V IE W O F C H EM ISTS’ B U ILD IN G FROM H O T E L B ELM O N T. M URRAY H IL L H O T E L AT T H E L E F T .
T H E J O U R N A L O F I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y
Vol. I I I . A P R I L , i 9 r i .
No. 4
T h e J o u r n a l o f I n d u s t r i a l
a n d E n g i n e e r i n g C h e m i s t r y
P U B L I S H E D BY
THE AM ERICAN CHEMICAL SOCIETY.
ŁDITORIAL5.
BOARD OF EDITORS.
E d ito r:
M. C. W hitaker.
Associate Editors:
Geo. P. A dam son, E . G. B ailey, H . E. Barnard, G. E.
Barton, W m. Brady, W m . Cam pbell, F. B. C arpenter, Virgil Coblentz, F ran cis I. D u pon t, W . C. E bau gh, W m . C. Geer, W. F. H illeb ra n d , W . D. H orn e, Karl L an gen - beck, A. D. L ittle, P. C. M c llh in e y , E . B. M cC ready, Win. McMurtrie, J. M erritt M atth ew s, T. J . Parker, J.
D. Pennock, W . D . R ichardson , G eo. C. S to n e, E rn st Twitchell, R obt. W ahl, W m . H . W alk er, W . R . W h itn e y .
P u b lis h e d m o n t h ly . S u b s c r ip ti o n p r ic e to n o n - m e m b e r s o f th e A m e r i c a n C h e m ic a l S o c ie ty J600 y e a r l y .
n te r e d a t t h e P o st-O ffic e , H a s t e n , p a ., a s S e c o n d - c l a s s M a t t e r F oreign p o s ta g e , s e v e n ty - f iv e c e n t s , C a n a d a , C u b a a n d M e x ic o e x c e p te d
Vol. III. A P R I L , 1911. No. 4
C O N T E N T S . Editoria ls:
C hem ists' B u il d in g ... 205
H is to ric a l... 207
A New D y e in g I n d u s t r y to b e E s ta b lis h e d ... 208
The Ch e m is t s' Bu i l d i n g... 208
Original Pa p e r s: P re lim in a ry R e p o r t o n th e T e r n a r y S y s te m C;iO-Al'vO;;-HiCK A S tu d y o f tile C o n s tit u tio n o f P o r tla n d C e m e n t C lin k e r. B y E . S. S h ep h erd a n d G . A . R a n k i n ... 211
M ethods f o r T e s tin g C o al T a r a n d R e fin e d T a r s , O ils a n d P itc h e s D eriv ed T h e r e f r o m . B y S . R . C h u r c h ... 227
The B e h a v io r o f H ig h -b o ilin g M in e ra l O ils o n H e a t in g in tile A ir. B y C. E . W a t e r s ... 233
The D e te rm in a tio n o f M a n g a n e s e b y th e S o d iu m B is m u t h a te M ethod. B y P a u l I I . M .-P . B r i n t o n ... 237
M ethod fo r T e s tin g G a lv a n iz e d I r o n to R e p la c c th e P r e e c e T e s t. . . B y W a lte r A . P a tr i c k a n d W illia m H . W a lk e r ... 239
I h e F o r m a tio n T e m p e r a t u r e o f C a r b o r u n d u m . B y H o r a c e W . G ille tt... 242
A nalyses o f C h a r a c te r is tic N o r th w e s t T i m b e r S o ils. B y J . S. J o n e s . . . ... 246
In v e s tig a tio n s o n th e E s ti m a t io n o f I n o r g a n ic P h o s p h o r u s in A n im al T iss u e s . B y A . C. W h i t t i e r ... 2 48 A V o lu m e tric M e th o d f o r A n tim o n y in A llo y s. B y G e o rg e S. J a m ie s o n . ... 2 50 D e te rm in a tio n o f C o co a S h e lls in C o c o a P o w d e r. B y W . L . D u bois a n d C. I . L o t t ... 251
The E s tim a tio n o f E s s e n tia l O ils. B y C h a rle s D . H o w a r d . . . . 252
R ep o rt On M e th o d o f D e te r m in a tio n o f G ly c e ro l. B y E u g e n e P ro b e c k ... 253
The P resen ce o f A r s e n ic in th e C o a tin g of T a b le ts . B y H . C. F u lle r . . ... 254
The D e te c tio n o f P r u n e J u ic e a n d C a ra m e l in V a n illa F la v o r in g E x tra c ts . B y W . D e n is ... 254
M ethod f o r N ic k e l-Z in c S e p a r a t io n in G e r m a n S ilv e r a n d O th e r A lloys. B y I . a V e rn e W . S p r in g ... 255
A R ap id V o lu m e tric M e th o d o f S u g a r E s ti m a t io n . B y W . F . S u tlie r s t... 256
The Im p o r ta n c e o f a S t a n d a r d T e m p e r a t u r e f o r S p ecific G r a v it y D e te rm in a tio n s , a n d f o r S ta n d a r d iz i n g M e a s u re s o f C a p a c ity . By G. W . T h o m p s o n ... 2 56 A M ethod fo r th e D e te r m in a tio n o f T in in C a n n e d F o o d s. B y H erm an S c h r e ib e r a n d W . C. T a b e r ... 257
Scientific So c i e t i e s... 261
Notesand Co r r e s p o n d e n c e... 262
Consularan d Tr a d e No t e sa n d Ab s t r a c t s. ... 265
Book Re v ie w s... 2 68 Abstractsan d Ne w Pu b l ic a t io n s. ... 2 7 0 Recent In v e n t io n s... 272
Market Re p o r t s... 278
CHEMISTS’ BUILDING.
The opening of the new Chem ists’ B uilding in New Y o rk C ity is an event of national, rather than m erely local significance, and the com m ittee in charge has done its d u ty b y seeking to express this fa ct in the program m e of the opening exercises. The social com forts of the Chem ists’ Club are b u t an incident in the general scheme, and, in fact, the physical transfer of th at organization from its present quarters to the splendid home now provided for it must, necessarily, aw ait the com pletion of its furnishings, a fter the building itself was declared ready for occupancy.
Hence, the C lub’s festivities were subordinated to the dedication ceremonies of the building, when due emphasis could be laid upon the serious aims of the enterprise, and to the scientific meetings, under the auspices of the local sections of our Society, the A m er
ican Electrochem ical Society and the Society of Chemical Industry. It w ill not be the fau lt of the speakers a t these meetings, if the general public fails to grasp the im portance of chem istry in our industrial developm ent, and also as a branch of pure science.
W e ourselves, and, especially, those of us who do not dwell in New Y o rk itself or its im m ediate vicin ity, m ight well take this occasion to reflect upon the practical significance of this undertaking.
