(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Biodiversity Heritage Library | Children's Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Studies in glazes. Part I. Fritted glazes"

>.**J^: 



■t-W 




^ 


f^^^^M 


tt^C5^---: 


'm^^y^ ^.'^^*p^m^. 


mas 


iHflA^^^^n^Dp^^^Vt^^fl&vl^^ *^^H(| ^^^wB 


^^rrj 




m^ 



:4ir^ 



V V 






'«^- 4^^-"^)^ V 



UNIVERSITY OF ILLINOIS 
LIBRARY 



Class 



Book 

1t^ 





W' 


Volume 


' v*/ ,-. 


f*^' 

h 

'" 










m?%S:^:; ^ 



V 4 > 



^^^■■.^''j.:^. 






4i<^' 



t: ^ •■/ 






'S^SSI^ 



^«.«. 






The person charging this material is re- 
sponsible for its return to the library from 
which it was withdrawn on or before the 
Latest Date stamped below. 

Theft, mutilation, and underlining of books 
are reasons for disciplinary action and may 
result in dismissal from the University. 

UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN ^ 




-7' 






"Y 



/ 



UNIVERSITY OF ILLINOIS BULLETIN 

Vol. 4. FEBRUARY 15, 1907 No. 12 

[Entered at Urbana, Illinois, as second-class matter] 



STUDIES FROM THE SCHOOL OF CERAMICS 

NUMBER TWO. 



STUDIES IN GLAZES. 



PART II. 



CRYSTALLINE GLAZES. 



BY 



R. C. PURDY AND J. F. KREHBIEL. 



PUBLISHED FORTNIGHTLY BY THE UNIVERSITY 



[Reprinted from ihe Tkansactions ok the Amkkican Ckramu- Society, Vol. IX. Paper 
read at St. Louis meeting, February, 1907.] 



CRYSTALLINE GLAZES. 

BY 
KUSS C I'UKDV AM) .lu.MLS F. KUKlllUEL,, 

Cliaiiipaigu, IlL 

Ciy.stalliiie j^Iazes are aiii()ii<i the most fasciuating of 
the many eerauiic problems wbicli the potter is today try- 
ing to solve. Tliej' are fascinating to the artist, because of 
the many perfect forms and the unexpected display of col- 
ors produced from the same glaze and in the same burn; 
to the cry stall ographer, because the conditions under which 
crystals are developed in a glaze are so radicalh' different 
from those under which nature has developed the larger 
part of her crystals; to the physical chemist, because of the 
insight they afford into the structure and nature of igneous 
solutions; to the ceramist because so little is known today 
of their range in composition, and the firing conditions re- 
quisite to their best development. 

Crystallographers find it exceeding difficult to explain 
the many curious phenomena of crystallization occurring 
in nature, although they have much knowledge concerning 
the fundamental facts of crystallization. The difficulties 
of the erystallographer are increased many fold for the 
ceramist. The erystallographer can readily obtain type 
specimens, determine their chemical and physical proper- 
ties, and in the end formulate a system of classification 
based upon them which enables him to identifj^ with com- 
parative readiness any inorganic crystalline substance. 
The ceramist, on the other hand, cannot collect type speci- 
mens of the crystals he produces, nor has he been able as 
yet to ascertain the conditions most suilalde for their pro- 
duction. The ceramist's problem is further complicated by 
the fact that the production of crystals is in truth but a 
minor consideration in the manufacture of salable ware, 



6 CRYSTALLINE ULAZKS. 

for above all things there must be iiiaintained perfect phy- 
sical harmony between the glaze and the body. 

Strange as it may seem, it is a fact that the ceramist 
can profit but very little, except for methods of study, by 
the mineralogist's experience. StulP has shown the futility 
of attempting to produce in a glaze matrix a condition 
that, theoretically considered, would cause the separation 
of material having the form and composition of some of 
nature's crystals. The fact of the case is, the ceramist has 
to deal with a set of conditions that require for their full 
understanding the establishment of physical and chemical 
laws which pertain to his special case. In the end, the 
science of producing crystals in glazes must be reduced to 
the same exactness as the science of metallurgy has been in 
the case of iron and steel, through the help of metallog- 
raphy. 

It is indeed a long road to a full understanding of even 
a few of the simplest and most fundamental phenomena in 
a new field. Empirical researches must be many before 
very much progress can be made toward the discovery of 
laws underlying physico-chemical phenomena. We have a 
few desultory researches in crystalline glazes, very scat- 
tered records of successful attempts in their production, 
but a resume of these several researches is today productive 
of nothing short of confusion of ideas. Published records 
of systematic empirical studies must precede attempts to 
formulate laws or deduce facts. It was with these consid- 
erations in mind that the studies here reported were under- 
taken. No attempt has been made to deduce definite laws 
from the results of these experiments, but contentment is 
found in observing and recording the facts, so that some 
day, after more of the field has been surveyed, we can cor- 
relate the results and deduce laws that will place the pro- 
duction of crystalline glazes beyond the realm of chance 
and freak. 



'Trans. Am. Cer. Soc, Vol. VI. p. 188 



i 



CKYSTA 1,1.1 NK GI.A/.KS. j 

TYI'E or (insTALIvINK CLAZKS STUDIED. 

Crystals Iiunc Ix'cii [n-oduccd in glazes lia\'iiii;, a w idc 
variety of coiiipositious, and at uearly all temperatures. 
It would be aside from the purpose of this paper to review 
all the types of glazes w hich have beeu reported as "crys- 
talline," except for the sake of establishing the fact that 
there may be a wide variation in their composition and heat 
treatment, and to demonstrate that in a thorough study, 
such as is indicated in the title, it would be impossible to 
deal with more than one type. The following formulje are 
suflQcient for these purposes : 

RIDDLE'S CRYSTALLINE GLAZE.' 
0.700 ZnO I I 1 .o75 SiO., 



. 075 K„0 ! 
0.075 Na.O 



U U Al.O, 



0.150 CaO J I 0.^25 hJJ, 



CONE 2 GLAZE.-' 

|'1.8U5 SiO 

I 
U.04o Al.,0. 



