could
hardly expect a basic substance to form such a salt as Kiister and Thiel's
with barium.

Some light upon the subject is to be had from the study of the quan-
titative results in the paper of Jannasch and myself before mentioned.
We found that the impure baric sulphate precipitated from a ferric solu-
tion (containing enough sulphuric acid to have yielded 1.1608 grams of
pure baric sulphate) as a matter of fact weighed on the average only
1.1170 grams, and contained 0.0144 gram of ferric oxide, hence the baric
sulphate actually present amounted to 1.1170 — 0.0144 = 1.1026 grams.
This is 0.0582 gram less than the amount which should have been
present ; hence, supposing that no sulphuric acid remained in the solu-

SO
tion, there must have been 0.0582 -r. ,^t, == 0.0199 gram of sulphur

trioxide in the iron complex. But 0.0144 gram of ferric oxide corre-
sponds to 0.0216 gram of sulphur trioxide in ferric sulphate; hence
there was a deficiency of nearly two milligrams. If sulj^huric acid
remained in solution, the deficiency in the complex must have been still
greater. It is true that the observed difference is only half so great as
that demanded by a formula analogous to Recoura's, Fe4(S04)5(OH)2,
and is perhaps not very much greater than a possible experimental error;
but, as far as the argument goes, it points toward the existence of a basic
(complex in the precipitate.

* A convenient resume of the argument in favor of this point of view is to be
found in a recent article bv Whitney, Jl. Am. Chem. Soc, 21, 1075 (1899).

RICHARDS. — SOLID SOLUTIOXS, 381

Turning to the results of Kuster and Tliiel's work, we find other
arguments indicating the same thing. On pages 439-440 of their third
paper they poiut out that the addition of acid diminishes very much the
occlusion of the iron complex. This is wholly consistent with the present
thesis : for (according to the IVIass Law) the addition of acid must
diminish the concentration of the basic complex in the solution, and
hence (according to the distribution law) the amount which finds its way
into the solid. It is true that this argument is somewhat diminished in
force by two circumstances : first, because Kuster and Thiel have not
here considered the fact that hydrochloric acid increases the weight of
the precipitate by introducing into it more baric chloride ; and second,
because the addition of hydrochloric acid undoubtedly causes a chloride-
complex which is not occluded to nearly so great an extent as the sulphate-
complex by the baric sulphate. In order really to study the effect of
the hydrogen ion one should not rely simply upon the total weight of
the mixed precipitate, but should actually determine the amount of iron
present in the precipitate and compare it with the amount present in a
precipitate obtained from a solution containing the same amount of iron
and an amount of sodic chloride equivalent to the hydrochloric acid
used in the first case. In spite of these objections, however, the efi'ect
observed bv Kiister and Thiel is so ^reat that one must ascribe a
portion of it to the action of the hydrogen ion, and hence, according to
Whitney and Recoura. infer the substance concerned to be a basic com-
plex. Kuster and Thiel oifered no explanation for the action of the
hydrogen ion.

Again, Kiister and Thiel point out (pp. 437-438) that the hydrolysis
caused by dilution prevents the addition of much water from causing as
considerable a decrease in the absorbed material as would otherwise have
been expected. This is equivalent to saying that the complex which is
concerned is a basic one : for the hydrolysis undoubtedly involves the
ionization of hydrogen.

Yet another, althouorh less cogent, arofument in favor of the basic
nature of the complex is the fact that Jannasch and I found the impure
baric sulphate to be capable of retaining about three per cent of water at
250°. "While it is not impossible that this should be at least in part
simply held in minute cells or as crystal-water, one is inclined to
ascribe part of it to hydroxy! existing in the precipitate. It may
well be, however, that the extra water is simply dissolved in the
precipitate, as other undissociated substances are dissolved by baric
sulphate. This tendency to dissolve water may be the reason why

382 PROCEEDINGS OP THE AMERICAN ACADEMY.

many precipitates require prolonged ignition in order to attain constant
weight.

Over a year ago I was attracted by the observation that green chro-
mic sulphate is plentifully occluded by baric sulphate,* into beginning
a study of the subject of occlusion with the help of this phenomenon.
Chromium was selected instead of iron because its complex is more
stable and has been more thoroughly studied. The research was begun
by Mr. R. P. Gushing and is being continued by Mr. F. Bonnet. Since
the precipitated baric sulphate is green, one must infer immediately that
the green modification (or complex) is responsible for the occlusion.
The work is not yet ready for publication, but our data agree well with
those of Kiister and Thiel, while our explanation accords with that
given in the present paper. We are also studying aluminum sulphate,
which likewise seems to form a complex very considerably occluded
by baric sulphate ; although this fact does not seem to be generally
known. t

This paper is intended to serve as an introduction to two other papers
upon specific cases of occlusion. The chief points upon which emphasis
is desired are the following : —

(1) Occlusion from solutions differs from inclusion by being a more
intimate and a less mechanical mixture.

(2) Occlusion differs from adsorption in concerning the whole mass of
the precipitate, and not merely its surface. Hence after a precipitate is
once formed occlusion takes place only with extreme slowness, because
of the rigidity of solids. New evidence of this will be given in one of
the following papers.

(3) Different precipitates have almost as widely varying powers of
occluding other substances as different liquids have widely varying powers
of dissolving other substances. The pi'operty is no more to be predicted
than the power of dissolving.

(4) The process of occlusion is probably the distribution of an electri-
cally neutral simple substance or complex between the solution itself
and the solid at the moment of its formation within the solution. Hence
any change in the substances present in the solution which tends to
diminish the concentration in the solution of the group under consider-
ation, tends to diminish the amount occluded.

* This circumstance has been independently observed by Whitney, Jour. Am.
Chem. Soc. 21, 1075 (1899).

t Compare Peckham, Jl. Am. Chem. Soc, 21, 772 (1899).

RICHARDS, — SOLID SOLUTIONS. 383

(o) The nature of the groups which are occluded iu the cases of iron,
chromium, aud aluminum is not yet determined, but they are probably
basic iu nature, and identical with the complexes of Kecoura and
Whitney.

Harvard University, Cabibridge, Mass.,
February 2, 1900.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. VJ. — March, 1900.

THE SPECIES OF THE ORTHOPTERAN GENUS

DEPOT MEM A.

By Samuel H. Scudder.

VOL. XXXV. — 25

THE SPECIES OF THE ORTHOPTERAN GENUS

DEROTMEMA.

By Samuel H. Scudder.

Received February 8, 1900. Presented February 14, 1900.

The genus Derotmema was founded by me in 1876 upon a single
species found in northern New Mexico. Eight years afterward Saussure,
who had not seen this species, added two others, regarded as new, one of
which has since turned out to be identical with a species described in 1871
by Thomas as an Qidipoda coming from Colorado and Wyoming. None
have since been added to the list, but I have for a long time had in my pos-
session other new species, mostly collected by me in the Rocky Mountain
region ; and recently Mr. A. P. Morse has brought home from the
Pacific coast still others, so that I am now able to refer eight species to
this sfeuus.

They easily fall into two groups, in one of which the insects are of
relatively small size with very prominent eyes and rather long antennae ;
the tegmina are distinctly tapering with nearly straight costa and the inter-
calary vein subequidistant from the median and ulnar veins ; while their
more conspicuous markings are confined to the costal and inner edges.
In the other, the insects are of larger size, with normal eyes and ratlier
shorter antennae ; the tegmina do not taper, the costa is arched, and the
intercalary vein approaches the median much more closely than the ulnar
vein ; while the markings are in no way especially attached to the
margins of the tegmina. In both, the wings are colored at the base and
crossed by an extramesial fuscous band.

Saussure (Prodr. QCdip., 154, note) has pointed out that a more
correct derivative from the Greek words forminsf the basis of the ojeneric
name would be Tmetodera, and accordingly I apply this term to one of
these two main groups, but not the typical one, for which latter the
original form is retained.

Derotmema is a desert-inhabiting genus of slender CEdipodinae, found
largely on and about sage-brush and often simulating it in color. It is con-
fined to the western half of the United States, from the western margin
of the Great Plains bordering the Rocky Mountains to the Pacific.

388 PROCEEDINGS OF THE AMERICAN ACADEMY.

Forms with the hind margin of the pronotum angulate predominate east-
wardly, those with this margin rounded or very obtusangulate predominate
westwardly. The prevalent color of the base of the hind wings is pale
yellow or citron, but one species in the extreme east has this area generally
red, while in one found in the extreme west it is usually pale blue, though
both species have sometimes the normal yellow of the genus.
The species may be separated by the following table : —

Table op the Species of Derotmema.

A^. Smaller forms ; eyes very prominent ; tegmina distinctly tapering, with nearly

straight costa, the intercalary vein subequidistant between median and ulnar veins;

the more conspicuous markings of tegmina generally confined, and that narrowly

and subequally, to the costal and inner edges. . . . [Derotmema proper]

^1. Metnzona broad, the hind margin rounded, rarely feebly obtusangulate.

c^. Dark band of wings normally broader than the metazona, reaching the

margin posteriorly and percurrent, or broken only in the posterior axillary area.

1. laticinctum.

r2. Dark band of wings narrower than the metazona, not reaching the margin

posteriorly and lacking almost throughout the axillary area. . 2. delicatulum.

62. Metazona, even in female, no broader or scarcely broader than at eyes, the

hind margin distinctly angulate.

ci. Rugosities of metazona wholly irregular and not longitudinally pectinate ;

dark band of wings relatively narrow, nowhere broader and generally narrower

than length of pronotum, more or less narrowed in the posterior axillary area,

and never occupying more but generally less than half the outer half of the

wing ; base of wings pale blue or citron.

c?i. Metazona rectangulate or feebly obtusangulate posteriorly (rarely in

female considerably obtusangulate) ; costal margin of tegmina with dul'

fuscous and ashen tints in no way conspicuously contrasted ; basal half of

wings normally pale citron, occasionally subliyaline. . . 3. cupidineum.

d". Metazona very obtusangulate posteriorly ; costal margin of tegmina

with vivid contrasts of black and pallid, giving it an ornate appearance ;

basal half of wings normally pale blue, occasionally pale citron.

4. saussurearmm.
c^. Rugosities of metazona forming longitudinal ridges giving it a pectinate
appearance ; dark band of wings broad, subequal, much broader than length
of pronotum and occupying more than half the outer half of wing ; base of

wings red or citron ' . 5. liajjdeni.

A-. Larger forms ; eyes less prominent, especially in female ; tegmina of sub-
equal breadth, with arched costa, the intercalary vein closely approaching the
median and widely separated from the ulnar vein ; costal and inner edges of
tegmina not marked distinctively from the rest of the wing. . [Tmetodera.]
h^. Disk of pronotum relatively smooth, posteriorly obtusangulate ; tegmina
with contrasted markings on costal half, in large patches as in Conozoa.

6. rileyanum.

SCUDDER. — THE GENUS DEROTMEM A. 389

62. Disk of pronotura tuberculate, its liind margin rectangulate, rarely feebly
obtusangulate; tegmina with almost uniformly distributed markings, consisting
largely of longitudinal streaks or fleckings following the longitudinal veins, rarely
so arranged as to form massive blotches and then inconspicuous and simulating
Trimerotropis rather tlian Conozoa.
c^. Hind wings broad in the radiate area, after the manner of Circotettix, so
that they are more than three fifths as broad as long ; tegmina flecked through-
out rather than streaked 7. lentii/inustan.

c2. Hind wings of normal form, rarely more than four sevenths as broad as
long ; tegmina beyond the basal fourth marked almost exclusively with
longitudinal streaks 8. lichenosum.

(Subgenus Derotmema.)

1. Derotmema laticinctum, sp. nov.

Of relatively small size, fusco-ciuereous, flecked with fuscous. Head
with prominent subglobose eyes, the occiput ascending, more or less
clearly separated from the descending fastigium of the vertex by a
transverse ridge ; fastigium rather broad and depressed between high
lateral walls, traversed by a slight median carina, the lateral walls con-
verging anteriorly and continuous with the lateral margins of the frontal
costa, which is narrow and deeply sulcate above, but broadened at
ocellus and particularly at base ; lateral carinte of face strongly divergent
below; antennre somewhat longer (^) or somewhat shorter (9) than
the hind femora, the joints, or most of them, alternately pallid and
fuscous. Pronotum cinereous, more or less flecked with fuscous, mesially
constricted so as to be subselliform, the metazona broad, generally
broader than at the eyes, its posterior margin very broadly rounded, with
no or scarcely a trace of angulation mesially ; disk of metazona generally
smooth or nearly so, but sometimes, at least in the female, distinctly
rugulose, the lateral carinas distinct, but fading on the prozona. Tegmina
long, slender and tapering, strongly rounded apically, cinereous, heavily
flecked with fuscous along front and hind margins, and with some slender
streaks apically along the longitudinal veins, the intercalary vein subequi-
distant from median and ulnar veins ; wings citron yellow at base,
hyaline apically with some fuscous clouds or maculations at extreme
apex, and with a broad extramesial fuscous band, normally broader than
the metazona, percurrent or broken only in the posterior axillary area,
reaching and including the hind border but not attaining the anal angle,
and leaving free three lobes of the radiate area apically, in the axillary
area sending a shoot half way to the base. Hind femora cinereous,
blotched or fasciate with fuscous.

390 PROCEEDINGS OF THE AMERICAN ACADEMY.

Length of body, J*, 14 mm., 9, 22mm.; antennae, J', 11 mm., 9,
9 mm. ; tegmina, ^ , 16.5 mm., 9> 22.25 mm. ; hind femora, $, 9 mm.,
9,12 mm.

30 (? , 28 9 . Las Cruces, N. Mex., T. D. A. Cockerel], C. H. T.
Townsend ; Mesilla, N. Mex., June 30, A. P. Morse ; Phoenix,
Ariz., Oct. 16, T. D. A. Cockerell ; B^ort Whipple, Ariz., E. Palmer ;
forty miles east of Tucson, Ariz., E. Palmer.

Mr. A. P. Morse found this species common at Mesilla on dry soil,
especially on sandy ground. It flies freely and is not easily caught.

2. Derotmema delicatulum, sp. nov.

Of relatively small size, pallid testaceous, occasionally tinged with
rufous, flecked more or less with fuscous. Head with the face often
hoary in the male, with prominent subglobose eyes, occiput, fastigium
and frontal costa exactly as in the last species; antennae as long as (J)
or considerably shorter than ( 9 ) the hind femora, pallid, interrupted with
fuscous, Pronotum testaceous, often more or less rufescent in female,
flecked more or less with fuscous, somewhat selliform especially in the
male, the metazona broad, generally broader, especially in the female,
than at the eyes, its posterior margin very broadly rounded, but
occasionally showing signs of angulation mesially, the disk nearly
smooth but generally with raised points and occasionally slightly rugose
in the male, distinctly and rather sharply rugose in the female, the lateral
carina? rather distinct but fading on the prozoua. Tegmina slender and
tapering, marked as in the last species and with similar intercalary vein ;
wings pale citron yellow at base, hyaline apically, with a slight infuscation
of the veins at extreme apex and a narrow and somewhat cloudy
extramesial fuscous band, narrower than the metazona, almost completely
interrupted in the axillary area though sending in its upper portion a
short shoot toward the base, and by a trifle failing posteriorly to include
the hind margin, leaving at least four lobes of the radiate area free.
Hind femora cinereo-testaceous, obliquely and obscurely fasciate with
fuscous.

Length of body, $, 14 mm., 9 5 22 mm.; antennas, (^ , 8.75 mm.,
9, 9 mm. ; tegmina, $, 15.5 mm., 9? 21 mm.; hind femora, J* ,
8.75 mm., 9, 12.5 mm.

5 cj , 3 9- Lancaster, Cal., July 31, A. P. Morse; Mohave, Cal.,
Aug. 1, A. P. Morse; between Gila Bend and Yuma, Ariz., July 4,
A. P. Morse.

SCUDDER. — THE GENUS DEROTMEMA. 891

3. Derotmema cupidineum.

Derotineina cupidineum Scudd., Ann. Rep. Chief Eng., 1876, 513 (1876) ; Sauss.,
Prodr. (Edip., 156-157 (1884).

The type of the genus, and originally described from northern New
Mexico. I have since taken it on different western trips at Green River
and Alkali, Wyo., July 27 ; at White River, Col., near the Utah
boundary, July 24, Aug. 3 ; and at Castle Gate, Utah, Aug. 22. It is
one of the smallest species of the genus.

4. Derotmema saussureanum, sp. nov.
Derotmema saussureanum Bruner !, MS.

Of relatively small or medium size, fusco-testaceous, much mottled
with fuscous. Head with prominent subglobose eyes, followed centrally
by a fuscous stripe, the face often wholly or largely hoary, the form and
sculpturing of the occiput, vertex and front agreeing perfectly with the
two species described above ; autennje as long as ((^) or somewhat
shorter than ( 9 ) the hind femora, rufous, banded with fuscous, distinctly
in the male, obscurely in the female. Prouotum fusco-testaceous more
or less ferruginous, the lateral lobes margined below with hoary and
marked from the centre toward the lower posterior angle with a bright
white attenuated oblique stripe margined with black ; prozona mesially
constricted, at least in the male, so as to be subselliform, the metazona
scarcely or not so broad as at the eyes, its disk sparsely and irregularly
rugulose, often in the male nearly smooth, the hind margin very
obtusangulate, the angle distinct. Tegmina long, slender, and tapering,
dull ferruginous becoming subpellucid apically, the fore and hind margins
flecked alternately with dark fuscous and pallid, the contrasts between
which are vivid on the costal margin ; intercalary vein subequidistant
between the median and ulnar veins : wings either pale blue * or pale
citron at the base, pellucid at apex, with some iufuscation of the veins at
the extreme tip, and crossed by an extramesial, modei-ately broad and
solid blackish band, narrowed at the axillary area, nowhere broader than
the length of the pronotum, occasionally broken at the lower margin of
the axillary area, in that area sending a short spur baseward, attaining
the hind margin but not the anal angle, leaving two to three lobes
of the radiate area free, and not occupying in all more than half the
outer half of the wing. Hind femora ferrugineo-testaceous, fasciate with

* The tint as in the European (Edipoda ccerulescens or paler than that of our
Leprus wheeleri.

392 PROCEEDINGS OF THE AMERICAN ACADEMY.

fuscous aud especially with a strongly oblique forward directed median
stripe.

Length of body, ^, 14 mm., 9, 19.5 mm.; antennae, <?, 9.5 mm.,
$, 10.5 mm.; tegmina, J', 15.5 mm., 9, 20.25 mm.; hind femora, ^,
9.5 mm., 9? 12.25 mm.

45 (J, 34 9- California, in the following localities : San Bernardino,
July 16; Colton, July 17; Gabon Pass, July 18 ; San Diego, July 22 ;
Coronado, July 24 ; Los Angeles, July 26 ; Rubio Wash, Altadena,
July 29, — all the above collected by A. P. Morse ; and Anaheim,
Coquillett (Bruner).

Specimens with blue hind wings come from all the above localities

except Cahon Pass and Anaheim ; those with yellow hind wings come

from these two localities, and also from San Diego, San Bernardino and
Altadena.

5. Derotmema haydeni.

(Edipoda hai/deni Thorn., Ann. Rep. U. S. Geol. Surv. Terr., v. 460-461 (1871).
Derotmema haydeni Brun., Ann. liep. Nebr. Bd. Agric, 1896, 132 (1897).
Derotmema brunnerianum Sauss., Prodr. (Edip., 155-156 (1884).

First described by Thomas from Colorado and Wyoming, aud credited
by Bruner to the Yellowstone and Nebraska and by Saussure to
Colorado. I have specimens before me from Colorado, 5500', Morrison ;
Canon City, Col., P. R. Uhler ; Alpine, Tex., June 26, and Sierra Blanca,
Tex., A. P. Morse; and have myself taken it at Lakin, Kans., Sept. 1,
and in the following localities in Colorado : Cucharas, July 8-9 ;
Pueblo, Aug. 31; between Pueblo and Veta Pass, Aug. 27; and
Garland, Aug. 28-29.

The color of the base of the hind wings is sometimes red and some-
times citron. Specimens at hand with the latter color come from
Cucharas and Garland, Col., and from the two Texan localities; those
with red wings from all the localities excepting Sierra Blanca, Tex.

A yellow-winged female of this species, taken at Garland, was observed
in flight to produce from six to ten short sharp clicks in rajiid succession,
like the ticking of a watch but much faster.

(Subgenus Tmetodera.)

6. Derotmema rileyanum.

Derotmema rileyanum, Sauss., Prodr. Q^dip., 156 (1884).

Originally described from Idaho. "Salm Co." is given by Saussure,
but there is no such county, and probably Salmon City in Lemhi Co.

SCUDDER. — THE GENUS DEROTMEMA. 393

■was meant. I have specimens from Humboldt Station, Nevada, Green
River, July 21-31, and Alkali. Wyo., July 27 ; Salt Lake Valley, Utah,
Aug. 1-4, and White River, Col., next the Utah border, July 24-Aug. 3,
all but the first taken by myself. I consider this, rather than either of
the following, to be the species described by Saussure, principally from
the wider angle of the hind border of the pronotum, and the band of the
wiugs more noticeably broken at its crossing by the veins.

7. Derotmema lentiginosum, sp. nov. .

Of relatively large size, fusco-testaceous, much flecked with fuscous.
Head not very prominent, much infuscated above the clypeus, the
fastigium of vertex deeply channelled between high walls which gradually
approximate both in front and behind, and with a distinct median
carina ; frontal costa relatively narrow, subequal but gently expanding
at base, sulcate throughout, its margins continuous with the walls of the
fastigium; eyes only modei'ately prominent; ant?enn£e scarcely so long
as (S) or much shorter than ( 9 ) the hind femora, ferrugineo-testaceous,
very obscurely banded with fuscous. Pronotum ferrugineo-fuscous,
flecked with fuscous and especially marked on the lateral lobes
with a broad fuscous bar, darkest below, inclosing a narrow one of
testaceous, the lateral carinte sharp and distinct on the metazona, the
disk of which is rather delicately rugnloso-tuberculate with a slight
pectinate appearance, its hind margin rectangiilate or in the male some-
times slightly acutangulate. Tegraina shaped as in Triraerotropis, with
arched costa, cinereo-fuscous, with nearly uniformly distributed rather
coarse fuscous fleckings, sometimes obscurely grouped into patches near
the base and middle, rarely very obscurely streaked longitudinally in the
apical half, the intercalary vein approaching the median much more
closely than the ulnar vein ; wings broad and full in the radiate area,
somewhat after the manner of Circotettix, being about three fifths as
broad as long ; they are very pale citron at base, apically hyaline but
with nearly all the veins and cross-veins irregularly infuscated, and with
a moderately narrow generally unequal and percurrent but more or less
interrupted extramesial fuscous band, narrowest just below the axillary
area, and in the latter sending an offshoot half way to the base. Hind
femora ferrugineo-testaceous, rather obscurely fasciate with fuscous.

Length of body, ^, 19.5 mm., 9, 25 mm. ; antenna?, ^ , 9 mm., 9>
8.5 mm,; tegmina, ^, 23 mm., 9? 26 mm. ; hind femora, ^, 10 mm.,
9 , 12 mm.

894 PROCEEDINGS OF THE AMERICAN ACADEMY.

10 (?, 2 9 . South Park, Col., Aug. 11-16; Garland, Col., Aug. 28-
29, collected by myself.

8. Derotmema lichenosum, sp. uov.

Of relatively large size, fusco-ciuereous, much marked with fuscous
and ferruginous. Head not very prominent, the face much mottled with
cinereous and fuscous, generally darker above the clypeus, the fastigium
and frontal costa much as in the last species ; eyes only moderately
prominent even in the male; antennae about as long as {^) or much
shorter than (?) the hind femora, ferruginous, distinctly banded with
fuscous. Pronotum cinereo-fuscous, often flecked or blotched with bright
ferruginous (simulating the rusty lichens of the sage-brush) on the disk,
the lateral lobes with markings similar to those of the last species ;
lateral carinas sharp and distinct, at least on the metazona, the disk of
which is granuloso-rugose, its posterior margin rectangulate, rarely
feebly obtusangulate. Tegmina shaped as in Trimerotropis, with arched
costa, cinereous, marked with fuscous nearly uniformly distributed in
slender streaks and patches, showing a decided longitudinal tendency,
especially in the distal three fourths where they follow interruptedly the
course of the longitudinal veins ; occasionally an obscure tendency is
seen to their grouping basally, as common in Trimerotropis ; the inter-
calary vein approaches the median much more closely than it does the
ulnar vein ; wings marked quite as in the last species, the bund as there
reaching the hind margin but not the anal angle, and leaving three lobes
of the radiate area free ; but they are of the normal form, not full in the
radiate area and rarely more than four sevenths as broad as long.
Hind femora cinereo-testaceous, rather obscurely fasciate with fuscous.

Length of body, ^, 21 mm., ?, 26.5 mm.; antennoe, ^, 10.5 mm.,
9, 10 mm.; tegmina, J' , 24 mm., 9; 28.25 mm.; hind femora, ^,
10.75 mm., 9, 13 mm.

40 (?, 14 9. Alkali, Wyo., July 27 ; Green River, Wyo., July 21-
31 ; between Green River and Evanston, Wyo., July 31 ; Evanston,
Wyo., Aug. 6; Fossil, Wyo., Sept. 2; Castle Gate, Utah, Aug. 22;
White River, Col., at the Utah border, July 24- Aug. 13 ; Florissant,
Col., Aug. 17-22; South Park, Col., Aug. 11-16; Garland, Col.,
Aug. 28-29, all collected by myself.

The species was much more common in the northern of these localities
than in the southern, and was found on and about the sage-brush, its
colors rendering it most inconspicuous and sometimes showing, especiidly
upon the thorax and occasionally upon the head, rusty patches quite

SCUDDER. — THE GENUS DEROTMEMA. 395

similar to those, of lichen origin, upon the sage-brush. One could not
collect them without observing it.

Both sexes of this grasshopper make a noise in flight. A male,
observed at Evanston and repeatedly followed and then captured, made
from six to twentv-four clicks in a uniform series at the rate of seventy-
two in a minute, or six in five seconds ; a female at Alkali made a
succession of equidistant clicks, four or five to a second, in a short flight
in one direction, lasting only one or two seconds. These observations
were made, and most of the specimens collected, in 1877.

Note. — Since this article was in type I have found that this last
species was described in 1871 by Thomas as Oedlpnda gracilis. My
name will therefore fall.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 20. — April, 1900.

VIEW OF THE CARBONIFEROUS FAUNA OF THE
NARRAGANSETT BASIN.

By Alpheus S. Packard.

PALEONTOLOGICAL NOTES.

IV.* VIEW OF THE CARBONIFEROUS FAUNA OF THE
NARRAGANSETT BASIN.

By a. S. Packard.

Rficeived February 21, 1900. Presented March 14, 1900.

While the flora of the Narragausett coal basin is abundant, remains
of about eighty-eight species of plants having been detected in the black
carbonaceous shales and lighter sandstones, but few traces of animal life
have been found, these being mostly the wings of cockroaches and other
net-veined insects.

The age of these beds was originally supposed by the elder Hitchcock
to be Lower Carboniferous, but from a collection from the black shales
near the western edge of the Narragausett coal basin, at Providence
and Pawtucket, in the Museum of Brown University, sent by us for
examination by Lesquereux, he referred the beds to the Upper Carboni-
ferous, stating in a letter to us :

" These specimens, taken together, are interesting, as indicating more
than any other lot I have seen of fossil plants of Rhode Island, the strati-
graphical relations of your coal strata to those of the upper part of the
anthracite measures of Pennsylvania, where, even, I have not observed
such a predominance of species of Odontopteris typically allied to
those described by Fontaine and White from the Upper Carboniferous
of Pennsylvania." f

Besides the fourteen species of insects and an arachnid described by
Scudder t from the plant beds of Rhode Island, we have previously
noticed § the discovery of other fossil animals, viz., Spirorbis carbonarius,

* Paleontological notes I. -III. appeared in Proceedings Boston Society of
Natural History, xxiv., 1889, pp. 209-216. See also Recent discoveries in the Car-
boniferous flora and fauna of Rliode Island (Amer. Journ. Sci., 3d Series, xxxvii.,
p. 411, 1889).

t Proceedings Boston Soc. Nat. Hist., xxiv., 1889, p. 214.

I Insect fauna of the Rhode Island coal field. Bulletin U. S. Geo!. Survey,
No. 101, 1893.

§ L. c, p. 214.

400 PROCEEDINGS OP THE AMERICAN ACADEMY.

the supposed impression of an Annelid, and the track either of a mollusc
or worm. We now have to announce the discovery of additional animal
remains and tracks detected by Mr. J. H. Clarke of Providence, who for
nearly thirty years has industriously collected in this coal field. These
comprise casts of a fresh-water bivalve, Anthracomya arenacea, small
footprints, possibly those of a shrimp-like animal {Protichnites carhon-
arius), and an impression which seems to be rather worm-like than plant-
like.

We will enumerate these forms in a way to give a conspectus of the
animal life of che Narragausett coal basin, so far as known up to the
present time.

Annelida.

Spirorbis carhonarius. Pawtucket plant beds. (Scholfield and Gorham) Proc. Bost.

Soc. Nat. Hist, xxiv, 1889, p. 214.
Impression of an Annelid? Pawtucket plant bed. (Scliolfield) Ibid., p. 215.

Impression of a plant or worm ? This marking or impression was
found by Mr. J. H. Clarke in a boulder of fine red shale at South Attle-
boro, Mass. Whether it is an impression of an aquatic plant or of a
worm I am not sure, but am rather inclined to regard it as a worm-cast.
It is serpentine, with from four to five curves, no two curves alike. In
front it ends broadly, is pointed triangularly, tapering more at what
appears to be the posterior end. Length 70 mm. ; greatest breadth
4 mm.

Sections of %corm holes. I am indebted to Mr. J. B. Woodworth for
the opportunity of visiting with him an interesting quarry, one mile south
of East Attleboro, in red and green shales and light conglomerates, the
greenish shales showing distinct coarse ripple marks, rain-drops, and mud
cracks, besides numerous sections of worm holes, perpendicular to the
bedding. The worm holes are abundant, some eight or ten to the square
inch, and varying in size from ^^ to J of an inch in diameter. The
round deep holes had been excavated in a fine mud, and then, after the
worms had left them or died, silted up with fine sand. If the worms
which made these deep holes were fresh-water forms they were much
larger than any Nais like Oligochetes known to us at the present day,
and were possibly therefore marine.

MOLLUSCA.

Anthracomya arenacea (Dawson) Hind. (Fig. 1, A, B, C.) These
occurred in a small boulder of fine black shale found by Mr. Clarke at

PACKARD.

RHODE ISLAND CARBONIFEROUS.

401

Valley Falls. In this specimen there were about a dozen casts of
valves of old and partially grown shells with the shape and markings in
some cases well preserved owing to the fineness of the shale. Another
specimen showing well the shape of the valves was detected in the black
shaly plant-beds enclosing a vein of coal just north of Silver Spring, East
Providence, by Prof. F. P. Gorham, associated with the verticillate leaves
apparently of Calamites. In these specimens the valves are elliptical,
long, narrow, pod-like, the anterior end but little larger and rounder than
the posterior end. The umbones are situated at or between the anterior
^ and ^ of the shell. There are about twenty-five fine lines of growth.
The fully grown specimens are narrower than the young and the anterior
is but little larger and rounder than the posterior end. Size and
proportions of the largest examples : length 22 mm., breadth 9 mm., being
about 2^ times as long as wide. Length of the East Providence
example, 17 mm.; breadth, 9 mm. In this example the umbones are
situated near the anterior fourth of the valve. The young in the loose
boulder were 10 mm. in length, 5 mm. in width, or one-half as wide as
long, with numerous fine lines of growth.

A

Figure 1. — Anthracomija arenacea. X 2|. m, umbo.

The Rliode Island specimens present no differences from the descrip-
tion and figures of Dawson (Acadian Geology, 3d edit., p. 20.5). It is an
entirely different species from Naladites elongatus and laevis Dawson, of

VOL. XXXV. — 20

402 PROCEEDINGS OP THE AMERICAN ACADEMY.

which I have examples, kindly sent me by the late Sir J. W. Dawson.
It is also different from any British species figured by Wheelton Hind in
his elaborate monograph. Desirous of comparing our specimens with
types from Nova Scotia, I applied to Dr. G. M. Dawson, Director of the
Geological Survey of Canada, who very kindly sent me specimens from
the museum of the Survey, labelled in Sir J. W. Dawson's own hand-
writing, from Sydney, C. B. I can see no specific differences, although
the Sydney examples are casts in a rather coarse micaceous sandstone, a
less favorable medium for the preservation of specific marks ; but the
general shape and proportions of the valves are the same. The Sydney
examples are small specimens from 8 to 14 mm. in length ; each end is
much alike, and the beaks are plainly situated at the anterior |-^ of the
shell ; in the largest one at the anterior third. In Fig. 1, A represents a
large, and B a much smaller specimen in the pebble found by Mr. Clark,
and Ca specimen intermediate in size found in place by Prof. Gorham.
All are drawn to the same scale and are enlarged.

Track of a gastropod mollasc'^ Pawtucket plant beds. (Scholfield)
Proc. Bost. See. Nat. Hist., xxiv., p. 215. This track is, of course, doubt-
ful, and might have been made by a worm.

Merostomata.

Protichnites narragansettensis, n. sp. This name is given to a new
kind of track discovered in a pebble of dark arenaceous shale taken from a
kame in north Providence by one of my class, Mr. H. H. Mason. It is
allied to and evidently made by a species of perhaps the same group as
made the tracks described as Proticlinitcs octonotatus Owen, and P. loga-
nanus Marsh from the Cambrian. Description and figures are reserved
for a future occasion.

Crustacea.

Reviains of a Crustacean'^ Three fragments of the remains of what
appears to be a macrurous crustacean were found in the black shales of
Valley Falls by Mr. Clarke, associated with the leaves of Calamites. The
better preserved fragment is square at the base, with one side produced
above and ending squarely ; the lower corners are truncated. On the
lower edae of this fragment is a distinct raised boss or tubercle, while
the two other fragments are not thus marked. The plates remind one
of the epimerura of a shrimp, which is wider on the ventral edge than
above. The surface is polished but has not the markings of a Pandalus.
The surface is however marked with very fine irregular raised lines

PACKARD. — RHODE ISLAND CARBONIFEROUS. 403

passing across the surface. The edge is margined somewhat as in
Pandalus.

I was at a loss to what group to refer these remains, but on showing
them to Prof. C. E. Beecher he suggested that they might be crustacean,
and I am inclined to agree with him. The sides of the segments of
Acanthotelson are no wider ventrally than tergally ; there is also no close
resemblance to the segments of Palieocaris, both being from the Carboni-
ferous beds of Mazon Creek, Illinois. It is possible that they may be-
long to some true shrimp such as Anthrapala^mon or other macrurau of
that period.

Ostrakichnites carbojiarius {Protichnites carbonarius) Dawson, Acadian Geology 3d
edit., 1878, p. 55. Fig. 9, a.

Dawson describes and figures certain tracks from the millstone grit
formation at McKay's Head in No^a Scotia which he refers to
Protichnites and whicli he supposes to have been made by a Limulus-
like animal.

Somewhat similar but much less regularly arranged tracks occurred in
a boulder of fine red shale found in a stone wall at South Attleboro, kindly
given me by Mr. J. H. Clarke. The tracks were associated with mud
cracks, raindrops, and the worm-like impression already mentioned.

They are of the same size as those figured by Dawson, but are not so
regularly arranged, being much more scattered, and with no median
linear tail-mark. Yet the individual impressions are of the same shape and
size, and so like those of Dawson's Protichnites that they were apparently
made by the same kind of animal and could perhaps have been made
by the e.xtremities of the feet of a small shrimp-like creature.

The impressions are in sets of three, each of which is round in front,
deep and succeeded behind by a shallow faint furrow, showing where the
tip of the foot or spine of the hinder feet had trailed over the mud,
before the final impress of the feet was made. The three impressions
are not arranged in a straight line, but in a slightly curved line, showing
that the middle spine or claw was longer than the lateral ones. In some
cases there are single impressions forming two series about 8 mm. apart,
but with no tail-mark between.

The Protichnites tracks figured by Dawson, could not have been made
by a full grown Euproops or Prestwichia, and it should be observed that
the set of three prints is quite different from the long oblique crescentic
tracks made by the hind cephalic legs of Limulus. The tracks might as
well have been made by the crustaceans Gampsoiiychus, Anthrapaleemon,

404 PROCEEDINGS OF THE AMERICAN ACADEMY.

or Acanthotelson, whose legs end in a sharp point. As the tracks were
evidently not made by any merostome, we have thought it well to refer
these trails to a new genus, for which we propose the name Ostrahich-
nites, although we are not fully persuaded that it is worth while to
bestow names on these tracks, except for convenience of reference.

Arachnida.
Anthracomartus woodruffi Scudd. Pawtucket plant beds. (Rev. E. F. Clark.)

Insecta.

Mylacris Packard a Scudd. Bristol plant beds. (Rev. E. F. Clark.)
Etoblattina iUustris Scudd. Pawtucket plant beds. (J. H. Clark.)
" sp. Silver Spring, East Providence. (H. Scholfield.)

" clarkii Scudd. Pawtucket plant beds. (Rev. E. F. Clark.)

" scholjieldii Scudd. East Providence plant beds. (H. Scholfield.)

" sp. Fenner's Ledge, Cranston, near Providence. (F. P. Gorliain and

H. Scholfield).
" gorhami Scudd. Pawtucket plant beds. (F. P. Gorhara.)

" exilis Scudd. From a boulder near Kettle Point, East Providence.

(H. Scliolfield.)
" sp. Pawtucket plant beds. (II. Scholfield.)

" reliqiia. Pawtucket plant beds. (F. P. Gorham.)

Gerablattina scaptdaris Scudd. Pawtucket plant beds. (H. Scholfield.)

" fraterna Scudd. Silver Spring, East Providence. (H. Scholfield.)

Rhaphidiopsis dii-ersipenna Scudd. Cranston plant beds. (Rev. E. F. Clark.)
Paralof/us ceschnoides Scudd. Silver Spring, East Providence. (F. P. Gorham.)

The presence of the Spirorbis and of the tracks of two marine
Arthropods suggest that the Rhode Island plant-beds, even if in general
of fresh water origin, were deposited where the sea had access to them.
The presence of these marine fossils, with the fresh water naiad,
Anthracomya arenacea, strongly suggests that the horizon of the black
shales of Providence and also of the red and greenish beds of Attleboro,
Mass., belong to the same horizon as those of the South Joggins of Nova
Scotia, which is Upper Carboniferous, the rocks there consisting of sand-
stones and dark carbonaceous shales, frequently becoming reddish. The
South Joggins shales also contain the remains of Anthrapaljeraon, which
should be looked for in the Narragansett coal measures. Thus far, then,
the animal remains confirm Lesquereux's reference of the dark plant-beds
to the Upper Coal Measures.

These beds also appear to be higher in the series than the ]\Iiddle
Carboniferous Mazon Creek beds of Illinois which contain a larger

PACKARD. — RHODE ISLAND CARBONIFEROUS. 405

number of marine animals, viz., Belinuridse (Euproops, Prestwichia and
Belinurus), besides Anthrapakiemon and Acanthotelson, together with
the impressions of marine annelid worms.

It should be observed that the black plant-beds of Providence on the
western side of Providence Bay, which dip southeastward at an angle of
45° -50°, extend over to the arkose conglomerates on the we.stern edge
of the basin, which are very highly inclined and dip westwards. The
thickness of the plant-beds, unless much folded, appears to be over a
mile. The exact relations of the plant-beds on the eastern side of
Providence to the arkose conglomerates we have thus far been unable to
definitely determine.*

Brown University.

* Since reading this proof I have, at a point about half a mile north of Xatick,
on the western edge of the coal basin, seen the gradual passage of the carbcuiferous
shales into the arkose.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. Xo. 21. — Apkil, 1900.

CONTRIBUTIONS FROM THE CRYPTOGAMIC LABORATORY OF
HARVARD UNIVERSITY. — XLII.

PRELIMINARY DIAGNOSES OF NEW SPECIES OF
LABO ULBENIA GE^. — II.

By Rol.\xd Thaxter.

CONTRIBUTIONS FROM THE CRYPTOGAMIC LABORATORY OF
HARVARD UNIVERSITY. — XLII.

PRELIMINARY DIAGNOSES OF NEW SPECIES OF
LABOULBENIACE^. — 2.

By Roland Thaxter.

Received March 19, 1900 ; Presented April 11, 1900.

Dimorphomyces Myrmedoniae nov. sp.

Male individual tinged with smoky brown, relatively small, similar to
those of the other species, except that the neck of the antheridium is pro-
portionately much shorter, its base inflated, its distal portion short and
attenuated ; the distal cell of the receptacle large, distally very thick-
walled, the thickened portion undergoing a gelatinous degeneration.
Total length to tip of antheridium, including foot, 65 fx, to tip of sterile
portion 40 /x. Antheridium 38 /a long, the venter 25X14 /x.

Female individual relatively large, tinged with smoky brown, the
sterile part of the receptacle terminated by a blunt distal cell, subtended
by a blackened septum, the distal portion of which is thick walled and
often swollen or disappearing through gelatinous degeneration; the fer-
tile portion resembling that of the other species in general structure but
developed almost at right angles to the axis of the sterile part on one
side only, as in D. Thleoporae, very long and nearly isodiametric throuofh-
out ; the perithecia from two to five in number and the appendages from
three to six, alternating as a rule, the two series diverging slightly from
opposite sides, the subtending cells relatively large and slightly oblique
in reference to the marginal portion, so that their size appears to be
greater when viewed on one side than on the other. Perithecium
rather long and slender, nearly straight, stouter in small individuals and
somewhat inflated ; the tip blunt or nearly truncate ; tinged with smoky
brown, borne on a short divergent stalk-cell (hardly visible), from which
it bends abruptly upward. Appendages relatively large, consisting of
two superposed cells constricted at the septa, the basal one longer, distally
more deeply suffused, the upper distally suffused and modified like the

410 PROCEEDINGS OF THE AMERICAN ACADEMY.

corresponding portion of the sterile portion of the receptacle. Spores
about 20 X 3 [X. Perithecia in well developed individuals 100 X 18 ^.
Lateral portion of receptacle 75 X 20 jj., the sterile portion 60 /j.. Ap-
pendages about 58 X 13 yu..

On Mijnnedonia jiavicornis Fauv., British Museum (Biologia Coll.),
No. 766, Guatemala.

Dimorphomyces Thleoporae nov. sp.

Male individual much as in D. muticus, the receptacle ending in a
short blackened cell, bluntly rounded or nearly truncate. Total length
to tip of autheridium 66 ,u, to tip of sterile cell 37 ^.

Female individual. Structure of the receptacle like that of the other
species, but only developed on one side of the median sterile portion,
which consists of three cells like that of the male individual and is black
tipped. The appendages and perithecia arising as in the other species,
the latter nearly symmetrically fusiform, bluntly rounded at the tip, tinged
with smoky brown, the tip undifferentiated, the appendages (broken)
relatively large and simple. Perithecia (not quite mature) 50 X 14 /a.
Foot to end of lateral portion 50 fx. Total length to tip of perithecium
80 II.

On TIdeopora corticalis Gz., Paris Museum, No. 297, Santa Anna,
Madeira. Ou inferior surface of abdomen.

Dimeromyces pinnatus nov. sp.

Male individual consisting of a basal cell more than twice as Ions: as
broad, the axis above of eight or nine cells separated by horizontal septa,
all but the lowest and the terminal cell separating a small cell on one side
which forms the base of an antheridium or of a sterile appendage, the
two organs diverging slightly from one another so as to form two vertical
rows. Antheridium compound, short and stout, the venter abruj^tly dis-
tinguished from the stout neck, the base of which is slightly enlarged
and purplish brown, the distal part tapering very slightly, the apex
blunt. Appendages consisting typically of six cells, including the basal
cell, constricted at the dark septa, the distal cells suffused with brownish,
the terminal one larger, longer, and more or less vesicular, the thick
walls tending to gelatinous degeneration. Total length of receptacle
100 X 12 yu.. Antheridia 35 X 12 ^. Appendages 50 X 7 /x.

Female individual. Basal cell large and stout, the cells above it about
twenty in number, greatly flattened, the septa horizontal, a few of the

THAXTER. NEW LABOULBENIACE^. 411

lower cells having appendages on only one side or none, the rest bearing
them on both sides. Appendages simple, consisting of from five to seven
cells including the basal cell, constricted at the dark septa, the terminal
cell hyaline and much larger, as in the male ; the rest, except the basal
one, purplish or the upper one tinged with brown. Perithecia one to
three in number, mostly rather slender, sHghtly curved outward; the
stalk portion about half as long as the remainder, which is purplish
brown, deeper distally ; the tip well distinguished when not distended
by spores, consisting of a basal portion larger and slightly inflated and
a distal one, formed by the lip-cells, abruptly distinguished, its external
margins generally symmetrically divergent, the four cells nearly equal
and symmetrical and ending distally in a corresponding number of
papilla3 about the pore. Perithecia including stalk, 125 X 20 /x. Re-
ceptacle 190-225 X 28 IX, not including basal cells of appendages.
Appendages 55 X 8 /x.

On Ardistomis sp., Hope Collection, No. 296. No locality, but
probably Mexico or South America. At base of elytra and on leg.

Dimeromyces nanomasculus nov. sp.

Male individual very minute, consisting of three superposed cells, the
upper bearing a terminal two-celled appendage with dark septa, the sub-
basal cell giving rise to a single antheridium like that of D. pinnatus,
smaller, the neck usually abruptly turned to one side. Total length to
tip of appendage 35 /it,; the antheridium about 15 X 5 /x.

Female individual resembling that of D. pinnatus in general structure,
the basal cell large and long, narrower below ; the cells above about ten
to fifteen in number, usually roundish ; the appendages mostly five-celled
above their basal cell, variably suffused with brown, the septa dark, some-
times curved or almost hooked distally, the distal cell not conspicuously
enlarged becoming brownish ; the subterminal cell of the single terminal
appendage examined producing a blackish-brown, lateral, irregular, spine-
like outgrowth. Perithecium straight, more commonly solitary near the
tip, brownish yellow to dark brown, not at all distinguished from the
stalk, which is hyaline only at its narrow base, the hyaline tip abruptly
distinguished by a slight subtending ridge, its margins usually converging
symmetrically to the truncate or blunt apex. Spores about 45 X 3 yu..
Perithecia, including stalk, 100-120 X 20-24 ix. Appendages longer
60 X 6 /x. Total length to tip of perithecium 245-285 X about 22 (t.

On Ardistomis viridis Say, Cocoanut Grove, Florida. November.
On A. educta Bates, British Museum (Biologia Collection), No. 676.

412 PROCEEDINGS OF THE AMERICAN ACADEMY.

MONOICOMYCES nov. genus.

Receptacle consisting of a basal and subbasal cell, above which it ter-
minates iu a small two-celled sterile portion, the terminal cell of which
may or may not be in the form of a short appendage ; the subbasal cell
giving rise to from one to several fertile branches, the habit becoming
thus unilateral, bilateral or subverticillate in different species. The fer-
tile branches consisting of from one to several cells in different species,
the terminal cell of each branch normally giving rise to a stalked peri-
thecium and a stalked antheridium ; the remainder, if there are more
than one, appendiculate on the upper side, rarely (abnormally ?) produc-
ing an additional antheridium. Antheridium of the compound type, con-
sisting of a stalk composed of a pair of cells, the antheridium proper
consisting of certain basal cells, two tiers of peripheral cells, which sur-
round (not on all sides?) numerous antheridial cells and a cavity above
them, and three or four terminal cells, which appear to surround an open-
ing through which the antherozoids are discharged, and which subse-
quently grow upward, forming terminal simple appendages of irregular
length.

Monoicomyces Homalotae nov. sp.

Receptacle consisting of very small hyaline basal and subbasal cells
surmounted by a distal portion, very much as in Dimorphomyces, wliich
is blackish brown, the lower cell larger and distally inflated. Fertile
branches normally two, when the individual is bilaterally symmetrical,
rarely three, consisting of a single cell which bears distally an antheridium
from its outer and a perithecium from its inner angle. Perithecium
amber brown slightly asymmetrical, relatively very large, shoi-t and
stout, the lower lialf greatly inflated, generally more so on the outer than
the inner side, distally conical, the small tip not distinguished, usually
abruptly truncate, the stalk-cell hyaline, narrowing to its base. Anthe-
ridium borne on a pair of rather short stalk-cells, the basal cells some-
what smaller and angular, forming part of the wall around the antheridial
cells, the wall cells somewhat smaller than the basal cells, but large and
distinct, the terminal cells apparently four in number, later forming stout
finser-like upgrowths of unequal length. Spores about 35 X 3 /x. Peri-
thecia including basal cells 100-120 X 30-35 /x, the stalk-cell 35-40 /i;
Antheridia including stalk-cells (not the terminal projections) 70-80
X 30-35 jU. Sterile part of receptacle about 70 fx.

On Homalota putrescens Woll., British Museum, No. 412, Azores.
On inferior surface of abdomen.

THAXTER. — NEW LABOULBENIACEJE. 413

Monoicomyces Brittanieus nov. sp.

Nearly or quite hyaline. General structure as in 31. Homalotee, the
sterile portion of the receptacle small and inconspicuous, the terminal
cell in the form of a short hyaline simple appendage. Fertile branches
consisting of a single cell bearing the antheridium and perithecium as in
J/. Homalotce. Perithecium rather long-stalked, slightly asymmetrical or
bent, the base inflated, tapering gradually to the blunt undifferentiated
tip. Antheridium borne on a pair of rather long stalk-cells, the basal
cells almost exactly similar to them and distally not enclosing any por-
tion of the antheridial cavity, the wall cells well developed, the cavity
within them relatively small, the terminal cells growing up into gen-
erally stout finger-like processes which may extend above the tip of the
perithecium. Perithecium, including basal cells, 90 X 30 /x. Antheridia
to base of terminal projections 80 X 20 /x.

On Homalota insecta Thom., British Museum, No. 454, Hammer-
smith, England. On superior surface of abdomen.

Monoicomyces St. Helenae nov. sp.

Superficially resembling Compsomyces in general habit. Pale yel-
lowish or straw colored. Receptacle consisting of a triangular basal and
a squarish or roundish subbasal cell which bears a small distal cell with
a short terminal appendage separated from it by a constricted blackish
septum; two to four branches arising from the subbasal cell, normally
fertile and consisting of from two to five, usually three, superposed cells
which are commonly somewhat inflated distally, more so on one side, an
upgrowth from which becomes separated so as to form a prominent small
cell lying close against the base of the axis-cell next above and is simple
or longitudinally divided ; in the former case bearing one, in the latter
a pair of peculiar short appendages from which they are separated by a
conspicuously blackened septum; the basal cell of this appendaofe usually
similarly blackened externally, as is the base of the cell above it, which is
usually characteristically geniculate, its terminal portion erect suffused
with brown or hyaline, and either terminating the appendage or followed
by two or three hyaline cells. Perithecium symmetrically inflated slightly
distinguished from the basal cells, the tip small straight truncate tapering
but slightly and abruptly distinguished. The antheridium relatively
small, its stalk and basal cells about equal, the antheridium proper about
as large as the basal part and hardly broader, its terminal cells develop-
ing as in the other species to long flexuous hyaline upgrowths. Spores

414 PROCEEDINGS OF THE AMERICAN ACADEMY.

38 X 3.0 fi. Perithecia 100-120 X 40-48^. The stalk 34-50 X 25-27
fi. Appendages of fertile branch 50-90 X 5 /x. Sterile part of recep-
tacle 50-50 /x. Greatest length to tip of perithecium 250-435 /x.

On Oxytelus alutaceifrons WolL, British Museum, No. 411, Island of
St. Helena. On abdomen and elytra.

Monoicomyces invisibilis nov. sp.

Hyaline. Basal cell of receptacle small subtriangular, the subbasal
cell rather long and narrow, bearing terminally a distally rounded cell
from which it is separated by an oblique septum and which is surmounted
by a short simple cylindrical appendage ; the fertile branch developed on
one side only, not distinguished from the receptacle and its appendage,
consisting of two or three obliquely superposed cells extending obliquely
upward in a divergent series, the terminal cell bearing a perithecium and
antheridium in the usual relative positions, the subterminal cell sometimes
apparently producing a second antheridium instead of the simple appen-
dage which terminates the lower cell of the series. Perithecium borne
on a rather short stout stalk-cell, its inflated basal half not distinguished
from the flattened basal cells, its slender distal half abruptly distinguished.
Antheridium apparently similar in general to that of the other species,
its detailed structure not recognizable in the types. Perithecia 84 X 30
iu,. The stalk-cell 20 XIO^. Receptacle, sterile part, about 40 /x.
Total length to tip of perithecium 110— 140 /x.

On Homalota putrescens Woll., British Museum, No. 412, Azores.

POLYASCOMYCES nov. genus.

Receptacle consisting of two superposed cells, the upper bearing a
perithecium laterally and an appendage terminally. Appendage consist-
ing of a series of superposed flattened cells, surmounted by a dome
shaped portion which is not persistent (a compound antheridium ?). Peri-
thecium with a distinct stalk-cell and well developed basal cells, the
supporting cell and the lower wall cells forming a broad base the upper
surface of which forms a broad ascigerous area, the asci arising from
great numbers of ascigerous cells.

It has not been possible from the material available to determine the
exact nature of the antheridium in this remarkable genus. The terminal
dome shaped portion of the appendage appears to consist originally of
several cells, but whether it constitutes the whole of the antheridium or
whether the latter is represented in part or wholly by the curious cells

THAXTER. — NEW LABOULBENIACE^E. 415

below it, was not shown by the material. The multiplication of ascigerous
cells of which there are not less and probably more than thirty-six, dis-
tinguishes it from all other known genera.

Polyascomyces Trichophyae nov. sp.

Pale dirtv brownish. Perithecium broadest in the asciiierous zone,
tapering thence to the distinctly differentiated neck-like tip, the subter-
minal wall cells enlarged distally, externally and laterally, the resultant
rounded protuberances forming a ridge about the tip just above its
middle, the distal portion, formed by the lip-cells, of which that on the
right is slightly longer than the rest, broad blunt brownish, the other
three somewhat shorter terminatinsr in narrow blunt extremities which
lie on three sides of the first. The stalk-cell similar to and lying beside
the subbasal cell of the receptacle to which it is united throughout, its
base being in contact with the distal end of the basal cell, while from its
distal end the large basal cells of the receptacle curve abruptly outward
and upward. The appendage consisting of from three to six flat super-
posed darker brown cells, constricted at the septa, looking as if they had
been made irregular by crushiug, the terminal portion (antheridium ?)
blunt, slightly longer than broad, with evidences of lateral apertures.
Spores 28 X 2.5 /x. Perithecia 175 X 50-65 ^. Basal cells 40-48 X 30
fi. Stalk 38 fx. Receptacle 70 /x. Appendage 48-60 X 20-24 /x.

On Trichophya piUcornis Gyll., British Museum, No. 453, Farnham,
England. On superior surface of abdomen.

Cantharomyces Platystethi nov. sp.

Yellowish with a brownish tinge. Receptacle consisting of a small
basal cell and a subbasal cell more than twice as large, bearing the
perithecium and appendage. Perithecium borne on a rather long stalk-
cell, the basal cells continuous with its main body which is inflated below,
conical above, the narrow apex truncate or bluntly rounded. The appen-
dage large, its subbasal cell nearly twice as long as the basal, bearing the
very small antheridium which forms a short cellular margin below its
upper inner angle and apparently consists of not more than ten cells ;
the subbasal cell terminated by an irregular series of small cells which
appear to produce a tuft of branches distally, and from which it may
sometimes be separated by a third cell similar to it. Perithecia 80-86
X ^o IX. The stalk-cell 55 X 20 /x. Receptacle 50-70 X 28 /x. Appen-
dages 140-170 /x.

416 PROCEEDINGS OF THE AMERICAN ACADEMY.

Oa abdomen of Plati/stethus cornutus Grav., British Museum, No. 449,
Kilburn, England.

Eucantharomyces Diaphori nov. sp.

Pale straw colored. Perithecium rather short and stout, its basal cells
small, slightly and usually symmetrically inflated; tapering from about
the middle to the broad blunt slightly asymmetrical tip, which is sub-
tended below the free lips, on the inner side, by the flat trichophoric
cell, just above which arises a very slender recurved rigid appendage
(not cellular) about 8-9 jx long. Receptacle rather short and stout, the
subbasal (anterior) cell somewhat stouter than the basal, the two together
somewhat larger than the short stalk-cell of the perithecium. Apj^en-
dage relatively large, its basal cell short subtriangular, the upper and
lower septa oblique, the subbasal cell slightly longer than broad, its upper
two thirds bordered by the marginal cell which terminates in a slender
stiff straight spine-like process about 11-12 /x long and slightly diver-
gent ; the antheridial cells in five rows of five, four, three, three, and two
cells respectively ; a single additional cell sometimes jDcrsistiug above the
antheridial cavity ; the discharge tube bent outward and slightly upward,
the tip bluntly conical with a slight basal enlargement. Spores 40 X
3.5 /x. Perithecia 120 X 30 /x. The stalk-cell 30 X 18 /x. Appendage
70 ,u long, the antheridium 28 X 21 /x. Receptacle 45 X 24 /x. Total
length to tip of perithecium 180 /x.

On Diophorus tenuieornis Chaud., British Museum (Biologia Coll.),
No. 714. Oaxaca, Mexico. On mid-elytron.

Eucantharomyces spinosus nov. sp.

Perithecium straw colored, rather stout, inflated, tapering to the broad
asymmetrical tip which is slightly sulcate ; the outer lips often larger
than the inner, the latter bordered or subtended by the more deeply
colored trichophoric cell which, in mature specimens, is not very con-
spicuous ; the stalk-cell rather short. Receptacle short, the cells nearly
equal. Appendage much as in E. Diaphori, more slender, the marginal
cell extending nearly to the base of the subbasal cell, distinctly enlarged
below a terminal spine-like process, which is usually nearly erect; the
antheridial cells in three rows of five, tliree (relatively large) and one to
two cells respectively, the discharge tube large and broad, nearly trun-
cate, bent abruptly upward from the base. Spores 35-40 X 3.5-4 /x.
Perithecia 138 X 41 /x. Appendage 70 X 14 /x, the antheridium 35 X

THAXTER. NEW LABOULBENIACE^. 417

IG/x. Receptacle 50 X 20 /x. Total length to tip of peritheciura 190-
207 /x.

Oa Drypta sp., Paris Museum, No. 80, Java. On elytron.

Eucantharomyces Euprocti nov. sp.

Perithecium straw colored to pale amber brown, slightly asymmetrical,
somewhat inflated below ; the upper half tapering gradually to the blunt
asymmetrical apex which is subtended on the inner side by the rounded
flat darker amber brownish trichophoric cell which may extend slightly
beyond the lip-edges, simulating a lip-cell ; the outline of the mature
perithecium becoming more or less corrugated through the appearance of
three to five rather broadly rounded successive elevations, corresponding
to the distal and basal septa of the two lower tiers of wall cells and to
a median protrusion of these cells where five are present; stalk-cell be-
coming slender, mostly slightly shorter than the receptacle ; the basal
cells small. Receptacle relatively rather long, the cells nearly equal.
Appendage generally longer than the receptacle, its basal and subbasal
cells nearly equal ; antheridial cells in three rows of five, three, and two
cells each, the marginal cell bluntly rounded above and extending nearly
to the base of the subbasal cell ; the discharge-tube large, bent outward
or obliquely upward. Spores 50 X 4.5 /x. Perithecia 160-170 X 48 yu,,
stalk-cell 70 X 15^. Receptacle 85-90 X 25-30 ju,. Appendage 110 /x,
antheridium 41 X 22 /x, the discharge-tube 30 /x. Total length to tip of
perithecium 310 /x.

On Euproctus quadrinus Bates, British Museum (Biologia Coll.),
No. 731. Volcan de Chiriqui, Panama.

Eucantharomyces Casnoniae nov. sp.

Perithecium relatively large, rather long, often slender, inflated below,
tapering to the relatively narrow blunt apex ; its outline corrugated
through the presence of from seven to eleven elevations varying in
prominence, the trichophoric cell simulating a lip-cell, the basal cells
elongated and as long or nearly as long as the rather stout stalk-cell.
Receptacle relatively small. The basal and subbasal cells of the appen-
dage relatively small and stout, nearly equal, or the latter somewhat
smaller, the marginal cell bordering its upper half and distally promi-
nent, partly free and slightly inflated, ending in a short spine-like tip ;
the antheridium consisting of three rows of nine, seven and five cells
respectively, the discharge-tube relatively short and stout, bent upward
VOL. XXXV. — 27

418 PROCEEDINGS OP THE AMERICAN ACADEMY.

and over the prominent tip of the marginal cell. Spores 45 X 3.5 /x,.
Perithecia 240-260 X 45-62 ya, stalk-cell 75-80 X 20-28 /x, basal cells
75-100 X 25-30//,. Receptacle 55-65 X 27//,. Appendage 85-103 yu,
the antheridium 50-GO X 24-28 p.. Total length to tip of peritheciura
375-450 /x.

On Gasnonia suhdistincta Chaud., British Museum (Biologia Coll.),
No. 704. Cordova, Mexico.

Eucantharomyces Callidae nov. sp.

Perithecium rather narrow, slightly inflated, tapering from about the
middle to the blunt tip which is rather abruptly distinguished externally,
sometimes bent outward, its distal margin outwardly oblique; the rela-
tively small trichophoric cell simulating a lip-cell, projecting slightly be-
yond the latter externally, but not abruptly distinguished on its inner
side ; the outline of the perithecium becoming inconspicuously corrugated
through the presence of sometimes as many as eleven successive eleva-
tions ; the basal cells elongated, the base of one of the outer external to
the stalk-cell from which it is separated by an oblique septum longer
than the width of the stalk-cell, which is narrower below and about equal
to the basal cells in length or somewhat shorter. Receptacle symmetri-
cally sulcate distally, rather long, the two cells nearly equal. Appen-
dage rather long, its basal cell extending downward and lying external
to the upper half of the subbasal cell of the receptacle ; the subbasal cell
more than twice as long as broad, the marginal cell reaching to its base
and distally prominent. Antheridium relatively small, the antheridial
cells in three rows of five, four, and three cells respectively, the discharge-
tube rather short and stout, erect or bent but slightly. Spores 40 X 4/x.
Perithecia 230-265 X 50 /x, basal cells 120//,, stalk-cell 103 /,i. Recep-
tacle 100-1 20 /t. Appendage 120-125//, antheridium 25 X 38 /;i. Total
length to tip of perithecium average 325 /x.

On Callida sp., Paris Museum, No. 68. Venezuela.

Eucantharomyces Africanus nov. sp.

Very similar to E. CnlUdfB. Amber brown. Perithecium large sub-
fusiform, the margins generally indistinctly corrugated, sometimes marked
by fine transverse striations which may be wholly absent, the tip rela-
tively small and rather abruptly distinguished, the trichophoric cell well
defined, projecting beyond the lip-cells so that the apex usually appears
oblique asymmetrical and slightly sulcate ; the basal cells somewhat

THAXTER. — NEW LABOULBENIACE^. 419

shorter than the stalk-cell, their lower septa nearly equal and symmetri-
cal, both slightly oblique. Receptacle of medium size, the basal cell dis-
tally enlarged. Appendage short, the two basal cells rather small and
nearly equal in length, consisting of three rows of six, four, and four cells
respectively, the discharge-tube slightly curved, abruptly nearly erect,
distally somewhat narrower and conical when young, the marginal cell
extending nearly to tlie base of the subbasal cell. Peritheeia 275-325
X 45-50 /A, the stalk-cell 100-130 /i, the basal cells 75-100 yu.. Recep-
tacle 100 X 26 fx. Appendage 100 /a. Antheridiura 45 X 21 /x, the
discharge tube 21 />i. Total length to tip of perithecium 575-600 /a.

On Callida Natalensis Hope, Hope Coll. No. 274, Natal, Africa. On
CaUida sp., Brit. Museum, No. 550, Angola, Africa. On elytra.

Eucantharomyces Catascopi nov. sp.

Straw colored becoming pale amber brown. Perithecium elongate
tapering but slightly toward the tip or becoming distally swollen through
the pressure of the spore mass, the margins corrugated through the pres-
ence of sometimes as many as seventeen or even more prominences,
which are mostly well defined, especially the distal one of the series,
above which the slightly bent tip is abruptly distinguished, its distal mar-
gin straight oblique, the lip-cells extending just beyond the small darker
trichophoric cell ; basal-cells very much elongated and often corrugated
through the presence of six or more elevation^ corresponding to those
of the perithecial wall-cells; the stalk-cell rather stout much shorter than
the basal cells, from which it is separated by an outer very oblique and
an inner short nearly horizontal septum. Receptacle relatively small,
the basal cell longer than the subbasal cell, distally enlarged so that it
almost coincides with the base of the stalk-cell, which is thus hardly in
contact with the subbasal cell from which it was originall}^ derived. The
basal cell of the appendage somewhat smaller than the subbasal cell,
the marginal cell bulging outward slightly distally and extending almost
to the base of the subbasal cell. Antheridial cells in five rows of eight,
seven, six, five, and four cells or the four inner rows somewhat variable.
Spores 50 X 4.5 p.. Peritheeia 400-475 X 60-70 /x, the stalk-cell 140-
200 X 35-40 IX, the basal cells 200-240 ix. Receptacle 100-110 X 38 jx.
Appendage 120 /x ; antheridium 60 x 32 /x. Total length 680-950 /x.

On Catascopns sp., Paris Museum, No. 117. lies des Moluques.
On the margin of the right elytron.

420 PROCEEDINGS OF THE AMERICAN ACADEMY.

Dichomyces Javanus nov. sp.

Perithecium as long or longer than the receptacle, clear dark reddish
brown, translucent, straight or slightly curved, rather slender, of about
the same diameter throughout, the tip usually abruptly distinguished,
and more or less conspicuously bent to one side, tapering but little to the
rather broad blunt undifferentiated apex. Receptacle rather narrow, the
basal cell dark red-brown below, nearly hyaline above; the central cell
of the lower tier dark red or red-brown, lighter or hyaline at the base ;
the cells on either side symmetrical blackish brown opaque, extending
upward so as to partly enclose the base of the second tier, the margins of
the two tiers coincident: the second tier composed of from seven to nine
cells, hyaline or becoming suffused below with reddish brown, bearing a
well defined sharply pointed purplish slightly asymmetrical antheridium,
on either side, which are subtended by from one to two typical rather
short appendages : the upper tier very similar to the middle or slightly
larger, nearly hyaline, the single perithecium rising to the right of the
median appendage, the right half of the tier thus somewhat larger and
higher than the left, three typical appendages usually present on either
side. Perithecia 145 X 26 /x. Spores about 36 X 4 ^m. Receptacle
120-140 X 50 /A. Total length to tip of perithecium 250-275 /x.

On Plulonthus sp. British Museum, No. 375. On abdomen.

Dichomyces exilis nov. sp.

Basal cell hyaline. Median cell of lower tier deeply suffused with
brown but not opaque ; marginal cells wholly opaque or translucent on
the inner margins, extending upward so as to enclose the base of the
second tier; second tier consisting typically of thirteen cells, colorless or
partly suffused with brownish, the antheridia large brownish straight or
slightly curved, the venter inflated, the cells external to them appendicu-
late, the outer three free above the marginal prolongation of the lower
tier and forming a short blunt projection on either side : upper tier like
the middle one mostly somewhat longer and narrower, consisting of from
thirteen to fifteen cells, the sub-median ones nearly triangular and for the
most part distally overlapped by the external cells next in order and the
basal cells of the perithecia. Perithecia typically two, pale brownish
amber, long and narrow, slightly if at all inflated, tapering gradually to
the undifferentiated broad nearly truncate apex. Spores 35 X 4 ju,.

THAXTEE, — NEW LABOULBENIACE^. 421

Perithecia 130-140 X 22 /x. Receptacle 130-140 X 22 jm. Total length
to tip of perithecium 250-275 /x.

On Philonthus xanthomerus Kraatz., British Museum (Biologia Coll.),
No. 751, San Andres, Vera Cruz. On antennae and anal appendages.

Dichomyces Angolensis nov. sp.

Basal cell hyaline. Lower tier opaque or the middle cell subhyaline,
the marginal cells opaque, extending up on either side of the middle tier :
middle tier relatively large, consisting of about thirteen to sixteen hyaline
cells, the three to four external ones continuing the margin of the first
tier directly and either subhyaline or blackened below, each bearing a
normal appendage; the antheridia of medium size, brownish: upper tier
smaller, shorter, and narrower than the middle one, consisting of from
thirteen to fifteen hyaline cells : distally slightly concave, bearing a pair
of perithecia, the appendages small, hyaline. Perithecium large and
stout straight, faintly brownish, slightly inflated, tapering distally to the
nearly symmetrical truncate apex. Perithecia 120-135 X 30 /x. Length
to the tip of perithecium 250 /a. Greatest length of receptacle 140
X 75 fx.

On Philonthus sp. indet. British Museum, No. 379. Angola, Afiica.
On elytra.

Dichomyces insignis nov. sp.

Basal cell sufi'used with reddish brown or partly hyaline; the lower
tier wholly opaque or translucent along the median line; the middle tier
consisting of about thirteen to seventeen cells, exclusive of those which
are indistinguishable in the slender fork-like prolongations which extend
on either side higher than the middle of the upper tier, the margin broadly
blackened, continuous with the opaque margin of the lower tier; the
lower portion of the three to five median cells marked by a few large
scattered transversely elongated brown patches which merg^e on either
side into the opacity of the marginal cells; antheridia very large, the
venter slightly inflated, the neck sharply pointed, conical, brown, often
abruptly contrasting, three to five of the cells immediately external to
them bearing normal brownish appendages : distal tier very large sub-
triangular, distally concave, consisting of from twenty-nine to thirty-nine
narrow and elongated cells and bearing from four to eight perithecia
with some irregularity ; the appendages brownish, paired above the sub-
tending cell, not as long as the perithecia. Perithecia relatively small.

422 PROCEEDINGS OF THE AMERICAN ACADEMY.

purplish brown, tapering almost continuously from die broad base; the
tip moderately well distinguished, the posterior lips prolonged to form
long nearly straight and horizontal slightly inflated appendages which
project from the tip on either side ; the anterior lips forming the truncate
apex, which consists of two distinct lateral projections with an interven-
ing convex portion having a median apiculus. Perithecia about 85-30
fjL, the appendages from tip to tip 35-39 yu. Antheridia 50 X 11 ^.
Receptacle 300-340 X 200-230 ^. Total length to tip of perithecium
375-400 n.

On an undetermined staphylinid collected by A. R. Wallace at
Sarawak, Borneo, Hope Coll. No. 218.

Dichomyces biformis nov. sp.

Basal cell hyaline or nearly so, usually somewhat enlarged and often
with a heel-like anterior projection; lower tier rather narrow, quite
opaque, the marginal cells extending up to the subterminal marginal cell
of the middle tier or to the cell next below it : the middle tier short and
stout, the nine to eleven cells hyaline or faintly reddish brown above,
usually becoming more or less suffused below and externally with brown ;
the median cells, where suffused, marked by darker transverse flecks on
their anterior face, the marginal cells ending in a blunt distal often
hyaline prominence on either side; antheridia short and stout subconical,
subtended by a single brown inconspicuous appendage : the distal tier
assuming in well developed individuals the form of a rather slender
crescent, the number of cells very variable, the maximum about fifty,
sometimes less than half this number, in which case the form is stouter,
the marginal cells rarely extending above the tips of the perithecia which
are four to eight in number and of two kinds which are not known to be
associated on the same individual ; in the one case they are stouter,
purplish brown, the basal third or more often abruptly hyaline or nearly
so, the much darker red brown tip tapering rather abruptly to the apex,
which is hyaline nearly truncate, with a well defined median blunt pro-
jection ; the 2)Osterior lip-cells prolonged much as in D. insignis to form
a long horizontal nearly cylindrical or slightly tapering bluntly tipped
hyaline appendage on either side ; the second type more often longer and
more slender than the first, pale reddish brown, the tip tapering, slightly
truncate or blunt, often with a blunt median projection as in the first ij\iQ,
but without appendages. Perithecia 105-110 X 20-35 /x,. Receptacle
200 X 100-300 X 270 ^.

On Philonthus sp., Niagara Falls, New York, Mr. Charles BuUard : on

THAXTER. — NEW LABOULBENIACE^. 423

Philonthus umbratilis Grav., British Museum, No. 3G2, Leicester, Eng-
land; Paris Museum No. 206 and British Museum, No. 407, Madeira;
Paris Museum, No. 175, St. Pierre et Miquelon.

Dicliomyces hybridus, nov. sp.

Basal cell small hyaline with a red brown suffusion near the base:
lower tier narrow and elongate, opaque or sometimes with a median trans-
lucent line : middle tier rather narrow, not more than five of the median
cells distinguishable, and more or less conspicuously marked on the ante-
rior side by dark transverse flecks or stride ; distally hyaline or merely
tinged with reddish brown above, becoming red-brown and finally opaque
below ; the margins opaque, continuous with those of the first tier and
extending upward to form fork-like opaque projections, as in D.furciferus,
which equal or exceed the upper tier in length ; a single appendage aris-
ing posterior to the rather small purplish antheridium : upper tier rela-
tively large, distally concave, composed of from fifteen to thirty-three
nearly hyaline cells with reddish brown shades along the septa, the median
cells sometimes flecked with reddish brown spots or transverse striae
towards the base, bearing two to six perithecia which may be of two types
associated on the same indi^'idual or occurring on different individuals :
the one type somewhat smaller, straighter and more erect, reddish brown,
the lower half often abruptly paler or nearly hyaline, tapering rather
abruptly to the tip, the lips of which are modified much as in P. farci-
feriis ; the other type larger, rather characteristically divergent, tapering
rather abruptly to the truncate unmodified apex ; appendages hyaline,
sometimes as long as or even longer than the perithecia. Spores 35 X
4;u. Perithecia 100-115 X 25-30 /x. Total length to tip of perithecium
250-300 /x. Receptacle 175-250 X 85-145 fi.

"With both types of perithecia : on Philonthus aeneipennis Boh., Paris
Museum, No. 203, Gulf of Oman, India ; on Philonthus sp., British Mu-
seum, No. 36G, Sylhet, Assam, India; on Philonthus sp., British Museum,
No. 368, Hong Kong, China.

With only one form of perithecium (not appendiculate) : on Philonthus
ventralis Grav., British Museum, Ealing, England ; Paris, No. 207,
Funchal, Madeira; British Museum, No. 426, Europe : on Philonthus sp.
British Museum, No. 495, Balthazar, Grenada, West Indies; British
Museum, No. 369, China ; on P. proximus Woll., British Museum, No.
403, Canaries ; on P. gemellus Kr., British Museum, No. 367, Ceylon ;
on Philonthus sp., Niagara Falls, N. Y. (C. Bullard).

424 PROCEEDINGS OF THE AMERICAN ACADEMY.

Dichomyces Madagascarensis nov. sp.

Basal cell deeply suffused with brown. Lower tier very long and
slender, opaque except for a faint median translucent line : middle tier
with three to five of the median cells distinguishable, red-brown ; the rest
indistinguishable in the opaque margins which extend upward to form
long fork-like outgrowths on either side that may reach nearly to the tips
of the perithecia ; antheridia not large, brownish : upper tier consisting
of about twenty-one to twenty-three cells, tinged with reddish brown,
relatively large, deeply concave distally ; the median cells like those of
the middle tier, marked by fine faint transverse stria3, bearing normally
two perithecia which are long and slender, often slightly curved and
divergent, pale reddish brown, the tip narrow, the posterior lip-cells form-
ing two small, slightly divergent projections (like those of D. farciferus
but relatively smaller) curved at the tips, the anterior lips meeting in a
point between them. Appendages hyaline, sometimes equalling the peri-
thecia in length. Spores very slender and abundant, 35 X 2 /a. Peri-
thecia 125-135 X 25 /;i. Total length to tip of perithecium 320-350 /t-
Receptacle 225-240 X 105 /x.

On Philonthus Sikorae Fauv., Paris Museum, No. 179, Tananarive,
Madagascar. On abdomen.

'O"

Dichomyces vulgatus nov. sp.

Receptacle short and stout, the basal cell small squarish hyaline ; the
lower tier externally opaque, except the whole or the middle of the median
cell or only its upper end, the opaque margin divergent extending above
the base of the second tier, the blackened margin of which is continuous
with that of the first tier ; sometimes, like it, divergent, more often
abruptly less divergent or even erect, extending upward to form on either
side free fork-like, usually opaque, sometimes hyaline projections as in Z>.
farciferus which may extend to a point somewhat above the base of the
perithecia or may be almost obsolete; the three middle cells of the middle
tier usually more or less conspicuously punctate below, with transversely
elongated blackish brown spots : the antheridia normally placed, unusu-
ally long and large, pointed, with two or three short, inconspicuous normal
appendages placed one behind, the rest external to it. The upper tier
distally concave, consisting of from fifteen to twenty-one cells, pi'oducing
normally four perithecia associated as usual with short stout typical
appendages. Perithecia usually erect, straight, rather stout, pale reddish

THAXTER. NEW LABOULBENIACE.E. 425

amber brown, the lower half or third often abruptly lighter, tapering to a
blunt tip which bears on either side a short, stout, often slightly recurved
ear-like outgrowth as in D. furciferus formed by the prolongation of the
anterior lip-cells, the posterior lips forming a usually angular, sometimes
sharply {)ointed projection between them. Antheridia purplish, nearly
straight or slightly curved, rather abruptly enlarged below the sharply
pointed apex, the venter somewhat inflated. Perithecia 80-1 00 X 25 /x.
Antheridia 35 X 7 /x. Total length to tip of perithecia 200-225 X 100
-115 /x. Appendages 35 /A.

On Philontfms jinvolimhatus Erichs., Panama, British Museum, No.
750 (Biologia Coll.) ; P. parvimanus Sharp, Chontales, Nicaragua, British
Museum, No. 74G (Biologia Coll.) ; Philonthus sp., Mt. Gay, Est Grenada,
"West Indies, British Museum. 489 (Smith Coll.) ; P. sabalarius Nord.,
British Museum, No. 406, Madeira ; P. longicornis Staph., British
Museum, No. 408, Island of St. Helena ; P. cruentatus Gmel., British
Museum, No. 358, Europe ; P. varians Peck, British Museum, No. 359,
Ealing, England ; P. dimidiatus Er., British Museum, No. 761, Not-
ting Hill, England. On abdomen. A form, apparently this species, also
from Hong Kong, on Philonthus sp., British Museum, No. 396.

Dichomyces Cafianus nov. sp.

Tinged with dull amber brownish throughout, the perithecia darker.
Basal cell nearly hyaline, the lower tier as in D. vidgatus, the opacity
involving in general but half of the upper (external) cells, the septa of
which are visible on the inner side, the median cell dark brown, its lower
half or more opaque : the middle tier consisting of typically thirteen cells,
the margins unmodified and ending in a short external rounded projection,
which does not extend beyond the base of the upper tier; the rather in-
conspicuous antheridium normally placed, concolorous, a single short
appendage close behind it : the upper tier consisting of from nineteen
to twenty-three, usually twenty-one, cells, forming an inverted crescent,
the short, stout, bladder-like appendages arranged as in D. vidgatus.
Perithecia normally two in number, somewhat inflated externally, nearly
straight, slightly asymmetrical, rather stout, tapering to the bluntly
pointed undifferentiated tip. Spores 45 X 4.5 /x. Perithecia 120-140
X 35-40/;.. Receptacle 200-250 X 100-140 /x. Total length to tip of
perithecium 310-350 /u. Appendages about 20 X 6/x.