N obody can review the history of Am erican progress during the past tw en ty-five years w ithout recognizing how much more intim ately chem istry is enmeshed in the general economic and sociological texture than a quarter century ago. Then, the Am erican student who w ent beyond the general chem ical courses pre
scribed for all freshmen and sophomores did so as a preparation for medicine or some other recognized profession, or w ith the definite purpose of entering an academ ic career. The possibility of establishing himself as an independent chem ical analyst or expert was certainly never placed before the student; and the works chemist, in the eyes of the industrial world, was held in the sort of regard which m ay be likened to Lincoln’s estim ate of the valu e of brigadier-generals, when he remarked, on hearing th a t a Confederate raider had cu t out a baggage train and captured three generals, “ I can m ake a brigadier any day, b u t mules cost m oney.” Under such conditions, chem istry as a profession could have slight standing, and the gregarious needs of its votaries were am ply m et b y the annual “ m eetings” of Section C of the Am erican Association for the A dvancem en t of Science, and b y occasional, local gatherings. The reorganization of the Am erican Chem ical Society on a national basis, in June, 1890, presaged the change which seems to have followed the W orld’s Fair at Chicago. W e need n ot inquire too closely w hether the lean years
2o8 T H E J O U R N A L O F I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y . April, i 9n Dr. Toch becam e President of the Club in 1907,
and in his retiring address, after review ing the success of the Club, he announced th a t the tim e was ripe and he was ready to undertake to develop w ays and means for the chemists to build a larger and better house of their own. Professor Loeb becam e interested, and suggested a plan to include special laboratories, offices, and other accom m odations for professional chemists.
Professor Loeb furtherm ore gave tangible form to. the enterprise b y offering to subscribe $50,000 which he subsequently increased to $75,000. Dr. Nichols and his associates subscribed $50,000. A Finance Com
m ittee was organized and various plans were discussed.
In 1908 the old Tilden Club Building a t 74th Street and B roadw ay was offered for sale, and was being con
sidered b y a com m ittee, when the present site at 52-54 E ast 41st Street came to the attention of Dr.
M cKenna, who again realized an opportunity and appreciated its advantages as a central location and arranged for its purchase.
The building was planned and erected under the supervision of the Directors of the Chem ists’ Building Com pany, Dr. Morris Loeb, Dr. Charles F. Chandler, A lbert P laut, W . H. Nichols, Jr., and Dr. L. H. B aeke
land.
The stead y progress of the Chem ists’ home in New Y o rk , from the single room in W ashington Square furnished w ith packing cases and strewn w ith books and journals, to the handsome eleven sto ry building a t F ourth A venue and 41st Street, w ith its auditorium , library, museum, laboratories, offices, social rooms, living quarters and restaurant will alw ays be inti
m ately coupled w ith the history of the developm ent of the profession, and m arked b y the unselfish de
votion of these and m any other loyal hard-w orking men to the interests of chemists, w hether th ey be n ative or foreign, young graduates, or men years in the service of the profession.
A NEW DYEING INDUSTRY TO BE ESTABLISHED.
The Bradford D yers’ Association, a v e ry large E n g
lish corporation, which p ractically controls the great bulk of dyeing in the United Kingdom , has definitely
decided to establish a branch works in this country.
T his has been brought about in one sense b y the opera
tion of the tariff which has pu t a higher rate of duty on dyed and finished cloth than on the unfinished material. In another sense it has also been brought about b y the increasing com petition of Am erican mills- in the production of novelties which have hitherto been specialties of the Bradford D yers’ Association.
The Englsh Corporation has q u ietly bought up about 800 acres of land along the P aw catu ck R iver at Ni- antic, R. I. Their purchases include also the small plant of the N iantic D yeing Co., together w ith flowage rights, docking privileges, and railroad facilities.
There is every indication th at a large plant is to be erected in the im mediate future w ith the idea of de
veloping a general commission dyeing business on a large scale. The Bradford D yers’ Association has now forty-six plants in operation in England, and as the com pany is backed b y alm ost unlim ited capital, and has under its control a large num ber of processes and valuable specialties, there is no doubt but that its influence will be largely felt in this phase of Amer- can industry. I t is expected to have a unit plant in operation a t N iantic b y the com ing summer;
that is to say, a plant of sufficient size to dye and finish about 1000 pieces per day. A s the business develops and grows this unit plan t will be duplicated as often as conditions demand. A s the circumstances and organization of the Am erican textile industries are som ewhat different than in England, it is a question open to considerable discussion as to whether this move of the Bradford D yers’ Association will prove to be a successful one. One factor which has prob
a b ly hastened the determ ination of the English firm is the circum stance th at a large French dyeing cor
poration has also decided to enter the Am erican field b y the erection of extensive plants in this country.
It is to be hoped th at the Bradford D yers’ Association will have a more successful venture than was the case w ith the earlier experiences of the British Cotton and W ool D yers’ Association, in their endeavor to establish a branch of their yarn-dyeing industry in America.
J . Me r r i t t Ma t h e w s.
THL CHLMI5T5’ BUILDING.
B y lv. L . El l i o t t, E d i t o r I llu m in a tin g E n g in eer.
It is a question which of the tw o sciences whose origin extends back into the ages possesses the richer lover of historical romance, astronom y or chem istry.
W hen in the dawn of civilization m an began to wan-
■der from place to place, seeking to destroy or to es
cape destruction, he learned to use the stars as guide posts, and his im agination discovered in the heavens the outlines of the beasts of the field and the other forms of nature w ith which he was most fam iliar; and so he came to believe th at his entire destiny'' lay w ithin the powers of this m ysterious dome above.
T h e race of astrologers is even not y e t extinct, while the modern science of astronom y has m apped and weighed the heavens w ith well nigh superhuman exactness.
The beginnings of chem istry came later, when man had learned to extract some of the commoner metals and turn them to his use; and when gold became the substance and the sym bol of w ealth, the search for m ethods of producing it from the baser m etals afforded a field of infinite possibilities, w ith the motive of cupidity for all th at wealth signifies as an instigator and prom oter of the work. Added to this was the equally powerful incentive of discovering a means of escaping "sickness and death. The universal belief in im m ortality is the child of a wish to defeat the inevi
table course o nature, to the end th at man m ay enjoy perpetually the delights of youth. To possess y o u th
forever and gold a t w ill! W ho would not strive to accomplish this end? The m odem science of chemis-
A U D ITO R IU M FROM T H E RO STRUM .
SOCIAL ROOM
T R U S T E E S ’ ROOM.
try thus has an origin quite as rom antic and fanciful as its sister science of astronom y.