0. 


.105 


K,0 


0. 


.189 


Na,0 


0. 


.052 


BaO 


0, 


.029 


CaO 


0, 


.625 


ZnO 



[O.'ilH H.,0, 



After reviewing the published records of many experi- 
ments, it was decided to limit the present studies to a type 
of glaze that is now popular in France, and which has been 
successfully used in this country to some extent; that is, a 
siniple alkali-alkaline-earth silicate, in which the crystal 
line effect is produced by some simple crystallizers, the 
whole designed to mature at cone 10. The types of tin- 
glazes studied are given in the table at top of page 8. 



'Trans. Am. Cer. Soc, Vol. VIII, p. 347. 
'Sprechsaal. 1903. 



CRYSTALLINE GLAZES. 



TABLE I. 
Crystalline Glaze P^ormulae. 



Name 


K,,0 


1 ZnO 


TiO., 


1 SiO., 


Ox. Ratio 


Bries glaze No. 1 


.300 


.700 





2.000 


4.000 






Bries glaze No. 2 


.300 


.700 


.300 


1.700 


4 . 000 






Heath's glaze 


.300 


.700 


.500 


2.000 


5 000 






Sevres glaze No. 1 


.333 


.666 




2.000 
1.94" 


4 000 






Sevres glaze No. 2 


.166 


.833 


3 870 






Sevres glaze No. 3 


.333 


.666 


.341 


1.666 


4 016 






w, ^ A ( 85% No. 1 ) 
Sevres blend A | ^^^^ ^^ ^ | 


.308 


.691 


1.991 


3.983 



Sevres blend B 



85% No. 3 
15% No. 2 



308 ! 691 



.290 



1.719 



3.998 



StuU's glaze, according to Gates..! .305 I .695 I .297 I 1.690 I 3.970 



SCOPE OF INVESTIGATION. 

Before atteiDpting to outline a series of studies on 
fritted glazes, it is essential to consider the conditions that 
must prevail, in order to have a crystallizing conii)ound sep- 
arating out from a molten matrix. It is especially essen- 
tial that we should understand the fundamental difference 
between a clear non-crystalline glaze, and a glaze in wliicli 
clusters of crystals appear throughout the whole mass. 

Primarily a glaze may be considered as a glass. Win- 
dow glass, for instance, is a perfectlj^ transparent and hom- 
ogeneous compound. It has been proven beyond a doubt, 
however, that window glass has a tendency to crystallize. 
Crystals of wollastonite have been developed in one in- 
stance, and crystals of quartz in another, showing that even 
in so simple a compound as glass, consisting in the main of 
calcium oxide, soda, potash, silica and in some cases lead, 
there is what may be styled latent crystallization. Curious 
indeed is the fact that the potter has learned by empirical 
trials that the best crystalline effects are produced in a 
glassy matrix, which in composition is very much like the 



CRYSTALLINE GLAZKS. 9 

mixture whicli the <>Iass manufacturer has learned also by 
similar empirical trials will produce the best glass. 

The belief that ordinary glass has a tendency to crys- 
tallize is strongly supported by the following quotation: 

'"If one considers that the contents of silica in commercial glass 
is very variable, and that a higher temperature more of this acid can 
be taken up than can be retained by the glass as this temperature falls; 
that further, feldspathic minerals may be taken up and again separated 
by a molten glass; finally, also sulphates, phosphates, borates, metallic 
oxides may enter into the constitution of glass, it would appear that 
these facts allow of no other explanation than that commercial glasses 
are quickly solidified solutions, not only of different silicates in each 
other, but even solutions of basic oxides or silica and other salts and 
metals, in molten silicates. Whether in all glasses some fundamental 
silicates, a compound say of Na,.0, CaO, SiO^, or Na^O, CaO, 4 SiOo, are 
to be taken as the solvent or not, must remain a question so long as 
nothing more definite is known than at present, yet many of the ob- 
servations appear to speak for such an assumption." 

Or the following :^ 

"A devitrified glass is a heterogeneous mixture, and while support 
is lent to the theory that the original undevitrified material is likewise 
a mixture, it cannot be accepted as conclusive proof. One cannot con- 
ceive the formation of such a heterogeneous mixture when a glass is 
devitrified by being kept for a long time at a high temperature, on the 
assumption that one is dealing with a mixture, more or less homo- 
geneous, and that an agglomeration or segregation of the constituents 
takes place, giving rise to crystalline forms. But on the other hand, 
it is also conceivable that there is a real solution, as described by 
Benrath, supersaturated at a high temperature with some definite 
silicate or other compound which gradually crystallizes out. Hence the 
facts of devitrification may lend support to either theory." 

Seydholt, Jackson and Ki«h, and others have etched 
iilass with concentrated hydrofluoric acid, developing on 
the ((niiiiioii sheet glass, crystals that are of various forms, 
some *'star shaped," some as "rods with crystalline radia- 
tion on both sides," and in tlie lead glasses, "superposed 
rings" (hat certainly cannot be due to accidental in-egular- 
ities in texture of the glasses. 

'Benrath — "Die Glasfabrikation," Jour. Soc. of Cheni., Ind., 1901. 
2Jotir. Soc. of Chem. Ind. 1901. 



10 CRYSTALLINE GLAZES. 

Daubree has shown that bv digestiou in a sealed tube 
with super-heated steam at a high temperature, ordinary 
glass is decomposed into quartz, wollastonite and alkaline 
silicate. 