On Cafius puncticeps White, British JMuseum, No. 381. Colenso (S.
Africa?).

426 PROCEEDINGS OF THE AMERICAN ACADEMY,

Dichomyces dubius nov. sp.

Receptacle much as in D. princeps and similarly colored, smaller,
shorter, and stouter, antheridia large purplish ; the distal tier of cells
producing typically two, rarely more, perithecia, which are pale brown-
ish and dimorphous ; usually rather slender, tapering slightly, the pos-
terior lip-cells producing ear-like outgrowths recurved or bent forward
as in D. vulgatus : more rarely larger and stouter, the blunt, often asym-
metrical tip witliout appendages ; the two forms sometimes, but not
usually, associated on the same individual : external appendages normally
large, long, colorless, reaching to the middle of the perithecium or even
to its tip. Individuals asymmetrical, with a single antheridium and peri-
thecia of the second type, are not infrequently met with on the legs of the
host. Perithecia 70-00 X 20 /x, those without appendages 70-105 X 30
-35 /x. Spores 35 X 4^. Receptacle about 120 X 75|ii. Total length
to tip of perithecium average 190 /x.

On Philonthus sp., Niagara Falls, New York. On all parts of host.
On hosts received from Mr. Charles Bullard. Possibly a variety of D.
princeps^ to which it is very closely allied. None of the abundant mate-
rial of the latter from different parts of the world, however, show any
tendency to produce an auricled type of perithecium.

Dichomyces Peruvianus nov. sp.

Receptacle with faint brownish shades especially along the septa,
almost in the form of two superposed isosceles triangles, the lower very
regular, including the basal cell and the first and second tiers, its distal
margin horizontal, the upper truncate at the base and distally concave.
The basal cell short, the lower tier consisting of from three to four cells,
nearly equal in length; the middle tier of typically thirteen cells, the
antheridia of medium size, the outer five cells distally appendiculate, one
of the appendages situated behind the antheridium as usual; the distal
series consisting of usually twenty-seven cells bearing typically four peri-
thecia, the appendages placed as usual, colorless, somewhat shorter than
the perithecia, which are mostly brownish externally and hyaline on the
inner side, the brown or reildish fawn color sometimes predominating,
asymmetrical, somewhat inflated, slightly bent inward near the tip which
is small, pointed, and well distinguished. Perithecia about 120 X 30/.*..
Receptacle 207-240 X 140-175 /x. Appendages 185 /x (longest). Total
length to tip of perithecium 300-350 ,«.

On Brachyderus simplex Sharp. In Dr. Sharp's Collection, Peru.
On elytra and abdomen.

THAXTER. — NEW LABOULBENIACE^. 427

Peyritschiella Amazonica nov. sp.

Perithecium trausluceut browu, about as long as the receptacle, sub-
clavate large, contracted below to form a neck-like base, somewhat in-
flated distally, the tip well though not abruptly distinguished, tapering to
the nearly truncate apex formed by the slightly expanded tips of the lip-
cells which are otherwise unmodified. Receptacle rather narrow, pale
translucent brown, consisting of a single basal cell followed by three
tiers of cells ; the lower symmetrical or nearly so consisting of three
long narrow nearly equal cells not appendiculate and not projecting
laterally : the middle tier asymmetrical, consisting of about twelve cells,
the series projecting distally on either side, all the cells except the three
larger median and the external ones producing distally short typical ap-
pendages, the third cell on the right from the median cell bearing a
prominent erect antheridium : the terminal tier very similar to the middle
one, consisting of about the same number of cells which produce short
typical appendages distally and (in the types) a single nearly median
perithecium. Perithecia 200-210 X 36 [x. Receptacle 225 X 70 ^.
Antheridium 45 ^a long. Total length to tip of perithecium about
400 IX.

On an undetermined staphylinid. British Museum, No. 400. Nanta,
Amazon River.

Peyritschiella protea nov. sp.

Perithecia translucent, brownish amber colored, rather stout and sym-
metrically inflated, the symmetrical tip tapering rather abruptly, the apex
rather narrow truncate, the lip-edges unmodified. Receptacle nearly or
quite hyaline, consisting of a single basal cell, above which the three
typical tiers of cells are very variously developed : the lowest of these may
rarely consist of a single cell, often of three which do not project laterally,
or in well-developed specimens of as many as twelve or more cells, those
external to the middle three forming on either side distal external angular
usually asymmetrical ^projections, one or both of which may bear termi-
nally one or even two perithecia and typical appendages : the middle series
like the lower when the latter is well developed, subtriangular in form,
consisting of sometimes as many as fifteen to eighteen cells, generally
somewhat asymmetrical ; a single perithecium usually arising distally from
the projecting portion on either side, together with numerous typical
appendages : the distal tier similar to the middle one, mostly smaller,
somewhat asymmetrical, bearing usually a single perithecium above the

428 PROCEEDINGS OF THE AMERICAN ACADEMY.

median cell, though not produced from it, the remaining cells bearing
typical appendages, often as long or longer than the peritheciura, the small
subtending cell being unusually well defined. Subject to great variation,
and sometimes producing more than one antheridiuin. Peritheciu 80-96
X 32 i^. Receptacle 270 X 80-100 to 120 X 45 fx. Total length to tip
of perithecium 200-350 fx.

On Bledius bicornis Germ., British Museum, No. 392, Europe (Thu-
ringia), No. 432, Europe; on Oxyfelus rugosus Fabr., British Museum,
450, Hampstead, England; on Acrognatlius mandibidaris Gyll, Brit-
ish Museum, No. 434, Europe. On legs, elytra, and prothorax. In small
specimens the two lower tiers may be but slightly developed, bearing
neither appendages nor perithecia, the middle producing one antheridium,
the number of cells and appendages on one side of the perithecium being
as in all cases greater than on the other,

LIMNAIOMYCES nov. genus.

Receptacle consisting of two ^lortions, a basal part below the perithe-
cium and a distal part united to its posterior margin ; the basal portion
consisting of a single basal cell, surmounted by two tiers of cells (some-
what as in Peyritschiella), the anterior cell of the upper tier giving rise
to a compound antheridium in structure similar to that of Peyritschiella :
the distal (marginal) portion consisting of an inner and an outer elon-
gated cell, the inner terminating in one of the bell-shaped appendiculate
cells characteristic of Chitonomyces, separated from the simple appendage
by a broad, constricted, blackened septum ; the outer by successive sub-
terminal external proliferations forming a series of cells from which a
smaller secondaiy appendiculate cell is separated above, the whole cor-
responding in development to the external portions of the tiers of cells
in Dichomyces, the proliferation taking place to the right and left succes-
sively, so that the appendages appear to arise in two rows.

A clearly defined genus apparently intermediate between Peyritschiella
and Chitonomyces.

Limnaiomyces Tropisterni nov. sp.

Perithecium amber brown, straight, erect, with a slight nearly median
inflation or tapering but very slightly to the undifferentiated tip; the
upper half free. Receptacle pale straw colored, distally dull amber
brownish, the foot minute, black ; the basal cell short and small, the
lower tier consisting of two cells which are nearly equal, several times

THAXTER. NEW LABOULBENIACE^. 429

as long as broad : the second tier consisting of three cells, the posterior
one longest, the median longer than the anterior, which terminates in
the antheridium, which is subtended by four basal cells, two of them
outer and lower and separated by oblique partitions, while a smaller
upper one lies on either side : above the antheridium two vertically elon-
gated cells form the clearly defined base of the perithecium ; external
to these cells and somewhat obliquely separated from them lies the broad
base of the inner marginal cell of the distal portion of the receptacle,
which lies next above the middle cell of the upper tier, its cavity nearly
obliterated above as the spores mature, the primary appendiculate cell
which terminates it rather elongate ; the proliferation of the outer mar-
ginal cell beginning quite near its base, forming a series of about eight
cells separated by oblique septa and terminated by small appendiculate
cells ; the appendages very small, vesicular, brownish below. Perithecia
127-175 X 35-37 /x. Receptacle, basal part, 75-110 /a, distal part
75-110 IX. Appendages 6 X 3 /a. Total length to tip of perithecium
240-375 fjL, to tip of receptacle 190-265 ^.

On Tropisternus sp. indet., Paris Museum, No. 47. Mexico. On
tip of abdomen.

Limnaiomyces Hydrocharis nov. sp.

Hyaline. Perithecium rather stout and short, somewhat inflated, its
tip abruptly bent outward, the apex bluntly rounded or nearly truncate ;
the tip and the appendiculate cell usually symmetrically divergent.
Basal portion of the receptacle relatively short and stout but otherwise
similar in structure to that of L. Tropisterni ; the two basal cells of the
perithecium almost obliterated at maturity so that its base appears to
rest immediately on the antheridium ; the distal portion of the receptacle
bordering the perithecium to its tip, the inner cell becoming almost wholly
obliterated in the middle and terminating in a short bell-shaped appen-
diculate cell which is slightly divergent: the outer marginal cell usually
proliferating three times ; of the three cells thus formed the two inner,
as a rule, produce well developed, long, simple, hyaline appendages;
not, however, as well developed as the primary appendage, which may
be twice as long as the perithecium. Spores 50 X 3 ,«. Perithecia
60-80 X 17-20 ,u. Receptacle, basal part, 50 X 20-26 //, distal part
50-62 ^. Appendages, longest, primary 140 ^m, secondary 70 ^u. Total
length to tip of perithecium 100-128 ju.

On Hydrocharis obtusatas Say, Cutts Island, Kittery Point, Maine.
At tip of abdomen.

430 PROCEEDINGS OF THE AMERICAN ACADEMY.

Chitonomyces Floridanus nov. sp.

Pale straw colored with a smoky, brownish tinge, the basal and subbasal
cells relatively large, the former rather elongate, the latter broader than
long, tlie distal cell erect, conical, appendiculate, its basal septum hori-
zontal. Perithecium relatively large, distally somewhat inflated, the
posterior margin to the apex nearly straight, the tip moderately well dis-
liijguished, the inner margin strongly convex between the tip and the
secondary appendage ; the lip-cells each forming a more or less distinct
papilla. Spores relatively large about 35 X 3 ^u. Perithecia 70 X 28 fi.
Receptacle, distal part, 62 ^, the two basal cells with foot 52 /<. Total
length to tip of perithecium 120-138 /a.

On Cnemidotus 12-puiictatus Say, Eustis Florida, October. On legs
and elytra.

Cliitonomyces aethiopicus nov. sp.

Perithecium red-brown, darker on the inner side, with faint transverse
striations, somewhat curved ; the inner lip-cells producing at the left a
blackish brown projection directed obliquely outward across the tip and
resembling a canine tooth, the inner lip-cell on the left producing a much
smaller, blackish, inconspicuous, tooth-like projection ; one of the inner
wall cells abnormally developed, bulging inward against and almost over-
topping the subterminal appendiculate cell, the greater portion of the
margin of this outgrowth nearly horizontal and extending from the apex
of the perithecium to the insertion of the subterminal appendage which
is sunk in an abrupt depression between it and the base of the ter-
minal cell of the receptacle. Receptacle nearly hyaline, strongly curved
throughout, consisting of a long basal and subbasal cell which appear to
lie side by side for nearly their whole length, the lower marginal cell of
the distal portion almost obliterated by the body of the perithecium, the
subterminal cell large triangular, the terminal cell about as large, and
separated from it by a nearly horizontal septum ; wholly free, abruptly
geniculate, the distal portion much narrower, erect and black (the tip
broken), abruptly distinguished above an external bulge of the portion
below it. Perithecia 128 X 40 ^, the tooth-like projection 18 ju. Re-
ceptacle to tip 275 jU, the basal and subbasal cells including the foot
140 X 35 fx. Total length to tip of perithecium 255 (i.

On Orectochilus specularis Aube, Paris Museum, No. 100, Gold Coast,
Africa. On elytra.

THAXTER, — NEW LABOULBENIACE^. 431

Amorphomyces obliqueseptata nov. sp.

Male individual uuknovvn.

Female individual, straw colored tinged with amber brown, the recep-
tacle consisting of a very small basal cell and a short, broad, subbasal cell
bearing the very large perithecium and without appendages. The peri-
tliecium broadly inflated at the base becoming gradually narrow distally,
the tip blunt asymmetrical ; the apex somewhat oblique, the asci and
spores filling the perithecium in great numbers, developed from a single
ascogenic cell. The spores obliquely septate 40 X 7 ^. Perithecium
200 X 55-GO ^. Receptacle without foot 35 X 27 ^u.

On the antennai of an undetermined staphiliuid, British Museum,
No. 398, Ega, Amazon River.

Teratomyces vulgaris nov. sp.

Perithecia one to three in number, usually symmetrical and straight,
becoming clear purplish brown, often considerably inflated below and
conical above ; the tip blunt or sometimes slightly pointed, the basal
cells variously elongated sometimes nearly as long as the perithecium
proper and often longer than the usually well developed stalk-cell.
Receptacle symmetrical, its basal cell nearly hyaline, the cell above it
tinged with reddish brown and somewhat larger, the third cell like the
subbasal, squarish and somewhat larger. Appendages nearly hyaline or
suffused, never deeply, with reddish brown, comparatively few in number,
rather stout and long in general, the curved beak-like terminations of
other species wholly wanting ; all the appendages or their primary
branches distinguished by a blackish brown basal septum, some, often
many of them distinguished by being closely septate above, the cells thus
formed producing a series of lateral outgi'owtbs projecting obliquely up-
ward and superposed (the antheridia?) Perithecia 140-200 X 45-60//,
their basal cells 40-120 ,«, the stalk-cell 35-126 X 25 /^. Receptacle
to base of appendages 70-100 /x. Appendages (longest) 175 /a. Total
length to tip of perithecium 325-450 /x.

On Qiied'ms fidgidns Fabr., British Museum, No. 354, Kiel, Germany ;
on Q. fuUginosus Grav., British Museum, No. 355, Europe; on Q. trim-
cicolas Fair. (= rentmlis Arag.), British Museum, No. 435, Great
Britain; on Q. cruentus Oliv., British Museum, No. 422, Europe ; on
Quedius sp. indet., British Museum, No. 356, Canada; on Q. fidgidus
Fabr., Hope Coll., No. 216, Europe; ou Philonfhus? sp. indet., British
Museum, No. 365, Hungary.

432 PROCEEDINGS OF THE AMERICAN ACADEMY.

Teratomyces Philonthi nov. sp.

Perithecia commonly two, long and slender, a basal middle and distal
portion distinguished, corresponding to the basal, middle and the distal
wall- and lip-cells, the basal portion slightly inflated, purplish, the middle
distinguished from it by a slight elevation at the septa ; the middle nearly
hyaline, rather abruptly narrowed, its margin slightly concave owing to
a slight distal enlargement, which, in mature specimens, distinguishes it
rather abruptly from the much shorter narrower subconical mostly sym-
metrically truncate colorless distal portion; the stalk-cell. rather short,
concealed by the appendages ; the basal cells forming a squarish base.
Receptacle relatively small, symmetrical or asymmetrical, the basal cell
translucent brownish, the subbasal cell very small, flatfish, wholly in-
volved by the deep nearly opaque suffusion of the lower half or more of
the upper cell, which is nearly hyaline above. Appendages short,
slightly exceeding the base of the perithecium, rather rigid, sliglitly
divergent, for the most part dark brown ; forming a rather dense tuft,
many ending in pointed cells, the slender terminations straight or bent
and forming the beak-like cells characteristic of tlie genus. Spores
36 X 4/x. Perithecia 140-175 X 25-30 ^u, the stalk-cell about 35 /x.
Eeceptacle about 85 X 35 /x. Appendages (longest) about 70 /*. Total
length to tip of perithecium 250-300 /x.

On Philonthus sp. indet., British Museum, No. 365, Hungary.

Corethromyces Brazilianus nov. sp,

Perithecium and receptacle much as in C. Cryptobii, but differing dis-
tinctly in the character of its appendage, the inner main branch of which
consists of from four to six cells, the others very short, all nearly opaque,
the branchlets long rigid divergent, curved abruptly outward at the tips.
Total length to tip of perithecium 200-375 /a. Spores 28 X 3 yu,. Peri-
thecia 90-175 X 28-38 /M. Appendages to tip of branchlets 140-200 /x.
Two specimens from Colombia, apparently identical, are much larger;
total length 610 yu, ; branches of appendages 540 /x ; perithecia 450 fx.

On O/'i/ptobium Brazillanum Lee, Paris Museum, No. 173, Brazil;
on C. fdsciatnm Erichs, Paris Museum. No. 197, Caracas, Venezuela;
on G. Ftohri Sharp, British Museum (Biologia Coll.), No. 762, City
of Mexico ; also from same collection on C venustum Sharp, No. 758,
Oaxaca, Mexico ; on G. simili2)enne Say, No. 761, Mexico. The larger
type on Gryptohium sp. indet., British Museum, No. 385, Colombia. On
all parts of host.

THAXTER. — NEW LABOULBENIACEiE. 433

Corethromyces purpurascens nov. sp.

Perithecia dull purple, mostly slender straight or slightly curved,
nearly isodiametric or the outer margin convex ; the base slightly broader,
the junction of the basal and subbasal and of the subbasal and subter-
minal wall cells indicated by a distinct protrusion in well developed indi-
viduals, in which the tip is thus moderately well distinguished although
in most cases, especially in smaller specimens, the margin forms an
unbroken line from base to apex, the perithecium being sometimes dis-
tinctly inflated basally ; the stalk-cell as in G. Gryptobii hyaline above,
becoming opaque brown below. Basal cell of the receptacle purplish or
brownish translucent, the rest opaque indistinguishable from the almost
wholly opaque main body of the appendage, the oblique inner margin of
which is followed by a series of hyaline or purplish cells, three or more
in number which give rise to the erect branches ; the primary branches
sometimes purplish near the base but producing an erect tuft of branches
and branchlets which are quite hyaline, more or less flexuous and taper-
ing. Perithecia 100-150 X 2b jx. Total length to tip of perithecium
175-275 /x. Longest branches of appendages about 140 /x..

On Cryptobiiim capitatum, Paris Museum, No. 172, Brazil; on Cryp-
tobhim sp. indet., British Museum, No. 494, Balthazar, Grenada, West
Indies.

EUCORETHROMYCES nov. genus.

General form as in Rhadinomyces, the receptacle consisting of two
superposed cells, the upper giving rise to the perithecium and appendage.
Perithecium as in Rhadinomyces, stalked. Appendage consisting of
several superposed cells the distal one bearing terminally a series of
branches which produce free flask shaped antheridia laterally, borne on
short lateral branchlets or sessile.

Eucorethromyces Apotomi nov. sp.

Hyaline becoming tinged, especially the perithecium, with pale amber
brown. Receptacle short, the subbasal cell usually smaller, its axis coin-
cident with that of the stalk-cell. Perithecium rather slender, inflated
toward the base, the distal half slender tapering slightly to the blunt
unmodified apex, the basal cells rather small, nearly equal, the stalk-cell
stout and well developed. Appendage divergent almost at ri^ht angles
to the axis of the receptacle, its basal cell usually more than twice as
large as the subbasal cell, which bears distally and anteroposteriorly a
single, or partly double, row of from four to six branches, some of them
VOL. XXXV. — 28

434 PROCEEDINGS OP THE AMERICAN ACADEMY.

often elongate, slender straight or curved, suffused with dark blackish
browD, hyaline along the inner margin at least toward the base, obliquely
septate, the septa dark ; the antheridia stout, flask shaped, subtended by
a dark septum, borne singly and laterally or several together on short
brauchlets near the base. Spores 2Q X 2 ^. Perithecia 100-125 X 25
-28 yu, 35-38 X 14-18(«. Receptacle 40 X 18 /a. Appendage without
branches 50 X 16/x. Total length to tip of perilhecium 190-207//.
The Celebes material somewhat smaller.

On Apotomus xanthotelus Bates, British Museum, No. 578, Celebes ;
on A. rufas Rossi, British Museum, No. 577, Europe. On elytra.

Rhizomyces crispatus no v. sp.

Perithecia brownish, rather stout : when viewed sidewise, the inner
margin strongly convex, the outer nearly straight with a general median
elevation or concave owing to a general outward curvature, tapering to
the undiiferentiated tip, the apex broad truncate, usually symmetrically
bisulcate : viewed at right angles to this position straight symmetrical,
abruptly enlarged below the narrow symmetrical abruptly distinguished
tip : the basal cells well defined nearly isodiametric, the stalk-cell large,
as long or longer than the perithecium. Receptacle two-celled, the foot
typically modified and blackened without rhizoids, distally geniculate
through a protrusion of the distal cell below the insertion of the stalk-
cell and opposite that of the appendage. Appendage erect, sometimes
exceeding the tip of the perithecium, consisting of a single series of super-
posed cells, the three or four lower suffused with smoky brown, the rest
subhyaline, each cell except the basal one giving rise directly and exter-
nally to a branch, the insertion in successive cells being somewhat to the
right and left of the median line so as to form two vertical rows, the
basul cells of alternate branches being superposed ; each branch con-
sisting of a basal cell externally blackened, which gives rise above to a
one-celled short branchlet, bearing usually a pair of long, slender anthe-
ridia, the remainder of the branch curved upward blackish brown except
its upper margin, and giving rise from its lower (external) side to a series
of close-set simple brauchlets, black, recurved, more abruptly at the tips
which are slightly enlarged and nearly hyaline, the whole suggesting the
margin of a curled black feather. Spores 20 X 2.8 fj,. Perithecia 65-
75 X 27-30 /x, the stalk-ceil 50-85 /a. Receptacle 30 /a. Appendages
140-175//.

On Diopsis sp., British Museum, No. 739, Natal, Africa.

THAXTER. — NEW LABOULBENIACE^. 435

Rliachomyces Philonthinus nov. sp.

Perithecia borne on a short broad hardly visible stalk-cell, reddish
brown, inflated toward the base, conical above, straight and nearly sym-
metrical, the tip blunt, undifferentiated symmetrical. Main axis of the
receptacle distinct, consisting of about twenty cells, including about eight
to ten cells which form its erect free termination beside the base of the
perithecium ; the three lower cells mostly suffused with red brown, those
above hyaline or partly suffused, increasing in size to about the eleventh
cell, above which they become successively smaller to the tip of the free
portion ; the septa for the most part marked by rather prominent con-
strictions. Appendages numerous but not obscuring the main axis of
the receptacle, slightly divergent, mostly tapering distally and slightly
bent below the straight hyaline tips ; those arising about the base of the
perithecium longer and stouter, brown and mostly blunt tipped, about
six in number and extending about to the middle of the perithecium.
Spores about 40-45 X 4/x. Perithecia 140-200 X 40-60^. Receptacle
220-340 /x. Total length to tip of perithecium 350-500 yu,. Longest
appendages about 100 yu..

On Fhilonthus longicornis Steph., British Museum, No. 408, Island
of St. Helena; on Fhilonthus sp. indet., Hope Coll., No. 225, British
Isles. On abdomen and elytra.

Rhachomyees velatus nov. sp.

Perithecium short stout straight symmetrical, evenly inflated pale
brownish, translucent ; the tip abruptly dark brown opaque or nearly so,
tapering symmetrically to the blunt rounded apex. Receptacle varying
in length, consisting of perhaps eighteen to twenty cells, the basal cell
and sometimes two or three of those above it hyaline or nearly so, the
rest indistinguishable, being concealed by the densely crowded ajipressed
appendages, which are rather short and slender, deep brown or opaque
except along the inner margin and at the tip ; those around the base of
the perithecium also densely crowded, subequal blunt-tipped, wholly suf-
fused, completely enveloping it and wholly concealing it till it is fully
developed when the tip alone projects beyond them. Spores about 35-
40 X 3-4 ju. Perithecia 175 X 75 ,u or smaller. Total length to tip of
perithecium 400-550 /z. The longer appendages about 120 /x.

On Colpodes agilis Chaud., British Museum (Biologia Coll.), No. 696,
Jalapa, Mexico ; on C. atratus Chaud., British Museum (Biologia Coll.),

436 PROCEEDINGS OF THE AMERICAN ACADEMY.

No. G98, Irazu, Costa Rica; on Gynandropus Mexicanus Putz., British
Museum (Biologia Coll.), No. 682, Cordova, Mexico. Usually on legs.

Rhachomyces Thalpii nov. sp.

Perithecium hyaline or straw colored, becoming faintly tinged with
brown, slender, inflated near the base ; the distal half or less mostly
curved away from the appendages, tapering gradually to the blunt
undifferentiated apex. Receptacle normally consisting of eleven cells
superposed to form the main axis, hyaline, their septa horizontal or but
slightly oblique, the basal cell subtriangular, tinged with reddish brown ;
the cells of the secondary series hyaline and proportionately rather large.
Appendages rather dense, almost opaque except the nearly or qnite
hyaline tip and inner margin ; rather short, about four to six of those
about the base of the perithecium much larger, longer, and stouter, reach-
ing somewhat higher than the middle of the perithecium, their tips at
first clavate becoming obliquely truncate or fan shaped through the
degeneration of the hyaline portion, the curved tips of the antheridia
projecting rather conspicuously. Perithecia 115 X 30 /a. Total length
of receptacle 140/x,. Longer appendages 90^.

On Thalpius riifulus Lee, Amer. Mus. of Nat. History. Texas.

Rhachomyces Zuphii nov. sp.

Perithecium relatively small, straw colored, somewhat inflated at the
base, the tip rather abruptly distinguished and slightly inflated. Axis
of the receptacle slender, consisting normally of about sixteen cells which
are nearly hyaline, or with brown shades below the septa. Appendages
nearly opaque, straight stout appressed, not elongate, more or less
swollen distally along the inner margin of the subhyaline tip ; eight to
ten about the base of the perithecium longer and stouter. Perithecia
110-140 X 25 /x. Longer appendages about 110-140 /x. Total length
to tip of perithecium 350-400 /x.

On Zuphkim Mexicanum Chaud., British Museum (Biologia Coll.),
No. 713. Cordova, Mexico.

Rhachomyces Canariensis nov. sp.

Perithecium pale straw colored, nearly straight, a median and subter-
minal well-defined broadly-rounded ridge marking the transverse septa
between the three lower tiers of wall-cells ; the tip tapering rather
abruptly, hyaline, the apex rounded. Receptacle rather slender, the

THAXTEB. — NEW LABOULBENIACE^. 437

basal and subbasal cells relatively large, hyaline, the rest pale straw
colored, the main axis consisting of fifteen or sixteen cells, the upper
five or six forming a free erect termination. The appendages not
numerous, appressed, brown ; those about the base of the perithecium
larger, distally blunt and hyaline, about two thirds as long as the peri-
thecium. Perithecia 90-130 X 27-30 /x. Receptacle 175-225 yi^. Ap-
pendages, longest, about 100 /x. Total length to tip of perithecium 250—
325 /A.

On Trechus flavomarginatus Woll., British Museum., No. 419. Ten-
erifFe. On elytra.

Rhachomyces tenuis nov. sp.

Perithecium relatively small, the lower half or more hidden by the
appendages, hardly inflated, faintly tinged with brown, tapering very
slightly to the tip, which is sutf'used with dark brown, broad, hardly dif-
ferentiated and slightly asymmetrical. Receptacle very long and slen-
der, the cells of the main axis thirty to forty in number, dark reddish
brown or nearly opaque, subhyaline below the somewhat oblique septa,
except the lower members of the series, which are as a rule wholly
opaque ; the cells increasing slightly in size from the base upward.
Appendages straight, narrower distally, rather short and appressed, not
very numerous ; those about the base of the perithecium, about twelve
in number, somewhat larger and longer than the rest, surrounding and
concealing it more or less completely ; some of the lower appendages
also longer and curved conspicuously outward, as are the antheridia.
Perithecia about 110 X 30 /i. Longer appendages about 140-1 60 /x.
Greatest width of receptacle about 20 /x. Total length to tip of peri-
thecium 800-1000^.

On the legs of a small carabid beetle, Paris Museum, No. 113. Java.

Rhachomyces Cryptobianus nov. sp.

Perithecium hyaline or pale straw colored, very long and slender,
nearly isodiametric throughout, almost straight, the tip apparently blunt
and not well differentiated. The main axis of the receptacle consists
of about sixteen cells ; the basal cell and those immediately above it
slender black and opaque ; the rest becoming larger upward, hyaline
suffused or mottled with reddish brown. Appendages numerous slightly
divergent, becoming longer from the base upward, nearly opaque except
along the inner margin and at the tip which is generally bent abruptly

438 PROCEEDINGS OP THE AMERICAN ACADEMY.

outward, a group of about six below the base of the perithecium much
longer than the rest and curved outward in a tuft, those arising about the
base of the perithecium very elongate, erect, with straight blunt tips,
reaching nearly to the apex of the perithecium.

Perithecia 490-450 X (about) 35 /x. Receptacle 275-430 /«. Total
length to tip of perithecium G50-800 ^m. Longest appendages 300-430 ^m.

On Cryptobium capitatum, Paris Museum, No. 172. Brazil.

Rhachomyces Cayennensis nov. sp.

Perithecium rather small, yellowish, the anterior margin nearly
straight, the posterior convex ; the tip clearly and abruptly differentiated,
concolorous, asymmetrical, somewhat bent. Main axis of the receptacle
rather strongly curved, consisting of about twelve cells ; the basal ones
slender, deeply suffused, those immediately above opaque slender, the
rest rather large with central brown suffusions ; the distal cells paler.
Appendages rather coarse, crowded, black brown, opaque or nearly so, the
tips mostly bent outward, appressed below, somewhat divergent distally ;
six or more about the base of the perithecium slightly longer than the
rest, nearly equalling, sometimes slightly exceeding the perithecium in
length. Perithecium 120-140 X 25-30 /i. Total length to tip of peri-
thecium about 350 |U (average). Longest appendages 140 ^.

On Cryptobium sp. indet., British Museum No. 387. Cayenne. On
the inferior surface of abdomen.

Rhachomyces stipitatus nov. sp.

Perithecium pale straw colored becoming tinged with brownish, much
darker toward the tip; broadly subfusiform, usually symmetrical, taper-
ing from about the middle to the small blunt usually symmetrical, liardly
differentiated, often hyaline tip; borne free on a stalk-cell which is con-
colorous, sometimes as long as the receptacle, in other cases but slightly
developed. Receptacle straw colored, or faintly brownish ; the main
axis consisting of about fifteen to seventeen cells, the septa rather oblique,
its distal portion, consisting of about two to four cells, erect and free :
the cells of the secondary axis relatively large, concolorous, that opposite
the subbasal cell of the main axis bearing a long opaque blackish brown
appendage curved toward the receptacle and often equalling it in length,
other similar appendages arising at intervals above it but not from all
the lower cells, becoming more numerous throughout the distal half and
in some instances extending to or beyond the tip of the perithecium even

THAXTER. — NEW LABOULBENIACE^E. 439

in the long stalked forms, associated throughout with shorter appendages
and antheridia. Some of the individuals on A. Lespezi small, the main
axis of the receptacle consisting of only seven cellsj the perithecia nearly
sessile and small in proportion. Spores 50-60 X 4 //. Perithecia 140-
150 X 45—69 fi (100 X 30 ju in small specimens), the stalk including basal
cells longest 220 X 47 ^. Total length to tip of peritheciam about 550 ^
(200-680 /i). Receptacle 325 ^u (1 10-350 ,«). Appendages longest
400//.

On Anophilialmns Ehadnmanthus Lind., Hope Coll. No. 306, Greece ;
on A. Lespezi Fair., Paris Museum; No. 185, Grotte des Capucini,
Seine et Garonne, France.

Compsomyces Lestevi nov. sp.

Receptacle consisting of a small basal and subbasal cell, the latter giv-
ing rise to rarely more than two branches ; one of which consists of a
basal cell, from the upper side of which the stalk-cell of the perithecium
arises; while externally it gives rise to a characteristic sterile branch,
simple, usually slightly upcurved, rather closely and somewhat obliquely
septate, commonly consisthig of about nine superposed cells tapering
rather abruptly at the tip. Perithecium borne on a well developed erect
stalk-cell, nearly symmetrical, tapering from about the middle to the
broad truncate undifferentiated tip ; the basal cells small, but slightly
distinguished from the inflated base of the ascigerous portion, the spores
few and relatively large. The other branch arising from the subbasal
cell of the receptacle, an antheridial branch, divergent, consisting typically
of four superposed cells above which it becomes furcate, dividing into
two branches which iire large stout tapering, distally curved ; the third
cell of the appendage producing a single short flask-shaped autheridium
distally on its inner side and sometimes giving rise to a branch similar to
those above. Spores 42 X 4^/. Perithecia 75-85 X 30-35 ju, the stalk
75 X 22 fi. Antheridial appendage including branches 275 ^, the basal
part about 60 X 20 ju.

On Lesteva siciila Erich, British Museum, Nos. 452 and 453, Paisley
and Red Hill, England. On abdomen and elytra.

CLEMATOMYCES nov. genus.

Receptacle consisting of a basal and a subbasal cell from which arises
distally a main axis bearing a terminal perithecium and formed by a double
row of cells ; the cells of the external row producing sterile appendages,

440 PROCEEDINGS OF THE AMERICAN ACADEMY.

those of the inner producing either secondary axes similar in structure to
the primary one, or antheridial brandies; the secondary axes producino-
autheridial or sterile branches on both sides, and like the primary ones
bearing a single terminal perithecium. The antheridia simple, borne as
in Compsomyces, usually several from the distal ends of successive cells.

Clematomyces Pinophili nov. sp.

Nearly hyaline or yellowish, the basal and subbasal cells small, the
cells of the main axis in six to twelve pairs more or less . alternate on
either side, each cell of the outer series giving rise to a three or four
celled usually simple generally appressed sterile appendage, the terminal
cell of which is often smoky brown, its basal cell almost wholly united to
the cell of the axis next above ; the secondary axes one to three in
number, usually with a single basal cell, the external branches more often
simple and sterile, the inner fertile ; the antheridial appendages of both
primary and secondary axes more often simple, sometimes sparingly
branched, those near the perithecia bearing the greatest number of an-
theridia which may arise singly or opposite in pairs, or in whorls of three
or four from the distal (one to four) cells of the appendage. Perithecium
solitary, sessile at the tips of the primary and secondary axes, often
straight and symmetrical tapering to the truncate unmodified apex, pale
becoming amber brown. Spores about 38 X 3 /.i. Perithecia 100-150
X 25-40 1«. Sterile appendages, longer, 100X7 fx. Greatest length
to tip of perithecium (main axis) 300-400 X 35 fi.

On Pinoph'ilus sp. indet., British Museum, No. 390, Burmah, India.
On inferior surface.

Sphaleromyces obtusus nov. sp.

Perithecia relatively large, clear dark brown becoming almost opaque ;
the inner margin nearly straight, the outer strongly convex ; tapering
very slightly basally and distally ; the tip paler brown, abruptly distin-
guished, and when viewed sidewise flaring, with straight divergent lateral
margins, the distal margin as broad as the portion of the perithecium
below the tip and slightly concave, the outer lips more prominent and
much broader than the inner : when viewed at right angles to this posi-
tion the tip appears in general bluntly rounded not expanded, the more
or less papillate tips of the lip-cells situated in asymmetrical pairs, which
are visible above and below a broad bluntly rounded median portion
between them : the basal cells colored like the perithecium, distinct,

THAXTER. — NEW LABOULBENIACEiE. 441

hardly broader tliaa the stalk-cell which is hyaline contrasting thick-
walled and about twice as long as broad. Receptacle small, suffused
with brown, two celled ; the septa somewhat oblique, the upper cell con-
trasting abruptly with the colorless stalk-cell, and giving rise laterally to
the slightly divergent appendage, which consists of from five to six
nearly opaque brown cells, separated by oblique septa ; each producing
distally on the inner side a short hyaline branch, sometimes once
branched. Spores about 40 X 3.5 /x. Perithecia 140-150 X 40-45 /a,
the stalk-cell 35 X 20 /a. Receptacle not including foot 27-35 X 10/*.
Appendage, mostly broken 70-100 /x. Total length to tip of perithecium
230-245/1.

On Lathrobium Illi/n'cnm Dej., British Museum, No. 384. Algeria(?).
On superior surface of abdomen.

Sphaleromyces propinquus nov. sp.

Like »S'. ohlusus in size, form, and color, except that the tip of the
perithecium is symmetrical or nearly so, the lips forming a broadly
rounded blunt terminal prominence with sometimes a slight median ele-
vation, while at the base the tip is characteristically broadened through
the presence of distinct lateral elevations on either side : the stalk-cell
rather abruptly swollen below the basal cells of the perithecium.

On Lathrobium, sp. indet., British Museum, No. 383, Europe. On
superior surface of abdomen.

This species is doubtfully separated from S. obtusus on account of the
very different conformation of the tip of the perithecium, which, as the
material in either case occurred in the same position on the host, can
hardly be due to position of growth.

Sphaleromyces atropurpureus nov. sp.

Perithecium large, purplish, more or less distinctly curved away from
the appendages, tapering below, often broader distally above the middle,
tapering thence slightly to the tip, which is usually not distinctly differ-
entiated ; the apex small truncate or slightly papillate ; the basal cells
larg'^, as long as or longer than the stalk-cell, dull amber brown. Basal
cell of receptacle large, not greatly elongated, tapering below, the nearly
hyaline distal portion obliquely distinguished from the deeply suffused,
partly opaque portion of the cell below ; the subbasal cell subtriangular.
Appendage consisting of about five cells decreasing in size from below
upward, the septa nearly horizontal ; those above the basal cell giving

442 PROCEEDINGS OF THE AMERICAN ACADEMY.

rise to a branch on the inner side, which in the subbasal cell and the
cell above it consist of a very large broad basal cell, from which arise
from two to four subhyaline branches which may be once branched, the
branches crossing the stalk and basal cells of the perithecium obliquely,
usually on the left side, so that when the perithecium lies at the left, the
appendages lie above them. Spores 35 X 3.5 yu,. Perithecia 175-200 X
30-35 /A, the stalk and basal cells together 50-70 X 17-20 /a. Receptacle
85-100 X 40 /x. Total length to tip of perithecium 270-350jU. Appen-
dage without branches 50-75 /x.

On Quedius graciUventris Sharp, British Museum, No. 740 (Biologia
Coll.), Volcan de Chiriqui, Panama ; on Q. basiventris Sharp, No. 741,
from same locality. On abdomen.

Sphaleromyces Brachyderi nov. sp.

Perithecium evenly suffused with brown, paler distally, somewhat
inflated at the base, tapering slightly toward the tip; an external deep
brown almost opaque appendage projects subtermiually, exceeding the
tip of the perithecium, broad with a nearly median indentation of the
inner side, the outer margin slightly, the upper strongly curved outward,
terminating in a short blunt point or slightly inflated portion rather
abruptly distinguished on its inner side; the tip asymmetrical, one of
the outer lip-cells extendin<j above and free from the others, forming a
hyaline bluntly pointed termination at the base of whiclj the tips of the
other lip-cells form irregular prominences; basal cells concolorous, stalk-
cell short and similar to the basal cells. Basal cell of the receptacle very
large, long, attenuated below and deeply blackened, as in Camptomyces,
the distal cell subtriangular concolorous with the stalk-cell. Appendage
consisting of four or five rather flattish brown cells, decreasing in size
from below upward, their septa directed obliquely outward and down-
ward, each producing a branch from its upper inner angle, which is
simple or one or more times branched, the branchlets subhyaline.
Spores about 30 X 3 /a. Perithecia 120-140 X 28-35 /x, its appendage
35 X 12^. Receptacle 85-100 X 30-35 ^w. Primary appendage about
35 jU, with branchlets about 120^*. Total length to tip of perithecium
225-2G0^/.

On Brachyderus antennatus Sharp, in Dr. Sharp's Coll. Peru.

THAXTER. — NEW LABOULBENIACE.E. 443

MISGOMYCES nov. genus.

Receptacle consisting of numerous cells superposed singly or in tiers
of two to three cells each, terminating in a more or less irregularly
cellular base bearing appendages singly or in groups. The solitary
perithecium arising beside the appendages, the two situated in relation
to one another as in Laboulbeuia.

The antheridia could not be recognized in the material examined, the
appendages being more or less broken in all cases, but are doubtless
simple, the genus being probably related to Laboulbenia, while at the
same time it suggests some forms of Ceratomyces.

Misgomyces Dyschirii nov. sp.

Rather rich amber brown, the receptacle consisting of from eight to
twenty-three superposed cells, the upper ones rarely divided longitu-
dinally, the distal cell lying between several, usually three, smaller cells
which become separated from it on either side, and which, together with
its base, are united to the base of the perithecium ; while above it, and
separated from it by a thin dark insertion, a cellular base gives rise to
the group of appendages, the irregular basal cells of which alone remain
in the material examined. Perithecium nearly oval or very slightly
pointed, the tip and lips undifferentiated. Spores, seen only in peri-
thecium, with base apparently abruptly recurved or bent, about 60 X
3.8/^. Perithecia 70-85 X 35-40 //. Receptacle 135-400 /x. Total
length to tip of perithecium 200-435 fx.

On Dyschirius globosus Herbst., Hope Coll., No. 349, England ; on
D. salinus Schaum., British Museum, No. 582, Europe.

Misgomyces Stomonaxi nov. sp.

Hyaline or pale straw colored. Receptacle consisting of a basal and
one or two more single superposed cells, the cells above these becoming
rather irregularly divided longitudinally to form a double row of variable
length, above which a second longitudinal division appears, the recepta-
cle in this region being made up of three-celled tiers as far as the base
of the perithecium, above which its distal part consists of several super-
posed pairs of cells, or of two rows of cells more irregularly distributed,
the insertion of the appendages rather indefinite, the cells composing it
producing irregular hyaline or brownish branches distally. Perithecium
externally nearly straight, the inner margin convex, the tip rather abruptly

444 PROCEEDINGS OF THE AMERICAN ACADEMY.

differentiated, straight or curved outward. Perithecia 90-100 X 25-30
l^i. Receptacle 300-335 fi. Total leugth to tip of perithecium 365-400
X 40-45 |U.

On Stomonaxus striaticollls Dej., Britisli Museum No. 593, China.
On elytra.

Ceratomyces Floridanus nov. sp.

Purplish brown. Perithecium much as iu G. confusus, the outer mar-
gin nearly straight, the inner somewhat convex, the two inner rows of
wall cells, about twenty-four in number, rich red brown with a blackish
tinge distally; the outer wall cells, about nineteen in number below the
perithecial appendage abruptly and evenly paler, yellowish straw color or
faintly brownish ; the twentieth cell (about) forming the base of the
perithecial appendage which is black, quite opaque, curved outward and
upward and geniculate near its base the inner margin of which is
abruptly distinguished (not continuous with the adjacent margin of the
tip as in C. confusus) ; the tip distally hyaline, the apex forming a
prominent symmetrical well defined rounded hyaline papilla. Recepta-
cle consisting of three superposed cells almost wholly black and opaque
except along their anterior margins and the distal margin of the upper
cell, which are pale sti'aw yellow or amber colored, the series surmounted
by two small flattened cells from which arise the perithecium and appen-
dage respectively. Appendage long tapering, consisting of seven or
eight superposed cells, clear reddish brown with a blackish tinge,
the inner margin as well as the distal portion yellowish or amber
colored. Perithecia 300-325 X 60 fx. The appendage 150 fi. Recep-
tacle 150-160 X 75 ju. Appendage about 175-200 ^.

On Tropisternus glaber Hb, Eustis, Florida, October. On margin
of left elytron.

Ceratomyces cladophorus nov. sp.

Perithecium very large with a slightly sigmoid curvature, the lower
half conspicuously inflated above the rather narrow base, the outer mar-
gin of the inflated portion strongly convex, the inner slightly concave ;
the distal half or third more nearly isodiametric, bulging subterminally on
the inner side, the margin curving thence abruptly outward to the short
broad beak-like tip; the apex sometimes apiculate ; about the fortieth
cell of the outer row of wall-cells forming the base of a subterminal
appendage which is curved upward, geniculate at its base, rather long
slender and tapering, amber brown becoming blackish below ; the peri-

THAXTER. — NEW LABOULBENIACE^. 445

thecium at first pale yellowish, the ijiflated portion becoming rich amber
brown, the distal portion much paler except in the region of the more
deeply suffused subterminal elevation on the inner side. The narrowed
base nearly hyaline, not differentiated from the receptacle. Receptacle
consisting of three superposed cells, short, narrow below, abruptly very
broad above; the foot relatively small, the basal cell small, at first
hyaline, later becoming tinged with smoky brown ; the two distal cells
relatively very small and broadly blackened except along the nearly hyaline
anterior margin, the opaque area extending obliquely so as to involve the
geniculate base of the appendage. Appendage relatively very large and
stout, tapering in very young individuals to a slender apex and consist-
ing of from fifteen to twenty superposed cells, many of which may be
once longitudinally divided, a subtriangular appendiculate cell being
separated from the inner side, or also from the outer distally ; the
branches numerous with very long and slender branchlets which may be
several times branched. Spores 95 X 4 ^u. Perithecia 475-550 X 90-
110 (inflated portion) y<i 70 fj. (distal portion). Receptacle including
foot 85 ju. Total length to tip of perithecium (longest) 635 ^. Appen-
dage 275-475 X 45 ^, its longest branches 550 X 3 ju.

On Tropisternus niinbatus Say. Eustis, P^lorida. On the inferior sur-
face of the thorax on the left side.

Ceratomyces denticulatus nov. sp.

Similar to C. rostratus. Amber brown, the ascigerous portion of the
perithecium slightly inflated and rather abruptly distinguished from the
elongate neck, which at maturity is straight or slightly sinuous ; the cell
rows containing about fifty-five cells, the neck more slender toward the
base, distally somewhat broader; successive cells in two adjacent rows in
this broader portion projecting to form well marked rather slender tooth-
like blunt outgrowths, directed obliquely upward and separated by a
basal septum, one series usually consisting of five cells, sometimes six,
the cells immediately below often bulging prominently or forming
shorter tooth-like otitgrowths ; the second series consisting of usually not
more than three well defined similar tooth-like outgrowths : above these
two series the upper fifth (about) of the neck is bent abruptly backward,
lying nearly parallel to the portion below it ; the tip broad snout-like,
the lip-cells forming a small papillate prominence above and external to
a broad rather distinctly differentiated cell, which terminates one of the
inner rows, and is almost as large as the lip-cells taken together. Re-

446 PROCEEDINGS OP THE AMERICAN ACADEMY.

ceptacle rather slender, tapering to the base, consisting of (invariably ?)
ten superposed cells, exclusive of the foot-cell, which is not always
wholly blackened. Appendage as in G. rostratus, the numerous branches
and branchlets rather slender, not very long, forming a rather compact
tuft more or less appressed around the base of the perithecium. Perithe-
cium, ascigerous portion about 85 X 35-40 /i, neck to recurved portion
475-500 X 20-25 fx, recurved portion about 125 n, tooth-like projec-
tions 15-35 X 6-7 (i. Receptacle (ten superposed cells) 130-150 n.
Appendages (longest branches) 175 X 3 ^t/.

On a small hydrophilid beetle, Paris, Mus., No. 11, lies Mariannes,
on under surface, and legs.

Ceratomyces elephantinus nov. sp.

Closely resembling O. denticulatus, rather faintly tinged with pale
amber brown, the neck proportionately somewhat broader ; the upper
three sevenths to four ninths abruptly recurved, certain adjacent cells of
two opposite rows just below this curvature producing broad rather short
blunt tooth-like outgrowths, one to two and three to four in each
respectively; the tip broad slightly and irregularly sulcate. Receptacle
consisting of from seventeen to twenty-two squarish or flattened cells,
sometimes hardly broader distally. Appendage producing numerous
long slender flexuous branches repeatedly branched. Perithecia, asci-
gerous part about 140 X 65 f.i, neck to recurved part 475-525 ^w, re-
curved part 390-400 ^. Spores 70 X 3.5 ^. Receptacle 375-550 (i.
Longest branches of appendage 600 fx.

On Hydrohiiis sp. ?. Eustis, Florida, October. On legs.

Ceratomyces rhynchophorus nov. sp.

Hyaline. Receptacle long slender, but slightly narrower below, con-
sisting of about forty (thirty-five to fifty) superposed cells, wider than
long ; those in the lower half more flattened, the foot small. Perithe-
cium lateral, nearly erect, slightly divergent ; a short but definite stalk-
cell ; the cells at the base greatly elongated, extending some distance up
around the ascus mass and forming- together with the large elongated
supporting cell a broad sterile base to the perithecium which is not differ-
entiated from its main body; the cell rows consisting of but five cells,
including the very small lip-cells, and the cells of the sterile base ; the
three upper tiers of cells forming an abruptly differentiated, thick walled,
long, tapering beak-like termination, curved outward or inward, often

THAXTER. — NEW LABOULBENIACE.E, 447

at right angles ; below which the distal end of the outer wall-cell forms
a slight rounded prominence, the very small lip-cells forming a slight
enlargement. Appendage similar to and continuing the axis of the
receptacle directly, or diverging very slightly ; the cells giving rise
to brandies on opposite sides which are subtended by small cells
obliquely separated at the distal angles, those from the lower cells short
(antheridial ?), those from the upper long and several times branched ;
the main appendage usually broken, but in young individuals consisting
of from twenty to twenty -five superposed cells. Spores 48X3 /a. Peri-
thecia, ascigerous portion 175 X 45-50 ^u, beak-like termination 140-
160 /i, sterile basal portion about 100^. Receptacle 270-430 X 30-35^.
Appendage (young individuals) 350 |U, the branchlets 200 X 6 a.

On Phcenonotum estriatum Say. Eustis, Florida, October. On legs
and inferior surface. A form growing on the lower surface of the apex
of the elytron has enormously developed perithecia with a maximum
length of one millimetre.

Ceratomyces reflexus nov. sp.

Closely allied to C. rhynchophorus. Hyaline with a few purplish or
reddish sutfusions on the receptacle, which is composed of from about
twenty-five to fifty superposed cells ; the foot hyaline, or slightly yel-
lowish, much enlarged, bladder-like or spherical ; the distal portion dis-
tinctly broader, its axis coincident with that of the erect appendage which
forms a direct continuation of it. Peritheciura small with few asci,
abruptly recurved at the base, its apex thus sometimes touching the in-
flated foot ; nearly straight, tapering almost symmetrically to the blunt
slender tip ; the ascigerous cells situated at the base just above the small
angular stalk-cell. Appendage usually flat and broader than the recep-
tacle towards its base, the superposed flat cells of which it is composed
producing appendages on either side much as in C. rhynchophorus. Spores
70 X 4/^. Perithecia 140 X 20 //. Receptacle 140-280 ^. Appendage
200-400 ^. Foot about 30 X 30-38 ^.

On Phcenonotum estriatum Say. Eustis, Florida, October. "With
C. rhynchophorus.

Ceratomyces acuminatus nov. sp.

Hyaline. Receptacle consisting of three superposed cells, the basal cell
partly suffused and continuous with the blackened foot. Perithecium
rather stout, the outer margin nearly straight, the inner strongly convex;
the seventh wall-cell of the inner row greatly enlarged, its outer wall

448 PROCEEDINGS OF THE AMERICAN ACADEMY.

very thick, fonuing an erect tapering bluntly pointed terminal appendage,
at the base of wbich the papillate apex of the lip-cells projects on the
right side ; the fifch cell of the external row of wall-cells growing out
to form a subterminal slender appendage, eight-celled in the type, distally
attenuated, its terminal cell bearing one or two slender branches. Ap-
pendages consisting of from four to five superposed cells, the distal ones
appendiculate (the branches mostly broken). Perithecium 185 X 40 ^u.
The appendage without branches 82 ^u, the branches 150//; the rostrate
terminal cell 50 X 17 ^a (at base). Receptacle 85 X 48 fu. Spores about
70 X 3.5 fi. Appendage, broken, without branches 70^.

On Berosus sp. indet; Eustis, Florida, October. On the inferior
surface of abdomen and thorax.

Ceratomyces Californicus nov. sp.

Allied to C. camptosporus. Amber brown. Receptacle relatively
slender, consisting of three small superposed cells surmounted by two
similar cells which form the base of the appendage and perithecium; the
foot small, normal. Perithecium short and stout, from two to three
times as broad distally as at the base ; about twenty cells in each of the
inner rows of wall cells, the inner margin convex, distally abruptly bent
inward to the short beak-like apex ; the conformation of the tip, the inner
margin of which is thus horizontal or even oblique, resembling that of
C. nrnithocephahis ; about the eighteenth cell of one of the outer rows
forming the base of the usually straight rather remotely septate perithe-
cial appendage which commonly diverges at an angle of forty-five degrees
or even at right angles. Appendage small and slender (the extremities
broken in the types) becoming lateral in position. Perithecia 185-200
X (base) 30-40 (distal portion) 70-85 ^u. Receptacle 50-70 X 25 jj.
Total length to tip of perithecium 250-300 u.

On Tropisternus dorsalis Brulle. California. On the left anterior
inferior angle of the prothorax.

Ceratomyces ornithocephalus nov. sp.

Allied to 0. furcatus. Hyaline or becoming more or less suffused
with amber brownish. Perithecium relatively rather small, the ex-
ternal margin somewhat concave, the inner convex, the four distal
cells of the eight external wall-cells rather abruptly enlarged, their ex-
ternal walls much thickened and forming an irregularly rounded crest-
like prominence, the distal half of the margin of which becomes abruptly

THAXTER. NEW LABOULBENIACE^, 449

almost horizontal, terminating near the base of the beak-like pointed
apex, which projects somewhat obli(iuely from the right side : the
external row of wall cells producing above the fourth cell a large append-
age, geniculate at its base, tapering distally where it curves outward,
cousisting of from ten to twenty cells, the terminal cell rarely bearing
one or more slender branches. Receptacle consisting of three superposed
cells, the basal one usually opaque, except distally, and continuous with
the foot, surmounted by two ceils from which arise the iierithecium and
the appendage. The appendage (usually broken) curved outward and
upward, consisting of about ten superposed cells, the upper ones giving
rise to a few branches on the inner side, which may be several times
branched, the branchlets slender, mostly erect and rather ricfid. Spores
about 70 X 30^. Perithecia 120-160 X 35-45^, the crest-like tip 38-
4:0 fi broad, the appendage 120-325 ju. Receptacle 85-1 20 |U. Append-
age, exclusive of branches 140-150, «. Total length to tip of perithe-
cium 210-290 ft.

On Berosus striatus Say. Kittery Point, Maine. On margin of right
elytron towards the apex (invariably).

EUZODIOMYCES nov. genus.

Receptacle elongate, multicellular; consisting of a large and indefinite
number of cells superposed above the single basal cell and distally be-
coming divided by few or many longitudinal septa ; the distal portion
bearing a unilateral series of perithecia and appendages. Perithecia
with from nine to ten wall cells in each row, borne on a three-celled
stalk.

Closely allied to Zodiomyces. Antheridia were not distinguished, the
material being scanty and in bad condition.

Euzodiomyces Lathrobii nov. sp.

Hyaline or faintly yellowish. Receptacle long and slender, or shorter
and stouter as in Zodiomyces, according as the longitudinal septa are few
or abundant ; the superposed cells and tiers of cells sometimes nearly
a hundred in number, the upper half or more producing a unilateral series
of perithecia and appendages. Perithecia distinctly broader distally, the
fourth or the fifth to the seventh wall cells inclusive, of two opposite
rows, growing upward and outward to form well developed prominences,
giving the margin on either side in this region a bluntly serrate appear-
ance; the lip-cells arched, forming a characteristic broad dome-like apex ;
VOL. XXXV. — 29

450 PROCEEDINGS OF THE AMERICAN ACADEMY.

the two lower stalk-cells small, the upper much larger, stout, and as broad
as the base of the perithecium. Appendages long slender cylindrical,
simple or sjiaringly branclied, flexuous. Perithecia 75 X 28-30 ^u (in-
cluding projections), stalk about 40 fj., the upper cell about 22 X 14 ^m.
Total length of receptacle 200-475 X 25-70 ^. Appendages 125-230
X 4^.

On Lathrohium punctatum Zett., British Museum No. 442, Netting
Hill, England; on L. multipunctatum Grev., British Museum No. 429,
Europe; on L. fill for me Grav. British Museum No. 443, Netting Hill,
England.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 22. — April, lUUO.

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

XLIX.— HIS TO RIO AL NOTES EEL A TING TO MUSICAL
PITCH IN THE UNITED STATES.

By Charles R. Cross.

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

XLIX. — HISTORICAL NOTES RELATING TO MUSICAL
PITCH IN THE UNITED STATES.

By Charles R. Cross.

Presented March 14, 1900. Received March 20, 1900.

•

Except in a very general way there is not much to be said regarding
the early history of musical pitch in this country. The different manu-
facturers and musical organizations necessarily followed the usage abroad,
and the same gradual rise in pitch that occurred there occurred here
also.

The great harm arising from the excessive height to which the pitch
had risen at the time was recognized by those interested in the procure-
ment of the Great Organ for the Boston Music Hall, and when it was
erected in 1863, it was tuned to the French pitch, A3 = 435 double
vibrations per second, corresponding to a tempered C3 of 258.65 vibra-
tions, which had been established in France four years before. It was
hoped and expected that the result of this would be the gradual accept-
ance of the "normal diapason" as a standard throughout the United
States. A second German organ, by the same maker, Walcker of Lud-
wigsburg, built a few years later for the First Church of Boston, was
tuned to the French pitch, at which pitch it has remained up to the
present time. Also in 1868, the French pitch was introduced as a
standard into the public schools of Boston, by vote of the School Com-
mittee, although it never obtained a firm foothold there. Meanwhile the
musical instruments in use by the various orchestras were still at the high
pitch, and opera troupes and other foreign musical organizations em-
ployed the same standard. Serious difficulty was experienced from this
cause, esj^ecially when the Great Organ was used in connection with an
orchestra. After a time, in fact, at two separate periods, the Harvard
Symphony Orchestra was furnished with instruments in accord with the
organ, but apart from the concerts of this society, at theatres and else-

454 PROCEEDINGS OF THE AMERICAN ACADEMY.

where, the jjerformers were still obliged to use instruments at the high
concert pitch, which naturally caused much annoyance. Moreover, the
large organs built in this' country during the years shortly following
1863, did not copy the example of the Music Hall Organ, and outside of
Boston the French pitch was nowhere adopted. After much discussion,
and not without strong opposition, it finally resulted that a decision was
made to retune the Music Hall Organ, and raise the pitch to that ordi-
narily in use ; the feeling of those who urged the change and finally
prevailed being that while the lower pitch was desirable, and might be
the pitch of the future in this country, they were concerned rather with
the present, and might better wait the result of the etForts to introduce it
abroad, which did not at first make rapid progress. The Great Organ
was retuned in 1871 and remained thereafter unchanged at the high
jDitch 63 = 271 vibrations, and tempered A3* = 455.8 vibrations, until it
was taken down in 1884.

No further serious attempt to lower the pitch in this country was made
for a considerable time. A number of years later, however, when the
French pitch had come to be quite generally adopted abroad, the subject
again attracted attention here.

In 1882, Professor Eben Tourjee, then Director of the New England
Conservatory of Music, determined to introduce the French pitch into
that institution. For some reason, not wholly clear at the present time,
the standard chosen was a C3, a true sixth below the French A3 and
giving 261 double vibrations per second, and a fork of this pitch was con-
structed and adopted as a standard. The organ of the Conservatory was
tuned to this pitch, and the fork continued to be used as the standard of
the Conservatory until the close of 1897, when it was replaced by a new
fork of 258.65 vibrations, a tempered sixth below the French Ag. The
older fork had the disadvantage that an instrument tuned in equal tem-
perament from it would differ somewhat in pitch from one tuned in the
same temperament from an A3 at French pitch, 435 vibrations.

Soon after this date several important orchestral organizations adopted
a lower pitch than the one then ordinarily in use in the United States.
In 1882 the orchestra of Theodore Thomas employed a sort of compro-
mise pitch, slightly higher than the French pitch, viz. : A3 = 437.4
vibrations. During the seasons 1881-2, 1882-3, the first and second of
its existence, the Boston Symphony Orchestra employed a high pitch, A3
= 448.5 vibrations, but in the fall of 1883 it adopted the French pitch
as a standard, a procedure which speedily became general among
American orchestras.

CROSS. — MUSICAL PITCH. 455

During the years preceding the installation of the Great Music Hall
Organ, the pitch of organs and pianos shared in the general upward
tendency, though instruments of the former class were not infrequently
tuned at a somewhat lower pitch than that used by the orchestra. Cabi-
net organs intended for export were also in certain cases tuned to the
French pitch. But the organ pitch commonly used was substantially
identical with the high orchestral pitch, and that habitually used by
piano manufacturers was often even higher.

The general lowering of the orchestral pitch in 1883 and the following
years, of course necessitated a corresponding lowering of the pitch of
pianos and organs used in concert with the orchestra, though it was a
number of years before any general action was taken by the manu-
facturers.

In 1889 the National Music Teachers' Association at its Philadelphia
meeting adopted the French pitch, and the National League of Musicians
at Milwaukee, in March, 1891, also urgently recommended the adoption
of this standard. For several years prior to this date the question of
bringing the standard pitch used for pianos and organs into unison with
the low pitch which had come to be the generally accepted jiitch for or-
chestral use, had been agitated by a number of persons engaged in the
manufacture of pianos and organs, and especially by the late Gov. Levi
K. Fuller, of the Estey Organ Co., of Brattleboro, and Mr. William T.
Miller of Boston. Finally at a meeting of the Piano Manufacturers'
Association, held in New York, March 31, 1891, it was unanimously
decided that it was desirable that a uniform pitch should be adopted in
the United States, and a Committee was appointed, of which Mr. "VYm.
Steinway was chairman, and Gov. Levi K. Fuller, Secretary, to consider
what standard should be adopted. This committee collected much evi-
dence relating to the subject, and in response to a request therefor,
received expressions of opinion from a large number of manufacturers
and others interested in the determination of a standard, together with
sample tuning-forks giving the pitch then in use by those sending them.
The Committee reported in favor of the adoption of the A of 435 double
vibrations per second as a standard of pitch, and their recommendation
was adopted by the Association. It was also decided to call the newly
adopted standard the " International Pitch."

The International pitch is of course i lentical with the French pitch,
each having an A3 of 435 double vibrations. Some confusion has arisen
at times from the fact that the official standard A3 made in 1859, and
intended to represent the '' diapason normal," is in fact somewhat sharper

456 PROCEEDINGS OF THE AMERICAN ACADEMY.

than it purports to be, making, according to Koenig, 435.45 double vibra-
tions per second instead of 435, when at the temperature of 15° C, and
making exactly 435 vibrations only at the temperature 24.26° C. But
the legal French pitch was defined by the rate 870 single vibrations, and
not by the fork constructed by the Commission. Moreover the standard
French forks made by Koenig wei*e substantially correct in rate. The
difference is, of course, too slight to be of any consequence in practice.

The International pitch has come to be generally adopted, so that it
is now the standard pitch of this country, although it seems to be custom-
ary to tune pianos for use at concerts somewhat sharp, even up to A3
440 vibrations, which is in fact the " Stuttgart pitch " of 1834.

At various times dux'iug the past twenty years the writer has taken the
opportunity to ascertain the rates of such tuning-forks and other stan-
dards of pitch as were accessible. The results of a considerable number
of these measurements were published in the "American Journal of
Otology" for October, 1880, in a paper "On the Present Condition of
Musical Pitch in Boston and Vicinity," by Charles R. Cross and William
T. Miller. The later measurements have not hitherto been 2:>ublished.
These have been made in part by the writer and in part by several of
his assistants in the Rogers Laboratory, Messrs. Goodwin, Mansfield,
Wendell, and Burgess. The present paper is intended to include such
results as are likely to be of general interest.

Table I. is reprinted from the paper of Messrs. Cross and Miller.
The tonometer forks available at the time of its publication were less
accurate than those which have been procured subsequently, so that in
certain cases, where the standards measured in 1880 were still accessible
a remeasuremeut has recently been made, the results of which will be
found in Table II. Where this has been done, it is indicated in the
tables by an asterisk prefixed to the number designating the standard.
By a comparison of Tables I. and II. it will be seen that the remeasure-
ments have not materially altered the values obtained in the eariier
measurements.

The standard C fork upon which the measurements of 1880 were
based was a C3 fork (No. 1 of Table I.) by Koenig, belonging to the
Massachusetts Institute of Technology, the rate of which had been deter-
mined by comparison with a C3 fork by Koenig belonging to the Stevens
Institute of Technology, which last fork had been very carefully rated
by Professor A. M. Mayer of that institution. The standard A used
was a fork by Koenig assumed to be exact. From these the forks of an
improvised tonometer were rated, the C forks being of pitch C3 and

CROSS. — MUSICAL PITCH.

45i

TABLE I.

No.

Designation.

Vibration

Remarks.

Frequency.

Cs

*1

Koenig, physical pitch . . .

256.1

Stamped 512 v. s.

*2

Koenig, French pitch (approxi-

mate) ........

260.2

Stamped 520 v. s.

*3

Koenig, German pitch . . .

264.2

Stamped 528 v. s.

*4

Ritchie, physical pitch . . .

256.2

5

Koenig, physical pitch . . .

256.2

Stamped 512 v. s.

6

Marloye, physical pitch . . .

256.4

Made between 1845-50.

7

Ritchie

259.1

Made about 1868.

8

Ritchie

259.4

tc it (i

*9

Ritcliie, copy of Chickering's

standard

269.0

Made about 1868.

*10

Mason & HaniHn, French pitch

259.1

Used for a few years only.

11

Hatchings, Plaisted & Co . .

264.0

Low organ pitch, C^ fork
measured.

12

Hook & Hastings, old flat organ

pitch

264.6

C4 pipe measured. Tempera-
ture, 69° F.

13

Organ in church of the Imma-

culate Conception, Boston .

266.7

C4 pipe measured. Tempera-
ture, 69° F.

14

Smith American Organ Co . .

267.2

C4 fork measured.

15

New England Organ Co . . .

268.2

C4 fork measured.

*16

Chickering's standard fork . .

268.5

C3 fork, marked " 1865, stand-
ard pitch."

*17

H. F. Miller, pianos ....

268.9

C4 fork measured.

*18

Mason & Hamlin, present stand-

ard pitch

269.0

C3 fork measured.

19

Fork of W. H. Clement, tuner .

269.2

C4 fork measured.

20

George Woods & Co., cabinet

organs

269.5

C4 fork measured.

21

Hook & Hastings, present stand-

ard pitcli

270.0

C3 and C4 pipes measured.
Temperature, 73° F.

22

CiiickeringpianousedatJoseffy

concerts, 1880

270.1

C4 fork of tuner measured.

*23

Covent Garden pitch, 1879 . .

270 3

C4 fork furnished by R. Spice.

24

Weber pianos

270.3

String of piano measured.

25

Thomas' pitch

271.1

C4 fork furnished to builders
of great Cincinnati organ.

26

Music Hall organ

271.2

Cg, principal, great. Tempera-
ture 70° F.

*27

Steinway's pitch

272.2

C4 fork furnished by R. Spice.

*28

Highest New York pitch . .

273.9

f^ It (t u u

29

Nichols' fork, Boston, Germa-

As

nia orchestra

448

Corresponding to untempered
C3, 269

*30

Marloye's A fork

426

Imported by Prof. Levering,
1845-50.

*31

Florence pitch, Marloye . . .

438

(C (( t(

*32

Vienna pitch, Marloye . . .

446

i< ii it

33

Milan pitch, Marloye ....

448

« iC <c

458 PROCEEDINGS OF THE AMERICAN ACADEMY.

mostly large forks, while the A forks were the ordinary musicians' small
A3 tuning forks, tuned by the authors to convenient pitches. The same
forks were used for several years subsequent to 1880.

In the recent measurements three large standards of Koenig have been
used as a basis, viz.: a C3 of 256 double vibrations, a tempered C3 of
258. G5 vibrations, and an A3 of 435 vibrations, at a standard temperature
of 20° C. From these were rated by the method of beats, a series of
large Co forks and also two sets of C4 small Scheibler's tonometer forks
by Koenig, and two sets of A3 forks of the same character. These small
tonometer forks were also compared with a series of large Scheibler
tonometer forks by Koenig, which last were assumed to be correct within
the limits sought in our measurements.

Table II. gives the results of ratings of various standards of epochs
indicated by the date. In all measurements later than 1891 the Koenig
standard tonometer forks have been employed. The data given in
Table II. are all in terms of the pitches of C4 and A3. The pitch of the
standards actually measured is specifically stated in all cases except when
it is C4 or A3.

Table III. contains the results of measurements of fifty-six forks sent
by various manufacturers to the Committee of the National Piano and
Organ Manufacturers' Association in response to their request already
referred to. A preliminary rating was made by me in 1890-91 with
such forks as I then possessed. The results of this were shown in a
circular privately printed in 1891 for the use of members of the Associa-
tion. A more exact determination was made by me a few months later
in 1891, using the Koenig tonometer forks as previously stated. The
results of these measurements are found in the table.

Certain of the forks and other standards referred to in Table II.
deserve special mention. Those numbered 4, 5, 6, 7, 9, 10, 11, 12, 13,
14, 15, 16, 17, 18, 19, 87, 88 were intended to give the physical pitch.
Nos 2 and 3 had been in the possession of their owners for many years,
and were authenticated as to the date assigned to them. Nos. 5 and 6
are two forks belonging to Harvard University, imported many years
since. No. 5 is of the early Marloye pattern, with inclined prongs, but
does not bear any mark to indicate the maker. No. 6 bears the mark
" R. K.," always employed by Koenig. Nos. 16 and 17 are also Koenig
forks of early date, belonging to Harvard University. No. 8 is a pitch
pipe formerly employed for church use, belonging to Mr. B. J. Lang of
Boston. It is a whistle with a movable plug, and the pitch can be varied
through an octave. When the plug is set for C4, according to the lines

CROSS. — MUSICAL PITCH.

459

TABLE II.
C4

No.

Vibration
Frequency.

Designations and Remarks.

1

506

Tonic Sol-Fa, 1882. N. E. Conservatory of Music. Rated
in 1882.

2

509.7

Old Fork of J. T. Batchelder, 1782. Rated in 1884.

3

510,8

Old Fork of G. A. Emery, 1840. Rated in 1881.

4

oH.l

E. S. Ritchie & Sons, Standard C4 ; early f(jrk.

5

511.5

C3, 255.75, Marloye pattern, Harvard University.

6

511.7

C3, 255.85, R. K., ; early fork, Harvard University.

7

511.8

C3, 255.9, Tufts College.

8

512

Early Pitch Pipe, Maker's Standard. T = 23^C. B.J.Lang.

9

512.1

C3, 256.05, R. K. " 512 u. s." slender pronged ; early fork,
M. I. T.

*10

512.1

C3, 256.06, R. K. " 512 v.s.," 1873, Basis of measurements
of 1877, M. I. T.

11

512.2

C3, 256.1, E. Greaves, " Scheibler Pitch."

12

512.2

R. K. " 1024 r. s.," iM. I. T.

*i:-]

512.3

E. S. Ritchie & Sons, Standard C3, 256.17.

14

512.4

E. Greaves, " Scheibler Pitch, 512."

15

512.5

W. T. Miller, 1880. Physical Pitch.

16

512.6

C3, 256.28, R. K., "512 v. s.," ; early fork (prior to 18G8),
Harvard University.

17

512.7

R. K., C4 " 1024 V. s." ; early fork, Harvard University.

18

514.7

Walmsley, Fuller & Co., Chicago, 1895.

19

514.8

Co, 128.7, E. S. Ritcliie & Sons, makers. University of
Virginia.

20

515.5

L. K. Fuller; electrically welded fork, 1893.

21

515.6

Tempered C4, French Pitch, 1883. Small Fork, M. I. T.

22

515.6

"517.3 Piano Mfrs. Ass'n." New England Conservatory,

1808.

23

517.0

"Philharmonic." Tufts College.

24

517.3

Pitch Pipe, 1826, Yale University.

25

517.3

Tempered C4, French Pitch, 1884. Small Fork. Rated in

1884.

26

517.4

^Miller's Standard Tempered Cg, 258.68, 1884. Rated in 1884.

27

517.5

Tempered C4, Small Fork, French Pitch, 1883, Boston
Symphony Concerts. M. I. T.

28

517.5

Cg, 258.78, Ritchie, early fork, Marloye pattern. M. L T.

29

517.5

Tempered C4, French Pitch, 1884. Large fork on box.
M. I. T.

30

517.6

E. S. Ritchie & Sons, Cg, 258.78, early fork, Marloye pat-
tern. M. I. T.

31

517.7

E. S. Ritchie & Sons, C3, 258.83 ; early fork. University
of Virginia.

32

517.8

Chickering & Co., new tempered C3, 258.0, French Pitch,
1884. Rated in 1884.

33

517.8

Tempered C4, French Pitch, 1883. M. I. T.

34

518

Early Pitch Pipe. Local Pitch, T = 23° C. B. J. Lang.

»35

518.5

Mason & Hamlin, Weighted Cg, 259.24. French " Stand-
ard 1866," Pitch.

30

518.7

G. S. Hutchings, "517.3," "International Pitch."

37

519.0

C4 Pipe giving Thomas' Pitch of 1883. T = 22° C.
Hook & Hastings.

460

PROCEEDINGS OP THE AMERICAN ACADEMY.

TABLE II. — Continued.

No.

Vibration
Frequency.

Designations and Remarks.

38

519.0

C3, 259.5. N. E. Conservatory Standard of 1882; on
box. Remeasuremeut, 1898.

39

519.2

Do. dismounted, 259.6. Do. ^

40

519.6

Do. on box, 259.8. Do.

*41

520.1

C3, 260.07. R. K. " 520 v. s." 1873.

42

520.4

Pitcli used by Patti, 1882. L. K. Fuller.

43

520.6

G. S. Ilutchinijs. International Pitch, C4, natural.

44

520.6

" 522. Estey Organ Co." N. E. Conservatory.

45

520.6

" N. E. Conservatory, 1892." L. K. Fuller. '

46

520.8

" 522, Estey Organ Co." G. S. Hutchings.

47

520.9

"International Pitch." Fork of M. Steinert & Sons Co.,
Boston, 1898.

48

521.4

C3, 260.7. French Pitch, on box, N. E. Conservatory
Standard of 1882. Rated, 1883.

49

521.8

C3, 260.9. Copy of N. E. Conservatory Standard of 1882,
M. I. T., on box. Rating of 1883.

50

521.8

C3, 260.9. E. S. Ritcliie & Sons, Standard C, "261."

51

522.1

True French C4, 1883. M. I. T.

52

522.2

C3, 261.1. Copy of N. E. Conservatory Standard of 1882.
M. I. T., off box. Rating of 1883.

53

522.4

C3, 261.1. Copy of N. E. Conservatory Standard of 1882.
M. I. T., on box. Rating of 1898.

54

523.5

Tiieodore Thomas' Pitch, 1882, N. E. Conservatory.
Rated in 1882.

55

527

Whitney & Raymond Organ Co., 1882. N. E. Conserva-
tory. Rated in 1882.

*56

528.2

C3, 264.11, R. K., 1873, " 528 v. s.," German Pitch. M. I.T.

57

532.8

C3, 266.4. Princeton University.

58

535.3

R. & M., Richmond, Va. L. K. Fuller, 1883.

59

536.0

Standard Pipe, Pitch of 1884, " 540." Hook & Hastings.

60

536.3

Hazelton, L. K. Fuller, 1883.

61

536.8

Fork of period 1880. W. T. Miller.

62

536.8

Decker, L. K. Fuller, 1883.

63

536.8

C3, 268.4. Copy of Chickering's Standard " 1865." N. E.
Conservatory. Rated in 1S82.

*64

536.9

C3, 268.44, Ciiickering Co., Standard fork, " 1865." " Con-
cert Pitch."

65

537.3

McPhail, L. K. Fuller, 1892.

*66

537.5

Miller, standard of 1880.

*67

537.9

C3, 268.96. E. S. Ritchie & Sons, 1870. Copy of Chick-
ering's Standard.

68

538.1

Hodge & Essex, L. K. Fuller, 1892.

*69

538.2

Cg, 269.1. Mason & Hamlin Co., " Standard Pitch, 1866."

70

538.3

Isotonic Fork, High Pitch, 1880.

71

538.6

" C, 540," Estey Organ Co. N. E. Conservatory.

72

538.9

"540," Estey & Co. G. S. Hutchings.

73

539

Chickering Concert Pitch, 1882. N. E. Conservatory.
Rated in 1882.

74

539.3

"Philharmonic." L. K. Fuller, 1883.

75

539.8

E. Greaves, " 540, Scheibler Pitch."

*76

540.2

" Covent Garden," 1879. R. Spice.

77

541.4

W. T. Miller, Fork of period, 1880.

78

542.5

Sanders, L. K. Fuller, 1883.

*79

544.5

Steinway, New York, 1880. M. I. T.

CROSS. — MUSICAL PITCH.

461

TABLE 11.— Continued.

No.

Vibration
Frequency.

Designations and Remarks.

80

545

Theodore Thomas' Higli Pitch prior to 1893. N. E. Con-
servatory. Hated in 1883.

81

545.1

Eeed Fitch Pipe, 1885. Rated in 1885.

*82

547.7

Cg, 273.85, Higiiest New York Pitch, 1878. R. Spice.

83

549.0

" Philharmonic Pitch," Standard Fork of Steinert & Sons
Co., 1898.

As

84

415.3

Ao, 207.7 " Shore Fork," 1715. L. K. Fuller.

85

422.3

" Handel Fork," 1749. L. K. Fuller.

86

428.7

Pjpe, Pitch of 1889, Hook & Hastings, T = 21.7° C.

87

426,2

A.i, 213.1, Ritchie early fork, University of Virginia.

*88

426.5

Marloye, " 853^ " Harvard University.

89

428.8

Tufts College.

90

432.8

French A, Boston Symphony Orchestra, Henschel's
original small fork, 1883.

91

433.0

Ritz, L. K. Fuller, 1883.

92

434.3

" Piano Mfrs. Ass'n." G. S. Hutchings.

93

434.5

" A, 435, Piano iMfrs. Ass'n." N. E. Conservatory.

94

434.6

" A, 435, Piano Mfrs. Ass'n." M. I. T., from L. K. Fuller.

95

434.7

E. Greaves " French Pitch."

96

434.8

R. K., French Pitcli, correct at 15° C. 1883, M. I. T.

97

434.9

Electrically welded fork, L. K. Fuller, 1893.

98

434.9

R. K., A 435 at 20=" C. Mason & Hamlin Standard, 1898.

99

435.0

Standard R. K., "870 v. s. at 20°." G. S. Hutchings.

100

435.0

R. K., A 435 at 20° C. H. F. Miller, Standard, 1891.

101

435 0

R. K., A 435 at 20° C. Chickering Co., Standard, 1898.

10-2

435

Pitch Pipe, 1826. Yale University.

103

435.4

Ritz. L. K. Fuller, 1883.

104

435.4

Small R. K. Fork, " 870 r. s.," Tufts College.