Today we think of the alchem ist as a m isguided philosopher of the Middle Ages, hopelessly striving to achieve the impossible, bu t who, building better than he knew, laid the foundations of the science of chem istry, which in its benefits to m ankind a t least equals all the inventions and im provem ents in the mechanic arts.
Nor need we turn to Europe or the D ark Ages to find an example of the popular distrust of the “ black a r t ” of alchemy. The first chem ical lab oratory ever insti
tuted in this country was p ractically a dungeon, being entered b y a trap door from the top, and occupying a subterranean room; such was the laboratory of Prof.
Silliman, of Y ale.
It has been well said th at the progress of science has proceeded with an accelerated velocity, a year now equaling the progress of a previous decade, and a decade surpassing the progress of a previous century.
Within the span of a single human life we have pro
gressed from the cellar of Prof. Sillim an to the m agnifi
cent modern fire-proof structure of this Club; and the Club itself has b u t barely reached its m ajority, and the building in which' it is now housed was onl}r a bold conception of one of its honored presidents but four years ago.
In exterior design the building m ay best be charac
terized as modern Am erican, the architecture being an adaptation of the best traditions of building to the necessities and conditions of modern construction and requirements. The facade is of light buff sandstone treated in a simple and wholesome manner w ithout any effort a t conspicuous decoration, the general motive being classical. The balcony at the top and the metal work about the windows is finished in verd bronze, giving a most pleasing harm ony of color.
The general plan of the building is unique in com bin
ing all of the social requirem ents of a Club w ith the practical utilitarian features of the professional chem ical laboratory. These two features, while intim ately combined, are a t the same time entirely separated for practical purposes. Thus, the auditorium and the laboratories can be reached w ithout in an y w a y intru
ding upon the portions of the building devoted to club use, while on the other hand, the Club sections of the building are in direct com m unication w ith the business portion. A detailed description of the building b y floors will make this arrangem ent clear.
The first, or street, floor is occupied b y the lobb y and office of the Club, and the auditorium . The lobb y occupies the front section, and is entered directly from the street. A s you enter, the office is a t the im m edi
ate left, a handsome fire-place a t the right, and the main hallway leading to the auditorium in the rear At the left of this hallw ay is the main stairw ay to the various club-rooms above, w ith a coat-room a t the entrance to the auditorium . The walls of the lobby are finished in plain oak paneling in antique finish, with pilasters of freestone; the ceiling is w hite w ith simple plastic decoration of conventional classic design.
The auditorium is two stories in height w ith a gallery m the rear. It has a seating cap acity of 300 on the
main floor and 50 in the gallery. The stage is lighted b y a skylight, and has preparation-room s on either side. R em ovable tables are provided, so arranged th at th ey can be readily connected to supply water, gas, electricity, compressed air, and vacuum ; when not required for experim ental or dem onstration pur
poses th ey can be removed, and the stage left entirely free. A panel a t the rear of the stage is finished white to serve as a stereopticon screen, and a stand for the lantern, w ith electric current a t hand, is provided in the gallery. A blackboard, arranged so th a t it can be drawn up into position for use or lowered when the screen is required for lantern views, is also provided.
A n ample hallw ay leading dircct from the street to the auditorium extends along the right side of the building, so th at it is possible to close the auditorium e n tir e ly . from direct com m unication with the lobb y when desired, thus m aking it available for use for other purposes w ithout in an y w ay encroaching upon the p rivacy of the club members. The' elevators to the upper floors are also reached directly b y this hallw ay, which opens into the lobby, thus perm itting the labora
tories and private rooms to be reached w ith equal fa
cility either through the lobb y or from the street direct.
The gallery of the auditorium is reached from a m ez
zanine floor, which is devoted to private offices. L a v atories are provided in the section between the stair
w ay and the auditorium .
The second floor is given over to the dining-room and the social room. The latter occupies the entire front section of the building, having large windows on the street side, and a handsome fire-place at either end.
T he general character of the architecture has been care
fu lly preserved in this as in the other features of the interior tre atm en t; there is simple elegance rather than ornate decoration. The woodwork is of m ahogany;
the side walls are covered w ith tap estry in natural color background w ith a simple figure in old blue; the ceiling is paneled in v e ry low relief, and given a cream tint. A s the rooms were not furnished a t the tim e of the dedication the photographs necessarily show them in this condition, and so, while giving an idea of their spaciousness and general appearance, the im agination must supply the air of homelike com fort which only a furnished and used room can possess.
The dining-room occupies the rear section of the floor, and is finished in m ahogany, w ith walls of dull red. The ceiling is perfectly plain and of a light cream tint. A t this height the rear of the building has no im m ediate obstruction, so th at the dining-room is beau tifully lighted b y large windows facing the south.
This gives the cheerfulness of such sunshine as is vouchsafed us in the winter months, while in the sum mer the projecting roof of the auditorium affords an open air ca.16.
A dining-room necessarily implies a kitchen, and this necessity has been m ost carefully provided for in a room of am ple dimensions and equipped w ith every m odem facility known to the culinary art.
There is a p articu larly close relation between chem istry and cooking; b u t whether the an alytical propensities of the chemists are sufficiently perfected to enable
2 I O T H E J O U R N A L O F I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y . A p ril, i 9n them to prepare viands synthetically remains .to be
seen. It is a fairly safe prophesy th at the savants of the test tube and wash bottle will be satisfied to follow Sancho P an za’s form ula in thier tests of the puddings.
On the third floor the front section has been set apart for the library. The finish here is quartered oak w ith plain walls. The bookstacks are arranged along the inner wall, leaving the front portion b y the windows for the reading tables. Dr. C. F. Chandler has signified his intention of donating his private li
b rary to the Club for im mediate use. This benefac
tion will afford a splendid foundation for a very com plete chem ical library. It is worth m entioning th at the gift comes from a teacher who m ay rightfully be called the Dean of Chem istry in this country, and a fine life-sized bronze of Dr. Chandler b y H artley fur
nishes a fitting and striking figure in this room, which is to be known as Chandler Hall.
The rear section of this floor contains a room which is to be used as a museum and reference room. It is finished in quartered oak and is left w ithout decora
tion. Adjoining this there is a small room which is perhaps the most unique feature of the entire build
ing. The official name of this is the Trustees’ Room, but it has already acquired the facetious title of “ The Chapel.’ ’ The room has been designed to represent the den of the ancient alchemist. There is a “ high- arched, vaulted ceiling” under which the im agination can readily depict Dr. Faustus himself ranging over the field of human knowledge, to find himself a t the end no wiser than before. A Gothic window w ith roundels set in leaded glass adm its subdued daylight, while a t the opposite end is a crude furnace, ready to receive the alem bic and the crucible. A small metal lantern of m ediaeval pattern hangs from the ceiling a t this end, while the traditional salamander, defying history b y grasping modern electric lamp bulbs in his claws, swings from chains at the other end. A n iron chest of ancient design, w ith ponderous key, occupies a corner of the room, and doubtless contains the secret of the philosopher’s stone and the elixir of life. A massive oaken table occupies the center of the room, and suggests the parchm ent tome in black-letter, but alas is more likely to witness the anachronism of a stenographer’s notebook.