True glasses can be made having a wide variation in 
composition, yet it is a fact that to produce the clearest and 
most weather-resisting glass, the manufacturers have 
learned that they cannot depai't \er\ much from a given 
composition. Oxygen ratio, or ratio of bases to acids on 
the basis of the oxygen content of each, is a factor of prime 
importance. Starting with a glass in which crystallization 
is not manifested under ordinary conditions, the points 
>\liicli tlie ceramist must strive to control arc: 

1st. The development of these latent crystallizing 
tendencies. 

2nd. To cause the development of crystals in a glassy 
matrix the physical properties of which, ])rinci])ally ex- 
pansion and contraction, agree with those of the clay body 
upon which it is fused, so that both glaze and body will re- 
sist all tendency to rupture. A glaze badly crazed, or a 
(lunted l)ody, makes the ware worthless, ii-respective of the 
character of crystals that may have been developed. 

As we know practically nothing about the crystalliz- 
ing value of the several oxides, or combinations of oxides 
in different proportions, it is very plain why the develop- 
ment of crystals in a glaze is reputed as "freakish," and 
Avhy but very little importance has been attached to the 
composition of the amorphous portion of the glaze matrix. 
Following first one plan of investigation and then another, 
promising results have been obtained with such varying 
ratios that experimenters have come to believe that the 
oxygen ratio or ratio of base to acid is of no consequence in 
crystalline glaze production, that CaO may or may not be 
]U'esent, and that alkaline earth may be used in any equiva- 
lent quantities compatible with the full maturity of the 
glaze under practical heat treatment. For the reason that 



I 



CRYSTALLINE GLAZES. 11 

such an unlimited range of choice in materials and ratios 
is not possible in any other type of glaze or even glass, if 
for none other, it was believed that before the problem of 
the production of crystals in a glaze could be studied with 
any degree of scientific understanding, the true facts con- 
cerning possible range in composition should be estab- 
lished. To this end, in the present studies, all other factors 
such as variation in heat treatment and body composition 
were eliminated by maintaining them constant throughout. 

The table on pages 12-13 gives a synopsis of the plan 
of investigation of the one factor, viz : Effect of variation 
in composition. 

Crystallizers. Nearly every oxide, and many of the 
elements are found in nature in a crystalline form. An im- 
mense variety of complex silicates are also found in nature 
as crystals. In ceramic art, the "gold stone," "tiger eye," 
"aventurine," etc., etc., are crystalline effects produced by 
supersaturation of a glaze by iron, chromium, etc., in ex- 
tremely fine particles intimately mixed. Even in the cruder 
glaze, such as the Bristol and ordinary matt majolica 
glazes, the opacity in the one and dimness in the other are 
attributed to crystallization of some element. Having 
known these facts for some time, ceramists are still unfa- 
miliar with the crystallizing value of the various elements, 
except in the case of two, zinc oxide and titanium oxide, 
which give the most pronounced effects. There are cases 
in which silica and others, lime, for instance, seem to be- 
have as crystallizers. It is certain that opalescent effects 
which are crystalline in nature, are produced by supersat- 
uration with either lime or silica, and that they occur in 
nature in a variety of crystalline forms; yet because they 
have not been known to produce crystals in a glaze, they are 
not classed as crystallizers. On the basis, then, of the in- 
tensity of their crystallizing tendency in a glaze, the list of 
crystallizers practically consists of but three: — 

(1) Zinc oxide. 

(2) Titanium oxide. 

(3) Manganese oxide. 

p. & K.— «. 



12 



CRYSTALLINE GLAZES. 



TABLE II. 
Synopsis of General Plan. 



Group number 


Limits of ZnO 


Limits ol MnU 


Limits ol K.,0 


Limits ot NajO 


[ 


1.0— .1 




•0— .9 




II 


1.0— .1 




0— .9 




III 


1.0— .1 






0— .9 


IV 


1.0— .1 






0— .9 


V 


1.0— .1 




0— .45 


0— .45 


VI 


1.0— .1 




0— .45 


0— .45 


VII 


1.0— .1 




0— .G 


0— .3 


VIII 


1.0— .1 




0— .G 


0— .3 


IX 


1.0— .1 




0— .3 


0— .6 


X 


l.O-.l 




0— .3 


0— .6 


XI 




l.O-.l 


0— .9 




XII 




1.0— .1 


0— .9 




XIII 




l.O-.l 




0— .9 


XIV 




1.0— .1 




a— . 9 


XV 




1.0— .1 


0— .45 


0— .45 


XVI 




l.O-.l 


0— .45 


0— .45 


XVII 


1.0— .1 




0— .45 


0— .45 


XVIII 


l.O-.l 




0— .45 


0—. 45 


XIX 


1.0— .1 




0— .45 


0— .45 


XX 


l.O-.l 




0— .45 


0— .45 


XXI 


l.O-.l 




0— .45 


0— .45 


XXII 


l.O-.l 




0— .45 


0— .45 


XXIII 


l.O-.l 




0— .45 


0— .45 


XXIV 


l.O-.l 




0— .45 


0— .45 


XXV 


l.O-.l 




0— .45 


0— .45 


XXVI 


l.O-.l 




0— .45 


0— .45 


XXVII 


l.O-.l 




0— .45 


0— .45 


XXVIII 


1.0— .1 




0— .45 


0— .45 


XXIX 


l.O-.l 




0— .45 


0— .45 


xxxx 


l.O-.l 




0— .45 


0— .45 


XXXI 


1.0— .1 




0— .45 


0— .45 


XXXII 


1.0— .1 




0— .45 


0— .45 


XXXIII 


1.0— .1 




0— .9 




XXXIV 


l.O-.l 







0— .9 


XXXV 


l.O-.l 




0— .45 


0— .45 


XXXVI 




l.O-.l 


0— .9 




XXXVII 




l.O-.l 




0— .9 


XXXVIII 




1.0— .1 


0— .45 


0— .45 



CRYSTALLINE GLAZES. 