105

437.4

Theodore Thomas' Pitch, 1883. M. I. T.

106

437.7

Theodore Thomas' Pitch, 1883. Georg Henschel. Rated
in 1883.

107

437.8

Ritz. L. K. Fuller, 1883.

108

438.1

" French Pitch." L. Waldo, 1898.

*109

438.4

" Florence Pitch." Harvard University.

110

441.3

E. Greaves, " Scheihler Pitch," German Pitch.

*lll

446.0

" Vienna Pitch." Harvard University.

112

446

Reed Pitch Pipe. Rated in 1885.

113

446.7

Carl Eichler, L. K. Fuller, 1883.

114

448.0

L. K. Fuller, 1883.

115

448.5

Boston Symphony Orchestra, Standard A, 1882-83.

116

449.1

L. K. Fuller, 1883.

117

451.5

Ciiickerine's Pitch, L. K. Fuller, 1883.

118

454.1

W. T. Miller, 1880.

119

457.5

L. K. Fuller, 1883.

120

457.7

Steinway, L. K. Fuller, 1883.

462 PROCEEDINGS OF THE AMERICAN ACADEMY.

made for the purpose, the rate is, as given in the table, 512 vibrations.
Notches have been cut in certain places, appareutly to give the "local
pitch" used by the choir. When adjusted by these (see No. 34), the rate
is 518, practically French pitch. The frequency of vibration of such a
pipe is of course greatly influenced by the temperature. This pitch pipe
was used in a church in "West Townsend, Mass., early in the present
century, and its pitch was adjusted to that in use in Boston.

No. 10 is a Koenig fork already referred to, made for the Massachu-
setts Institute of Technology in 1873, and used as the basis of measure-
ments by Cross and Miller in 1880. Like all of Koenig's forks, prior to
the establishment of his new standard in 1880, it is a little sharp. Nos.
11, 14, 95, 110 are forks made by Mr. E. Greaves. No. 13, the stan-
dard Cg of E. S. Ritchie and Sons, was procured from Duboscq of Paris
at a date prior to 1870. Jt has the inclined prongs of the Marloye forks.
Nos. 19, 31, 87, are large forks made for the University of Virginia by
E. S. Ritchie and Sons more than twenty-five years since. No. 24 is a
pitch pipe deposited in the library of Yale University by the class of
1826, and kindly rated by Professor A. W. Wright. No. 102 is the
same pipe when set to give A3. When properly blown it gives, substan-
tially, French pitch. No. 26 is a fork used by Miller and Sons as their
standard for piano pitch in 1884. No. 29 is one of a pair of large forks
on resonating boxes made in 1883, for the purpose of tuning the Chicker-
ing piano when used in the Boston Symphony Concerts, the orchestra
having adopted the French pitch. It was tuned from No. 27. The firm
desired to tune the piano from C rather than from A. No. 32 is a large
tempered C fork, adopted in 1884 by the Chickerings for tuning pianos
to be used with orchestras employing the French pitch. No. 35 is the
standard French Pitch of the Mason and flamlin Co. The fork No. 69
is lowered in pitch by attaching to each prong by wax a small rectangular
piece of steel. No. 36 is a fork of G. S. Hutcliings and Co., used in tun-
ing the organs made by them. No. 37 is a flue pipe belonging to the
Hook and Hastings Organ Co., and giving the pitch proposed by Theodore
Thomas in 1883. Nos. 38, 39, 40 give ratings under difl^erent conditions
of a large standard fork made by Ritchie and Sons for the New England
Conservatory of Music in 1882. No. 48 is the same as measured in 1883.
The fork has flattened by a considerable amount since its manufacture,
from unknown causes. It has apparently been kept with care. The box
has a considerable influence upon the j^itch of the fork. No. 42 is a fork
belonging to Mr. Levi K. Fuller, giving the pitch which was used in
opera in 1882 by Patti. It is a little sharp of French pitch, though

CROSS, — MUSICAL PITCH. 463

far below the high pitch then habitually used in this country. No. 47
is a small fork, the standard for "International Pitch" of the M.
Steinert and Sons Co., Boston. Nos. 1, 54, 55, 63, 71, 73, are forks
belonging to Mr. F. W. Hale of the New England Conservatory of
Music. No. 57 is an old C3 fork belonging to Princeton University, and
probably dating from the time of Prof. Joseph Henry. It is considerably
rusted. No. 59 is a flue pipe giving the high pitch used by Hook and
Hastings in 1884. No. 64 is the former standard fork of the Chickerings,
made in 1865, and giving the high concert pitch then in vogue. It is a
large C3 fork, marked " 1865, Standard Pitch," with its prongs inclined
towards one another. No. 67 is a copy of the Chickering fork made by
E. S. Ritchie and Sons for their own use in 1870. It has sharpened
somewhat after tuning. No. 69, the Mason and Hamlin Co.'s former
standard, is a large fork, almost a counterpart of the Chickering fork.
It is marked " Standard Pitch, 1866." No. 76 is a small fork furnished
by Mr. Robert Spice of Brooklyn, in 1879, and giving the pitch then
used in the Covent Garden Theatre. No. 82 is a bell metal fork made
in 1878 by Mr. Spice, and giving the highest pitch then used in New
York. No. 83 is a small fork at the high " Philharmonic Pitch " of the
Steinert and Sons Co.

Among the A3 forks. No. 84 is a very old and low-pitched fork, pro-
cured by Mr. Levi K. Fuller in England, and purporting to have been
made in 1715 by John Shore, the inventor of the tuning-fork, and to be
the oldest fork in existence. No. 85, which also belonged to Mr. Fuller,
is supposed at one time to have been used by Handel. No. 86 is a pipe
giving the pitch settled upon by Hook and Hastings in 1889. Nos. 88,
109, 111 are small forks imported by Prof Joseph Levering for Harvard
University, between 1845 and 1850. No. 90 is one of a number of
small forks made for Mr. Georg Henschel in 1883 as a basis for the
pitch of instruments to be made for the Boston Symphony Orchestra.
It was tuned from tlie forks of a Valentine and Carr tonometer (see No.
3, Table V.), and owing to the extreme flatness of the ''A,432" and
"A, 436" used in the comparison, it is considerably below French pitch.
No. 96 is a standard A3 '• French Pitch," by Koenig, imported by the
Massachusetts Institute of Technology in 1883. It is correct at 15° C,
and hence at 20° C, the present standard temperature, is slightly flat.
Nos. 98, 99, 100, lOl are large Koenig standard forks, mounted on boxes.
No. 102 is the same pitch pipe as No. 24 when adjusted to give the Note
A3. Nos. 105, 106 are small forks giving the medium pitch proposed and
used by Theodore Thomas in 1883. No. 106 was given to Mr. Henschel

464

PROCEEDINGS OF THE AMERICAN ACADEMY.

by Mr. Thomas. No. 105 was copied from this. No. 113 is a fork
giving the pitch used by the Germania Orchestra in 1883. No. 115 is
a high A3 fork, made by Mr. R. Spice, and belonging to the Massachu-
setts Institute of Technology, which was used by the Boston Symphony

TABLE III.
C4 Forks.

No.

1

2

3

4

5

6

7

8

9
10
11
12
13
14
15
16
17
18
191
20
21
22
23
24
251
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40

Designation.

Knabe & Co., Baltimore (Low Pitch)

Roosevelt Organ Co., N. Y

C. B. Snyder, VVinfield, Kan

J. H. &. C. S. Odell, N. Y

Chickering & Sons, Boston (Low Pitch) ....

J. & C. Fischer, N. Y

Wilcox & White Organ Co., W. Meriden, Conn .

Jewett & Co., Leominster, Mass

Shoninger Organ & Piano Co.. New Haven, Conn.

Gallup & Metzger, Hartford, Conn

J. H. Foote, N. Y

Francis Bacon, N. Y

Dyer & Hughes, Foxcroft, Me .

A. Weber, N. Y

Mason & Hamlin Organ & Piano Co., Boston . .

Vose & Sons, Boston

Benning & Sons, N. Y

C. C. Briggs & Co., Boston

A. M. McPhail Piano Co., N. Y

Chickering & Sons, Boston (High Pitch) . . .

Clough & Warren Co., Detroit

Atlanta Piano Co., Atlanta, Ga

Geo. Steck & Co., New York

Wm. E. Wheelock & Co., N. Y

Boardman & Gray, Albany, N. Y

Estey Piano Co., N. Y

Decker Bros.. N. Y

Mehlin & Sons. N. Y

R. M. Bent & Co., N. Y

Pease Piano Co., N. Y

A. B. Chase Co., Norwalk, Ohio

Newby & Evans, N. Y

Knabe & Co., Baltimore (High Pitch) . . . .

Stirling Co., Derby, Conn

Hazelton Bros., N. Y

Decker & Son, N. Y

Conover Bros., N. Y. . ,

Sherman Clay & Co., San Francisco

Lester Piano Co., Philadelphia

Leicester Piano Co., Leominster, Mass

Vibration
Frequency.

508.
516.
518,
£20.
521.
526.
531.
532.
533.
534,
534,
535,
536,
536,
537,
537,
537,
537.
537,
538,
538,
538.
539,
539,
539,
539,
539
539,
539.
540.
540.
540.
510.
541.
542.
542.
542
545
547.
549

Forks very poor and hard to rate.

CROSS. — MUSICAL PITCH.

465

TABLE III. — Continued.
A3 Fork&

No.

41

42

43

441

451

46

47

481

49

60

51

52

53

54

55

56

Designation.

F. Werlein, New Orleans, La

Mason & Hamlin Organ & Piano Co., Boston . .

Philadelpliia

C. H. W. Ruhe, Pittsburg, Pa

Hook & Hastings, Boston

Clougli & Warren Organ Co., Detroit

Kranich & Bach, N. Y

Geo. Jardine & Son, N. Y

J. H. Foote, N. Y

W. \V. Kimball Co., Chicago

Hallet, Davis, & Co., Boston

Soluner & Co., N. Y

Krakauer & Bros., N. Y

Steinway & Sons, N. Y

Chas. M. Stieff, Baltimore

Keller Bros. & Blight, Bridgeport, Conn. . . .

Vibration
Frequency.

431.
434.
439.
441,
443.
444.
444.
447
449
451
453
454
454
456,
45G
458,

TABLE IV.
Forks issued in 1892 as representing " International Pitch."

A3

C4

M

424.1

M

434.4

M

445.0

M

516.3

M

429.8

M

434.6

M

445.5

M

517.2

N

432.3

M

434.8

M

491.9

M

517.7

M

433.1

M

435.4

M

498.1

M

519.4

N

433.3

0

435.7

M

520.2

Orchestra as a standard in 1882-83, the year prior to the introduction of
the French pitch. In the following years an exact copy of No. 96, a
large Koenig A3 435 fork mounted on a resonating box was used, which
was subsequently replaced by an electrically driven fork made by
Wolters of Vienna. The orchestra also possesses a Koenig A3 fork

1 Forks very poor and hard to rate.

VOL. XXXV. — 30

466

PROCEEDINGS OP THE AMERICAN ACADEMY.

mounted on a box and giving International (French) pitch, with which
the last-mentioned fork can be compared.

In Table IV. are given the rates of a number of small musicians' forks
loaned by Mr. L. K. Fuller, which were sold in 1892 as representing the
new " International Pitch." The attached letters, M, N, O, denote the
different dealers who furnished them. It will be seen that the C4 forks
range from, approximately, 516 to 520 vibrations, the true value being
517.3 vibrations, and the A3 forks from 424 to 498 vibrations, the differ-
ence in the latter case amounting to over a major tone, and showing
an extreme of carelessness in tuning and comparison that is almost
incredible.

TABLE V.
Tonometer Forks. Valentine and Carr.

Designated
Frequency.

Measured Frequency.

1

2

3

4

5

420

419.6

417.5

424

423.6

421.5

428

427.7

425.5

432

431.8

429.5

429.8

430.3

436

435.7

433.4

433.7

434.4

440

439.8

437.5

437.4

438.3

444

443.4

441.2

441.2

442.5

448

447.3

445.6

445.5

446.6

452

451.1

449.3

456

454.9

453.3

256

255.8

255.8

512

511.2

511.8

512.0

516

515.4

536

535.3

540

539.2

539.0

Ratings have also been made of forks selected at random from several
tonometers. In Table V. will be found the results of measurements upon
forks from four tonometers by Valentine and Carr. The forks are all
of the small pattern ordinarily used by that firm. The numerals 1, 2, 3,
4, 5, heading the corresponding columns indicate the particular set of

CROSS. — MUSICAL PITCH. 467

forks referred to. With the exception of No. 5, the tonometers are long
range ones, running from C3 to C4. No. 5 is 11 short range tonometer
of 12 forks, from 512 to 544 vibrations, purchased by the Massachusetts
Institute of Technology, in 1882. No. 1 belonged to Mr. Levi K. Fuller
and was purchased by him in 1892. No. 2 was made for the Massachu-
setts Institute of Technology in 1882. No. 3 refers to the same tono-
meter. The values under (2) were obtained by comparison of the forks
with those of (1). The values under (3) were obtained by direct com-
parison with Koenig's forks, as were all the results in Table V. except
those under (2). No. 4 is a tonometer belonging to Harvard University,
and which was purchased by Professor Lovering about 1883. The forks
taken for comparison were A3 and C4 forks. It will be seen that the
error is in some cases very considerable, amounting for some of the A3
forks to several vibrations.

The small tonometer forks of Koenig are usually closely in accord with
his large standard and tonometer forks, not often deviating from the num-
bers stamped upon them by as much as one tenth of a double vibration.

In all the measurements referred to in the present paper time-intervals
were measured by the use of an accurate stop-watch reading to one-fifth
of a second.

Rogers Laboratory of Physics,

March, 1900.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. Ko. 23. — May, 1900.

THE DRIVING ENERGY OF PHYSICO-CHEMICAL

REACTION, AND ITS TEMPERATURE

COEFFICIENT.

By Tukodoke William Richards.

Investigations os Light axd Heat, made and publisued wrollt or in part with Appropriations

from the rumford fcnd.

THE DRIVING ENERGY OF PHYSICO-CHEMICAL

REACTION, AND ITS TEMPERATURE

COEFFICIENT.

By Theodoke William Richards.

Received April 10, 1900. Presented April 11, 1900.

Evert beginner in Physical Chemistry of course notices the similarity
in form of the equations of Clausius and van't Hoff, - = ^

and 7- = , . The differences in the interpretation of the sym-

bols are the only hindrance to the welding of the two equations into
one. P usually consists of a single quantity, a pressure, while K is
usually a product of concentrations. On the other hand, the second
member of the equation sometimes includes work (A. = ^ — U) and
sometimes contains simply the diminution of the internal energy (U)
of a process conducted at constant volume.

Since these distinctions are so confusing that errors from the lack of
proper discrimination have crept into some of our best and most helpful
treatises, the present paper, which is a treatment of the subject from a
single point of view, may be of assistance to students who have not mas-
tered the matter. I have found it advantageous, both in teaching and in
thinking, to look at the whole field from the point of view of pressure.
This practice has recently received support in an interesting article by
Arrhenius * on the speed of reactions.

That the pressures affecting a given reaction should exert the deciding
influence in determining the magnitude and direction of a chemical change,
is no more astounding than that the resultant of mechanical forces should
be the agent determining the change of position of finite masses. Hence
the consideration of pressure, which has the dimensions of an intensity,
seems to be a more direct method of studying the progress of a reaction

* Zeitschr. phys. Chem. 28, 317 (1899).

472 PROCEEDINGS OF THE AMERICAN ACADEMY.

than does the consideration of volume or entropy, which have the dimen-
sions of caj^acities merely.

There are of course many possible ways of attacking the subject math-
ematically. Perhaps the simplest is given below.

In his recent comprehensive and exceedingly interesting paper on
equilibrium and free energy,* Gilbert N. Lewis has shown that the
following equation is a very general one, applying to both homogeneous
and heterogeneous systems, but rigorously accurate only when the sys-
tem is composed of ideal gases, ideal solutions, and "■ condensed phases "
of constant volume. This is of course the equation of van't Hoff.

»5 "i ■>}' "'i . . ^ "2 /•' "'2 rT

Vi y J 1 . . . ^ Co ■= C 2 - • ■ • _ (J .^ .

dT dT ' RT^

In this expression v means molecular volume ; T, absolute temperature ;

c, concentration f = - J ; w, the number of reacting molecules of any

given molecular species ; U, the diminution in the internal energy of the
process, or the heat of reaction at constant volume ; and R, the gas con-
stant (1.98 calorie-units, or better 8.31 c. g. s.). The products of a
reaction are indicated by a subscript 2, the factors by 1. The chief dis-
advantage of the equation for practical purposes is the fact that the
numerator of the second member does not always represent the actual
heat of the reaction, since it does not take account of the work which the
reaction may involve on account of changing volume.

Let us now introduce pressure instead of concentration into this

expression, c = 777^, hence the expression becomes
R I

d (R Tf^- + "'' • • •>j9i"'/i"'' ■ . . _ U

* These Proceedings, 35, 1 (1899) ; Z. phys. Chem. 32, 364 (1900).

I am much indebted to Dr. Lewis for valuable mathematical criticism of the
present paper. To Dr. Edgar Buckingham also I owe thanks for his interest,
and for valuable suggestions as to matters of detail. Since reading the manuscript
the latter has derived the " reaction-isobar " according to the method of Duhem ;
but in the present exposition I have retained the original derivation, because it is
much simpler than Ids more rigorously exact method, as well as because most
readers of physical chemistry are accustomed to the method and nomenclature of
Nernst.

RICHARDS. ENERGY OF PHYSICO-CHEMICAL REACTION. 473

Cancelling the unnecessary powers of R T, we have

JT p^"^-p\-'- ... - RT^' ^ ^

Let us consider for the sake of argument a case in which

(rio + n\ ... — Ml — n\) = N

is positive, that is, a case in which the total number of gram-molecules in
the product is greater than that in the reacting mixture before the reac-
tion took place. The expression then becomes

dT j9/'2//^... RT^' ^^

From this pressure-equation at constant volume we may omit c?ln^,
because ^ is a constant. Thus

« 111 } ; y-r ' /A\

jOo"2;>V'^-- • • U (4)

dT R T' '

Here two cases may arise. If iV = 0, that is, if the osmotic or gas
pressure (or total number of molecules) does not change during the
reaction, the pressure remains constant. Mathematically, T'' becomes 1
and hence disappears. On the other hand, if iV^ 0, T'^ becomes a
serious factor in the equation, affecting immensely the temperature coeffi-
cient of the " mass-law constant."

The numerator of the second term of this modified equation of van't
Hoff still consists merely of U; hence if the equation is to be used with
data obtained under constant pressure, the observed heat of reaction must
be corrected for the work done during expansion. For this reason its
prototype has been called the " reaction isochor " by Nernst ; it repre-
sents immediately the observed conditions only when the reaction takes
place in constant volume. The introduction of the correction causes an
interesting simplification.

At constant pressure the heat evolved during the reaction would be

less than U, because the work JSfR T is done against constant pressure ;

N R T N
hence we shall be obliged to subtract = ^ from each side of

the equation in order that the numerator of the second member may
truly indicate the actual conditions. But

474 PROCEEDINGS OP THE AMERICAN ACADEMY.

N _Nd\nT
T~ dT '

hence by simply subtracting the equations we obtain

d rV %2"=/2"'= ... ) dT ~ RT- RT''

Simplifying, and noting that — C + N R T =^ X, the heat actually ab-
sorbed in the reaction, we may write

_^, Px"^p\"'^ ... _ - A

dT p."2p'/2...- RT^' ^^^

Here — X signifies heat given out when the reaction takes place un-
der constant gas or osmotic pressure, because X signifies heat absorbed,
or latent heat. The equation states that when pressures are being com-
pared, the latent heat under constant pressure is the term to be used in
calculation.

Hence at constant outside pressure the heat observed in any reaction
bears a very simple relation to the acting pressures which take part in
that reaction, just as at constant volume the concentrations are similarly
related to the observed evolution of heat. This equation may be called
the " Reaction-isobar," on account of the fact that constancy of total
pressure is the condition essential to its immediate application. The
equation of Clausius is a special case of this law.

Evidently these reactions isobar and isochor may now be combined
into one, for the obstacles to this union are now removed. Before com-
bining them, it is advantageous to examine into the meaning of

d\n(R7y=d\n T'' ,

which constitutes the only puzzling difference between them.

The physical meaning of this important factor in the equation is inter-
esting. From its mode of introduction into equation (2), this term is
clearly the representative of the pressure caused by an addition to the
total number of molecules in a given space. Wheu the total pressure is
artificially maintained at a constant value, In T^ disappears from the
equation, just as it does when the pressure remains constant because the
total number of molecules present is not increased by the reaction. This
is equivalent to defining N in its present situation as a direct function of
the increase in the total pressure of the reacting svhstances during the
reaction. In order to avoid conflict with the original definition of N, we

RICHARDS. — ENERGY OF PHYSICO-CHEMICAL REACTION. 475

shall call this newly defined exponent N'. Mathematically, the inspec-
tion of the equations (4) and (5) shows that the expression

(where ^ is the total number of grara-molecules in the initial reacting
mixture, and Pi and P2 the initial and final total pressures respectively)
is true either for constant pressure or constant volume. Whether this
still holds true for the ill-defined cases where neither volume nor pressure
is constant, we shall not at present determine. In constant volume,
according to this definition, N' = N ; under constant pressure, likewise
according to this definition, N' = 0. With the help of this new defini-
tion, together with the use of the symbol Q to signify in general the
actually observed latent heat of the reaction, it becomes possible to com-
bine the reactions-isochor and isobar into one equation, from which either
may be logically obtained again, according to the circumstances attending
the reaction. Thus, we may write : —

which is a general expression for the temperature coefficient of the equi-
librium ratio * of dilute reactions.

If the total osmotic or gas pressure is kept constant, N' = 0, hence
T''' — 1, and disappears, and Q = X = NRT — U. This is the reac-
tion isobar. If, on the other hand, the volume is kept constant, T''' rep-
resents the increase in pressure which at constant pressure would have
represented work, and — Q ^ U. This is the reaction isochor. If ^'
is 0, and — Q = U, the equation is both isochor and isobar.

For the general expression (6) which defines the temperature-coeffi-
cient of the equilibrium ratio I propose the name "reaction metatherm"
(fieTOL, between). The chief advantage which it possesses over the con-
centration equation is the fact that its term — ^ is always the heat
actually evolved, whether work is done or not, as well as the fact that it
deals with the more rational dimension, — pressure.

Since Q is the heat really observed, it is clear that the expression
must be the analytical statement of the theorem of Maupertuis or Le

* This satisfactory name is due to Lewis. " Mass law constant " is less satisfac-
tory, because the quantity is not constant, and the law seems to be due to pressure
rather than to mass.

476 PROCEEDINGS OF THE AMERICAN ACADEMY.

Chatelier as far as heat is concerned. In order to trace this evident
necessity, one must assume a somewhat puzzling inverted attitude. The
pressures indicated by the formula define a condition of equilibrium, not
a condition of action. It is clear, then, that a small pressure in the
numerator of the logarithmic expression means a great tendency towards
the denominator. That is, a growing denominator means an increasing
tendency to change from factors to products. But the logarithms in the
denominator take the minus sign, or the sign of Q, which represents
heat absorbed. Hence a reaction which absorbs heat evidently must be
pushed farther by increasing temperature.

The inverted attitude just mentioned may be easily remedied by con-
sidering carefully the nature of the quantities involved. We have seen
that the expression

P2'^P 2 = • • •

the familiar " mass-law " expression for equilibrium expressed in terms
of pressure, seems to represent the resultant reacting tendency of a given
reaction at constant temperature ; because it is this quantity which is
concerned with the theorem of Maupertuis. When concentrations are
used, it is difficult to imagine any physical meaning in this equilibrium
ratio ; but when the expression is conceived of in terms of pressure,
we may look upon k as an opposing tendency which has been balanced
by the ratio of the pressures observed. That is, we may call k the
"reaction tendency." This means not merely pressure, but work ; be-
cause the expression R T \ ~ from which it was originally derived

means work. It represents then the variable factors in the " driving
enerfry " of the reaction.

AVe may conceive of this reaction tendency as consisting of a number of
individual reacting tendencies, one for each substance. But the pres-
sures in the pressure-equilibrium ratio do not directly represent the
individual reacting tendencies of the substances represented. They are
only the pressures which remain in equilibrium when all the reacting
forces have been balanced. When under the circumstances a given
individual pressure is small, we must ascribe to that substance a great
reacting tendency, and vice versa. Thus it seems to me probable that
each of these pressures must have a term in the function k which corre-
sponds to its tendency to react, and this individual tendency we shall
call In 2, — a term which will be in the numerator when the correspond-

RICHARDS. ENERGY OF PHYSICO-CHEMICAL REACTION. 477

ing pressure is in the denomiuator, and vice versa. The reasoning may
be stated in the form of an equation : —

rt «i rt' «'i ■r >h 9-' "'2

This equation is simply a definition of the vahies z, which are made
logarithmic from analogy to the other pressures. One must bear in
mind, however, that this z, or •' physico-chemical potential," need not be
the reciprocal of p ; for the condition of equilibrium demands only that the
total sum of the logarithms on each side must be equal, and not that the
individual opposites are immediate functions of one another. The p
values depend of course upon the amount of substance present ; while
the z values are constant for any given temperature, because they are
by definition the constant factors of a constant. Any constant tendencies,
not given by the differential equation, may hence be included among the
z values.

Transposing the second member, we obtain

In ;^i-;>i;---^i--i ••• ^ o. (7)

This expression may be written
«i In i^pi 2i) + n\ 1q {p'\ z'l) — «2 111 {p^'^i) — ^'2 lo (,p'2. ^'2) . . . = 0.

Except for its logarithmic form and the substitution of pressure for
mass, this equation reminds one of Berthollet's old statement concerning
chemical action. It is a fundamental equation of chemical equilibrium in
dilute or ideal mixtures.

Stated in words, the equation reads : Each molecule taking part in a
reaction may be said to possess at any given temperature a reacting ten-
dency which is the logarithm of the product of its constant physico-chemical
potential and variable observed pressure. Obviously the logarithmic
arrangement is so convenient as almost to demand its adoption, although
the same idea might have been expressed otherwise. One may say, for
example, that in equilibrium the algebraic sum of the opposing energies
concerned is zero, — almost an axiom. The logarithmic equation is a
plausible hypothesis which is concordant with the well known Nernst
equation, and with many other natural tendencies.

The substitution of the new value for k instead of its pressure equiva-
lent in the equation gives us a less inverted view of the theorem of
Maupertuis : —

478 PROCEEDINGS OP THE AMERICAN ACADEMY.

d\n

z^H z\^'y ... - Q

dT RT'

It will be seen at once that the sum of the reacting tendencies of the
initial substances In z^\ In 2:'i"'i, etc., agree in sign with Q ; that is, a
reaction which absorbs heat will have a greater tendency to take place as
the temperature rises.

Besides comprehending the partial pressures within a single phase, and
the " physico-chemical potentials " of the various substances which are
in it, the pressure equations may be applied to the relations between two
2)hases ; since the equation from which they were derived is true for ideal
heterogeneous equilibrium. We may then introduce into the equation
phase-pressures, such as vapor or solution tensions, which are exerted
in opposition to gas or osmotic pressures respectively. Such phase-
pressures will be represented, for the sake of convenience, by the usual
designation P.

A few simple physical examples may now be cited, to show that the
point of view is not inconsistent with the facts in these cases. For
example, in the case of an evaporating liquid, we may say that the force
concerned is the vapor tension, which therefore takes the place of z-^. It
is placed in the denominator because the process absorbs heat. There
are no opposing forces except the constant gas pressure against which
the evaporation proceeds \ hence the equilibrium equation becomes

lo^=0,

— an obvious truth ; and the reactiou-metatherm reduces at once to the
equation of Clausius.

A more complicated case is the evaporation of water from a crystal-
lized salt ; but the treatment is equally simple. The water h'as a certain
reacting tendency or driving energy which manifests itself as a definite
pressure at each temperature. Precisely the same explanation applies
to calcic carbonate, or any other similar case. The simplicity of the
conception of this reacting tendency seems to commend itself as an
improvement over the usual involved treatment, since it does not demand
that all the solid should be vaporized in order to effect the change.

A still more complicated case is the evaporation of ammonic sulphy-
drate. Here we have to consider the physical tendency of the evaporat-
ing substance, which is balanced by the pressure of the ammonic sulphy-

RICHARDS. — ENERGY OF rnYSICO-CHEMICAL REACTION. 479

drate, as well as the physico-chemical potentials and actual pressures of
the three molecular species. The equation of equilibrium would then be

laiHlM^li^i^ = 0.

This expression is an analysis of the forces which are at work in
maintaining equilibrium, p^, the pressure of the uudissociated NH4SIT,
occurs first in the numerator as opposing P, the sublimation tension of
the solid substance, and then in the denominator as a factor of the chem-
ical reacting tendency of the ammonic sulphydrate. These two of course
cancel, and since at constant temperature P and all the z values are
constant, we obtain for the chemical part of the equilibrium the familiar
expression

p^p'z — k' .

This is entirely in accord with the results of Isambert, and is undoubtedly
true.

Many other examples of the application of the pressure equations
might be given. For example, the dissociation of nitric peroxide is
easily treated by the reaction-isobar, while that of hydriodic acid needs
merely the simplest form of all, which is both isochor and isobar.
Moreover, it is obvious that upon introducing electrolytic solution ten-
sions and osmotic pressures into the reaction isobar, it becomes at once
the well known expression for the temperature coefficient of electro-
motive force. In short, whenever pressures are under consideration, one
of the two special forms of the equations above affords a ready means
of treatment.

The results of this paper may be summarized as follows : —

1. It is pointed out that the study of pressure affords a more direct
method of analyzing the progress of a reaction than the study of volume,
concentration, or entropy.

2. An expression called the " reaction-metatherm " has been evolved
which represents in terms of pressure the temperature coefficient of the
equilibrium ratio of ideal physico-chemical reaction.

3. Since this equation always contains in its second member the heat
actually evolved in the reaction, whether under constant volume or con-
stant pressure, it is a mathematical expression of the Theorem of
Maupertuis or Le Chatelier.

480 PROCEEDINGS OP THE AMERICAN ACADEMY.

4. Moreover, since the equation represents this fundamental theorem,
its logarithmic quantity must represent the variables in the reacting
tendency, or the driving energy of reaction.

5. When analyzed, this equation shows that the part played by each
substance in a reaction may be considered as the logarithm of the product
of its " physico-chemical potential " and its actually present pressure.

G. The reaction metatherm may be simplified into a reaction-isobar
and a reaction-isochor, according as the pressure or volume is kept
constant during the reaction.

7. While the reaction-isobar, expressed in terras of pressure, offers the
most convenient basis for the calculation of the cases to which it is
applicable, results obtained under constant volume are more conveniently
calculated if the reacting substances are expressed in terms of concentra-
tion, according to the equation of vau't Hoff.

Harvard University, April 3, 1900.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 24. — May, 1900.

SUPPLEMENTARY NOTE ON THE CHIEF THEOREM

OF LIE'S THEORY OF FINITE

CONTINUOUS GROUPS.

By Stephen Elmer Slocum.

SUPPLEMENTARY NOTE ON THE CHIEF THEOREM OF
LIE'S THEORY OF FINITE CONTINUOUS GROUPS.

By Stephen Elmer Slocum, Clark University.

Presented by Henry Taber, April 11, 1900.

On pages 239-250 of the current volume of these Proceedings, in a
paper entitled " Note on the chief theorem of Lie's theory of continuous
groups," I pointed out an error in Lie's demonstration of the first funda-
mental theorem of his theory. In what follows I indicate how this error
may be avoided and the demonstration completed.

Lie's error in the demonstration of the first fundamental theorem con-
sists in neglecting conditions imposed at the outset upon certain auxiliary
quantities jUi, /Uo, • • . introduced in the course of the demonstration. Thus
in the Continuierliche Gruppen, pp. 372-376 (and substantially in Trans-
formationsgruppen, vol. HI., pp. 558-564) Lie proceeds as follows: —

Being given a family with an oo'' of transformations T^i defined by
the equations

^'> =/. {^x- • • ^n^ «! . . . «,) (/ = 1, 2 . . . n),

containing the identical transformation, and, moreover, such that the a;"s
satisfy a certain system of differential equations, he defines by the intro-
duction of new parameters /x a family of transformations B^,

x'i = Fi (x'l . . . a;',,, fii. . . /J.,) (^■ = 1, 2 . . . n),

each of which is generated by an infinitesimal transformation ; Lie then
establishes the symbolic equation

* If the equations defining: the families of transformations Ta and Efx are re-
spectively,

■r'i =fi (Ti . . . Xn, rti . . . a,.) (/ = 1, 2 . . . n),
and

T'i = Fi (x\ . . . x'n, fj-i . . . /A,.) (/ = 1,2 ... n),

the symbolic equation Ta E/j. = Ta is equivalent to the simultaneous system of
equations

484 PROCEEDINGS OF THE AMERICAN ACADEMY,

where the a's and /u's are arbitrary, and

«* = ^fc (l"i . • • /"r, «i • . • a,) {k = 'i, 2 . . . r),

the <I>'s being independent functions of the fx's.

For a^ — a^"" (k = 1, 2 . . . r), the transformation Ta becomes the iden-
tical transformation ; and therefore we have

£J^ = T-o^ E, - Ta, *
where

Thus every transformation of the family ^^ is a transformation of the
family T„. If, conversely, every transformation T„ belonged to the
family E^, it would follow that

Ta Ta = Ta,f

that is to say, we should have shown that the family of transformations
Ta forms a group.

But, although the 4>'s are independent functions of the /x's, nevertheless
the /x's in certain cases become infinite for certain systems of values of
the a's ; and infinite values of the /x's, by their definition, are excluded at
the outset, t We cannot then assume that every transformation T„
belongs to the family E^.

We may, however, proceed as follows : — For all values of the a's for
which the functions

X'i =fi{Xi . . . Xn, «i . . . a,.),

X'i = Fi (x\ . . . x'r,, /Xi . . . /x,.), (/ = 1, 2 . . . «)

x'i —fi (.Ti . . . .r„, (7, . . . a,),
or to the functional equations

Fi (/i i^', «) ■ ■ -fn (■'^. «)> A*! • • . Mr) =/ (ti . . . .r„, rtj . . . a,-) {i-\,1 . . . n).

* That is,

Fi {x\ . . . x„, Ati . . . fir) = Fi{f {.r, a'"') . . . /„ (,r, o(O'), ^j . . . /jl,-) =fi(xx . ■ . r„, a^ . . . a,.)

0' = 1, 2 . . . n),
since !

^' =fi (^-1 ■ ■ ■ ^n, «! <»'... a,(0 ) (/ = 1, 2 . . . n). I

t That is, I

fi (/i {^> <') ■ • ■/' (■'■. "), oi . . . a,.) =/(ti . . . T„,a-x . . . Or) (/ = 1, 2 . . . n).
I Tliese Proccedinjrs, p. 217.

SLOCUM. — FINITE CONTINUOUS GROUPS. 485

^i. = Mj ((xi ... a,., oi"" . . . «,."") (J — \, 2 ... r) are finite, we have

that is,
/.• (/l {^, «)•••/. (^J «) , ai . . . a,) =

^i (/l (^> f~<) ■ ■ -fn (a^i «), ii\ ■ • • f^r) =fi(Xl'-- x,n «i- • • «,)
(/= 1, 2 . . . 7l).

Let y3i, ^82 ... be a system of values of the as for which one, or more,
of the corresponding /x's is infinite in all branches. Also let bi, 5o . . .
be the system of values assumed by the a's for a^ = yS^ (^' = 1, 2 . . . ?•).
Since the functions f are continuous functions of the variables and
parameters, and we assume that the system of parameters (3 give a defi-
nite transformation Tp of the family, we have

/■(/i(-^» «)•••/, {^, a), A • • • A) = lim./i (/i (a:-, ~a) . . ./, (.r, a), f^i . . . a,)
= lim./^ (Xi . . . a:„, Oi . . . a,) = ft {x^. . . x„, by . . . b,) (i =1,2... n).

which is equivalent to the symbolic equation

Ta 7> = Ta lim. 7; = lim. 7', ^ = lim. Ta = Ti,.

Consequently, the composition of two arbitrary transformations Ta and
T^ of the family is equivalent to a transformation 7\ of this family ;
that is to say, the family of transformations T,, forms a group. The
transformation T^, however, may not be a transformation of the group
that can be generated by an infinitesimal transformation of this group.
Thus, every transformation of a group with continuous parameters is uot
necessarily generated by an infinitesimal transformation of the group.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. Xo. 25. — Juxe, 1900.

CONTRIBUTIONS FROM TPIE GRAY HERBARIUM OF
HARVARD UNIVERSITY.

New Series. — No. XIX.

Bv M. L, Fernald.

I. A Synopsis of the Mexican and Central American Species of
Salvia.

II. A Revision of the Mexican and Central American Solanums of
the Subsection Torvaria.

III. Some undescribed Mexican Phanerogams, chiefly Labiatae and
,1^^ Solanaceae.

MARINE BIOLOGICAL LABORATORY.

Received

Accession No

o

Given by ^^
Place,

/(>.,J?.M

'?^L

* *

"Mn 'WnnO n** -nn .-v^ h V. I a*. :» \ 1- ,

CONTRIBUTIONS FROM THE GRAY HERBARIUM OF HARVARD
UNIVERSITY, NEW SERIES, NO. XIX.

By M. L, Fernald.

Presented by B. L. Robinson, March 14, 1900. Received April 19, 1900.

I.— A SYNOPSIS OF THE MEXICAN AND CENTRAL
AMERICAN SPECIES OF SALVIA.

In his monograph of the Labiatae in De Caudolle's Prodromus, Ben-
tham (1848) recognized 118 species of Salvia in Mexico and Central
America, several of them, however, being known only from the original
descriptions.

In the botanical portion of Biologia Central i- Americana, Mr. Hemsley
enumerated 135 species, but a number of them were unknown to him,
seven have proved to be identical with others there listed, and the records
of two are based upon their occurrence in adjacent portions of the
United States but not in Mexico proper. Thus excluding these nine
species, there were recorded by Mr. Hemsley, in 1887, 126 Mexican
and Central American Salvias.

During the past quarter century the unprecedented activity in the
botanical exploration of those countries has brought together in Salvia,
as in many other large genera, an abundance of material for study.
Many of these recent collections have been critically examined, and a
large number of species based upon them have been described. The
collections of the past decade, furthermore, contain very many unique
plants which cannot be referred to any of the species already published.
These species, together with those described in the last half-century,
since the publication of Bentham's treatment of the group in the Pro-
dromus, increase the number of known Mexican and Central American
Salvias by nearly one hundred.

In order to place before students of Mexican botany many previously
undescribed forms and to show more clearly than could otherwise be
done their affinities, a synopsis has been prepared of all the known
Mexican and Central American species. In the preparation of this
work the general divisions of Bentham have been adopted, though with

490 PROCEEDINGS OF THE AMERICAN ACADEMY.

more material at hand it has been necessary in many cases to amplify
or alter tlie limits of groups as defined by him.

It was hoped that tlie treatment of the genus published by M. le
Professeur Jean Briquet in " Engler & Prantl's Natlirlicheu Planzen-
familien " ^ would be of assistance in preparing this synopsis ; but, except
for the introduction of somewhat helpful minor divisions of the groups,-
that work adds little to the earlier conclusions of Bentham. In fact, so
far at least as the Mexican species are concerned, Professor Briquet's
translations of Bentham's sectional and subsectional diagnoses are most
unfortunate, often so far so as quite to contradict the true characters of
the plants he is supposed to be describing, and entirely to mislead the
student who attempts to follow his synopsis. In the description of the
very first group, the § Micranthae, for example, Bentham says : " Corolkx
vix S-Iiuearis calyce dimidio vel rarius suhduplo longior^^ (corolla about
3 lines long, once and a half or rarely almost twice as long as the calyx),
proportions which are maintained almost without exception by the
species of that section. Yet this is rendered by Briquet " Blkr. [Blumen-
krone] klein, die Hdlfte der Ldnge des Kelches erreichend, seltener 2 mal
grosser " (corolla small, half the length of the calyx, rarely twice longer)
although the species which constitute the section have the corolla as
defined by Bentham. Briquet's description of the § Microsphaceae, in-
cluded by Bentham in the Prodromus under § Micranthae, reads : " Blkr.
klein, ^-aMm die Hdlfte der Ldnge des Kelches erreichend" (corolla small,
scarcely half as long as the calyx), thus suggesting plants with the calyx
definitely exceeding the corolla, instead of the species, enumerated by
him, with the corolla distinctly exceeding the calyx. Again in the
§ Brachyanthae Bentham describes the lower lip of the corolla as fol-
lows : " labium patens . . . galea longius " (lip spreading . . . longer
than the galea), while Briquet, rendering it into German says "...
ausgebreiteter Unterlippe, diese nicht liinger als die Oberlippe " (. . .
the spreading lower lip, this not longer than the upper lip [galea]), thus
absolutely contradicting the character of the corolla as shown by the
species included by him in the section.

Of the 217 Salvias recognized in the present paper, specimens — or in
seven cases merely authentic plates — of 174 species have been ex-
amined. Of the remaining 43 species very many, although well
described, were unknown to Bentham and have not been identified with
recently collected material. Others recognized by Bentham as of doubt-

1 Engl. & Prantl, Nat. Pflanzenf. iv. Ab. 3, 270-286.

FERNALD. — MEXICAN SALVIAS. 491

ful status are here noted, although they are so incompletely character-
ized as to make their identification without access to the types quite
impossible. In such cases the traditional conception of the plant has
been maintained as far as possible, though it is highly probable that
future study of these little-known types will identify some of them with
better known species.

The descriptions of two species recently published from old manu-
scripts contain so little of specific significance that it is impossible to
say upon what plants they were based. These are S. azurea and S.
dichroma, La Llave in La Naturaleza, vii (1885) Apend. 82.

A European species, »S'. Sclarea, L., is often cultivated in central
Mexico and is sometimes distributed in exsiccatae as if an indigenous
plant (for example, see SchaSfner's no. 49 from the mountains of San
Luis Potosi).

In the study of species of § Membranaceae Mr. N. E. Brown of the
Royal Gardens at Kew has rendered very valuable service by comparing
specimens submitted to him with the types of Bentham's species.

SYNOPSIS OF SPECIES.

(As far as possible the sectional numbers and descriptions have been
maintained as given in the Prodromus.)

Section VII. CALOSPHACE, Benth. Calyx ovate, tubulose or
campanulate, the upper lip entire or shortly tridentate, the teeth approxi-
mate, and the middle one longest. Corolla-tube exserted or included,
not annulate within, but occasionally furnished with 2 teeth near the
base. Upper lip (galea) straight or concave, entire or often short-
emarginate; the lower with spreading lobes. Anterior portion of the
connective deflexed, linear, longitudinally connate or approximate, oc-
casionally a little dilated, and rarely bearing an empty adnate anther-cell.

§ 1. MicuANTHAE, Benth. Bracts small, mostly deciduous. Corolla
blue or white, short, 8 mm. or less (in one species nearly 1 cm.) in
length, very slightly or rarely almost twice exceeding the calyx; the
tube generally ventricose ; the galea straight.

* Corolla very small, 3 to 5 mm. long : calyx glandular.

H- Calyx-lobes blunt.

1. S. occiDENTALis, Swartz, Fl. Ind. Occ. i. 43; Benth. I.e. 296;
Gray, Syn. Fl. ii. pt. 1, 370 ; Hemsl. 1. c. 562 ; Briq. 1. c. 277. S. pro-
cumhens, Ruiz & Pav. Fl. Per. & Chil. i. 27, t. 39, fig. a. S. radicans,
Poir. Diet. vi. 621. Verbena minima chamaedryos folio, Sloane, Jam. i.

492 PROCEEDINGS OF THE AMERICAN ACADEMY.

172, t. 107. Hyptis glandulosa, Sieb. Fl. Mart. Exs. no. 151, fide Benth.

1. c. — Common in tropical and subtroj^ical America, extending north to

Vera Cruz. Vera Cruz, Mt. Orizaba {Botteri, no. 394) ; Cordoba {Asa

Gray) : Yucatan, waste ground near Merida, April, 1887 {Porjirio

Valdez, no. 55) : Guatemala, Cobau, Dept. Alta Verapaz, alt. 1,320 m.,

Nov. 1886 (H. von Tuerckheim in exsicc. J. D. Smith, no. 1090); Santa

Rosa, alt. 770 to 920 m., May, Dec, 1892 {Jleyde & Lux in exsicc.

J. D. Smith, nos. 3014, 4399) : Honduras, San Pedro Sula, Dept.

Santa Barbara, alt. 250 m., May, 1888 (C. Thieme in exsicc. J. D.

Smith, no. 5422) : Nicaragua {Charles Wright) : Costa Rica, Ujaras

{Oersted) ; Cartago, alt. 1,300 m., Dec, 1887 {Juan J. Cooper) in exsicc

J. D. Smith, no. 5901 ; roadsides San Jose, Jan., 1893 {Ad. Tonduz, no.

7280).

■I- +- Calyx-lobes subulate-mucronate.

*+ Bracts deciduous.

2. S. MiSELLA, HBK. 1. c 290 ; Benth. 1. c 297 ; Hemsl. 1. c 561 ;

Briq. 1. c. — Described from near Acapulco : perhaps the same as the

next.

** *-*■ Bracts persistent.

3. S. obscura, Benth. Lab. 245, & in DC. 1. c'297; Millspaugh,
Field Columb. Mus. Bot. ser. i. 43. S.-privoides, Gray, Syn. Fl. I.e. 371,
not Benth. S. occidentah's, Millsp. 1. c, in part, not Swartz. S. occi-
dentalis, var. ? Garberi, Chapm. Bot. Gaz. iii. 10. — Range similar to
that of S. occidentalis, extending north to Sinaloa and Lower California.
Lower California, La Paz, 1890 {Edw. Palmer, no. 10): Sinaloa,
Mazatlan, Dec, 1894 {F. H. Lamb, nos. 311, 317) : Guerrero, Aca-
pulco, 1895 {Edw. Palmer, no. 244) : Vera Cruz, Valley of Cordova,
Dec. 18, 1865 {Bourgeau, no. 1504): Yucatan, roadside, near Izamal,
Jan. 14, 1895 {Millspaugh, no. 90) ; Island of Cozumel, 1895 {G. F.

Gaumer, no. 394).

* * Corolla 6 to 8 ram. long.

•(- Calyx glandular.

++ Bracts caducous.

= Stem villous : leaves ovate-lanceolate, rufous-tomentose beneath : verticels
4-6-flowered : calyx with obtuse lobes.

4. S. iNCONSPtCUA, Benth. Lab. 247, & in DC. 1. c 298; Hemsl.

1. c 558 ; Briq. 1. c. 278. — Described from Mexico. Not seen.

= = Stem pubescent with long spreading gland-tipped hairs : leaves broad-ovate,
slightly puberulent beneath: verticels 2-flovvered : calyx *with acuminate lobes.

5. S. PODADENA, Briq. Ann. Conserv. Jard. Bot. Geneve, ii. 131. —
Described from Oaxaca. Not seen.

FERNALD. — MEXICAN SALVIAS. 493

** -M- Bracts persistent.
= Leaves villous or long-pilose beneath.

a. Leaves canescent on both faces, I to 3 cm. long, on slender naked petioles : calyx-

lobes acutish.

6. S. SEROTINA, L. Mant. 25 ; Jacq. Ic. Rar. i. 1, t. 3 ; Benth. 1. c;
Hemsl. 1. c. iv. 107 ; Briq. in Engl. & Prantl, 1. c. .S'. dominica, Vahl.
Euum. i. 233; Swartz, Obs. 18, t. 1. fig. 1, not L, — Florida and the
West Indies. Reported from Cozumel Island, Yucatan.

b. Leaves ferrugineous-villous beneath, 1 dm. or less long, cuneate to winged

petioles : calyx-lobes setaceous-mucronate.

7. S. PRivoiDES, Benth. Bot. Sulph. 150, & in DC. I.e. 297 ; Hemsl.
1. c. ii. 563 ; Briq. 1. c. — Southern Mexico and Central America.
Jalisco, Guadalajara, Sept., 1886 {Edio. Palmer, no, 498) ; cool shady-
places, barranca near Guadalajara, Nov. 5, 1888 (C. G. Pringle, no.
1727) ; San Sebastian, alt. 1,185 to 1,540 m., March 16, 1897 {E. W. Nel-
son, no. 4072) : Morelos, Cuernavaca, Nov. 14, 1865 {Bourgeau, no.
1239) : Oaxaca, Jayacatlan, alt. 1,320 m., Sept. 10, 1894 (Z. C
Smith, no. 174).

= = Leaves short-pilose or glabrate beneath.

a. Inflorescence an oblong raceme, 4 to 7 cm. long, not secund : calyx tubulose-

campanulate, in fruit 7 mm. long.

8. S. micrantha, Vahl. Enum. i. 235 ; Benth. 1. c. 298 ; Hemsl.

1. c. 561 ; Briq. 1. c. S. bulla9a, Ort. Dec. ix. 109 ; Jacq. Hort.

Schoenb. iv. 41, t. 48L S. serotina, Vahl. 1. c. 232, not L. S. occi-

dentalis, Millsp. 1. c. in part, not Swartz. — Tropical America, in Mexico

known only from the Yucatan coast. Yucatan, streets of Dolores,

Island of Mugeres, Jan. 1, 1895 (^MiUspaugh, no. 18) ; without locality,

1896 {Porjirio Valdez, no. 31). S. orhicidaris, Benth. Bot. Sulph.

151, from Panama seems to differ from the species only in its fruticose

base.

b. Inflorescence secund, loosely-flowered.

1. Calyx-tube bearing long straight spreading setiform glandless hairs mixed with
the sliorter gland-tipped ones ; upper lip entire : racemes simple, elongated, be-
coming 1 to 1.5 dm. long.

9. S. setosa. Annual, 2 to 5 dm. high ; the stems slightly appressed-
retrorse-setulose on the angles : leaves ovate or rhombic-ovate, thin,
acute or blunt at tip, cuneate at base to winged petioles, crenate-serrate,
8 cm. or less in length, sparingly appressed-setulose above, paler and
minutely puberulent or glabrate beneath : racemes stiff ; verticels mostly
2-flowered, all becoming remote, the lowest 1.5 to 2.5 cm. apart: bracts

494 PROCEEDINGS OP THE AMERICAN ACADEMY.

lance-subulate, 5 to 7 mm. long : pedicels 2 to 4 mm. long : calyx in
anthesis 5 mtn., in fruit 9 mm. long; the lobes equalling the tube,
narrow-ovate, with long subulate aristiform tips, the pubescence less
spreading than on the tube : corolla 7 to 8 mm. long ; the sparingly
pilose lips deep blue. — S. privoides, Gray in Wats. Proc. Am. Acad.
xxi. 435; Rose, Contrib. U. S. Nat, Herb. i. 110; not Benth. — North-
western Mexico. Chihuahua, Hacienda San Jose, 1885 {Edw. Palmer,
no. 64) : Sonora, by shaded water-courses and in canons. Alamos, Sept.,
1890 {Edw. Palmer, no. 680, 681).

2. Calyx pubescent with gland-tipped hairs ; upper lip generally tridentate :
racemes paniculate, 5 cm. or less in length.

10. S. lateriflora. Bushy-branched annual, 2.5 to 3 dm. high: stems
sparingly retrorse-pilose: leaves ovate or rhombic-ovate, thin, blunt or
acutish at tip, cuneate to slender petioles 5 cm. or less long, coarsely
crenate-dentate, minutely setulose on both faces or glabrate : flowers
solitary or in 2's, all remote, the lowest 0.7 to 1 cm. apart: bracts ovate,
acuminate, 1 to 2 mm. long : pedicels 2 to 4 mm. long : calyx in
anthesis 2 to 3 mm., in fruit 5 to 6 mm. long ; the tube twice exceeding
the broad blunt subulate-tipped lobes. — Sonora, about abandoned gar-
dens, Guaymas, Oct., 1887 {Edw. Palmer, no. 320). Habitally re-
sembling Scutellaria lateriflora. A fragmentary specimen collected by
Xantus at Cape St. Lucas, Lower California, may belong here.

H- -t- Calyx not glandular.

*+ Leaves thin, membranaceous, the primary ones 4 to 9 cm. long : verticels 6-

many-flowered.

11. S. TILIAEFOLIA, Vahl. Leaves broad- or rhombic-ovate, cordate
truncate or rounded-cuneate at base, sparingly pubescent on both faces,
dark green above. — Symb. iii. 7 ; Jacq. Hort. Schoenb. iii. 2. t. 254;
Benth. I.e. 299; Hemsl. 1. c. 566. S. flmhriata, HBK. I.e. 299, t.
149. — Common in tropical America, extending northward through
Mexico. Chihuahua, Valley of Chihuahua, Sept. 17, 1885 (0. G.
Pringle, no. 556; damp places, Caiion de Pilares, Sept. 22, 1891 (C. V.
Hartman, no. 749) : Coahuila, Saltillo, 1848-49 {J. Gregg, no. 542) ;
Soldad, Sept., 1880 and in shaded places, Saltillo, Sept., 1880, 1898
(Ediv. Palmer, nos. 1062, 335) : Durango, abundant in shade near
Durango, Sept., 1896 (Edw. Palmer, no. 572): San Luis Potosi, in
deep shade near the city, 1876 (Sehajfner, no. 674), alt. 1,850 to 2,460
m., 1878 (Parry & Palmer, nos. 743, 746^) : Aguas Calientes (Hart-
weg, no. 159) : Mexico, Valley of Mexico, May 5, 1865 (Bourgeau, no.

FERNALD. — MEXICAN SALVIAS. 495

122) : Vera Cruz, Orizaba (Botteri, nos. 533, 869), Aug. 13, 1891

(Seaton. no. 304) ; Cordoba (Asa Gray) : Oaxaca, Moute Alban, alt.

1,900 m., Oct. 11, 1895 (L. C. Smith, no. 949) : Guatemala, Santiago,

Dept. Zacatepequez, alt. 2,000 m., 1891 (Rosalia Gomez in exsicc. J. D.

Smith, no. 81 G) ; Castillas, Dept. Santa Rosa, alt. 1,230 m., Sept.,

1892 (Heyde & Lux in exsicc. J. D. Smith, no. 4055) : Costa Rica,

Cartago, alt. 1,300 m., Nov., 1888 (Juan J. Cooper in exsicc. J. D. Smith,

no. 5900) ; San Jose, July, 1892, and San Francisco de Guadalupe,

Dec, 1893 (Ad. Tonduz, nos. 701, 8456). See note under S. Chia. S.

flexuosa^ Presl in Benth. Lab. 248, is perhaps only a glabrate form of

this species, apparently represented by Botteri's no. 395 from Orizaba.

Var. cinerascens. Whole plant cinereous with fine puberulence. —

Jalisco, barranca near Guadalajara, Oct. 3, 1891 (G. G. Pringle, no.

5176).

++ ++ Leaves firm, small, 1 to 2.5 cm. long : verticels 2-4-flowered.

= Herbaceous, erect : leaves broad-ovate, cordate, minutely pubescent : lower
verticels in axils of upper foliar leaves.

12. S. HUMiLTS, Benth. Lab. 247, & in DC. 1. c. 298; Hemsl. 1. c.
558 ; Briq. 1. c. — Described from " Mexico." Not seen.

= = Diffusely branching from woody base : leaves narrowly rhombic-ovate, or
oblong-ovate, cuneate, densely pilose-setulose beneath : racemes short-peil uncled.

13. S. pusilla. Stems 1 to 2.5 dm. long, slender and wiry, puberu-
lent and hoary with fine spreading setulose hairs ; more or less diffusely
branched : leaves blunt or acutish, irregularly more or less serrate,
green above, pale beneath, long-setulose on both faces or glabrate above,
on short slender petioles: peduncles 1.5 to 4 cm. long: racemes 1 to 8
cm. long; the verticels all remote, the lowest becoming 2 to 2.5 cm.
apart : bracts ovate, 3 to 4 mm. long, firm and essentially persistent :
pedicels very short : calyx white-setulose, in anthesis 2 to 3 mm., in
fruit 5 mm. long ; the tube once and a half longer than the acuminate
lobes : corolla blue, 7 mm. long ; the tube short-exserted ; the pilose
galea twice exceeded by the lip: style densely bearded. — Oaxaca,
vicinity of Yalalag, alt. 1,230 to 2,400 m., July, 1894 (E. W. Nelson,
no. 958); Sierra de San Felipe, alt. 2,160 m., Sept. 23, 1895 (C.
Conzatti, in exsicc. L. C. Smith, no. 708).

* * * Corolla 9 to 10 mm. long.

-1- Leaves cuneate or rounded-cuneate at base : calyx subglabrous.

++ Leaves oblong-lanceolate.

14. S. OUROPHYLLA, Briq. Ann. Conserv. Jard. Bot. Geneve, ii.
126. — Described from Costa Rica. Not seen.

496 PROCEEDINGS OP THE AMERICAN ACADEMY.

++ -M- Leaves ovate or ovate-elliptic.

15. S. PERMiXTA, Briq. 1. c. 129. — Described from Costa Rica.

Not seen.

t- -t- Leaves hastate at base : calyx glandular.

16. S. COSTARICENSIS, Oersted ic Kjoeb. Vidensk. Meddel. (1853)
39; Hemsl. I. c. 555. — Costa Rica, Cartago, alt. 1,300 m,, Dec,
1887 {Juan J. Cooper in exsicc. J. D. Smith, no. 5895) ; San Jose,
Oct., 1892 {Ad, Tonduz, no. 7158).

§ 2. Membranaceae, Benth. Flowers as in the Micranthae or a
little larger. Bracts suborbiculate, persistent, membranaceous, veiny,
equalling or exceeding the calyx,

* Leaves narrow-ovate, cuneate or narrowed at base.

■\- Calyx glabrous, subtruncate, with very short teeth : verticals approximate,
forming a spiciform raceme 4 to 8 cm. long: bracts and calyces blue or roseate.

17. S. LOPHANTHA, Benth. in DC. 1. c. 301 ; Hemsl. 1. c. 560, in
part ; Briq. in Engl. & Prantl, 1. c. — Guatemala, Santa Rosa, alt.
1,230 m., Oct., 1892 {Heyde & Lux in exsicc. J. D. Smith, no. 4051).

-I- H- Calyx pubescent.

■w Calyx canescently short-pilose, not viscid : verticels remote, forming a raceme

1 to 2.5 dm. long.

18. S. MOCiNOi, Benth. Lab. 271, & in DC. 1. c. 300; Hemsl. 1. c.
561 ; Briq. 1. c. — Guatemala, Laguna de Ayarza, Dept. Jalapa, alt.
2,460 m., Sept., 1892 (Bei/de & Lux in exsicc. J. D. Smith, no. 4048).

■w -M. Calyx viscid, short-pilose.

= Branches pilose ; the hairs spreading.

a. Racemes simple, or very slightly branched ; verticels tending to become remote;
the primary racemes becoming 0.8 to 2 dm. long : lowest bracts crenate-serrate.

19. S. RUBiGiNOSA, Benth. 1. c. 301. Mature leaves pubescent
beneath only on the nerves. — Hemsl. 1. c. 565 ; Briq. 1. c. S. lophantha,
Donnell Smith, Enura. PI. Guat. ii. 62, not Benth. S. Mocinoi, Donnell
Smith, 1. c. iv. 125, 188, in part, not Benth. — Southern Mexico and
Central America. Guatemala, Pinula, alt. 1,350 m., Feb., 1890 (/. D.
Smith, no. 1911) ; Chiapas, Dept. Santa Rosa, alt. 1,080 m., Dec, 1892
(Heyde & Lux in exsicc. J. D. Smith, no. 4400).

Var. hebephylla. Leaves velutinous beneath. — S. lophantha,
Hemsl. 1. c. 560, in part, not Benth. — Vera Cruz, region of Orizaba,
Oct. 11, 1866 (Bourgeau, no. 3215): Chiapas, among the mountains
(Ghieshreght, no. 745) : Guatemala, Volcan Fuego, Zacatepequez, alt.
1,540 m., March, 1892 (J. D. Smith, no. 2597).

FERNALD. — MEXICAN SALVIAS. 497

b. Inflorescences paniculate ; all the verticels remote : bracts entire.

20. S. cladodes. Stems glabrate below, sordid-pilose above : leaves
oblong to uarrow-ovate, 1 dm. or less long, 5 cm. or less wide, long-
aciimiuate at tip, cuneate at base, finely crenate-serrate, short-velutinous
or glabrate on both surfaces; petioles 2 cm. or less long: panicle with
slender ascending densely pilose branches 1 to 2 dm. long : bracts purp-
lish, broad-ovate or suborbicular, acuminate, more or less pilose, ciliate :
verticels 1 to 2.5 cm. apart, 3-9-flowered ; pedicels 4 to 7 mm. long,
spreading and nodding at tips : calyx purplish, in anthesis 8 to 9 mm.
long, broadened upward ; the tube glandular-pilose ; the glabrate limb
with ovate-acuminate lobes : corolla-tube iiacluded ; the lips one half
longer than the calyx. — Northwestern Mexico. Without locality (See-
mann) : Tepic, near Compostela, alt. 1,540 to 1,850 m., Apr. 7-8, 1897
(B. W. Nelson, no. 4171).

= = Branches canescent with appressed strongly recurved hairs : only the lowest
verticels remote, the spiciform racemes 1 to 6 cm. long : bracts entire or un-
dulate, ciliate.

21. S. saltuensis. Stems 1 to 1.5 m. high, bearing solitary ter-
minal racemes or many inflorescences of 1 to 3 verticels on short leafy
branchlets : leaves 6 cm. or less long, acute, finely crenate-serrate, dull
green and finely pubescent above, canescent-tomentulose beneath, on
slender petioles 1.5 cm. or less in length: bracts reniform, acuminate,
brownish or slightly rosy tinged, minutely pubescent on the nerves, 1 to
1.5 cm. long: calyx purple-tinged, appressed-hirsute, in anthesis 7 mm.
long, with broad blunt or short-acuminate lobes : corolla 1.3 cm. lons^,
blue, the pilose galea one half as long as the lip. — Morelos, in woods
of Sierra de Tepoxtlan, alt. 2,310 m., Feb. 8, 1899 {G. G. Pringle, no.
8035) : Jalisco, between San Sebastian and the summit of Mt. Bufa
de Mascota, alt. 1,380 to 2,300 m., March 20, 1897 {E. W. Nelsoii, no.

4103).

= = = Of close affinity to the preceding is

22. S. BUPLEUROiDES, Presl in Benth. Lab. 271, a glabrous plant
with fascicled peduncles each bearing solitary many-flowered verticels.
Not identified.

* * Leaves broad-ovate (narrow-ovate in .S'. nitidd) or rhombic-ovate, rounded
to the subcordate truncate or subcuneate base.

t- Verticels remote (appaoximate in a form of S. kyptoides) : leaves 5 cm. or less

in length.

VOL. XXXV. — 32

Leaves glabrous or only minutely puberulent.
= Leaves ovate-lanceolate, glossy.

498 PROCEEDINGS OF THE AMERICAN ACADEMY.

23. S. NiTiDA, Benth. in DC. 1. c. 300 ; Herasl. 1. c. 562 ; Briq. 1, c.
Hyptis nitida, Mart. & Gal. Bull. Acad. Brux. xi. pt. 2, 189. — De-
scribed from Oaxaca. Not seen.

= = Leaves rhombic-ovate, dull.

24. S. galinsogifolia. Stems 1.5 to 6 dm. high, branching, minutely
puberulent, the young parts canescent : leaves 5 cm. or less in length,
blunt or acutish, minutely and sparingly puberulent or glabrous, the
lowest shorter than the slender petioles: peduncles 1.5 dm. or less in
length; verticels from 1 to 5, the lowest becoming 3 to 5 cm. apart;
rhachis canescent : bracts reuiform, short acuminate, pale-brown or
roseate-tinged, puberulent, ciliate-margined : calyx white-villous, with
bluntish ovate-lanceolate lobes : corolla 8 to 9 mm. long, puberulent ;
the galea thrice exceeded by the lip. — S. hyptoides, Gray in Wats.
Proc. Am. Acad. xxi. 435 ; Rose, Contrib. U. S. Nat. Herb. i. 110; not
Mart, and Gal. — Northwestern Mexico. Chihuahua, Hacienda San
Miguel, 1885 {Edw. Palmer, no. 205) : Sonora, in shade, mountain-
canon. Alamos, Sept., 1890 {^Edw. Palmer, no. 682) ; Huehuerachi, alt.
1,230 m., Dec, 1890 ((7. V. Hartman, no. 323, F. E. Lloyd, no. 452).
Resembling S. hyptoides.

++ ++ Leaves setulose-hirsute.

= Calyx white-lanate, with short deltoid lobes.

25. S. LASIOCEPHALA, Hook. & Am. Bot. Beech. 306; Benth. 1. c. ;
Hemsl. 1. c. 559; Briq. 1. c. — Tepic, Tepic (the type station), Feb.,
1895 {F. K La7nb, no. 621).

= = Calyx-tube hirsute ; lobes lanceolate.

26. S. HTPTOIDES, Mart. & Gal. I.e. 74; Benth. I.e.; Hemsl. I.e. 558;
Briq. I.e. S. elsholtzioides, Benth. Bot. Sulph. 152, t. 50. — Southern
Mexico to Venezuela. Jalisco, bluffs of the Rio Grande de Santiago,
near Guadalajara, Oct. 19, 1889 (C G. Pringle, no. 2297) : Vera Cruz,
Jalapa, alt. 1,230 to 1,390 m., 1894 {C. L. Smith, no. 1664) : Oaxaca,
dry banks, Sierra de San Felipe, alt. 2,160 m., Oct. 11, 1894 {G. G.
Pringle, no. 5624) ; Jayacatlan, alt. 1,320 m., Nov. 4, 1894 (Z. C. Smith,
no. 269); El Fortin, alt. 520 m., 1897 (C. Oonzatti & V. Gonzalez,
no. 478): Guatemala, Coban, Dept. Alta Verapaz, alt. 1,320 m., March,
1886 (^H. von Tuerchheim in exsicc. J. D. Smith, no. 299) ; Jumayte-
peque, Dept. Santa Rosa, alt. 1,850 m., Sept., 1892 (Heyde «&; Lnx, in
exsicc. J. D. Smith, no. 4047 : Costa Rica, Volcan el Viejo (^Oersted).

Var. subspicata. Plant simple, 1 to 2 dm. high : verticels 2 to 5,

FERNALD. — MEXICAN SALVIAS. 499

approximate in a spiciform raceme 1 to 3.5 cm. long, or the lowest a little
lemole. — CosTA Rica, near San Francisco de Guadalupe, Jan. 4, 1893
{Ad. Tonduz, no. 7228).

t- -1- Verticels approximate in a spiciform raceme: principal leaves 6 to 8 cm. long.

•

27. S. lophanthoides. Stem tall, 1 m. or so high, glabrous, or
minutely pilose at the tip; internodes 1 to 2dm. long: leaves broad-
ovate, short-acumiuate, glabrous above, villous beneath on the nerves, and
especially along the midrib, finely crenate-serrate, on slender petioles
3cm. or less in length: peduncles 1.5 cm. or less long; racemes dense,
the primary ones 1.3 dm. or less in length, 1.5 cm. thick : bracts brownish,
ciliate-margined : calyx viscid, pilose on the strong nerves, in authesis
5 to 6 mm. long, slightly enlarged at the throat, with broad-deltoid sub-
ulate-tipped lobes: corolla blue, slightly viscid, 1.2 to 1.4cm. long, the
lip twice exceeding the galea. — Oaxaca, mountains near Tlapancingo,
alt. 1,850 to 2,460 m., Dec. 7, 1894 {E. W. Nelson, no. 2086).

§ 3. Brachtanthae, Benth. Bracts small, after authesis falling
away, rarely persisting. Corolla middle-sized, conspicuously exceeding
the calyx, 1 to barely 2 cm. long (rarely longer), generally blue, very
rarely white flesh colored or crimson, never scarlet, with the tube gene-
rally ventricose and often contracted at the throat ; with straight or
scarcely arched galea (upper lip), and broader 3-lobed lower lip gene-
rally exceeding the galea, the middle lobe broadest and emarginate.
(Species with exceptionally large flowers, but with their affinities here
are S. heterotricha, S. flaccida, S. cedrosensis, S. semiatrata, and »S'. sidae-
folia with corollas fully 2 cm. long ; S. anyustifolia with corolla 2.5 cm.
long ; and S. sessilifolia with corolla 3 cm. long.)

A. AngustifoUae, Benth. Leafy-stemmed herbs or half-shrubs with
slightly branching stems : leaves linear, lanceolate, or narrowly ovate-
elliptic, narrowed cuneate or rounded-truncate at base.

* Bracts orbicular, acuminate, about equalling the calyx, persistent : verticels
becoming remote : calyx campanulate, green, strongly nerved, subinflaterl,
becoming 1 cm. long, the broad upper lip suberect. Erect annual 2 to
4 dm. higli.
•

28. S. HiRSUTA, Jacq. Hort. Schoenb. iii. 1, t. 252 ; Benth. in DC.

Prodr. xii. 301 ; Hemsl. 1. c. 557 ; Briq. in Engl. & Prantl, 1. c. 278. S.
phlomoides, Cav. Ic. iv. 10, t. 320. S. ciliata, Poir. Diet. vi. 588. S.
bracteata, Poir. 1. c. 622. S. sideritidis, Vahl, Enum. i. 250. S.
cryptanthos, Schultes, Obs. 12, ace. to Benth. S. nepedfolia, Poir.
Suppl. v. 48. S. ciliaris, Sesse & Moc. PI. Nueva Espana, ed. 2, 7. —

500 PROCEEDINGS OF THE AMERICAN ACADEMY.

Central Mexico. Durango, infrequent on shady slopes of ravines,
Santiago Fapasquiaro, Aug., 189 Q (Udw. Palmer, no. 4:51) : San Luis
PoTOSi, alt. 1,850 to 2,460 m., 1878 {Parry & Palmer, no. 736) :
Mexico, barranca near Santa Fe, July 6, 1865-66 (Bourgeau, no.
'490) ; Santa Fe, Sept. 3, 1899 {C. G. Pringle, no. 7991).

* * Bracts less conspicuous, soon deciduous : calyx campanulate, conspicuously

bilabiate, subinflated.

H- Leaves sessile or subsessile or narrowed to short inconspicuous petioles : racemes
elongated, at least the lower verticels becoming remote : perennials. {S.
assurgens may be looked for here.)

++ Stems leafy nearly or quite to the inflorescence.

= Without glandular hairs on stem and calyx.

29. S. ANGUSTiFOLiA, Cav. Stems hispid, bearded at the nodes, 2 to
7 dm. high, very leafy: leaves mostly hispid on the margins and nerves
— Ic. iv. 9, t. 317; Lindley, Bot. Reg. xviii. t. 1554; Sweet, Brit. FL
Gard. n.s. iii. t. 219; Benth. I.e. 301; Hemsl. I.e. 552; Gray, Syn. Fl.
N. A. ii. 369 ; Briq. 1. c. S. reptans, Jacq. Hort. Schoen. iii. 38, t. 319.
S. virgata, Ort. Dec. i. 3. — Central and Southern Mexico. Zaca-
TECAS, near Plateado, Sept. 2, 1897 (/. N. Rose, no. 2745) : Jalisco,
Valley of Mexico, Tizapan, June 26, 1865-66 {Bourgeau, no. 125, in
part) : Michoacan, TIalpujahua 1828 (Graham) ; plains near Patzcuaro,
Aug. 7, 1892 (C. G. Pringle, no. 4163). S. Unifolia, Mart. & Gal.
I. c. 70, and Benth. 1. c. 302, from Michoacan is probably only a form
with rose-colored corolla.

Var. GLABRA, Gray, 1. c. Stems and leaves glabrous. — S. leptophjlla,

Benth. Lab. 249, & in DC. 1. c. 299. S. azurea, Torr. Bot. Mex. Bound.

131, in part. — From Texas through central and southern Mexico, of

broader range than the hispid type. San Luis Potosi, in sand near the

city of San Luis Potosi, 1876 {Schaffner, no. 672) : Aguas Calientes

{Hartweg, no. 163) ; Jalisco, edge of swamp, Guadalajara, July, 1886

(Edw. Palmer, no. 226): Chiapas, without locality, 1864-70 {Ghies-

breght, no. 751) ; valley of Jiquipilas, alt. 650-1,100 m., Aug. 16, 1895

{E. W.Nelson, no. 2922).

= = Glandular-hairy.

30. S. heterotricha. Stems erect from a rather woody base, 2.5 to
4.5 dm. high, puberulent, and, especially above, bearing slender jointed
glandular hairs : leaves linear-attenuate, strongly 1-3-nerved, 3 to 8 cm.
long, puberulent or glabrate or rarely with some slender glandular hairs :
raceme 2 cm. or less in length; verticels all remote, the lower 3 to 4 cm.
apart, 3-9-flowered : bracts lanceolate to ovate-lanceolate, glandular-

FERNALD. — MEXICAN SALVIAS. 501

clliate, mostly shorter than the calyx: pedicels 2 to 4 mm. long: calyx
tubular-campanulate, in authesis 8 to 10 mm. long, glandular-ciliate on
the strong nerves; lobes about 3 mm. long; the upper lip broadly ovate,
bluntish, entire; lower lip with 2 slightly narrower and more pointed
lobes : corolla blue or violet as in S. angustifolia, but the galea more
pubescent : style bearded. — S. angustifolia, Gray in Wats. Proc. Am.
Acad. XX. 445, in part, not Cav. — Jalisco, in bottoms, Rio Blanco,
June, 18G6 {Edw. Palmer, no. 53) ; plains near Guadalajara, July 1,
1889 {O. G. Pringle, no. 2913).

Var. multinervia. Leaves lance-linear to oblanceolate, mostly with
5 parallel nerves. — Tepic, foothills between Acaponeta and Pedro
Paulo, Aug. 2, 1897 (/. N. Rose, no. 1934).

-w- ++ Raceme long-pedunculate : leaves confined to the lower half of the plant.

= Upper lobe of the calyx tridentate.

a: Calyx-tube pilose-hispid, not glandular : attenuate bracts glabrous or glabrate.

31. S. COMOSA, Peyr. Stem 2 to 6 dm. high, glabrous or glandular-
pilose : leaves membranaceous, linear or lanceolate, the upper half crenate-
serrate, generally pubescent beneath on the veins. — Linnaea, xxx. 32 ;
Hemsl. 1. c. 555. S. glechomaefolia, Wats. Proc. Am. Acad, xviii. 137,
in part, not HBK. — Central and southeastern Mexico. Without lo-
cality {Coulter, no. 1115): San Luis Potosi, without locality, 1878
{Parry & Palmer, no. 761) : Mexico, Santa Fc, July 6, 1865-66
{Bourgeau, n. 396) : Vera Cruz, Mt. Orizaba, alt. 2,460 to 2,770 m.,
Aug., 1891 {Seaton, nos. 259, 323). Originally described from Toluca,
Mexico.

Var. hypoglauca. Similar, glabrous : leaves slightly petioled, very
glabrous, glaucous beneath, entire or serrulate at tip. — S. hypoglauca,
Briq. Ann. Conserv. Jard. Bot. Geneve, ii. 136. — Michoacan, moun-
tains about Patzcuaro, July 30, 1892 {G. G. Pringle, no. 4155). A
glabrous and glaucous extreme, not separable otherwise from S. comosa.

b. Calyx and attenuate bracts glandular-hispid.

1. Leaves linear, revolute, with one conspicuous broad nerve.

32. S. unicostata. Slender, 2 to 4 dm. high : stem sparingly gland-
ular-hispid : leaves 3 to 6 cm. long, 2 mm. or less wide, glabrous : pedun-
cles 0.5 to 1 dm. long; raceme 1 to 1.5 dm. long, the lower verticels
0.5 dm. apart, the upper approximate; verticels 3-6-flowered : bracts
short, ovate, subulate-tipped : calyx in anthesis 5 mm. long, the tube
equalling the lance-subulate lobes : corolla 1.3 cm. long, puberulent. —

502 PROCEEDINGS OP THE AMERICAN ACADEMY.

S. angustifolia, var. glabra, Wats. Proc. Am. Acad, xviii. 138, in part,
not Gray. — San Luis Potosi, without locality, alt. 1,840 to 2,460 m.,
1878 {Parry & Palmer, no. 760).

2. Leaves oblanceolate, pinnately nerved.

33. S. firma. Steins decumbent at base, glabrous below, glandular-
pubescent in the inflorescence, 4 to 5 dm. high : leaves 4 to 8 cm. long, 0.5
to 2 cm. broad, acute or blunt at tip, cuueate at base, crenulate-serrate,
thick and firm, glabrous, above sublucid : peduncles elongated, 1-2-brac-
teate ; racemes becoming 3 dm. long ; verticels all distinct, 3-6-flowered,
the lowermost 5 to 8 cm. apart : bracts short, broad-ovate, subulate-tipped :
pedicels short : calyx campanulate, in anthesis 5 to 6 mm., becoming 1 cm.
long ; tube twice exceeding the lobes ; upper lip broad-ovate, the teeth
subulate; lobes of the lower lip with long subulate tips: corolla 1.5 cm.
long, the tube slightly exserted, the dark blue lower lip with a pale spot
in its centre and twice exceeding the puberulent blunt galea : style
bearded. — S. glechomaefolia, Wats. I.e. not HBK. — Jalisco, on hill-
sides, Rio Blanco, June, 1886 (Udw. Palmer, no. 61).

= = Upper lip of calyx entire.

a. Glabrous : leaves oblong-lanceolate : bracts round, obtuse or short-mucronate.

34. S. LAEVis, Benth. Lab. 251, & in DC. I.e. 303; Hemsl. I.e.
559; Briq. I.e. & in Engl. & Prantl, I.e. — Not seen. Described from
Mexico and from near Tlalpujahua, Michoacan.

b. Pubescent : leaves elliptic-oblanceolate or narrowly obovate : bracts ovate, long-
attenuate.

35. S. sinaloensis. Stems slender, 2 to 2.5 dm. high, slightly
branched, below short- hirsute, above and in the inflorescence densely
pubescent with long straight fine viscid hairs : leaves acutish, the upper
half appressed-serrate, the lower half subentire, subcuneate to a sessile
base or obscurely short-pe tided ; those of the main stem 3 to 4 pairs, the
uppermost largest, 3.5 to 6 cm. long, 1,25 to 2 cm. broad, above appressed
short-setulose or glabrate, beneath pale, minutely setulose on the nerves :
peduncle 6 to 9 cm. long, about equalling the leafy lower portion of the
plant; raceme becoming 1 dm. or so long; verticels 3-6-flowered, the
lower 4 cm. apart, the upper rather approximate : bracts pilose-hispid,
colored, soon deciduous : calyx densely pilose-hispid with fine viscid hairs,
in anthesis 5 to 7 mm. long; the tube twice as long as the broad abruptly
subulate-tipped lobes : corolla 1.5 cm. long, the tube slightly exserted, the
dark blue lip with a pale spot in its centre twice exceeding the puberu-

FEIINALD. — MEXICAN SALVIAS. 503

lent blunt oblong galea : style bearded. — Sinaloa, foothills of the
Sierra Madre near Colomas, July 14, 1897 (^J. N. Rose, no. 1727).