The fifth and sixth floors are divided into single rooms and suites for the use of the Club members and their guests. In the furnishing of these some genius hit upon a very happy idea, viz., to have a num ber of the alum nae of colleges th at are well represented in the membership each furnish a room or suite. The fol
low ing institutions have accepted this idea: U niver
sity of Pennsylvania, New Y o rk U niversity, Cornell U niversity, Massachusetts In stitute of Technology, Y a le U niversity, College of the C ity of New Y ork , U niversity of Michigan, U niversity of Virginia, Johns H opkins U niversity, German Universities, Swiss Uni
versities, British Schools and Universities, U niversity
of Tokio, and the C ity of Chicago. T he walls are hung w ith view s of the fam iliar scenes on the campus and in the buildings, together w ith the emblems that are dear to the heart of the college man. Thus, added to all the com forts to be found in the best modern club or hotel is the memory of the good old college times, when it was considered a m ark of special prowess to blow the gas pipes in the chem ical “ la b ” full of air, or produce reactions between N H tOH and H C1 th at were not prescribed in the course, and which resulted in a general clouding of the vision, or perhaps to perfume the instructor’s overcoat with valerianic acid.
The remaining six floors of the building are designed for laboratory use. A ll the facilities required by the chemist are provided, and the spaces so arranged th at th ey can be subdivided to suit the exact require
ments of the tenants. The floors are of cement, and are made continuous w ith the baseboard, and are sup
plied w ith a drain a t either corner so th at a stream of running w ater could be left on the floor continuously w ithout doing the slightest dam age to the floor itself or the room beneath. There are arrangements for hoods with separate outlet and inlet openings reaching to the roof. These rooms offer the combined advan
tages of com plete modern equipm ent, unsurpassed light and ventilation, the most central location in the city, and immediate connection w ith an organization representing the entire chem ical profession of the country. It is small wonder therefore th at th ey have been eagerly taken up b y consulting and professional chemists.
A feature deserving of special mention is the com
plete equipm ent, including all the regular chemical ap
paratus, of two small laboratories, which will be let to com petent chemists who wish to conduct temporary experim ents or demonstrations. These are on the ninth floor, and are known as the R obert Bunsen and the W olcott Gibbs laboratories. The difficulty of se
curing such facilities is a common experience with the chem ical departm ents of universities, and w ith those who have found it necessary to m ake demonstrations of new processes or products to prospective investors.
If there ever was a case of “ filling a long-felt want”
this is certainly one.
The building is served b y three elevators which im
m ediately reach all offices and laboratories, and also b y stairw ays a t each side.
The building throughout shows the utm ost care and thought in the planning of every detail. It should perhaps rather be called a growth than a plan, for the arrangem ent and the construction have received the unrem itting thought and attention of a number of the best chemists, in connection with the architects, Messrs. Y o rk and Sawyer. B u t an original concep
tion worked out to so com plete a success as this is worth a lifetime of labor and thought. The Chem
ists’ Club stands tod ay as the most unique building of the kind in this or any other country.
ORIGINAL PAPLR5.
p r e l im in a r y r e p o r t o n t h e t e r n a r y s y s t e m CaO-Al.0.,-Si02. A STUDY OF THE CONSTITUTION
’ OF PORTLAND CEMENT CLINKER.
B y E . S . Sh e p h e r da n d G . A . Ra n k i n. I
W i th O p tic a l S t u d y b y Fr e d E - Wr i g i i t. R e c e iv e d F e b r u a r y 2 4 , 191 l . v
Two earlier papers from this lab orato ry1 having prepared the w ay for a system atic stu d y of the ternary systems, we desire to present the results thus far obtained with the interesting oxides, lime, silica and alumina.
It would be better for a num ber of reasons to de
lay publication until a final paper could be presented, but the active interest of technical men in the P ort
land cement portion of the diagram, as well as the necessarily time-consuming nature of the observa
tions, seems to ju stify offering a prelim inary report without further delay. To fix the position of all the fusion surfaces of this diagram will require con
siderably more time, b u t it is hoped th at the facts thus far obtained w ill suffice as a starting point for those primarily interested in technical cem ent prob
lems.
It will be well to begin b y review ing briefly the results obtained in the stu d y of the binary system s, partly in order th at we m ay have clearly before us the conclusions upon which the present work is based, and partly because the work w ith three components has revealed some im portant new developm ents in the relations which obtain in the two-com ponent systems.
Limc-Silica Scries (Fig. i) .— It will be recalled th at pure lime and pure silica were found to combine to form two definite compounds, the m etasilicate (Ca0.Si02) and the orthosilicate (2 C a0 .S i0 2). In mixtures of pure lime and silica, no trace of the gen
erally accepted tricalcic silicate (¿CaO.SiOj) w-as found, nor of a compound of the ty p e 4 C a0 .3 S i0 2, which is fam iliar to m ineralogists under the name akermannite. More recently, we have found th a t a small addition of alum ina is sufficient to bring out the tricalcic silicate, which proves to be a compound with peculiar properties and lim itations which will be discussed presently, ju st as V o g t has found2 that the addition of enough m agnesia to an appropriate mixture of pure lime and silica will bring out the akermannite. It is interesting to note in passing that magnesia in m ixtures of tricalcic silicate com po
sition also develops the tricalcic silicate compound, but good crystals are not obtained b y using oxide of iron, vanadic acid or other usual laboratory fluxes.
Of the pure silica, six different crystalline forms with definite regions of stab ility were described in the earlier work, and one of lime, b u t w ith the im portant corollary, in the case of the latter, th at the proper
ties of the so-called amorphous (burned) lime differ
1 “ T h e I.im e -S ilic a S e rie s o f M in e ra ls .” A r t h u r L. D a y . E . S. S h e p h e r d
•Hid F. E . W rig h t, A m . J . S c i .. [4 ] 2 2 , 265 (1 9 0 6 ). " T h e B in a r y S y s te m s of A lumina w ith S ilica. L im e a n d M a g n e s ia ,” E . S. S h e p h e r d , G . A . R a n k in :ind F. E. W rig h t. I b id .. [4 ] 2 8 , 293 (1909J.