13 



TAHLE II. 
Syuopsis (if (icncral Plan 



Limits of TiOj Limits of SiO, 


Limits Ox. ratios 


Limits of uifmbers 


Crystallizing 
Agents 




1 . 5—1 . 


1—2 


1— 110 


ZnO 






; 1 . 0—2 . 

1 


2—4 


111— 220 


ZnO 






1 .5—1.0 


1—2 


221— 330 


ZnO 






1.0—2.0 


2—4 


331— 440 


ZnO 






j . 5—1 . 


1—2 


441— 550 


ZnO 






1 1.0—2.0 


2—4 


551— 060 


ZnO 






j . 5—1 . 


1-2 


6G1— 770 


ZnO 






1 1 . 0—2 . 


2—4 


771— 880 


ZnO 






.5—1.0 


1—2 


881— 990 


ZnO 






] 1 . 0—2 . 


2-4 


991—1100 


ZnO 






! ..5—1.0 


1—2 


1101—1210 


MnO 






1.0—2.0 


2-4 


1211—1320 


MnO 






. 5—1 . 


1-2 


1321—1430 


MnO 






i 1 . 0—2 . 


2—4 


1431—1540 


MnO 






1 .5—1.0 


1—2 


1541— 1G50 


MnO 






' 1.0—2.0 


2—4 


1G51— 17G0 


MnO 




.5 


— 


.0— .5 


1 


1761—1870 


ZnO 


TIO, 


.G 


— 


.0— .G 


1.2 


1871—1980 


ZnO 


TiO, 


7 


— 


.0— .7 


1.4 


1981—2090 


ZnO 


TiO, 


.8 


-0 


.0— .8 


l.G 


2091—2200 


ZnO 


TiO., 


.f> 


— 


.0— .9 


1.8 


2201—2310 


ZnO 


TiO, 


1.0 


-0 1 .0—1.0 


2.0 


2311—2420 


ZnO 


TiO., 


1.0 


— 1 .2—1.2 


2.4 


2421—2530 


ZnO 


TiO.. 


1.0 


— 1 .4—1.4 


2.8 


2531—2640 


ZnO 


TiO.. 


1.0- 


-0 ! .6—1.6 


3.2 


2641—2750 


ZnO 


TiO... 


1.0 


— i . 8—1 . 8 


3.G 1 


2751—2860 


ZnO 


TiO., 


1.0 


— : 1.0—2.0 


4.0 


2861—2970 


ZnO 


TiO, 


1.0 


— 1 1.2—2.2 


4.4 


2972—3076 


ZnO 


TiO., 


1.0 


— : 1.4—2.4 


4.8 


3082—3186 


ZnO 


TiO, 


1.0 


— ! 1.6— 2. G 


5.2 1 


3192—3296 


ZnO 


TiO., 


1.0 


— 1 1.8—2.8 


5.6 


3302—3406 


ZnO 


TiO, 


1.0 


-0 ' 2.0—3.0 


G.O 1 


3412—3512 


ZnO 


TiO., 




1 2.0—3.0 


4—6 




ZnO 






1 2 . 0—3 . 


4—6 




ZnO 






2.0—3.0 


4—6 




ZnO 






2 . 0—3 . 


4—6 




MnO 






2 . 0—3 . 


4—6 




MnO 






•} n — 2 


4-6 1 




MnO 















14 CRYSTALLINE GLAZES. 

titer glaze ingredients. With the other ingredients, 
there is not much choice. Alkalies, alkaline earths and 
acids must be employed. Potash and soda are the only 
alkalies practically available for this purpose. Calcium, 
magnesium and barium are the alkaline earths that have 
been experimented with sufficiently to know their relative 
value as glaze ingredients. Calcium seems to be the best 
of the alkaline earth fluxes for glaze purposes. Practical 
limitations, however, prevent the study of the influence of 
lime at this time. 

Points considered. The whole burden of this investi- 
gation is directed toward the settlement of the following 
questions : 

1. The effect of the alkalies, considered singly and 
together in various proportions, on the development of 
crystals. 

2. The proportional equivalent content of ZnO and 
conducive to the best development of crystals. 

3. The relative crystallizing tendency of zinc, man- 
ganese and titanic acid. 

4. The character and shape of the crystals induced 
by manganese. 

5. The limits of oxygen ratio permissable in crystal- 
line glazes. 

CONSTRUCTION OF THE GROUPS. 

The general scheme of construction of the various 
groups is shown in Table III. 



CRY8TALI.INK (;I-AZES. 15 



TABLE III. 



Showing the relation of glaze formulae to the serial numbers of the 

glazes. 

GROUP I. Oxygen Ratios 1:1—1:2. 



I ZnO K,0 I 



•o C 

C K 
M S 









1.00:0.0 
0.9 :0.1 
0.8 :0.2 
0.7 :0.3 
0.6 :0.4 
0.5 :0.5 
0.4 :0.6 
0.3 :0.7 
0.2 :0.8 
0.1 :0.9 pon ^ \ '\ \ \~W\ j \ TllO" 



1 1 


11 


21 


31 


41 


51 


61 


71 


81 


91 


101 


1 ^ 












62 








102 


1 '^ 












63 








103 


1 '^ 


14 


24 


34 


44 


54 


64 


74 


84 


94 


104 


1 ^ 












65 








105 


1 ^ 












66 






106 


1 ' 












67 








107 


1 8 












68 








108 


1 9 












69 








109 



'0.5010. 5510. 60i0. 6510. 7010. 7510. 8010. 8510. 9010. 951 1.00 



Equivalents of Silica. 