M- ^- Leaves nearly all with definite petioles. (S. comosa and S. sinaloensis may

be looked for here.]

-w Annuals, more or less busliy-branched : racemes elongated, at least the lower

verticels remote.

= Leaves linear-lanceolate to oblong-linear, obscurely serrate : bracts lanceolate.

36. S. LANCEOLATA, Brouss. App. Elench. PI. Hort. MoDsp. (1805)
15 ; Willd. Eimm. Hort. Berol. i. 37 ; Jacq.f. Eclog. i. 22, t. 13 ;
Benth. I.e. 299; Gray, 1. c. 369; Hemsl. I.e.; Briq. I.e. aS\ rejlexa,
Horn. Hort. Hafn. (1807), i. 34, *S'. lanceifolia, Poir. Suppl. v. 49.
S. aspidojihyUa, R. & S. Syst. Mant. i. 206. ^S*. trichostemoides, Pursh,
Fl. i. 19. aS'. aegyptiaca, Sesse & Moc. I.e. 6, not L. — Florida and
S. W. United States to Central Mexico. Without locality, 1848-49
{^Gregg, no. 541) : Chihuahua, low ground near Chihuahua, Oct. 1852
{Geo. Thurher, no. 821) ; plains near Chihuahua, Sept. 11, 1885 ( C G.
Pmigle, no. 654); damp places near Pilares, Sept. 23, 1891 (O. V.
Hartman, no. 744) : Coahuila, abundant in abandoned fields and
bottom-lands, Saltillo, Sept., 1898 (Edw. Palmer^ no. 336) : Durango,
abundant in rich bottom-lands, near Durango, July, 1896 {Edw. Palmer,
no. 327) ; along arroyos, Santiago Papasquiaro, Aug., 1896 {Edw.
Palmer, no. 446) ; between Cerro Prieto and La Providencia, Sept. 11,
1898 {E. W. Nelson, 4969) : San Luis Potosi, damp places about the
city, 1876 {Schaffner, no. 673); alt. 1,840 to 2,460 m., 1878 {Parry &
Palmer, no. 744) : Guanajuato, Presa de la olla, 1893 {A. Duges) :
QuEKETARO, Nov. 19, 1827 {Berlandier, no. 1279).

= = Leaves lanceolate or oblong-lanceolate, coarsely subincised-dentate : bracts

broadly ovate.

37. S. SUBINCISA, Benth. PI. Hartw. 20, & in DC. I.e. 303;
Gray, I.e.; Ilemsl. I.e. 565; Briq. I.e. — Texas to Central Mexico.
Chihuahua, Pilares, Sept. 18, 1891 {O. V. Hartman, no. 776): Du-
rango, rich low ground near Durango, July, 1896 {Edw. Palmer, no.
305): Aguas Calientes, in fields near the city {Hartweg, no. 160).
Originally described from Hartweg's plant.

= = = Leaves ovate or narrowly rhombic-ovate, crenate-serrate : bracts narrowly
ovate with long attenuate barbulate tips.

38. S. Chia. Similar to the two preceding : about 6 dm. high, with
long internodes (the lower 5 to 7 cm. long) : stem strongly quadrangular,
puberulent with appressed white hairs, densely white-pilose at the nodes :

504 PROCEEDINGS OF THE AMERICAN ACADEMY.

leaves bluntish at tip, cuneate at base to a long slender petiole, coarsely
crenate-serrate, especially above the subentire base ; margin and petiole
densely short-pilose ; upper face dark green, jjuberulent or glabrate ;
lower face paler, minutely pubescent : racemes very short-pedunculate,
0.5 to 2 dm. long; verticels 3-6-flowered, the lower 1.5 era. apart, the
upper approximate : pedicels 3 mm. long, minutely white-pilose : calyx
ciliate on the strong nerves, narrowly campanulate, iu anthesis 8 mm.
long ; the tube twice exceeding the ovate acuminate lobes ; upper lip
entire: corolla 1.3 to 1.5 cm. long, the white tube somewhat exserted ;
lips blue, the lower pubescent beneath, twice as long as the pubescent
upper one: style glabrous. — Coahuila, damp bottom-lands, Saltillo,
Sept., 1898 {Edw. Palmer, no. 334). This as well as three other species,
S. lanceolata^ S. tiliaefolia, and S. hispanica, are called chia by the Mex-
icans, and a cooling beverage known likewise by that name is prepared
from the seed. (See Rose, Coutr. U. S. Nat. Herb. v. 225.)

•<-<• ■*-*• Perennials, mostlj' decumbent at least at base.

= The lower verticels becoming remote.

a. Corolla white or pale : stem strongly decumbent or subrepent : leaves ovate-
elliptic to oblong, glabrate : peduncle U.5 to 1.5 dm. long.

39. S. ASSURGENS, HBK. Nov. Gen. & Spec. ii. 293 ; Benth. 1. c.
304 ; Hemsl. 1. c. 553 ; Briq. 1. c. — Miciioacan, grassy hills near Patz-
cuaro, July 18, 1892 (C. G. Pringle, no. 4150); Oct. 20, 1898 {E. W.

D. Holway, no. 3184) ; originally collected near the same town, alt. 2,100
m., by Humboldt & Bonpland. According to Kunth the color of the
corollae is " pallide violacea ? (carnea ex Bonpl.)," but neither of the
recently collected specimens shows any violet tinge.

h. Corolla blue or violet.

1. Pilose-hirsute: internodes short: leaves oblong to rhombic-ovate : calvx hispid
below, in anthesis 5 to 7 mm. long; tube equalling the lobes; upper lip triden-
tate or entire.

40. S. PRUNELLOiDES, HBK. 1. c. 289; Benth. I.e. 305; Hemsley,
I.e. 563; Briq. I.e. aS. glechomaefolia, Wats. Proc. Am. Acad, xviii.
137, not HBK. S. tricandra, Briq. Ann. Conserv. Jard. Bot. Geneve,
ii. 133. — Coahuila to Oaxaca. Coahuila, mountains 64 km. south of
Saltillo, July, 1880 {Edw. Palmer, no. 1098) ; limestone hills, Carneros
Pass, Sept. 27, 1890 {G. G. Pringle, no. 3681) : Nuevo Leon, Lerios,
July, 1880 {Edw. Palmer, no. 1097) : Durango, Cacaria, Aug. 5, 1898

E. W. Nelsoyi, no. 4651): Mexico, without locality, 1848-49 {Gregg,
no. 406) ; cool slopes, Sierra de las Cruces, Aug. 21, 1892 ((7. G. Prin-

FERNALD. — MEXICAN SALVIAS. 605

gle, no. 4200) : Oaxaca, Cuilapan, alt. 1,840 m., June 27, 1895 (Z. C.
Smith, no. 778). Originally from Volcan de Jorullo, Michoacan.
Briquet bases bis species S. trichandra upon Pringle's no. 4200, stating
in his description of tbe calyx tbat the upper lip is entire. In the speci-
mens of this number, as represented in the Gray Herbarium, tbe upper
lip is usually tridentate, thus placing the plant distinctly with S. pru-
nelloides.

2. Pulierulent : internodes longer : leaves oblong or narrowly ovate-oblong : calyx
puberulent, tubular-campanulate, in anthesis 7 to 8 mm. long, the tube one
half longer than the lobes ; upper lip entire.

41. S. OBLONGiFOLiA, Mart. & Gal. I. c. 79 ; Benth. 1. c. ; Hemsl. 1. c.
562; Briq. in Engl. & Prantl, I.e. S. reticulata, Mart. & Gal. I.e. 64,
ace. to Benth. — Southern Mexico. Chiapas, without locality, July,
Aug., 1864-70 (Ghiesbreght, nos. 61, 750) ; near San Cristobal, alt.
2,150 to 2,460 m., Sept. 18, 1895 (K W. Nelson, no. 3191). Originally
described from Oaxaca.

= = Verticels congested in a long-peduncled head : stem pilose-hispid, 2 to 2.5
dm. high: leaves ovate-elliptic or ovate-oblong, obtuse: heads 1 to 1.5 cm.
high.

42. S. Tatei, Briq. Ann. Conserv. Jard. Bot. Geneve, ii. 135. —
Mexico, without indicated locality in herb. Delessert. Not seen.

* * * Bracts deciduous : calyx cylindric, not becoming inflated, less conspicu-
ously bilabiate : stems numerous from a woody base.

f- Bracts very early deciduous : calyx blue-tinged, conspicuously nerved, puberu-
lent or short-pubescent.

•w- Verticels, or all but the lowermost, aggregated, forming a rather dense head.

= Stems 1 m. or less high, finely canescent: leaves 2 to 8 cm. long, soft-canescent
beneath : heads simple or branched, 2 to 12 cm. long : calyx, in anthesis, 5 to
7 mm. long; the tube thrice exceeding the deltoid subulate-acuminate lobes:
corolla blue rose or white.

43. S. LAVENDULOiDES, HBK. Leaves lanceolate, oblanceolate, or
oblong-lanceolate, acutish' or blunt, crenulate. — Nov. Gen. & Spec. ii.
287 ; Benth. 1. c. 303 ; Hemsl. 1. c. 559 ; Briq. I. c. S. Humboldtiana,
R. & S. Syst. Mant. i. 183. S. lavendulaefolia, Spreng. Syst. i. 58, not
Vahl. S. purpurina, La Llave, La Nat. vii. 82. S. stricfa, Sesse & Moc.
I.e. 8. — Southern Mexico and Central America. Mexico, Valley of
Mexico {Bourgeau, no. 1110, Schafner, no. 410) : Michoacan, hills of
Patzcuaro, Nov. 21, 1891 (C G. Pringle, no. 3954) : Morelos, Cuer-
navaca, alt. 2,300 m., Jan. 4, 1899 (01 C. Deam, no. 2) : Oaxaca, N. W.
slope of Mt. Zempoaltepec, alt. 2,460 to 3,000 m., July 10, 1894 {E. W.
Nelson, no. 701) : Cuyamecalco, alt. 2,000 m., Sept. 4, 1895 (Z. C. Smith,

506 PROCEEDINGS OF THE AMERICAN ACADEMY.

no. 601): Chiapas, without locality, 1864-70 {Ghiesbreght, nos. 738,
741, 747); near San Cristobal, alt. 2,150 to 2,460 ra., Sept. 18, 1895
{E. W. Nelson, no. 3142) : Guatemala, Santa Rosa, Depart. Baja
Verapaz, alt. 1,530 m., April, 1887 {H. von Tuerckheim in exsicc. J. D.
Smith, no. 1193); Santiago, Depart. Zacatepequez, alt. 2,000 m., 1891
(^Rosalio Gomez in exsicc. J. D. Smith, no. 823) ; Sacabaja, Depart.
Quiche, alt. 1,230 m., March, 1892, and Castillas, Depart. Santa Rosa,
alt. 1,230 ra., Dec, 1892 {Heyde & Lux in exsicc. J. D. Smith, uos.
3128, 4395). Originally described from Patzcuaro, Michoacan.

Var. LATiFOLiA, Benth. Leaves elliptic-oblong, acute. — PI. Hartw.
21, & in DC. I.e. — Jalisco, Bolaiios {Hartweg, no. 171).

= = Similar: stems minutely retrorse-pubescent : leaves not canescent beneath.

a. Leaves oblong or narrowly ovate-elliptic, 1.5 to 3 cm. long, tliickish, rugose,

green, sometimes a little pubescent on the nerves beneath, obscurely crenate or
entire : lower verticels a little remote : calyx 5 to 6 mm. long the upper lip very
short, the lower with 2 ovate lobes 1 mm. long.

44. S. GUADALAJARENSis, Briq. Ann. Conserv. Jard. Bot. Geneve ii.
132. S. helianthemifolia^ van, Gray in Wats. Proc. Am. Acad. xxii. 445,
not S. helianthemifolia, Benth. — Jalisco, among rocks, Rio Blanco,
Sept., 1886 {Edw. Palmer^ no. 556) ; dry rocky hills near Guadalajara,
Nov. 1, 1893 (a G. Pringle, no. 4624).

b. Leaves lanceolate or oblong-linear, bluntish, dull, minutely puberulent or gla-

brate : deltoid acuminate calyx-lobes subequal.

1. Stems numerous, assurgent, 2 to 2.5 dm. high, leafy chiefly near the base : leaves

entire, glaucous : calyx white-puberulent, in anthesis 6 mm. long.

45. S. Teresae. Leaves short-petioled, 2 to 2.5 cm. long, 0.25 to
0.5cm. broad: peduncles 1 to 1.5 dm. long; raceme rather loosely flow-
ered ; the lower verticels 1 cm. apart, 2-6-flowered : bracts ovate-lance-
olate, acuminate : calyx dark blue : corolla 1.3 cm. long, the galea densely
pubescent. — Tepic, near Santa Teresa, top of Sierra Madre, Aug. 13,
1897 {J. N. Rose, no. 2233).

2. Stems few, erect, 6 to 8 dm. high : leaves remote, serrate, not glaucous : calyx

minutely pubescent, not whitened, in anthesis 4 to 5 mm. long.

46. S. muscarioides. Stems simple or sparingly branched ; inter-
nodes 1 dm. or less long; leaves short-petioled, 5 to 6 cm. long, 0.75 to
1 .5 cm. wide : peduncles 2 dm. or less long : bracts minute, caducous :
lower verticels 2 or 3 cm. apart, the others crowded, 8-20-flowered :
pedicels spreading and drooping: calyx dark blue and green: corolla
1.2 cm. long; the lower lip much exceeding the pubescent galea: style
slightly exserted, bearded. — Northwestern Mexico. Chihuahua, base

FERNALD. — MEXICAN SALVIAS. 507

of Mt. Mohinora, 13 km. from Guadalupe y Calvo, alt. 2,150 to 2,310 m.,

Aug., 1898 {E. W. Nelson, no. 4850). Inflorescence, as also that of

related species, suggesting Muscari botryodes.

.^ *♦ Verticels mostly remote, forming an elongate loose raceme.

= Stem 5 to 6 dm. high : leaves oblong, blunt or acutish : racemes 2.5 dm. long or
loss ; verticels G-20-flo\vered, lower 4 to 5 cm. apart, upper approximate :
calyces retiexed : corolla 1 cm. long, the tube included or short-exserted.

47. S. HELiANTHEMiFOLiA, Benth. Lab. 254, & in DC. 1. c. 304 ;
Ilemsl. 1. c. 557 ; Briq. in Engl. & Prantl, 1. c. — South-central Mexico.
S.\N Luis Potosi, alt. 1,840 m., 1878 {Parry &, Palmer, no. 729) :
Hidalgo, Sierra de Pachuca, alt. 3,000 m., Sept. 14, 1899 (C. G.
Pringle, no. 8222): Michoacan, near Tlalpujabua {Graham). De-
scribed from Graham's material from Tlalpujahua and from San Martin,

PUEBLA.

= = Similar : leaves oblong-lanceolate, acute, serrate : verticels subequally re-
mote : calyces less retiexed : corolla-tube equalling the calyx.

48. S. RE5I0TA, Benth. I.e. ; Hemsl. I.e. 564; Briq. 1. c. Described
from Mexico without definite locality.

-1- 1- Bracts hardly persistent, lower verticels becoming slightly remote : nerves
of calyx hidden by the long appressed silky pubescence.

49. S. cryptodonta. Resembling S. lavendidoides : stems canes-
cent : leaves narrow-oblong, blunt, crenate-serrate, 2 to 5 cm. long, 0.5 to
1.5 cm. wide, rugose, green and appressed-setulose above, white-tomentose
beneath, short-petioled : raceme 2 to 5 cm. long ; bracts ovate-lanceolate,
acuminate, ascending : calyx blue, in anthesis 5 mm. long, the short
teeth obscured by the dense silky hairs : corolla 1 cm. long, the tube
barely exserted ; galea pilose, half as long as the lip. — Durango,
Aug. 16, 1897 (/. N. Rose, no. 2338).

t- -1- -t- Bracts persistent through anthesis, their slender tips conspicuous : nerves
of calyx somewhat hidden by long hairs : heads very dense.

= Leaves lanceolate, slightly canescent: the long-peduncled heads 3 to 5 cm. in
lengtii : bracts ascending : calyx white-villous.

50. S. STACHYOiDES, HBK. 1. c. t. 138; Benth. I.e. 303; Hemsl.
I.e. 5G5 ; Briq. I.e. — Southern Mexico. Oaxaca, 29 km. S. W. of
the city of Oaxaca, alt. 2,300 to 2,920 m., Sept., 1894 {E. W. Nelson,
no. 1387). Originally from Los Joares and Santa Rosa.

= = Leaves oblong, glabrous or slightly puberulent : heads 2 to 13 cm. long :
bracts more spreading : calyx pilose-hirsute.

51. S. ELONGATA, HBK. 1. c. t. 139 ; Benth. I.e.; Hemsl. I.e. 556.
S. Betonica, R. & S. I. c. 188. S. simplex, Spreng. Syst. i. 58. — South-

508 PROCEEDINGS OF THE AMERICAN ACADEMY.

central Mexico. Mexico, Valley of Mexico, Sept. 7, 1865-66 (Bour-
ffeau, no. 859); wooded canons, Sierra de las Cruces, Oct. 2, 1892, and
Serrania de Ajusco, alt. 3,075 m., Sept. 11, 1897 (C G. Pringle, nos.
4278, 7457). Originally from the region of Ario, Michoacan.

B. Acaulae. Similar to AngustifoUae but tending to be acaulescent or
subacaulesceut, the obovate canesceut basal leaves forming a rosette :
peduncle 0.5 to 2.5 dm. high ; raceme as long ; the verticels remote.

52. S. NANA, HBK. 1. c. 289 ; Benth. in DC. 1. c. 304 ; Hemsl. 1. c.
561 ; Briq. 1. c. S. pi-unelloides, Benth. PI. Hartw. 90, 351, not HBK.
S. rhomhifoUa, Sesse & Moc. I.e. 8. — Northern Mexico to Central
America. Durango, El Salto, July 12, 1898 {E. W. Nelson, no. 4566) :
Zacatecas, near San Juan Capistrano, Aug. 18, 1897 (/. N. Rose, no.
3534) : San Luis Potosi, rare in the mountains, San Rafael, 1876
{Schafner, no. 680) ; alt. 1,840 m. {Parry & Palmer, no. 745, 746) :
Guanajuato, 1893 (A. Duges, no. 228 B) : Oaxaca, Boca de Leon,
Telixtlahuaca, June 27, 1895 (Z. G. Smith, no. 414). Originally from
Guanajuato.

C. Vulgares, Benth. Branching or sometimes simple herbs, rarely
half-shrubs : leaves petioled, ovate, rarely oblong, membranaceous,
rounded round-cuneate or subcordate at base. (>S'. prunelloides and
*S'. Martensii may be looked for here.)

« Annuals.

•<- Coarse more or less canescent plant with long-petioled pale-green leaves and
peduncled spiciform heads with persistent foliaceous broad bracts.

53. S. HiSPANiCA, L. Spec. 25 ; Edw. Bot. Reg. v. t. 359 ; Benth. in
DC. 1. c. 308 ; Hemsl. 1. c. 558 ; Briq. in Engl. & Prantl, 1. c. 279.
S. tetragona, Moench, Meth. 373. S. prismatica, Cav. fide Hemsl. 1. c.
S. neo-hispaiiica, Briq. Ann. Conserv. Jard. Bot. Geneve, ii. 137. —
From western Texas and Coahuila to northern South America and the
West Indies. Introduced into southern Europe, whence the specific
name. Coahuila, Carneros Pass, Sept. 27, 1890 {G. G. Pringle, no.
3683) : Durango, sides of arroyos, Santiago Papasquiaro, Aug., 1896
{Edw. Palmer, no. 967); bottom-lands, Durango, Oct., 1896 {Edw.
Palmer, no. 757) : San Luis Potosi, sand near the city, 1876
{Schaffner, nos. 675, 1053) : Guanajuato, 1895 {A. Duges) : Jalisco,
Rio Blanco, Oct., 1886 {Edw. Palmer, no. 659) : Vera Cruz, Orizaba
{Botteri, no. 534) : Mexico, Santa Fe, Oct. 15, 1865-66 {Boiirgeau, no.
1109): Oaxaca, near Rej^es, alt. 1,800 to 2,600 m., Oct. 20, 1894
{E. W. Nelson, no. 1782); Jayacatlan, alt. 1,350 m., Nov. 4, 1894

FERNALD. — MEXICAN SALVIAS. 509

(Z. G. Smith, no. 2S8) : Guatesiala, Buena Vista, Depart. Santa
Rosa, alt. 1,700 m., Dec, 1892 {Heyde & Lux in exsicc. J. D. Smith, no.
4-101).. Frequently cultivated in Mexico as " Chia-blanco " — see note
under S. Chia above.

-t- H- Delicate, not canescent : leaves short-petioled : bracts small, deciduous.

54. S. flaccida. Slender, 2 to 3 dm. high, simple or branching from
the base, puberulent below, more or less glandular-pilose above : leaves
very thin and flaccid, glabrous and lucid or minutely hispidulous on the
nerves beneath, from rhombic-ovate to oblong-ovate, short-acuminate,
coarsely appressed-serrate, cuneate and entire at base, the uppermost
largest, 4 to 6 cm. long, 2 to 3.5 cm. bi'oad ; petioles slender, puberulent,
0.5 to 1.5 cm. long: peduncle 4 to 5 cm. long; raceme short, 2 to 3
cm. Ions:, of 2 or 3 remote 3-6-flowered verticels : bracts lanceolate :
pedicels 3 mm. long : calyx minutely glandular-hispidulous, narrovr-
campanulate, in anthesis nearly 1 cm. long ; the upper lip bluish, entire,
ovate, bluntly mucronate ] the lower lip paler, with 2 lance-atteuuate
lobes : corolla 2 to 2.5 cm. long, the white tube one half longer than
the calyx, hardly ventricose; galea blue, puberulent, one half as long
as the broad white lip. — Southern Mexico and adjacent Central America.
Chiapas, between Tumbala and El Salto, alt. 460 to 1,380 m., Oct.
29, 1895 {E. W. Nelson, no. 3374) : Guatemala, in woods, Rubeleruz,
Depart. Alta Verapaz, alt. 770 m., Oct., 1885 {H. von Tuerckheim in
exsicc. J. D. Smith, no. 780).

* * Perennials.

■»- Erect or suberect, branching mostly above the base, leafy nearly or quite to

the inflorescence.

++ Leaves rounded rounded-truncate or subcordate at base, not tapering to the

distinct petioles.

= Calyx in anthesis about 3 mm. long.

a. Racemes terminal and from the upper axils, panicled, and often branching:
leaves pale green, glaucous beneath : calyx minutely puberulent, with very
short broad-deltoid teeth.

1. Racemes 2 to 7 cm. long, compactly flowered : pedicels 1 mm. long.

55. S. BREViCALYX, Benth. in DC. 1. c. 309 ; Hemsl. 1. c. 553. —
Southern Mexico. Guerrero, between Ayuainapa and Petatlan, alt.
1,530 to 2,150 m. Dec. 14, 1894, and at top of Sierra Madre near Chil-
pancingo, alt. 2.770 to 3,140 m., Dec. 24, 1894 {E. W. Nelson, nos.
2146, 2220) : Oaxaca, mountains near Tlapancingo, alt. 1,840 to 2.460
m., Dec. 7, 1894 {E. W. Nelson, no. 2064). Originally from Oaxaca.

510 PROCEEDINGS OF THE AMERICAN ACADEMY.

2. Racemes longer, slender : pedicels 2 to 3 mm. long, equalling the calyx.

56. S. FiLiPES, Benth. 1. c. ; Hemsl. I.e. 556. S. polystachya, var.,
Benth. PI. Hartw. 50, not S. poJijstachya, Ort. Described from Regla.
Not seen.

b. Racemes short, simple, solitary, or paniculate, with some remote lower verticels
in tiie axils of the upper foliar leaves : calyx hispidulous, with short deltoid-
subulate teeth.

57. S. menthiformis. Tall herb, 6 to 8 dm. high : stems strongly
angular and furrowed, cinereous-puberuleut, freely branching : leaves
ovate or orbicular-ovate, acuminate, round-cordate at base, crenulate-
serrate, 1.5 to 5 cm. long, 1 to 3.5 cm. broad, dark green and glabrate
above, pale and minutely puberulent or glabrate beneath, on slender
ciuereous-puberulent petioles 1 to 2 cm. long ; raceme, excluding the
remote axillary verticels, 2 to 5 cm. long ; the crowded verticels 6-20-
flowered : pedicels barely 1 mm. long : corolla blue, pilose, 1.2 cm. long;
the tube twice exceeding the calyx ; the lip hardly equalling the galea :
style glabrous or slightly bearded. — S. polystachya, Donnell Smith,
Enum. PL Guat. iv. 126, not Ort. S. purpurea, Donnell Smith, 1. c. in
part, not Cav. — Costa Rica, Cartago, alt. 1,300 m., Oct., 1887 {Juan
J. Cooper in exsicc. J. D. Smith, no. 5902) ; San Francisco de Guada-
lupe, alt. 1,200 m., Jan., 189^ {A. Tonduz, no. 1781); Rio Turrialba,
Prov. Cartago, alt. 500 m., March, 1894 {J. D. Smith, no. 4919).
Habitally resembling forms of Mentha aquutica.

c. Racemes simple, with no axillary lower verticels.

1. Raceme spiciform, densely flowered, 4 to 10 cm. long: leaves broad-ovate, 1
dm. long, 6 dm. wide, on long petioles: corolla white.

58. S. Cataria, Briq. Ann. Conserv. Jard. Bot. Geneve, ii. 142.
Described from Costa Rica, No specimen seen.

2. Raceme elongate, loosely-flowered, at least the lower verticels remote.

O Leaves broad-ovate, glabrous, coarsely and irregularly serrate : pedicels 2 to 4

mm. long.

59. S. brachtodonta, Briq. 1. c. 149. S. albijlora, var. caerules-
cens, Gray in Wats. Proc. Am. Acad. xxii. 445, in part. — Jalisco,
on sides of cafions, Rio Blanco, Sept., 1886 {Edw. Palmer, no. 598);
hillsides near Guadalajara, Sept. 27, 1889 {G. G. Pringle, no. 2463).
Placed by Briquet under the group Cordifoliae, Benth., but the leaves
are rounded-truncate at base, not cordate, and in habit as well the plant
much more resembles members of the Vulgares.

FERNALD, — MEXICAN SALVIAS. 511

O O Leaves narrow-ovate, lanate beneatli in the axils of the nerves, finely and
regularly serrate : pedicels 1 mm. long.

60. S. Ghiesbreghtii. Tall, 1 m. (?) or less high: stems puberu-
lent, with long asceudiug branches : leaves loDg-acuminate, rounded at
base, 3 to 6 cm. long, 1 to 3 cm. wide, puberulent above, pilose or
glabrate beneath, except for the often densely lanate mid-nerve, on
petioles 1 cm. or less in length : racemes slender, flexuous, 1 to 2.5 dm.
long; verticels 10-20-flo\vered, the lower 1.5 cm. apart: bracts small,
ovate-attenuate, caducous : calyx hispid, with short-deltoid ciliate subu-
late-tipped teeth : corolla pale blue, pilose, 1.2 cm. long; the ventricose
tube twice exceeding the calyx ; galea and lip subequal : style bearded.
— ^. polystachya, Hemsl. 1. c. 563, in part, not Ort. — Chiapas, among
the mountains, July, 1864-70 {Ghiesbreght, nos. 129, 743).

= = Calyx in anthesis about 5 mm. long.

a. Lower surfaces of firm leaves tlie pedicels and the cal^'ces permanently canes-
cent with fine stellate pubescence.

61. S. Palmeri, Gray, Proc. Am. Acad. xxi. 408. — Chihuahua,

hillsides at the Frailes, on the mountains above Batopilas, 1885 {Edw.
Palmer, no. 259).

6. Lower surfaces of leaves the pedicels and the calyces canescent with appressed
short hairs : racemes very elongated, the verticels subequally remote : broad
lip of corolla much exceeding the galea.

%-2. S. LEPTOSTACHYS, Benth. Lab. 258, «fe in DC. 1. c. 308; Hemsl.
I.e. 560; Briq. in Engl. & Prantl, I.e. — South central Mexico.
Jalisco, near Plateado, Aug. 31, 1897 (/. N. Rose, no. 2682) : More-
Los, hills near Cuernavaca, Nov. 10, 1895 (C. G. Prlngle, no. 7078).

c. Lower surfaces of leaves canescent-tomentose, glabrate or glabrous : calyx vil-
lous or hirsute : galea and lip of corolla subequal.

63. S. POLYSTACHYA, Ort. Leaves broad-ovate, tomentose beneath :
racemes panicled : calyx canescent, villous. — Dec. 55 j Cav. Ic. i. 17,
t. 27; HBK. I.e. 296; Benth. I.e.; Hemsl. I.e. 563; Briq. I.e. S.
Unenrif'}Ha, Lag. Nov. Gen. & Spec. 2. S. Durandiana (as subspecies),
Briq. Bull. Soc. Bot. Belg. xxx. 238, & Ann. Conserv. Jard. Bot.
Geneve, ii. 138. — An extremely variable species, the typical form seen
only from Michoacan, Tlalpujahua {Graham).

Var. CAESiA, Briq. Leaves narrow-ovate to ovate-lanceolate, glaucous,
glabrous or minutely pubescent, not tomentose, beneath : inflorescence as
in the type but racemes more elongated, the lower verticels often remote :

512 PROCEEDINGS OF THE AMERICAN ACADEMY.

calyx generally with shorter pubescence. — Bull. Soc. Bot. Belg. xxx.
236 (fe Ann. Conserv. Jard. Bot. Geneve, 1. c. S. caesia, Willd. Enum.
i. 40; HBK. 1. c. 295. — Central Mexico to Central America. Hidalgo,
Guadalupe, Valley of Mexico, Aug., 18G5-66 (Bourgeau, nos. 721, 854),
Aug. 17, 1865 {Bilimek, no. 316) : Mexico, Takubaya, Aug. 28, 1865
(Bilhnek, no. 315) : Oaxaca, near Puebla, alt. 2,150 m., Nov. 9, 1895
(L. C. Smith, no. 908).

Var. philippensis. Leaves as in the species : racemes simple,
elongated, 1 to 2 dm. long: calyx short villous. — Oaxaca, Sierra de
San Felipe, alt. 2,300 m., Sept. 1, 1894 (E. W. Nelson, no. 1175), Oct.
5, 1894 {C. G. Pringle, no. 4953).

Var. seorsa. Similar to the last: leaves glabrate beneath: racemes
simple, elongated ; verticels many-flowered, the lower 2 to 3 cm. apart. —
Mexico, Valley of Mexico, Sept. 19, 1889 (C. G. Pringle, no. 2818).

Var. albicans. Leaves ovate, thinner than in the other forms,
slightly canescent above, very white-tomentose beneath : racemes simple,
1 to 1.5 dm. long : calyx densely white-villous. — S. purpurea, var.
pubens, Donnell Smith, I.e. iii. 67, not Gray. — Guatemala, Santa
Rosa, Depart. Santa Rosa, alt. 920 m., Nov., 1892 {Heyde & Lux in
exsicc. J. D. Smith no. 4394). An extremely canescent plant.

Var. POTOSINA, Briq. 1. c. Leaves ovate-lanceolate or lanceolate,
long-acuminate, pubescent as in the species: racemes simple, 1 dm. or
less long : calyx hirsute. — San Luis Potosi, San Jose Pass, July 23,
1890 ((7. G. Pringle, no. 3224).

d. Leaves pilose-hispid or glabrate beneath : stems and calyces pubescent with

slender spreading glandular hairs.

64. S. aequidistans. Erect or ascending: stems 1.5 to 4 dm. high,
glandular-hirsute : leaves rather remote, the lower small, the upper
much larger, oblong-ovate to broadly deltoid-ovate, obtuse, the upper-
most 3 to 4.5 cm. long, crenate-serrate, more or less setulose above:
racemes slender, elongated, 0.5 to 2 dm. long ; the verticels 2-6-flowered,
all remote, the lower equidistant, 2 to 3 cm. apart : pedicels filiform, 2
to 4 mm. long : calyx campanulate, strongly bilabiate, the tube twice or
thrice as long as the ovate lips ; upper lip entire, lower short-cleft :
corolla 1.3 cm. long, the pilose or glabrate tube ventricose, nearly twice
exceeding the calyx ; lip a little exceeding the pilose galea : style
bearded. — Sinaloa, between Rosario and Colomas, July 12, 1897 (/.
N. Rose, no. 1609).

FERNALD. MEXICAN SALVIAS. 513

= ==== Calyx in anthesis 6 to 8 mm. long.

a. Leaves broad-ovate, crenate-serrate, obtuse : bracts large, broad-ovate : upper

lip of ]uirple glandular-hairy calyx tridentate : corolla dark-blue, with the
calyx, bracts, &c., red-dotted.

65. S. TRicusPiDATA, Mart. & Gal. I.e. 78 ; Benth. I.e. 311 ; Hemsl.
1. c. 566. — Oaxaca, Sierra de San Felipe, alt. 3,080 m., Aug. 28, 1894
{C G. Pringle, no. 4845). Originally collected by Galeotti in the
same mountains.

b. Leaves incised-serrate, acuminate : bracts minute, lance-subulate : calyx blue

and green, minutely liispidulous, upper lip entire : corolla paler blue without
red dots.

66. S. PRASiiFOLiA, Benth. Bot. Sulph. 151, & in DC. I.e. 310;
Hemsl. 1. c. 563. S. aliena, Greene, Pittonia, i. 157. — Tepic, Maria
Madre Island, April, 1877 {W. T. Fisher), May, 1897 {E. W: Nelson,
no. 4247). Originally described from Tepic.

++ ++ Five species are not identified with recent material.

From the descriptions alone it is impossible to make out their exact
relationships, or whether they are all distinct from the species here
defined. These five plants are: — S. gracilis, Benth. Lab. 258, & in
DC. 1. c. 307, described from " New Spain " ; S. protracta, Benth. in
DC. 1. c. 309, described from Oaxaca ; S. membranacea, Benth. Lab.
259, & in DC. 1. c. 310, described from " Mexico " ; S. glabra. Mart. &
Gal. 1. c. 68, Benth. 1. c, described from Oaxaca ; S. herbacea, Benth.
Lab. 720, &, in DC. 1. c. 311, described from Tehuantepec, Oaxaca.

4^. ++ ++ Leaves cuneate at base, or at least attenuate to the petioles.

= Leaves white-villous or pilose beneath.

a. Herbaceous.

1. Raceme dense, 4 to 7cm. long: leaves densely villous beneath.

O Leaves thick, crenulate-serrate.

67. S. XALAPENSis, Benth. I.e. 308; Hemsl. I.e. 566; Briq. in
Engl. & Prantl, 1. c, S. polystachya, Mart. & Gal. 1. c. 77, not Ort. —
Southern Mexico. Vera Cruz, Jalapa, alt. 1,230 to 1,380 m., .Jan. 1 6,
1894 (C. L. Smith, nos. 1749, 1756) ; near Motzorongo and Omealca,
Feb. 22, 1894 {E. W. Nelson, nos. 132, 177).

O O Leaves thin, sharply serrate.

68. S. LONGiSPiCATA, Mart. & Gal. 1. c. 73 ; Benth. 1. c. 307 ; Hemsl,
1. c. 560; Briq. 1. c. — Southwestern Mexico. Guerrero, between

VOL. XXXV. — 33

514 PROCEEDINGS OF THE AMERICAN ACADEMY.

Coiiala and Juchitango, alt. 60 to 185 m., Feb. 9, 1895 {E. W. Nelson,
no. 2300). Originally collected by Galeotti in Michoacan.

2. Racemes looser, elongated, the terminal one 1 to 3 dm. long : leaves less

densely villous.

O Calyx in anthesis 4 to 5 mm. long : corolla 1 cm. long.

69. S. CORDOBENSIS, Briq. Ann. Conserv. Jard. Bot. Geneve, ii.
140. — Vera Cruz, Valley of Cordova, Aug. 24, 1866, and region of
Orizaba, Aug. 12, 1866 (Bourgeau, nos. 1591, 2857); Cordova, alt.
790 m., Aug. 20, 1891 (Seaton, no. 431) : Pdebla, near Metlaltoyuca,
alt. 250 m., Feb. 27, 1898 (B. A. Goldman, no. 72).

O O Calyx in anthesis 8 to 9 ram. long : corolla 2 cm. long.

70. S. monclovensis- Stems sparingly pilose or glabrate : leaves
ovate, acuminate, coarsely crenate-serrate, the primary ones 7 to 8 cm.
long, unequally cuneate or rounded-cuneate at base, dark green and mi-
nutely puberulent above, canescent beneath, on petioles 3 to 4 cm. long :
peduncles 1 dm. or less long: verticels 4-8-flowered, the lowest 1.5 to
3 cm. apart : bracts narrow-ovate, long-acuminate, the lowest 1.5 cm.
long, sometimes leaf-like and slightly toothed : calyx minutely appressed-
pubescent ; the tube thrice exceeding the broad-ovate ciliate mucronate-
short-acuminate lobes : corolla blue, sparingly pilose ; the tube one half
longer than the calyx ; the lip twice exceeding the galea : style bearded.

— Coahuila, Caracol Mts., 33.8 km. southeast of Mouclova, Aug., 1880
(Edtv. Palmer, no. 1096).

b. Shrubby at base : leaves thinly villous.

1. Stem erect, pubescent: leaves crenate-serrate: calyx-lobes broad ovate: galea

of corolla glandular-pilose.

71. S.WARSZEWicziANAjllegel, Flora, xxxii. 184; Walp.Ann. iii.257;
Hemsl. 1. c. 566. — Described from Guatemala. No specimen seen.

2. Stem ascending, branches sordid-villous : leaves sharply serrate : lower calyx-

lobes lance-ovate, long-attenuate : galea villous not glandular.

72. S. Sanctae-Luciae, Seem. Bot. Herald, 327 ; Hemsl. 1. c. 565.

— Western Mexico. Tepic, Tepic, Jan. & Feb., 1892 (Edw. Palmer,
no. 1964). Originally from Santa Lucia in the Sierra Madre.

= = Leaves glabrous or puberulent beneath, only the nerves sometimes hispid or
pilose, or tlie youngest minutely canescent-tomentulose.

a. Calyx in anthesis 3 to 4 mm. long.

1. Leaves green and glabrous on both sides : calyx campanulate and angulate
below, enlarged-cupulifonn above.

73. S. JuRGENSENii, Briq. 1. c. 144. Described from Oaxaca.

FERNALD. — MEXICAN SALVIAS. 515

2. Young leaves canescent-tomentulose at least beneath: calyx simply

carapanulate.

74. S. mazatlanensis. Stems slender, miuutely puberulent or
glabrate : leaves rhombic-ovate or deltoid-ovate, with long acuminate
entire tips, and entire cuneate bases, otherwise coarsely appressed-serrate,
thin, dark green above, pale beneath, 2 to 4 cm. long, 1 to 3 cm. wide,
on slender petioles 2 cm. or less in length : racemes elongate, 0.5 to 2
cm, long ; verticels 2-10-flowered, all somewhat remote, the lowest 1 to
1.0 cm. apart: bracts lance-subulate, caducous : pedicels 2 to 4 mm. long:
calyx glabrate ; the tube twice exceeding the lobes ; upper lip broad-
ovate, entire, blunt and submucronate, the lower with ovate-lanceolate
acuminate lobes : corolla blue, 1 to 1.2 cm. long; galea pilose, nearly
equalling the lip. — Sinaloa, head of Mazatlan River, Jan., 1889 ( W. G.
Wright, no. 1298); Mazatlan, Dec. 29, 1894 (F. H. Lamb, nos.
351, 355). Lamb's no. 395 from Villa Union has thicker leaves and
puberulent calyx, and may not belong with the Mazatlan plant,
though it is nearer related to that than to any other species.

h. Calyx in anthesis 5 to 6 mm. long.

1. Branches cinereous : leaves cinereous-puberulent beneath, rhombic-ovate,
coarsely crenate-serrate : racemes rather loosely flowered, 1 to 3 dm. long;
verticels 2-15-flowered : calyx tubulose-campanulate, somewhat pilose on the
nerves.

75. S. JALISCANA, Briq. I. c. 141. S. alhijiora, Gray, Proc. Am.
Acad. xxi. 408, not Mart. &, Gal. S. albljiora, forma caerulescens,
Gray, 1. c. S. albijlora, var. caerulescens, Gray, ace. to Wats. Proc. Am.
Acad. xxii. 445, in part. — Western Mexico. Chihuahua, shady ravines
near Batopilas, 1885 {Edw . Palmer, nos. 96, 154, 157) : Jalisco, in
ravines and by shaded roadsides, Guadalajara, Sept., 1886 {Edw. Palmer,
no. 488) ; by streams near Guadalajara, Nov., 1888 ( C. G.Pringle, no.
1798).

2. Leaves glabrous, glabrate, or merely pubescent on the nerves beneath.

O Leaves and calyces strictly glabrous : leaves with long-acuminate tips : calyx
with sliort broad generally mucronate-tipped lobes.

76. S. ALBiFLORA, Mart. & Gal. 1. c. 76 ; Benth. in DC. 1. c. 307 ;
Gray, Syn. Fl. 1. c. 370 ; Hemsl. 1. c. 552 ; Briq. in Engl. & Prantl,
1. c. — Arizona to Venezuela. Sonora, moist places, Magdalena, Oct.,
1857 (Geo. Tlmrher, no. 907) : Vera Cruz, Tolosa, Dec. 24, 1898
(C. C. Beam, no. 57): Oaxaca, Santo Domingo, alt. 290 m., June 12,
1895 (E.W. Nelson, no. 2668).

516 PROCEEDINGS OF THE AMERICAN ACADEMY.

O O Lower faces of leaves and nerves of calyces pubescent.
+ Verticels crowded into a dense spiciform raceme 5 to 6 cm. long: stem shrubb5\

77. S. PTEROURA, Briq. Ann. Conserv. Jard. Bot. Geneve, ii. 139.
Described from Costa Rica.

+ + Verticels becoming remote.
X Branches slender, glabrate : calyx tubular-campanulate, not enlarged above.

78. S. Seemannii. Frutescent (?) : branches minutely pilose or gla-
brate, purplish ; leaves elliptic-ovate, acuminate at the tip and at base to
a slender petiole 1.5 to 3.5 cm. long, sharply serrate, above sparsely
appressed-setulose, beneath minutely so on the nerves or glabrate, 4 to 7
cm. long, 2 to 4 cm. wide : raceme 1 dm. or less in length ; verticels
about 6-flowered, all a little remote, the lowest 1 cm. apart : bracts
purple, lance-attenuate, persistent, slightly exceeding the setulose pedi-
cels, 2 to 4mm. long: calyx purple-tinged ; the tube barely twice exceed-
ing the lips ; the upper lip broad-ovate, acuminate-subulate, the lower with
lance-ovate acuminate lobes : corolla 1 cm. or so long, the tube a little
exserted ; galea pilose, equalling the lip: style bearded. — S. Jlexuosa,
Seem. I. c. not Presl; Hemsl. 1. c. 556, in part, with doubt. — North-
western Mexico, in the Sierra Madre (Seemann).

X X Stems herbaceous, stout, densely puberulent in decussating bands : calyx

slightly flaring above.

79. S. fluviatilis. Stems 1.5 m. or less high, very quadrangular :
leaves rhombic-ovate, short-acuminate at tip, cuneate at base, crenate-
serrate, the upper 5 to 9 cm. long, the lowermost broader, subtruncate at
base, 1.5 dm. long, nearly as broad, puberulent above, especially on the
nerves, cinereous beneath ; petioles cinereous-puberulent, the lowest 5
cm. long, the others shorter : racemes slender, flexuous, the terminal
becoming 2 dm. long ; verticels 10-20-flowered, the lower remote, the
lowest 1 cm. apart: bracts ovate, acuminate : pedicels filiform, becoming
2 to 3 mm. long : calyx green, strongly bilabiate, the tube twice exceed-
ing the lips ; the upper lip ovate, acuminate, entire, the lower with
narrower attenuate lobes : corolla deep blue, pilose, 1 cm. long ; the tube
slightly exserted ; galea about equalling the lip : style bearded. — Mo-
RELOS, by streams near Cuernavaca, alt. 1,540 m., May 16, 1898 (0. G.
Pringle, no. 6850), Sept. 29, 1898 {E. W. D. Holway, no. 3028).

c. Calyx in anthesis 7 to 9 mm. long.
1. Stem and leaves glabrous : the latter 1 to 1.3 dm. long, serrate.

80. S. alvajaca, Oerst. Vidensk. Meddel. 1853, 38 ; Hemsl. I. c.
552. — Described from Costa Rica. No specimen seen.

FERNALD. — MEXICAN SALVIAS. 517

2. Stems and leaves pubescent, the stems glandular above.
O Primary leaves 4 to 5 cm. long, crenate.

81. S. RHOMBiFOLTA, Ruiz & Pav. Fl. Per. & Chil. i. 26, t. 36, f. b. ;
Bentli. I.e. 310; Hemsl. I.e. 564. — Doubtfully reported from Mexico
by Beuth. 1. c.

O O Primary leaves 1 dm. long, serrate.

82. S. roscida. Herbaceous ? (only the top of the plant seen) :
stem closely glaudular-puberulent above, the tip sparingly pilose : leaves
very thin, elliptic-ovate, acuminate at both ends, coarsely and irregularly
serrate, setulose-pilose above, sparingly so and minutely glandular-puber-
ulent below ; petioles slender, 5 cm. or less in length : racemes 1 to 1 .5
dm. long; the verticels G-12-flo\vered, all becoming remote, the lowest
3 to 4 cm. apart : bracts rhombic-ovate, long-acuminate, 6 to 8 mm. long :
pedicels filiform, very glandular, 1 cm. or less in length : calyx glandular-
puberulent, slightly setulose on the nerves, the long-attenuate lobes one
half shorter than the cylindric-campanulate tube : corolla 1.8 cm. long,
the tube exserted and slightly curved, glabrous, white ; the blue lower
lip somewhat exceeding the pilose galea : style included, bearded. —
Northern Mexico. Durango, Chacala, alt. 920 m., March 5, 1899
{E. A. Goldman, no. 340).

I- Generally branching from the base, the stems slender and mostly decum-
bent: racemes long-peduncled. (Forms of S. prunelloides might be looked for
here.)

++ Stems and leaves very setulose with long straight slender hairs.

= Leaves cuneate at base to short petioles or subsessile, obscurely crenate : bracts

broad-ovate.

83. S. VERONICAEFOLIA, Gray in Wats. Proc. Am. Acad. xxii. 444.
— Jalisco, on moist hillsides, Rio Blanco, June, 1886 (Edw. Palmer,
no. 28) ; hillsides near Guadalajara, July 2, 1889 (C. G. Pringle, i\o.
2555).

= = Leaves truncate at base, coarsely crenate-dentate, long-petioled : bracts

ovate-lanceolate.

84. S. oreopola. Stems repent at base, more or less assurgent,
about 5 dm. long : leaves deltoid-ovate, dark green above, pale beneath,
setulose on both surfaces, 2 to 3.5 cm. long, 1.5 to 3 cm. wide, on setulose
petioles 1.5 to 2.5 cm. long: peduncles 7 to 10 cm. long; verticels 3-6-
flowered, all becoming remote, the lowest 6 cm. apart : pedicels pubescent,
2 to 4 mm. long: calyx setulose, open-campanulate, strongly bilabiate ;
the broad purple upper lip blunt or acute, the lower with narrower

618 PROCEEDINGS OF THE AMERICAN ACADEMY.

ovate-lanceolate acute lobes : corolla blue, 2 cm. long, the slightly pubes-
cent tube twice exceeding the calyx ; the blunt pilose galea one third as
long as the broad lip: style densely bearded. — Morelos, mountain
side at 2,150 m. alt., near Cueruavaca, Aug. 9, 1898 (C. G. Pringle,
no. 7G43).

•^+ t-H. Stems and leaves glandular-pilose, hardly setulose.

85, S. villosa. Stems 2.5 to 3 dm. high, glandular-pilose below,
densely glandular-villous above and on the rhachis : leaves thick, pale
green or cinereous, deltoid-ovate, mostly blunt at tip, subtruncate at base,
entire or obscurely undulate, 1.5 to 3 cm. long, on cinereous-pubescent
petioles 0.5 to 1.5 cm. long, the upper subsessile : peduncles 0.5 dm.
long ; verticels 3-6-flowered, all becoming remote, the lowest 2.5 to 3
cm. apart : bracts ovate, acuminate, setulose : pedicels slender, 2 to 3
mm. long: calyx blue-tinged, viscid, pilose-setulose, in anthesis 6 to 7
mm, long ; the tube slightly longer than the lips ; upper lips ovate, acumi-
nate, entire, lower with two ovate acuminate lobes: corolla violet, 1.8
cm, long; the glabrous tube slightly exserted ; the puberulent galea one
third as long as the broad lip : style bearded, — San Luis Potosi, rare
in the mountains, San Miguelito, 1876 (Schaff'ner, no. 678),

+++-(■++ Stems and leaves mostly glabrous or puberulent or only sparingly

setulose.

= Upper lip of calyx tridentate.

a. Calyx-tube hispid : leaves from orbicular to ovate-triangular, 3 to G cm. long,

2 to 4 cm. broad.

86, S, GLECHOMAEFOLiA, HBK. 1. c. 290, t. 141 ; Benth. in DC. I.e.
306 ; Hemsl, 1. c. 556 ; Briq. 1. c. — South central and southern Mexico.
San Luis Potosi, alt, 1,850 to 2,460 m., 1878 (Parry & Palmer, no.
756) : Oaxaca, near Reyes, alt. 2,060 to 3,070 m„ Oct, 20, 1894 {K W.
Nelson, no. 1794). Originally from near Guanajuato,

b. Calyx glabrous or minutely puberulent : leaves narrowly ovate-triangular, 1 to

1.75 cm. long.

87, S, FoRRERi, Greene, Pittonia, i. 156. — Durango, Sierra Madre,
west of Durango, alt. 2,500 m., Sept., Oct., 1881 {A. Forrer).

= = Upper lip of calyx entire, subulate-tipped ; calyx densely white-pubescent
with long fine viscid spreading bairs.

88, S, prunifolia. Stems puberulent, decumbent, assurgent only at
the tips, 3 to 5 dm. long : leaves orbicular to obovate, rounded or acutish
at tip, round or cuneate at base, crenate-serrate, the uppermost largest,
3 to 5 cm, long, 2.5 to 3,5 cm. broad, the lowest much smaller, dull green,

FERNALD. MEXICAN SALVIAS. 519

minutely puberulent above and on the nerves beneath, on puberulent
petioles 0.2 to 1 cm. long: peduncles 7 to 10 cm. long, pubescent above
with long fine spreading hairs ; verticels 4-8-llovvered, all becoming
remote, the lowest 1.5 to 2 cm. apart: bracts narrow-ovate, acuminate :
pedicels 2 to 3 mm. long : calyx in anthesis 5 to 6 mm. long, the tube
twice exceeding the subulate-tipped lobes: corolla 1.3 to 1.5 cm. long,
the glabrous tube one half exceeding the calyx ; galea puberulent, half
as long as the dark blue lip. — Tepic, Santa Gertrudis, in the Sierra
Madre, Aug. 8, 1897 {J. N. Rose, no. 3407).

D. ArilUjiorae, Benth. Small much branched and depressed suffru-
tescent plants: the flowers borne in the axils of foliar leaves : corolla-
tube scarcely ventricose.

* Leaves oblong-linear oblanceolate or spatiilate, 0.5 to 1.5 cm. long, entire, canes-

ceiitly rough-liispid.

89. S. AXILLARIS, Moc. & Sess. in Benth. Lab. 270; Benth. in DC.
I.e. 313; Hemsl. I.e. 553; Briq. 1. c. — South central Mexico. With-
out locality (^Coulter, no. 1118): San Luis Potosi, in the mountains,
1876 {Schaffner, no. 648), alt. 1,850 to 2,500 m., 1878 {Parry & Palm-
er, no. 698) : Guanajuato, 1893 {A. Duges, no. 228 A).

» * Leaves cuneate-obovate, 5 to 7.5 mm. long, 3-toothed, appressed-setulose.

90. S. cuNEiFOLiA, Benth. Lab. 270, & in DC. I.e.; Hemsl., I.e.
555; Briq. I.e. — Southern Mexico. Hidalgo, bare hills above Pa-
chuca, alt. 2,600 m., July 18, 1898 (C. G. Pringle, no. 6905). Orig-
inally described from Oaxaca.

E. Candicantes, Benth. Shrubs : leaves oblong or lanceolate, rarely
cordate, or when ovate not cordate, beneath tomentose or densely fine
canescent, rarely pale and glabrate. (Some species approaching the Scoro-
doniae, others the Erianthae.)

* Calyx canescent short-pubescent or tomentose, not densely woolly.

t- Leaves entire or essentially 80. (Reduced forms of S. chamaedryoides may be

looked for here.)

•w Leaves sessile or subsessile, narrow-oblong, strongly revolute.

91. S. Coulteri. Branches ligneous, covered with gray-brown
bark and numerous approximate leaf-scars: leaves densely cinereous-
tomentulose with stellate hairs, strongly revolute, blunt at tip, truncate
or subcordate at base, 1 to 1.5 cm. long, 2 to 4 mm. wide : racemes 2 to
3 cm. long; the rhachis pilose with long branching hairs ; verticels about
6-flowered, slightly remote : calyx tubulai'-campanulate, glandular-pilose,

520 PROCEEDINGS OF THE AMERICAN ACADEMY.

in anthesis G to 7 mm. long ; the tube twice or thrice exceeding the
lance-subuhite teeth; upper lip tridentate : corolla 1.2 cm. long, short-
pilose, the galea about equalling the lip : style slightly bearded. —
Mexico, without locality {^Coulter, no. 1120).

++ 4H- Leaves narrowly triangular-ovate, truncate or slightly cordate at base, on
thickish winged petioles, closely and densely white-pubescent beneath.

92. S. CANDiCANS, Mart. & Gal. 1. c. 61 ; Benth. I.e. 315; Hemsl.
I.e. 554; Briq. 1. c. — Puebla, limestone hills between Tehuacan and
Esperanza, alt. 2,000 m., Dec. 21, 1895 ((7. G. Pringle, no. 6245).
Essentially the type station of the species.

++■»-+++ Leaves narrow-ovate or oblong, attenuate at base to short petioles, or the

upper subsessile.

= Calyx ovate tubular, with sliort blunt lobes, glandular-pubescent, in anthesis 5
to 6 mm. long : leaves closely white-pubescent beneath.

93. S. THYMOiDES, Benth. Lab. 255, & in DC. 1. c. 314; Hemsl. 1. c.
566; Briq. I.e. — Southern Mexico. Puebla, limestone hills between
Tehuacan and Esperanza, alt. 2,000 m., Dec. 21, 1895 (C. G. Pringle,
no. 6251) : Oaxaca, Mitla ( G. Andrieux, no. 150) ; vicinity of Coixt-
lalmaca, alt. 2,150 to 2,300 m., Nov. 12, 1894 {K W. Nelson, no. 1915).
Species based in part on Andrieux's plant.

=: = Calyx tubular-campanulate, with acuminate lobes, not glandular-pubescent,
in anthesis 6 to 8 mm. long: leaves sparingly stellate puberulent or glab-
rate.

a. Leaves oblong, 1 to 2.5 cm. long, the uppermost subsessile.
1. Leaves finely stellate-puberulent at least when young.

94. S. coahuilensis. Freely branching mostly from the stout
woody base ; branches assurgent, 0.5 to 7 dm. high ; bark pale brown,
finely stellate-puberulent or glabrate : leaves confined mostly to the lower
portions of the branches : racemes elongated, becoming 0.5 to 3.5 dm.
long, the rhachis often glandular-puberulent ; verticels mostly 2-flowered,
all remote, the lowest 2 to 3 era. apart : bracts ovate-oblong, acuminate,
caducous : pedicels filiform, 2 to 4 mm. long : calyx granulose-puberu-
lent ; tube twice exceeding the ovate-lanceolate subequal lobes ; upper
lip entire: corolla 1-5 cm. long, blue with white centre; the glabrous
tube nearly twice exceeding the calyx ; the slightly pilose galea one half
as long as the broad lower lip. — S. chamaedryoides, Wats., Proc. Am.
Acad, xviii. 137, in part, not Cav. — Coahuila, rocky hill, Saltillo,
1878 {Parry, no. 29), May, 1898 {FAw. Palmer, no. 194) ; Parras,
June, 1880, and Lerios, July, 1880 {Edw. Palmer, nos. 1072, 1064).

FERNALD. — MEXICAN SALVIAS. 521

2. Leaves glabrous.

95. S. LYCioiDES, Gray, Proc. Am. Acad. xxi. 408. — Chihuahua,
ca,nons, Santa Eulalia Mts., May 1, 1885 {C. G. Priiujle, uo. 52).

b. Leaves ovate or ovate-oblong, 3 to 7 mm. long, all slender petioled.

96. S. serpyllifolia. Erect (?), 5 or 6 dm. high, the ascending
branches brown, short-pubescent iu decussating bands with minute wliite
spreading or slightly recurved hairs ; leafy to the inflorescence, the
internodes 1 to 1.5 cm. long: racemes becoming 1 to 2 dm. long; verti-
cals 2-6-flowered, all becoming remote, the lowest 1 cm. or so apart :
bracts ovate, acuminate, ciliate, caducous : calyx short-hispidulous on
the veins; tube thrice exceeding the ovate lobes ; upper lip blunt, entire,
lower with 2 short acuminate lobes : corolla 1.2 cm. long ; the glabrous
tube one-half longer than the calyx ; the lower lip slightly exceeding
the pilose galea : style slightly bearded. — S. charnnedryoides, Wats.
Proc. Am. Acad, xviii. 137, in part, not Cav. — San Luis Potosi, with-
out locality, alt. 1,850 to 2,460 m., 1878 {Parry & Palmer, no. 751).

•*- -1- Leaves crenate or crenate-serrate. (A form of S. scorodoniaefolia may be

looked for here.)

*+ Calyx lobes normal, not conspicuously broadened.

= Stems chalky-white with ver}' close hirsute indumentum : leaves ovate, thin,
slender-petioled, minutely white stellate beneath: verticels 6-10-flowered:
calyx stellate-tomentulose ; the tube thrice exceeding the short ovate lobes ;
the upper lip blunt, erect.

97. S. CEDROSENSis, Greene, Bull. Cal. Acad. i. 212. — Lower
California, Cedros Island, March, 1889 (Edio. Palmer, no. 684),
March to June, 1897 (A. IV. Anthony, no. 294) ; Magdalena Bay, Jan.
14, 1889 (T. S. Brandegee).

= = Stems and leaves (at least beneath) more or less cinereous-tomentulose with
stellate hairs : verticels 2-6-flowered : calyx-tube twice exceeding the acu-
minate lobes.

98. S. CHAMAEDRTOiDES. Cav.^ Leaves elliptic-ovate or narrowly
deltoid-ovate, green above, slender-petioled, 1 to 2.5 cm. long. — Ic. ii.
77, t. 197 ; Sims, Bot. Mag. t. 808 ; Benth. 1. c. 314 ; Hemsl. 1. c. 554;
Briq. 1. c. S. chamaedri folia, Andr. Bot. Rep. vi. t. 416. S. Chamaedrys,

1 S. ramosissima. Similar to S. chamaedryoides, very freely branching
above : young branches slightly canescent with short spreading or somewhat re-
curved hairs : leaves paler beneath than above, more or less scabrous on both faces
with short stiff simple hairs : calyx with similar pubescence on the nerves. — 5.
chamaedrt/oides, Gray, Syn. Fl. 1. c. 371, not Cav. — Cafions of the Rio Grande, R.W.
Texas, 1849 ( C. Wright, no. 472 a) : Organ Mts., New Mexico, 1881 (G. R. Vasey).

522 PROCEEDINGS OF THE AMERICAN ACADEMY.

Willd. Hort. Berol. i. 29, t. 29, — Central Mexico. Zacatecas, plains,
La Honda Station, Aug. 18, 1890 ( (7. G. Pringle, no. 3170): San
Luis Potosi, alt. 1,850 to 2,4G0 m., 1878 {Parry & Palmer, no. 753) :
Hidalgo, bare hills above Pachuca, alt. 2,600 ft., July 18, 1898 {C. G.
Pringle, no. 6907) : Mexico, mountains near Mexico and Guadalupe,
Aug. 24, 1865 {Bourgeau, no. 855).

Var. isochroma. Leaves reduced, 0.5 to 1.5 cm. long, short-petioled
or subsessile, mostly crowded, canescent on both faces. — San Luis
Potosi, in the mountains, San Rafael and San Miguelito.1876 {Schaff-
ner, uos. 663, 664) ; without locality, alt. 1,850 to 2,460 m., 1878 {Parry
& Palmer, nos. 750, 751^).

= = = Stems witli spreading pubescence : leaves 2 to 3 cm. long; blades little

exceeding tlie petioles.

99. S. PAUCiFLORA, HBK. I.e. 303; Benth, I.e. 315; Hemsl. I.e.
563. — A doubtful species from " New Spain."

= == = = Of this section but not identified.

S. SPiCATA, R. & S. Syst. Mant. i. 202 {S. pulchella, HBK. 1. c. 288,
t. 140, not DC. S. jiomifera, Sesse & Moc. 1. c. 7, ex. char., not L.) and S.
BREViFLORA, Moc. & Sesse in Benth. Lab. 274, species described from
"New Spain" are placed by Bentham in this section. S. spicata, Gray,
Syn. Fl. 1. c. 461, is very different from the plate of S. pulchella, HBK.
*+ -M- Calyx fuunelform, the lobes becoming ampliate.

= Leaves ovate or ovate-orbicular, dark green above, much paler beneath, regularly

crenate.

100. S. BALLOTAEFLORA, Benth. Leaves rugose, densely white-tomentu-
lose beneath, 0.5 to 3 cm. long, rarely larger : racemes 0.5 to 8 cm. long.
— Lab. 270, & in DC. I.e. 313 ; Torn Bot. Mex. Bound. 131 ; Gray,
1. c. ; Hemsl. 1. c. 553 ; Briq. 1. c. S. laxa, Benth. 1. c. — Texas to San
Luis Potosi. Mexico, without locality {Gregg, no. 322) : Coahuila,
Monclova, Aug., 1880, Parras, June, 1880, Lerios, July, 1880, Saltillo,
May, 1898 {Edw. Palmer, nos. 1069, 1067, 1068, 107): Tajiaulipas,
1843 {Berlandier, no. 3186): San Luis Potosi, San Rafael, 1876,
{Schaffner, no. 670) ; near Matehuala, June 18, 1898 (E. IF. Nelson, no.
4527).

Var. Eulaliae. Leaves large, 3 cm. long, broad-ovate, rugose, very
slightly pubescent and hardly canescent beneath : racemes very numerous,
4 to 9 cm. long. — Chihuahua, Santa Eulalia Mts., Sept. 1, 1885
{C. G. Pringle, no. 659).

FERNALD. — MEXICAN SALVIAS. 523

Var. pinguifolia. Leaves large, as ia the last, scarcely rugose, very
pale beneath with a close very minute indumeutuni, not tomentulose,
greasy to the touch : racemes all many-flowered. — Neav Mexico,
1851-52 ((7. Wright, no. 1524): Arizona, San Francisco Mts., back
of Clifton, Sept. 3, 1880 ( E. L. Greene, no. 300).

= =: Leaves very pale on both surfaces, thick, hardly rugose, greasy to the

touch, irregularly crenate.

101. S. PLATYCHEILA, Gray, Proc. Am. Acad. viii. 292, & Syn. Fl.
I.e. — Lower California, Carmen Island, 1870, Nov. 1890 {Edw.
Palmer, nos. 7, 878).

* * Calyx densely woolly : verticels in short spiciform racemes : lip of corolla
distinctly exceeding the galea (otherwise approaching the Eriantliae).

102. S. confinis. Branches canescent with fine stellate pubescence,
leafy to the inflorescence : leaves oblong, blunt or acutish, rounded-trun-
cate or subcuneate at base, very short-petioled, thick and rugulose
especially along the closely crenulate margin, closely canescent on both
surfaces, or ferrugineous-tinged beneath, 1.5 to 4.5 cm. long, 0.5 to 2 cm.
wide : spiciform racemes simple or slightly paniculate, 1 to 4 cm. long,
lowest verticels slightly remote: bracts large, broad-ovate, acuminate,
somewhat persistent, 0.5 to 1 cm. long, stellate-tomentose : calyx dark
blue, mostly hidden by dense white-lanate pubescence : corolla blue,
scarcely 1 cm. long. — S. spicata, Gray, 1. c. 461, not R. & S. — Southern
Arizona and adjacent Mexico. Arizona, mountain pass near Fort
Huachuca, 1882 (^Lemmon, no. 2861) : Sonora, Fronteras, alt. 1,400
m., Sept. 25, 1890 {C. V. Hartman, no. 43).

F. Scorodoniae, Benth. Shrubs with ovate (in S. thyrsijlora ovate-
lanceolate) rugose leaves generally cordate, rarely cuneate, at base.

* Leaves white-tomentose beneath with simple hairs.

■*- Leaves very rough-rugose above, mostly broad-ovate (except in variety of the
first species) with obtuse or blunt tips.

++ Pedicels very short, at most 3 mm. long : racemes rather dense, often branch-
ing : calyx somewhat cuneate-carapanulate, in anthesis 4 to 5 mm. long, with
short flaring obtuse lobes.

= Pubescence of calyx glandular-villous.

103. S. scoRODONiAEFOLiA, Poir. Leaves ovate, subcordate. —
Suppl. V. 46: Benth. I.e. 316; Hemsl. I.e. 565; Briq. I.e. S.
rnelissodora, Lag. Gen. & Spec. Nov. 2. S. hirta, Schranck, Syll. PI.
Soc. Eatisb. ii. 60 ? according to Benth. S. scorodonin, Benth. Lab.
264. — Central and southern Mexico. San Luis Potosi, alt. 1,850

524 PROCEEDINGS OF THE AMERICAN ACADEMY.

to 2,460 m., 1878 {Parry & Palmer, no. 730) : Mexico, Guadalupe
June 21, 18G5 {Bourgeau, no. 295) : Oaxaca, Valley of Cuicatlan,
alt. 2,000 to 2,460 m., Nov. 10, 18D4 {E. W. Nelson, no. 1898) ; San
Juan del Estado, alt. 1,800 m., Oct. 20, 1895 {L. C. Smith, no. 930) ;
Piipalo, Cuicatlan, alt. 1,650 m., Dec. 9, 1895 (V. Gonzalez, no. 42).

Var. crenaea. Leaves narrower, oblong-lanceolate, blunt, rounded
or subcuneate at base. — *S^. scorodonia, Benth. PL Hartw. 20, not Benth.
Lab. 264. — A northern narrow-leaved extreme. Aguas Calientes
{Hartweg, no. 164): Chihuahua, southwestern section, without locality,
1885 {Ediv. Palmer).

= = Pubescence of calyx long-villous, not glandular.

104. S. LASIANTHA, Benth. Lab. 276, & in DC. 1. c. 321 ; Hemsl.
1. c, 559 ; Briq. 1. c. — Range of the last from which it differs only in
the more villous glandless calyx. Perhaps only a variety of that.
D.URANGO, Ramos to Inde, Aug., 1898 (E. W. Nelson, no. 4679) : San
Luis Potosi, in the mountains, San Rafael, 1876 {Schaffner, no. 671);
without locality, alt. 1,850 to 2,460 m., 1878 {Parry & Palmer, no.
731) ; limestone ledges, San Jose Pass, July 23, 1890, and limestone
hills, Las Canoas, Oct. 8, 1890 {0. G. Pringle, nos. 3206, 3273):
Hidalgo, calcareous hills near Tula, alt. 2,160 m., Oct. 5, 1896 (C G.
Pringle, no. 6538) : Oaxaca, near Dominguillo, alt. 2,000 m., Oct. 3,
1894 {E. W.Nelson, no. 1595). Placed by Bentham and by Briquet in
the section Erianthae.

++ -M- Pedicels longer, 8 to 9 mm. long: racemes loosely flowered; verticels re-
mote : calyx tubular-campanulate, in anthesis 6 mm. or more long.

= Calyx purplish-violet, glandular-villous, in anthesis nearly 1 cm. long : lower
lip of corolla almost black ; tube and galea pale.

105. S. SEMIATRATA, Zucc. Abhandl. Baier. Akad. Wiss. i. 298;
Benth. in DC. I.e. 316; Hemsl. 1. c. 565; Briq. I.e. — Oaxaca, with-
out locality, June {Andrieux, no. 149); calcareous hills. Las Sedas, alt.
1,850 m., Aug. 2, 1894 {C. G. Pringle, no. 4763), alt. 2,150 m., June
27, 1895 {L. C. Smith, no. 413); near city of Oaxaca, alt. 1,600 to
2,100 m., Oct. 2, 1894 {E. W. Nelson, no. 1511).

= = Calyx green (or blue-tinged), glandular-puberulent and minutely hispidulous
on the nerves, in anthesis about 6 mm. long : coralla blue.

106. S. Gonzalezii. Branches very slender, somewhat villous, the
youngest parts and the rhachises glandular-puberulent : leaves broad-
ovate, bluntish at tip, truncate or subcordate at base, green and very
rugose, slightly pubescent above, white-villous beneath, crenulate, 1 to

FERNALD. — MEXICAN SALVIAS. 625

2.5 cm. long, 0.75 to 1.75 cm. broad, ou villous petioles 1 cm. or less in
length: racemes simple, 0.5 to 1.5 cm. loug ; verticels 3-8-flowered,
all remote, the lowest 2 to 2.5 cm. apart: pedicels 3 to 5 mm. loug,
glaudular-piiberuleut : calyx tubular-campanulate ; the tube twice or
thrice exceeding the ovate mucronate lobes ; upper lip entire : corolla
deep blue, 1.5 cm. long; the glabrate tube twice exceeding the calyx;
the pilose galea slightly exceeding the dark lip : style bearded. —
Oaxaca, El Pariau-Etla, alt. 370 m., Nov., 1898 ( V. Gonzalez & C.
Conzatti, no 903).

-(- t- Leaves less rugose, narrow-ovate, mostly acuminate at tip.
■w- Flo\v.ers in small cymes forming a terminal thyrsus : leaves ovate-lanceolate.

107. S. THYRSiFLORA, Benth. Bot. Sulph. 151, & in DC. I.e.;
Hemsl. 1. c. 566 ; Bricj. 1. c. — Southwestern Mexico, Tepic to Michoacan.
Jalisco, vicinity of Mascota and San Sebastian, alt. 1170 to 1540 m.,
March, 1897 {E. W. Nelson, nos. 4057, 4081): Michoacan, mountains
near Patzcuaro, Nov. 10, 1890, Dec. 21, 1891 (C G. Pringle, uos.
3593, 4097). Originally from Tepic.

++ -M- Flowers in simple or more or less branclied generally paniculately or virgately

disposed racemes.

= Branches short-pilose with recurved spreading hairs.

a. Calyx tubular, in fruit 5 mm. long, the tube four times exceeding the very

short broad lobes.

108. S. ALAMOSANA, Rose, Contrib. U. S. Nat. Herb. i. 110. —
SoNORA, Sierra de los Alamos, 1890 {Edw. Palmer, no. 345).

6. Calyx campanulate, in fruit 6 to 7 mm. long, the tube twice exceeding the

narrower lobes.

109. S. CHAPALENSis, Briq. Ann. Conserv. Jard. Bot. Geneve, ii.
145. — Jalisco, rich wooded carious, mountains near Lake Chapala,
Nov. 22, 1892 (C. G. Prhujle, no. 4351).

= = Branches pubescent with ascending soft hairs.

110. S. multiramea. Stems about 1 m. high, freely paniculate-
branched, the upper branches leafless and bearing racemes : leaves
narrow-ovate, acuminate, rounded or subcordate at base, somewhat rugose
and scabrous-hispidulous above, densely tomentose beneath, short-petioled,
2 to 5 cm. long, 1 to 2.5 cm. broad : inflorescence paniculate ; lower
branches leafy and axillary, upper naked ; terminal racemes 1.2 dm. or
less in length, the others shorter; verticels 2-12-flowered, all becoming
remote : bracts lance-subulate, minute, early deciduous : pedicels 2 to

526 PROCEEDINGS OF THE AMERICAN ACADEMY.

4 mm. long, spreading : calyx tubular, in antbesis 3.5 to 4 mm. long,
appressed-i)ilose ; the tube 3 or 4 times as long as the short blunt lobes :
corolla blue, 1 to 1.2 cm. long; the glabrous tube one half longer than
the calyx ; the pilose galea twice exceeded by the exteriorly j^ilose lip :
style bearded. — Southern Mexico. Oaxaca, near city of Oaxaca, alt.
2,300 to 2,920 m., Sept., 1894 {E. W. Nelson, no. 1448), alt. 1,850 m.,
Oct. 25, 1894 (f, G. Pringle, no. 6013); near Reyes, alt. 1,700 to
2,060 m., Oct. 20, 1894 {E. W. Nelson, no. 1783) ; mountains of Jaya-
catlan, alt. 1,300 m., Nov. 4, 1894, alt. 2,000 m., Oct. 18, 1895 (Z. C.
Smith, nos. 272, 860) ; near Chilpanciugo, alt. 2,770 to 3,140 m., Dec.
24, 1894 {E. W. Nelson, no. 2235) ; Chiapas, without locality ( Ghies-
breght, no. 764).

* * Leaves, at least when young, pale beneath with stellate hairs.

1- Calyx densely invested with close stellate indumentum, not long woolly (ex-
cept at the base in no. 114).

++ Leaves 1 to 1.5 cm. long.

111. S. FRDTicuLOSA, Benth. Lab. 721, & in DC. I.e. 315; Hemsl.

I.e. 556; Eriq. in Engl. & Prantl, I.e. — Oaxaca, near city of Oaxaca

{Andricux, no. 151).

++ -^ Leaves 2 to 6 cm. long.

= Branchlets and nerves of young leaves with sulphur-yellow indumentum.

112. S. Conzattii. Branching above, the younger parts densely
stellate-tomentulose : leaves ovate, 1.5 to 6 cm. long, 0.75 to 3.5 cm.
broad, rounded at tip, rounded-truncate or subcuneate at base, dark
green and very rugose above, pale and very strongly reticulate-rugose
beneath; margin finely crenulate ; petioles 1.5 cm. or less in length: ra-
cemes spiciform, 2 to 6 cm. long, at first dense, the 3-many-llowered
verticels later becoming a little remote, all the j^arts more or less red-
punctate: bracts ovate-acuminate, exceeding the calyx, white-tomentose
at base, yellowish at tip, soon deciduous : calyx ovate-campanulate,
densely tomentose with pale blue freely-branching hairs, with very short
blunt inconspicuous lobes: corolla 1.2 cm. long, blue, red-punctate and
whitened above with stellate tomentum ; tube somewhat exserted ; the
galea hardly equalling the lip: style bearded. — Oaxaca, El Parian-
Etla, alt. 370 m., Nov. 1898 ( V. Gonzalez & C.Gonzatti, no. 902).

= = Branchlets and lower faces of leaves with white indumentum.
a. Leaves broad-ovate, sliarply crenate-dentate : calyx-lobes blunt.

113. S. pruinosa. Rather stout shrub with pale brown bark, the
younger parts densely pruinose with stellate hairs : leaves 2.5 to 5 cm.

FERNALD. — MEXICAN SALVIAS. 527

long, 1.5 to 3.5 cm. wide, blunt at tip, rounded or subcuneate at base,
green very rugose and glabrate above, very white beneath, irregularly
crenate-deutate ; petioles 1.25 cm. or less in length: racemes 1 dm. or
less long; verticels many-flowered, rather crowded, the lowest 1 to 1.5
cm. apart; all the parts red-punctate; rhachis pruinose-pubescent : bracts
ovate-acuminate, bluish, finely stellate-pubescent : calyx ovate-campauu-
late, closely pruinose-pubescent, bluish-tinged, in anthesis 4 to 5 mm.
lon<r; the tube twice or thrice exceedins; the deltoid blunt lobes : corolla
1.2 cm. long, pruinose-stellate ; the tube one half longer than the calyx;
the galea about equalling the lip : style bearded. — Jalisco, road between
Mesquitec and Monte Escobedo, Aug. 26, 1897 (J. N. Rose, no. 2601).

b. Leaves oblong-ovate, bluntly crenate : calyx-teeth subulate.

11 4. S. Goldraanii. Stem stout, 6 or 7 dm. high, herbaceous, or
fruticose at base, puberulent, with numerous ascending branches above :
leaves dull-green, slightly rugose above, cinereous and red-punctate be-
neath, 1 dm. or less long, bluntish at tip, rounded-cuneate to petioles
2 cm. or less in length: racemes 0.5 to 1.5 dm. long; verticels 6-12-
flowered, the lower 2 cm. apart, the upper approximate ; rhachis lanate :
bracts lance-ovate, caducous: calyx in anthesis 5 to 6 mm. long, lanate
at base, the deltoid subulate-tipped lobes one half as long as the tube :
corolla red-punctate, 1.5 cm. long; the tube somewhat exserted, glabrous
below ; the lips stellate-pilose without, the lower exceeding the galea :
style barely exserted, bearded. — Northwestern Mexico. Chihuahua,
near Batopilas, alt. 1690 to 2000 m., Oct. 4, 5, 189S {E. A. Goldman,
no. 214).

-^ -4- Calyx blue, invested in long white wool : leaves becoming glabrate, conspic-
uously red-punctate beneath.

115. S. RUBROPUNCTATA, Robinsou & Fernald, Proc. Am. Acad,
XXX. 121. — SoNORA, iu the canon, Huehuerachi, Dec. 12, 1890 {F. E.
Lloyd, no. 451).

* * * Leaves glabrous, or the youngest slightly puberulent on the nerves

beneath.

116. S. Nelsonii. Shrub with smooth reddish-brown bark, the
younger branchlets slightly puberulent : leaves ovate, blunt or acutish at
tip, rounded at base, 1.5 to 5 cm. long, 1 to 3 cm. broad, dull green and
obscurely rugose above, paler beneath, crenate-serrate: petioles slender,
puberulent, 1 cm. or less in length : inflorescence more or less paniculate ;
the racemes 1 dm. or less long, the upper terminating leafless branches ;
verticels 6-12-flowered, becoming a little remote, the lowest 1 to 2 cm.

528 PROCEEDINGS OP THE AMERICAN ACADEMY.

apart : calyx cuneate-carapauulate, resembling that of aS^. scorodoniaefolia,
densely short white-villous, iu anthesis 5 to 6 mm. long ; the tube three
or four times exceeding the short blunt lips : corolla blue, pilose, 1.2 cm.
long; the tube included or barely exserted; the lip slightly exceeding
the galea: style bearded. — Puebla, between Acatlan and Piaxtla, alt.
1,230 to 1,380 m., Nov. 22, 1894 {E. W. Nelson, no. 2005). The calyx
has a strong ginger-like fragrance.

* * * * Leaves pubescent beneath : plant not identified.

117. S. Keerlii, Benth. Lab. 263, & in DC. I.e. 316, with ovate
cordate very rugose leaves canescent-tomeutose beneath, and with dense
paniculate-branched racemes, the calyx pilose-hirsute, not canescent-
tomentose, described from Tlalpujahua, Michoacan, and from Regla.