■ J. H . L . V o g t, P i e S iiik a tsc h m e h lo s u n Q e n , I I , 22.
in some im portant particulars from the more coarsely crystallized form. Tw o crystalline forms of calcium m etasilicate were established and designated a (some
times called pseudowollastonite), and [i (corresponding to the natural mineral wollastonite). Three stable crystalline forms of the orthosilicate w ith definite properties were established, defined and given the names a-, 0-, and r-orthosilicate, the a-form being
F ig . 1.
the first crystalline form to appear in the melt. ■ In addition to these, a fourth and apparently instable crystalline m odification has now been discovered, which m ay be of im portance in the constitution of Portland cement. T o avoid confusion, the new m odification has been designated ft'. The new com pound, tricalcic silicate, has so far been found in but a single stable crystalline form.
These compounds when brought together in varyin g compositions yield eutectics as follows: (i) A t 37 per cent, lime, 'between pure silica and the m eta
silicate; (2) a t 54 per cent, lime, between the m eta
silicate and orthosilicate; (3) 67.5 per cent, lime, be
tween the orthosilicate and pure lime. The tricalcic silicate becomes unstable in the v icin ity of 1900 o C.
and does not appear to enter or come out of the binary m elt às such. F or this reason it does not enter into eutectic or other relations w ith the adjacent compounds, calcium orthosilicate and lime. Since Fig. 5 was . drawn, a sm all field has been discovered in which there
2 1 2 T H E J O U R N A L O F I N D U S T R I A L AND. E N G I N E E R I N G C H E M I S T R Y . April, i 9n is a crystal sim ilar to akerm annite but of a composition,
3C a0 .aS i0 j. The relation of this phase to the binary system is quite sim ilar to th a t of the tricalcic silicate.
The m elting tem peratures of the compounds and eutectics are all relatively high. Pure silica melts to a liquid of extrem e viscosity a t about 1600°, the eutectic between silica and calcium m etasilicate melts a t 1426°, the m etasilicate itself ( a-form) at 1540°, the eutectic between the a-m etasilicate and the a-orthosilicate a t 1440°, a-orthosilicate at 2086°, and, finally, the eutectic between a-orthosilicate and pure lime a t 2015 °. The m elting point of pure lime is beyond the reach of accurate tem perature measure
ment. The tricalcic silicate becomes unstable and dissociates into the orthosilicate and pure lime be
fore the m elting tem perature is reached.
In entering upon the three-com ponent diagram from the lime-silica side, we have to take account, therefore, of fifteen possible phases: Pure silica (six forms), calcium m etasilicate (two forms), 3 C a0 .2 S i0 2 (one form), calcium orthosilicate (four forms), tricalcic sili
cate (one form), pure lime (one form). M any of these possible m odifications are of minor significance for the stu d y of the constitution of Portland cement.
Alum ina-Silica Series (Fig. 2). — The alumina- silica series presents only one compound, A l203.Si02—
sillim anite— which forms eutectic m ixtures w ith the pure components on either side of it. The eutectic between sillim anite and corundum occurs a t 64 per cent. A ljO ,, tem perature slig h tly ’ below the m elting
/A // // c t/
0 F ’
✓
x
s A n s '
"
G
-S i 02 A l2S¡ O5 AL2O3
F ig . 2 .
point of pure sillim anite (1816°). The eutectic between sillim anite and cristobalite also occurs ju st below 1600°. Tw o other crystal forms of A l203.Si02 (andalusite and cyanite) occur in nature, b u t th ey have neither been successfully formed in the laboratory nor has their relation to sillim anite been satisfactorily
established beyond the fa ct th at th ey go over into sillim anite on heating.
E ntering the triangular diagram from this side, there is, therefore, bu t one stable phase which may be expected to participate in addition to the pure components, alum ina and silica.
Lim e-Alum ina Series (Fig. 3). — This series was found to contain four compounds in addition to the
Fí e. 3 .
original components, and three eutectics. The com
pound 3 C a0 .A l20 3 is unstable at its m elting point, in consequence of which the invariant system CaO, 3 C a0 .A l20 3, liquid and vapor, does not reach a minimum tem perature. The system 3 C a0 .5A l20 3,A ls0 3, liquid and vapor, seems to present a sim ilar case, though the tem perature is too high to m ake this determina
tion certain. There is a monotropic form of both the s C a 0 .3 A lj0 s and 3 C a0 .5A l20 3. 5C a0 .3A l20 3 and C a0 .A l203 present normal m axim a.
Entering the three-com ponent diagram from the lim e-alumina side, four stable phases m ay therefore be expected to participate. The tw o unstable forms thus far discovered do not appear w ith any regularity in any of the three-com ponent m ixtures studied.
New Phases.— Before entering upon the detailed dis
cussion of the three-com ponent system , it remains to describe the new two-com ponent phases which have been discovered since the publication of the earlier papers referred to. Of these two, the tricalcic silicate and the p' form of the orthosilicate, m ay be im
portant constituents of Portland cement. Of these, the ft' form of the orthosilicate is w holly new, but the tricalcic silicate was sought for m ost diligently in the earlier stud}" of m ixtures of pure lime and silica, because of the ve ry general opinion prevailing among students of cement problem s th at this compound existed and formed one of the most im portant com
ponents of Portland cem'ent; bu t no trace of it was
ever found. The exam ination of the m ixtures of this composition made w ith great care in this labora
tory, together with some preparations made b y such distinguished experts as N ew berry and Richardson, yielded one of the forms of the orthosilicate and an excess of free lime in every instance. The invari
able reappearance of m ixtures of orthosilicate and free lime in preparations of tricalcic silicate com po
sition, in place of the homogeneous product necessary to define and establish the properties of a new com
pound, when combined w ith the equally persistent reappearance of the orthosilicate inversion a t 1410°
throughout the region where the tricalcic silicate should appear if such a compound existed, and the complete absence in a v e ry extensive literature of any adequate evidence of hom ogeneity in the jjroducts to which the name tricalcic silicate has been given, led us to the positive conclusion th at no compound of tricalcic silicate composition existed. A ll this evidence still remains unshaken, and y e t the com pound tricalcic silicate has recently been prepared in a state of high p u rity and homogeneous, except for a persistent trace of the orthosilicate, or lime, or both, which varies in am ount from one to two per cent, in individual cases. This som ewhat contra-' dictory situation requires the addition of b u t a single experimental observation to clear it up com pletely:
the tricalcic silicate has been found to be unstable at its melting tem perature and for some distance below, so that a melt of this composition invariab ly crystal
lizes on cooling to orthosilicate and lime, which then appear with the properties and in the successive modi
fications already described. In the presence of alum ina or even alone, if held for a sufficient tim e a t tem pera
tures in the vicin ity of 1800°, the orthosilicate and lime combine to form a new com pound w ith new and independent properties and homogeneous within the limits above noted.