There are no i:;roups in which titanium is the sole acid, 
but there are individual glazes in a number of the groups 
having an oxygen ratio equal to 2 or less, in which Ti02 is 
the only acid present. 

I'UKPARATION OF THK GLAZES. ^ 

All glazes were fritted at least once before application 
to the ware. A few of tlie glazes had to be renielted two, 
and in a few cases Three times, before all of the materials 
were in (•om])lete solution. The case was not uncommon 
wlien some of the more infusible material remained in sus- 
pension, altliougli the whoh' was fluid enough to tlow freely 
from the crucible. 

Where the alkali <-ontcut is high and the oxygen ratio- 
low, the fritts are (piite s<(luble in water. In fact, in many 
cases the fritts were too soluble to allow wet grinding. It 
seems that some of the soluble salts in the more soluble 



'The open spaces in the table are supposed to be filled in in arith- 
metical sequence with those indicated. 

p. K K— 3. 



16 CRYSTALLINE GLAZES. 

fritts are deliquescent to such an extent that in grinding, 
the powder absorbed moisture from the atmosphere, until 
they adhered to the pebbles and lining of the mills, making 
their preparation by either wet or dry grinding impossible. 
The soluble glazes are designated in the curves given 
later, as occurring in the area bounded by solid black lines. 
These were not made up or tested. 

BLENDING CALCULATIONS. 

The glazes of each group were obtained by blending 
between extremes. Since in each group there were two 
variable factors, i. e., variations in RO composition, and 
variation in oxjgen ratio with each change in RO, there 
are four extremes. 

Referring to the chart given above showing the varia- 
tion in composition and glaze numbering, the intermediate 
glazes are obtained by blending glazes 1^ 20, 101, and 110. 
The blending was first made between 1 and 101 and also 
20 and 110, i. e., on the basis of oxygen ratio, by the fol- 
lowing formula : 

a(u+l) b(n — 1) ' 

c = 



N — 1 N — I 

Where a =r combining weight of glaze No. 1 
b = combining weight of glaze No. 101 
c = combining weight of requireil glaze 
N ^= number of variations in oxygen ratio 
n =^ position of required glaze to the right of the left 
extreme 

After having made the blends of the extremes on the 

l)asis of oxygen ratio as shown in the top and bottom lines 

'of the chart for group 1, the corresponding glazes in each 

line are re-blended on the basis of the RO variation by the 

same formula as given above, except that the notations in 

the second case are as follows : 

a = combining weight of top extreme shown in chart 
b = combining weight of bottom extreme 
c =^ combining weight of required glaze 
N = number of variations in RO content 
n := position of glaze in vertical column 



CRYSTALIilNE ULAZES. 17 

Eaamplc of horizontal hlcnding. Given the combiniug 
weight of No. 1 as 111, and 101 as 111 ; required blending 
mixture for glaze No. 41, where 

a = 111, b = 141, N == 11, u = 5 

111 X 6 141 X 4 

C — 1 =0(i.6 + 66.4 = 123 

10 10 

We would therefore take the proportion of 66.6 grams 
of No. 1 and 56.4 grams of No. 101 to make No. 41 which 
has a combining weight of 123. 

EaampJe of vertical blending. Given the combining 
weight of No. 1=111, and No. 7=118.8. To find the blend 
for No. 3. Here a^lll, b=:=llS.8, N=T, n=3. 

Ill X 4 118.8 X 2 
C = \ =74 + 39 6 = 113.6 

6 6 

Therefore the proportion of extreme 1 and 7 to make 
3 is 74 grams of 1. and 39.6 grams of 7. The amount of 
glaze necessary to cover the trial vase was found to be 16 
grams, and if the group is complete, it requires 110x16 or a 
total of 1760 grams of glaze for each group. 

BODY USED. 

The body used in these experiments was not chosen 
because of any special property other than vitreousness at 
cone 10. It possessed a stony fracture, and was opaque, 
even when thin. The color generally was pale blue, due to 
reduction in the biscuit burn. Occasionally, however, it 
had a pale yellowish tint. No notice was taken of the color 
of the body, i. e., no discrimination was made in the use of 
the biscuit, on account of its color. 

The composition of the body was as follows : 

Florida ball clay 20 

Tennessee ball day' 11 

Georgia kaolin 21 

Flint 13 

Spar (Brandywine) 35 

'This clay had theapproxlmate formula, O.2K2O 1.0Al2O.j 3.9Si02. 



18 CRYSTALIvlNK GLAZES. 

The body was bluuged in the form of a slip in ball 
mills having a capacity of about 4 gallons. The mills were 
run for about 30 minutes at 45 revolutions per minute, this 
amount of grinding being sufficient to slake the clay, and 
thoroughly blend all the body ingredients. 

After grinding, the body was sieved through a 40-mesh 
screen and stored in large stoneware jars. No attention 
was paid to the length of time that the slip remained in 
the jars, hence some portions of the mixture as used were 
aged longer than others, obtaining as a consequence slight 
variation in character of biscuit. 

SHAPE OF TRIAL PIECE. 

For the development of crystals, a heavy coating of 
glaze must be applied to the biscuit. As a consequence of 
this application, and of the excessive fluidity of the glazes 
when at the state of fusion required to permit the best de- 
velopment of crystals, provision must be made for the large 
amount that naturally flows off the vertical sides of the 
vases. Further, as was shown by the saggar bottom ex- 
hibited by Mr. Gates at the Birmingham meeting of the 
American Ceramic Society, crystals seem to develop best 
on flat surfaces where the glaze accumulates in thick layers 
and flowage is impossible. It was tliought imperative, 
therefore, that both a vertical and a flat surface should be 
provided. These conditions were attained in the combina- 
tion of vase, ring stilt, and saucer. 