G. OordifoUae, Benth. Tall herbs or shrubs (ours herbs) with broad
cordiform herbaceous leaves green or pale beneath and cordate at base,
at least the lowermost long-petioled. (^S. purpurasce7is & S. Martensii
might be looked for here.)

* Stems pilose-hispid : leaves crenate, pale beneath : calyx pilose-hispid, more or

less glandular.

118. S. amarissima, Ort. Dec. 4; Edw. Bot. Reg. iv. t. 347 ; Benth.
1. c. 317 ; Hemsl. 1. c. 553 ; Briq. 1. c. 280. S. circinata, Cav. Ic. iv. 9, t.
318, S. amara, Jacq. Hort. Schoenb., iii. 2, t. 255. S. hirsuta, Sesse
& Moc. 1. c. 8. S. amethystina, Donnell Smith, Enum. PI. Guat. i. 35,
ii. 62, in part, not Smith. — S. cyanea, Donnell Smith 1. c. iii. 67, iv. 125,
not Benth. — A common species from Central Mexico to Central
America and Panama. Without locality (^Uhde, no. 791): San Luis
POTOSI, alt. 1,850 to 2,460 m., 1878 {Parry & Palmer, no. 752):
Hidalgo, hills above Pachuca, alt. 2,550 m., July 18, 1898 ((7. G.
Pringle, no. 6914) : Mexico, Guadalupe, Aug. 17, 1865 (Bilimek, no.
304); Valley of Mexico, May 11, 1865 (Bowgeau, no. 125, iu part),
Sept. 19, 1889 (C. G. Pringle, no. 2817); Rio Hondo Canon, Aug. 22,
1890 (C G. Pringle, no. 3153) : Michoacan, Tlalpujahua {Graham) :
Oaxaca, Monte Alban, alt. 1,790 m., July 20, 1894 (Z. C. Smith, no.
171) ; Valley of Oaxaca, alt. 1,540 to 1,790 m., 1894 {E. W. Nelson,
nos. 1103, 1289) : Guatemala, Coban, Dept. Alta Verapaz, alt.
1,320 m., July, 1885, and S;uita Rosa, Dept. Baja Verapaz, alt. 1,540 m.,
July, 1887 {H. von Tuerckheim iu exsicc. J. D. Smith, nos. 110. 1192);
San Lucas, Dept. Zacatepequez, alt. 1,700 m., April, 1890 {J. D. Smith,
no. 2187); Estanzuela, Dept. Santa Rosa, alt. 770 m., Aug., 1892,
Santa Rosa, alt. 925 m., May, 1893, and Laguna de Ayarza, Dept. Ja-

FERNALD. — MEXICAN SALVIAS. 629

lapa, alt. 2,460 m., Sept., 1892 {Heyde & Lux in exsicc. J. D. Smith,
nos. 4050, 4564, 4056 ; Panama, Gatuu {Sutton Hayes).

* * Stems puberulent above and slightly short-pilose on the angles, densely woolly
at the nodes : leaves crenate-serrate : calyx puberulent and minutely pilose.

119. S. SIDAEFOLIA, Mart. & Gal. I.e. 67; Beiith. I.e. 318; Hemsl.
1. c. 565 ; Briq. 1. e. — Oaxaca, wet ravines, Sierra de San Felipe, alt.
2,600 m., Jan. 3, 1895 {C. G. Pringlc, no. 5643). Originally eoUected
ill the eordillera of Oaxaca.

H. Rudes, Benth. Herbs with ovate leaves cordate at base, subsessile
or very short-petioled.

* Bracts ovate-lanceolate, 1 to 1.5 cm. long, firm and persistent : calyx in anthesis

1 cm. long : corolla crimson, 3 cm. long : plant weak, pilose-hispid, the leaves
firm and lucid.

120. S. SESSiLiFOLiA, Gray in "Wats. Proc. Am. Aead. xxii. 445;
Briq. Aim. Couserv. Jard. Bot. Geneve, ii. 152. — Jalisco, bottom of
ravine, Rio Blaneo, July, 1886 (Edw. Palmer, no. 184) ; ravines near
Guadalajara, July 1, 1889 ( C. G. Pringle, no. 3058). In its corolla
approaching the Fulgentes.

* * Bracts minute, caducous.

-t- Stems simple or subsimple from woody base, more or less pilose : leaves green
on both faces, blunt or obtuse : calyx in fruit 7 to 8 mm. long.

121. S. PLATTPHYLLA, Briq. I.e. 50. ^S". aman'ssima, Wats. I.e.,
rot Ort. — Jalisco, hills, Rio Blaneo, July, 1886 (Edw. Palmer, no.
183) ; hillsides near Guadalajara, July 3, 1889 ( C. G. Pringle, no. 2560).

I- -t- Stem freely branching above, pilose with somewhat recurved hairs : leaves
canescent, pilose beneath, acuminate : fruiting calyx about 5 mm. long.

122. S. NEPETOiDES, HBK. I.e. 299, t. 150. -S. am«msi/?ia, Benth.
I.e. 317, as syn. ; Hemsl. 1. e. 553, as syn. ; not Ort. — South central
Mexico. MoRELOS, mountain side near Cuernavaea, alt. 2,160 m., Aug.
9, 1898 (C. G. Pringle, no. 7612). Originally from Guanajuato.

§ 4. Erianthae, Benth. Shrubs generally with rugose leaves (not
in *S'. popidifolia). Calyx densely lauate. Lips of the corolla subequal
or the galea longer ; the tube a little exserted, straight, ventrieose or
enlartred above.

* Pubescence of calyx purple or rose-colored.

-I- Leaves lanceolate or linear-lanceolate, green above, white-lanate beneath : ra-
ceme elongated, 4 dm. or less in length : corolla white, lanate.

123. S. leucantha, Cav. Ic. i. 16, t. 24; Hook. Bot. Mag. t. 4318;
Benth. I.e. 321; Hemsl. I.e. 560; Briq. in Engl. & Prantl, 1. c. S.

VOL. XXXV. — 34

530 PROCEEDINGS OF THE AMERICAN ACADEMY.

bicolor, Sesse & Moc. 1. c. 8. — Coahuila to Oaxaca, often cultivated.
Without locality {Graham): Coahuila, garden at Saltillo, 1848-49
(Greffff, uo. 507) : Zacatecas, in plaza, Moute Escobedo, Aug. 27,
1897, and on the Sierra de los Morones, near Plateado, Sept. 1, 1897
(J. N. Rose, nos. 2641, 3627) : San Luis Potosi, near Santa Maria,
1876 {Sdiaffner, no. 669).

•t- -1- Leaves ovate, subcordate: corolla reddish.
++ Leaves very rugose, 2.5 to 4 cm. long, on short petioles 4 to 8 mm. long.

124. S. lantanakfolia, Mart. & Gal. I.e. 69; Benth. I.e. 322;
Hemsl. I.e. 559; Briq. I.e. — Puebla, Atlixco, July 25-Aug. 1, 1893
{E. W. Nelson). Originally from Puebla. Seaton's no. 500 from
Maltrata, Veua Cruz, is near this species but not satisfactorily referable
to it.

t-c ++ Leaves hardly rugose, 0.5 to 1 dm. long, on petioles 1.5 to 2.5 cm. long.

1 25. S. populifolia. Shrub with gray bark : branchlets and petioles
lanate with yellowish white pubescence: leaves ovate, bluntly acuminate,
crenulate, green and puberulent above, white-velutinous beneath : raceme
1.5 dm. or less in length, simple or with a tendency to branch ; verticels
about 6-fiowered, becoming a little remote : bracts ovate, acuminate, 1 cm.
long, purple-lanate without, soon deciduous : calyx campanulate, in an-
thesis 1.8 cm. long, the tube twice or thrice exceeding the deltoid acu-
minate lobes ; upper lip entire, lower 2-lobed : corolla pilose, reddish,
3.5 cm. long, the galea slightly exceeding the lip : long-exserted por-
tion of style glabrous. — Jalisco, Bolanos, Sept., 1897 (J.N. Rose,
no. 2862).

* * Pubescence of dorsal and ventral portions of calyx dark red, of tlie sides bril-

liant yellow : corolla dark red ; the lip much shorter than the galea : leaves
ovate4anceolate, acuminate at base, very rugose above.

126. S. PiTTiERi, Briq. Bull. Soc. Bot. Belg. xxx. pt. 1, 237. De-
scribed from Costa Rica, and doubtfully referred by its author to this
group.

§ 5. Macrostachyae, Benth. Pubescent large-leaved shrubs or
herbs. Racemes dense, thick, with large thick firm leaf-like j^ersistent
bracts. Corolla blue or white, twice or thrice as long as the calyx.

* Villous herb: leaves ovate, rounded or subcuneate to a winged petiole : verti-

cels very many-flowered, somewhat remote.

127. S. clinopodioides, IIBK. 1. c. 294, t. 145 ; Benth. I.e.; Hemsl.
I.e. 554; Briq. I.e. S. parquerensis, Moc. & Sess. in Benth. Lab. 273;

FERNALD. — MEXICAN SALVIAS. 531

Benth. in DC. I.e.; Hemsl. I.e. 562; Briq. I.e. S. patzcuarensis,
Sess. & Moc. PI. Nueva Espana, ed. 2, 8. — Sierra Madre of Western
Mexico. Chihuahua, 100 km. S. of Guadalupe y Calvo, alt. 2,300 to
2,600 m., Aug., 1898 (E. W. Nelson, no. 4807). Michoacan, sandy
fields, hills of Patzcuaro, Oct. 11, 1892 (C. G. PringJe, no. 4258).
Originally collected by Humboldt & Bonpland near Patzcuaro.

* * Canescent-puberulent shrub : tlie firm leaves pale beneath : corolla pale blue

or whitish.

128. S. Shannoni, Donnell Smith, Bot. Gaz. xix. 256. — Southern
Mexico and Central America. Chiapas, top of ridge back of Tonala,
alt. 370 to 770 m., Aug. 10, 1895 {E. W. Nelson, no. 2898) : Guate-
mala, Volcan Chingo, Dept. Jutiapa, alt. 925 m., Oct., 1892 (IF". C.
Shannon in exsicc. J. D. Smith, no. 3612); Chupadero, Dept. Santa
Kosa, alt. 1,540 m,, Oct., 1892 {Heyde & Lux in exsicc. J. D. Smith no.
4052). The Chiapas plant has broader leaves and bracts than those
from Guatemala, but they undoubtedly belong together.

§ 6. LoxGiFLORAE, Benth. Corolla from 2 to 12 cm. long, of vari-
ous culors. (See note under § Brachyantliae.)

A. Lufeae, Benth. Shrubs with axillary or racemose flowers : corolla
yellow, ventricose below the throat.

* Flowers axillary, solitary: corolla 2 to 3 cm. long: leaves cordate-ovate, 2.5
cm. or less long, very rugose, canescent-tomentose beneath.

129. S. ASPERA, Mart. & Gal. I.e. 71; Benth. in DC. I.e. 327;
Hemsl. 1. c. 553 ; Briq. 1. c. — Southern Mexico. Puebla, dry calcare-
ous hills, Tehuacan, alt. 1,700 m. [the original station of Galeotti], Nov.
27, 1895 (C. G. Pringle, no. 6240) : Oaxaca, Jayacatlan, alt. 1,320 m.,
June 3, 1894, and Nov. 18, 1895 {L. C. Smith, nos. 17, 897) ; near
Dominguillo, alt. 2,000 ra., Oct. 3, 1894 (E. W. Nelson, no. 1587).

* * Flowers racemose.

f- Calyx invested with yellow or yellow and reddish tomentum : corolla yellow,
red within, 2 to 3 cm. long: leaves rugose, canescent-tomentose beneath.

130. S. chrysantha. Mart. & Gal. 1. c. 72 ; Benth. 1. c. 326 ; Hemsl.
I.e. 554; Briq. I.e. — Southern Mexico and Central America. Guer-
rero, between Tlapa and Tlaliscatilla, alt. 1,200 to 1,380 m., Dec. 5,
1894 (E. W. Nelson, no. 2049) ; Oaxaca, Monte Alban, alt. 1,850 m.,
Sept. 4, 1894 (C. G. Pringle, no. 4862), Oct. 11, 1895 {L. G. Smith,
no. 950); Valley of Oaxaca, alt. 1,540 to 1,630 ra., Sept. 25, 1894
{E. W. Nelson, no. 1244): Chiapas, without locality {Ghiesbreght, no.

632 PROCEEDINGS OF THE AMERICAN ACADEMY.

765) ; between San Cristobal and Teopisca, alt. 2,060 to 2,600 m., Dec.

4, 1895, and between Hacienda Juucana and San Vincente, alt. 1,300 to
1,850 m., Dec. 12, 1895 {E. W. Nelson, nos. 3429, 3507).

■*- -1- Calyx glandular-puberulent : corolla 3.5 to 4 cm. long: leaves cordate, acu-
minate, the lowest often 3 dm. long.

131. S. MADRENSis, Seem. Bot. Herald, 327, t. 70; Hemsl. 1. c. 560.
— Described from the Sierra Madre of northwestern Mexico. Not
seen.

B. Injlatae, ^enth. Shrubs: calyx infiated-companulate, brightly col-
ored : corolla scarlet ; the tube subventricose. (Calyx and corolla by
rare exception white.)

* Leaves ovate-lanceolate to oblong, cuneate or narrowed at base.
1- Calyx and corolla glabrous or merely puberulent : style glabrous.

132. S. Sessei, Benth. Lab. 288, & in DC. 1. c. 332 ; Hemsl. 1. c.
565; Briq. I.e. 282. S. Roezli, Scheidw. Fl. des Serres, xiv. 31, t.
1407. S. semperjlorens, La Llave in La Nat. vii. Apend. 81, ex char.

5. fastuosa, Sesse «&. Moc. 1. c. 7. Rhodochlamys speciosa, Schauer, Lin-
naea, xx. 707. — Southern Mexico. Morelos, Cuernavaca, Nov. 14,
1865 (Botirgeau, no. 1238).

-1- -1- Calyx and corolla glandular-pilose : style pilose.

133. S. "Wagneriana, Polak. Linnaea, xli. 591. — Described from
Costa Rica. Not seen.

* * Leaves ovate-lanceolate, cordate at base : stems viscid-hirsute.

134. S. ADGLUTiNANS, Lag. Gen. & Spec. Nov. 2; Benth. I.e.;
Hemsl. I.e. 552 ; Briq. 1. c. — Described from " New Spain." Not seen.

* * * Leaves ovate, acuminate, rounded at base, white-tomentose beneath, 1 dm.
or less in length : branches tonientose : shrub or small tree ( 1 to G m. high).

135. S. PUBESCENS, Benth. Lab. 723, & in DC. 1. c. ; Hemsl. 1. c. 563 ;
Briq. 1. c. — Southern Mexico. Morelos, near Cuernavaca, Oct. 27,
1895, and Nov. 9, 1895, calyx and corolla white (C. G. Pringle, nos.
7080, 7065) : Oaxaca, San Dionicio (A?idrieux, no. 143) ; Jayacatlan,
alt. 1,380 m., Sept. 10, 1894 (L. 0. Smith, no. 170) ; Monte Alban, alt.
1,790 m., Oct. 4, 1894 (C. G. Pringle, no. 4960) ; near Oaxaca, alt. 1,600
to 1,930 m., Oct. 2, 1894, near Huajuapam, alt. 1,470 to 2,000 m., Nov.
19, 1894, and near Tlapancingo, alt. 1,850 to 2,460 m., Dec. 7, 1894
{E. W. Nelson, nos. 1972, 1521, 2091).

FERNALD. — MEXICAN SALVIAS. 533

* * * * Leaves deltoid-ovate to orbicular, coriaceous, pubescent on the nerves
beneath, 4 cm. or less long : branches slender, purplish, glabrous or glabrate.

136. S. Regla, Cav. Ic. v. 33, t. 455 ; Beuth. Lab. 288, & in DC.
I.e., 333; Lindl. Bot. Reg. xxvii. t. 14; Hemsl. I.e. 564; Briq. I.e.
S. deltoidea, Pars. Syn. i. 28. S. negla, Pers. 1. c. as syn. S. crenafa,
Mart. & Gal. 1. c. 74 ; Benth. 1. e. 348 ; Hemsl. 1. c. 555 ; Briq. Ann.
Conserve. Jard. Bot. Geneve, ii. 157. — Central and southern Mexico.
Without locality {Coulter, no. 1099, Gregg, no. 340) : Coahuila, Lerios,
July, 1880 (Edw. Palmer, no. 1079) ; shaded gulches of limestone hills,
Carneros Pass, Sept. 1, 1889 ( C G. Pringle, no. 2384): Durango,
Santiago Papasquiaro, 1896 {Edw. Palmer, no. 404), Aug. 7, 1898
{E. W. Nelson, no. 4663) : San Luis Potosi, in woods, San Rafael,
1876 {Schaffner, no. 665) ; without locality, 1878 {Parry & Palmer, no.
741); San Luis Potosi, 1898 {Edw. Palmer, no. 726): Aguas Calien-
TES, by brooks {Hartweg, no. 165): Guanajuato, 1880 {A. Dnges):
Oaxaca, near Coi.xtlahuaca, alt. 2,150 to 2,300 m., Nov. 12, 1894
{E. W. Nelson, no. 1917).

C. Fulgentes, Benth. Shrubs, rarely herbs : calyx tubulose-campan-
ulate : corolla scarlet or crimson ; the tube straight, ventrieose.

* Freely branching shrubs (rarely herbs) with leaves 1 to 4 cm. long.

■t- Flowers axillary.

137. S. disjuncta. Branches ferrugineous-pubescent: leaves cordate,
ovate, blunt or acuminate, 4 cm. or less long, rugose, crenulate, dark
green and hispid above, canescent and pilose beneath, on slender petioles :
verticels mostly 2-flowered, in the axils : pedicels 6 mm. or less in length,
usually twisted: calyx pilose on the nerves, in anthesis 1.2 to 1.5 cm.
long; the tube twice exceeding the ovate acuminate lobes; upper lip
entire : corolla 3 to 4 cm. long, minutely pilose, the ventrieose tube
twice exceeding the calyx ; galea and lip subequal : style bearded. —
Chiapas, among rocks, Aug., Sept. {Ghiesbreght, nos. 76,753); near
San Cristobal, alt. 2,150 to 2,460 m., Sept. 18, 1895 {E. W. Nelson, no.
3166). A unique species.

t- -I- Flowers in terminal racemes.
■w- Leaves ovate or broad oblong.
= Pubescence of branches and leaves simple.

a. Calyx liispidulous on the nerves.
1. Leaves glabrous or at least glabrate beneath.

138. S. Grahami, Benth. Bot. Reg. xvi. t. 1370, & in DC. 1. c. 335 ;
Lodd. Bot. Cab. xviii. t. 1798; Ortgies, Gartenfl. vii. t. 242; Hemsl.

534 PROCEEDINGS OF THE AMERICAN ACADEMY.

I.e. 557; Briq. in Engl. & Prantl I.e. aS*. microphylla, Gray, Proc.
Am. Acad. xxi. 407, in part, not Benth. — Northern and central
Mexico, not clearly separable from S. microjihylla. Chihuahua, 100
km. south of Guadalupe y Calvo, alt. 2,300 to 2,620 m., Aug., 1898
{E. W. Nelson, no. 4812^); Batopilas, alt. 1,690 to 2,000 m., Oct. 1898
{E. A. Goldman, no. 212) : Coahuila, hillside, Saltillo, May, 1898 {Edw.
Palm-r, no. 161): Durango, by stream, Durango, Nov. 1896 {Edw.
Palmer, no. 844) ; near El Salto, July 12, 1898, and in the Sierra
Madre, 25 km. north of Guanacevi, alt. 2,300 to 2,620 m., Aug. 17, 1898
{E. W. Nelson, nos. 4557, 4755) : Zacatecas, hills of Zacatecas, Oct.
26, 1888 ( C. G. Priagle, no. 1764) ; in the Sierra Madre, Aug. 18, 1897,
and near Plateado, Sept. 4, 1897 (/. N. Rose, nos. 2402, 2804) : San
Luis Potosi, in the mountains about San Luis Potosi, 1876 {ScJiaffner,
no. 660), 1878 {Parry & Palmer, nos. 747, 748): Vera Cruz, Mt.
Orizaba, alt. 2,770 ra., Aug. 8, 1891 (Seaton, no. 260) : Hidalgo, Zima-
pan {Coulter, no. 1093) : Mexico, ledges and banks, Sierra de las
Cruces, Oct. 23, 1892, and Serrania de Ajusco, alt, 2,770 m., May 23,
1898 {G. G. Pringle, nos. 4298, 7548).

2. Leaves pubescent beneath.

O Shrubs.

+ Leaves mostly less than 2 cm. long,

139. S. MICROPHYLLA, HBK. 1. c. 295. Leaves rounded at tip,
slightly pubescent. — Benth. 1. c, ; Hemsl. 1. c. 561 ; Gray, 1. c, in part ;
Briq. 1. c. — Central and southern Mexico. San Luis Potosi, San
Miguelito, 1876 {Schaffner, no. 662) ; in market of San Luis Polosi,
1898 {Edw. Palmer, no. 640^-) : Mexico, Valley of Mexico {Schajfner,
no, 394); Toluca, Sept. 17, 1898 {E. W. D. Holway, no. 3136):
Oaxaca, Tehuacan, alt. 615 m., June 24, 1896 {C.Conzatti, no. 163).

Var. CANESCENS, Gray, 1. c. Leaves and branchlets densely canes-
cent-pubescent. — Chihuahua, in shade of cliffs, hills near Chihuahua,
Oct. 1 7, 1885 ( G. G. Pringle, no. 637) : Zacatecas, hills near Zacate-
cas. Mar. 17, 1885 {C. G. Pringle, no. 239), alt. 2,460 m. Dec. 8, 1898
{O.C.Denm, no. 139).

Var. WiSLiZENi, Gray, 1. c. 408. Puberulent : leaves narrowed to
acutish tips. — Chihuahua, Sierra Madre west of Chihuahua ( Wislizenus,
no. 152) ; north of Batopilas, 1885 {Edw. Palmer, no. 379).

+ + Leaves larger.
X Leaves subsessile or short (3 mm.)-petioled.

140. S. tricolor, Lemaire, 111. Hort. iii. Misc. 71, t. 120, & Fl. des

FERNALD, — MEXICAN SALVIAS. 535

Serres, xii. 109, t. 1237; Hemsl. I.e. iv. 81. — Introduced iuto culti-
vatiou from Mexico.

X X Leaves long-petioled.

-If Corolla 2.5 to 2.75 cm. long.

□ Leaves thin, hardly rugose, with subacuminate tips, pubescence rather sparse.

141. S. Lemmoni, Gray, Proc. Am. Acad. xx. 309, & Syn. FI. 1. c.
461 ; Hemsl. 1. c. — Arizona and Chihuahua. Chihuahua, Canon de
San Diego, Sept. 17, 1891 (C. V. Hartman, no. 738) ; foothills of Sierra
Madre, near Colonia, Aug. 29, 1899 {E. W. Nelson, no. 6315).

n n Leaves rugose, mostly rounded at tips : pubescence more abundant.

142. S. neurepia. Freely branching shrub: branches puberulent,
the younger parts pilose or densely villous : leaves ovate or round-ovate,
truncate or subcordate at base, coarsely crenate, thick, rather densely
short-pubescent on both faces, 2.25 to 2.75 cm. long, twice exceeding
the slender petioles : racemes 1 dm. or less in length ; verticels mostly
2-flowered, all becoming a little remote : bracts ovate, ciliate, caducous :
calyx glandular-puberulent, a little hispidulous on the nerves, tubular-
campanulate, in anthesis 1 to 1.3 cm. long ; the tube twice exceeding
the ovate-acuminate lobes : upper lip entire : corolla-tube pale, glabrous,
twice exceeding the calyx ; the pilose galea hardly equalling the red lip :
style bearded. — Central Mexico. San Luis Potost, in woods near
Morales, 1876 (Schaffner, no. 661) ; without locality, alt. 1,850 to 2,460
m., 1878 {Parry & Palmer, no. 754) ; in the market of San Luis Potosi,
1898 {Ediv. Palmer, no. 640). Graham's no. 1096 without locality
apparently belongs here, and Bourgeau's no. 856 from near Guadalupe,
Mexico, though with larger thinner leaves, may be an extreme form.
This species, as well as S. micropkylla, is sold in the market of San Luis
Potosi as a Mirto, decoction from them being used to allay fever, head-
aches, and nervous irritations.

-H- -H- Corolla 3.25 to 3.5 cm. long.

143. S. Schaffneri. Branches glandular-puberulent and slightly
pilose : leaves thickish, rugose, deltoid-ovate, acuminate, rounded-trun-
cate at base, puberulent on both faces, 2.5 to 4 cm. long, 1.5 to 3 cm.
wide, irregularly crenulate, on slender petioles 1 to 2.5 cm. long :
racemes becoming 1 to 1.5 dm. long; verticels mostly 4-flowered, all
becoming remote, the lowest 2.5 cm. apart: bracts 5 mm. long, ovate,
long-acuminate : pedicels 3 to 5 mm. long : calyx tubular-campanulate
and minutely hispidulous, in anthesis, 1 to 1.2 cm. long; the tube twice

536 PROCEEDINGS OF THE AMERICAN ACADEMY.

exceeding the lobes ; upper lip oblong-ovate, entire, blunt, lower with
two acuminate narrower lobes : corolla-tube glabrous, twice exceeding
the calyx ; galea slightly pilose at tip, shorter than the lip. — San Luis
PoTOSi, in the mountains near Morales, 1876 {Schaffner, no. 6G7),
called " mirto grande." Habitally this plant suggests aS". fidgens, var.
Boucheana, Benth., but it is readily distinguished by its less pubescent
leaves, shorter calyx and slightly pilose galea.

O O Pubescent herb with ovate obtuse leaves canescent beneath.

144. S. OBTUSA, Mart. & Gal. 1. c. 72 ; Benth. in J)C. 1. c. 334;
Van Houtte, Fl. des Serres, xiv. t. 1412 (?) ; Hemsl. 1. c. ii. 562. —
Described from Oaxaca. Various j^lants are cultivated under this
name, but the figure above cited perhaps best represents the species.

h. Calyx glabrous or barely puberulent at base.

145. S. oresbia. Small shrub, 2 to 3 dm. high : branches purplish,
puberulent in lines or ghibrate : branchlets short, leafy mostly at the
tip : leaves firm, ovate, acute at each end, appressed-serrate, glabrate,
1 to 2.5 cm. long: racemes 1 dm. or less in length; verticels 2-4-
flowered: bracts firm, ovate, acuminate, 1 cm. or less long, persistent:
pedicels 3 to 7 mm. long, puberulent: calyx purple-tinged, in anthesis
1 to 1.2 cm. long, puberulent within ; tlie tube thrice exceeding the
broad ovate mucronate-acuminate lobes: corolla 2 cm. long; the gla-
brous strongly ventricose tube one half longer than the calyx ; the pilose
galea hardly equalling the lip : style glabrous. — San Luis Potosi,
in the mountains, San Rafael, 1876 {Schaffner, no. 666) ; without lo-
cality, alt. 1,850 to 2,460 m., 1878 {Parry & Palmer, no. 740).

= = Pubescence of branches and leaves stellate.

146. S. oaxacana. Very branching shrub : the bark pale gray, the
young branches and branchlets white with stellate tomentum : leaves
ovate, truncate or cordate at base, 1 cm. or less in length, i)ale green and
strongly rugose above, white-tomentose beneath, crenate, one-half longer
than the tomentose petioles : racemes 6 cm. or less long ; the verticels
mostly 2-flowered, 1 to 1.5 cm. apart : bracts ovate-lanceolate, caducous :
pedicels 3 to 6 mm. long, glandular-pilose : calyx open-campanulate,
glandular-pilose, in anthesis 8 to 10 mm., in fruit becoming 1.3 to
1.5 cm. long : the upper lip broad-ovate, mucronate, somewhat bidentate,
the lower lip with 2 ovate-lanceolate acuminate lobes : corolla somewhat
glandular-pilose ; the strongly ventricose tube twice exceeding the calyx ;
the galea and lip subequal : style bearded. — Oaxaca, La Hoya Caiion,

FERNALD. — MEXICAN SALVIAS. 537

alt. 1,380 m., Nov. 2, 1894 ((7. G. Pringk, no. 5G54). Resembling •S'.
aspern, JNIart. & Gal., from which it differs iu its short terminal racemes
of red flowers.

■w ++ Leaves from narrow oblong to linear-oblanceolate.

147. S. Greggii, Gray, Proc. Am. Acad. viii. 369, & Sjn. Fl. 1. c
368; Hemsl. I.e. 557, & iv. 81 ; Hook. Bot. Mag. t. 6812. S. micro-
phijJla, Torr. Bot. Mex. Bound. 131, not HBK. — Texas to Durango.
CoAHUiLA, canon above Palomas, 1848-49 {Gregg, no. 335) ; high
dry ridge, battlefield of Buena Vista {Gregg); Saltillo, 1878 {Parry,
no. 26) ; Lerios, July, 1880, and near Saltillo, May, 1898 {Edw. Palmer,
nos. 1071, 153): Durango, shady hillsides and river-banks, Santiago
Papasquiaro, Aug., 1896 {Edw. Palmer, no. 417).

* * Branching or subsimple shrubs or half-shrubs or tall herbs, leafy nearly or
quite to the inflorescence : leaves larger, only the smallest 4 cm. long.

-I- Leaves pubescent beneath.

** Calyx4obes with long setaceous-acuminate tips: corolla 2.5 to 4 cm. long:
brandies densely tomentose, viscid above.

= Leaves densely tomentose beneath, minutely so above.

148. S. LiNDENii, Benth. in DC. I.e. 333; Hemsl. I.e. ii. 560;
Briq. 1. c. — Southern Mexico and adjacent Central America. Oaxaca,
between Pluma and San Miguel Suchistepec, alt. 2,460 to 3,070 m.,
March 21, 1895 {E. W. Nelson, no. 2508) : Chiapas, without locality
{Gkieshreght, nos. 122, 755): Guatemala, Volcan de Agua, Dept.
Zacatepequez, alt. 2,770 m., Apr., 1890 {/. D. Smith, no. 2169).

= = Leaves greener, thinly pubescent beneath, becoming glabrate above.

149. S. RECTiFLORA, Vis. Sem. Hort. Pat. 1839, ex ejusd. Ort. Bot.
Pad. 145; Benth. I.e. 334; Hemsl. I.e. 564, excl. syn. — South-central
Mexico. Guanajuato, rocky mountain-slopes, 1880, 1894 {A. Duges).

t-f t-f Calyx-lobes blunt or mucronate or merely short-acuminate.

= Corolla 4 to 6 cm. long.

a. Leaves densely white-tomentose beneath, finely crenate-serrate, and with sharp-
acuminate tips : calyx glandular-puberulent.

150. S. FULGENS, Cav. Leaves cordate-ovate. — Ic. i. 15, t. 23;
Sweet, Brit. Fl. Gard. ser. 2, i. t. 59; Lindl. Bot. Reg. xvi, t. 1356;
Benth. I.e. 333 ; Hemsl. I.e. 556; Briq. 1. c. S. cardinalis, HBK. 1. c.
300, t. 152. S. grand (flora, Sessc & Moc. I.e. 7. — Central Mexico.
Zacatecas, on the Sierra de los Morones, near Plateado, Sept. 1, 1897
(/. N. Rose^ no. 2743) : Jalisco, between San Sebastian and the sum-

538 PROCEEDINGS OF THE AMERICAN ACADEMY.

mit of Mt. Bufa de Mascota, alt. 2,150 to 2,300 m., March 20, 1897 {E.

W. Nelson, no. 4093): Michoacan, Tlalpujahua {Graham) : Mexico,
Toluca {Andrieux, no. 144) ; Desierto Viejo, Sept. 7, 18G5, and San
Nicolas, Oct. 26, 1865 {Bourgeau, nos. 853, 1106); Valley of Mexico
(Schnffner, no. 396); cool forests. Sierra de las Cruces, alt. 3,230 m.,
Aug., 1892 ((7. G. Pringle, no. 4227) ; Amecameca, Feb. 3, 1893 {E.

W. Nelson, no. 6) ; foothills, Mt. Ixtaccihautl, Jan. 5, 1899 (0. C.
Deam, no. 124) : Morelos, Serrania de Ajusco, alt. 2,460 m., May 26,
1898 {0. G. Pringle, no. 6862).

Var. BoucHEANA, Benth. 1. c. 334. Leaves narrowly deltoid-ovate,
truncate or subcordate at base. — *S'. Boucheana, Kunth, Ind. Sem. Hort.
Berol. 1845. — Michoacan, Tlalpujahua {Graham): Puebla, calca-
reous hills near Tehuacan, Aug. 30, 1897 (C. G. Pringle, no. 7449).

h. Leaves merely pilose on the nerves beneatli, coarsely crenate, and with blunt
tips : calyx not at all or very minutely glandular-puberulent.

151. S. orizabensis. Branches short-pilose with recurved hairs,
minutely glandular above : leaves ovate, thin, minutely pilose above, 4
to 8 cm. long ; the base subcordate with a short acumination at the junc-
tion with the slender retrorse-pilose petiole: raceme 1 dm. or so long;
verticels 4-10-flowered, all remote, the lowest 4 to 5 cm. apart: bracts
ovate-lanceolate, acuminate : pedicels 3 to 5 mm. long, sordid-viscid :
calyx in anthesis 1.5 to 1.75 cm. long, slightly pilose on the nerves ; the
tube nearly twice exceeding the ovate bluntish lobes; upper lip entire:
corolla-tube glabrate, twice exceeding the calyx ; galea, long-bearded,
equalling the lip: style somewhat bearded. — Vera Cruz, Mt. Orizaba
{Botteri, no. 1168).

= = Corolla 2.5 to 2.75 cm. long.

152. S. adenophora. Shrub (" sarmentose " according to Smith)
with brown bark ; the branches glandular-pilose : leaves cordate-ovate,
acuminate (rarely blunt), crenate-serrate, green and sparingly pubescent
above, more or less white tomentose beneath, 4 to 10 cm. long, on
slender petioles 3 cm. or less in length : racemes 3 dm. or less in length ;
the rhachis very glandular-hairy ; verticels 6-20-flowered, all remote,
the lowest 2 to 3 cm. apart : bracts ovate, glandular, caducous : pedicels
6mm. or less in length, glandular : calyx campanulate, very glandular,
in anthesis 8 to 10 mm. long, the tube twice exceeding the broad ovate
mucronate or short-acuminate lobes ; upper lip entire : corolla glandular-
pilose, the tube once and a half longer than the calyx ; galea and lip sub-
equal : style bearded. — Oaxaca, Jayacatlan, alt. 1,230 m., Sept. 11,

FERNALD. — MEXICAN SALVIAS. 539

1894, and Nov. 17, 1895 (Z. C. Smith, nos. 1G8, 875) ; between Tlapan-
cingo and TIalixtaquilla, Guerro, alt. 1,G00 to 1,230 m., Dec. 9, 1895
{E. W. Nehon, no. 2093, in part) ; El Parian-Etla, alt. 1,200 m., Nov.,
1898 ( V. Gonzalez & C. Conzatti, no. 900).

= = = Related to the above, but not identified, is

153. S. LiNEATA, Benth. Lab. 724, with the branches pubescent in
lines, obtuse leaves slightly pubescent, truncate at base, and with the
flowers somewhat smaller than in S.fidgens, in 4-6-flowered verticels.

-t- -I- Leaves glabrous beneath.

■»-f Calyx herbaceous ; the tube becoming: corrugated ; the 3 acuminate lobes large
and subequal, long-ciliate on the margins.

154. S. Pringlei, Robinson & Greenman, Proc. Am. Acad. xxix.
391. — Jalisco, under cliffs, barranca of Tequila, Oct. 2, 1893 (C. G.
Pringle, no. 45G4).

f-f ++ Calyx less herbaceous ; the tube sub-hiflated, not corrugated ; the lobes

unequal.

= Leaves cuneate at base.

155. S. INVOLUCRATA, Cav. Ic. ii. 3, t. 105; Lindl. Bot. Reg. xiv. t.
1205 ; Bot. Mag. t. 2872; Benth. 1. c. 333; Hemsl. 1. c. 558, in part;
Briq. 1. c. aS. laevigata, HBK. 1. c. 295, t. 147. S. Tonduzli, Briq.
Ann. Conserv. Jard. Bot. Geneve, ii. 157. — Southern Mexico and Cen-
tral America. Guatemala, Coban, Dept. Alta Verapaz, alt. 1,320 m.,
Nov. 1886 (^H. von Tuerckhelm in exsicc. J. D. Smith, no. 345) ; Jumay-
tepeque, Dept. Santa Rosa, alt. 1,850 m., Jan., 1893 {Heyde & Lux in
exsicc. J. D. Smith, no. 4397).

= = Leaves cordate-ovate.

a. Calyx in anthesis 0.6 to 1 cm. long.

156. S. PULCHELLA, DC. Cat. Hort. Monsp. 142 j Colla, Hort.
Ripul. 125, t. 16; Benth. 1. c. 334; Hemsl. 1. c. 563 ; Briq. in Engl. &
Prantl, 1. c. — Southern Mexico and Central America. Chiapas, with-
out locality (^Gliiesbreglit, nos. 71, 758) ; near San Cristobal, alt. 2,150 to
2,460 m., Sept. 18, 1895, and between San Cristobal and Teopisca, alt.
2,060 to 2,610 m., Dec. 4, 1895 {E. W. Nelson, nos. 3171, 3427) :
Guatemala, Purula, Dept. Baja Verapaz, alt. 1,540 m., Apr. 1887
(^H. von Tuerckheim in exsicc. J. D. Smith, no. 1194).

b. Calyx in anthesis 1.5 to 2 cm. long.

157. S. puberula. Stem herbaceous, simple, 8 dm. high, from a
woody base, minutely cinereous-puberulent : leaves 7 to 10 cm. long.

540 PROCEEDINGS OP THE AMEEICAN ACADEMY.

acuminate, appressed-serrate quite to the base, dull aud minutely
puberulous above, pale aud glabrous beneath, on slender puberuleut
petioles 4.5 cm. or less in length : raceme simple, 1.3 dm. long ; the
rhachis and pedicels glandular and sparingly pilose; verticels 6-10-
flowered, all remote, the lowest 2.5 cm. apart : bracts caducous (not
seen) : calyx tubular-campanulate, slightly pilose on the nerves ; the
lobes setaceous-acuminate, the upper lip twice exceeding the 2-lobed
lower one, and twice exceeded by the tube : corolla 3 to 4 cm. long, the
moderately ventricose glabrous tube twice as long as the calyx ; the
pilose galea equalling the lip : style bearded. — ^S*. involucrata, Hemsl.
I. c. 558, in part, not Cav. — San Luis Potosi, alt. 1,850 to 2,460 m.,
1878 {Parry & Palmer, no. 755).

* * * Stoloniferous herb ; the subdeltoid or cordiform leaves confined chiefly to
the lower half: jjeduncle elongated, simple; verticels 2-G-flowered, remote:
corolla 5 cm. long.

158. S. STOLONIFERA, Benth. PI. Hartw. 70, & in DC. 1. c. 333 ;
Ilemsl. 565; Briq. 1. c. — Oaxaca, Sierra de San Felipe, alt. 2,300 to
3,080 m., June 23, 1894 {G. G. Pringle, no. 4705, E. IF. Nelson, no.
1183); Telixtlahuaca, alt. 2,300 m., July 27, 1895 (Z. G. Smith,
no. 477).

D. Gyaneae, Benth. Calyx tubulose-campanulate : corolla blue violet
or rose, straight ; the tube ventricose.

* Corolla rose or rose-purple.

H- Leaves hispid or pilose beneath.

*+ Leaves rounded-cuneate at base: corolla 1.5 to 1.8 cm. long.

159. S. PURPURASCENS, Mart. & Gal. 1. c. 69 ; Benth. in DC. 1. c.
335 ; Hemsl. 1. c. 5G4 ; Briq. 1. c. 283. — Described from Mt. Orizaba,
Vera Cruz. Not seen.

*+ *-* Leaves cordate at base.
= Verticels 6-flowered : calyx 1.3 cm. long: corolla 3 cm. long.

160. S. iodochroa, Briq. Ann. Conserv. Jard. Bot. Geneve, ii. 161.
— Described from Costa Rica. Not seen.

= = Verticels 10-30-flowered : mature calyx 8 mm. long: corolla 1.5 cm.

long.

161. S. irazuensis. Tall and freely branching, the stem purple,
slightly crisp-pilose in lines, or glabrate ; the branchlets glandular-jiilose :
internodes long (1 dm.) : leaves cordate-ovate, acuminate, serrate, the

FERNALD. — MEXICAN SALVIAS. 541

larger 5 to 6 cm. long, green and glabrate above, pale and appressed-
pilose on the nerves beneath : racemes elongated, 3 to 4dm. or less long;
verticels all remote : pedicels becoming 6 to 8 mm. long, glandular-pilose :
calyx puberulent, short pilose on the nerves ; the upper lip ovate, short-
acuminate, one half as long as the tube, the lower lip with 2 lance-
ovate setaceous-acuminate lobes : corolla-tube once and a half longer
than the calyx ; the pilose galea shorter than the lip : style bearded. —
S. tiliaefolia, Donnell Smith, 1. c. v. 71, in part, not Vahl. — Costa
Rica, Volcan Irazu, Prov. Cartago, alt. 3,000 m.. Mar., 1894 {J.
Donnell Smith, no. 4920). Near the last species, but with much smaller
flowers, in denser verticels.

^_ -t- Leaves glabrous or merely puberulent beneath, rounded or subcordate at

base: corolla 1.3 to 2.5 C|ii. long.

162. S. Martensii, Gal. Bull. Acad. Brux. 1. c. 77 ; Benth. 1. c;
Hemsl. 1. c. 560 ; Briq. 1. c, & in Engl. & Prantl, 1. c. S. membranacea,
var. villosa, Benth. Lab. 720. S. villosula, Benth. in DC. 1. c. ; Briq.
Ann. Conserv. Jard. Bot. Geneve, ii. 160. — Southern Mexico.
Oaxaca, Sierra de San Felipe, alt. 3,170 m., Aug. 8, 1894 ( C. G.
Pringle, no. 4772), alt. 2,920 to 3,380 m., 1894 {E. W. Nelson, no.
1100), alt. 3,000 m., Nov. 14, 1897 {C. Conzatti and V. Gonzalez, no.
539) : Chiapas, among the mountains {Ghiesbreght, no. 742). Ghies-
breght's plant with rounded, not subcordate leaf-bases and corolla only
1.3 cm. long, probably represents S. villosula which Briquet maintains
at least as a subspecies.

* * Corolla blue or violet.

H- Leaves cuneate or narrowed at base.

*+ Leaves lanceolate or oblong-lanceolate.

= Leaves oblong-lanceolate, acuminate, crenate-serrate, more or less pubescent
beneath : corolla-tube and galea white ; lip dark blue.

163. S. ALBO-CAERULEA, Linden, Belg. Hort. vii. 199, & Gart-
enfl. vii. 55, 97, t. 221 ; Hemsl. 1. c. iv. 80. — South-central Mexico.
MoRELOS, wet mountain caiion above Cuernavaca, alt. 2,000 m.. May
15, 1898, Feb. 15, 1899(0. G. Pringle, nos. 7615, 8020), distributed
under an unpublished manuscript name. Originally from Michoacan.

= = Leaves elliptic-lanceolate, caudate-acuminate, sharply dentate-serrate,
glabrous beneath : corolla violet-purple.

164. S. PANSAMALENSis, Donuell Smith, Bot. Gaz. xxiii. 249. —
Guatemala, Pansamala, Dept. Alta Yerapaz, alt. 1,230 m., June, 1886
(^H. von Tuerckheim in exsicc. J. D. Smith, no. 933).

542 PROCEEDINGS OF THE AMERICAN ACADEMY.

++ ++ Leaves ovate or ovate-oblong, crenate-serrate : corolla deep blue.

= Leaves thin, glabrous beneath : raceme slender, lax ; the remote verticels 2-8-
flowered : lower lip of calyx entire.

165. S. PHAENOSTEMMA, Doimell Smith, 1. c. 13. — Guatemala, be-
tween Rodes and Malacate, Dept. San Marcos, alt. 430 to 1,080 m.,
Jan. 31, 1896 {E. W. Nelsoii, no. 3736).

— = Leaves tomentulose beneath : raceme ratlier dense, the verticels many-flow-
ered : lower lip of calyx 2-lobed.

a. Bracts ovate-acuminate, herbaceous, cinereous-pubescent, 8 mm. or less in
lengtli: corolla-tube glabrous; galea short-pilose.

166. S. MEXiCANA, L. Racemes becoming lax, 4dm. or less long:
calyx cinereous-pubescent on the nerves, in anthesis 1.5 cm. long. — Sp.
23 ; Cav. Ic. i. 16, t. 25 ; Benth. in DC. 1. c. 337 ; Hemsl. 1. c. ii. 561 ;
Briq. in Engl. & Prantl, 1. c. S. inelissifoUa, Desf. Cat. Hort. Par. Ed.
3, 94. Jungia altissima, Moench, Meth. 379. Sclarea mexicana, Dill,
Hort. Elth. 339, t. 254, f. 330. — Central Mexico. Without locality
{Coulter, no. 1100): Guanajuato, 1880 (A. Duges): Zacatecas, near
Monte Escobedo, Apr. 26, 1897 (J". N. Rose, no. 2635) : Mexico, To-
kobaya, Aug. 10, 1865 (Bilimek, no. 308); Valley of Mexico, July 21,
1865, and Aug. 1, 1865 {Bourgean, nos. 993, 619), Sept. 19, 1889
(C. G. Pringlc, uo. 2824); Rio Hondo Canon, 1890 ( C. G. Pringle,
no. 3157).

Var. MAJOR, Benth. 1. c. Similar but with glabrate calyx. — S. pa-
jyilionacea, Cav. Ic. iv. 9, t. 319. S. nitidifolia, Ort. Dec. 53. —
"Without locality {Coulter, no. 1101): Mexico, near City of Mexico
( Graharn).

Var. MINOR, Benth. I.e. Racemes shorter: calyx pubescent, in
anthesis 0.9 to 1.3 cm. long. — Ocimum micranthum, Wats. Proc. Am.
Acad. xxi. 435, not Willd. — Extending further north than the species.
Chihuahua, Frayles, 1885 {Edw. Palmer no. 290): near Batopilas,
alt. 1,690 to 2,000 m., Oct. 4-5, 1898 {E. A. Goldman, no. 205) : San
Luis Potosi, near Morales, 1876 {Schaffner, no. 677); without local-
ity, 1878 {Parry & Palmer, nos. Ttbl, 758) : Zacatecas, Plateado,
Sept. 3, 1897 (/. N. Rose, no. 3638): Mexico, Toluca {Andrieux, no.
148); San Angel, May 31, 1866 {Bourgeau, no. 126) ; Amecameca,
Jan. 5, 1899 {G. C. Beam); Tizapan, alt. 2,310 m., Sept. 23, 1899
{C. G. Pringle, no. 7936).

h. Bracts oblong-ovate, acuminate, membranaceous, veiny, glabrate, 2 cm. long :

corolla long-pilose.

167. S. lupulina. Stems cinereous-puberulent : leaves ovate or

FERNALD. — MEXICAN SALVIAS. 543

rhombic-ovate, acuminate at tip, rounded-cuneate at base, about 1 dm.
long, appressed-serrate, green and sparingly puberulent above, canescent-
puberuleut beneath, on slender canescent petioles 8 cm. or less in length :
racemes in bud resembling the fruit of the hop (JIumulus lupulus) ; the
membranaceous veiny bracts caducous : racemes becoming 1.5 dm. long;
the verticels lO-raany-flowered, the lowest a little remote : pedicels 1
cm. or less long, puberulent: calyx hispidulous at base, glabrate above,
violet-tinged, in anthesis 1.3 to 1.7 cm. long; the tube 3 or 4 times
exceeding the lobes ; upper lip broad-ovate, short-acuminate, entire,
lower with 2 narrower acuminate lobes: corolla 3 to 4.5 cm. long, the
tube once to once and a half longer than the calyx ; galea slightly ex-
ceeding the lip: style exserted, densely bearded. — Oaxaca, mountains
of San Juan del Estado, alt. 2,310 m., Sept. 11, 1894 {L. C. Smith,
no, 167); near Tiapancingo, alt. 1,850 to 2,460 m., Dec. 7, 1894
(£J. W. Mlson, no. 2068).

•*- -I- Leaves round round-truncate or cordate at base.

■*-f Leaves glabrous beneath.

= Bracts somewhat persistent, lasting through anthesis : calyx in anthesis 2 cm.

long : corolla 5 cm. long, violet.

168. S. lANTHiNA, Otto & Dietr. Allgem. Gartenz. xv. 362; De-
caisne, Fl. des Serres, ix. 73, t. 884. — Habitat unknown, supposed to
come from Mexico or Peru.

= = Bracts caducous.

a. Calyx villous.

1. Corolla violet, 3.75 to 5 cm. long, three or four times exceeding the calyx :
leaves thin, lung-acuminate, glandular-punctate.

169. S. RECURVA, Benth. in DC. I.e. 336 ; Hemsl. 1. c. 564 ; Briq.
I.e. — Chiapas, very rare in the forests (G/desbreght, nos. 24, 756).
Originally collected in the Sierra de San Pedro Nolesco.

2. Corolla twice as long as the calyx.

170. S. CONCOLOR, Lamb, in Benth. Lab. 297; Benth. in DC. 1. c. ;
Hemsl. 1. c. 555. — Described from "Mexico."

b. Calyx glabrate, with acuminate-attenuate lobes : corolla 2.5 cm. long, thrice

exceeding the calyx.

171. S. GLABRATA, HBK. 1. c. 299 ; Benth. I.e. ; Hemsl. 1. c. 556 ;
Briq. 1 c. — Described from Venezuela, but reported from " Mexico"
by Benth. 1. c.

54J: PROCEEDINGS OP THE AMERICAN ACADEMY.

*+ +-<■ Leaves pubescent beneath.

= Stem glandular-setulose : leaves setulose-hispid above, coarsely and irregularly
serrate : calyx setulose-hispid : corolla barely 2 cm. long.

172. S. BiSEKRATA, Mart. & Gal. 1. c. 66 ; Benth. 1. c. 335 ; Hemsl.
I.e. 553; Briq. I.e. — Chiapas {Ghieshreght, no. 763). Described
from Mt. Orizaba.

= = Stem villous above, glabrate below : leaves minutely pubescent above, finely
and regularly crenate-serrate : calyx pilose: corolla 2.5 to 3 cm. long.

173. S. CYANEA, Benth. Lab. 296, &, in DC. I.e. 336; Hemsl. I.e.
5)5. S. cyanijiora, Dietr. in Otto & Dietr. I. c. i. 301. ^S". cyanifera,
Otto in Benth. I.e. as syn. — Central Mexico. Mexico, Desierto
Viejo, Sept. 7, 1865 (^13ourgeau, no. 837).

E. Tubiflorae, Benth. Shrubs (rarely herbs) with tubulose-campanu-
late calyx : corolla scarlat crimson flesh-colored or purple, rarely blue-
violet ; the tube straight or curved, cylindric or ampliate (not ventricose)
above.

« Leaves cuneate or narrowed at base. (Forms of S. cinnabarina, elegans, and

nervat.a may be looked for here.)

-I- Leaves pubescent beneath, corolla rose-purple.

++ Leaves softly canescent-tomentose or cinereous-pilose beneath.

= Calyx minutely pilose, in anthesis 9 to 10 mm. long; the tube once and a half
exceeding the ovate-lanceolate acuminate lobes : corolla 2.5 to 4 cm. long.

174. S. CURVIFLORA, Benth. Lab. 284, & in DC. 1. c. 340 ; Hemsl. 1. c.
555 ; Briq. 1. c. — Hidalgo, Zimapan {Coulter, nos. 1088, 1089, 1090) :
MiCHOACAN, Tlalpujahua {Graham).

= = Calyx hispidulous: corolla 1.5 to 1.8 cm. long.

175. S. chiapensis. Fruticose (?), the tall simple branches puber-
ulent belovF, pilose-hispid above : leaves thick, ovate or oblong-ovate,
appressed-serrate, acuminate at tip, subcuneate at base, 5 to 12 cm, long,
green and minutely pilose above, canescent and subvelutinous beneath, on
puberulent petioles 3.5 cm. or less in length: raceme 3 to 4 dm. or less
long ; the verticels 6-lO-flowered, all becoming remote, the lowest 3 to
4 cm. apart: bracts oblong-ovate, acuminate, short-pubescent, 1.5 cm. or
less in length, caducous : pedicels becoming 5 mm. long : calyx in
anthesis 7 to 8 mm. long; the tube nearly twice exceeding the ovate-
lanceolote acuminate lobes: corolla-tube about twice exceeding the calyx ;
the densely villous galea about equalling the lip : style bearded. —
Chiapas, without locality {Ghieshreght^ nos. 57, 761); near San Cris-
tobal, alt. 2,150 to 2,460 m., Sept. 18, 1895 {E. W. Nelson, no. 3201).

FERNALD. — MEXICAN SALVIAS. 545

= = = Calyx long-pilose : corolla 2.5 to 3 cm. long.

176. S. venosa. Herbaceous, with erect branches ; stem hollow,
strongly quadrangular, pilose : leaves thin, ovate or rhombic ovate, cren-
ulate-serrate, short-acuminate at tip, rounded-cuueate at base, 3.5 to
8 cm. long, green and short-pilose above, conspicuously veiny and cine-
reous-pilose beneath ; the upper most often sessile, the others on slender
pilose petioles 2.5 cm. or less in length: racemes 0.6 to 1.2dm. long;
the rhachis white-pilose, densely so in the axils ; verticels 2-8-flowered,
all remote, the lowest 1 to 2 cm. apart : pedicels filiform, white-pilose,
becoming 5 mm. long : calyx in authesis 7 mm. long ; the tube thrice
exceeding the ovate short-acuminate lobes : corolla villous ; the tube twice
longer than the calyx ; the galea 5 mm. long, arcuate-cucullate above the
short lip : style bearded. — Chiapas, on plains, Aug. ( Ghiesbreght, no.
749); near San Cristobal, alt. 2,150 to 2,460 m., Sept. 18, 1895 {E.
W. Nelson, no. 3138).

t-f- •»-(- Shrub with oblong-lanceolate leaves green and sparingly pilose beneath :
rhachis and calyx glandular-pilose ; the upper calyx-lip S^dentate, the middle
tooth filiform-appendiculate : corolla 1.5 cm. long.

177. S. ANTEXNiFERA, Brlq. Ann. Conserv. Jard. Bot. Geneve, ii. 168.
— Described from Chiapas. Not seen.

t- 1- Leaves glabrous beneath or merely puberulent on the nerves,
t-f Leaves ovate : corolla scarlet or carmine.

178. S. TUBiFERA, Cav. Ic. i. 16, t. 25 ; Lindl. Bot. Reg. xxvii. t. 44 ;
Benth. I.e. 341; Hemsl. I.e. 566; Briq. in Engl. & Prautl. I.e. S.
longijlora, Willd. Spec. i. 141. — Described from Mexico. Not seen.

++ t-*. Leaves oblong, 2 cm. long : corolla blue-violet.

179. S. ZACUALPENSis, Briq. Ann. Conserv. Jard. Bot. Geneve, ii.
166. — Described from Chiapas. Not seen.

-w. ++ ++ Leaves oblong-lanceolate, 1.5 to 2 dm. long.

= Verticels 2-flowered : calyx 1.7 cm. long: corolla crimson-violet.

180. S. BELLA, Briq. I.e. 169. — Described from Costa Rica. Not
seen. •

= = Verticels 4-6-flo\vered : calyx in anthesis 9 mm. long : corolla vermilion.

181. S. miniata. Glabrate shrub: branches very slender, deeply
furrowed: leaves thin, minutely puberulent on the nerves above, other-
wise glabrous, long-acuminate at both ends, sharply serrate ; petioles 1
to 2 cm. long : peduncle 1 dm. long ; raceme slightly longer ; verticels

VOL. XXXV. — S5

546 PROCEEDINGS OF THE AMERICAN ACADEMY.

remote, the lowest 3 to 4 cm. apart : pedicels puberulent, 3 mm. loog :
calyx minutely puberulent or glabrate ; the tube thrice longer than the
lips ; upper lip broad-ovate, blunt or short-mucronate, lower with two
short incurved subulate teeth : corolla 3 to 3.25 cm. long, slightly curved,
short-jjilose ; the tube twice longer than the calyx ; the galea and lip sub-
equal : style glabrous or very slightly bearded. — Chiapas, in the forests,
flowering in June {Ghiesbreght, no. 760).

* * Leaves rounded or cordate at base. {S. curvlflora may be looked for liere.)

H- Leaves rounded or rounded-truncate at base, not definitely cordate.

■t-f Leaves lanceolate, 1 to 1.5 dm. long.

182. S. perlonga. Shrub with pilose-hispid branches : leaves thick,
with long-attenuate tips, crenulate-serrate, rugose and covered with short
dense plush-like indumentum above, canescent-velutinous beneath, on
canescent petioles 2 cm. or less long : racemes 1 to 2 dm. long ; verticels
2-8-flowered, all remote ; jjedicels and rhachis sordid glandular-pilose :
calyx purplish, glandular, pilose-hispidulous, in anthesis 1.3 cm. long, the
tube twice exceeding the blunt ovate entire lips : corolla vermilion, 3 cm.
long, the slightly ventricose tube twice exceeding the calyx ; the minutely
pilose lips subequal : style slightly bearded. — Guerrero, northeast
slope of Sierra Madre, near Chilpancingo, alt. 2,150 to 2,620 m., Dec. 24,
1894 {E. W. Nelson^ no. 2186). Further material may show this to
belong with the Fulgentes.

++ ++ Leaves ovate or ovate-lanceolate, shorter.

= Corolla purple rose-purple or flesh-colored.

a. Leaves glabrous or only minutely puberulent beneath, not pilose (except slightly

so in var. of S. purpurea).

1. Branches canescent-velutinous : verticels 2-flowered.

183. S. GRACiLiFLORA, Mart. & Gal. I.e. 75; Benth. I.e. 342;
Hemsl. 1. c. 557. — Described from Zacuapan. Not seen.

2. Branches slightly pilose, puberulent or glabrous : verticels several-many-
flowered.

O A pair of glands generally present at the base of the petiole : corolla' slender,

puberulent, 2 to 2.75 cm. long.

184. S. PURPURKA, Cav. Minutely puberulent or glabrous. — Ic. ii.
52, t. 166; Jacq. Hort. Schoenb. iii. 2, t. 253; Benth. I.e. 341; Hemsl.
I.e. 564; Briq. I.e. S. affinis, Cham. & Schl. Linnaea, v. 99; Benth.
1. c. ; Hemsl. 1. c. 552 ; Briq. 1. c. — Southern Mexico and Central Amer-
ica. Jalisco, Bolaiios {Hartweg., no. 161) : Morelos, Cuernavaca,

FERNALD. — MEXICAN SALVIAS. 547

Nov. 14, 1865 {Bourgeau, no. 1244) : Vera Cruz, region of Orizaba,
without date {BotteiH, no. 576), Oct. 2, 1866 {Bourgeau, no. 3162),
Feb.-May, 1885 {A. Gray), March 10, 1894 {E. W. Nelson, no. 189):
Jalapa, alt. 1,230 to 1,380 m., Jan., 1894 {C. L. Smith, no. 1781):
Oaxaca, near Oaxaca, alt. 1,600 to 2,090 ra., Oct. 2, 1894, and near
Tlapaucingo, alt. 1,850 to 2,460 m., Dec. 7, 1894 (E. TV. Nelson, nos.
1501, 2087); Sierra de San Felipe, alt. 2,000 m., Sept. 23, 1895 {C.
Conzatti, no. 717): Chiapas, without locality {Ghieshreght, no. 128) ;
near Yajalon, Nov. 21, 1895, between San Cristobal and Teopisca, alt.
2,060 to 2,610 m., Dec. 4, 1895, and valley of Comitan, alt. 1,780 to
2,000 m., Dec. 8, 1895 {E. W. Nelson, nos. 3401, 3478, 3484) : Gcate-
MALA, Coban, Dept. Alta Verapaz, alt. 1,320 m., Nov., 1886 {H. von
TuercJcheim in exsicc. J. D. Smith, no. 500) ; Castillas, Dept. Santa
Rosa, alt. 1,230 m., Jan., 1893 (^Heyde & Lux in exsicc. J. D. Smith,
no. 4398) : Costa Rica {Oersted).

Var. PUBENS, Gray. Stems and leaves slightly pilose. — Proc. Am.
Acad. xxii. 446. — Jalisco, deep ravines and shaded hillsides, Rio
Blanco, Oct., 1886 (Edia. Palmer, no. 662) ; near Guadalajara, Oct. 2,
1889 ((7. G. Pringle, no. 2458).

O O No glands at base of petiole : corolla hardly slender, villous, 2.5 to 4 cm.

long.

185. S. LiTTAE, Vis. 111. Plant. Nuov. Ort. Padov. 15; Benth. I.e.;
Hemsl. I.e. 560; Briq. I.e. — Oaxaca, near Oaxaca, alt. 2,300 to
2,930 m., Sept., 1894 {E. W. Nelson, no. 1342) ; Sierra de Clavellinas,
alt. 2,770 m., Oct. 18, 1894 {0. G. Pringle, no. 4991); Rancho de
Caciques, alt. 1,540 m., Sept. 5, 1895 (Z. C. Smith, no. 698).

b. Leaves more or less pilose or tomentose beneath.

1. Pubescence simple.

O Leaves merely pilose beneath.

+ Leaves broad-ovate : corolla deep purple-violet.

186. S. iodantha. Stem 0.9 to 1.5 m. high, puberulent, much
branched above : leaves short-acuminate, serrate, the larger ones 7 cm.
long, 4.75 cm. broad, sparingly pubescent above, canescent beneath,
slender-petioled : racemes long, rather loosely flowered, subsecund, 1 to
2 dm. long; verticels 6-10-flowered, all remote, the lowest 3 to 4 cm.
apart : pedicels slender, about equalling the calyx : calyx in anthesis
5 mm. long, slightly hispid on the nerves and glandular-dotted ; lobes
very short, broad-ovate, mucronate-acuminate : corolla villous, 2 cm.

54:8 PROCEEDINGS OF THE AMERICAN ACADEMY.

long ; the galea exceeding the drooping lip : stamens and style exserted,
glabrous. — Morelos, mountain side above Cuernavaca, alt. 707 m.,
Feb. 5, 1899 {G. G. Pvingle, no. 8039).

+ + Leaves ovate-lanceolate ; corolla flesh-colored.

187. S. michoacana. Shrub 1.5 to 3 m. high, the branches puber-
ulent: leaves 0.3 to 1.2 dm. long, 1 to 6 cm. broad, long-acuminate,
minutely jjilose or glabrate above, canescent or glabrate beneath, slender-
petioled : racemes dense, 1.5 dm. or less long; the lower verticels often
in the axils of the upper leaves : calyx in anthesis 4 mm. long, otherwise
as in the last: corolla villous, slender, 2 to 2.5 cm. long; the galea ex-
ceeding the lip : stamens and style exserted, glabrous. — Michoacan,
dry wooded hills near Patzcuaro, Nov. 12, 1890, and Nov. 24, 1891 (C.
G. Pringle, nos. 3600, 3946) : Jalisco, between San Seb?.stian and the
summit of Mt. Bufa de Mascota, alt. 1,380 to 2,150 m., March 20, 1897
{E. W. Nelson, nos. 4102, 4104).

O O Leaves velutinous or lanate beneath.

+ Leaves serrate : calyx setulose-hispid or glabrate, with short broad raucron-

ate lobes.

188. S. NERVATA, Mart. & Gal. 1. c. 77 ; Benth. 1. c. 339 ; Hemsl.
I.e. 561 ; Briq. I.e. S. monochila, Donnell Smith, Bot. Gaz. xxiii. 13. —
Southern Mexico and adjacent Central America. Guatemala, Tondos
Santos, Dept. Huehuetenango, alt. 3,080 m., Dec. 26, 1895 (K W. Nel-
son, no. 3635). Originally from Oaxaca. Capt. Smith distinguishes
his S. monochila from S. nervnta by its pubescent calyx and scarlet cor-
olla. In the specimen of Nelson's no. 3635 at hand, however, the corolla
is distinctly purple, not scarlet, and there seems little else to keep the
two apart.

+ + Leaves crenate : calyx glandular-villous ; the lobes long-acuminate.

189. S. Karwinskii, Benth. Lab. 725, & in DC. I.e. 345; Ilemsl.
1. c. 558 ; Briq. 1. c. — Described from " Mexico." Not seen.

O O O Of close affinity to the preceding but not identified is

190. S. EXCELSA, Benth. Bot. Reg. xxvii. Misc. 90, & in DC. I.e.
342, described from Guatemala and said to be near S. tubifera but
with the 2-4-flowered verticels in an elongated loose raceme.

2. Pubescence stellate.

191. S. Rosei. Shrub with dark gray bark ; branches canescent with
minute stellate pubescence: leaves ovate, appressed-serrate, blunt or

FERNALD. — MEXICAN SALVIAS. 549

acute, 3.0 to 6 cm. long, dull and sparingly puberulent above, closely
canescent-pubenileut beneath ; petioles canescent, 2 cm. or less long :
racemes simple, 1dm. or less iu length ; verticels 2-6-flowered : pedicels
slender, 2 to 4 mm. long: calyx canescent, in anthesis 1.2 to 1.3 cm.
long, the tube becoming corrugated at base, broadened above, tvpice ex-
ceeding the ovate-lanceolate acuminate subequal lobes : corolla 2.5 to
3 cm. long, purple-violet, villous, the tube twice exceeding the calyx ;
the lip one half as long as the galea : style long-exserted, villous, — Ja-
lisco, between Colotlan and Bohmos, Sept. 8, 1897 (J. N. Rose, no.
2844).

=: = Corolla scarlet or vermilion, without purple tinge.

a. Upper lip of calyx prolonged into a subulate awn, nearly or quite equalling the
calyx-tube ; lower lip with 2 short subulate teeth.

192. S. CINNABARINA, Mart. & Gal. 1. c, 63 ; Benth. 1. c. ; Hemsl. 1. c.
554 ; Briq. I.e., & Ann. Conserv. Jard. Bot. Geneve, ii. 169 (including
varieties). — Southern Mexico and Central America. Oaxaca, near
Oaxaca, alt. 2,310 to 2,920 m., Sept., 1894 {E. W. Nelson, no. 1345)
Sierra de San Felipe, alt. 460 m., Oct. 5, 1894 (C. G. Pringle, no. 4947) ;
San Juan del Estado, alt. 2,300 m., Nov. 4, 1894 (Z. G. Smith, no.
273): Chiapas {G/iiesbreght, nos. 72,757): Guatemala, San Miguel
Uspantan, Dept. Quiche, alt. 1,850 m., Apr., 1892, and Buena Vista,
Dept. Santa Rosa, alt. 1,700 m., Dec, 1892 (ffeyde & Lux in exsicc. J.
D. Smith, nos. 3118, 4381).

b. Calyx-lips subequal.

1. Calyx-lobes subulate tipped.

193. S. ELEGANS, Vahl. Leaves mostly broad-ovate, subtruncate
or rounded at base, rarely exceeding 7 cm. in length : rhachis
and calyces sordid glandular-villous. — Enum. i. 238, 362 ; DC. 1. c.
342; Hemsl. I.e. 555; Briq. in Engl. & Prantl, I.e. *S'. incarnata,
Cav. Ann. Cienc. Nat. ii. 112; HBK. I.e. 293, t. 144; not Etling. .S".
microcuUs, Poir. Diet. vi. 614. aS. punicea, Mart. & Gal. I.e. 65.
S. microcalyx, Scheele, Linnaea, xxii. 589. S. longijiora, Sesse & Moc.
1. c. 8. — Central and southern Mexico. Without locality (Coulter, no.
1097) : Guanajuato, without locality, Dec, 1827 (Berlandier, no.
1319); oak woods {Hartweg, no. 162): Queretaro, Nov. 13, 1827
{Berlandier, no. 1245): Mexico, Santa Fe, July 24, 1865 {Bourgeau,
no. 489) ; woods of Maromenas, Oct. 11, 1865 {Bourgeau, no. 1111);
Valley of Mexico {Schaffner, no. 395) ; woods of Eslava, alt. 2,400 m.,
Jan. 31, 1899 (C G. PringJe, no. 7711): MiCHOACAN, Tlalpajahua

550 PROCEEDINGS OF THE AMERICAN ACADEMY.

{Graham) ; rich canons, hills near Patzcuaro, Dec. 22, 1891 (C. G.
Pringle, no. 3981): Vera Cruz, Mt. Orizaba {Botteri, no. 578), alt.
3,080 m., Aug. 5, 1891 (Seaton, no. 514) ; Cruz Verde, near Jalapa,
alt. 2,130 m., 1894 {C. L. Smith, no. 1480): Guerrero, near Chil-
pancingo, alt. 2,770 to 3,140 m., Dec. 24, 1894 {E. W. Nelson, no.
2245) : Oaxaca, Sierra de San Felipe, alt. 3,080 m., Aug. 8, 1894
((7. G. Pringle, no. 5779).

Var. sonorensis. Leaves long-acuminate, mostly cuneate at base,
about 1 dm. long : rhachis and calyces pilose, only slightly glandular. —
SoNORA, Alamos, 1890 {Edw. Palmer, no. 292); Huehuerachi, alt.
1,230 m., Dec. 6, 1890 (C V. Hartman, no. 312) : damp canon, Rin-
canardo Mts., Dec. 27, 1890 {F. E. Lloyd, no. 449): Chihuahua,
Hacienda San Miguel, 1885 {Ediv. Palmer, no. 270) ; near Batopilas,
alt. 1,G90 to 2,000 m., Oct. 4, 5, 1898 {E. A. Goldman, no. 206).

2. Calyx-lobes not subulate-tipped.
O Bracts 2 to 3 cm. long, persistent.

194. S. MOLLissiMA, Mart. & Gal. 1. c. 71 ; Benth. 1. c. 342 ; Hemsl.
1. c. 561. — Described from Oaxaca. Not seen.

O O Bracts smaller, deciduous.

195. S. COCCINEA, Juss. Stems canescent-puberulent. — Juss. ex.
Murr. Comm. Gott. i. (1778) 86, t. i. ; Benth. I.e. ; Kegel, Gartenfl.
vii. t. 232; Morren, Belg. Hort. ix. t. 5; Gray, Syn. Fl. I.e. 368;
Hemsl. I.e. 555 ; Briq. 1. c. S. rosea, Vahl, I.e. 244. aS'. glaucescens,
Pohl, Fl. Bras. Ic. ii. 136, t. 192. — Of broad range in the Southern
United States, the West Indies, South America, aud India, rare in
Mexico. Reported by Benth. I.e. from Matamoros, Tamaulipas.

Var. PSEUDO-cocciNEA, Gray. Stems conspicuously hirsute. —
Syn. Fl. 1. c. S. pseudo-coccinea, Jacq. Coll. ii. 302, & le. Rar. ii. 2.
t. 209 ; Hook. Bot. Mag. t. 2864 ; Benth. 1. c; Paxton, Fl. Card. ii. t. 40;
Briq. I.e. S. ciliata, Benth. Lab. 286. S. Galeottii, Mart., in Mart. &
Gal. 1. c. 75. — Central and southern Mexico and Central America. Nuevo
Leox, Papagallas, Dec, 1852 (Geo. Thurher, no. 872) ; Guajuco, March,
1880 {Edw. Palmer, no. 1074) : San Luis Potosi, near Tancanhuitz,
alt. 370 m., May 1, 1898 {E. W. Nelson, no. 4376) ; Oaxaca, Chiqui-
huitlan, alt. 1,230 m., Aug. 16, 1895 (Z. C. Smith, no. 619): Chiapas
( Ghiesbreght, no. 752) ; valley of Jiquipilas, alt. 675 to 820 m., Aug.,
1895 (E. W. Nelson, no. 2942) : Yucatan, Chichen Itza, Jan. 18,
1895 (C. F. Millspaiigh, no. 116); Izamal, 1895 (G. F. Gaumer,
no. 453).

FERNALD. — MEXICAN SALVIAS. 551

Var. minima. Low, subsimple, resembling the last, but the leaves
very small, rarely exceeding 2 cm. in length, densely long-setulose above.
— Chiapas, table-land about Ocuilapa, alt. 1,040 to 1,170 ra., Aug. 21,
1895 {E. W. Nelson, no. 3062).

•t- -t- Leaves cordate. {S. coccinea may be looked for here.)

++ Corolla purple or flesh-colored : calyx in anthesis 2 to 2.5 cm. long, with

elongate setiform tips to the lobes.

196. S. ARiSTULATA, Mart. & Gal. I.e. 67; Benth. I.e. 340;
Hemsl. I.e. 553, Briq. I.e. S. longistyla, Benth. I.e. 344; Hemsl.
I.e. 560; Briq. I.e. — Southern Mexico. Jalisco, mountains near
Lake Chapala, Dec. 16, 1889 ( C. G. Pringle, no. 2421): Oaxaca,
Valley of Oaxaca, alt. 1,570 m., Nov. 14, 1894 {L. C. Smith, no. 306);
near Tlapancingo, alt. 1,850 to 2,460 m., Dec. 7, 1894 {E. W. Nelson,
no. 2085).

*+ •*-+ Related to the above but of doubtful affinity is

197. S. Ottonis, Lehm. in E. Otto, Hamb. Gartenz. vi. 350;
Regel, Gartenfl. vii. 129, & Bot. Zeit. xi. 335.

++++++ Corolla scarlet or vermilion.

= Calyx in anthesis 1.75 to 2 cm. long.

a. Leaves white-tomentose beneath.

198. S. INCANA, Mart. & Gal. I.e. 68; Benth. I.e. 344; Hemsl.
1. c. 558. — Described from Puebla. Not seen.

b. Leaves slightly pilose on the nerves beneath.

199. S. TUBiFORMis, Link, Kl. & Otto, Ic. Rar. 70, t. 28 ; Benth.

I.e. 334, under S. rectiflora. S. rectijlora, Hemsl. I.e. 564, in part, not

Vis. — Described and figured from garden specimens. Perhaps native

of Mexico.

= = Calyx in anthesis 1 cm. long.

200. S. cyclophylla. Shrub with dark brown bark ; branches
glandular-hispid : leaves from suborbicular to broad-ovate, rounded or
bluntish at tip, crenate, 6 to 7 cm. across, minutely pilose above, more
densely so beneath, all but the uppermost on slender petioles 3 to 4 cm.
long: raceme 5 to 6 cm. long, the 4-8-flowered verticels 1.5 to 2 cm.
apart: pedicels 3 mm. long, hirsute: calyx glandular-hirsute on the veins ;
the tube twice exceeding the lobes ; upper lip suborbicular, blunt or
mucronate, lower with ovate mucronate-acuminate lobes : corolla pilose,
2.5 to 2.75 cm. long; the tube once and a half longer than the calyx;

552 PROCEEDINGS OF THE AMERICAN ACADEMY.

the short blunt galea equalling the lip : style barely esserted, bearded. —
Between Tlapancingo, Oaxaca and Tlalixtaquilla, Guerrero, alt. 1,240
to 1,600 m., Dec. 9, 1894 {E. W. Nelson, no. 2093, in part.)
++++++++ Related to the above is

201. S. GLUMACEA, HBK. I.e. 298, a very doubtful species probably
of this group, the corolla, however, being unknown. Described from
"New Spain."

F. Hastatae, Benth. Perennial herbs : leaves mostly or entirely basal,
hastate or angulate-cordate : racemes simple or commonly branching,
glutinous-villous : corolla blue; the tube ventricose, ampliate above.

* Calyx in anthesis 1.3 to 1.8 cm. long : corolla 4.5 to 6 cm. long : leaves from
broad-hastate to rounded at base, pubescent on both faces.

202. S. PATENS, Cav. Ic. v. 33, t. 454 ; Lindl. Bot. Eeg. xxv. t. 23
Hook. Bot. Mag. t. 3808; Paxton, Mag. vi. 1; Benth. I.e. 348
Planch. Fl. des Serres, v. t. 503; Hemsl. I.e. 562; Briq. I.e. 284
S. grandijlora^ Nee ex Cav. I.e. S. spectahills, HBK. I.e. 304. S.
mncrantha, Schl. Ind. Sem. Hort. Hal. 1841. S. decipiens, Mart. & Gal
I.e. 64. S. sfaminea, Mart. & Gal. I.e. 65. S. glutinosa^ Sesse «fe
Moc. 1. c. 7, not L. — Central Mexico. Without locality ( Coulter, no.
1102) : San Luis Potosi, mountains of San Miguelito, 1876 {Schaff"-
ner, no. 676); without locality, alt. 1,850 to 2,400 m., 1876 {Parry &
Palmer, no. 759) : Guanajuato, rocky hillsides, Guanajuato, Aug.,
1894 (A. Duges, no. 228).

* * Calyx in anthesis 5 to 8 mm. long.

•*- Leaves villous beneath ; the margins crenate.

++ Calyx-lobes elongated, subulate-aristate : leaves angulate-lobed.

203. S. viTiFOLiA, Benth. Lab. 724, & in DC. 1. c. ; Hemsl. 1. c. 566 ;
Briq. I.e. — Southern Mexico. Vera Cruz, Maltrata, alt. 1,700m.,
Aug. 16, 1891 {Seaton, no. 376) : Oaxaca, Sierra de San Felipe, alt.
2,310 m.. May 26, 1894 {C. G. Prmgle, no. 4659), Sept. 1, 1894 (K
W. Nelson, no. 1169) ; San Juan del Estado, alt. 2,160 m., June 4, 1894
(L. C. Smith, no. 169) ; Caiiada de San Gabriel, Etla, alt. 600 m., June
13, 1897 {G. Conzatti, no. 328).

•w -w- Calyx-lobes short-aristate: leaves generally slightly or not at all an-
gulate-lobed.

204. S. CACALTAEFOLIA, Benth. 1. c. ; Hemsl. 1. c. 554 ; Briq. 1. c. —
Chiapas, in pine forests ( Ghiesbreght, no. 754) ; near San Cristobal, alt.
2,160 to 2,700 m., Sept. 18, 1895 {E. W. Nelson, nos. 3230, 3236 ^r).

FERNALD. — MEXICAN SALVIAS. 553

■*- •»- Leaves sliort-pilose beneath ; margins subentire.

205. S. atriplicifolia. Tall, 1 m. high : stem glandular-pilose,
especially above, leafy nearly to the inflorescence : leaves broadly augu-
late-deltoid, acuminate, glabrate above, pale beneath, mostly 1 dm. long
and broad, on pilose petioles 5 cm. or less in length : inflorescence pan-
iculate, 2.5 to 3 dm. long ; the verticels mostly 2-flowered, a little remote :
bracts linear, glandular-hispid, 2 mm. long : pedicels 5 mm. long : calyx
glandulur-puberulent, with ciliate margins, in antliesis 5 mm. long, the
ovate-lanceolate lobes with short-subulate tips ; upper lip 2-3-dentate :
corolla pilose, dark blue, 3 cm. long : style glabrous. — Chiapas, among
the mountains, flowering in October (^Gliieshreght, no. 759).