The new compound has an index of refraction practically identical w ith /3-orthosilicate, bu t in the binary system we have never been able to develop crystals of it which were large enough to allow a satisfactory determ ination of its optical character.
In the ternary system , fair-sized crystals have been obtained and prove to be w eakly birefracting and optically negative. The relations in the binary system were further com plicated b y the discovery of the apparently instable ft' form of the orthosilicate which is also weakly birefracting, bu t optically7 positive.
After a long search for the composition of this new optically negative phase, we were able to assure ourselves, by means of a large num ber of preparations varying in composition b y one per cent, and often by only half of one per cent., th at preparations ap proaching 3 C a 0 .S i0 2 were richer in this new phase.
Ihis seemed to point definitely to the existence of a new compound in the binary series, bu t the 3 : 1 and 4 : 1 m ixtures of the pure components when fused in the iridium furnace (2 o o o °-2 io o ° C.) gave only well crystallized lime and orthosilicate, as before.
Preparations made from ve ry finely ground lime and silica, or even from precipitated (gelatinous)
silica and hydrated lime, did not }rield a homogeneous preparation w ith good optical properties.
W e had already observed th a t the ternary prepara
tions, after being heated in the gas furnace, or baked (w ithout melting) for days or weeks in the platinum furnace a t tem peratures near 1500°, gave large am ounts of this optically negative phase. B u t the same prepa
ration when melted in the iridium furnace gave large quantities of lime and orthosilicate and much less of the o ptically negative material. It was here th at we began to suspect th at the op tically negative phase was instable in contact w ith the melt.
W e then tried com bining lime and silica in the 3 : 1 ratio b y baking them for a long tim e in the platinum furnace. B u t even so we were unable to obtain crystals whose optical character could be satisfactorily determined. Subsequently, the prepara
tions were baked for several hours in the iridium furnace a t about 1800-19000 C. B y this means we were able to obtain a nearly homogeneous m aterial which upon fusion w ent over into lime and orthosilicate bu t of which again the optical character could not be definitely ascertained. The 74.5 per cent. CaO prepara
tion had alw ays lime in excess, while the 72.5 per cent, preparation contained orthosilicate in excess. On the other hand, the presence of 0.5 per cent. A 1203 proved to be quite sufficient to develop the crystals, bu t the preparation is not homogeneous. Finally, m ixtures of the 3 : 1 composition were baked until th ey were p ractically homogeneous and consisted of the w eakly birefracting m aterial. Then th ey were fused and allowed to recrystallize, whereupon th ey yielded products in which only CaO and 2 C a0 .S i0 3 could be discovered. T he reaction is evid en tly som ewhat slow especially before the appearance of the liquid phase. I t is not impossible, therefore, th at we have placed the dissociation tem perature a little too high.
A s the result of all the observations (some 800 in number), one m ay conclude th at 3 C a 0 .S i0 2 belongs to th at class of compounds which form b y reaction between the solid components, b u t which decompose before the m elting tem perature is reached, i. <?., th ey are w h olly unstable in contact w ith the m elt. This relation is indicated b y the line N L M in figure 1, (3 C a0 .S i0 2 + CaO + Ca2Si04). L is the quadruple point at which 3 C a 0 .S i0 2 decomposes to form lime and C a,Si04; thus
3 C a 0 .S i0 2 > 2 C a0 .S i0 2 + CaO.
The reaction is reversible though it m ay be expected to proceed slow ly b y reason of the phases being solid and vapor. The time factor becomes a serious con
sideration once the preparation has been m elted and allowed to crystallize again into lime and ortho
silicate crystals of relatively large dimensions. These w ill combine to form tricalcic silicate under favor
able conditions but w ith extrem e slowness. In the ternary system the liquid phase appears a little above 1400° C. and greatly facilitates the reaction. I t is a noticeable fa ct th at while certain ternary concen
trations containing several per cent, of A 1203 when suddenly cooled from 16000 C. show a certain q u an tity of free lime, the same preparations show a much
2 1 4 T H E J O U R N A L O F I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y . April, i 9„
greater q u an tity when cooled from 1750o. On being reheated a t 1600 o the q u an tity of lime diminishes, showing a read y reversibility, i. e., the concentration of the melt follows along the boundary curve ju st as would be expected.
If this phase were m erely instable at its m elting point like the tricalcic alum ínate, then after fusion, one m ight expect a m ixture of CaO, 3CaO.SiO, and CajSiO^, b u t lime and orthosilicate only are found, so th at the evidence is all in support of the view th at tricalcic silicate in the bin ary series is never stable in contact w ith the m elt. The observations in the ternary system lead to the same conclusion. The binary diagram (Fig. 1) therefore has the new lines N -L -M 3CaO.SiO, added and the dotted lines O P and K ' Iv", which indicate th at the inversions a < > ft and ft c y y in the orthosilicate are instable extensions of these inversions into the field belonging to 3 C a 0 .S i0 2 + CaO.
A New Form of C a ,SiO t.— A n old and hydrated sample of orthosilicate when quickly cooled from a tem perature of 14 2 s0 yielded a w eakly birefracting and optically positive crystal which could not be obtained w ith fresh preparations. A ttem p ts to locate the inversion tem perature b y therm al means showed two heat changes about 300 apart, one of which disappeared after a few heatings and could not again be obtained. E x cep t for w ater and probably some carbon dioxide, the preparation was pure. W e re
gard this form as instable. I t appears occasionally in the ternary m ixtures, bu t with no regularity.
In brief résumé then, we find in the binary series betw een pure lime and pure silica four compounds:
CaO SiO ,— two forms a and ft. M.P. a = i5 4 o ° C . 3 C a0 .2 Si0 2— one form.
2CaO.SiO„— four forms, a, ft, ft' and r of which ft' is monotropic.
3CaO.SiOs— instable in contact w ith the melt, the quadruple point falling a t about 1900 o C.
There are three eutectics:
aCaO.SiO, + SiOj 1426° C.
aCaO.SiOj + a2CaO.SiO, i4 4 o °C .
a2CaO.SiOj -f CaO 2 o i5 ° C . t
T h e inversion ¿CaO .SiO, ( y /?Ca0 .Si02 occurs at about ' 1190 0 C.
The inversion a2CaO.SiO. foC aO .SiO , occurs at about 1410 o C.
The inversion A C a O .S iO , ( > r2CaO.SiO, occurs at about 675o C.