This combination of vase, ring and saucer was 4i/i>" 
over all in height and 3'' in diameter. The vases were cast 
in a two-part mould, the ring pressed in a mould similar 
to handle moulds, and the saucer batted out in a one-piece 
mould. The shape of the vase produced is shown in the 
following plate: 



CRYSTALLINE GLAZES 10 

TRAN5.AM.CER.S0C VOL. IX PURDYAND KREHBIEL PLATE I 




Mold employed in making the Vases used as Trial Pieces in tlie 
Crystalline Glaze Experiments. 

r.ISCUIT FlUIXfJ. 

Aside from the tirst few biirus, wliicli of uece.ssitv were 
all biscuit, no separate biscuit burns were made. After a 
stock of biscuit liad l)een accuiinilated, one saiijicr of biscuit 
in each <^lost burn was found sufficient to furnish all that 
was required. Six saggers were placed in the kiln, one for 
the cone and pyrometer, one and sometimes two saggers for 
biscuit, the remaining three or four l)eing reserved for the 
glaze. In the biscuit the vases were placed touching one 
another. The biscuit rings were generally pla<(Ml in the 
cone sagger on top of a bung exposed to the full play of the 
flames, and the saucers were placed wherever opportunity 
offered. 



20' CRYSTALLINE GLAZES. 

GLOST FIRING. 

^iaggers. The saggers used in these experiments were 
made with extra thick walls to prevent sudden fluctuation 
in temperature at times when the fire doors were opened, as 
well as to insure the slow steady cooling so necessary to 
the best production of crystals. Each sagger is large 
enough to hold eleven or twelve vases. The sides of the 
glaze saggers were washed with a mixture of alkaline flux, 
zinc and flint. The bottom of the sagger was coated with a 
mixture of equal parts of bone ash and flint.^ 

Placing of ivare in lihi. The saggers were placed in 
the kiln in two bungs, the side of the front bung being next 
to the wicket. Tlie bungs were three saggers high, nearly 
filling tlie space in the firing chamber. The cone or trial 
sagger was placed in the middle of the front bung. Inas- 
much as the top sagger of the front bung seemed to be in 
the place least favorable for slow cooling, it was used ex- 
clusively for biscuit. The bottom sagjger in the front and 
all the saggers in the rear bung were used for glaze. 

Number of trials in hum. With one sagger for the 
cone, and one for biscuit, it was possible to burn only 44 to 
48 glost trials in a single burn. 

Heat Treatment. The cones placed in the trial sagger 
were Nos. 5, 7, 8, 9 and 10. In each sagger of glost, there 
was placed a shortened cone 10. There was developed a 
difference of two cones between the inside and outside of 
the sagger, due probably to the difference in rate of cooling. 
It was found that when the heat Avas raised, as shown in 
tlie firing curves, until cone 8 in the trial sagger was fused, 
cone 10 would be down in all of the glost saggers. This 
regularity in heat distribution was obtained in practically 
every burn. 

Firinf/. Time was not taken in the beginning to thor- 
oughly study the conditions favorable to the best develop- 



'This mixture lias been used by Professor Bimis to advantage in pre- 
venting tlie slagging of the sagger bottoms by flowing glaze. 



CRYSTALLINE (JLAZES. 21 

ment of crystals. Tlic iiu'tluKl of firing in the first few 
burns ditYered therefore from tlie nietliod used in the later 
burns, as shown in the following summary: 

First far Burns. 

Length of time expendetl in firing and cooling, 18 

to 22 hours. 
Fuel used — soft coal. 
Thickness of bed — thin. 
Cooled b}' slow firing. 
Results — irregularity. 

Later Burns. 

Length of time expended in burning and cooling 

to 200" C, 15 to 16 hours. 
Fuel used — coke. 
Thickness of bed — 3 to 4 inches. 
Cooled by sealing kiln, and pulling fire. 
Results — Perfect control. 

Time — Temperature Treatment. The accompanying 
curyes, Plate II, show the time-temperature treatment, 
which was obtained in practically eyery burn. Consider- 
able importance is justly attached to the time-temperature 
treatment of crystalline glazes. In a preliminary burn, 
good crystals were developed in glazes that had been suc- 
cessfully used on a commercial scale, and the time-tempera- 
ture treatment plotted. It was found, however, that this 
curve could not be followed strictly in the later burns, and 
hence there were considerable irregularities in this regard 
from burn to l)ni'n. Fui-ther work with the moi-e promising 
glazes is planned to determine the time-temperature treat- 
ment best suited to the deyelopment of crystals in the 
University of Illinois trial kiln. (For di-awing, see page 
25, r. of i. IJulletiii Xo. 12. 



22 



CRYSTALLINE GLAZES. 



Ul 

a 

_i 

d) 

X 
UJ 


/ 


^ 






















1 
























/ 
























o: 

X. 


L--' 


#• 
f 






















< i 


/ 
























a 8f 
a T 




























\ 


























\ 


























\ 


t 


























K 


























A 


























N 


k 




















X 






1 


\»^ 




























'%, 


*^ 












O 

in 








Cij 


^ 


--%■ 
























%. 


^ 


Ss 


k- 








< 

iri 
z 
< 
















*^ 


.^ ' 


^; 




























*v. 


c 
c 


I 1 


> c 

) c 
c 


5 C 

3 C 
> 


i I 


> c 

i c 


> c 

> c 
< 


) c 
) c 
) u: 


3 C 


> c 


o 




8 



in 


rC 


!t 


•^ 


- -) 


fl 


-0 


"^ 


I 


bt 














o / 


f— H 








o 


Z 


U 


c 


_, 


0>D 


ri 


(U 


a 


li. 