Section IX. PYCNOSPHACE, Bentli. Bracts imbricated. Calyx
ovate ; the upper lip tridentate, the lower bifid, all the teeth spinescent.
Corolla-tube pilose-annulate within; the galea erect, emarginate-bifid ;
the lip with small lateral lobes, the larger middle one lacerate-fimbriate
or dentate. Herbs with flowers in dense glomerules or verticels.

* Leaves deeply pinnatifid : verticels 1 or 2.

206. S. COLUMBARIAE, Benth. Lab. 302, «& in DC 1. c. 349 ; Gray,

Syn. Fl. 1. c. 367; Briq. 1. c. — Southwestern United States and adjacent

Mexico. Lower California, San Quentin, 1889 (Edw. Palmer,

no. 620).

* * Leaves crenate.

207. S. Leonia, Benth. Lab. 303, & in DC. 1. c. ; Hemsl. 1. c. 559 ;
Briq. 1. c. Leonia salvifolia, Llav. & Lex. Nov. Veg. Descr. fasc. 2, 6.
— Described from " Santa Rosa."

Section X. HETEROSPHACE, Benth. Calyx tubulose, the upper
lip truncate, tridentate. Corolla-tube pilose-annulate within or rarely
subnaked ; the galea short, erect, the lower lip with small somewhat
spreading lateral lobes. Herbs with loose racemes of few-flowered
verticels.

* Corolla scarlet : stems hirsute : leaves petioled, reniform-cordate, repand-toothed

or pinnatifid.

208. S. RoEJiERiANA, Scheele, Linnaea, xxii. 586 ; Torr. Bot. Mex.
Bound. 132; Gray, Syn. Fl. 1. c. 367 ; Hemsl. 1. c. 564 ; Briq. 1, c. 285.
S. porphyrantha, Decne. Rev. Hort. ser. 4, iii. 301, t. 16; Planch. Fl.
des Serres, xi. t. 1080. S. porphyrata. Hook. Bot. Mag. t. 4939. —
Texas and adjacent Mexico. Chihuahua, Sta. Eulalia IMts., March 27,
1885 (C. G. Prinyle): Coahuila, mountains near Saltillo, July, 1880

554 PROCEEDINGS OF THE AMERICAN ACADEMY.

(Edw. Palmer, no. 1073) : Nuevo Leon, limestone hills, near Monterey,
July 16, 1889 {C. G. Pringle, no. 2869).

* * Corolla blue : stems lanate : leaves sessile, coarsely and pungently dentate.

209. S. CALiFORNiCA, Brandegce, Proc. Cal. Acad., ser. 2, ii. 197. —
Lower California, Calmalli, Cardon Grande, Apr. 23, 1889 i^T. S.
Brandegee).

Index to the Numbered Specimens of Salvia cited in this

PAPER.

(The first number indicates the specimen ; the second, in parentheses, the num-
ber of the species in the preceding synopsis.)

Andrieux, 143 (135); 144 (150); 148 (166) ; 149 (105); 150 (93) ; 151.(111).

Anthony, 294 (97).

Berlandier, 1245 (193) ; 1279 (36) ; 1319 (193) ; 8186 (100).

Bilimek, 304 (118); 308 (166); 315 (03); 316 (63).

Botteri, 394 (1); 395 (11); 533 (11); 534 (53); 576 (184); 578 (193); 869 (11);

1168 (151).
Bourgeau, 122 (11) ; 125 (29, 118) ; 126 (166) ; 295 (103) ; 396 (31) ; 489 (193) ; 490

(28) ; 619 (166) ; 721 (63) ; 837 (173) ; 853 (150) ; 854 (63) ; 855 (98) ; 856 (142) ;

859 (51); 993 (166); 1106 (150); 1109 (53); 1110 (43); 1111 (193); 1238 (132);

1239 (7); 1244 (184); 1504 (3); 1591 (09) ; 2857 (69) ; 3162 (184) ; 3215 (19).
Conzatti, 163 (139) ; 328 (203) ; 717 (184).

Conzatti & Gonzalez, 478 (20) 539 (162) ; 900 (152) ; 902 (112) ; 903 (106).
Coulter, 1088 (174) ; 1089 (174) ; 1090 (171) ; 1093 (138) ; 1097 (198) ; 1099 (136) ;

1100 (166) ; 1101 (166) ; 1102 (202); 1115 (81); 1118 (89); 1120 (91).
Deam, 2 (43) ; 57 (76) ; 124 (150) ; 139 (139).
Duges, 228 (202) ; 228 A (89) ; 228 B (52).
Gaumer, 394 (3) ; 453 (195).
Ghiesbreght, 24 (169) ; 57 (175) ; 61 (41) ; 71 (156) ; 72 (192) ; 76 (137) ; 122 (148) ;

128 (184) ; 129 (60) ; 738 (43) ; 741 (43) ; 742 (162) ; 743 (60) ; 745 (19) ; 747

(43); 749 (176); 750 (41); 751 (29); 752 (195); 753 (137); 754 (204); 755

(148) ; 756 (169) ; 757 (192) ; 758 (156) ; 759 (205) ; 760 (181); 761 (175) ; 763

(172); 764 (110); 765 (130).
Goldman, 72 (69) ; 205 (166) ; 206 (193) ; 212 (138) ; 214 (114) ; 340 (82).
Gonzalez, 42 (103).
Graham, 1096 (142).

Gregg, 322 (100) ; 335 (147) ; 340 (136) ; 406 (40) ; 541 (36) ; 542 (11) ; 597 (123).
Greene, 300 (100).

Hartman, 43 (102) ; 312 (193) ; 323 (24) ; 738 (141) ; 744 (36) ; 749 (11) ; 776 (37).
Hartweg, 159 (11); 160 (37); 161 (184); 162 (193); 163 (29); 164 (103); 165

(136); 171 (43).
Holway, 3028 (79) ; 3136 (139) ; 3184 (39).
Lamb, 311 (3) ; 317 (3) ; 351 (74) ; 355 (74) ; 395 (74) ; 621 (25).
Lemmon, 2861 (102).

FERNALD. — MEXICAN SALVIAS. 555

Lloyd, 440 (103) ; 451 (115) ; 452 (24).

Millspaugh, 18 (8) ; 00 (3) ; 110 (105).

Kelson, 6 (150); 132 (67); 177 (67); 180 (184); 701 (43); 058 (13); 1100 (162);
1103 (118); 1160 (203); 1175 (63); 1183 (158); 1244 (130); 1280 (118); 1342
(185); 1345(102); 1387 (50); 1448(110); 1501 (184); 1511 (105); 1521 (135);
1587 (120); 1505 (104); 1782 (53); 1783 (110); 1704 (86); 1898 (103); 1015
(03) ; 1017 (136); 1072 (135) ; 2005 (116) ; 2040 (130) ; 2064 (55) ; 2068 (167) ;
2085 (106); 2086 (27); 2087 (184); 2001 (135); 2003 (152, 200); 2146 (55);
2186 (182); 2220 (55); 2235 (110); 2245 (103); 2300 (68); 2508 (148); 2668
(76); 2898 (128); 2922 (29); 2942 (105); 3062 (195); 3138 (176); 3142 (43);
3166 (137); 3171 (156); 3101 (41); 3201 (175); 3230 (204); 3236 g (204);
3374 (54); 3401 (184); 3427 (156); 3420 (130); 3478 (184); 3484 (184); 3507
(130) ; 3635 (188); 3730 (165); 4057 (107); 4072 (7) ; 4081 (107); 4003 (150);
4012 (187); 4103 (21) ; 4104 (187); 4171 (20); 4247 (66); 4376 (195); 4527
(100); 4557 (138); 4566 (52); 4651 (40); 4603 (130); 4070 (104); 4755 (138);
4807 (127) ; 4812 a (138) ; 4850 (40) ; 4969 (36) ; 6315 (141).

Palmer [1870], 7 (101).

Talmer [1880], 1062 (11) ; 1064 (94) ; 1067 (100) ; 1008 (100) ; 1069 (100) ; 1072
(04) ; 1073 (208) ; 1074 (105) ; 1079 (130) ; 1096 (70) ; 1097 (40) ; 1008 (40).

Palmer [1885], 64 (9) ; 96 (75) 154 (75) ; 157 (75) ; 205 (24) ; 259 (61) ; 270 (193) ;
290 (100); 370 (139).

Palmer [1886], 28 (83); 53 (30); 61 (33); 183 (121); 184 (120); 226 (29); 488
(75) ; 498 (7) ; 556 (44) ; 598 (59) ; 659 (53) ; 662 (184).

Palmer [1887], 320 (10).

Palmer [1880], 620 (206) ; 684 (07).

Palmer [1890], 10 (3) ; 202 (193); 345 (108); 680 (9); 681 (9); 682 (24); 878
(101).

Palmer [1892], 1964 (72).

Palmer [1895], 244 (3).

Palmer [1896], 305 (37) ; 327 (36) ; 404 (130) ; 417 (147) ; 446 (36) ; 451 (28) ; 572
(11) ; 757 (53) ; 844 (138) ; 967 (53).

Palmer [1808], 107 (100) ; 153 (147) ; 161 (138) ; 194 (94) ; 334 (38) ; 335 (11) ; 336
(36) ; 640 (142) ; 640^ (130) ; 720 (136) ; 1071 (147).

Parry, 26 (147); 29 (94).

Parry & Palmer, 098 (89) ; 729 (47) ; 730 (103) ; 731 (104) ; 730 (28) ; 740 (145) ;
741 (130) ; 743 (11) ; 744 (30) ; 745 (52) ; 740 (52) ; 746^ (11) ; 747 (138) ; 748
(138); 750 (98); 751 (96); 751^ (08); 752 (118); 753 (08); 754 (142); 755
(157) ; 756 (86) ; 757 (166) ; 758 (166) ; 759 (202) ; 760 (32) ; 761 (31).

Prlngle, 52 (95) ; 239 (139) ; 556 (11) ; 637 (130) ; 654 (36) ; 659 (100) ; 1727 (7) ;
1764 (138); 1708 (75); 2297 (26); 2384 (136); 2421 (106); 2458 (184); 2463
(59); 2555 (83); 2560 (121); 2817 (118) ; 2818 (63) ; 2824 (166) ; 2869 (208) ;
2013 (30) ; 3058 (120); 3153 (118) ; 3157 (100); 3170 (98) ; 3206 (104) ; 3224
(03) ; 3273 (104) ; 3593 (107) ; 3600 (187) ; 3681 (40) ; 3683 (53) ; 3940 (187) ;
3954 (43) ; 3981 (193) ; 4097 (107) ; 4150 (39) ; 4155 (31) ; 4163 (29) ; 4200 (40) ;
4227 (150); 4258 (127) ; 4278 (51); 4298 (138); 4351 (109); 4504 (154); 4624
(44); 4059 (203) ; 4705 (158); 4763(105); 4772 (102); 4845 (05); 4862 (130);
4947 (192); 4953 (63); 4900 (135); 4991 (185); 5176 (11); 5024 (20); 5643
(119); 5654 (140); 5779 (193); 0013 (110); 0240 (129); 0245 (92); 6251 (93);

556 PROCEEDINGS OP THE AMERICAN ACADEMY.

6538 (104); 6850 (79); 6862 (150); 6905 (90); 6907 (98); 6914 (118); 7065
(135); 7078 (62); 7080(135); 7449 (150); 7457(51); 7548(138); 7612(122);
7615 (163) ; 7643 (84) ; 7711 (193) ; 7936 (166) ; 7991 (28) ; 8020 (163) ; 8035
(21) ; 8039 (186) ; 8222 (47).

Kose, 1609 (64); 1727 (35); 1934 (30); 2233 (45); 2338 (49); 2402 (138); 2601
(113); 2635 (166); 2641 (123); 2682 (02); 2743 (150); 2745 (29); 2804 (138);
2844 (191) ; 2862 (125) ; 3407 (88) ; 3534 (52) ; 3027 (123) ; 3638 (166).

Schaffner, 394 (139); 395 (193); 396 (150); 410 (43); 648 (89); 660 (138); 661
(142); 662 (139); 663 (98); 664 (98); 665 (136); 666 (145); 667 (143); 669
(123) ; 070 (100); 671 (104) ; 672 (29); 673 (36) ; 674 (11) ; 675 (53) ; 676 (202);
677 (166) ; 678 (85) ; 680 (52) ; 1053 (53).

Seaton, 259 (31) ; 260 (138) ; 304 (11); 323 (31) ; 376 (203) ; 431 (69) ; 500 (124) ;
514 (193).

C. L. Smith, 1480 (193) ; 1664 (26) ; 1749 (67) ; 1756 (67) ; 1781 (184).

J. D. Smitli, 110 (118); 299 (20); 345 (155); 500(184); 780 (54); 816 (11); 823
(43); 933 (164); 1090 (1); 1192 (118); 1193 (43); 1194 (156) ; 1911 (19) ; 2169
(148); 2187 (118); 2597 (19); 3014 (1); 3118 (192); 3128 (43); 3612 (128)
4047 (26); 4048 (18) ; 4050 (118) ; 4051 (17) ; 4052 (128) ; 4055 (11) ; 4056 (118)
4381 (192) ; 4394 (63) ; 4395 (43) ; 4397 (155) ; 4398 (184) ; 4399 (1) ; 4400 (19)
4401 (53) ; 4564 (118) ; 4919 (57) ; 4920 (161) ; 5422 (1) ; 5895 (16) ; 5900 (11)
5901 (1); 5902 (57).

L. C. Smith, 17 (129) ; 167 (167) ; 168 (152) ; 109 (203) ; 170 (135) ; 171 (118) ; 174
(7) ; 268 (53) ; 269 (26) ; 272 (110) ; 273 (192) ; 306 (196) ; 413 (105) ; 414 (52) ;
477 (158); 601 (43); 619 (195); 698 (185); 708 (13); 778 (40); 860 (110); 875
(152) ; 897 (129) ; 908 (63) ; 930 (103) ; 949 (11) ; 950 (130).

Tliurber, 821 (36) ; 872 (195) ; 907 (76).

Tonduz, 701 (11) ; 1781 (57) ; 7158 (16); 7228 (26); 7280 (1); 8456 (11).

Uhde, 791 (118).

Valdez31 (8) ; 55 (1).

C. Wright, 1524 (100).

W. G. Wright, 1298 (74).

Wislizenus, 152 (139).

FERNALD. — MEXICAN SOLANUMS. 557

II. — A REVISION OF THE MEXICAN AND CENTRAL
AMERICAN SOLANUMS OF THE SUBSECTION

TORVARIA.

Dunal's subsection Turvaria of the genus Solanum has never been
unilerstood in America. Almost without exception the many diverse
forms from equally diverse regions have been in American herbaria
placed under the type species Solanum torvum, Swartz. This treatment
as a single polymorphous species of all the plants of similar subgeneric
character, — a course by no means without precedent in other sections
of Sulanuni and scores of other tropical American genera, — has been
due to a lack of authentic material and of any more concise statement of
the specific characters than can be found in the rather ponderous mono-
graph of Dunal. An accumulation of specimens from many sources has
made it possible to divide the Mexican material passing in the Gray
Herbarium as Solanum torvum into species of marked morphological
characters and restricted geographical ranges. Most of these plants
thus separated are found to agree very well with the descriptions of
different species recognized from Mexico by Dunal in his monograph,
though three species there characterized have not yet been identified with
modern herbarium material. Doubtless these identifications of modern
Mexican specimens with the old descriptions cannot all be taken as
final, and a study of the type specimens, when it is possible to examine
them, may prove the present conclusions to be in some cases inaccurate.
Yet confidence is felt that the present understanding of the group is
much clearer than that which has prevailed among recent students of
JNIexican botany. With the hope of simplifying the future study of the
group the following synopsis is presented of the Mexican species of the
section as now interpreted.

* Pubescence of flowerin;? branches densely stellate-tomentose, hairs sliort and

fine.

+- Pedicels bearing simple gland-tipped liairs among the stellate ones. (See also

S. ochraceo-ferrugineum.)

S. TORVUM, Swartz. Branches slightly armed, canescent-ochraceous,
the young parts, especially, ochraceous : leaves subcordate-ovate, shal-
lowly sinuate-lobed, olive-green and stellate-scabrous above, canescent
and stellate-tomentose beneath, 1 to 1.5 dm. long, 6 to 12 cm. broad,
often sparingly armed on the midrib beneath, more rarely so above :

558 PEOCEEDINGS OP THE AMERICAN ACADEMY.

corymb generally bifid or trifid, many-flowered, cymose, the fruitinw
pedicels mostly ascending. — Prodr. 47 ; Duual, Sol. 203, t. 23, & in
DC. Prodr. xiii. 260. — Originally described from the West Indies,
where it is common. In Mexico known only from the Sonth, Chiapas,
near liuehuetan, alt. 150 to 615 m., Feb. 22, 1896 {E. W. Nelson, no.
3830).

Var. RUBiGiNOSDM, Dunal, I.e. 261, described from Guatemala, ap-
parently ditfers from the species only in its more ferrugineous pubescence.
— Guatemala, Rio Dulce, Depart. Livingston, March, 1889 (J. Donnell
Smith, no. 1840): Nicaragua {0. Wright).

S. IIeknandesii, Moc. & Sesse. More spiny throughout, the branches
more loosely sordid-tomentose : leaves deeply and sinuately 5-7-lobed,
the lobes sometimes pinnatitid : inflorescence at first subcorymbose. simple
or bifid, becoming distinctly racemose : fruiting pedicels spreading. —
Moc. & Sessc! in Dunal, 1. c. 266. — Chiapas, mountains near Tonala,
alt. 600 to 1,050 m., Aug. 14, 1895, table land about Ocuilapa, alt. 1,050
to 1,170 m., Aug. 21, 1895 {E. W. Nelson, uos. 2904, 3029): Guate-
mala, San Miguel Uspantan, Depart. Quiche, alt. 1,850 m., April,
1892 {Heyde & Lux in exsicc. J. D. Smith, no. 3446) : Nicaragua (C.
Wright) .

S. madrense. Shrubby, the young branches loosely stellate-tomentose
with canescent or ochraceous hairs, and armed with straight or slightly
curved prickles : leaves thick, ovate to ovate-lanceolate, acuminate, un-
equally subcordate at base, subentire or bluntly and shallowly sinuate-
angulate, without the petiole (1 to 3 cm.) 5 to 15 cm. long, 4 to 13 cm.
wide, above olive-green, stellate-scabrous, beneath a little paler and
tomentulose, sometimes slightly armed ; young leaves often aureate-
tinged, and velutinous : inflorescences extra-axillary, in maturity 4 to 9
cm. long, simple or bifid, scorpoid-racemose ; the pedicels glandular
hairy, spreading and reflexed in fruit : calyx-lobes lanceolate, long-
acuminate : corolla 2 cm. long, 3 cm. broad, whitish, deeply lobed, the
lobes lanceolate or lance-ovate, acutish : filaments very short ; anthers
subequal, slender, 8 or 10 mm. long: the style somewhat longer: berry
apparently black, glabrous, 1 to 1.5 cm. in diameter. — S. diver sifolium,
Wats. Proc. Am. Acad. xxi. 434, not Schl. S. torvum, Wats. 1. c. xxii.
441, not Swartz. — The common representative of the group in the
Sierra Madre and westward to the Pacific. Sonora, Sierra de los
Alamos, March 25 to Apr. 8, 1890 {Edw. Palmer, nos. 363, 364): Chi-
huahua, Hacienda San Miguel, 1885 {Ediv. Palmer, no. 22): Tepic,
San Bias, June 6, 1897 {E. W. Nelson, no. 4335): Sinaloa, Mazatlan,

FERNALD. — MEXICAN SOLANUMS. 659

Dec. 1894, and Isla Pietlra, uear Mazatlan, Dec. 31, 1894 {Frank H.

Lamb, nos. 336, 330"); Rosario, June 20, 1897, and foothills of the

Sierra Madre, near Colomas, July 20, 1897 (/. N. Rose, nos. 1402,

1777): Jalisco, barranca near Guadalajara, June, 1886 (Edw. Palmer,

no. 106); uear Guadalajara, Nov. 14, 1888, May 27, 1891 (C. G.Fringle,

nos. 2193, 5140 [type]) ; mountains near Talpa, alt. 1,200 to 1,540 m.,

March 7, 1897, Maria Madre Tsl., May 3-25, 1897 {E. W. Nelson,

nos. 4040, 4185): Colima, Colima, Jan. 9-Feb. 6, 1891 {Ediv. Pol-

mer, no. 1179): Guerrero, Acapulco, Nov. 1894 (AVw. Palmer, no.

148).

-1- •<- Pedicels not glandular.

++ Pubescence of branches and lower face of leaves whitish-lanate.

S. Hartwegi, Benth. Essentially unarmed, branches rarely with a
few spines : leaves from oblong-lanceolate to ovate, entire or sinuate-
repand, acuminate, pale green above, the youn^^er stellate-velutinous, the
older scabrous: cymes many-flowered, the branches scorpoid; peduncles
pedicels and calyx white-lanate ; fruiting pedicels mostly erect: ripe
fiuit red, glabrous, 1 cm. or more in diameter. — PI. Hartvv. 68;
Duual, I.e. 262, S. torvum, var. lanatum, Dunal, I.e. 261. — Hart-
weg's original specimen was from Hacienda del Carmen. Dunal's
variety was based upon a Guatemala specimen. The following are
referred here: Mexico, Valley of Mexico, Nov. 27, 1865 or 1866
{Bourgeau, no. 725): Vera Cruz, Valley of Cordova, Jan. 16, 1865
or 1866 {Bourgeau no. 1671) ; region of Orizaba, May to July {Bour-
5'eaM, nos. 2408, 2556, Botteri, nos. 82, 1083, Seaton, no. 142); Santa
Lucretia, Isthmus of Tehuantepec, Feb., 1895 (C. L. Smith, no. 1071) :
Oaxaca, hills near Oaxaca, alt. 1,850 m., Sept. 8, 1894 (C. G. Pringle,
no. 4891) ; Monte Alban, alt. 1,690 m., Nov. 24, 1894 (Z. C. Smith,
no. 341) ; near Totontepec, alt. 1,690 m., vicinity of Choapam, alt.
1,170 to 1,385 ra., vicinity of Yalalag, alt. 1,230 to 2,400 m., July,
1894 (^. W. Nelson, nos. 788, 835, 952); Tillantongo, Dec. 12, 1895
{Ed. Seler, no. 1591) : Guatemala, San SigUc4n, Depart. Quiche, alt.
1,785 m., May, 1892 {Heyde & Liix in exsicc. J. D. Smith, no. 3446) :
Costa Rica, Cartago, Prov. Cartago, alt. 1,300 m., Nov., 1887 {Juan
J. Cooper in exsicc. J. D. Smith, no. 5872). Palmer's no. 637 from San
Luis Potosi, 1878, may be an attenuated form of this.

++ ++ Pubescence ochraceous or fuscous.
= Branches more or less armed.
S. Fendleri, Van Heurck & Miill. Arg. Sparingly armed : pubes-

560 PKOCEEDINGS OF THE AMERICAN ACADEMY.

cence of young branches leaves and calyx ocliraceo-ferrugineous, pul-
verulent ; leaves rhombic-ovate, sballowly sinuate-angled, 6 to 8 cm. long,
4 to 5.5 cm. broad : inflorescence closely flowered. — Van Heurck &
Mull. Arg. in Van Heurck, Obs. 130. — Panama, Chagres, Feb.,
March, 1850 (A. Fendler, no. 254).

S. ochraceo-ferrugineum. Stems 1.5 to 3 m, high, armed with
stout broad-based deltoid prickles: young branches leaves and calyces
ochraceo-ferrugiueous, densely velutinous-tomentose rarely a little gland-
ular: leaves oblong-ovate, shallowly siuuate-angled or lobed, 7 to 16 cm.
long, 4.5 to 12 cm. broad, paler and more tomentose beneath than
above : inflorescence loosely many-flowered : calyx in anthesis 8 or 10
mm. long, deeply 5-parted into deltoid-lanceolate acuminate lobes :
corolla 3 cm. broad : anthers slender, subequal, 6 or 7 mm. long : berries
glabrous, 1 cm. or more in diameter. — S. ohtmifolium, Benth. PI.
Hartw. 23, not HBK. S. torvum, var. ochraceo-ferrugineum, Dunal,
1. c. 260. — Guanajuato, Guanajuato, 1837 {Hartweg, no. 204), 1880,
1883 {A. Duges) : Tamaulipas, San Luis Potosi to Tampico, Dec,
1878 to Feb., 1879 (Edw. Palmer, no. 639^) ; Vera Cruz, old fields
and recently cleared grounds, Wartenburg, near Tantoyuca, 1858
(Ervendberg, uos. 285, 485).

S. DiVERSiFOLiUM, Schlecht. Sparingly armed : branches petioles
and inflorescences cinereous stellate-tomentulose : leaves oblong-ovate,
bluntly acuminate, with rounded bases, olive-green and stellate-scabrous
above, cinereous-tomentulose or pulverulent beneath, 7 to 20 cm. long, 4
to 12 cm. broad, subentire or slightly undulate, rarely shallowly sinuate:
inflorescence subaxillary, simple or branched, becoming 6 to 9 cm. long,
extremely floriferous, the pedicels after anthesis mostly strongly reflexed
or drooping. — Linnaea, xix. 297 ; Dunal, 1. c 262. — Originally described
from Papantla, Vera Cruz. The following are referred here ; Mexico,
without locality (Coulter, no. 1245) : San Luis Potosi, hills. Las
Canoas, Aug. 21, 1891 (C. G. Pringle, no. 3901, doubtfully referred
here) : Vera Cruz, Valley of Cordova, Dec. 26, 1865 or 1866 (Bour-
geau, no. 1608); Orizaba (Botteri); near Motzorongo, Feb. 22, 1894
(E. W. Nelson, no. 149) : Oaxaca, Paso de Canoa, Tuxtepec, alt. 150
m., Aug. 28, 1895 (Conzatti, no. 133, L. C. Smith no. 669).

= = Branches unarmed.

S. Hayesii. A small tree : the young branches finely but densely
invested with short ochraceo-cinereous stellate hairs : upper leaves gemi-
nate and unequal, ovate to ovate-oblong, acuminate, unequal and sub-

FERNALD. — MEXICAN SOLANUMS. 561

cordate or rounded at base, entire or slightly undulate, the larger 1.7 to
2.4 dm. long, 1.1 to 1.4 dm. wide, at first stellate-pubescent above, soon
quite glabrous and lucid, cinereous-tomentose beneath ; primary nerves
5 to 7 pairs ; petioles thick, 2 to 4 cm. long : inflorescence extra-axillary,
bifid or trifid, becoming 6 or 8 cm. long, the branches scorpoid-racemose,
densely flowered, ochraceo-cinereous ; pedicels in fruit becoming strongly
deflexed : calyx in anthesis 2 mm. long, with short rounded lobes :
corolla densely stellate without, G or 7 mm. high, barely 1 cm. broad,
with long lanceolate lobes : anthers subequal, lanceolate, truncate, 3 or 4
mux, long: style slightly longer, sparingly stellate-pubescent : berry
glabrous or sparingly puberulent, 1 cm. or more in diameter. — Panama,
Chagres, Jan., 18J0 (A. Fendler, no. 24G) ; near Gatun, Dec, 1859
(Stilton Hayes).

* * Short stellate pubescence of flowering branches mixed with long slender

naked hairs.

S. erythrotrichum. Strongly armed with deltoid mostly hooked
stout prickles : branches densely reddish-tomentose with appressed
stellate hairs and longer naked jointed trichomes : leaves oblong, acumi-
nate, rounded at the base, subentire, 1 to 1.7 dm. long, 3 to 8 cm. wide,
ferrugineous-green sparingly stellate and strongly rugose above, toment-
ulose beneath, at first rufescent, later cinereous, rarely armed on the
midrib beneath; primary nerves 5 to 7 jjairs ; petiole 1.5 to 2.5 cm.
long, rufescent: peduncle extra-axillary, 3.5 cm. long, few-flowered,
rufescent ; pedicels short (7 mm.), thickish : calyx in fruit deeply cut
into lanceolate lobes G or 7 mm. long: fruit puberulent, becoming glab-
rate, about 1 cm. in diameter. — Guatemala, Coban, Depart. Alta Vera-
paz, alt. 1,325 m., Feb., 1888 (H. von Tuerckheim in exsicc. J. D. Smith
no. 1381).

* * * Branches clothed with long stellate-tipped hairs with broad bases.

S. HisPiDuii, Pers. Armed with long stout hooked-prickles, and
densely pubescent with ochraceous or fuscous distinct stiff trichomes :
leaves ovate, subcordate or atteyiuate at base, 0.5 to 2 dm. long, subentire
or with 5 to 11 sinuate lobes; above ferrugineous with distinct stipitate
stellate hairs, or glabrate, and often with long sharp prickles on the
nerves ; beneath cinereous with similar hairs, the nerves often spiny.
— Syn. i. 228 ; Dunal, 1. c. 275. S. stellatum, Ruiz & Pav. Fl. Per. ii.
40, t. 176. S. chrysotrichum, Schlecht. Linnaea, xix. 304 ; Dunal, 1. c.
276. — From the variable material now at hand it seems impossible to

VOL. XXXV. — 30

562 PROCEEDINGS OF THE AMERICAN ACADEMY.

keep S. chrysotrlchuni separate from Persoon's species. Vera Cruz,
Cordoba, alt. 830 m., Aug. 20, 1891 {H. E. Seaton, no. 395) ; Jico, July
16, 1893 (E. TV. Mlson, no. 24) : Orizaba, Jan. 28, 1894 (E. W. Nel-
son, no. 45) : OxVxaca, Calderon, alt. 1,850 m., San Juan del Estado,
Juue IS, 1894 (Z. C. Smith, no. 3G) ; near Reyes, alt. 1,785 to 2,0G0
m., Oct. 24, 1894 {E. W. Nelson, wo. 1785) : Chiapas, near Tumbala,
alt. 1,230 to 1,690 m., Oct. 20, 1895 {E. W. Nelson, no. 3331): Guate-
mala, Santa Rosa, alt. 925 m., June, 1892, San Miguel Uspantiiu, alt.
1,850 m., Apr., 1892 {Heyde & Lux in exsicc. J. D. Smith, nos. 3441,
3448): Costa Rica, Cartago, alt. 1,650 m., Dec, 1887 (^Jnan J.
Cooper, in exsicc. J. D. Smith, no. 5870).

Three species of the subsection Torvaria, described from Mexico, are
still obscure and perhaps not recently collected. These are S. amlctum,
Moric. in Duual 1. c. 263; S. rude-pannum, Dunal, 1. c. 264; S. Lani'
berti, Dunal, 1. c. 268.

III. — SOME UNDESCRIBED MEXICAN PHANEROGAMS,
CHIEFLY LABIATAE AND SOLANACEAE.

Pelexia Pringlei. Roots clustered, cylindric, whitish, tuberiform,
0.5 to 1 dm. long : leaves 3, at the base of the bracteate scape, long-
petioled, with smooth entire oblong to ovate-lanceolate acuminate blades
8 to 12 cm. long, about 4 cm. wide: scape about 3 dm. high, sparingly
pubescent above, with about 8 sheathing lanceolate bracts : spike 1 dm.
long, rather loosely flowered : bracts lance-acuminate, 3 cm. long, much
exceeding the ovary : sepals greenish, lanceolate, acuminate, two of
them free and somewhat di-ooping, 2 cm. long, the others united to form
a blunt galea; lip short and rounded, exceeded by the beak of the
stigma ; spur adnate to the ovary : anther ovate, bluntish, 5 m. long.
— Vera Cruz, wooded hills near Ja-lapa, alt. 1,230 m., April, 1899
(C. G. Pringle, no. 8122). Habitally resembling P. setacea, Lindl.
{Neottia calcarata, Hook. Bot. Mag. t. 3403), but with adnate spur, and
shorter blunter sej^als galea and lip.

Scutellaria distans. Stems slender, flexuous, 3 to 4 dm. long,
somewhat recurved-pilose on the angles ; internodes rather long (3 to
5 cm.) : leaves firm, sparingly appressed-pubescent or glabrate, fan-
shaped, broader than long, upper from deltoid-ovate to rhombic-ovate,

FERNALD. — MEXICAN PLANTS. 563

short-acuminate, coarsely crenate above the rounded or truncate base, 2
to 5 cm. long, 2 to 3.5 cm. wide: flowers solitary in the axils, on slightly
pilose pedicels 0.5 cm. long : calyx pilose in lines, in anthesis 4 to 5 mm.
long : corolla dark blue, arcuate, 2 to 2.5 cm. long, puberulent without ;
the lower lip slightly longer than the upper, with a white spot in the
middle. — Jalisco, in the Sierra Madre, west of Bolaiios, Sept. 15-17,
1897 (J. N. Rose, no. 2951). Related to the northern S. galericidata, L.

S. Rosei. Apparently tall (upper portions of plants at hand 6 to
7 dm. high) : stem short cinereous-pubescent : lower leaves suborbicular,
rounded at tip, upper ovate, acuminate, coarsely crenate-dentate, rounded
or subcordate at base, minutely soft-pubescent on both sides, 4 to 6 cm.
long, 3 to 4 cm. wide, on very short pubescent petioles 1 to 3 mm.
Ions: : raceme elongated, 2 to 2.5 dm. or more in length : the flowers
irregularly scattered, alternate, opposite or fascicled : bracts lanceolate,
about equalling the glandular-hirsute pedicels : calyx somewhat glandu-
lar-hirsute, in anthesis 3 to 4 mm. long : corolla 2 to 2.5 cm. long,
rose-purple above ; the pubescent tube pale ; the lower lip somewhat
exceeding the upper. — Sinaloa, foothills of the Sierra Madre, near
Coloraas, July 20, 1897 (/. N. Rose, no. 1784). Ghiesbreghfs nos. 88
and 803 from Chiapas are probably forms of the species, though with
smaller corollas.

S. Pedicularis. Stems slender, minutely cinereous-puberulent :
leaves oblong or ovate-oblong, coarsely crenate-dentate, short-acuminate
and blunt at tip, rounded or subtruncate at base, sparsely pubescent
above with short appressed hairs, glabrous beneath, 2 to 3.5 cm. long,
1 to 2 cm. wide, on slender finely-puberulent petioles 1.5 to 2 cm. long :
raceme 15-25-flowered, at first dense, resembling Pedicularis canadensis :
bracts linear or lanceolate, equalling the pedicels (3 to 5 mm. long) :
calyx puberulous, in anthesis 3 to 3.5 mm. long, in fruit becoming 8 to
9 mm. long: corolla yellowish, 2.5 cm. long, the tube barely 2 mm. in
diameter nearly to the slightly broader throat ; lower lip a little longer
than the upper : seeds roughish, not margined. — Chiapas, near Tum-
bala, alt. 1,230 to 1,700 m., Oct. 20, 1895 {E. W. Nelson, no. 3342).
The specimens are rather fragmentary, so that the base of the plant can-
not now be described. Nearly related to the showy S. Mociniana,
Benth., which has much longer scarlet corollas.

Stachys (Stachyotypus) flaccida. Ascending or spreading, freely
branched, 4 dm. high or more (?) : stems minutely pilose or glabrate ;
internodes 1 dm. long: leaves thin and flaccid, glabrous or minutely
appressed-pubescent beneath, the lower on slender petioles, the upper

564 PROCEEDINGS OP THE AMERICAN ACADEMY.

becoming sessile, deltoid-ovate to oblong, cordate at base, coarsely
crenate, 3 to 7 cm. long: verticels 3-6-flowered, remote, only the lower
conspicuously leafy-bracteate : pedicels very short : calyx minutely puber-
uleut, ill anthesis (i mm. long, cleft half way to the base into lance-subul-
ate teeth: corolla apparently pale, 1.1 to 1.3 cm. long; the tube once
and a half as long as the calyx; the 3-lobed lower lip twice as long as
the entire short pubescent galea. — Oaxaca, Hacienda de Caciques, Dis-
trict of Cuicatlan, Aug. 4, 1895, alt. 2,150 m. {L. G. Smith, no. 612).
Apparently related to S. Drummondii, Benth.

S. (Calostachys) oaxacana. Stems slender, 2.5 to 6 dm. high,
densely retrorse-hispid below, sparingly so or only glandular-puberulent
above : leaves deltoid-ovate to deltoid-lanceolate, coarsely crenate,
cordate or truncate at base, 1.5 to 3.5 cm. long, above somewhat strl-
gose-pubescent, beneath more or less pubescent, often white with
appressed longitudinally spreading stiffish hairs ; petioles slender, the
lower equalling the blades, the upper shorter : verticels 4-6-flowered,
rather remote, only the lowest conspicuously bracteate : pedicels about
equalling the calyx : calyx green, glandular-puberulent, in anthesis
5 ram. long, in fruit becoming as broad as long, with 5 deltoid-subulate
teeth : corolla scarlet, 2 to 2.5 cm. long, minutely puberulent or glabrate
without ; the 3-lobed lower lip slightly exceeding the entire or emargi-
nate galea. — Oaxaca, near Reyes, alt. 2,060 to 3,080 m., Oct. 20,
1894 {E. W. Nelson, no. 1795) ; Sierra de San Felipe, alt. 2,150 m.,
Sept. 23, 1895 (C. Gonzatti, in exsicc. L. C. Smith, no. 709) ; mountains
of San Juan del Estado, alt. 2,300 m., Oct. 21, 1895 (Z. G. Smith,
no. 924) ; Cuicatlan, alt. 1,675 ra., Dec. 9, 1895 ( F. Gonzalez, no. 47).
Closely related to S. coccinea, Jacq., with which it has formerly been
confused, but from which it is distinguished by its very slender habit,
smaller calyx and different pubescence.

Calamintha oaxacana. Shrub with smooth brown bark, and very
slender minutely puberulent or glabrate branchlets : leaves thin, elliptic-
ovate, 1 to 1.5 cm. long, finely and sharply serrate, acute at tip, nar-
rowed or rounded at base ; petioles filiform, about equalling the leaf-
blades : flowers solitary, axillary, on slender 2-bracteate pedicels 5 to
8 mm. long : calyx tubular, 13-costate, in anthesis 6 to 7 mm. long ;
the tube twice exceeding the lance-subulate lobes ; the upper lip with
3 upturned lobes, the two lobes of the lower straighter : corolla red,
3 to 3.25 cm. long, short-pubescent without. — Oaxaca, El Parian-
Etla, alt. 1,200 m., Nov. 1898 ( F. Gonzalez & G. Gonzatti, no. 901).
Nearest related to G. coccinea of the southern United States. Habit-

FERNALD. — MEXICAN PLANTS. 565

ally somewhat resembliog Gardoquia mexicana, but with the definitely
bilabiate lip of Calaminiha.

Cunila tomentosa. Stem tall (probably 8 dm. high), sharply quad-
rangular, freely paniculate-branched, densely tomentose above and on
the younger parts : leaves lanceolate to ovate-lanceolate, acuminate,
sessile or short-petioled, entire or sparingly appressed-serrate, above
puberulent, beneath densely white-tomentose, 2 to 5 era. long, 1 to 1.5 cm.
wide : racemes dense, spiciform, 1 to 4 cm. long, solitary or in 3's, ter-
minating the short branches: pedicels slender, about equalling the calyx :
calyx 2 to 3 mm. long ; the tube twice or thrice as long as the lauce-tub-
ulate teeth, densely villous especially within the throat : corolla villous,
slightly exceeding the calyx : stamens mostly included. — Oaxaca,
between Pluma and San Miguel Suchistepec, alt. 1,850 m., March 21,
1895 {E, W. JSTelson, no. 2495).

Hyptis (Cephalohyptis) madrensis. Stems procumbent, puberu-
lent (densely short-pilose-hirsute on the younger parts), freely branching ;
internodes 0.5 to 1 dm. long : leaves thickish, oblanceolate to elliptic-
obovate, coarsely crenate-dentate above, the subcuneate entire base nar-
rowed gradually to a short petiole, appressed-pilose on both faces, dark
green above, pale beneath, 2,5 to 5 cm. long, 1 to 2 cm. broad : peduncles
axillary, 3 to 5 cm. long: fruiting head 2 cm. in diameter : bracts folia-
ceous, lanceolate to oblong, entire or coarsely dentate, 1 cm. or less long :
calyx villous, the long subulate teeth setulose. — Tepic, foothills of the
Sierra Madre, near Pedro Paulo, Aug. 3, 1897 (J. JV. Rose, no. 1958).
Resembling, apparently, II. Parkeri, Benth., of South America but dif-
fering from that in its pubescent leaves and calyx.

H. (Minthidium) Pringlei. Stem 1 m. (?) high, freely branching,
more or less pubescent with pilose or somewhat strigose hairs : leaves
lanceolate, acute, subsessile or short-petioled, entire, essentially glabrous,
2 cm. or less long : flowers axillary, 2 to 6 in a fascicle: pedicels slender,
glabrous, becoming 2 mm. long in fruit : calyx glabrous, campanulate, in
anthesis 1.5 mm., in fruit 3 to 4 mm. long, cleft one-third to the base
into narrowly deltoid subacuminate lobes: corolla slightly exceeding the
calyx. — San Luis Potosi, Tamasopo Canon, Aug. 5, 1890 {C. G.
Pringle, no. 3223). Related to H. verticillaris, with which it has been
confused, but with axillary flowers.

H. (Minthidium) axillaris. Similar to the preceding : leaves lan-
ceolate or rhombic-lanceolate, 7 cm. or less in length, acuminate, coarsely
and irregularly serrate above the middle, entire below and cuneate to
petioles 1 cm. or less long, puberulent above, pilose beneath on the

566 PROCEEDINGS OF THE AMERICAN ACADEMY,

nerves : calyx sparingly pubescent or glabrate, salverform, with shorter
broader lobes. — Puebla, near Metlaltoyuca, alt. 250 m., Jan. 31, 1898
{E. A. Goldman, no. 48).

H. (Hypenia § Laxiflorae) Nelsonii. Tall, stem smooth and
glossy below, minutely puberuleut above : leaves thick, glabrous, rather
glaucous, lance-acuminate, slightly auriculate-clasping at base, those of
the stem 1.5 to 2 dm. long, with fine short teeth along the margin, the
upper much shorter and entire : panicle 4 to 5 dm. long, dichotomous ;
the lower ascending branches 3 dm. long : bracts ovate-lanceolate, acum-
inate, puberulent, 1 cm. or less long: ultimate pedicels 0.5 to 1 cm.
long: calyx puberulent, campanulate, in anthesis 5 to 6 mm., in fruit
1 cm. long, strongly 13-nerved, slightly bilabiate ; the deltoid acuminate
lobes one-half as long as the tube : corolla 2 to 2.25 cm. long ; the tube
constricted below, tubular and slightly enlarged upward; the blunt lobes
only 3 or 4 mm. long : styles and anthers exserted, glabrous : nutlets
oblong-obovate. — Jalisco, between San Sebastian and the summit of
Mt. Bufa de Mascota, alt. 1,850 m., March 20, 1897 {E. W. Nelson,
no. 4108). A unique plant among the Mexican species, belonging
to a section hitherto known only from Brazil and adjacent South
America.

Lycium genictilatum. Branches slender, geniculate, covered with
pale gray bark ; spines slender, on the flowering branches about 8 mm.
long, barely 1 cm. apart : leaves very glaucous, glabrous, oblong to
obovate, blunt or acute, 2.5 cm. long or less, on slender petioles 1 cm. or
less in length : flowers abundant in small cymes, terminating the rather
crowded short ultimate branchlets : pedicels 7 to 9 mm. long : calyx
glabrous, short campaimlate, 2 mm. high, slightly broader, with 5 spread-
ing lance-subulate teeth 1 to 1.5 mm. long: corolla 1.2 cm. long, funnel-
form, with broad-cordate lobes 4 to 5 mm. long, pubescent within the
tube : stamens slightly unequal, a little exserted ; filaments pubescent
below: fruit 5 to 8 mm. in diameter, red with a bloom. — Pdebla, near
Tehuacan, Nov. 27, 1895 (C. G. Pringle, no. 7000). Nearest related,
apparently, to L. cestroides, Schl., of Brazil.

Margaranthus sulphureus. Annual, glabrous ; stem stout and
rather fleshy, 4 dm. high, branching above : lower leaves alternate, the
upper and those of the branches geminate and unequal, from ovate to
broadly rhombic-obovate, subentire or shallowly and bluntly sinuate, the
larger 3 to 6 cm. long, 3 or 4 cm. wide, narrowed below to winged pe-
tioles varying from 1 to 4 cm. long; upper leaves smaller, sometimes
subsessile : flowers generally single from each of the upper axils ; pe-

FERNALD, — MEXICAN PLANTS. 567

duncle slender, somewhat arcuate, in fruit becoming 0.5 to 1.5 cm. long:
calyx in authesis 3 or 4 mm. long, sparingly pilose, with 5 short deltoid
ciliate lobes : corolla campanulate-urceolate, 7 to 10 cm. long, nearly as
broad, sulphur-yellow, with purple patches at base : fruiting calyx gla-
brate, globose, 1 to 1.5 cm. in diameter, the short calyx-tips closely con-
nivent. — Mexico, borders of ditches. Valley of Mexico, July 10, 18G5
or 1866 {Bourgeau, no. Ill), Valley of Mexico, alt. 2,250 m., Oct. 4,
1899 (C. G. Pringle, no. 8215). A much coarser plant than the similar
M. solanaceus, with larger yellower corolla.

Athenaea Nelsonii. Stems petioles and pedicels glandular-villous,
fuscous : leaves solitary or geminate and very unequal, on petioles 8 cm.
or less in length, broadly ovate, acuminate, equally or unequally cordate
at base, entire, undulate or slightly sinuate-toothed, the larger 2 dm. long,
thin, dark green and minutely pubescent above, pale and appressed-pilose
beneath, especially on the nerves : flowers in fascicles of 5 to 15 : pedi-
cels at first rather short, in fruit becoming 2 to 3 cm. long : calyx thin,
glandular-pilose, short-campanulate, in anthesis 5 to 6 mm. high and
broad, rounded or subtruncate at base, the tube hardly equalling the 5
deltoid lobes: corolla yellowish, 1.5 to 2 cm. long, deeply cut into oblong
ciliate lobes 1 cm. long : filaments broad at base, inserted midway up the
throat of the corolla, rather shorter than the oblong anthers (3.5 to 4 mm.
long) : fruiting calyx becoming 1 cm. in diameter, inclosing the scarlet
berry. — Chiapas, between Tumbala and El Salto, alt. 460 to 1,385 m.,
Oct. 29, 1895 {E. W. Nelson, no. 3395). A large-leaved species habit-
ally resembling A. viscosa, Benth. & Hook. (^Saracha viscosa, Schrad.,
Sweet, Brit. Fl. Gard. ser. 2, t. 323), but with smaller more regular
calyx, more deeply lobed corolla and much shorter filaments.

Physalis subintegra. Perennial from a very slender root-stock 3
or 4 dm. long, ascending, slightly branching from near the base : stems
tough, dark green, sparingly appressed-pubescent with simple hairs or
glabrate, about 3 dm. high : leaves from ovate-lanceolate to broadly ovate^
3 to 6 cm. long, 1.5 to 3 cm. broad, entire or slightly repand-dentate,
somewhat appressed-pubescent on both surfaces, tapering above to a blunt
tip and slightly rounded below to a petiole 0.5 to 2 cm. long: peduncle
short, 0.5 to 1 cm. long : calyx in anthesis 7 to 10 mm. long, pubescent,
with triangular-lanceolate or ovate acute lobes : corollas pubescent with-
out, the earlier 2.5 cm. broad, the later smaller, sulphur yellow with
large dull brown markings at base : one filament slightly exceeding the
others, 5 mm. long; anthers oblong, purplish, 2 or 3 mm. long: fruiting
calyx globose-ovoid, a little sunken at base, slightly 5-augled, firm and

568 PROCEEDINGS OF THE AMERICAN ACADEMY.

Strongly veiny, 3 to 3.5 cm. long.- — Mexico, Sierra de las Cruces, alt.
3,080 m., Sept. 19, 1899 (C. G. Pringle, no. 8225); Vera Cruz,
Orizaba {BoUeri, no. 207): Oaxaca, slopes of Mt. Zempoaltepec, alt.
2,400 to 3,080 m., July 5-10, 1894 (K W. Nelson, nos. 587, 681):
Jalisco, between Huejuilla and Monte Escobedo, Aug. 25, 26, 1897
(J. N. Hose, nos. 2577, 2606). Nearly related to P. areiiicola, Kear-
ney, of the southeastern United States.

P. philippensis. Freely branching from a stout woody caudex ; the
slender branches somewhat ligneous at base, about 2 dm. in length, long-
pilose with crisp white jointed hairs : leaves appressed-pubescent above,
pilose beneath on the veins and ciliate on the margins, broadly rhombic-
obovate, coarsely and bluntly angulate-sinuate, 1 to 3 cm. long, 1 to 2.5
cm. wide, broadly subcuneate below to a broad long-pilose petiole 1
to 1.5 cm. long: peduncles 1 cm. long, ^jilose : calyx during anthesis
about 8 mm. long, white-pilose, cleft into narrowly ovate acutish or
blunt lobes: corolla short-campanulate, 1.5 cm. long, 2 to 2.5 cm.
broad, whitish or pale yellow, with very broad purplish markings : 2
filaments longer than the others, 7 mm. long ; anthers short-oblong,
purple and yellow, 3 or 4 mm. long : fruit not seen. — Oaxaca,
Sierra de San Felipe, alt. 2,617 m., June 1, 1894 ( C. G. Pringle, no.
5621).

P. sordida. Perennial, densely pubescent all over with short sordid
glandular-pilose hairs : the stems depressed, 4 or 5 dm. long, freely
branching : leaves from suborbicular to ovate, coarsely but bluntly
sinuate-dentate, subcordate or truncate below, pubescent on both faces,

1 to 5 cm. long, 1 to 3 era. broad, on slender glandular-pilose petioles 1
to 3 cm. long: peduncles short, barely 1 cm. long in fruit: calyx in
anthesis 6 or 7 mm. long, sordid-pubescent, with triangular-lanceolate
acute lobes: earlier corollas 1.5 cm. broad, sulphur-yellow, with small
drab or brownish patches at base : filaments somewhat unequal ; anthers
yellow and green, narrowly oblong, 3 or 4 mm. long: fruiting calyx
broadly short-ovate, glandular-pilose, not conspicuously angled, 1.5 to

2 cm. long. — Apparently in sand (the entire plant coated with fine sand),
Oaxaca, Boca de Leon, Telixtlahuaca, alt. 2,310 m., Aug. 8, 1895
{Albert L. Smith, no. 637). Nearest related, apparently, to P. rotundata,
Rydberg, of the southwestern United States.

P. saltillensis. Puberulent throughout with fine white stellate hairs :
branches slender, terete below, sub-angulate above : leaves elliptic-ovate,
coarsely and bluntly sinuate-dentate, the upper unequal and geminate, 3
to 6 cm. long, 2 to 4 cm. wide, unequally subcuneate or rounded at base

FERNALD. — MEXICAN PLANTS. 569

to a sleuder petiole 1.5 to 3.5 cm. long: peduucle filiform, usually curved
at tip, iu fruit 2 to 3.5 cm. long : calyx iu anthesis G ram. long, with
short narrowly triangular acutish or blunt lobes: corolla 1 to 1.5 cm.
broad, sulphur-yellow, with purple patches at base: fruiting calyx ovate,
1.5 to 2 cm. long, obscurely 5-angled, the tips connivent : mature berry
orange, edible. — Coahuila, iu shade, Saltillo, Sept. 1898 {Edw.
Palmer, no. 332).

P. Rydbergii. Perennial (?) : stem slender, tough, much branched,
finely glandular-puberulent : leaves lanceolate to ovate-lanceolate, sub-
entire or shallowly sinuate-dentate, finely glandular-pubescent on both
faces, 0,5 to 1.5 cm. long, 3 to 5 mm. broad, acute at tip, tapering un-
equally at base to slender petioles about half as long : peduncle filiform,
in fruit becoming 1 to 1.5 cm. long: calyx open-campanulate, the lobes
not connivent at tip, in anthesis 0.5 cm. long, with ovate-lanceolate blunt
lobes : corolla barely 1 cm. iu diameter, yellow, with 5 acute lobes :
fruiting calyx broadly open-campanulate, 1 cm. long, 5-10-angled,
strongly reticulate. — Mexico, Ymala, Sept. 25-Oct. 8, 1801 {Edw.
Palmer, no. 1713). Dedicated to Per Axel Rydberg, whose critical
studies of this genus have greatly cleared the confusion which prevailed
in the American species, and to whose unique section Crassifoliae the
present species is an addition.

Solanum (Anthoresis) plumense. Shrub with cinereous smooth-
ish bark: young branches finely pulverulent: leaves ovate-lanceolate,
long-acuminate, acutish or rounded at the base, 1 to 1.8 dm. lono-, 4 to 6
cm. wide, entire or slightly undulate, minutely cinereous-stellate on both
faces ; petioles slender, 2 to 5 cm. long : corymb on a long cinereous-
pulverulent peduncle, in fruit 1 to 1.5 dm. broad, many-flowered: calyx
2 mm. high, cinereous-stellate, with 5 ovate-deltoid bluntish lobes : cor-
olla white, stellate-pulverulent without, 0.5 cm. high, 1.5 cm. broad, with
ovate-lanceolate lobes: stamens subequal, oblong, truncate, 2 to 2.5 mm.
long: fruit subglobose, sparingly pulverulent, 1 cm. in diameter. —
Oaxaca, about Pluma, alt, 925 to 1,480 m., March 17, 1895 (E. W.
Nelson, no. 2493).

S. (Polymeris) dejectum. Branching a little above the stout
woody root : stems slender, herbaceous, ascending, decumbent at base,
branching, cinereous-pilose above with fine stellate hairs: upper leaves
geminate, unequal, from subreniform to broadly ovate, rounded or taper-
ing to a bluntish tip, subtruncate or gradually tapering below to a petiole
1 to 3.5 cm. long, finely and closely cinereous-stellate on both faces,
more strongly so beneath, entire or slightly undulate, 2 to 5 cm. long;

570 PROCEEDINGS OP THE AMERICAN ACADEMY.

peduncles slender, in anthesis erect, 0.5 to 1 dm. long, soon strongly de-
jected: calyx white-stellate, with 10 long linear-subulate divisions:
corolla 3 to 4 cm. broad, plicate, 5-angled, violet or violet and white :
filaments short, one of them twice or thrice exceeding the others, 5 mm.
lone; anthers broad-lanceolate, blunt, 5 mm. long: style 10 to 12 mm.
long, glabrous: fruit said to be yellow. — Durango, rare in crevices of
rocks, Iron Mt., &c., near Durango, July, 1896 {Ediv. Palmer, no. 347) :
Mexico, Mount Zacoalco, July 10, 1865 or 1866 (Balm, no. 543) ; San
Juan del Rio, July 14, 1896 (G. G. Pringle, no. 7202) : Guanajuato,
in fields, 1895 {A. Dugh). Related, apparently, to ^S". somniculentum ,
Kunze, but differing in its closer whitish pubescence, longer petioles,
longer more slender peduncles, and generally blunter leaves.

S. (Polymeris) nocturnum. A woody climber with the habit of
S. lentiim, Cav. : branches glabrous, or minutely stellate-puberulent at
tips: leaves solitary or geminate, ovate, 2.5 to 5 cm. long, 2 to 3 cm.
wide, with acute or obtuse tips, rounded or acutish below to slender
petioles 1 cm. or less in length, subentire, glabrous on both faces, or
sparingly puberulent when young : peduncles solitary or in 2's, 1 to 2 cm.
long, bearing solitary nocturnal blue flowers : calyx, in anthesis sparingly
puberulent, later quite glabrate, a shallow cup with 10 lance-subulate
teeth; corolla minutely and sparingly puberulent without, 12 mm. high:
berry apparently red, glabrous, 1 cm. in diameter. — Guerreuo, low
sandy soil, Acapulco, Jan., 1895 (Edtv. Palmer, no. 533) : Oaxaca,
river-bed above Tehuantepec, Jan. 10, 1896 (Ed. Seler, no. 1625).

S. (Cryptocarpum) macroscolum. Plant with the habit of
S. rostratum, Dunal : stem puberulent, slightly glandular above, armed
with rather scattered yellowish straight stout spines 1 to 1.75 cm. long:
leaves very deeply pinnatifid, the segments again cut into rounded lobes,
sparingly pilose above or glabrate, armed with long stout spines : calyx
hirsute, becoming glabrate, armed with several long (1.5 cm.) slender
spines and many shorter finer ones, deeply cut into lanceolate lobes :
corolla as in S. rostratum : 4 stamens subequal, 8 or 10 mm. long, the
other larger, arcuate, 18 mm. long: fruiting calyx, including the persist-
ent lobes, 15 to 18 mm. long. — Oaxaca, near the town of Oaxaca,
Sept. 19, 1895 (L. C Smith, no. 814), Nov. 18, 1895 (Ed. Seler, no.
1394) : originally collected by Haenke, presumably between Acapulco
and the City of Mexico.

S. (Nycterium) azureum. Stem fruticose, the young branches,
with peduncles pedicels and calyx, cinereous with dense but fine stellate
pubescence ; the branches sparingly beset with remote straight slender

FERNALD. — MEXICAN PLANTS. 571

brown spines 7 mm. or less in length : leaves solitary or the upper
geminate, thin, elliptic-ovate, 4 to 9 cm. long, 2 to 3.5 cm. broad, deeply
sinuate-piunatifid into 5 or 7 unequal oblong or obovate entire or un-
dulate obtuse lobes ; above green and thinly pubescent with appressed-
stellate hairs, beneath slightly more pubescent, especially on the promi-
nent sometimes sparingly short-setulose nerves ; petioles slender, 1 to 3.5
cm. long, often sparingly setulose : peduncles lateral, 5-10-flowered,
becoming 6 or 7 cm. long: calyx campanulate, in the staminate flower
unarmed, in the pistillate with slender spines 5 mm. or less in length,
the 5 lance-subulate teeth 8 or 9 mm. long : corolla 4 or 5 cm. broad,
plicate, stellate-tomentose without, apparently sky-blue, slightly irreg-
ular, deeply cut into 5 ovate-lanceolate curved segments : anthers
yellow, subsessile, in the staminate flower the two shorter ones 1 to
1.5 cm. long, the three longer 3 to 3.5 cm. long, strongly curved above ;
in the hermaphrodite flower the two short anthers scarcely 1 cm. long,
the three long ones straight, 1.5 cm. long: style 2.5 cm. long, strongly
curved above: fruit globose, 1.25 cm. in diameter, in herbarium speci-
mens olive-brown ; the fruiting-calyx deeply cleft, excluding the fragile
elongated lance-subulate tips, about half the height of the fruit. — SiN-
ALOA, Topalobampo, Sept. 15-25, 1897 {Edw. Palmer, no. 178).
Related to the common Solanum amazonium, Ker, and to S. obtusi-
folium, Mart. & Gal. From *S^. amazonium it differs principally in its
thin deeper-lobed and greener leaves, its longer stamens and shorter
fruiting-calyx. From of S. obtusifoUum as described it is separated
by its larger leaves which are not " cordate-ovate," and its much larger
flowers.

Datura villosa. Annual, 1.5 to 3 dm. high: young branches and
petioles villous with long white hairs: leaves 0.5 to 1 dm. long, petio-
late, ovate, sinuate-angulate or deeply cut, villous above and especially
beneath on the veins, becoming glabrate: calyx cylindric, 2.5 to 3.5 cm,
long, somewhat villous, with 5 short teeth : corolla narrow, twice ex-
ceeding the calyx, pale below, purplish at edge, with 5 short lance-
subulate teeth : capsule puberulent, about 5 cm. across, armed with
comparatively few coarse and unequal flat prickles. — .Jalisco, Bolaiios,
Sept. 10-19, 1897 (// N. Rose, no. 3G80) : San Luis Potosi, in
sand near the city of San Luis Potosi, 1876 {J. G. Schnffner, no. 70G) ;
in the mountains at 1,850 to 2,450 m., 1878 {Parry & Palmer, no. 658,
in part). — Related to D. quercifolia, HBK., with which it has been
confused.

Cestrum elegans, Schlecht., var. truncata. Like the species as

572 PROCEEDINGS OP THE AMERICAN ACADEMY.

represented iu Bot. Reg. xxx. t. 43, but with the corolla limb merely
short-toothed or subeutire, uot definitely lobed. — Vera Cruz, San
Miguel del Soldado, alt. 1,850 m., April 20, 1899 (C. G. Prinyle, no.
7800).

C. fulvescens. Branches stout; cortex yellowish brown, pruinose
or glabrate : leaves lanceolate or ovate-lanceolate, short-acuminate,

4 to 7 cm. long, 1 to 2 cm. wide, on slender petioles 1 to 2 cm,
long, dark green and minutely puberulent above, yellow-green and
puberulent beneath : inflorescences on short (1 dm. or les.s) leafy fulvous-
pubescent branches ; the flowers in small axillary clusters or more abun-
dant at the naked tip ; pedicels 3 to 7 mm. long, jointed below the very
slender neck (3 mm. long) of the calyx : tubular-campanulate calyx
broadened above, 1.5 to 2 cm. long, glabrous", conspicuously nerved, with

5 (or 6) unequal deltoid acute ciliate lobes 4 mm. or less in length :
corolla 2.5 to 3 cm. long, the yellow tube slender, slightly exceeding the
calyx, clavate above and enlarged at the throat; limb of 5 (or 6)
elongate-lanceolate strongly involute ascending yellow-brown lobes, short-
pubescent especially within. — Oaxaca, mountains of Telixtlahuaca,
alt. 2,000 m., Apr. 5, 1895 {L. C. Smith, no. 343). Apparently a mem-
ber of the section Habrothamnus, but with paler flowers than in any
described species of that section.

C. Bourgeauianum. Branches slender, glabrous, covered with
a grayish brown cortex : leaves glabrous, entire, oblong-lanceolate
to oblong-elliptic, acutish, including the short petiole 4 to 6 cm. long,

1 to 2.5 cm. wide : inflorescence leafy-paniculate, of mostly compound
sparingly bracteate flexuous or drooping long-peduncled racemes 4 to
9 cm. long: pedicels slender, 3 to 9 mm. long, gradually thickened to
the glabrous campanulate shallowly 5-cleft calyx (in anthesis 4 to 5 mm.
long. 2 to 3mm. broad): corolla yellow, glabrous tliroughout, 1.7 to

2 cm. long, narrowly funnelform below, gradually broadened at the throat ;
limb with 5 generally broad-flaring somewhat rounded lobes nearly
0.5 cm. long: mature fruit nearly 1 cm. in diameter; calyx becoming as
broad as long. — Mexico, Valley of Mexico, June 12, 18G5 or 18G6
{Bourgeau, no. 57) : Oaxaca, Mts. southeast of Miahuatlan, alt. 2,950
to 3,235 m., 1895 {E. W. Nelson, no. 2525). This is perhaps C. pedun-
culm-e, Pavon, but that species is described as having glandular reddish
branches, purplish-veined leaves, and acuminate corolla lobes.

Aster jalapensis. Perennial from slender elongated rootstocks :
stems decumbent at base. 1 to 2.5 dm. high, slender, very leafy, pilose in
lines : leaves oblanceolate, bluntly and shortly acuminate, glabrous or

FERNALD. — MEXICAN PLANTS. 573

minutel}^ puberulent, the upper half coarsely appressed-serrate, the lower
half entire and cuneate to a narrow-winged subpetiolar strongly ciliate
base, including the subpetiolar portion 3 to 5 cm. long, 0.5 to 1 cm, wide ;
the upper smaller : peduncles slender, bracteate, elongate, 2 to G cm. long,
bearinw 1 or 2 heads about 1 cm. broad : bracts of involucre 3-seriate,
linear-oblong, the inner 6 mm. long, the outer much shorter, slightly
ciliate on the margins, with short bluntish herbaceous appressed tips :
rays white : achene 2Jubescent. — Vera Cruz, wet ledges, Barranca de
Chavarrillo, alt. 920 m., Apr. 24, 1899 ((7. G. Pringle, no. 8118).
Somewhat resembling A. bullatus, Klatt, but a lower more slender plant
with shorter broader leaves, broader involucral bracts, and more slender
peduncles.

Bncelia Pringlei. Stem subterete, scabrous-puberulent: leaves oblong
or ovate-lanceolate, acuminate, the lower rather abruptly narrowed below
the middle into an oblong portion which is somewhat auriculate at base ;
the larger 1.2 to 1.4 dm. long, 4 to 6 cm. wide, above tuberculate-
scabrous ; beneath scabrous, glandular-dotted, and somewhat villous on
the nerves ; petioles very short: inflorescence terminal, corymbose, 1 to
1.5 dm. across: heads 1 to 1.5 cm. broad, excluding the rays 1.5 cm.
long : bracts of involucre 2-3-seriate, the outer successively shorter, linear-
oblong, bluntish, appressed-pubescent and somewhat glutinous : rays 8
or 9, ovate-oblong, 1 cm. long, orange-yellow : achene 6 mm. long, sil-
very-silky ; the pappus of two lanceolate awn-tipped pales 4 or 5 mm.
long, and two shorter broader somewhat lacerate ones. — Hidalgo, hill-
sides above Pachuca, alt. 2,770 m., Sept. 14, 1899 (C. G. Prwgle, no.
8248). With remarkably paleaceous pappus for an Encelia, but appar-
ently very closely related to E. glutinosa, Rob. & Greenman.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 26. — June, 1900.

ON THE SINGULAR TRANSFORMATIONS OF GROUPS
GENERATED BY INFINITESIMAL
TRANSFORMA TIONS.

By IIexuy Tabek.

MARINE BIOLOQIGAL LABORATORY.

Received < <^yyr\^.. l^^.J^.^...,

Accession No. //>3..'/~..

Given by U?4-kw^.. 6^
Place,

%*Ho book OP pamphlet is to be removed from the Liab-
opatopy uiithout the pepmission of the Tpustees.

ON THE SINGULAR TRANSFORMATIONS OF GROUPS

GENERATED BY INFINITESIMAL

TRANSFORMATIONS.

By Henry Taber.

Presented April 11, 1900. Received May 1, 1900.
§ 1.

In what follows Xi, X,, . . . X,., will denote r differential operators
defined thus :

n Q

1 O *^i

(i=l,2, . . . r),

where the |'s are analytic functions of n independent variables x. It
will be assumed that the X's are independent, that is to say, that no
system of quantities aj, ao, . . • a^, independent of the x's and not all
zero, can be found for which

(ai Xi + ao X, + . . . + a, X,)/= a^ X,f+ a. X.f+ . . . + a, X,/= 0,
for all functions f of the a:'s ; that is, for which

ai ^1,- 4- as c^2i + . . . + a^ tri = 0,

simultaneously, for e = 1, 2, . . . n. By means of these different oper-
ators we may construct a family with oo'" of transformations

(1) X i =^ ji (xi, . . • a*,,, Gi, . . . «„)

(/:=!, 2, . . . n),

where the cr's are arbitrary parameters, and/^ {x, a) is defined for values
of the a's sufficiently small by the series

X, + kj a J X, X, + 1 kj 5, a J a, Xj X, + etc *

For assigned values of the a's the transformation defined by these
e(|uations may be denoted by T„.

* Lie: Transforniationsgruppen, I. pp. 61, 62.
VOL. XXXV. — 37

578 PROCEEDINGS OF THE AMERICAN ACADEMY.

Among the transformations of this family is an oo'"""^ of infinitesimal
transformations (that is, of transformations infinitely near the identical
transformation), obtained by making the a's infinitesimal. Thus let

tti =^ ai o t, CI2 ^=^ a.2 0 t, . . . a,. = a,, o t,

where the a's are arbitrary finite quantities independent of the x's, and
St is an infinitesimal constant. The system of equations defining this
00'""^ of infinitesimal transformations is then

r

(2) x'i =fi (xi, . . . a?„, ai 8 <, . . . a^ 8 <) = Xj + 8 < 2^- a^- Xj . Xi

{i = 1, 2, . . . n).

YoY assigned values of the a's, the continued applications to the
manifold (xi, x^, . . . x,) of the infinitesimal transformations

r

Xi-\- St ^j aj Xj . Xi,

r

of which 2^- aj Xj is said to be the symbol, generates a group Ci^") with a
single parameter t of transformations

(3) x'i=f^(xi, . . . x„,tai, . . . ta„)

(i = 1, 2, . . . n).
Thus, if

(4) x",=fi{x\, . . . x'„,t'ai, . . . t: a^)

{i = 1, 2, . . . n),

we derive by the elimination of the x"s

(5) x"i = fi (xi, . . . x„, t" ai, . . . 1!' a„)

0" =1,2,.. . n),

where f — t-\-t' . In particular, if <' = — t, x'\ = a-^ for ^ = 1, 2, . . . w.
Therefore, each transformation of G^i^"^ is paired with its inverse and,
for ^ = 0, we have the identical transformation.* In accordance with
the notation adopted, the general transformation of (?i^"^ is denoted by
Tta, ; and, by what precedes, if Tta~^ denotes the transformation inverse
to Tu, we have T/^-' = T^i^.

As t approaches infinity the transformation of group G-^""^ defined by (3)
may approach a definite finite transformation T. But, although for t
infinite, Tta. = T may be non-illusory, it cannot be said to be generated
by the infinitesimal transformation of G^°^. The conception of the

* Lie : Transformationsgruppen, I. pp. 52, 55.

TABER. — ON SINGULAR TRANSFORMATIONS. 579

generation of a finite transformation by an infinitesimal transformation is
not applicable in this case. Moreover, for t ^= oo, the resulting trans-
formation of (7i(") has properly no inverse.

For assigned finite values of the a's, the transformation T^ of the
family defined by equations (1), if not illusory, belongs to the group
G^i*"' with a single parameter generated by the infinitesimal transformation

whose symbol is 2^ cij Xj. Thus the totality of transformation with finite
1

parameters of the family (1) separate into anoo'""^ of groups Gi'-"K In
consequence of what has been said relative to the group G^i'"', it follows
that each transformation of the family (1) with finite parameters is
paired with its inverse, and we have 7^, ~^ = T_a.

As the «'s approach certain limiting values, of which some are
infinite, the transformation T„ may approach a definite finite transforma-
tion Z' as a limit. This transformation may be equivalent to a
transformation 7|, of the family (1) with finite parameters bi, b.2, . . . b,.
In this case T is generated by an infinitesimal transformation of the

r

family, namely IjbjXj, but not otherwise.*

1

The composition of two arbitrary transformations T^, Tj of the family,
defined, respectively, by the equations

(6) x%=fi(xi, . . . x„,ai, . . . a^) .

(i=l,2, . . . «),

(7) a:",=/(x'i, . . . x',„b,. . . b,)

('■=1,2,. . . n),

gives a transformation which may be denoted by T^ T^, and is defined by

(8) x'\ =f, (f, (x, a), . . . /„ (X, a), b^ . . . b,)

(/ = 1, 2, . . . n).

This transformation is not, in general, a transformation of the family.
It will, however, be assumed throughout this paper that

Let fli = a^ <, 02 = oo^ . . . ar = art. It is obviously necessary to distin-
guish between the equations of transformation which result from assigning definite
finite values to the o's, and then increasing t without limit, and those which result
when fli, do, . . . Or (without preserving the same ratio) approach severally cer-
tain limiting values some of which are infinite. The transformation which results
in the first case has properly no inverse. It transforms every point on any one of
the path curves of the group G-^ia) into invariant points of such curves. The trans-
formation which results in the second case if non-illusory may possess an inverse.

580 PROCEEDINGS OF THE AMERICAN ACADEMY.

r

the c's being quantities independent of the a:;'s. In which case, from

(6) x'i =fi(xi, . . . x„,ai, . . . a,.)

(i = 1, 2, . . . n),

(7) x"i —fi (x\, . . . x'„, bi, . . . b^)

{i = 1, 2, . . . «),

we shall obtain (by the chief theorem of Lie's theory)
(J) X j =y,- (a^i, . . . x„y Ci, . . . c^)

x'

!
%

= /■ (a^i,

.

•

• '^ni ^1)

(1 =

1,

2,

• •

.n),

^^5

\

<jpj («i,

«r,

b„

0 =

1

2,

. .

■ r).

where

(10) Cj = cpj (ai, . . . a„bi, . . . b,)

0 = 1,2, . . . r).

Whence it follows that T^ T^^ is also a transformation of the family,
which, tlierefore, constitutes a group, — denoted in what follows by G.*
The equivalence between the transformation T^ and the transformation
resulting from the composition of T,, and Tf, may be denoted by writing

T, = T, T,,

In general there is more than one system of functions cp^{a, h), gjjC^' ^)'
etc., such that T^ = T^ T„ if Cj = cpj (a, S) for^' = 1, 2, . . . r. For finite
values of the o's and b's it may happen that every value of some one (or
more) of the c's is infinite. In this case, from what precedes, it follows
that, while both T^ and T^, are generated by infinitesimal transformations
of the group, the transformation T^ T„ resulting from their composition
cannot be generated thus, and the group cannot properly be said to be
continuous. But, if each of one (or more) of the systems of values of
the C'S, is finite for the assigned values of the o's and b\,T^ T^ can be
generated by an infinitesimal transformation of the group. f A transfor-
mation of G which cannot be generated by an infinitesimal transformation
of G may be termed essentially singular.

In what follows I shall modify the preceding notation, restricting the
use of the symbols T^-, T^, etc., to denote, unless otherwise stated, trans-
formations of group G with finite parameters, and therefore generated,

r r

respectively, by the infinitesimal transformations Sa^- X^, %bj Xj, etc.

* Lie: Transformationsgruppen, I. p. 158.

t Rettger : American Journal of Mathematics, XXII. p. 62.

TABEll.

ON SINGULAR TRANSFORMATIONS.

681

§2.

The trausformation obtained by the successive application to the mani-
fold Xi, Xo, . . . x„, in the order named, of the transformation Ta~ ^ = T_ „,
inverse to T^, and the transformation Ta^say where Sa^, S 02, . . . 8 a^ are
infinitesimal (consequently, Ta-\-sa is infinitely near to T^), is one of the
QO '""^ of infinitesimal transformations of G. If we denote the parameters
of this infinitesimal transformation by Stbj (J ^ 1, 2, . . . r), 8t being
an infinitesimal constant, we have

(H)
or

(11a)

That is to say,

^Stb Ta — Ta-i^

Sw

. x,„ Stbi, . . . 8 tb^)

= fi (/i (^) — «),•••/« (^, — «), «! + 8ai,
(1 = 1,2, . . . n)

a,. + 8 a,)

From this system of equations, which hold for all values of the x's, we
derive, for the determination of bi, b^, etc., r equations independent of
the x's and linear in Soi, 803, etc., namely,

(12)

Stbj — Aji 8ai + Aj2 802 + • • • + ^jr 8a^

(j = 1, 2, . . . r),

where the -4's are functions of «!, a^, . . . a^. These equations written
in Cayley's " abbreviated notation " are

(12 a) 8t {bu bo, . . . b,) = (^u J12 . . . ^i,()Sa.i, 802, .. . 8 a,)*

-^21 -^^22 • • • -^2r

^7-1 -^7-2 * ' ' ^ ■

* In this paper I employ the notation of Cayley's "Memoir on the Theory of
Matrices," Fhilosopliical Transactions, 1858, with the exception that the identical
transformation will be denoted by 7, whereas Cayley denotes this transformation
by the symbol 1. In the notation and nomenclature invented by Cayley a linear
substitution and a bilinear form is each represented by the square array of its
coefficients, the matrix of the bilinear form or of the linear substitution. In
accordance with Cayley's theory, if A denotes the matrix of the linear substitution

n

x'i = 2^ aiv Xi ((' = 1, 2, . . . n),
1

and B the matrix of the linear substitution

582

PROCEEDINGS OF THE AMERICAN ACADEMY.

Let cj),, denote the matrix of the bilinear form —2^2^(2 (t-j Cj v ^) y^ z,,
namely,

/— 2a^-(?^.ii, — 2ajCj.2i, . . . — 2aj-c^-,i

^ Oj Cj 12, — 2 (Ij Cj 22» • • • 2 tlj Cj J. 2

Let /denote the matrix unity (the identical transformation), and let e'^"'
denote the series I + 4> „ + h 4>a' + • • •> which is convergent for any
matrix (/>„, Then it will be found that

■^115 -^12?
-^21) -^22)

. . AJ =— I = /+ *</>„ +

\

^a

Let now A„ denote the determinant of

/-7

^a

This determinant

vanishes if and only if the a's are so chosen that

2 Clj Cjii — 2 k-n- V — 1, 2 Ctj Cyi2 , • • •

2 aj Cji2 , 2 cij Cj„o — 2 A" TT V — 1 •

0,

where k is some integer not zero. The values of the parameters a for
which A„ vanishes may be termed critical values of the parameters.
The critical values of the parameters a are, therefore, those values of
the a's for which one or more of the roots of the characteristic equation
of the matrix <^„ is an even multiple, not zero, of tt ^y/ — i.

If A„ =}; 0, we may take the b's arbitrarily, and then, from equations
(12), derive expressions for Sa-i^Sa^, . • • 8 «,., as linear functions of
^1, b.^, . . . b,.. Thus, if A„ 4= 0, we have

x'i = 2k hiv Xi (i =1,2,... n),
1

A ±. B denotes the matrix of the linear substitution

x'i = 2" {aiv ± bh) Xi (i = 1,2, . . . n),
1

and A B the matrix of the linear substitution

tt n

x'i = 2*^ 2>/ «)> b^v Xv {i = 1, 2, . . . n).
1 1

We shall then have A {B C) - {A B) C, A {B ± C) -A B + A C, etc., but in
general AB :^ B A.

TABER. — ON SINGULAR TRANSFORMATIONS.

583

(13) (8 «!, 8 a„, ... 5 Or) =

e" - I

{K h,, . . . b,)

. .(J^i, 6.,, . . .b,.),

or

(13a)

+ ajr ^r)

(7 = 1.2,

»■).

where a^v is the first minor of A„ relative to A^fj.. The quantities
8ai, 8«2j S^ri as determined by these equations, are infinitesimal if
A„ ^ 0, since then the constituents a^„ A„~ ^ of the matrix <^„ (e '' — /) ~ ^
are finite. Therefore, if the parameters a are so chosen that A,, 4= Oj we
may take ^i, bo, . . . b,., arbitrarily, and, if Soi, hao, ... S«y, are
determined by equations (13), we have

Tub Ta = Ta + Sa,

where 8«i, Sa2> • • • S a^. are infinitesimal.

On the other hand, if the values assigned to the parameters a are
critical values of the parameters, that is, if A„ = 0, it will certainly iu
general, for arbitrary values of the S's, be impossible to determine infin-
itesimal increments 8 a-^, 8 a<^^ ... 8 a,., of the pai'ameters a to satisfy
the symbolic equation

■htb

T — T

-t a — -'■II

a -\- Sw

In this case, it may, nevertheless, be possible to find a finite system of
values Ci, c^, . . . c,., of the parameters such that T^h Ta = T ; but group
G may be such that, for at least special systems of values of the b's, no
finite system Ci, c^, . . . c,., of the parameters can be found to satisfy this

5 -,, 9 . 9
9x^'

2 TT V— 1, A„ = 0 ; and if b^ ^0,b.2 = 0, T^ T^;, is essen-
0.

symbolic equation. E.