I t will be recalled th at this last inversion is accom panied b y a 10 per cent, volum e change which causes the “ d u stin g” of the orthosilicate and preparations containing it in considerable quantity. In our e x perience, however, the absence of “ d u stin g” does not prove th at orthosilicate is not present in a consider
able quantity.
The Ternary System, Ca0 - A l 20 3~Si0 2.— The nature of the binary system s haying been cleared up,, we are now in a position to m ake a system atic exam ination of the ternary system . W hile one cannot predict from the nature of the binary series, ju st w h at will be the ternary relations, it is of advantage to know
th at binary compounds m ay occur as stable phases within the diagram.
In Fig. 4 the results of the binary studies are pro
jected upon- the sides of the ternary diagram. Since the compositions CaO.SiO,, 2 C a0 .S i0 2, sCaO^AljO,, CaO .A ljO , and A l203.Si02 are a t m axim um tempera
tures w ith eutectics a t A , B, C, E , F, G, I and J, boundary curves lim iting the fields for the stable
CaOAljO,
F ig . 4.
existence of the compounds (m onovariant systems) w ill extend out into the diagram at A A ', B B ', CC', E E ', F F \ G G ', I I ' and JJ'. Sim ilarly, mono
varian t system s beginning a t D, H and K will be found on the lines D D ', H H ' and K K ', Nothing definite as to the course of these boundary curves can be predicted from the binary systems.
W e also find among the natural minerals the ternary compounds anorthite (An), Meionite (Mi), Gehlenite (Gh) and Grossularite (Gr), all lime-silica-alumina minerals. Of these, pure grossularite decomposes on m elting into a m ixture of anorthite w ith (probably) gehlenite (?) and pseudo-wollastonite, and does not appear on the fusion surface of the ternary .system.
W e have prepared it b y the reaction of A 1C13 on Ca.SiO, in w ater under pressure. Meionite does not form by fusing together the pure oxides, b u t a crystal similar to gehlenite occurs, and must be included in the equi
librium diagram.
W e m ay, therefore, anticipate the appearance of a t least two fields for ternary compounds and ten fields for the binary compounds belonging to this ternary system . W ollastonite becomes stable over a small region of the ternary diagram where the fusion surface for CaSiO, falls below the inversion tempera
ture . of this compound. The /9-orthosilicate of cal
cium m ay also have a small field of stab ility depending on the slope of the orthosilicate fusion surface.
A fter this general reconnaissance of the field, it m ay be studied in detail experim entally. For re
search of this kind there are three general methods of attack.
1. One m ay m ake up the ternary m ixtures ob
serving the usual precautions' and then from the
1 S ee " B i n a r y S y s te m s ,** etc., loc- c it., p . 2 0 6 .
A N O R T H IT E
¡SLAG
f o ? \ o PROVISIONAL D IA G R A M \
A \G EO PH YSIC AL LABORATORY
o \ o
\ 1910
energy changes in the system which are registered by the thermoelement, endeavor to unravel the phase relations. This m ethod is invaluable, b u t like all methods employed in the exploration of new fields, it gains strength when em ployed in conjunction w ith other methods even though the latter be inferior in the scope of their practical application. W ith sili
cates as with alloys, it is much easier to interpret the thermal data if one knows the nature of the reacting phases.
2. The second method is to bake the different preparations for long periods of tim e at appropriate
is usually done, either b y adding the phase under investigation and finding out w hether or not it dis
solves in the saturated solution or the saturated solution is allowed to crystallize slightly in order th at the first crystals to form m ay be identified.1 The latter m ethod best serves our purpose here. A small am ount of m aterial of the desired composition is placed in a furnace and held at constant tem perature until on quenching (sudden cooling b y dropping into a bath of m ercury or water), only one kind of crystal is present, and the rest of the solution is cooled to glass before it has time to crystallize. The
PiK. 5.
temperatures in order to allow time for the reactions to become complete, and the crystals to grow to measurable size, which renders microscopic exam ina
tion much easier. W ith these two methods together, we have covered the m ajor portion of the lime-silica- alumina diagram.
3. The third method has for its object the determ i
nation of the boundaries of the various fields in which particular phases have their stable existence, i. e., of the location of. the boundary curves. In the case of ternary system s a t ordinary tem peratures, this
exact tem perature for any given m ixture m ust be determined b y a series of trials w ith sufficient va ria tion of the conditions of experim ent to rule out h yster
esis effects. Tem peratures will thus be found where the charge is w holly glass, and also where the product obtained b y quenching is w holly crystalline. The method also allows the crystals tim e to grow to meas
urable size, thus assisting the microscopic identi
fication. In other words, the m ethod develops the prim ary phase (Bodenkorper) embedded in th e glass
1 Cf. R o o z e b o o m .
2 l 6 T H E J O U R N A L O F I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y . April, ign (solution) and allows a rapid and ve ry satisfactory
m apping of the fields of sta b ility for each phase.
The method fails when the tem perature rises too high for the platinum furnace (1600°). The iridium fur
nace (i6 o o °-2 io o °) does not serve well for this method. N aturally, the method does not work so w ell where two kinds of crystals are present. T o get the most satisfactory results, i. e., fair-sized crystals of the prim ary phase, the charge should be held at a tem perature ju st below th at of com plete fusion for the composition in question.
In this w ay, the lim its of the various fields were found as indicated in Fig. 5. The dots given in the diagram do not indicate all of the concentrations examined nor the m any heat treatm ents and exam i
nations required to determ ine the relations of any one concentration. The entire ternary system thus far has necessitated the preparation of about five hundred different concentrations and fu lly five thou
sand heat treatm ents and microscopic exam inations.
W hile the location of the fields and boundary curves is now fixed w ithin ± 2l /1 per cent., the exact location of the ternary eutectics and other quintuple points other than points 13, 14, 15, 16 and 17, which are now de
termined, remains to be established by later work. T h ey have been located in this diagram b y following the general slope of the boundary curves, and b y applying the theorem of van R ijn van A lkem ade1 wherever practicable.
The phase 2CaO.Al3O3.SiOj (first observed b y Boud- ouard) is ve ry sim ilar in optical properties to the min
eral gehlenite, but its composition is quite different.
This phase appears at a m axim um tem perature in the fusion surface and is the pure typ e for this ternary system . M ixtures of the composition of typ ical gehlen
ite do not show a m axim um tem perature nor are th ey ever homogeneous. T h ey also show therm al changes corresponding to the boundary curves.
The diagram here given is one of m oderate com
plexity, bu t the crystallization curves for a n y given composition follow the general laws for such curves, set forth b y Geer2 in his paper on crystallization in three-com ponent systems.