OJQ 








Z 


(^ 


n 


"( 1 


^P 


o 


< 


<B 


CE 


08 
P5 



SBBtisaa 3avb)oij.Ki3^ ni 3anj.va3diAi3j. onuju 



CRYSTALLINE GLAZES. 28 

Oue fact in connection with the time-temperature 
treatment was developed in these studies, which permitted 
the shortening- of the total time required for each burn, 
viz., that the rate of raisino- heat did not seem to affect the 
character of crystallization, unless the heat raised so rap- 
idly as not to permit complete fusion of the entire glaze 
mass. Inasmuch as each glaze used in the various blends 
had been rendered homogeneous by fritting, and the thick 
saggers and heavy brick work of the kiln necessarily made 
the raising of heat a comi)aratively slow process, it may be 
said that under these conditions the heat may be raised as 
rapidly as possible. 

The cooling. The question of cooling seems the all- 
important one. Slow cooling was accomplished by two 
methods, first, by continuing the firing with a bare spot on 
the grate, and, second, by drawing the fire and sealing the 
kiln completely. The second method proved to be the bet- 
ter, for the results by the first method were far from con- 
sistent from burn to burn. 

DifficiiJtics encountered. First, each glaze was worked 
up into a paste with gum of tragacanth by spatula on a 
glass slab, without having previously mixed the blended 
portions. This method of treatment gives good results 
when sufficient pains are taken, but it sometimes happened 
that a careless operator would not thoroughly blend the 
several portions of the glaze, and conse(iuently the more 
fusible portions would be unevenly distriljutcd, giving rise 
to a warty appearance of the glaze when burned. 

Second . The base of the vase was so small tiiat if not 
setting squarely on the ring, or if the ring was not well 
made, the vase would often tund)le over. 

Third : Because each vase was placed on a ring and 
then the whole placed on a saucer, it was found necessary 
to set each piece .separately. The kiln walls being IS inches 
in thickness, the four burns per week kept the kiln so hot 
that it made the placing of these 48 trial pieces a very 
difficult task. It often hapiMMied, therefore, that the vases 



24 CRYSTALLINE GLAZES. 

were not set squarely in the saucers, and consequently were 
easily jarred over wlieu the sagger was covered either by 
another sagger, or by a hiller. 

ZINC OXIDE AS A CRYSTALLIZING AGENT. 

From researches in the past, it has been demonstrated 
that zinc oxide is, in small quantities, not only as active a 
flux as lime, but also that as a flux, it is more favorable to 
the development of crystallization. Seger's cone 4\ for ex- 
ample, is rendered considerably more fusible by the addi- 
tion of only a "trace" of zinc oxide. Bristol glazes, which 
are similar in composition to cone 4, have zinc oxide sub- 
stituted in part for lime. In amounts less than 0.3 equiva- 
lents it acts as a flux, producing clear glazes, but any addi- 
tion over this amount produces incipient crystallization 
as indicated by the resulting opacity. Whether or not it 
is a zinc compound that crystallizes, is not known. It is 
sufficient for the purpose at hand to note that incipient 
crystallization is a consequence of the presence of zinc 
oxide in excess of the amount that produces maximum fusi- 
bility in the glaze mixtures. It is therefore obvious that 
instead of the alkaline earth CaO, as in ordinary glass, it 
would be more conducive to the development of cryst<ils to 
use ZnO. 

Noting that the successful fritted glazes as given in 
Table T contain no ('aO, and not knowing its influence on 
the development of crystals, it was deemed advisable not 
to complicate our preliminary survey of this field by incor- 
porating it in our glaze compositions. After having deter- 
mined the alkali-zinc-silica mixture that has the greatest 
crystallizing tendency, then, it was thought, would be 
earl^' enough to begin to study the influence of lime. In 
Groups I to VI inclusive, therefore, the study is confined 
to alkali-zinc-silica combinations. 



iTrans. Amer. Cer. Soc. VoL YIII, p 163. 



CRYSTALLINE GLAZES. 



26 



Subdivisions into Series 
on Basis of SiO, Molec- 
ular Content 



Series No. 



1 

2 
3 
4 
5 
6 
7 
8 
9 
10 
11 



SiO, 
Content 



O.BO 
55 
0.60 
0.66 
0.70 
0.75 
0.80 
0.85 
0.90 
0.95 
1.00 



Variation in RO 


Content 


within each Series 


No. 


ZnO 


K,0 


a 


1.0 


0.0 


b 


0.9 


0.1 


c 


0.8 


0.2 


d 


0.7 


0.3 


e 


0.6 


0.4 


f 


0.6 


05 


g 


0.4 


0.6 


h 


0.3 


0.7 


i 


0.2 


0.8 


J 


0.1 


0.9 

i 



Oxygen Ratio of 
Each Seiies 



Series No. 



1 

2 
3 
4 
5 
6 
7 
8 
9 
10 
11 



O. R. 



1.0 
1.1 
].2 
1.3 
1.4 
1.5 
1.6 
1.7 
1.8 
1.9 
2.0 



Glaze Number 
in Each Series 



1- 10 
11- 20 

21- 30 
31- 40 
41- 50 
51- 60 
61- 70 
71- 80 
81- 90 
91-100 
101-110 



Fiisibilitu. When the RO contains less than 0.4 KoO, 
irrespective of acid content, the glazes are infusible at 
cone 10. 

t^oluhilitij. With more than 0.6 K2O, in the case of 
oxjgen ratio 1 : 1 and 0.7 K^O, in the case of oxygen ratio 
1 : 2, the glazes are soluble. 

Luster. All the matured glazes in this group have 
bright glossy luster. No matts are developed. Glaze No, 7 
having a molecular formula of 0.4 K^O, 0.6 ZnO, 0.5 SiO? 
approached the nearest to a matt finish. 