^., let )' = 2 and X^ = ^^--, ^2 = ^ 1^2 ^ — •

Then, if (to

tial singular for all values of t

From what precedes we have therefore the following theorem :
If T ^5 an arbitrary transformation of G for which A„ 4^ 0, the trans-
formation Ti^T,,, tlie iKtrameters bi, bg, . . . b^, being arbitrary, can be
generated by an infinitesimal transformation of the group, provided t is

584 PROCEEDINGS OF TUE AMERICAN ACADEMY.

sufficiently small* On the other hand if A„ = 0, and the b's 'arc prop-
erly chosen, the transformation T^^ T„ may be essentially singular however \
small t may be taken. Such a transformation T^ 1 term non- essentially ,
singular. i

If Ta is non-essentially singular, that is, if T'^j T^ is essentially singu-
lar however small t may be, a system of values b\, b'^, . . . 5',., of the
parameters can be found such that, however small t may be, T^ Tt^, is ,
essentially singular ; and conversely.

§3.

Let «!, a^, . . . a^, and ^i, b^, . . . b^, be any two systems of finite
arbitrarily chosen values of the parameters of G, and let the transforma-
tion Ta be defined by the symbolic equation

(14) T„ = n Ta,
where i is a variable quantity independent of the x's. We then have

(15) aj = (pj(ai, . . . a„tbi, . . . tb,)

0 = 1,2, ... r).

The differential equations satisfied by the a's are

See Bulletin of the American Mathematical Society for February, 1900,
p. 202.

Two groups G and (?'^' are of the same structure (Zusammensetzung)
if the structural constants [Zusammensetzungconstanten) Cj^i and cfli are
identical. For two groups of the same structure, the system of differ-
ential equations satisfied by the r dependent variables Oj ==: (pj (a, t b)
are the same. But the equations of the group may restrict the number
of systems of the functions aj, which differ in the initial values of the
a/s, in certain cases so that there shall be but one system of functions (p.
Consequently, in the case of two groups of the same structure, one
may contain essentially singular transformations and the other may
contain no essentially singular transformation. Two such groups cannot
properly be said to be isomorphic, since one is continuous and the other
discontinuous.

* If An i= 0, Ttb Ta may be essentially singular for an infinite number of values
of t. But this assemblage of values of t has no derived assemblage.

TABER. — ON SINGULAR TRANSFORMATIONS. 585

It will* be found that one or more of the roots of the characteristic

equation of the matrix 6*^^*6**" is equal to unity, irrespective of the value
of t. If Ta is non-singular, and for every value of t each root of this
equation is equal to unity, 7^,j7a is non-singular for every value of t.
Let it be assumed that Ta is non-singular, and that just s < r of the

roots of the characteristic equation of the matrix e*^''-' e^°- are equal to
unity, irrespective of the value of t. Then the values of t for which
7\ = T,t, Ta is singular (essentially or non-essentially) are included
among those for which one, or more, of the remaining r-s roots of this
equation is equal to unity.

§4.

The infinitesimal transformation 2 a^ Xj of group G, where the a's are
quantities independent of the x's, is said to be derived lineally from the r
independent infinitesimal transformations X^, X2, . . . X,. which gene-
rate G. The r infinitesimal transformations

ai"--'Xi -f a^^^X^ + . . . + a/*' Z,(^ = 1, 2, . . . r)

are independent if the determinant

(A)

to.

Any r independent infinitesimal transformation derived linearly from the
^'s also generates group G and may be substituted for the ^'s.*

Group G may contain an infinitesimal transformation 2 a^ Xj commu-
tative with each of the r infinitesimal transformations Xj which generate
G, and, therefore, commutative with every infinitesimal transformation
of G. Such a transformation Lie terms an ausgezeichnete injinitesimale
Tra7isformation.-\ In what follows it will be termed an exti-aordinary in-
Jinitesimal transformation.

Let G contain just s independent extraordinary infinitesimal transfor-
mation. In this case, from what has been said, we may suppose the A''3
so chosen that

Xj X^ = X^ Xj

(j=l,2, . . . s fc = l,2, . . . r),
but that

* Lie: Transformationsgruppen, I. p. 276.
t Lie : Continuinerliche Gruppen, p. 465.

586

PROCEEDINGS OF THE AMERICAN ACADEMY.

-^) ^k =r ^k Xj

(j,k = s+l,s + 2,... r)

We then have Cj^i = 0 for / = 1, 2, . . . s, aud k, 1= 1^2, . . . r.
And, as a consequence of the differential equations satisfied by the func-
tions Qj = (fj (a, t h), it will be found that

5 (fj (a, b) _ ^

9 cfj (a, ^) __ Q

9 b,

r k=zl,2, . . . s).

9 (h
{j = s+l,s + 2,

Moreover, we shall have

% {a, b) = ttj + b^ + ipj (a,4.i, . . . a„ b, + i . . . b,)
U = 1,2, ... s).

From the differential equations satisfied by the functions Oj = q)- (a, t h)
it also follows that, if Cj^i = 0 forj/', ^- = 1, 2, . . . r, we then may put

(Pi (a, h) =7ti + bi. *

§5.

If r = 2, group G either contains no extraordinary infinitesimal trans-
formation or two linearly independent extraordinary infinitesimal trans-
formations. In the first case, the infinitesimal transformation 2 Oy A^ is
commutative with no other infinitesimal transformation of G. In the
second case, every two transformations of G are commutative.

If r = 3, and the structural constants are such that

^122> ^\"2i ^-232
•^123' ^l-«' ^2.'!.3

* This theorem, for the case in which (r is a sub-group of the projective group,
was given by Mr. Rettger in tlie American Journal of Mathematics, XXII. p. 73.

As an example of this theorem let X, = oc-, -z — , A'o = a:., -r — , Xo = x.y -^ — .

Then
And if

Cj-ti = 0, C;i2 = 0 0', ^•= 1, 2, 3).

TbTn= Tr, c■^ — a-^ + h-^ + ^lcn y^ —\, c., — a^ + h2 + 2h'i7 y/ —\, where h and
k' are integers which may both be taken equal to zero.

t E. g., .Y,

dx.

, X:

()x.

X, = x{^

dx{

TABER, — ON SINGULAR TRANSFORMATIONS. 587

the infinitesimal transformation '^aj Xj is commutative with no other in-
finitesimal transformation of G. But, if ?• = 3 and this determinant van-
ishes, it is alvvay possible to find two distinct infinitesimal transformations
2cf, Xj and 2 bj Xj which shall be commutative.

Again, if r > 3, it is always possible to find two distinct infinitesimal
transformations of G which shall be commutative.

The condition necessary and sufficient that two infinitesimal transfor-
mations 2 a J Xj and 2 bj Xj shall be commutative is that

(17) </>„(^i, ^o, . . . (i,) =0;
or, what is the same thing, that

(18) cf>f, («i, a.2, . . . a^) = 0.

If A„ 4= 0> tl^6 necessary and sufficient condition that every transforma-
tion of the group Ci**' with a single parameter t shall be commutative
with Ta, that is to say, that 7\ T,,^ = T^f^ T^ for every value of t, is tliat
the infinitesimal transformations 2 cij Xj and 2 b^ Xj shall be commutative.

In certain groups G, whatever the transformation T^, provided i\a = O5
it is always possible to find an infinitesmal transformation 2 bj Xj, not
commutative with 2 Oj Xj, which shall, nevertheless, generate a group
Ci'*' with a single parameter t, every transformation J'^j of which shall
be commutative with T^. In other groups this is possible for certain
transformations 7^, for which A„ = 0.

As an example of tlie former we have the group

V — ^ \- — ^ V _ ^ V — ^

^^-'■^cTTi' ^^2--i-2^^, A3_^3^^, A,_X3^^

For this group A,, = 0 if a^ or (i., is an even multiple, not zero, of tt /^— 1. Let
«i. «3. «4 he arbitrary, and a., = "J tt \/ —1, and let n/ig — og hj^ = 0. Then Ta Ttb
= TtbTa ior all values of t; but 'S.hjXjXS not commutative with 2 a> A7 unless
"2 64 — «4 ^2 = 0. If, however, a\ — a-y, a'. 2 — 0, a'g = a^, a\ — 0, Ta' = Ta and 2 hj Xj
is commutative 2 a'j Xj.

If A„ = 0, the necessary and sufficient condition that T^ T^f, = Ttt, T^
for all values of t is

(19) (A - /(J^i, 5,, . . . 5,) =0.

It is to be noted that this condition is always satisfied if 2 a^ Xj and
2 bj Xj are commutative. For then

Cf>a (5i, b,, . . . b,) = 0

.•.4>\{b,,b,, . . . b,) = {)

^\ (b,, b„. . . b,) = 0

588 PROCEEDINGS OP THE AMERICAN ACADEMY.

Consequently, if A„ 4^ 0, and every transformation of the group Ci**' is
commutative with 7^„, the above condition is satisfied.

§6.

Let
(20) 7; = T„..

Then every transformation of the sub-group G^i'"'' is commutative with
7^,. Therefore, if A„ 4^ 0, 2 a! ^ X^ is commutative with 2 a^ Xj. Whence
it follows that

(21) J a - a' ^= -^ a ^ - a' ^ -'a -* a'

is the identical transformation.

If, however, A„ — 0, it does not necessarily follow that 2 Oj Xj and
2 a' J Xj are commutative ; and therefore we do not necessarily have

rp rp rp— 1

a — a'

E. g., let r = 5, and let

V— '^ V- ^ Tf- <^ It'- ^ A'— ^

"^i-^iJ7i' "^•^-''2^; ^3-'"«J7i' ^"-""3(^x7 ^^-'^*(^:r;

Then A„ = 0 if eitlier Oj or a^ is an even multiple, not zero, of ir /y/ — 1.

Let oj = 0, ao = 2 ^- T vA^ 4: 0, rtg = 0,

a'l = 2 k' TT /y/ — 1 , a'.2 = 2 A; TT y^ — 1 4^ 0, a '5 = fls,
where A: and ^•' are integers. Then A„ = 0, A,,/ = 0 ; and Ta = T^a'- But 2 «;■ X,- is
not commutative with 2 a'j Xj unless (('4 = a^. Moreover Ta- a' is not the identical
transformation (i. e., Ta-a< 4= Ta Ta'~^) unless a\ — a^.

When T,^ = T„, and A„ = 0, it does not necessarily follow that A„, = 0.
Thus, in the case of the group just considered, if

«i = 2k7r ^/^-^ ^ 0, a2 = 2 k' TT ^^ ^ 0,
a\ = 0, a',, = 0, a'a = 0, a\ = 0, a\ — 05,
we have T^ — ?„-, and A„ = 0 ; but A^, ^ 0.

§7.

The equations of the general infinitesimal transformations of the ad-
joined group r of group G are in Cayley's matrical notation

(22) (d'l, a\, . . . a',) = (I + 8<(^aO«n ^2, • • • «r)-

The successive application of this infinitesimal transformation to the
manifold cii, a^, . . . a^ gives the general transformation of F, namely,

(23) (a/i', «,", . . . a,^^) = (e<^«0^i, <t., . . . «,).

TABER. — ON SINGULAR TRANSFORMATIONS. 589

If

(24) (ax'^ a^'\ . . . «,.'=') = (.^^^a,<^>, a^\ . . . a;%

we have

(25) («!<-% «/', . . . a/2)) = (e-^^ A^«^, „^^ . . . ^^).

Thus the general transformation of the adjoined may be represented by
the matrix e*^" ; and the result of the two successive transformations
e^"* and e*^^ is represented by the matrix e^^ e^" obtained by their com-
position.

If yi, yo, . . . Yr are so chosen that

(26) T, = Tp T^,
that is, if

(27) yj = cpj (ai, . . . a„ ^1, . . . )8,)

U = 1, 2, . . . r),

then ,

(28) e'^y = e'^P e^-.

If G contains no extraordinary infinitesimal transformation, V has the
same structure as G and the r parameters a are all essential.* In this
case if we put

(29) e*^" = e'^'^ e*^",

the a's, as functions of t, are defined by the differential equations of
p. 584 ; thus we have

It will be found, however, that the symbolic equation (29), in general,
defines more systems of functions a than the symbolic equation

E. g., let r =.2, and

■^ _ d ^ _ d d

Then, if Tc=Tn Ta,

_ ^2 + ^2

^ «i e*'- —r-, — + ^1 — 1 — .

* Lie : Transformationsgruppen, I. p. 277.

690 PROCEEDINGS OF THE AMERICAN ACADEMY.

C2= a.2 + 1)2-
For this group

an(], therefore, the equations of r are

{a\, a'.,) =

(ai,fl2)= (e"2, ai

Therefore, if
we have

^2_1

O.I

0, 0

^«i, a^)-

6*7 = e*^e'*

7i =

[^ + ^2 + 2 A: TT -y/- 1

a, e^

where k is any integer.

7-2 = Oi + ^2 + 2 A T /y/— 1,

1 , «P= - 1

If G coutains one or more extraordinary infinitesimal transformations,
and yi, 72, . . • y,- are determined by equations ('27), we sliall still have
e^y =^ e^^ e^°- ; but these conditions, though sufficient, are not all neces-
sary. If G contains just s independent extraordinary infinitesimal trans-
formations, just r — s of the parameters a are essential.

If the s independent extraordinary infinitesimal transformations of G
are X^, X.^ . . . X^, then ai, ao, . . . a^ do not appear in ^a ; and if

we have for the determination of the remaining parameters a the differ-
ential equations

c^ «.s + 1

dt

do,
dt

'« + 1. s + 1

-, . .

s + 2, « + 1

^^^

6 + 2,

^,.),

Aa

where f^'" denotes the result of substituting as + 1, . . . a,, for a^^. i, . . . a^,

Aa

respectively in the functions -^ of p. 583.

Afl

If the group G is continuous, the adjoined group V is continuous.
Therefore, if T is discontinuous, group G and every group of the same
structure is discontinuous.

TABER. — ON SINGULAR TRANSFORMATIONS. 591

§8.

For certain types of structure every root of the characteristic equation
of the matrix cf)„ is zero irrespective of the values of the parameters a.
In this case A„ = 1, and the group is continuous.

For two groups of the same structure, but such that A„ ± 1, one may
be continuous and the other discontinuous. As remarked above, if the
adjoined is discontinuous, group G and every group of the same structure
is discontinuous.

I give below a table exhibiting the result of an investigation by my
pupil, Mr. S. E. Slocum, of the continuity of all types of groups with
either two, three, or four parameters. It will be seen that for every
type of structure for which A„ ^ 1, there is at least one group which is
discontinuous. Mr. Slocum has found that every real group is continu-
ous which possesses any one of the several structures distinguished in the
table by an asterisk ; and that the real group which possesses the struct-
ure marked in the table thus (f) is discontinuous. Also that every real
group of the structure {Xi, X,) = 0, (Xi, Xg) = — X,, (X.,, X3) = — Xi,
is discontinuous. For this group

_ (e"' ^^^ _ 1) (e - «3 i/~i _ 1)

« i

[In what follows (Xj, X^) denotes Xj X^ — X^ Xj.^

Groups with two Parameters (r = 2).

e"- — 1

Type I. (X„X,) = X,.^- A„

Adjoined group continuous.

Parameter group discontinuous ; also group

d 9 9

9 X2 9 X2, 9 Xi

Type II. (Xi,X2.) = 0.* A„=l.
All groups of this type are continuous.

Groups with three Parameters (r = 3).
Type I. (X„ X,) = X„ (X„ X,) = 2 X„ (X„ X3) = X,.t

y' Uo^ — 4 «! as ~ V "2^ — 4 ai 03

592 PEOCEEDINGS OF THE AMERICAN ACADEMY.

Adjoined group continuous.

Group Xi ^ — , Xi 7^ — — Xo-^ — , Xo -^ — , discontinuous.

d X2 a Xi d Xo " d Xi

Type II. (Zi, X,) = 0, {X,, X,) = X„ (X, X^) = /3 X, {(3 :\^ 0, 1)*
Adjoined group ~ a^ - — , — a^ (3 ^r— , Oj ^r \- cto (3

cy «! d a2 cy «i d a^

discontinuous. Tlierefore, all groups of this type are discontinuous ;

e. g., group

9 9,^9,9 9

Xi ^ h P Xo -^ h a

d Xi] d Xi d x^ d Xs d X2

Type III. (Zi, X,) = 0, (X^, X,) = X„ {X„ X^) = X,*

Adjoined group continuous.

Parameter group discontinuous ; also group

9 9 9 9 9

9 Xi " 9 x-i 9 X3' 9 Xi 9 Xo

Type IV. (Xi, X,) = 0, {X,, X,) = X,, (X„ X,) =X,\ X,*
Adjoined group

9 9 9 , ^9 9

— «3 ^ , — «3 ^ «3 7i , («1 + «2) 7=j 1- «2 Tj— -,

d ai d tti d a^ d fi\ d «2

discontinuous. Therefore, all groups of this type are discontinuous ;
e. g., group

c, 9 9 9 9 9 9 9
^ ^3 :s 1" ^2 7^ h ^1 7^ 1- ^2 n 1" ^3 ^=r~"5

d Xi d X3 d Xi d X2 d x^ d Xi d X3

Type V. (Zi, X,) = 0, {X„ X,) = X„ {X„ X^) = 0*

6*"^- 1

a

■3

Adjoined group continuous.

TABER. — ON SINGULAR TRANSFORMATIONS. 693

Parameter group discoutinuous ; also group

9 d 9,9

^IT^ • +

9 Xi 9 x^ 9 Xi 9 X3
Type VI. (Xx, X) = 0, (X\, X,) = 0, (X„ X,) = Xi*

All groups of this type are continuous.

Type VII. (Xx, X,) = 0, (X„ X,) = 0, (X„ X,) = 0*

All groups of this type are continuous.

Groups with four Parameters (r = 4).
A. Without three-parameter involution group.

( (X„ X,) = X„ (Xx, X,) = 2 X,, {X,, Z3)
^yP^ ■"• 1 (Xx, X,) = 0, {X,, X,) = 0, (X3, X,) =

g V «2^ — 4 fli as _ 1 g — V 32^ — 4 a-, flj _ 2^

V a^^ — 4 Ox 03 ~ V "s"^ — 4 ax «:

3

Adjoined group continuous.

Group Xi ^ — , Xi 7;; X2 :^ — , x^ ^ — , X3 ^ — , discontinuous.

^9x2 9xi 9x2 9 xi 9x3

( (Xx, X) = 0, (Xu X,) = 0, (X2, X3) = Xx,
lype 11. j ^^^^ ^^^ ^ ^ ^^^ ^^^^ -^^^ ^ ^^^ (^^^ ^^)

= (;3-l)X3(/3 4^1).

^«-V «, JV «4/3 JV «4(/3-l)/
Adjoined group

o5 5 P 9 /oix^ /jc'

9^,'~'^'9a, '''9a.: '^'9a, "* ^^ ^ 9 a^ ^' ^ 9 a.

5 f'.2 5 f/s

discontinuous. Therefore, all groups of this type are discontinuous;
e. g., group

9 9 9 J_M R -L^A^

9^: 9^; '''9x: ^^5xx"^^^'5x,'^5x;

VOL. XXXV. — 38

594 PROCEEDINGS OP THE AMERICAN ACADEMY.

lype lii. I ^x„ X,) = 2 X„ (X, X,) = X,, (X3, X,) = 2X, + X,.^

Adjoined group

5 9 9 9^9 9

2 «i 7=r— + («2 + 2 ag) ^ h ag

9 cii 9 a» 9 0,3

discoiitiuuous. Therefore, all groups of this type are discontinuous ;
e. g., group

c 3 , ^ c^ 9 , 9,9 9 9
2 Xs h X2 ^ — , 2xi^ \- X3 ^ h

' 9 Xi 9 x^ 9 Xi 9 x-i 9 x.^ 9 Xi 9 x^

j {X„ X,) = 0, (X„ X3) = 0, (X, X3) = Xi,
^^^ '■ I {X„ X,) = X„ {X,, X,) = X„ iX„ X,) = 0*

Adjoined group continuous.

^ 5 9 9 9 , c),5,. ^.
Group ^ — , -^ — , Xi TT — , Xi ^ h Xo ;r 1- ^ — , discontinuous.

9 x^ d Xi d X2 9 Xi 9X2 d x.^ ■

( {X„ X) = 0, (X„ X3) = 0, (X, X) = x„

e"^ - 1 \ /e"^-l\

«4 / \ «3 /

Adjoined group continuous.

Parameter group discontinuous ; also group

9 9.99 9

Xi :r — +

9 Xi 9 Xi 9 X3' 9 X.2 "9 X.2

B. With three-parameter involution group.

( {X„ X,) = (X, xo = (X3, xo = 0,

J, X4) = y A3

(a ^ /? 4^ y).

'^^P^ ^' "I (Ai', X]) = a xi, (X, A4) = (S X,, (A3, -^4) = 7 X3*

A.,

//^"-I\ //*' -1\ // ^" - 1 \

\ «4a y V «4/^ / \ ''47 /"

TABER. — ON SINGULAR TRANSFORMATIONS. 595

Adjoined group

c/ «! ' cy «2 d a-i d tti d (to d «3

discontinuous. Therefore, all groups of this type are discontinuous;
e. g., group

9 9 9 ^ j_ R _^ _L __^ _L _^

5^' 9^: 9^: ''''' 9V,^ '-''' 9x,^'^''' 9x,^ 9x,

i (X„ X,) = {X„ X,,) = (A-3, X\) = 0,
lype 11. <j ^^^^ ^,^ ^ ^ ^Y„ (A'2, XJ = (3 X„ (X„ X,)

= X, + ^ X,*

Adjoined group

(a ^ ft).
e"4«_ i\ /e"*^ _ 1

ri o

9 r, 9 9 r> d C

+ (rt2 /3 + fig) ^^ \- as(3

5«3

discontinuous. Therefore, all groups of this type are discontinuous ;
6. g., group

5 9 9 5, 5^„/5_^ _P_\ , _5_

5^/ 5^2' 5^3' ''^5^2'^"'''5^i"^^V''5^2'^'"'5:r3;+5:r,'

^ ^^^ J (X„ X2) = (X2, X3) = (X3, Xi) = 0,

lype 111. I ^^^,^^ ^.^^ ^ ^^.^^ ^^^.^^ -^.^^ ^ ^^.^ ^ ^^^^ (^^^ Y,)

= Xo + A'g.
//*- 1\3

*«

a^

Adjoined group

9 9 9 9 5 / , N 5

* 5 «! 5 f/i 9 (In * d do d Us ^ d I'l

9 a.^ ■'9 (I3

discontinuous. Therefore, all groups of this type are discontinuous;
e. g., group

9 9 9 9 9 9 9 9

2 Xo 7c — 4" Xo ^ — + Xi ^ 4" X2 -^ ~r X3 - , - , _ , _
^ 9 Xi 9 x.^ 9 x^ 9 x^ d x^ d Xi d Xo 9 Xg

596 PROCEEDINGS OF THE AMERICAN ACADEMY.

All groups of this type are continuous.

Type IV. f (^'i' ^2) = (^2, ^3) = (^^^1) = 0, ^ ^

(«4=r)

\ Cda J \ Ui y J

Adjoined group

a a d d B d

discontinuous. Therefore, all groups of this type are discontinuous ;

e- g., group

9 9 9 (9 9 \ 9 9

-^ — J Tij— , ^ — ? "i^i 0^ + ^20 — +y2^3o — +n — •

( (Xi, X2) = (X2, X3) = (A'3, A'O = 0,

Adjoined group

5 9 9 9 9 , , , ^ 9 , 9

c/«i 5a2 5ao 5a.g c/('i 5^/2 i? «3

discontinuous. Therefore, all groups of this type are discontinuous;
e. g., group

9 9, 9 9 9 9 9

d Xi d X2 9X3 d x^ d Xi d X2 d x^

, f (x„ X2) = (X2, X3) = (X3, xo ~ 0,

/P^ • 1 (Xi, X4) = 0, (X, X4) ^ 0, (X3, X4) = X2.*

A„ = 1.

All groups of this type are continuous.

f (Xi, X2) = (X2, X3) = (X3, XO = 0,

#

A„

Adjoined group continuous.

TABER. — ON SINGULAR TRANSFORMATIONS. 697

Parameter group discontinuous ; also group

9 9 9 9 9 9 9

0" Xi cy aro d x^ d Xi d Xo 9 x^ 9 x^

xype i. I ^^^.^^ ^^^^ _ ^^ ^^^^ ^-^.^^ _ ^^ ^^^^ ^^^ _ ^^

All groups of this type are continuous.

P. S. Since what precedes was written, I have found that the infiu-
itesinaal transformations of the parameter group, given in equations (13),
§ 2 (from which follow the differential equations (16), §3), had already
been obtained by Schur and Engel ; and that to Engel is due the deter-
mination of the values of the a's for which A„ = 0. See Lie : Trans-
formationsgruppen, III, pp. 760, 794.

June 8, 1900.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXV. No. 27. — June, 1900.

PROCEEDINGS OF THE ACADEMY, 1899-1900.

A TABLE OF ATOMIC WEIGHTS. By Theodore William
Richards.

REPORT OF THE COUNCIL: BIOGRAPHICAL NOTICES.
Epes Sargent Dixwell. By Charles P. Bowditch.
John Cummings. By William II. Niles.
John Codman Ropes. By John Fiske.
John Lowell. By Thornton K. Lothrop.

OFFICERS AND COMMITTEES FOR 1899-1900.

LIST OF THE FELLOWS AND FOREIGN HONORARY
MEMBERS.

STATUTES AND STANDING VOTES.

INDEX.

(Title-Page and Table of Contents.)

PROCEEDINGS.

Nine hundred and eighth Meeting.

May 10, 1899. — Annual Meeting.

Vice-President Hyatt in the chair.

In the absence of the Recording Secretary, William E. Story
was elected Recording Secretary pro tempore.

The Chair announced the death of Alvan Wentworth Chap-
man, of Appalachicola, Associate Fellow in Class II., Section 2.

The Corresponding Secretary read letters from Felipe Valle,
announcing his appointment as Director of the Astronomical
Observatory of Tacubaya ; from Angel Anguiano, announcing
his appointment as Director of the Mexican Geodetic Commis-
sion ; from the Geographical Society of Madrid, announcing the
death of its President, Francisco Coello de Portugal y Quesada ;
from H. P. Talbot and O. F. Wadsworth, accepting Fellowship
in the Academy ; from Charles D. Walcott, acknowledging his
election as Associate Fellow ; and from Oliver Heaviside,
acknowledging his election as Foreign Honorary Member. A
letter from the University of Cambridge inviting the Academy
to appoint a delegate to the jubilee of Sir George Gabriel
Stokes, Bart., in June, 1899, was read, and, on motion of
Charles R. Cross, it was

Voted, That the President and Corresponding Secretary send
a letter of congratulation to Sir George G. Stokes.

A letter from the Commission on Atomic Weights of the
German Chemical Society, inviting " Chemical Societies and
similar Institutions of all countries to assist in the formation of
an International Commission," was read, and, on motion of the
Corresponding Secretary, it was

602 PROCEEDINGS OF THE AMERICAN ACADEMY

Voted, That a committee of three be appointed to represent
the Academy in the International Commission on Atomic
Weights, The Cliair appointed Theodore W. Richards, Wol-
cott Gibbs, and Ira Remsen members of this committee.

The Corresponding Secretary announced that the Committee
to select a delegate to the International Congress of Orient-
alists at Rome in October next had appointed Charles Rock-
well Lanman of Cambridge, and that he had accepted.

The Corresponding Secretary i-ead the report of the Council.*

The Treasurer presented his annual report of which the
following is an abstract : —

General Fund.

Heceipts.

Balance, April 30, 1898 $3,481.95

Assessments $990.00

Sale of publications 48.52 $1,038.52

Income from investments 4,850.63

Return of bank tax 45.00

Donations 105.00 0.039.15

$9,521.10
Expenditures.

General expenses $2,023.69

Library expenses 1,559.70

Publishing expenses 2.071.95 $5,655.34

Investments 3,833.15

Balance, April 29, 1899 • 32.61

$9,521.10

RuMFORD Fund.

Receipts.

Balance, April 30, 1898 $1,778.44

Income 2,266.03

Return of bank tax 93.53

Return of Investment 1,034.99

$5,172.99

* See Proceedings, XXXIV. p. 639.

OF ARTS AND SCIENCES. 603

Expenditures.

Books $57.01

Publishing 790.88

Investigations 400.00

Rent 10.00 $1,257.89

Balance, AprU 29, 1899 . . . 3,915.10

$5,172.99

Warren Fund.

Receipts. ^

Balance, April 30, 1898 $925.95

Income 853.48

$1,779.43

Expenditures.

Investigations $630.00

Balance, April 29, 1899 1,149.43

$1,779.43
Building Fund.

Receipts.

Balance, April 30, 1898 $1,268.14

Income 504.83

$1,772.97

Expenditures.

Investment $1,466.94

Balance, April 29, 1899 306.03

$1,772.97

The annual report of the Librarian was presented and showed
that 3284 books and pamphlets had been added to the Library
during the year, of which 2432 were obtained by gift and ex-
change and 852 purchased, and 317 volumes were bound at an
expense of 8419.70. The total expenditure for books, periodicals,
and binding amounted to 81004.17. There were borrowed 235
books by 30 persons, of whom 18 were Fellows of the Academy.

The following reports were also presented: —

604 PROCEEDINGS OF THE AMERICAN ACADEMY

Report of the Rumfoed Committee.

Boston, May 10, 1899.

During the past year the Rumford Committee has made the following
appiopriations from the fund of $1000 placed at its disposition at the
last Annual Meeting, for the furtherance of researches in light and heat.

To Professor Theodore W, Richards of Harvard University, $200
for the construction of a micro-kinetoscope, the immediate application of
which is to be the study of the birth and growth of crystals.

To Professor W. C. Sabine of Harvard University, $200 for the con-
tinuation of his researches on jthe wave-leugths of ultra-violet radiations.

To Professor Henry Crew of the Northwestern University, a sum not
exceeding $200 for the continuation of his researches upon the spectrum
of the electric arc.

To Professor Arthur G. Webster of Clark University, $200 for a
research upon the distribution of energy in various spectra by means of
the Michelson interferometer and the radiometer.

At a meeting held on April 12, 1899, it was voted that the Committee
recommend to the Academy that the volumes necessary to complete the
set of the " Fortschiitte der Physik " in the library of the Academy, be
purchased from the Rumford Fund.

At the same meeting it was also voted that the Committee recommend
to the Academy the appropriation of one hundred and twenty dollars
from the Rumford Fund for the purchase and binding of the usual peri-
odicals of the current fiscal year.

At a meeting held on May 3, 1899, it was voted that the Academy be
asked to make the usual appropriation of $1000, at the Annual Meet-
ing, to be expended at the discretion of the Committee in furtherance of
research.

At the April meeting of the Committee it was voted for the first time,
" that the Rumford Committee recommend to the Academy the award of
the Rumford Medal to Mr, Charles F. Brush, for the Practical Develop-
ment of Electric Arc Lighting." At the May meeting of the Committee
the same resolution was voted for the second time.

In order to ascertain the stage of advancement of the various re-
searches in aid of which appropriations have been granted from the
Rumford Fund, a request for information was sent to such grantees as
were understood not to have made a final report, together with the fol-
lowing extracts from the records of the Committee : —

Nov. 10, 1897. "It was voted that, in future, recipients of grants for

OF ARTS AND SCIENCES. 605

investigations be requested to make a report annually as to the state of
the work for which the grant was made."

June 8, 1898. " It was voted that in the judgment of the Committee,
persons carrying on researches with the aid of the Rumford Fund should
submit to the Academy an account of their researches not less complete
than that published elsewhere. These researches may be published in
any place or form, with the proviso that due recognition be made of the
grant, and the presentation of the paper to the Academy."

In answer to this request, reports have been received of which the
following are summaries : —

Professor Henry Crew has published a paper " On the Sources of
Luminosity in the Electric Arc " in the Proceedings of the Academy for
June, 1898. He expects to continue the prosecution of his research
during the coming summer with the aid of the further appropriation
made by the Rumford Committee for that purpose.

Professor B. 0. Peirce reports that a continuation of his work upon
the thermal conductivity of poor conductors is in progress, the substance
vulcanite being at present a particular subject of study. An extended
paper " On Thermal Conductivities of Certain Poor Conductors " was
published in the Proceedings of the Academy for August, 1898.

Professor E. H. Hall is pursuing his studies upon the thermal con-
ductivity of wrought iron, having published in the Proceedings for
February, 1899, a paper " On the Thermal Conductivity of Cast Iron."

Professor Edward L. Nichols reports that his research upon the radia-
tion from carbon at high temperatures is making good progress. The
study of the acetylene flame as a standard of light, the calibration of
thermo-elements, and the application of the platinum-rhodium thermo-
element to the determination of the actual temperature of the carbon,
have occupied much time. Measurements with the spectro-photometer
are now in progress upon the visible radiations from carbon rods as com-
pared with those of like character with the acetylene flame. A study of
the distribution of energy in the spectrum of the acetylene flame and in
that of incandescent carbon will follow.

Professor W. C. Sabine says with regard to his investigation upon the
study of very short light waves : " In order to explain the advance which
he has made [Mr. Theodore Lyman, who has made the actual measure-
ments], it will be necessary to refer to the work of Schumann. This
work was done with a fluorite prism, and wave lengths were found by
extrapolation, the statement being made that speculum metal would not
reflect the shorter waves. Mr. Lyman has been using a concave grating

606 PROCEEDINGS OF THE AMERICAN ACADEMY

with success, and has got as low as wave length 900 tenth-meters,
Schumauu's estimated limit being 1000. The latter worked entirely
with the spectra of gases, and stated that he could not go below wave
length 1600 for metals."

Professor George E. Hale states that the spectro-heliograph, in the
construction of which he has been aided by a grant from the Rumford
Fund, is approaching completion. He writes that " devices have been
introduced whereby photographs of prominences or faculae can be taken
simultaneously in two different lines of the spectrum. In the case of
eruptive prominences the comparison of photographs made in this way
may prove to be instructive."

Professor Theodore W. Richards states that he has begun his investi-
gation of the birth and growth of crystals as studied by the micro-
kinetoscope, and has already obtained various excellent photographs
illustrating these phenomena, which also give promise of interesting
results as to the rate of growth of different crystals.

Charles R. Cross, Chairman.

Report of the C. M. Warren Committee.

10 May, 1899.

At the last Annual Meeting of the Academy the sum of $600 from the
income of the Warren Fund, was granted to Professor C. F. Mabery,
of Cleveland, Ohio, in furtherance of his researches on petroleum.
Several papers explanatory of Professor Mabery's results have been
published during the year; and it is well understood that his work is
being prosecuted all the while with ardor and success.

A research by Professor F. C. Phillips, of Allegheny City, in aid of
which a grant of $200 was made in 1896, has been in so far completed
that an account of it was published, in November last, in the Proceed-
ings of the Academy, under the title " On Fluctuations in the Composi-
tion of Natural Gas."

A grant in the sum of $200 made to Professor H. O. Hofman, of
Boston, in 1897, and supplemented in 1898 by an additional grant of
130, has also borne good fruit. Professor Hofman's subject was "The
Fusibility of Slags." He has explained to me that interesting and
important results have been obtained, which he intends to publish in the
near future.

F. H. Storer, Chairman.

OF ARTS AND SCIENCES. 607

Report of the Committee of Pdblication.

Boston, May 10, 1899.
The Publisbiug Committee begs leave to report that there have been
issued during the last academic year ten numbers of Vol. XXXIII. and
the first twenty numbers of Vol. XXXIV. of the Proceedings, aggregat-
ing 796 pages and 17 plates; besides one number of Vol. XII. of the
Memoirs containing 36 pages and 7 plates. Five numbers of the Pro-
ceedings have been printed at the cost of the Rumfox'd Fund. The
expenditure for the remaining publications was $2334.14. The appro-
priation from the General Fund was $2500 and the sales $48.52, making
a sum available for publication of $2548.52, and leaving therefore an
unexpended balance of $214.38. The Academy has never before pub-
lished so many pages in a single year.

Samuel H. Scddder, Chairman.

m

On the recommendation of the Committee of Finance, it was
Voted, To make the following appropriations from the income
of the General Fund for the ensuing year : —

For general expenses $2000

For the library 1500

For publishing 2400

Voted, That the assessment for the ensuing year be five
dollars.

Voted, That the Treasurer be authorized to pay from the
funds of the Academy any bills approved by the Librarian in-
curred on account of the expense of moving the library.

On the recommendation of the Rumford Committee, it was

Voted, That the sum of one thousand dollars (11000) from
tlie income of the Rumford Fund be placed at the disposal of
the Rumford Committee to be expended in aid of investigations
on Light and Heat, payments to be made on the order of the
Chairman of the Committee.

Voted, That the volumes necessary to complete the set of the
" Fortschritte der Physik " in the library be purchased at the
expense of the income of the Rumford Fund ; also, that one
hundred and twenty dollars ($120) be appropriated from the

608 PROCEEDINGS OF THE AMERICAN ACADEMY

income of the Rumford Fund for the purchase and binding of
periodicals.

Ou the recommendation of the C. M. Warren Committee, it
was

Voted, That the sum of six hundred dollars Ci^600) from
the income of the Warren Fund be granted to Charles F.
Mabery, of Cleveland, Ohio, in aid of his researches on the
chemistry of petroleum.

In accordance with the recommendation of the Rumford
Committee, it was

Voted, To award the Rumford Premium to Charles F. Brush
for the practical development of electric arc lighting.

The annual election resulted in the choice of the following
officers and committees : — ^

Alexander Agassiz, President.

John Teowbridge, Vice-President for Class I.

Alpheus Hyatt, Vice-President for Class II.

Augustus Lowell, Vice-President for Class III.

Samuel H. Scudder, Corresponding Secretary.

William Watson, Recording Secretary.

Francis Blake, Treasurer.

A. Lawrence Rotch, Librarian.

Councillors.

Henky Taber,

Theodore W. Richards, \ of Class I.

Harry M. Goodwin,

Benjamin L. Robinson,

William T. Councilman, \ of Class IL

John E. Wolff,

Barrett Wendell,

Edward Robinson, \- of Class HL

'J

James B. Ames,

Member of the Committee of Finance.
Augustus Lowell.

OF ARTS AND SCIENCES. 609

Riaiiford Committee.

Erasmus D. Leavitt, Amos E. Dolbear,
Edward C. Pickering, Arthur G. Webster,
Charles R. Cross, Theodore W. Richards,

Thomas C. Mexdenhall.

0. M. Warren Committee.

Francis H. Storer, Henry B. Hill,

Charles L. Jackson, Leonard P. Kinnicutt,
Samuel Cabot, Arthur M. Comey,

Robert H. Richards.

The Chair appointed the following standing committees : —

Committee of Publication.

Samuel H. Scudder, Seth C. Chandler,

Crawford H. Toy.

Committee on the Library.

A. Lawrence Rotch, Henry W. Haynes,

Samuel Henshaw.

Auditing Committee.
John C. Ropes, Eliot C. Clarke.

The following gentlemen were elected members of the
Academy : —

AVilliam Elwood Bjerlj'^, of Cambridge, to be a Resident
Fellow in Class I., Section 1 (Mathematics and Astronomy).

William Henry Pickering, of Cambridge, to be a Resident
Fellow in Class I., Section 1.

Henry Lefavour, of Williamstown, to be a Resident Fellow
ill Class I., Section 2 (Physics).

Charles Russell, Baron Russell of Killowen, of Tadworth, to
be a Foreign Honorary Member in Class HI., Section 1 (Philos-
ophy and Jurisprudence).
VOL. XXXV. — 39

610 PROCEEDINGS OP THE AMERICAN ACADEMY

The following papers were presented by title : —

On the Thermal Conductivity of Vulcanite. By B. O. Peirce.

Ferrous Iodide. By C. Loring Jackson and Ira H. Derby.

Nine hundred and ninth Meeting.

June 14, 1899.

Tlie President in the chair.

The Corresponding Secretary read letters from W. E. Byerly
and W. H. Pickering, accepting Fellowship, and from Lord
Russell of Killowen, acknowledging his election as Foreign
Honorary Member. Letters from Marion T. Hosmer, soliciting
a subscription on behalf of the Rumford Historical Society,
and from the University of Pennsylvania, the American Philo-
sophical Society, and other Philadelphia Societies, inviting at-
tendance at the presentation of the Franklin statue, were
referred to the Corresponding Secretar3\

The Chair announced the death of Francis Minot, Resident
Fellow in Class II., Section 4, and of Manning Ferguson Force,
Associate Fellow in Class III., Section 3.

The Rumford Medal awarded to James Edward Keeler at
the annual meeting of 1898 was presented, Edward C. Picker-
ing acting as Professor Keeler's proxy.

The President spoke of the proposed exploring expedition to
the mid-Pacific under his direction, the United States Commis-
sion of Fish and Fisheries having placed the Albatross at his
disposal. The vessel will be thoroughly equipped with the
newest apparatus for deep-sea investigations, and special ap-
pliances will be constructed for use in very deep water. The
expedition will leave San Francisco about the middle of August
for Tahiti, in the Society Islands, which will be the hea-dquarters
during the six or eight weeks required for exploration of the
Paumoto Islands. Afterwards a week or ten days will be spent
among the Tonga or Friendly Islands, and tlie expedition will
then proceed to the Fiji Islands, where a short stay will be
made. After visiting some of the Ellis and Gilbert Islands, six
or seven weeks will be devoted to the exploration of the Mar-

OP ARTS AND SCIENCES. 611

shall Islands. Between these islands and the Hawaiian Islands,
and between the latter and San Francisco, a distance of four
thousand miles, a line of deep-sea dredgings will be run, deep-sea
tow-nets being used while the dredging is going on. The
Albatross is expected to return to the United States in April
next.

The following papers were presented by title : —

The Development and Application of a General Equation for
Free Energy and Physico-chemical Equilibrium. By Gilbert
Newton Lewis. Presented by T. W. Richards.

The Electro-chemical Equivalents of Copper and Silver. By
T. W. Richards, E. Collins, and G. W. Heimrod.

A Revision of the Atomic Weight of Calcium. By T. W.
Richards.

Short Studies of North American Tryxalinae. By Samuel
H. Scudder.

The Recording Secretary read the agreement with the Mas-
sachusetts Historical Society in regard to quarters for the
Academy in the new building of the Society. Remarks on this
subject were made by the President and Augustus Lowell.

Nine hundred and tenth Meeting:-

October 11, 1899. — Stated Meeting.

The Academy met at Ellis Hall.

ViCE-PiiESiDENT Hyatt in the chair.

The Corresponding Secretary read letters from Henry Lefa-
vour, accepting Fellowship in the Academy ; from the Royal
Academy of Sciences of Turin, announcing the death of Pro-
fessor Cesare Nani ; from the National Society of Horticulture
of France, announcing the death of its first Vice-President,
Henri Ldveque de Vilmorin ; from the Students' Union at the
Polytechnic School at Zurich, inviting attendance at the
twenty-fifth jubilee of Professor A. Heim ; from Captain Con-
stantin Edler von Pott, announcing his appointment as Director
of the Hydrographic Bureau of the Imperial and Royal Navy
at Pola ; and from Charles R. Lanman, reporting his inability

612 PROCEEDINGS OF THE AMERICAN ACADEMY

to represent the Academy at the International Congress of
Orientalists. A letter from the Secretary of the organizing
committee of the International Congress of Phvsics at Paris in
1900, enclosing a prospectus and requesting that it be brought
to the notice of members, was referred to the members of the
Council from Class I.

The Corresponding Secretary reported that, in response to
letters of invitation received during the summer, the President
had appointed George P. Fisher, Associate Fellow, as a delegate
to the celebration of the one hundredth anniversary of the
Connecticut Academy of Arts and Sciences, and Vice-President
Trowbridge had appointed T. C. Mendenhall, Resident Fellow,
to represent the Academy at the celebration of the seventy-fifth
anniversary of the founding of the Franklin Institute.

The Chair announced the death of John Harrison Blake, of
Boston, Resident Fellow in Class I., Section 2; and of Robert
Wilhelm Bunsen, of Heidelberg, Foreign Honorary Member
in Class I., Section 3.

On the motion of the Corresponding Secretary, a quorum for
business not being present, it was

Voted, To meet on adjournment on the second Wednesday
in November.

Arthur G. Webster read and explained a paper presented by
A. A. Michelson on the Echelon Spectroscope.

The following papers were read by title : —

Peripheral Distribution of the Cranial Nerves of Spelerpes
hilineatus. By Mary A. Bowers. Presented by E. L. Mark,

Note on tlie Finite Continuous Groups of the Plane. B}'
F. B. Williams. Presented by Henry Taber.

Two Genera of North American Decticinae. By S. H.
Scudder.

Nine hundred and eleventh Meeting.

November 8, 1899. — Adjourned Stated Meeting,

In the absence of the regular presiding officers, Henry P.
BowDiTcn was chosen President j»ro tempore.

OP ARTS AND SCIENCES. 613

The Chair announced the death of John Codman Ropes, Res-
ident Fellow in Class III., Section 3.

Samuel IT. Scudder tendered his resignation as Corresponding
Secretary, to take effect in January, and it was accepted.

The Chair appointed from the next retiring Councillors

Theodore W. Richards, of Class L,
Benjamin L. Robinson, of Class II.,
Barrett Wendell, of Class III.,

a committee to nominate a candidate for the office of Corre-
sponding Secretary.

On the motion of the chairman of the C. M. Warren Com-
mittee, it was

Voted^ To grant permission to H. O. Hofman to publish in
the Transactions of the American Institute of Mining Engineers
or elsewhere the results of his research, for which aid was
granted by the Academy from the C, M. Warren Fund.

On the motion of the Librarian, it was

Voted, To authorize the expenditure of a sum not exceeding
two hundred dollars ($200) for furniture for the Library.

The following gentlemen were elected members of the
Academy : —

Maxime Bocher, of Cambridge, to be a Resident Fellow in
Class I., Section 1 (Mathematics and Astronomy).

William Fogg Osgood, of Cambridge, to be a Resident Fellow
in Class I., Section 1.

John Singer Sargent, of London, to be an Associate Fellow
in Class III., Section 4 (Literature and the Fine Arts).

Sir Benjamin Baker, of London, to be a Foreign Honorary
Member in Class I., Section 4 (Technology and Engineering),
in place of the late Sir Henr}^ Bessemer.

Rudyard Kipling, of Rottingdean, to be a Foreign Honorary
Member in Class III., Section 4 (Literature and the Fine Arts).

Arthur G. Webster described Maxwell's electric top and
exhibited it in operation.

The following papers were presented by title: —

Contributions from the Cryptogamic Laboratory of Harvard

614 PROCEEDINGS OF THE AMERICAN ACADEMY

University, XLI. Preliminary Diagnoses of New Species of
Laboulbeniaceae. — I. By Roland Thaxter.

jSTote on the Chief Theorem of Lie's Theory of Continuous
Groups. By S. E. Slocum. Presented by Henry Taber.

Nine hundred and twelfth Meeting.

December 13, 1899.

The Corresponding Secretary in the chair.

Letters were received from the Royal Academy of Sciences
of Turin, announcing the death of Domenico Perrero ; from the
International Congress of Ethnographical Sciences, enclosing
programmes and an invitation to attend its meetings at Paris
during the summer of 1900 ; from Maxime Bocher, John S.
Sargent, and Rudyard Kipling, accepting membership.

A letter was read from the Royal Prussian Academy of
Sciences of Berlin, announcing the celebration of the two hun-
dredth anniversary of its foundation on the 19th and 20th of
March, 1900, and inviting the American Academy to send dele-
gates. It was accordingly

Voted, To appoint John Williams White and John Eliot
Wolff, delegates to this celebration.

The chair announced the death of Epes Sargent Dixwell,
Resident Fellow in Class III., Section 2.

The following papers were read : —

Experimental and Statistical Studies on the Influence of Cold
on the Bacillus of Typhoid Fever and its Distribution ; with
special Reference to Ice Supply and the Public Health. By
William T. Sedgwick and Charles-Edward A. Winslow.

The Electrical Resistance of the Human Body. By William
L. Hooper.

The following papers were presented by title : —

A Revision of the Atomic Weight of Iron. By T. W. Rich-
ards and C. P. Baxter.

Note on the Constitution of Diparabrombenzylcyanamide. By
C. Loring Jackson and R. W. Fuller.

OF ARTS AND SCIENCES. 615

On Certain Colored Substances Derived from Nitro Com-
pounds. Third paper. By C. Loring Jackson and F. H.
Gazzolo.

Certain Derivatives of Metadibrombenzyl. By C. Loring
Jackson and W. P. Cohoe.

Nine hundred and thirteenth Meeting.

January 10, 1900. — Stated Meeting.

Vice-President Hyatt in the chair.

The Corresponding Secretary read letters from Sir B. Baker
and William Fogg Osgood, acknowledging election. He also
read circulars from the Royal Academy of Sciences of Turin,
announcing the terms of award of the Vallauri prizes, and from
the International Congress of Comparative History and the
International Congress of Horticulture, Arboriculture, and
Pomology, enclosing programmes.

A letter was received from Seabury C, Mastick, Secretary of
the committee on the modification of the Federal Legacy Tax,
enclosing a petition setting forth the desired modification of the
law. On the motion of William E. Story, it was

Voted, That the Acting President be instructed to sign this
petition on behalf of the Academy.

The Chair announced the followinsf deaths : —

Sir John William Dawson, of Montreal, Associate Fellow in
Class II., Section I.

William Alexander Hammond, of Washington, Associate
Fellow in Class II., Section 4.

Sir James Paget, Bart., of London, Foreign Honorary Mem-
ber in Class II., Section 4.

The vacancy occasioned by the resignation of Samuel H.
Scudder was filled by the election of

William M. Davis, Corresponding Secretary.

On the motion of Augustus Lowell, it was

Voted, That the thanks of the Academy be tendered to

616 PROCEEDINGS OF THE AMERICAN ACADEMY

Samuel Hubbard Scudder for his faithful and efficient services
as Corresponding Secretary.

The following gentlemen were elected members of the
Academy : —

James Mason Crafts, of Boston, as Resident Fellow in Class
I., Section 3 (Chemistry).

Joseph Hodges Choate, of New York, as Associate Fellow in
Class in., Section 1 (Philosophy and Jurisprudence), in place
of the late Thomas Mclntyre Cooley.

William Wirt Howe, of New Orleans, as Associate Fellow in
Class in., Section 1.

William Mitchell, of Saint Paul, as Associate Fellow in Class
III., Section 1, in place of the late Ezekiel Gilman Robinson.

Sir George Otto Trevelyan, Bart., of London, as Foreign
Honorary Member in Class III., Section 3 (Political Economy
and Historj').

The vacancies in the Auditino- Committee occasioned bv the
resignation of Eliot C. Clarke and the death of John C. Ropes
were filled by the appointment of the following

Auditing Committee :
Henry G. Denny, William L. Richaedson.

On the recommendation of the Rumford Committee, it was
Voted, That the Academy appropriate one hundred dollars
($100) to Theodore W. Richards for the prosecution of a re-
search on the transition point of crystallized salts.

Voted, That the Treasurer be authorized to make arrange-
ments for the sale to the public of the Life and Works of Count
Rumford.

John S. Kingsley read a paper entitled " The Ancestry of the
Mammalia."

The following papers were presented by title : —
Contributions from the Gray Herbarium of Harvard Univer-
sity. New Series. — No. XVIII.

I. New Species and Varieties of Mexican Plants. By J.
M. Greenman.

OP ARTS AND SCIENCES. 617

11. Synopses of the Genera Jaegeria and Russelia. By B.
L. Robinson.

III. New Dioscoreas from Mexico. By E. B. Uline.

IV. New Phanerogams chiefly Gamopetalae from Mexico

and Centi'al America. By B. L. Robinson.

Nine hundred and fourteenth Meeting.

February 14, 1900.

In the absence of the regular presiding officers, Samuel H.
ScuDDER acted as President pro tempore.

The chair announced the death of the following Foreign
Honorary Members : —

James Martineau, of London, Class III., Section 1.

Carl Friedrich Rammelsberg, of Berlin, Class II., Section 1.

John Ruskin, of Coniston, Class III., Section 4.

The following letters were received : from Mathilde Rammels-
berg, announcing the death of Carl Friedrich Rammelsberg,
Foreign Honorary Member of the Academy in Class II,, Section
1 ; from W. M. Davis, accepting his election as Corresponding
Secretary ; from John Williams White, accepting his appoint-
ment as Delegate to the celebration of the 200th anniversary of
the foundation of the Royal Prussian Academy of Sciences.

The following papers were presented by title : —

The Metamerism of the Hirudinea. By W. E. Castle. Pre-
sented by E. L. Mark.

The Freshwater Tertiary Formations of the Rocky Mountain
region. By W. M. Davis.

On the Determination of Sulphuric Acid in the Presence of
Iron : a Note on Solid Solutions. By Theodore William
Richards.

The Species of the Orthopteran Genus Derotmema. By
Samuel H. Scudder.

Crawford H. Toy gave an informal account of a new theory
of totemism.

618 PROCEEDINGS OF THE AMERICAN ACADEMY

Nine hundred and fifteenth Meeting.

March 14, 1900 — Stated Meeting.

The Academy met, by invitation of John E. Hudson, at the
Algonquin Club.

Vice-President Trowbridge in the chair.

The Corresponding Secretary read letters from Joseph H.
Clioate and Sir G. O. Trevelyan acknowledging their election
into the Academy. He also exhibited a bronze copy of the
medal struck in honor of Sir George G. Stokes's Jubilee in 1899,
presented to the Academy by the University of Cambridge.

On motion of the Recording Secretary, it was

Voted, To meet, on adjournment, on the second Wednesday
in April.

The following gentlemen were elected members of the
Academy : —

Arlo Bates, of Boston, to be a Resident Fellow in Class IH.,
Section 4 (Literature and the Fine Arts).

Liberty Hyde Bailey, of Ithaca, to be an Associate Fellow in
Class n., Section 2 (Botany), in place of the late Alvaii Went-
worth Chapman.

Friedrich Kohlrausch, of Berlin, to be a Foreign Honorary
Member in Class I., Section 2 (Physics).

The Chair appointed the following Councillors to act as
Nominating Committee : —

Theodoee W. Richards, of Class L,
Benjamin L. Robinson, of Class IL,
Barrett Wendell, of Class KL

At the request of the Chair, Charles R. Cross, Chairman of
the Rumford Committee, stated the grounds for the award of
the Rumford Premium to Charles F. Brush for his researches in
electric arc lighting.

The Acting President then presented the medal to Mr. Brush,
who, in response, gave an account of his early experiments in
electricity.

Elihu Thomson described new electric apparatus for high

OF ARTS AND SCIENCES. 619

potentials, and exhibited an improved form of his dynamo-elec-
tric machine.

The folio wing, papers were presented by title : —
View of the Carboniferous Fauna of the Narragansett Basin.
By A. S. Packard.

Contributions from the Gray Herbarium of Harvard Univer-
sity. New Series. — No. XIX. By M. L. Fernald. Presented
by B. L. Robinson.
I. Synopsis of the Mexican and Central American Species of

Salvia.
n. Revision of the Mexican and Central American Solanums
of the Subsection Torvaria.
ni. Some Undescribed Mexican Phanerogams, chiefly Labiatae
and Solanaceae.
Historical Notes relating to Musical Pitch in the United
States. By Charles R. Cross.

Nine hundred and sixteenth Meeting.

April 11, 1900. — Adjourned Stated Meeting.

The President in the chair.

The Chair announced the following deaths : —

Silas Whitcomb Holman, Resident Fellow in Class I., Section 2.

Edward John Phelps, Associate Fellow in Class III, Sec-
tion 3.

The Rumford Committee made a preliminary announcement
that it would recommend, at the next annual meeting, the award
of the Rumford Premium to Carl Barus, of Providence.

The following motion was offered by W. E. Story : —

That a committee consisting of the President and two others
selected by him be appointed to consider the propriety of amend-
ing the first section of Chapter I. of the Statutes, with refer-
ence to the classification of the Fellows and Foreign Honorary
Members.

The following recommendation was offered by the Rumford
Committee : —

The Rumford Committee recommends that the Academy

G20 PROCEEDINGS OP THE AMERICAN ACADEMY

appropriate the sum of two hundred and fifty dollars (1250)
from the income of the Rumford Fund to Arthur L. Clark, of
Worcester, to aid in the prosecution of researches on the prop-
erties of vapors near the critical point.

W. M. Davis read a paper entitled " The Freshwater Tertiary
Formations of the Rocky Mountain Region."

The following papers were presented by title : —

Contributions from the Cryptogamic Laboratory of Harvard
University XLII. : Preliminary Diagnoses of New Species of
Laboulbeniacese. — II. By Roland Thaxter.

The Driving Energy of Physico-chemical Reaction and its
Temperature Coefficient. By Theodore W. Richards.

Supplementary Note on the Chief Theorem of Lie's Theory of
Finite Continuous Groups. By Stephen Elmer Slocum. Pre-
sented by Henry Taber.

On the Singular Transformations of Groups generated by
Infinitesimal Transformations. By Henry Taber.

RICHARDS. A TABLE OF ATOMIC WEIGHTS.

621

A TABLE OF ATOMIC WEIGHTS

OF Seventy-four Elements.

Compiled in April, 1900, y"rom the most Recent Data.

By Theodore William Richards.

Name.

Symbol.

Atomic
Weight.

Name.

Symbol.

Atomic
Weight.

Aluminium . .

Al

27.1

Molybdenum

Mo

96.0

Antimony . .

Sb

120.0

Neodymium . .

Nd

143.6

Argon . .

A

39.9?

Nickel ....

Ni

58.70

Arsenic . .

As

75.0

Niobium . . .

Nb = Cb

94.

Barium . .

Ba

137.43

Nitrogen . . .

N

14.04

Beryllium .

Be=Gl

9.1

Osmium . . .

Os

190.8

Bismuth . .

Bi

208.

Oxygen (standard)

0

16.000

Boron . . .

B

11.0

Palladium . . .

Pd

106.5

Bromine . .

Br

79.955

Phosphorus . .

P

31.0

Cadmium . .

Cd

112.3

Platinum . . .

Pt

195.2

Caesium . .

Cs

132.9

Potassium . . .

K

39.14

Calcium . .

Ca

40.1

Praseodymium .

Pr

140.5

Carbon . . .

C

12.001

Rhodium . . .

Rh

103.0

Cerium . .

Ce

140.

Rubidium . . .

Rb

85.44

Clilorine . .

CI

35.455

Ruthenium . .

Ru

101.7

Cliromium

Cr

52.14

Samarium ? . .

Sm

150.

Cobalt . .

Co

59.00

Scandium . . .

Sc

44.

Columbium

Cb = Nb

94.

Selenium . . .

Se

79.2

Copper . .

Cu

63.60

Silicon ....

Si

28.4

" Didymium "

Nd + Pr

142±

Silver ....

Ag

107.93

Erbium

Er

166.

Sodium . . .

Na

23.05

Fluorine .

F

19.05

Strontium . . .

Sr

87.68

Gadolinium .

Gd

156. ?

Sulphur . . .

S

32.065

Gallium .

Ga

70.0

Tantalum . . .

Ta

183.

Germanium

Ge

72.5

Tellurium . . .

Te

127.5?

Glucinum .

Gl = Be

9.1

Terbium ? . . .

Tb

IGO.

Gold . . .

Au

197.3

Thallium . . .

Tl

204.15

Helium

He

4.0?

Thorium . . .

Th

233.

Hydrogen .

H

1.0075

Thulium ? . . .

Tu

170.?

Indium . .

In

114.

Tin

Sn

119.0

Iodine . .

I

126.85

Titanium . . .

Ti

48.17

Iridium

Ir

193.0

Tungsten . . .

W

184.

Iron . . .

Fe

56.9

Uranium . . .

u

240.

Lanthanum

La

1.38.5

Vanadium . .

V

51.4

Lead . .

Pb

206.92

Ytterbium . .

Yb

173.

Lithium

Li

7.03

Yttrium . . .

Yt

89.0

Magnesium

Mg

24.30

Zinc

Zn

65.40

Manganese

Mn

55.02

Zirconium . .

Zr

90.5

Mercury .

Hg

200.0

AMERICAN ACADEMY OF ARTS AND SCIENCES.

-♦-

Report of the Council. — Presented May 9, 1900.

BIOGRAPHICAL NOTICES.

Epes Sargent Dixwell Charles P. Bowditch.

John Cummings William H. Xiles.

John Codman Ropes John Fiske.

John Lowell Thornton K. Lothrop.

REPORT OF THE COUNCIL.

The Academy has lost nineteen members by death since the
annual meeting of May 10, 1899, as follows : Six Resident Fel-
lows,— John Harrison Blake, Epes Sargent Dixwell, Charles
Franklin Dunbar, Silas Whitcomb Holman, Francis Minot,
John Codman Ropes ; eight Associate Fellows, — Albert Nich-
olas Arnold, Frederic Edwin Church, Sir John William Dawson,
Manning Ferguson Force, Daniel Raynes Goodwin, William
Alexander Hammond, Edward John Phelps, George Clinton
Swallow; and five Foreign Honorary Members, — Robert Wil-
helm Bunsen, James Martineau, Sir James Paget, Carl Fried-
rich Rammelsberg, and John Ruskin.

EPES SARGENT DIXWELL.

Epes Sargent Dixwell was born in l5oston, on the 27th of Decem-
ber, 1807, and died in Cambridge, on the 1st of December, 1899,

He was the son of Dr. John Dixwell, who graduated from Harvard
College in 1796 and received the degree of M.D. in 1811, and of Esther
Sarsent, his wife. Dr. Dixwell was a descendant of the regicide.

Mr, Dixwell was educated at the Boston Latin School, and entered
Harvard College when he was not yet sixteen years old. In college he
was recognized as an admirable scbolar, and the interest which he then
showed in literature and music continued through his life, and afforded a
solace to his declining years. Graduating from college in the class of
1827, he turned his attention to teaching, though perhaps not then
realizing that this was to be the profession of his life. The two years
during; which he was sub-master of the Boston Latin School were fol-
lowed by several years spent in the study of law in the office of one of
the most eminent lawyers of Boston, Charles G, Loring. He was
admitted to the bar in October, 1833, and for three years he practised

VOL. XXXV. — 40

626 EPES SARGENT DIXWELL,

his profession; but. in the autumn of 1836 he was chosen head-master of
the Boston Latin School, and at once took up liis work as an instructor
of youth, a work which was to occupy his best thoughts and efforts for
over a generation. In 1851, owing to certain municipal regulations, he
felt called upon to resign his position, and at once opened " The Private
Latin School " in Boylston Place, which from its start became very
popular. Here he labored for twenty-one years, and when iu 1872 he
gave up the school with which his name had been so long associated,
he found that he had taken part during the course of his teaching ia
preparing between four and five hundred graduates of Harvard for
admission to college, besides teaching many others whose lives bear
witness to his instruction.

After relinquishing the active duties of his school, Mr. Dixwell led a
quiet and retired life in Cambridge, which he had chosen for his resi-
dence in 1842. Here he spent the remainder of his long life, except
during two trips to Europe and his summer outings in the mountains or
at the seashore. He found in Cambridge the congenial literary atmos-
phere which he enjoyed, and he contributed his share to the social and
scientific life of Cambridge and of Boston. He was a member of various
societies, and among others of the American Oriental Society, the Har-
vard Musical Association, and of our own Society, having been elected
to the Academy in August, 1848. But in "The Scientific Club" of
Cambridge he took more pleasure, perhaps, than in any other, for here
he enjoyed the social intimacy of Agassiz, Peirce, Gray, Quincy,
Sparks, Walker, Hill, Everett,* Fel ton, Wyman, and of many others who
have held distinguished rank in literary, professional, and scientific
circles. With President Felton he had the still closer tie which came
from their having been roommates in college.

Though Mr. Dixwell's life was a quiet one by choice, he did not
neglect his duties as a citizen and church-goer. He early saw the
necessity of manual training in the public schools, and served for many
years as a trustee of the Parish fund of his church.

He thoroughly enjoyed his travels in Europe, as it gave him the
opportunity which he had long wished for, of wandering through classic
scenes and of seeing the wonders of classic and modern art, with which
he had already become acquainted through his studies. He is said to
have been more familiar with the topography of Rome than were many
who had lived there for many years.

He was an admirable classical scholar, and his translations into Latin
verse give evidence of his thorough knowledge of the language and of

EPES SARGENT DIXWELL. 627

his ready skill in its use. A little volume of these verses called " Otia
Seuectutis," publislied iu 1885, is a proof that old age had not dulled
the critical keenness of his mind. English verse also flowed freely from
his pen, and the lines which for many years he was accustomed to write
for the annual family gatherings, in which he always took the greatest
interest and pleasure, will long be remembered by those who had the
privilege of listening to them, for their tender pathos, their wit and
their humor.

But it is as a teacher that Mr. Dixwell is best known. As an assis-
tant instructor in the public schools of Boston, as the head-master of the
Boston Latin School, and as the master of his own private school for
many years, his influence over the education of young men was very
wide. His coming to the Latin School was welcomed as a sign that
accurate scholarship and a high moral tone would be combined with a
strict discipline in the management of that institution, and these expec-
tations were justified in both the great schools with which he was con-
nected. One of his early scholars at the Boston Latin School recalls
him as a stimulating and encouraging teacher, always insisting on
accuracy, but trying to make his pupils see the beauties of the author
whose work they were reading ; as a strict disciplinarian, possessed of a
dignified bearing, something of a martinet, and inspiring awe in those
who were not able to see beneath the somewhat cold exterior the real
kindliness of his nature.

He was not only a good classical scholar but also was well versed in all
the branches which were taught in his school ; and the pupil must have
been dull indeed who, needing an explanation of a difficult point, left
Mr. Dixwell's presence without having the difficulties thoroughly cleared
away.

Ready to excuse errors of judgment or of ignorance, he was unwilling
to condone moral faults, and his standard in the studies which were
taught under his care was high. Bright scholars were encouraged and
assisted in their work, and to dull scholars he extended sympathy and a
helping hand, but to wilfid idleness he was sternly severe. He was
anxious to deal justly with all, and if in any individual case he failed to
do so, it was not owing to any want of goodwill or kind-heartedness
towards the scholar.

With a very high standard of honor in his daily life and in his com-
munion with his fellowmen, he was not always able to make due allow-
ance for the errors of others, and this led him at times to take a somewhat
narrow view of his duty; but all who have known him well, will recall

628 JOHN CUMMINGS.

with pleasure the dignified cordiality of his greeting, his kindly sympa-
thy in others' joys and sorrows, his tenderness of heart, and his wide-
spread interest in all that was going on about him. These traits
continued to the last, even gaining in intensity as his long life approached

its end.

Charles P. Bowditch.

JOHN CUMMINGS.

Hon. John Cummings of Woburn, Mass., was elected Resident
Fellow of the Academy, in Class III., Section 3, on the 12th of
October, 1881. He was at the time well known in this section of the
State as generously promoting the teaching of Natural Science in the
public schools, and as liberally aiding institutions and individuals in
the prosecution of more advanced scientific work. He was Vice-Presi-
dent of the Boston Society of Natural History, and had materially con-
tributed to its museum by the gift of valuable collections, while the
botanical specimens were being arranged and multiplied by his liberality.

One of his noteworthy contributions was the entire financial support
he gave to the " Teachers' School of Science " for the first two years of
its existence. When he was elected to the Academy this school had
become of established value in the diffusion of scientific knowledge and
in advancing the true method of teaching from objects and natural
features.

At the same time he was identified with those who were establishing
and building up the Massachusetts Institute of Technology. He was
early made a member of its Corporation and was serving as its Treasurer,
an office which he filled with great credit during a long period in the his-
tory of the institution when it most needed the services of a treasurer
who believed in it and who had the energy and the courage to struggle
with many difficulties and discouragements. To him the Institute of
Technology is deeply indebted for its present condition.

He was filling many positions of confidence and responsibility and was
highly esteemed in business circles. He was President of the Shawmut
National Bank ; he had served as President of the Boston Board of
Trade and of the Shoe and Leather Association ; and he had held impor-
tant offices in the Massachusetts Charitable Mechanic Association, Massa-
chusetts Horticultural Society, and the Perkins Institution for the Blind;
he was State Director of the Boston and Albany Railroad, also a director
of the Eastern Railroad. He had rendered valuable service as a mem-

JOHN CODMAN ROPES. 629

ber of the Board of Finance of the Centennial Exposition at Philadelphia,
and he had most liberally served his native town of Woburn in perform-
ing the duties of many offices and in the support of its public schools.

Such are some of the facts in the life and activities of Mr. Cummings
which made him worthy of fellowship in the American Academy of Arts
and Sciences.

Besides contributing to the support of institutions and working zeal-
ously to make the teachings of science more widely known, he manifested
a very unusual interest in the study of nature. He made himself very
familiar with the various species of plants growing in the region of his
large farm in "Woburn, and he profited by every opportunity to become
acquainted with the minerals, the rocks, and the physical features of the
vicinity. He was a good example of that notable class of men who in
the earlier days of science devoted as many hours to the study of nature
as their secular duties would permit. In all the various fields of work
with which he became associated he was appreciated for the readiness
and correctness of his judgment, for the energy with which he labored in
the causes he espoused, for his invariable adherence to the highest stand-
ard of integrity and of right doing, and for his strong individuality of
character.

Wm. H. Niles.

JOHN CODMAN ROPES.

JoHX CoDMAN Ropes was born in St. Petersburg, Russia, April 28,
1836, and died at his house, 99 Mt. Vernon Street, Boston, early in the
morning of October 28, 1899. He was elected a Fellow of the Academy
in May, 1885. His father was William Ropes, a native of Salem, and
his mother was Mary Anne Codman, daughter of Hon. John Codman.
William Ropes was for some time engaged in business in St. Petersburg,
but removed to London in 1837 and lived for some time at Islington,
where a younger son, the late Dr. F. C. Ropes, was born.

After the return of the family to Boston John Ropes studied for a
while at the Chauncy Hall School, but at about the age of fourteen he
was obliged to leave school on account of a physical infirmity. Up to
that time he had been perfectly well and his figure was erect and shapely.
But at about that time a slight curvature of the spine became apparent,
which increased rapidly until it became a noticeable malformation. This
physical deformity did not embarrass the action of heart or lungs, and
during his entire life his health was remarkably good. But nevertheless

630 JOHN CODMAN ROPES.

the deformity was a very serious burden and prevented Ropes from
eneacriii'; in activities which would have been most cono-enial to him. I
might add that to those who loved him — and no one knew him who did
not — this malformation was simply non-existent. In sitting and talking
with him one never thought of him as different from other men.

After leaving the Chauncy Hall School, Ropes was for a while under
the care of Dr. Buckmiuster Brown. He then resumed his studies under
Professor Goodwin, who acted as his private tutor and fitted him for
college. He was graduated at Harvard in 1857 and soon afterwards
entered the Law School, where he received his LL.B. in 1861. While
he was proficient in the work of the Law School, it is interesting to
observe that in that early time he also took a deep interest in questions
of philosophy and religion. He was always a man of profoundly reli-
gious nature, with all the strength and earnestness of Puritanism, but
without its ascetic features. In the year of his graduating at the Law
School he received the Bowdoin prize for an essay on " The Limits of
Religious Thought," — a title which strongly suggests that his mind had
been exercised by the famous book of Dean Mansel which we were all
then reading. For a short time Ropes studied in the office of Peleg
W. Chandler and George 0. Shattuck. He was admitted to the bar
November 28, 1861, and continued to practise law in Boston until the
time of his death. In 1865 he formed a partnership with John Chipman
Gray of the class of 1859 ; and thirteen years later W. C. Loring of the
class of 1872 was added to the firm, which has since been known as
Ropes, Gray and Loring. Ropes' professional work was almost entirely
confined to the office. Possibly his physical difficulty may have had
something to do with this. He had all the qualities which might have
placed him in the very highest ranks as an advocate before the court.
He had an almost infallible scent for the essential points in a case, he
could disentangle the most complicated details, he could hunt for evidence
with a kind of cosmic patience that took everything with the utmost
deliberation but never let slip the minutest detail, and he could marshal
his arguments with a logical power that was equalled only by the artistic
beauty of statement. To hear him argue any point was a genuine delight
both to one's reason and to one's aesthetic sense. With all these rare
endowments as an advocate. Ropes confined himself principally to busi-
ness that could be done in the office, especially to the care and manage-
ment of trust estates. At the time of his death there were more than
a hundred trust estates, large and small, in his hands. He had long ago
established his reputation as a safe person fur taking care of money. He

JOHN CODMAN ROPES. 631

always showed sound judgment in making investments, and I suspect
that die secret of his success was that minute and systematic attention to
detail which characterized everything that he did.

The high qualities which might have made him a great advocate found
a rich field for their employment in work done outside of office iiours ;
and it is after all hy that literary work that he will be longest and most
widely known. The recollection of his professional work will of course
pass away or be confined to very few persons after the present genera-
tion. But his contributions to history have excellences which are likely
to secure for them a very long life. His published writings relate almost
entirely to military history, in which his two chief topics were the career
of Napoleon and the Civil War in America I think there was in
Ropes' nature an infusion of the true soldier. Had he been physically
competent for service, he would probably have taken part in the Civil
War, like his younger brother Henry, whose brief life was ended at
Gettysburg. I fancy that the incapacity for service was a real grief to
John Ropes, but it never seemed to disturb his serenity of spirit. If he
could not be useful in one way he could in another. If he could not
follow in the footsteps of Alexander, he might at least in those of Arrian.
The thought of writing a history of the Civil War was one which grew
with him into a settled purpose, and very admirable was the sort of prep-
aration which he made for it. It was natural that the subjects upper-
most in his mind should come up for discussion in the pleasant evening
hours at the club. Gradually there grew up a habit of holding meetings
at his house, meetings in which veterans of whatever rank could compare
their experiences and discuss mooted questions. Ropes strongly encour-
aged the preservation of every scrap of experience that could be put upon
record, and thus grew up the habit of preparing historical papers to be
read and discussed at these informal meetings. In this way Ropes
became the founder of a most valuable institution, — the Military His-
torical Society of Massachusetts. For several years this body held its
meetings at Ropes' house, where the speaker of the evening was apt to
dine before the meeting and where the sessions were sure to end with a
social glass and abounding good-fellowship. The publications of this
Society, though few in number, are of great value. In recent years it
has found a permanent habitation in one of the rooms of the Cadet
Armory where Ropes, some time ago, placed the larger part of his valu-
able historical library.

One of the first literary results of these studies was an elaborate exam-
ination of the Virginia Campaign of General Pope in 1862, a summary

632 JOHN CODMAN ROPES.

of which was furnished by Ropes in his volume entitled '• The Army
under Pope," being one of the volumes of Scribner's series on the Civil
War. Among other things it may be said of this book that it comjjletely
exonerates General Fitz John Porter from the charo^es brought aarainst
him after the second battle of Bull Run and npon which he was so
unjustly and cruelly condemned. I have been told that Ropes' weighty
presentation of the case exerted no small influence upon the final verdict
which declared General Porter innocent and went as far as possible
toward repairing the grievous wrong that had been done. If no other
result had come from founding the Military Historical Society, this alone
would have more than justified its existence.

But Ropes' magnum opus, " The Story of the Civil War," was unfor-
tunately never completed. It would have filled four volumes, and death
removed the author soon after the publication of the second. The loss is
one that can never be made good. Other writers of course may go over
the period which Ropes failed to cover, but nobody can complete his
book, for it is a case in which the writer's individual characteristics and
personal experience are the all-important features. We have heard
much in recent years of the advantages of the co-operative method in
writing history, whereby a hundred experts may take each a small frag-
ment of the ground to be covered. The merits of such a method are not
denied, but it has one great defect : it gives us Hamlet with the Prince
of Denmark left out. In an historical narrative nothing can make up for
the personality of the narrator. A hundred experts on the Civil War
would not fill Ropes' place for the simple reason that their hundred
individual experiences cannot be combined in the same stream of con-
sciousness. Ropes had gathered experience from every quarter ; he had
not only read pretty much everything worth reading on his subject, he
had not only delved with endless patience in the original documents, but
he had obtained through social intercourse with soldiers now passed away
a truly enormous fund of information, a great part of which has surely
perished with him. I remember that during the last two or three years
tlie thoufjht sometimes occurred to him that he might not live to finish
his book. He told me one day that he only lacked eight years of being
three score and ten, and that eight years were all too short a period for
finishing the two volumes that remained to be done ; he must therefore
''scorn delight and live laborious daj's." He was always extremely fond
of society ; no man more keenly enjoyed a dinner-party or an evening at
tlie club, and I can testify that sometimes after club hours were over we
used to enjoy prolonging our friendly chat (juite into the morning hours ;

JOHN CODMAN ROPES. 633

but in these latter days Ropes became much more chary of his time and
subjected himself to a kind of discipline in order that his work might be
finishecU

In another direction and in dealing with a more limited theme, he
achieved a finished piece of work. He had always entertained a warm
admiration for the First Napoleon. It was natural that such an acute
military critic should admire such transcendent military genius. But
Ropes carried his admiration to an extent with which not all his friends
found it easy to sympathize. In his little book entitled " The First
Napoleon" Ropes appears as the great Corsican's advocate, and his case
is presented with consummate skill. It has all the more weight because
the author is far too skilful to weaken his case by over statement or by
any too conspicuous warmth of enthusiasm. It is a masterly piece of
writing, although in its philosophic grasp of the man and the period it is
surely far inferior to the book published about the same time by the late
Sir John Seeley.

It was in relation to the Waterloo Campaign that Ropes produced the
completely finished work already alluded to. No battle of the nine-
teenth century has called for so much discussion as Waterloo ; and most
of the discussion has centred about the question, " Why did Napoleon
lose the battle ? " The books on this subject are legion, and they present
us with an English view of the situation and a Prussian view, and ever so
many French views, according to the political and personal predilections
of the writers. Usually we find some particular antecedent selected as
explaining the mighty result, while other antecedents receive inadequate
attention or are passed over. One writer is impressed with the ineffi-
ciency of Grouchy, another one traces the catastrophe to the aimless
wanderings of Erlon's corps on the sixteenth of June, and so on. But in
Ropes' monograph what chiefly impresses us is the fact that he weighs
every circumstance with the greatest care and puts real mental effort
into the work of estimating the precise share which each circumstance
took in the general mass of causation. In the first place the quality of
the French army is duly considered and compared with the quality of the
allied forces. Then such facts as the Emperor having Soult for Chief of
Staff, an unaccustomed position for that able marshal, his feeling it nec-
essary to leave at Paris the invincible Davoust, and other like circum-
stances, receive due attention. The mysterious movements of Erlon,
which prevented his being of any use either to Ney at Quatre Bras or to
Napoleon at Ligny, are more acutely analyzed than in any other book.
Then the consequences of the very incomplete defeat of Blucher on the

634 JOHN LOWELL.

sixteenth are carefully considered. Then Napoleon's great and unusual
blunder in assuming an eastward retreat for the Prussians and acting
upon the assumption without verifying it, is properly characterized. Tlie
share wrought by the muddy roads and the rains is not forgotten, nor the
physical weaknesses which hampered the great general and allowed him
uow and then to be caught napping for a moment ; the masterly position
taken by Wellington ; the effects of the topography ; the extent to which
the Emperor's attention was diverted early in the afternoon in the direc-
tion of Planchenoit, — not one of these points is forgotten or slurred over.
It is this minute quantitative consideration of details that impresses upon
Ropes' historical writings their truly scientific character, and no theme
could have been better calculated to exhibit it in its perfection than the
campaign of Waterloo. One cannot read the book carefully without
feeling that for once in the world something has been done so exhaust-
ively that it will not need to be done again. It would seem almost
impossible for the most fertile mind to offer a suggestion of anything
actual, probable, or possible about W^aterloo that our author has not
already brought forward and considered. Those who write such books
are few, and to study them is a great and profitable stimulus. As this
monograph on Waterloo related to a subject already well understood in
Europe, it immediately gave Ropes a high reputation in European circles,
and I believe he is regarded by experts as one of the soundest military

critics since the days of Jomini.