From this diagram, it appears th at cement clinker, i. e., clinker whose essential composition falls approxi
m ately within the area v x y s, should consist chiefly of ¿CaO .SiOj and 2CaO.SiO, w ith a small am ount of 3CaO.Al2O s. W ith the exception of the small com er lyin g below the broken line connecting 3CaO.SiO.
and 3 C a0 .A l20 3 (Fig. 10), all compositions w ithin the field v x y s freeze solid a t point 16. Those lyin g w ithin th at corner of the rectangle v x y s will freeze solid at point 17 and contain free lime.
L et us n ext consider the results obtained b y pro
longed annealing of these various m ixtures a t high temperatures.
C arefully prepared and thoroughly combined charges
1 A . C. v a n R ij n v a n A lk e m a d e , Z . p h y s . C h em ., 1 1 , 28 9 (1 8 9 3 ).
2 “ C r y s ta lliz a tio n in T h r e e - C o m p o n e n t S y s te m s /* W . C. G e e r, J . P h y s . C h e m .. 8 , 257 (1 9 0 4 ). T h is is a v e r y im p o r t a n t p a p e r f o r th o s e w h o w is h to s tu d y th e r e la tio n s e x is tin g in s u c h a s y s te m a s th is . A s G e e r ’s d e d u c tio n s h a v e n o t y e t f o u n d th e ir w a y i n t o th e te x t b o o k s , w e v e n t u r e to c a ll p a r t ic u l a r a t t e n t i o n to th e p a p e r.
were enclosed in smalll platinum cups which were placed in the platinum furnace and heated for periods of from two to three weeks a t tem peratures ranging from 1300“-15 0 0 ° C., the tem perature depending on the com position of the charge. A fter this treatment, the m aterial was exam ined w ith the microscope and
I0 0 % Ca0 9 0 8 0 70 | 60 0
3CaO*AlzOj F ig . 6.
the different compounds identified. Crystallized CaO (free lime) was p ositively identified in all of the compo
sitions indicated b y dots in Fig. 6. This does not mean (as in Fig. 5) th at CaO was the only solid phase present, b u t th at it was one of the phases present.
From the general theory, free lime should not occur to the right of the line joining the compositions 3 C a0 .A l20 3 and 3 C a 0 .S i0 3, and this is found to be approxim ately true.
In Fig. 7 are shown some of the compositions in which orthosilicate is present. The small amounts
of orthosilicate which still exist in samples below the boundary curve 18 -17 can only mean that equi
librium has not become fu lly established. Even where a large excess of lime is present, the ortho
silicate does not alw ays com pletely change to tricalcic silicate.
¿CaO.SiO, in the Ternary System.— In Fig. 8 a small section (the triangle CaO 65 per cent., Si02 35 per c e n t.;
CaO 80 percent., SiO. 20 p e r c e n t.; and CaO 65 per cent., A1203 15 per cent., Si02 20 per cent.) of the ternary diagram is plotted on a much larger scale. The dots indicate some of the concentrations in which the
20%si0j zo%5iOi zoy.siot
F ig . 8.
optically negative crystals with low birefringence (tricalcic silicate) occur. The numbers enclosed in circles represent the approxim ate percentage of this phase, in the charge examined. Thus ninety means that under the microscope this composition appeared to consist of about 90 per cent. 3 C a 0 .S i0 2 and varyin g amounts of orthosilicate or lime. The estim ate is merely approximate. It will be noticed th a t while at 0.5 per cent. A 1203 the tricalcic silicate amounts to 90-95 per cent, or more of the total charge, at 5 per cent. A 1,03 it has fallen to about 70 per cent.
In the binary series a change of 1 per cent, in the concentration on either side of 3 C a 0 .S i0 2 gave a small excess of lime or of orthosilicate, while the body of the charge consisted of the w eakly birefracting, optically negative phase. W e m ay, therefore, regard the 3CaO.SiO, as a definite compound, although we have not yet been able to obtain it entirely free from pure lime, orthosilicate, or in well crystallized units.
In fact, observations of its optical character w ith the microscope become increasingly difficult as the composition approaches to hom ogeneity b y reason of the extremely fine-grained and closely interwoven character of the crystalline mass. This situation, by the way, has frequently been encountered before, both in our own Work and in the literature of sim ilar studies at low tem peratures, the purer the substance the more difficult it often becomes to secure well developed individual crystals of it. W e have no evidence that tricalcic silicate forms solid solutions
"’ith alumina, lime or orthosilicate, nor is any solid solution forrried with ferric oxide.
With the present equipm ent it has not y e t proved possible to determine the ex a ct course of the boundary curves 18-16, 18 -17, b u t in the lower portion of this field, i. c., in the neighborhood of the quintuple points
16 and 17, compositions which showed 3 C a 0 .S i0 2 as the prim ary phase were obtained.
I t is desirable to know the exact direction which these boundary curves take and the exact location of quintuple point 18. This cannot y e t be done w ith precision, b u t from theoretical considerations, point 18 should fall at or ve ry near to the tem perature corresponding to the point L in the bin ary system , i. e., about 1900° C., and quenchings from the iridium furnace indicate th at the path given is about right.
From the lower portion of the fields where we have been able to locate the boundary curves definitely, and from the general slope of the surfaces, it seems probable th a t the field for tricalcic silicate will be approxim ately th at which has been assigned to it.
I t m ay prove to be sligh tly broader a t the higher tem peratures, b u t experim ental evidence of this is at present very difficult to obtain, and the indications above mentioned do not support this view . A s far as the cem ent problem is concerned, the practical results w ill be unaffected either b y shifting the location of point 18 or broadening the upper p art of the field for 3 C a 0 .S i0 2.
In Fig. 9, the dots indicate the compositions in which 3C aO .A l2Os occurs as one of the phases. A circle instead of a dot indicates th at the identification was not absolute b u t is highly probable. The crystals in such cases were too small for positive identification b u t such optical properties as could be determined
F ig . 9.
agree w ith the properties of 3CaO .Al2Os. A s this phase is present in ve ry small am ount as a p art of the eutectic residue,' it is natu rally fine-grained and difficult to identify .positively, bu t w ith tw o of the three probable phases p ositively identified, a high degree of probability attaches to the indications of the third.
G E N E R A L O B S E R V A T I O N S O N T H E T E R N A R Y S Y S T E M .
Crystallization Curves for Cement Clinker.— If a charge of the com position 8.5 per cent. A 1203, 68.6 per cent.
CaO and 22.9 per cent. Si02 (Point M., Fig. 10) be com pletely fused and then allowed to freeze norm ally, i. e., w ithout undercooling, the following changes