John Fiske.

JOHN LOWELL

John Lowell, the fourth of that name in direct descent from the first
minister of Newbury port, who died in 1767, was born in Boston on the
18th of October, 1824. He was elected, in October, 1877, a Fellow of
this Academy, of which his great-grandfather had been one of the original
incorporators in 1780, and both his father and grandfather Fellows.

At the time of his birth his father was living on the lower (southerly)
corner of what are now Bedford and Chauncy streets ; but at that time,
between Bedford and Summer streets, at the points where Cliauncy
Street now turns off, there was on Summer Street a place called Chauncy
Place, running about two-thirds of the way through, and then closed
by a brick wall with two openings for foot passengers, but none for
vehicles, and turning up from Bedford Street a similar place called
Bedford Place, on the upper side of which and next to the wall stood
the house of Judge Charles Jackson, having a large garden and pear

JOHN LOWELL. 635

orchard extending down the phxce until it touched the garden of Judge
Prescott, which stretched from there up Bedford Street towards Wash-
ington. The lower side of Bedford Place was occupied by a row of
brick houses, and in the one at the corner, as has already been said,
Joiui Lowell was born.

In Chauncy Place, next to the dividing wall between that and Bedford
Place, stood the Chauncy Hall School, then under the management of a
well-known teacher, Gideon F. Thayer. To this school, partly perhaps
from its nearness as well as for its reputation, Judge Lowell was sent,
as soon as he was old enough to go to any man's school. He was fitted
for college in the private school of Daniel G. lugraham, who kept for
more than twenty years the leading private classical school in Boston,
and graduated at Harvard in 1843 at the age of eighteen years, with
high distinction in Greek, Latin, Philosophy, and History, — indeed it
would seem that he must have been recognized as especially proficient
in Greek, as at the Sophomore Exhibition in October, 1841, he was
one of the speakers in A Greek Dialogue, " Extract from Shakspeare's
Henry the Fourth : Glendower and Hotspur ; " and at the corresponding
Exhibition in his Junior year, had a Greek oration, " Bvpwv iv 'EXXdSi."
He was anions the first scholars in his class : of his Commencement
part, an English oration, the Rev. Dr. John Pierce wrote in his diary :
" The subject of this oration was ' The Battle of the Nile,' — eight
minutes in length, a summary of the principal facts." *

After leaving college Lowell entered the Dane Law School, and
remained there for the full course of two years. He was then for a
year in the office of Mr. Charles G. Loring, and was admitted to the
bar in 1846. Directly afterwards he went to Europe with his father
and family, and was absent about a year. On his return he formed a
connection with his brother-in-law, Mr. William Sohier, and began the
practice of the law. This connection lasted until 1857, when he took
an office by himself.

The business in Mr. Sohier's office was mostly chamber practice, and
largely connected with the management of trust estates, and Mr.
Lowell's work there was not such as to give him the opportunity for any
especial display of ability, or to attract any one's attention. His name,
I think, appears only twice in the Massachusetts Reports for this period.
In the autumn of 1857 he separated from Mr. Sohier almost at the
moment when the financial difficulties of that time were reaching their

* Mass. Hist. Soc. Proceedings, Series II., V. p. 237.

636 JOHN LOWELL.

crisis. These difficulties affected very seriously the manufacturing
interests of this part of the Comnaonwealth, and many of the large
mercantile firms who had been the agents of the factories. The family
and business connections of Mr. Lowell and his father naturally caused
him to be consulted in some of these matters, and he thus became
interested in the consideration of questions arising in bankruptcy and
insolvency, a branch of the law in which as a judge he subsequently
acquired a well-deserved and widespread reputation.

He continued in the practice of the law until 1865. During part of
this time he had charge of the " Law Reporter," a monthly journal then
published in Boston. From May, 1856, he was for two years its sole
editor, and afterwards until April, 1860, joint editor with Mr. Samuel
M. Quincy of the Boston bar. Four volumes (Vols. 19, 20, 21, and 22)
were published while he was editor.

Upon the resignation of Judge Sprague of the United States District
Court of Massachusetts, in March, 1865, Mr. Lowell was appointed his
successor. The appointment was wholly unsolicited by Mr. Lowell.
It was made upon the recommendation of a few of the leaders of the
Suffolk bar. The letter which they addressed to the President was
written by Mr. Charles G. Loring, and was signed by himself, Mr.
Charles B. Goodrich, Mr. Sidney Bartlett, Judge Josiah G. Abbott,
Mr. Samuel E. Sewall, Mr. Edward D. Sohier, Mr. George Bemis, and
Mr. Dwight Foster. It was carried to Washington at the same time
with Judge Sprague's resignation by Mr. Richard H. Dana, Jr., who
was then the United States attorney here. The nomination was sug-
gested to Mr. Lincoln by Mr. Sumner and Mr. Dana, and was made
and confirmed on the same day.

The appointment was one peculiarly gratifying to Mr. Lowell, for his
great-grandfather, Judge John Lowell, had been appointed in 1789 by
General Washington to this same office, and was the first District Judge
of the United States for the District of Massachusetts, and Mr. Lowell
felt a natural and proper pride and pleasure in succeeding to the honor-
able position and duties of his ancestor. In 1878, on the death of the
Honorable George F. Shepley, Judge of the Circuit Court of the United
States for the First Circuit, he was made Circuit Judge, thus again suc-
ceedino- his sreat-grandfather, who had been appointed by President John
Adams to a similar position in the court which the Federalists created in
the last year of President Adams's administration, only to have the act
creating it repealed and the court destroyed by the Democrats under
Jefferson, as the simplest mode of getting rid of the Federalist judges
whom Adams had appointed.

JOHN LOWELL. 637

la May, 1884, a little more than nineteen years after his first appoint-
ment to the bench, Judge Lowell resigned his position as Circuit Judge,
and resumed the practice of the law in Boston. His long period of ju-
dicial service had given the public ample opi)ortuuity to recognize his
legal ability and wisdom, and from the time he returned to the bar until
his last short illness, he was fully occupied with professional work. He
served as arbitrator, or sat as a commissioner or special master in many
causes and matters of importance ; he was retained as counsel in litiga-
tions involving large interests and difficult questions, and his advice was
constantly sought in solving the perplexities that arise in the administra-
tion of the law of bankruptcy. Fortunately for him this professional
work was interesting and stimulating and sufficiently engrossing to dis-
tract his mind from domestic griefs and anxieties, which weighed upon
him heavily at intervals.

During these years he two or three times visited Europe in the summer
with his family, before a severe accident to his wife rendered it difficult
for her to travel.

He married, in 1853, Lucy Buckminster, the only daughter of Mr.
George B. Emerson and of Olivia Buckminster, his wife, daughter of
the Rev. Joseph Buckminster of Portsmouth, N. H., and sister of the
Rev. Joseph Stevens Buckminster of Brattle Square Church, Boston.
For some time after his marriage Judge Lowell lived at No. 11 Chest-
nut Street, in the house in which his wife was born, but in 1858 he
purchased a farm lying between Chestnut Hill and Hammond's Pond,
where he lived until his death on the 14th of May, 1897. He had
inherited a taste for country life and an interest in trees and plants,
shrubs and flowers, from which he derived during these years great enjoy-
ment ; and the time which he passed in going about his place, in looking
at the improvements he had made, and considering those which he pro-
posed to make, undoubtedly contributed to the preservation of his health
and the prolongation of his life. He was not a man of remarkable
physique, yet until his last illness it is doubtful whether he was ever
detained from his office for two days at a time by any physical ailment.

In addition to the labors of his profession, he filled during these latter
years many positions of importance and responsibility. In 1896 he was
appointed by Governor Wolcott chairman of the commission to revise
the laws of Massachusetts on the subject of taxation, and held this office
at the time of his death. He was also Vice-President of the Massachu-
setts General Hospital, of which he had been for many years one of the
Trustees ; was President of the Trustees of the Peabody Fund, an Overseer

G38 JOHN LOWELL.

of Harvard University, and a Member of the Massachusetts Historical
Society. He received the degree of Doctor of Laws from Williams Col-
lege in 1870, and from Harvard in 1871.

Before his appointment as District Judge, Mr. Lowell's work at the
bar had been chiefly office practice. His clients, if not very numerous,
were warmly attached to him, were continually demanding his counsel,
uud had the most implicit confidence in the opinions he gave them. When
consulted by any one of them he rarely looked at a book ; but after hear-
ing what the client had to say, would sit perfectly quiet, until he had
sufficiently considered the question, and then advise him in a way show-
ing intuitive sagacity and familiarity both with the fundamental principles
of the law and their a^jplication by the courts. He was always a diligent
student, reading carefully, digesting slowly, and assimilating thoroughly
what was new or seemed to him important in the volumes of reports that
from time to time appeared, so that when his judicial work began he was
well equipped for the duties that devolved upon him.

The District Court of the United States, of which he was at first ap-
pointed Judge, is a court of limited and peculiar jurisdiction ; it has the
exclusive cognizance of admiralty and maritime cases, and of all crimes
committed upon the high seas ; it is also a prize court, and sits as a
criminal court, for the trial of offences against the laws of the United
States, such as violations of the post-office laws, revenue laws, etc. The
number of jury trials in this court is, however, very limited and mostly
confined to criminal cases, usually of great simplicity. Judge Lowell's
want of experience in practice before a jury, and in dealing with juries,
was for this reason a source of much less embarrassment to him in this
court than it would have been in most others. He had, in fact, before
his appointment tried only one jury case, and was perfectly aware of the
difficulties under which he labored in public speaking. Time and con-
stant practice diminished these difficulties, and made it more easy for him
to preside at jury trials, as he gradually overcame the obstacles which
sprang largely from his want of training and experience, and a lack of
confidence in himself The civil war was not over when he was appointed
to the bench, and the effect of the destruction of our merchant marine by
the guns of the Confederate cruisers, and of the transfer of our ships to
foreign owners, now so evident in many other ways besides that of the
sadly diminished number and importance of the suits in Admiralty, had
not then made themselves thoroughly felt. At the time of his appoint-
ment there were many causes pending which Judge Sprague's long illness
had prevented his hearing, and Judge Lowell had at first no laclj of

JOHN LOWELL. G39

Admiralty suits of all kinds, iucluding some difficult questions ia prize
causes. The principles and practice of the Admiralty Courts were espe-
cially congenial to him. He was in full sympathy with the tendei'ness
which that court has always shown for the ignorant and improvident
sailor ; and the cardinal rule of the Admiralty that no error of statement,
or technical defect, or mistake in the kind of relief asked for, should inter-
fere with or prevent such a decision of a cause as would work substantial
justice between the parties, was in entire harmony with his view as to
what should be the aim of all courts and judges so far as the established
rules of law and the decided cases would permit, — an opinion which he
held so strongly, that for his efforts to attain this end, he was at times
called " wayward " in his decisions.

At the time of Judge Lowell's appointment, there was no national
bankrupt act in force, although there was an insolvent act in Massachu-
setts with which and its operation he was very familiar ; two years after
he became the judge of the District Court, Congress passed the National
Bankrupt Act of 1867, in the administration of which he won a distinc-
tion which placed him among the foremost, — it is perhaps not too much
to say, at the head of the judges in this country, having original juris-
diction of cases in Bankruptcy and charged with the administration of
this law.

As has already been said, the subject of Bankruptcy had interested
him many years before his appointment to the bench ; he had begun the
preparation of the treatise on this branch of the law which since his death
has been published by his son, and the work he had done on this book
had made him unusually familiar with the underlying principles of the
Bankrupt Laws and the adjudicated cases. The purposes of a bankrupt
act, — to secure the equitable distribution of an insolvent debtor's assets
among his creditors, to grant him his discharge from his antecedent lia-
bilities if he has honestly surrendered his property to be distributed for
the payment of his debts, — and, on the other hand, to prevent such a
debtor, who has fraudulently concealed for his own advantage a part
of his estate, or unfairly preferred some of his creditors, from receiving
the benefit of the act, — these purposes equitable in both aspects, and the
simplicity and directness of the methods of procedure under the practice
in Bankruptcy, were all in harmony with his convictions of the useful
and beneficent purposes of the law and the courts of justice, and made
the administration of this law a congenial task, which he discharged not
merely to the satisfaction of the bar, but to the approval of the mercantile
class of the community whose interests are most affected by this law, who

640 JOHN LOWELL.

are perfectly able to understand it and to judge whether it is wisely ad-
ministered, and in the spirit of justice and fair dealing upon which it is
founded. Upon Judge Lowell's retirement from the bench in May, 1884,
the merchants of Boston invited him to a public dinner, that they might
have an opportunity of testifying their high esteem and affectionate con-
sideration for him as an administrator of the laws which most nearly con-
cerned them in their business relations, whose decisions and interpretations
of these laws had received the approval of the mercantile community
throughout the land. This unusual tribute, coming from the source from
which it emanated, was most gratifying to Judge Lowell as a proof that
his administration of the Bankrupt Law had not only been in accord
with the strong common sense of the business men of the community, but
also with the principles of justice and equity which it has been the aim
of every well-intended law of Bankruptcy to carry out. The mercantile
community still further showed its appreciation of him by requesting him
at a later date to prepare a new Bankrupt Act to be laid before Congress.
But this act was unfortunately not passed.

The jurisdiction of the Circuit Court of the United States for the First
Circuit, of which Judge Lowell was appointed Judge in 1878, includes
the States of Maine, New Hampshire, and Rhode Island, as well as
Massachusetts, and his duties as Circuit Judge brought him in contact
with the members of the bar and the people of all these States ; that he
discharged these duties to their satisfaction was shown by the attendance
of lawyers from every State in the circuit, at the bar meeting in Boston
after his death, and by the abundant testimony they gave of their recog-
nition and thorough appreciation of his worth and ability, both as a mag-
istrate and a man. The business of the Circuit Court and the class of
cases tried there is quite different fi'om that of a District Court. It has
no original jurisdiction in Bankruptcy or Admiralty, but in these matters
sits only as a Court of Appeals. It deals more largely with cases at
common law and with civil suits for violation of the revenue law. It has
also jurisdiction of patent causes, a branch of the law which is thought
to require some special aptitude for and knowledge of mechanics as well
as of the physical sciences. Neither Judge Lowell's professional work
at the bar nor on the bench had ever brought him to the study of this
branch of the law, and he was at the outset and continued to be unduly
distrustful of his ability to master and properly decide the patent cases
that were brought before him. As he occasionally jocosely expressed it,
" he was afraid of only one thing in the law, — those infernal machines."
But he applied himself with the conscientiousness and thoroughness that

JOHN LOWELL. 641

distinguished all his work to the study and investigation of these cases,
and by the common consent of the counsel practising in patent cases,
proved himself an excellent patent judge.

His judicial opinions are marked by clearness of thought, and the
direct, terse, and vigorous expression of it, with no digressions, no dis-
cussions or dicta not absolutely necessary to the determination of the
precise points before him. lu deciding a cause he took the line of least
resistance ; if it were a short cut he availed himself of it, and this not
so much because this course saved him labor, as because he intended his
judgment to be confined to the determination of the case actually before
him. He was careful, therefore, not to stray from the real questions at
issue, and to avoid being beguiled into by-paths of interesting legal
investigation which had but little or no tendency to enlighten or aid him
in arriving at his conclusions, and were not important in determining
his final results.

He had a natural aptitude for the law, one might almost say a legal
instinct, and this was strengthened and quickened by thought and study ;
he possessed as a judge a remarkable intuition " for perceiving on which
side lay the justice of any case and an equally remarkable ingenuity in
showinsf that a decision in favor of that side was in accordance with
the settled rules of law." When he found himself obliged to yield to
precedents which prevented what justice seemed to him to require, he
did so with a reluctance which he never tried to conceal and sometimes
openly expressed. His quiet humor not infrequently enlivened the trial
of a cause, and occasionally appeared in his Opinions, as when in denying
a motion to set aside a verdict because one juror had been asleep during
part of the trial, he said : " If one of the jurors was asleep, the defend-
ant should have called attention to the fact at the time. There is no
suggestion that it is newly discovered, and I cannot now say that the
defendant may not have thought his interests were promoted by the
actual course of the trial in this respect." Or, to quote one more
instance, — in a suit where a ship owner contended that the master had
forfeited his wages by taking on board some casks of Madeira wine, the
ship articles prohibiting him from carrying distilled spirits under penalty
of forfeiting his pay, Judge Lowell demolished the elaborate argument
of the counsel for the ship owner by saying, " Wine is not distilled
spirits, and cannot be made so by a usage of the port of New Bedford or
any other process that I am acquainted with, except distillation."

Any attempt to characterize Judge Lowell's judicial qualities would
be imperfect if it failed to recognize and call attention to the fact that

VOL. XXXV. — 41

642 JOHN LOWELL,

he had not merely the respect, but the warm affection of the members of
the bar. His simple and unassuming manner claimed nothing, yet no
one ever thought of treating him with discourtesy. His natural kindness
led him to treat the youngest members of the bar and to listen to their
arguments, however crudely presented, with a never-failing patience and
consideration ; and both on and off the bench, he was always, in his
relations with them, cordial, friendly, and ready to give them the
benefit of his wisdom and learning in any difficulty in which they might
ask his advice, doing this in such a manner that he rather seemed to be
receiving than conferring a favor. He was not merely a learned
lawyer ; he was also a lover of books, a great reader, catholic in his
tastes, and spending in his library many of his hours of rest and
recreation. In private life he was an agreeable companion, a loyal
friend, a most devoted husband and father. He was not a demonstra-
tive person, but " a quiet, self-contained, thoughtful, courageous,
affectionate man, who kept his eye single for the right and the true, who
did every duty with keen thoroughness, and who found his greatest
pleasure iu the companionship of those nearest and dearest to him."

Thornton K. Lothrop.

Other notices are postponed.

The number of new members elected during the year is as
follows : Resident Fellows, 4 ; Associate Fellows, 5 ; Foreign
Honorary Members, 4.

The roll of the Academy at present includes 194 Resident
Fellows, 89 Associate Fellows, and GQ Foreign Honorary
Members.*

* By transfer of an Associate Fellow to Resident Fellowship, and by election
of new members at the annual meeting, the roll stands at date of publication 198
Resident Fellows, 93 Associate Fellows, and 69 Foreign Honorary Members.

i

American Academy of Arts and Sciences.

OFFICERS AND COMMITTEES FOR 1899-1900.

Class I.
John Trowbridge.

president.

Alexander Agassiz.

VICE-PRESIDENTS.
Class II.

Alpheus Hyatt.

Class III.

Augustus Lowell.

CORRESPONDING SECRETARY.

Samuel H. Scudder, William M. Davis,

May, 1899 — January, 1900. January — May, 190O.

RECORDING SECRETARY.

William Watson.

TREASURER.

Francis Blake.

librarian.
A. Lawrence Rotch.

COUNCILLORS.
Class I. Class II.

Henry Taber, Benjamin L. Robinson,

Theodore W. Richards, William T. Councilman,
Harry M. Goodwin. John E. Wolff.

Alexander Agassiz,

Erasmus D. Leavitt,
Amos E. Dolbear,

COMMITTEE OF FINANCE.

Francis Blake,

Class III.
Barrett Wendell,
Edward Robinson,
James B. Ames.

Augustus Lowell.

RUMFORD COMMITTEE.
Edward C. Pickering, Charles R. Cross,
Arthur G. Webster, Theodore W. Richards,
Thomas C. Mendenhall.

Francis H. Storer,
Henry B. Hill,

c. m. warren committee.
Charles L. Jackson,
Leonard P. Kinnicutt,
Robert H. Richards.

Samuel Cabot,
Arthur M. Comey,

committee of publication.
Samuel H. Scudder, Seth C. Chandler,

committee on the library.
A. Lawrence Rotch, Henry W. Haynes,

Crawford H. Toy.

Samuel Henshaw.

auditing committee.
Henry G. Denny, William L. Richardson.

LIST

OF THE

EELLOWS AND EOEEIGN HONORAEY MEMBERS.

(Corrected to May 20, 1900.)

RESIDENT FELLOWS. — 198.

(Number limited to two liundred. )
Class I. — Mathematical and Physical Sciences. — 79.

Section L — 20.

Mntlip.matics and Astronomy.
Solon I. Bailey, Cambridge.

Maxime Bocher, Cambridge.

William E. Byerly, Cambridge.
Seth C. Chandler, Cambridge.
J. Rayner Edmands, Cambridge.
Gustavus Hay, Boston.

Henry Mitchell, Nantucket.

William F. Osgood, Cambridge.
James Mills Peirce, Cambridge.
Edward C. Pickering, Cambridge.
William H. Pickering, Cambridge.
John Ritchie, Jr., Boston.
John D. Rankle,
T. H. Safford,
Edwin F. Sawyer,
Arthur Searle,
William E. Story,
Henry Taber,
O. C. Wendell,
P. S. Yendell,

Cambridge.
Williamstown.
Brighton.
Cambridge.
Worcester.
Worcester.
Cambridge.
Dorchester.

Section H.— 21.
Physics.
A. Graham Bell, Washington, D.C.
Clarence J. Blake, Boston.
Francis Blake, Weston.

Charles R. Cross,
Amos E. Dolbear,
H. M. Goodwin,
Edwin H. Hall,

Brookline.
Somerville.
Boston.
Cambridge.

Hammond V. Hayes, Cambridge.
William L. Hooper, Somerville.
William W. Jacques, Newton.

Frank A. Laws,
Henry Lefavour,
T. C. Mendenhall,
Benjamin O. Peirce,
A. Lawrence Rotch,
Wallace C. Sabine,
John S. Stone,
Elihu Thomson,
John Trowbridge,
A. G. Webster,
Robert W. Willson,

Boston.
Williamstown.
Worcester.
Cambridge.
Boston.
Cambridge.
Boston.
Swampscott.
Cambridge.
Worcester.
Cambridge.

Section III. — 24.

Chemistry.

Samuel Cabot, Boston.

Arthur M. Comey, Cambridge.

James M. Crafts, Boston.
Thos. M. Drown, So. Bethlehem, Pa.

Charles W. Eliot, Cambridge.

Thomas Gaffield, Boston.

Henry B. Hill, Cambridge.

RESIDENT FELLOWS.

645

Charles L. Jackson, Cambridge.
AValtcr L. Jeiniiiigs, AVorcester.
Leonard P. Kinnicutt, Worcester.
Charles F. INlabery, Cleveland, O.
Arthur Michael, Boston.

George D. Moore, Worcester.
Charles E. Munroe, Wash'gton, D.C.
John U. Nef, Chicago, lU.

Arthur A. Noyes, Boston.

Robert H. Richards, Boston.
Theodore W. Richards, Cambridge.
Charles R. Sanger, Cambridge.
Stephen P. Sharpies, Cambridge.
Francis H. Storer, Boston.
Henry P. Talbot, Newton.

Charles H. Wing, Ledger, N. C.
Edward S. Wood, Boston.

Section IV. — 14.

Technology and Engineering.
Eliot C. Clarke, Boston.

L-a N. Hollis, Cambridge.

Gaetano Lanza, Boston.

E. D. Leavitt, Cambridgeport.

AVilliam R. Livermore, Boston.
Hiram F. Mills, LowelL

Cecil H. Peabody, Boston.
Alfred P. Rockwell, INI au Chester.
Andrew H. Russell, Wash'ton, D.C.
Peter Schwamb, Arlington.

Charles S. Storrow, Boston.
George F. Swain, Boston.

AMlliam Watson, Boston.

Morrill Wyman, Cambridge.

Class II. — Natural and Physiological Sciences. — 64;

Sectiox I. — 13.

Geology, Mineralogy, and Physics of

the Globe.

H. H. Clayton, Milton.

Algernon Coolidge, Boston.

AVilliam O. Crosby, Boston.

William M. Davis, Cambridge.

Benj. K. Emerson, Amherst.

O. W. Hmitington, Newport, R. I.

Robert T. Jackson, Boston.

AVilliam H. Niles, Cambridge.

John E. Pillsbury, Boston.

Nathaniel S. Shaler, Cambridge.

Rol)ert DeC. Ward, Cambridge.

John E. Wolif, Cambridge.
J. B. Wood worth.

Cambridge.

Section II. — 11.
Botany.
Geo. E. Davenport, Medford.
William G. Farlow, Cambridge.
Charles E. Faxon, Boston.
Merritt L. Fernaid, Cambridge.
George L. Goodale, Cambridge.

H. H. Hunnewell, Wellesley.

John G. Jack, Boston.

B. L. Robinson, Cambridge.

Charles S. Sargent, Brookline.

Arthur B. Seymour, Cambridge.

Roland Thaxter, Cambridge.

Section HI.

.'o.

Zoology and Physiology.
Alexander Agassiz, Cambridge.
Robert Amory, Boston.

James M. Barnard, Milton.
Henry P. Bowditch, Boston.
William Brewster, Cambridge.
Louis Cabot, . Brookline.

William E. Castle, Cambridge.
Sanmel F. Clarke, Williamstown.
W. T. Councilman, Boston.
Charles B. Davenport, Chicago, II!
Harold C. Ernst, Boston.

J. Walter Fewkes, Washington,D.C.
Edward G. Gardiner, Boston.
Samuel Henshaw, Cambridge.
Alpheus Hyatt, Cambridge.

64G

RESIDENT FELLOWS.

Cambridge.

Boston.

Salem.

Cambridge.

Boston.

Cambridge.

John S. Kingsley, Somerville.

Edward L. Mark,

Charles S. Minot,

Edward S. Morse,

George H. Parker,

James J. Putnam,

Samuel H. Scudder,

William T. Sedgwick, Boston.

James C. White, Boston.

William M. Woodworth, Cambridge.

Section IV. — 15.
Medicine and Surgery.
Samuel L. Abbot, Boston.

Edward H. Bradford,
Arthur T. Cabot,
David W. Cheever,
Frank W. Draper,
Thomas Dwight,
Reginald H. Fitz,
Charles F. Folsom,
Frederick I. Knight,
Samuel J. Mixter,
W. L. Richardson,
Theobald Smith,
O. F. Wadsworth,
Henry P. Walcott,
John C. Warren,

Boston.

Boston.

Boston.

Boston.

Boston.

Boston.

Boston.

Boston.

Boston.

Boston.

Boston.

Boston.

Cambridge.

Boston.

Class III. — Moral and Political Sciences. — 55.

Section I. — 11.

Philosophy and Jtirisprudence .

James B. Ames, Cambridge.

Charles C. Everett, Cambridge.

Horace Gray, Boston.

John C. Gray, Boston.

G. Stanley Hall, Worcester.

Nathaniel Holmes, Cambridge.

John E. Hudson, Boston.

Francis C. Lowell, Boston.

Josiah Royce, Cambridge.

Jeremiah Smith, Cambridge.

James B. Thayer, Cambridge.

Section H. — 20.

Philology and Archceology.

William S. Appleton, Boston.
Charles P. Bowditch, Boston.

Lucien Carr,

Franklin Carter,

Joseph T. Clarke,

Henry G. Denny,

AVilliam Everett,

William W. Goodwin, Cambridge.

Henry W. Haynes, Boston.

Cambridge.
Williamstown.
Boston.
Boston.
Quincy.

Charles R. Lanman,
David G. Lyon,
Bennett H. Nash,
Frederick W. Putnam,
Edward Robinson,
F. B. Stephenson,
Joseph H. Thayer,
Crawford H. Toy,
John W. White,
John H. Wright,
Edward J. Youni;,

Cambridge.

Cambridge.

Boston.

Cambridge.

Boston.

Boston.

Cambridge.

Cambridge.

Cambridge.

Cambridge.

Waltham.

Section HI. — 11.

Political Economy

Charles F. Adams,
Edward Atkinson,
Andrew M. Davis,
John Fiske,
A. C. Goodell,
Henry C. Lodge,
A. Lawrence Lowell
Augustus Lowell,
.Tames F. Rhodes,
Denman W. Ross,
Charles C. Smith,

and History.

Lincoln.
Boston.
Cambridge.
Cambridge.
Salem.
Nahant.
, Boston.
Boston.
Boston.
Cambridge.
Boston.

RESIDENT FELLOWS.

G47

Section IV. — 13.

Literature and the Fine Arts.

Francis Bartlett, Boston.

John Bartlett, Cambridge.

Arlo Bates, Boston.

George S. Boutwell, Groton.

J. Elliot Cabot, Brookline.

T. W. Higginson,

Cambridge.

George L. Kittredge,

Cambridge.

S. R. Koehler,

Boston.

Charles G. Loring,

Boston.

Percival Lowell,

Boston.

Charles Eliot Norton,

Cambridge.

Horace E. Scudder,

Cambridge.

Barrett Wendell,

Boston.

648

ASSOCIATE FELLOWS.

ASSOCIATE FELLOWS. — 93.

(Number limited to one hundred. Elected as vacancies occur.)
Class L — Mathematical and Physical Sciences. — 34.

Section I. — 14.

Mathematics and Astronomy.

Edward E. Barnard, Williams Bay,
S. W. Burnham, Chicago. [Wis.
George Davidson, San Francisco.
Fabian Franklin, Baltimore.
Asaph Hall, Cambridge, Mass.

George AV. Hill, W. Nyack, N.Y.
E. S. Holden, Washington.

James E. Keeler, Mt. Hamilton, Cal.
Emory McClintock, Mon'istown,N.J.
Simon Newcomb, Washington.
Charles L. Poor,
George M. Searle,
J. N. Stockwell,
Chas. A. Young,

Baltimore.
Washington.
Cleveland, O.
Princeton, N. J.

Section II. — 6.
Physics.

Carl Barus, Providence, R.I.

J. Willard Gibbs, New Haven.
S. P. Langley, Washington.

A. A. Michelson, Chicago.
Ogden N. Rood, New York.
H. A. Rowland, Baltimore.

Section III. —7.

Chemistry.

Wolcott Gibbs, Newport, R.I.
Frank A. Gooch, New Haven.
S. W. Johnson, New Haven.
J. W. Mallet,
E. W. Morley,
J. M. Ordway,
Ira Remsen,

Ch arlottesville , Va,
Cleveland, O.
New Orleans.
Baltimore.

Section IV. — 7.

Technology and Engineering.

Henry L. Abbot, New York.
Cyrus B. Comstock, New York.
W. P. Craighill, Charlestown, W.
F. R. Button, New York.[Va.

George S. Morison, Chicago.
William Sellers, Edgemoor, Del.
Robt. S. Woodward, New York.

Class II. — Natural and Physiological Sciences. — ■ 30.

Section I. — 14.

Geology, Mineralogy, and Physics of
the Globe.

Cleveland Abbe,
George J. Bru.sh,
Edward S. Dana,
Walter G. Davis,

Washington.
New Haven.
New Haven.
Cordova, Arg.

George M. Dawson, Ottawa.

G. K. Gilbert,
Clarence King,
Joseph LeConte,
J. Peter Lesley,
S. L. Penfield,
J. W. Powell,
R. Pumpelly,
A. R. C. Selwyn,
Charles D. Walcott.

Washington.
New York.
Berkeley, Cal.
Milton, Mass.
New Haven.
AVashington.
Newport, R.I.
Ottawa.
Washington.

ASSOCIATE FELLOWS.

649

Section II. — 5.
Botany.

L. H. Bailey,
D. II. Campbell,
J. M. Coulter,
John D. Smith,
W. Trelease,

Ithaca.

Palo Alto, Cal.

Chicago.

Baltimore.

St. Louis.

Section III. — 5.
Zoology and Physiology.
Joel A. Allen, New York.

W. K. Brooks, Lake Roland, Md.

S. Weir Mitchell,
A. S. Packard,
A. E. Verrill,

Philadelphia.
Providence, R.I.
New Haven.

Section IV. — 6.

Medicine and Surgery.

John S. Billings, New York.
Jacob M. Da Costa, Philadelphia.
William Osier, Baltimore.

Alfred Stille, Philadelphia.

Wm. H. Welch, Baltimore.

H. C. Wood, Philadelphia.

Class III. — Moral and Political Sciences. — 29.

Section I. — 7.

Philosophy and
James C. Carter,
Joseph H. Choate,
Melville W. Fuller,
William W. Howe,
William Mitchell,
Charles S. Peirce,
T. R. Pynchon,

Jurisprudence.
New York.
New York.
Washington.
New Orleans.
St. Paul.
Milford, Pa.
Hartford, Conn.

Section II. — 8.

Philology and Archceology.
Timothy Dwight, New Haven.
B. L. Gildersleeve, Baltimore.

D. C. Oilman, Baltimore.
T. R. Lounsbury, New Haven.
Rufus B. Richardson, Athens.

E. E. SaUsbury, New Haven.
Thomas D. Seymour, New Haven.
A. D. White, Ithaca, N.Y.

Section HI. — 6.

Political Economy and History.

Henry Adams, Washington.

G. P. Fisher, New Haven.

H. E. von Hoist, Chicago.

Henry C. Lea, Philadeljihia.

Henry M. Stevens, Ithaca.

W. G. Sumner, New Haven.

Section IV. — 8.

Literature and the Fine Arts.

James B. Angell, Ann Arbor, Mich.
L. P. di Cesnola, New York.
H. H. Furness, Wallingford, Pa.
R. S. Greenough, Florence.
Augustus St. Gaudens, New York.
John S. Sargent, London.
E. C. Stedman, Bronxville, N.Y.
W. R. Ware, New York.

650

FOREIGN HONORARY MEMBERS.

FOREIGN HONORARY MEMB ERS. — 69.

(Number limited to seventy-five. Elected as vacancies occur.)

Class I. — Mathematical and Physical Sciences. — 23.

Berlin.
Cambridge.

Sf:cTiON I. — 7.
Mathematics and Astronomy

Arthur Auwers,

George H. Darwin

H. A. E. A. Faye, Paris.

Charles Herniite, Paris.

Sir William Iluggins, London.

Otto Struve, Karlsruhe.

H. C. Vogel, Potsdam.

Section II. — 6.
Physics.
Ludwig Boltzmann, Vienna.
A. Cornu, Paris.

Oliver Heaviside, Newton Abbot.
F. Kohlrausch, Berlin,

Lord Rayleigh, Witham.

Sir G. G. Stokes, Bart., Cambridge.

Section III. — 6.

Chemistry.

Adolf Baeyer, Munich.

Marcellin Berthelot, Paris.
J. H. van't Hoff, Berlin.

D. Mendeleeff, St. Petersburg,

Sir H. E. Pvoscoe, London.

Julius Thomseu, Copenhagen.

Section IV. — 4.
Technology and Engineering.
Sir Benjamin Baker, London.

Lord Kelvin,
Maurice Levy,
William C. Unwin,

Largs.
Paris.
London.

Class U. — Natural and Physiological Sciences. — 22,,

Section I. — 6.

Geology, Mineralogy, and Physics of
the Globe.

Sir Archibald Geikie, London.

Albert Heim,
Sir John Murray,
A. E. Nordenskiold,
Henry C. Sorby,
Ileinrich Wild,

Zurich.

Edinburgh.

Stockholm.

Sheffield.

Zurich.

Section II. — 6.

Botany.

J. G. Agardh, Lund.

E. Bornet, Paris.

Sir Joseph D. Hooker, Sunningdale.

W. Pfeffer, Leipsic.

H. Graf zu Solms-

Laubach, Strassburg.

Eduard Strasburger, Bonn.

FOREIGN HONORARY MEMBERS.

651

Section III. —7.

Zoolofjii and Physiology.

Sir Michael Foster,
Carl Gegenbauer,
Ludimar Hermann,
A. von KiJUiker,
A. Kovalevsky, St,

H. de Lacaze-Duthiers,
Elias Metschnikoff,

Cambridge.

Heidelberg.

Konigsberg.

WUrzburg.

, Petersburg.

Paris.

Paris.

Section IV. — 4.

Medicine and Surgery.

W. Kuhne, Heidelberg.

Lord Lister, London.

F. V. Recklinghausen, Strassburg.
Rudolph Virchow, Berlin.

Class III. — Moral and Political Sciences. — 23.

Section I. — 5.
Philosophy and Jurisprudence.

Heinrich Brunner,

F. W. Maitland,

Sir Frederick Pollock,

Bart. ,
Baron Russell of Kil-

lowen.
Henry Sidgwick,

Berlin.
Cambridge.

London.

Tadworth.
Cambridge.

Section II. — 7.
Philology and Archceology.

Ingram Bywater, Oxford.

W. Dorpfeld, Athens.
Sir John Evans, Hemel Hempstead.

J. W. A. Kirchhoff, Berlin.

G. C. C. Maspero, Paris.

]\Iax Miiller, Oxford.

Karl Weiuhold, Berlin.

Section III. — 6.
Political Economy and History,

Due de Broglie,

Paris.

James Bryce,

London.

Herman Grimm,

Berlin.

Tlieodor Mommsen,

Berlin.

William Stubbs,

Oxford.

Sir G. 0. Trevelyan,

Bart.,

London.

Section IV

. — 5.

Literature and the

Fine Arts.

Georg Brandes,

Copenhagen.

F. Brunetiere,

Paris.

Jean Leon Gerome,

Paris.

Rudyard Kipling,

Rottingdean.

Leslie Stephen,

London.

STATUTES AND STANDING VOTES.

STATUTES.

Adopted May 30, 1854 : amended September 8, 1857, November 12, 1862,
May 24, 1864, November 9, 1870, May 27, 1873, January 26, 1876,
June 16, 1886, October 8, 1890, January 11 and May 10, 1893, April
11, il/rt?/ 9, and October 10, 1894, an(/ March 13, ^jortY 10, awcZ May
8, 1895.

CHAPTER I.

Of Fellows and Foreign Honorary Members.

1. The Academy consists of Felloivs and Foreign Honorary Mem-
bers. They are arranged in three Classes, according to the Arts
and Sciences in wliich they are severally proficient, viz. : Class I.
The Mathematical and Physical Sciences ; — Class II. The Nat-
ural and Physiological Sciences ; — Class III. The Moral and
Political Sciences. Each Class is divided into four Sections, viz. :
Class L, Section 1. Mathematics and Astronomy ; — Section 2.
Physics ; — Section 3. Chemistry ; — Section 4. Technology and
Engineering. Class II., Section 1. Geology, Mineralogy, and
Physics of the Globe ; — Section 2. Botany ; — Section 3. Zoology
and Physiology ; — Section 4. Medicine and Surgery. Class III.,
Section 1. Philosophy and Jurisprudence ; — Section 2. Philol-
ogy and Archaeology ; — Section 3. Political Economy and
History ; — Section 4. Literature and the Fine Arts.

2. Fellows, resident in the State of Massachusetts, only, may
vote at the meetings of the Academy.* Each Eesident Fellow
shall pay an admission fee of ten dollars and such animal assess-
ment, not exceeding ten dollars, as shall be voted by the Academy
at each Annual Meeting.

* The number of Eesident Fellows is limited by the Charter to 200.

654 STATUTES OF THE AMERICAN ACADEMY

3. Fellows residing out of the State of Massachusetts shall be
known and distinguished as Associate Fellows. They shall not
be liable to the payment of any fees or annual dues, but on remov-
ing within the State shall be admitted to the privileges,* and be
subject to the obligations, of Resident Fellows. The number of
Associate Fellows shall not exceed one hundred, of whom there
shall not be more than forty in either of the three Classes of the
Academy.

4. The number of Foreign Honorary Members shall not
exceed seventy-five ; and they shall be chosen from among per-
sons most eminent in foreign countries for their discoveries and
attainments in either of the three departments of knowledge
above enumerated. And there shall not be more than tldrty
Foreign Members in either of these departments.

CHAPTER II.

Of Officers.

1. There shall be a President, three Vice-Presidents, one for
each Class, a Corresponding Secretary, a Recording Secretary, a
Treasurer, and a Librarian, which officers shall be annually elected,
by ballot, at the Annual Meeting, on the second Wednesday in
May.

2. At the same time, and in the same manner, nine Councillors
shall be elected, three from each Class of the Academy, but the
same Fellows shall not be eligible on more than three successive
years. These nine Councillors, vrith the President, the three
Vice-Presidents, the two Secretaries, the Treasurer, and the
Librarian, shall constitute the Council. It shall be the duty of
this Council to exercise a discreet supervision over all nomina-
tions and elections. With the consent of the Fellow interested,
they shall have power to make transfers between the several
Sections of the same Class, reporting their action to the
Academy.

3. If any office shall become vacant during the year, the
vacancy shall be filled by a new election, and at the next stated
meeting, or at a meeting called for this purpose.

* Associate Fellows may attend, but canuot vote, at meetings of the Academy.
See Chapter I. 2.

OP ARTS AND SCIENCES. 655

CHAPTER III.

Of Nominations of Officeks.

1. At tlie stated meeting in IVLarch, the President shall appoint
from the next retiring Councillors a ^Nominating Committee of
three Fellows, one for each class.

2. It shall be the duty of this Nominating Committee to
prepare a list of candidates for the offices of President,
Vice-Presidents, Corresponding Secretary, Recording Secretary,
Treasurei", Librarian, Councillors, and the Standing Committees
which are chosen by ballot ; and to cause this list to be sent by
mail to all the Resident Fellows of the Academy not later than
four weeks before the Annual Meeting.

3. Independent nominations for any ofiice, signed by at least
five Resident Fellows and received by the Recording Secretary
not less than ten days before the Annual Meeting, shall be in-
serted in the call for the Annual Meeting, which shall then be
issued not later than one week before that meeting.

4. The Recording Secretary shall prepare for use, in voting
at the Annual Meeting, a ballot containing the names of all
persons nominated for ofiice under the conditions given above.

5. When an ofiice is to be filled at any other time than at the
Annual Meeting, the President shall appoint a Nominating Com-
mittee, in accordance with the provisions of Section 1, which shall
announce its nomination in the manner prescribed in Section 2
at least two weeks before the time of election. Independent
nominations, signed by at least five Resident Fellows and received
by the Recording Secretary not later than one Aveek before the
meeting for election, shall be inserted in the call for that
meeting.

CHAPTEE IV.

Of the President.

1. It shall be the duty of the President, and, in his absence,
of the senior Vice-President present, or next officer in order as
above enumerated, to preside at the meetings of the Academy ;
to summon extraordinary meetings, upon any urgent occasion ;
and to execute or see to the execution of the Statutes of the

656 STATUTES OF THE AlVIERICAN ACADEMY

Academy. Length of continuous membership in the Academy
shall determine the seniority of the Vice-Presidents,

2. The President, or, in his absence, the next officer as above
enumerated, is empowered to draw upon the Treasurer for such
sums of money as the Academy shall direct. Bills presented on
account of the Library, or the Publications of the Academy, must
be previously approved by the respective committees on these
departments.

3. The President, or, in his absence, the next officer as above
enumerated, shall nominate members to serve on the different
committees of the Academy which are not chosen by ballot.

4. Any deed or writing to which the common seal is to be
affixed shall be signed and sealed by the President, when thereto
authorized by the Academy.

CHAPTER V.

Of Standing Committees.

1. At the Annual Meeting there shall be chosen the following
Standing Committees, to serve for the year ensuing, viz. : —

2. The Committee of Finance, to consist of the President,
Treasurer, and one Fellow chosen by ballot, who shall have
charge of the investment and management of the funds and trusts
of the Academy. The general appropriations for the expendi-
tures of the Academy shall be moved by this Committee at the
Annual Meeting, and all special appropriations from the general
and publication funds shall be referred to or proposed by this
Committee.

3. The Eumford Committee, of seven Fellows, to be chosen
by ballot, who shall consider and report on all applications and
claims for the Eumford Premium, also on all appropriations from
the income of the Eumford Fund, and generally see to the due
and proper execution of this trust.

4. The C. M. Warren Committee, of seven Fellows, to be
chosen by ballot, who shall consider and report on all applica-
tions for appropriations from the income of the C. M. Warren
Fund, and generally see to the due and proper execution of this
trust.

5. The Committee of Publication, of three Fellows, to whom
all memoirs submitted to the Academy shall be referred, and to

OP ARTS AND SCIENCES. 657

whom the printing of memoirs accepted for publication shall be
intrusted.

6. The Committee on the Library, of three Fellows, who
shall examine the Library, and make an annual report on its
condition and management.

7. An Auditing Committee, of two Fellows, for auditing the
accounts of the Treasurer.

CHAPTER VL

Of the Secretaries.

1. The Corresponding Secretary shall conduct the correspond-
ence of the Academy, recording or making an entry of all letters
written in its name, and preserving on file all letters which are
received ; and at each meeting he shall present the letters which
have been addressed to the Academy since the last meeting.
With the advice and consent of the President, he may effect
exchanges with other scientific associations, and also distribute
copies of the publications of the Academy among the Associate
Fellows and Foreign Honorary Members, as shall be deemed
expedient ; making a report of his proceedings at the Annual
Meeting. Under the direction of the Council for Nomination,
he shall keep a list of the Fellows, Associate Fellows, and Foreign
Honorary Members, arranged in their Classes and in Sections in
respect to the special sciences in which they are severally profi-
cient ; and he shall act as secretary to the Council.

2. The Eecording Secretary shall have charge of the Charter
and Statute-book, journals, and all literary papers belonging to
the Academy. He shall record the proceedings of the Academy
at its meetings ; and after each meeting is duly opened, he shall
read the record of the preceding meeting. He shall notify the
meetings of the Academy, and apprise committees of their ap-
pointment. He shall post up in the Hall a list of the persons
nominated for election into the Academy ; and when any indi-
vidual is chosen, he shall insert in the record the names of the
Fellows by whom he was nominated.

3. The two Secretaries, with the Chairman of the Committee
of Publication, shall have authority to publish such of the pro-
ceedings of the Academy as as may seem to them calculated to
promote the interests of science.

VOL. XXXV. — 42

658 STATUTES OF THE AMERICAN ACADEMY

CHAPTER VII.

Of the Teeasueee.

1. The Treasurer shall give such security for the trust reposed
in him as the Academy shall require.

2. He shall receive officially all moneys due or payable, and
all bequests or donations made to the Academy, and by order of
the President or presiding officer shall pay such sums as the
Academy may direct. He shall keep an account of all receipts
and expenditures; shall submit his accounts to the Auditing
Committee ; and shall report the same at the expiration of his
term of office.

3. The Treasurer shall keep a separate account of the income
and appropriation of the Eumford Fund, and report the same
annually.

4. All moneys which there shall not be present occasion to
expend shall be invested by the Treasurer, under the direction of
the Finance Committee, on such securities as the Academy shall
direct.

CHAPTER VIII.

Of THE LiBEAEIAN AND LiBKAEY,

1. It shall be the duty of the Librarian to take charge of the
books, to keep a correct catalogue of same, and to provide for
the delivery of books from the Library. He shall also have the
custody of the publications of the Academy.

2. The Librarian, in conjunction with the Committee on the
Library, shall have aiithority to expend, as they may deem ex-
pedient, such sums as may be appropriated, either from the Rum-
ford or the General Fund of the Academy, for the purchase of
books, and for defraying other necessary expenses connected with
the Library. They shall have authority to propose rules and
regulations concerning the circulation, return, and safe-keeping of
books ; and to appoint such agents for these purposes as they
may think necessary.

3. To all books in the Library procured from the income of
the Rumford Fund, the Librarian shall cause a stamp or label to
be affixed, expressing the fact that they were so procured.

OP ARTS AND SCIENCES. 659

4. Every person who takes a book from the Library shall give
a receipt for the same to the Librarian or his assistant.

5. Every book shall be returned in good order, regard being
had to the necessary wear of the book with good usage. And if
any book shall be lost or injured, the person to whom it stands
charged shall replace it by a new volume or set, if it belongs
to a set, or pay the current price of the volume or set to the
Librarian ; and thereupon the remainder of the set, if the volume
belonged to a set, shall be delivered to the person so paying for
the same.

6. All books shall be returned to the Library for examination
at least one week before the Annual Meeting.

CHAPTEE IX.

Of Meetings.

1. There shall be annually four stated meetings of the Acad-
emy ; namely, on the second Wednesday in May (the Annual
Meeting), on the second Wednesday in October, on the second
Wednesday in January, and on the second Wednesday in March.
At these meetings only, or at meetings adjourned from these
and regularly notified, shall appropriations of money be made,
or alterations of the statutes or standing votes of the Academy
be effected.

2. Fifteen Fellows shall constitute a quorum for the transac-
tion of business at a stated meeting. Seven Fellows shall be
sufficient to constitute a meeting for scientific communications
and discussions.

3. The Recording Secretary shall notify the meetings of the
Academy to each Fellow residing in Boston and the vicinity ; and
he may cause the meetings to be advertised, whenever he deems
such further notice to be needful.

660 STATUTES OF THE AMERICAN ACADEMY

CHAPTEE X.

Of the Election of Fellows and Honorary Members.

1. Elections shall be made by ballot, aud only at stated
meetings.

2. Candidates for election as Eesident Fellows must be pro-
posed by two or more Eesident Fellows, m a recommendation
signed by them, specifying the Section to which the nomination
is made, which recommendation shall be transmitted to the
Corresponding Secretary, and by him referred to the Council for
Nomination. No person recommended shall be reported by the
Council as a candidate for election, unless he shall have received
a written approval, signed at a meeting of the Council by at least
seven of its members. All nominations thus approved shall be
read to the Academy at a stated meeting, and shall then stand on
the nomination list during the interval between two stated meet-
ings, and until the balloting. No person shall be elected a Eesident
Fellow, unless he shall have been resident in this Commonwealth
one year next preceding his election. If any person elected a
Eesident Fellow shall neglect for one year to pay his admission
fee, his election shall be void ; and if any Eesident Fellow shall
neglect to pay his annual assessments for two years, provided
that his attention shall have been called to this article, he shall be
deemed to have abandoned his Fellowship ; but it shall be in the
power of the Treasurer, with the consent of the Council, to dis-
pense (sub silentio) with the payment both of the admission fee
and of the assessments, whenever in any special instance he shall
think it advisable so to do.

3. The nomination of Associate Fellows shall take place in
the manner prescribed in reference to Eesident Fellows; and
after such nomination shall have been publicly read at a stated
meeting previous to that when the balloting takes place, it shall be
referred to the Council for Nomination ; and a written approval,
authorized and signed at a meeting of said Council by at least
seven of its members, shall be requisite to entitle the candidate
to be balloted for. The Council may m like manner originate
nominations of Associate Fellows, which must be read at a stated
meeting previous to the election, and be exposed on the nomina-
tion list during the interval.

OP ARTS AND SCIENCES. 661

4. Foreign Honorary Members shall be chosen only after a
nomination made at a meeting of the Council, signed at the time
by at least seven of its members, and read at a stated meeting
previous to that on which the balloting takes place.

5. Three fourths of the ballots cast must be affirmative, and
the number of affirmative ballots must amount to eleven to effect
an election of Fellows or Foreign Honorary Members.

6. Each Section of the Academy is empowered to present lists
of persons deemed best qualified to fill vacancies occurring in the
number of Foreign Honorary Members or Associate Fellows
allotted to it ; and such lists, after being read at a stated meeting,
shall be referred to the Council for Nomination.

7. If, in the opinion of a majority of the entire Council, any
Fellow — Resident or Associate — shall have rendered himself
unworthy of a place in the Academy, the Council shall recommend
to the Academy the termination of his Fellowship ; and provided
that a majority of two thirds of the Fellows at a stated meeting,
consisting of not less than fifty Fellows, shall adopt this recom-
mendation, his name shall be stricken off the roll of Fellows.

CHAPTER XT.

Of Amendments of the Statutes.

1. All proposed alterations of the Statutes, or additions to
them, shall be referred to a committee, and, on their report at a
subsequent meeting, shall require for enactment a majority of
two thirds of the members present, and at least eighteen affirma-
tive votes.

2. Standing Votes may be passed, amended, or rescinded, at
any stated meeting, by a majority of two thirds of the members
present. They may be suspended by a unanimous vote.

CHAPTER XII.

Of Literaey Performances.

1. The Academy will not express its judgment on literary or
scientific memoirs or performances submitted to it, or included in
its publications.

662 STATUTES OF THE AMERICAN ACADEMY

STANDING VOTES.

1. Communications of which notice had been given to the
Secretary shall take precedence of those not so notified.

2. Kesident Fellows who have paid all fees and dues charge-
able to them are entitled to receive one copy of each volume or
article printed by the Academy, on application to the Librarian
personally or by written order, within two years from the date
of publication. And the current issues of the Proceedings shall
be supplied, when ready for publication, free of charge, to all the
Fellows and members of the Academy who desire to receive them.

3. The Committee of Publication shall fix from time to time
the price at which the publications of the Academy may be sold.
But members may be supplied at half this price with volumes
which they are not entitled to receive free, and which are needed
to complete their sets.

4. Two hundred extra copies of each paper accepted for publi-
cation in the Memoirs or Proceedings of the Academy shall be
placed at the disposal of the author, free of charge.

5. Resident Fellows may borrow and have out from the
Library six volumes at any one time, and may retain the same
for three months, and no longer.

6. Upon special application, and for adequate reasons assigned,
the Librarian may permit a larger number of volumes, not exceed-
ing twelve, to be drawn from the Library for a limited period.

7. Works published in numbers, when unbound, shall not be
taken from the Hall of the Academy, except by special leave of
the Librarian.

8. Books, publications, or apparatus shall be procured from the
income of the Eumford Fund only on the certificate of the Rum-
ford Committee that they, in their opinion, will best facilitate
and encourage the making of discoveries and improvements which
may merit the Rumford Premium.

9. The Annual Meeting and the other stated meetings shall be
holden at eight o'clock, P. M.

10. A meeting for receiving and discussing scientific commu-
nications may be held on the second Wednesday of each month
not appointed for stated meetings, excepting July, August, and
September.

OP ARTS AND SCIENCES.

663

RUMFORD PREMIUM.

In conformity with the terms of the gift of Benjamin, Count
Rumford, granting a certain fund to the American Academy of
Arts and Sciences, and with a decree of the Supreme Judicial
Court for carrying into effect the general charitable intent and
purpose of Count Rumford, as expressed in his letter of gift, the
Academy is empowered to make from the income of said fund, as
it now exists, at any Annual Meeting, an award of a gold and
silver medal, being together of the intrinsic value of three
hundred dollars, as a premium to the author of any important
discovery or useful improvement in light or in heat, which shall
have been made and published by printing, or in any way made
known to the public, in any part of the continent of America, or
any of the American islands ; preference being always given to
such discoveries as shall, in the opinion of the Academy, tend
most to promote the good of mankind ; and to add to such
medals, as a further premium for such discovery and improve-
ment, if the Academy see fit so to do, a sum of money not
exceeding three hundred dollars.

INDEX.

Acentetus, 44.
carinatus, 45.

Agassiz, A., Proposed Exploring Ex-
pedition to the Mid-Pacific, 610.

Amorphomyces obliqueseptata, 431.

Anthracomya arenacea, 4U0.

Arbutus glandulosa, 309.

Aster jalapensis, 572.

Athenaea Nelsonii, 567.

Atomic Weight of Cobalt, 59 ; of
Iron, 251.

Atomic Weights, Table of, 041.

Aulocara, 54.
elliotti, 56.
femoratura, 55.
rufum, 55.
parallelum, 57.

Baxter, G. P. See Richards, T. W.,
and Baxter, G, P.

Biographical Notices, List of, 623.
John Cummings, 628.
Epes Sargent Dixwell, 625.
John Lowell, 634.
John Codman Ropes, 629.

Bowers, Mary A., Peripheral Distri-
bution of the Cranial Nerves of
Spelerpes biliueatus, 612.

Bruneria, 49.

Cacopteris, 87.

aequalis, 92.

ephippiata, 91.

femorata, 93.

fuscopunctata, 89.

inermis, 89.

nevadensis, 91.

sinuata, 90.
Calamintha oaxacana, 564.
Cantharomyces Platystethi, 415.
Carboniferous Fauna of Narragansett

Basin, 397.
Castle, W. E., The Metamerism of the

Hirudiuea, 283-303.

Ceratomyces acuminatus, 447.

Californicus, 448.

cladophorus, 444.

denticulatus, 445.

elephantinus, 446.

Floridanus, 444.

ornithocephalus, 448.

reflexus, 447.

rhynchophorus, 446.
Cestrum Bourgeauianum, 572.

elegans, 571.

fulvescens, 572.
Chemical Laboratory of Harvard

College, Contributions from,

1, 59, 121, 211, 229, 251, 261,

375.
Chitonomyces aethiopicus, 430.

Floridanus, 430.
Chloealtis, 47.
Clematomyces, 439.

Pinophili, 440.
Clinocephalus, 47.
Cobalt, Atomic Weight of, 59.
Cobaltous Chloride, 59; Oxide, 59.
Cohoe, W. P. /See Jackson, C. L., and

Cohoe, W. P.
ColUns, E. ^ee Richards, T. W.,

Collins, E., and Heimrod, G. W.
Committee of Publication. Report of,

607.
Committees, List of, 643.
Compsomyces Lesteri, 439.
Continuous Groups, 95, 237, 481.
Copper, Electrochemical Equivalents

of, 121.
Coreopsis rhyacophila, 313.
Corethromyces Brazilianus, 432.

purpurascens, 433.
Correspondence, 601, 610, 911, 614,

615, 617, 618.
Council, Report of, 623.
Cross, C. R., Historical Notes relating

to Musical Pitch in the United

States, 451-467.
Cunila tomentosa, 565.

666

INDEX.

Cryptogamic Laboratory of Harvard
University, Contributions from,
151, 407.

Cuphea lleipublicae, 325.

Datura villosa, 571.
Davis, W. M., Tbe Freshwater Terti-
ary Formations of the Rocky

Mountain Region, 343-373.
Decticinae, 81.
Derby, I. H. See Jackson, C. L., and

Derby, I. H.
Derotmenia, 385.

cupidineum, 391.

deiicatuhim, 390.

haydeni, 392.

laticinctum, 389.

lentiginosum, 393.

lichenosum, 394.

rileyanum, 392.

saussureanum, 391.
Dichomyces Angolensis, 421.

biformis, 422.

Cafianus, 425.

dubius, 426.

exilis, 420.

hybridus, 423.

insignis, 421.

Javanus, 420.

Madagascarensis, 424.

Peruvianus, 426.

vulgatus, 424.
Dichromorpha, 46.
Dimorphomyces Myrmedoniae, 409.

nanomascuhis, 411.

pinnatus, 410.

Thleoporae, 410.
Dioscorea lobata, 323.

oaxacensis, 322.

pumicicoUi, 322.
Diparabrombenzylcyanamide, 229.
Dysodia Seleri, 314.

Fchelon Spectroscope, 109.

Electrochemical Equivalents of Cop-
per and Silver, 121.

Encelia Pringlei, 573.

Energy, Free, 1.

Eucaiitliaromyces Africanus, 418.
Callidae, 418.
Casnoniae, 417.
Catascopi, 419.
Diaphori, 416.

Eucantharomyces Euprocti, 417.

spinosus, 416.
Eucorethromyces, 433.

Apotomi, 433.
Eupatorium adenachaenium, 329.

adspersum, 329.

aegirophyllum, 329.

albicaule, 330.

amblyolepis, 330.

areolare, 330.

badium, 331.

Bertholdii, 331.

brick elloides, 331.

capnoresbium, 331.

cardiophyllum, 331.

chapalen.-^e, 332.

chiapense, 332.

crassirameum, 332.

crenacum, 333.

desquamans, 333.

glaucum, 333, 335.

heterolepis, 335.

Heydeanum, 335.

lanicaule, 336.

Loesenerii, 336.

micranthum, 334.

Nelsonii, 337.

Oerstedianum, 337.

oresbium, 337.

papacanum, 338.

phaenicolepis, 338.

photinum, 338.

platyphyllum, 339.

Saltivarii, 339.

scabfelUnn, 339.

scorodonioides, 340.

Selerianura, 340.

Smithii, 340.

Thieleanum, 341.

Valverdeanum, 341.

Vitalbae, 341.
Euzodiomyces, 449.

Lathrobii, 449.

Fellows, Associate, deceased, —

Alvan Wentworth Chapman, 601.

Sir John William Dawson, 615.

Manning Ferguson Force, 610.

William Alexander Hannnond,
615.

Edward John Phelps, 619.
Fellows, Associate, elected, —

Liberty Hyde Bailey, 618.

Joseph Hodges Choate, 616.

INDEX.

667

Fellows, Associate, elected, —

William Wirt Howe, 616.

William Mitchell, 616.

John Singer Sargent, 613.
Fellows, Associate, List of, 648.
Fellows, Resident, deceased, —

John Harrison Blake, 612.

Epes Sargent Dixwell, 614.

Silas Whitcomb Ilolman, 619.

Francis Minot, 610.

John Codman Ropes, 613.
Fellows, Resident, elected, —

Arlo Bates, 618.

Maxime Bocher, 613.

William Elwood Byerly, 609.

James Mason Crafts, 618.

Henry Lefavonr, 609.

William Fogg Osgood, 613.

William Henry Pickering, 609.
Fellows, Resident, List of, 644.
Fernald, M. L., A Synopsis of the

Mexican and Central American

Species of Salvia, 489-556; A

Revision of the Mexican and

Central American Solanums of

the Subsection Torvaria, 557-562 ;

Some Undescribed INlexican Phan-
erogams, chiefly Labiatae aud

Solanaceae, 562-573.
Ferrous Iodide, 211.
Finite Continuous Groups, 95, 237,481.
Foreign Honorary Members, de-
ceased, —

Robert Wilhelm Bunsen, 612.

James Martineau, 617.

Sir James Paget, 615.

Carl Friedrich Rammelsberg, 617.

John Ruskin, 617.
Foreign Honorary Members, elected, —

Sir Benjamin Baker, 613.

Rudyard Kipling, 613.

Friedrich Kohlrausch. 618.

Baron Russell of Killowen, 609.

Sir George Otto Trevelyan, 616.
Foreign Honorary Members, List of,

650.
Free Energy and Equilibrium, 1 .
Freshwater Tertiariesof Rocky Moun-
tains, 343.
Fuller, R. W. See Jackson, C. L.,

and Fuller, R. W.

Galium triflorum, 313.
Gazzolo, F.H See Jackson, C. L.,
and Gazzolo, F. H.

Gomphocerus, 49.

Gray Herbarium of Harvard Univer-
sity, Contributions from, 305, 4S7.

Greenman, J. M., New Species and
Varieties of Mexican Plants, 307
-315.

Hechtia sphaeroblasta, 323.
Heiunod, G. W. See Richards, T. W.,

Collins, E., and Heimrod, G. W.
Hirudinea, Metamerism of, 283.
Hooper, W. L., The Electrical Resist-
ance of the Human Body, 614.
Horesidotes, 48, 49.

cinereus, 49.
Hosackia oaxacana, 309.

repens, 308.
Hyptis axillaris, 565.

madrensis, 565.

Nelsonii, 566.

Pringlei, 565.

Infinitesimal Transformations, 575.

Iodide, Ferrous, 211.

Iron, Atomic Weight of, 251.

Jackson, C. L., and Cohoe, W. P.,

Certain Derivatives of Metadi-

brombenzole, 615.
Jackson, C. L., and Derby, I. H.,

Ferrous Iodide, 211-227.
Jackson, C. L., and Fuller, R. W.,

Note on the Constitution of Di-

parabrombenzylcyanamide, 229-

236.
Jackson, C. L., and Gazzolo, F. H.,

On Certain Colored Substances

derived from Nitro Compounds,

261-281.
Jaegeria, 315.

discoidea, 318.

gracilis. .319.

hirta, 317.

macrocephala, 317.

minioides, 317.

pedunculata, 317.

petiolaris, 316.

prorepens, 318.

purpurascens, 316.

Kingsley, J. S., The Ancestry of the
Mammalia, 61G.

668

INDEX

Laboulbenia acrogenis, 154.
adunca, 154.
Aerogenidii, 155.
AuapJogenii, 156.
Anchonoderi, 156.
angularis, 157.
auomala, 157.
a(iuatica, 158.
ai'istata, 158.
Asiatjca, 159.
Assameiisis, 159.
barbata, 160.
bicornis, 160.
bidentata, 161.
Bracliionychi, 162.
Cafii, 162.
celestialis, 163.
ceratophora, 163.
Ceyloiiensis, 164.
Chiiiqueusis, 164.
Clivinalis, 165.
coarctata, 165.
Colpodis, 166.
constricta, 167.
Copteae, 167.
corethropsis, 168.
corrugata, 168.
Cubeiisis, 169.
dactylophora, 169.
Darwinii, 170.
denticulata, 170.
Dercylii, 171.
Dineutis, 171.
distincta, 172.
drepanalis, 173.
Egae, 173.
equatorialis, 174.
erecta, 175.
falcata, 175.
fallax, 176.
finitima, 176.
fissa, 177.
forficulata, 178.
geniculata, 178.
gibbifera, 179.
heterocheila, 179.
imitans, 180.
insularis, 181.
intermedia, 181.
Italica, 182.
Javana, 182.
lencophaca, 183.
Loxandri, 183.
niaculata, 184.
Madagascarensis, 184.

Laboulbenia Madeirae, 185.

Malayensis, 185.

melanopus, 187.

melanaria, 186.

microsoopica, 187.

microsoma, 187.

minimalis, 188.

Misceli, 189.

obtusa, 189.

Oedodactyli, 189.

Oopteri, 190.

Ophoni, 190.

Orectochili, 191.

orientalis, 121.

Orthomi, 192.

pallida, 193.

Papuana, 193.

Pericalli, 194.

platystoma, 194.

Polyhirniae, 195.

proininens, 195.

protrudens, 196.

Pseudomasci, 196.

punctata, 197.

punctulata, 197.

pygmaea, 198.

rhinophora, 199.

rostellata, 199.

separata, 200.

Serrimai-ginis, 201.

speciosa, 201.

spiralis, 202.

strangulata, 202.

subconstricta, 203.

Sumatrae, 203.

Taenodemae, 204.

tenuis, 204.

Thyreopteri, 205.

tibialis, 205.

tortuosa, 206.

Trichognathi, 206.

triordinata, 207.

tuberculifera, 208.

uncinata, 208.

verrucosa, 209.
Laboulbeniaceae, 151, 407.
Lamourouxia Couzattii, 312.

tenuifolia, 313
Lewis, (j. N., The Developmetit

and Application of a General

Equation for Free Energy and

Physico - chemical Equilibrium,

1-38.
Liabum Pringlei, 342.
Librarian, Report of, 603.

INDEX.

669

Lie's Tlieory of Continuous Groups,

•237, 481.
Liinnaioiiiyces, 428.

Hydrocharis, 429.

Tropisterni, 428.
Lycium geuiculatum, 506.
Lyo-odesiuia ramosissiraa, 315.

Mao-aranthus sulphureus, 566.
Meriniria, 41.

alacris, 43.

bivittata, 42.

intertexta, 42.

neomexicana, 43.

rostrata, 44.

texaiia, 42.

vigilans, 43.
]Metamerism of Hirudinea, 283.
Mexican plants, 305.
Michelson. A. A., The Echelon Spec-
troscope, 109-119.
Mikania eriophora, 341.
Mimosa Deamii, 324.

Psilocarpa, 325.
Misgoniyces, 443.

Dyschirii, 443.

Stomonaxi, 443.
Monoicomyces, 412.

Biitannicus, 413.

Homalotae, 412.

invisibilis, 414.

St. Helenas, 413.
Musical Pitch, 451.

Napaia, 48.

Narragansett Basin, 397.
Nitro Compounds, 2G1.
Nominating Committees, 613, 618.

Oeonomus, 47.
altus, 47.
Officers elected, 608, 615; List of, 643.
Opeia testacea, 46.
Orphulella, 46.
Ostrakichnites carbonarius, 403.

Packard, A. S. , View of the Carbon-
iferous Fauna of the Narragan-
sett Basin. 397-405.

Peirce, B. O., On the Thermal Con-
ductivity of Vulcanite, 73-80.

Perezia Nelsonii, 342.

Perlexia Pringlei, 562.

Peyritschiella Amazonica, 427.
protea, 427.

Physalis acuminata, 311.
philippensis, 568.
Pringlei, 311.
Rydbergii, 569.
saltillensis, 568.
sordida, 568.
subintegra, 567.

Physical Laboratory of the Massachu-
setts Institute of Technology,
Contributions from, 451.

Physico-chemical Equilibrium, 1 ; Re-
action, 469.

Piptothrix Goldmanii, 328.

Pitch, Musical, 451.

Platybothrus, 49.

Polyascomyces, 414.
Trichophyae, 415.

Polygonum Meisnerianum, 324.

Porophyllum brachypodum, 341.

Protichnites narragansettensis, 402.

Psoloessa, 51.

Rhachomyces Canariensis, 436.
Cayennensis, 438.
Cryptobianus, 437.
Philonthinns, 435.
stipitatus, 438.
tenuis, 437.
Thalpii, 436.
velatus, 435.
Zuphii, 436.

Rhizomyces crispatus, 434.

Richards. T. W.,On theDeterraination
of Sulphuric Acid in the Presence
of L-on, 375-383; The Driving
Energy of Physico-chemical Re-
action and its Temperature Coeffi-
cient, 469-480 ; A Revision of the
Atomic Weight of Calcium, 611 ;
A Table of Atomic AYeights, 621.

Richards, T. W., and Baxter, G. P.,
A Revision of the Atomic Weight
of Cobalt ; third paper : The An-
alysis of Cobaltous Chloride and
Oxide, 59-72 ; A Revision of tlie
Atomic Weight of Iron, 2.')l-260.

Richards, T. AV., Collins, E , and
Heimrod, G. W.. The Electro-
chemical Equivalents of Copper
and Silver, 121-I.jO.

670

INDEX,

Robinson, B. L., Synopsis of the Gen-
era Jaegeria and Russelia, 315-

321 ; New Phanerogams, chiefly

Gamopetalae, from Mexico and

Central America, 323-342.
Rocky Mountain, Tertiaries of, 343.
Rumford Committee, Report of, 604.
Rumford Fund, Paper published by

Aid of, 109.
Rumford Premium, Award of, 608,

610, 618, 619.
Russelia, 319.

alata, 319.

equisetiformis, 319.

floribunda, 320.

jaliscensis, 319.

multiflora, 321.

polyedra, 320.

rotundifolia, 321.

sarmentosa, 320.

subcoriacea, 319.

syringaefolia, 320.

tepicensis, 321.

ternifolia, 320.

verticillata, 320.

Salvia, species of, 489.
adenophora, 538.
adglutinans, 532.
aequidistans, 512.
alamosana, 525.
albiflora, 515.
albo-caerulea, 541.
alvajaca, 516.
amarissima, 528.
angustifolia, 500.
antennifera, 545.
aristulata, 551.
aspera, 531.
assurgens, 504.
atriplicifolia, 553.
axillaris, 519.
ballotaeflora, 522.
bella, 545.
biserrata, 544.
brachyodonta, 510.
brevicalyx, 509.
bupleuroides, 497.
cacaliaefolia, 552.
californica, 554.
candicans, 520.
C atari a, 510.
cedrosensis, .521.
chamaedryoides, 521.

Salvia chapalensis, 525.
Chia, 503.
chiapensis, 544.
chrysantha, 531.
cinnabarina, .549.
cladodes, 497.
clinopodioides, 530.
coahuilensis, 520.
coccinea, 550.
Columbariae, 553.
comosa, 501.
concolor, 543.
confinis, 523.
Conzattii, 526.
cordobensis, 514.
costaricensis, 496.
Coulteri, 519.
cryptodonta, 507.
cuneifolia, 519.
curviflora, 544.
cyanea, 544.
cyclophylla, 551.
disjuncta, 533.
elegans, 549.
elongata, 507.
excelsa, 548.
filipes, 510.
firma, 502.
flaccida, 509.
fluviatilis, 516.
Forreri, 518.
fruticulosa, 526.
fulgens, 537.
galinsogifolia, 498.
Gbiesbreghtii, 511.
glabrata, 543.
glechomaefolia, 518.
glumacea, 552.
Goldmanii, 527.
Gonzalezii, 524.
graciliflora, 546.
Grahami, 533.
Greggii, 537.
guadalajarensis, 506.
helianthemifolia, 507.
heterotricha, 500.
hirsuta, 499.
hispanica, 508.
humilis, 495.
hyptoides, 498.
ianthina, 543.
incana, 551.
inconspicua, 492.
involucrata, 539.
iodantha, 547.

INDEX.

GTl

Salvia iodochroa, 540.
irazuensis, 540.
j.iliscana, 515.
Jiirgensenii, 514.
Karwinskii, 548.
Keerlii. 528.
laevis, 502.
lanceolata, 503.
lasiantha, 524.
lasiocepliala, 498,
lantauaefolia, 530.
lateriflora, 494.
lavendiiloides, 505.
Lemmoni, 535.
Leonia, 553.
leptostachys, 511.
leucantha, 529.
Lindenii, 537.
lineata, 539.
Littae, 547.
longispicata, 513.
lophantha, 496.
lophanthoides, 499.
lupulina, 542.
lycioides, 521.
madrensis, 532.
Martensii, 541.
mazatlauensis, 515.
menthiformis, 510.
mexicana, 542.
michoacana, 548.
micrantha, 493.
microphylla, 534.
miiiiata, 545.
misella, 492.
mociuoi, 496.
mollissima, 550.
moticloveusis, 514.
multiramea, 525.
muscarioides, 506.
nana, 508.
Nelsonii, 527.
nepetoides, 529.
nervata, 548.
neurepia, 535.
nitida, 498.
oaxacana, 536.
oblongifolia, 505.
obscura, 492.
obtusa, 536.
occidentalis, 491.
oreopola, 517.
oresbia, 536.
orizabensis, 538.
Ottouis, 551.

Salvia ourophylla, 495.
Palmeri, 511.
pansainalensis, 541.
patens, 552.
pauciflora, 522.
perlonga, 546.
permixta, 496.
phaenostemma, 542.
Pittieri, 530.
platycheila, 523.
platyphylla, 529.
podadena, 492.
polystachya, 511.
populifolia, 530.
prasiifolia, 513.
Pringlei, 539.
privoides, 493.
pruinosa, 526.
prunelloides, 504.
prunifolia, 518.
pteroura, 516.
puberula, 539.
pubescens, 532.
pulchella, 539.
purpurasceus, 540.
purpurea, 546.
pusilla, 495.
ramosissima, 521.
rectiflora, 537.
recurva, 543.
Regla, 533.
remota, 507.
rhombifolia, 517.
Roeraeriana, 553.
roscida, 517.
Rosei, 548.
rubiginosa, 496.
rubropunctata, 527.
saltuensis, 497.
Sanctae-Luciae, 514.
Schaffneri, 535.
scorodoniaefolia, 523.
Seemannii, 516.
seuaiatrata, 524.
serotina, 493.
serpvUifolia, 521.
Sessei, 532.
sessilifolia. 529.
setosa, 493.
Shannon!, 531.
sidaefolia, 529.
sinaloeusis. 502.
spicata, 522.
stachyoides, 507.
stolonifera, 540-

672

INDEX.

Salvia subincisa, 503.

Tatei, 505.

Teresae, 506.

thymoides, 520.

thyrsiflora, 525.

tiliaefolia, 494.

tricolor, 5-34.

tricuspidata, 513.

tubifera, 545.

tubiformis, 551.

unicostata, 501.

venosa, 545.

veronicaefolia, 517.

villosa, 518.

vitifolia, 552.

Wagneriana, 532.

Warszewicziana, 514.

xalapensis, 513.

zacualpensis, 545.
Scudder, S. H., Short Studies of North

American Tryxalinae, 39-57 ;

Two Genera of North American

Decticinae, 81-93; The Species

of the Orthopteran Genus Derot-

mema, 385-395.
Scutellaria distans, 562.

pedicularis, 563.

Rosei, 563.
Sedgwick, W. T., and Winslow, C.

E. A., Experimental and Statisti-
cal Studies on the Influence of

Cold on the Bacillus of Typhoid

Fever and its Distribution, with

Special Reference to Ice Supply

and the Public Health, 614.
Silver, Electrochemical Equivalents

of, 121.
Slocum, S. E., Note on the Chief

Theorem of Lie's Theory of Con-
tinuous Groups, 237-250, 481-485.
Smilax domingensis, 323.
Solanum azureum, 570.

dejectum, 569.

diversifolium, 560.

erythrotrichum, 561.

Fendleri, 559.

riartwegi, 559.

Ilayesii, 560.

Ilerriandesii, 558.

hispidum, 561.

macroscolum, .570.

niadrense, 558.

nocturnum, .570.

ochraceo-ferrngineum, 560.

plumense, 569.

Solanum torvum, 557.
Solid Solutions, 375.
Spectroscope, Echelon, 109.
Sphaleromyces atropurpureus, 441.

Brachyderi, 442.

obtusus, 440.

propinquus, 441.
Spilanthes filipes, 314.
Spiranthes Nelsonii, 307.

Pringlei, 307.

tenuiflora, 308.
Stachys flaccida, 563.

oaxacana, 564.
Statutes and Standing Votes, 652.
Stemmadenia macrophylla, 310.

tomentosa, 310.
Stevia arachnoidea, 326.

Aschenborniana, 326.

bicrenata, 326.

elatior, 326.

madrensis, 326.

monardaefolia, 326.

neurophylla, 327.

Rosei, 327.

Seleriana, 327.

serrata, 328.

tephra, 328.
Stirapleura, 51.

pusilla, 52.

tenuicarina, 53.
Styrax Ramirezii, 309.
Sulphuric Acid, 375.

Taber, H., On the Singular Transfor-
mations of Groups generated by
Infinitesimal Transformations,
575-597.

Telanthera mollis, 324.

Teratomyces Philonthi, 432.
vulgaris, 431.

Thaxter, R., Preliminary Diagnoses of
New Species of Laboulbeniaceae,
151-209, 407-450.

Thermal Conductivity of Vulcanite,
73.

Thomson, E., New Electric Appara-
tus for High Potentials, 618.

Tmetodera, 388.

Torvaria, 557.

Toy, C. H., A New Theory of Totem-
ism, 617.

Treasurer, Report of, 602.

Tropizaspis, 83.
castanea, 84.

INDEX.

673

Tropizaspis diabolica, 86.

ovata, 84.

picturata, 85.

steindachiieri, 86.
Tryxalinae, 39.

Uliue, E. B., New Dioscoreas from
Mexico, 322-323.

Vernonia liatroides, 325.

serratuloides, 313.
Viburnum microcarpum, 313.
Vulcanite, Thermal Conductivity of,

73.

Warren (C. M.) Committee, Report

of, 6U6.
Warren (C. M.) Fund, Aid from, 608.
Webster, A. G., Maxwell's Electric

Top, 613.
Williams, F. B., Note on the Finite

Continuous Groups of the Plane,

95-107.
Winslow, C. E. A. See Sedgwick,

W. T., and Winslow, C E. A.

Zoological Laboratory of the Museum
of Comparative Zoology at Har-
vard College, Contributions from,
283.

MBL WHOI LIBRARY

UH lAfll D

1 3'7f