Crazing and shivering. All the matured glazes were 
crazed in such extremely fine meshes, irrespective of acid 
content, as to render this defect almost unnoticeable to the 
naked eye. 

Crystallization. In glazes having an RO of 

r 0.4 — 0.6 ZnO 
1 0.6 — 0.5 K2O 

irrespective of acid content, minute crystals develop, es- 
pecially on horizontal surfaces where the glazes accumu- 
late in thick masses. The.'je minute needle-like crystals are 
so interlaced and so thickly buuclied together, that in many 
eases thev lose their individualitv. 



26 CRYSTALLINE GLAZES. 

None of the glazes of this group were fired in a cheek 
burn, but judging from our experience with the titanium 
crystalline glazes, it is thought that under certain firing 
conditions (not now known) there would be developed on 
the vertical surface of the vase the same thick and irregular 
mixture of minute needle-like crystals that in our own 
burns appeared only in the saucers, producing either a 
matt or aventurine effect. It is, however, evident from 
the glossiness of the glazes on the vases, and the consistent 
absence of crystals on the upright faces throughout the 
whole series, that the conditions of firing would be too 
exacting to warrant the use of any of the glazes of this 
group for matt, crystalline or aventurine effects. 

(fKoup n 



Sub-division into Series 


Variation in RO 


Content 


Oxygen 


Ratio ot 


Glaze Numbers 


on Basis of SiO„ Molec- 




cries 


Each Series i 


in Each Series 


ular Content 












Series No. 


SiO„ 
Content 


No. 


ZnO 


K.O 


Series No. 


O R. 


a-j 


1 


1.0 


a 


1.0 


0.0 


1 


2.0 


111-120 


2 


1.1 


b 


0.9 


O.l 


2 


2.2 


121-130 


3 


1.2 


c 


0.8 


0.2 


3 


2.4 


131-140 


4 


1.3 


d 


('.7 


0..S 


4 


2.6 


141-160 


6 


1.4 


e 


0.6 


04 


5 


2.8 


151-160 


6 


1.5 


f 


06 


0.6 


6 


3.0 


161-170 


7 


1.6 


E 


0.4 


0.6 


7 


3.2 


171-180 


8 


1.7 


li 


0.3 


0.7 


8 


34 


181-190 


9 


1.8 


i 


0.2 


0.8 


9 


3.6 


191-200 


10 


1 9 


j 


0.1 


09 


10 


3.8 


201-210 


U 


2.0 








11 


4.0 


211-220 



Fusihilitij. When the RO contains less than 0.2 KgO, 
irrespective of the acid content, the glazes are infusible. 
With acid content of 1.0—1.2 SiOo (oxygen ratio 2.0 to 
2.4) and RO composition of 0.2 KoO, 0.8 ZnO, the glazes 
while glossy are not sufficiently fusible to flow smoothly. 

Soluhilitii. The only glazes of this group that are so 
soluble as to render them impractical on a commercial 
scale are those in which the RO contains 0.9 KgO. 



CBY6TALL1NK (iLAZKB. 27 

J.iiiitvr. All the matured j;lazes had l)ii<iht j;lossy 
luster, except No. 116, which lias the molecular formula 

Crazing and shireriug. All the matured glazes in 
this group crazed. With oxygen ratio of 1.2 the crazing 
was in fine meshes, while with oxygen ratio of 4 the crazing 
was relatively coarse. 

CryHtaUizat'wu. As shown in Plate No. Ill, the 
glazes in which crystals are developcnl have an RO range of 



0.3 — 6 K2O 
0.7 — 0.6 ZnO 



irrespective of acid content. 

The character and mode of occurrence of crystals are 
tlie same in this group as in Group I, except in the case of 
glazes No. 155 and 184, having respectively the molecular 
formula 

0.6 ZuO ) , . Q.^ , , 0.7 ZnO ) , _ rjir» 

0.4 K2O } 1-^ ^^^^ "''^ 0.3 K2O P-^ ^'^' 

in which there is a slight tendency to segregate into radial 
clusters. Most of the crystals on these two vases are indi- 
vidual bars scattered without definite direction throughout 
tlie glaze matrix. 

Conclusions froin Croups 1 (ind II. i'ntin the fact 
that the glazes of these two grouj)s were burned in five 
separate burns, in allotments of about 40 glazes to each 
burn, and gave consistent results throughout each of the 
two groups, it would seem that the following statements 
were substantiated beyond a doubt: 

First. While crystals were developed at all oxygen 
ratios between 1 and 4, the most extensive devclojuiieiit and 
the largest crystals cxMurred with oxygen ratios ranging 
from 2.2 to 3.0 and with RO content of 

f 0.3to()6K2O 
1 0.7 to O.B ZuC) 

Within this oxygen ratio range, the glazes on the vertical 
fa<es of the vases are crystalline to some extent, while at 



28 



CRYSTALLINE GLAZES. 




i3xmd oa Ni aaixo wnissvxod ox aaixo oniz jo soixva 



CRYSTALLINE G1AZES. 



Other oxygen ratios tlie crystals occur only on the hori- 
zontal faces, where the glazes have accumulated in thick 
masses. 

Second. Xot withstanding that commercial crystal- 
line glazes free from TiOa, quoted in Table I, fall in Group 
II, it is a fact that in no case are crystalline effects devel- 
oped to a commercially note-worthy extent. 

Third. The high coefficient of expansion and contrac- 
tion of the potash-zinc glazes is strikingly shown in the 
fact, that even with an oxygen ratio of 4, the craze lines are 
fine and interlace, forming meshes about the size of those 
in 20 mesh screens. This may explain in part the reason 
that the commercial glazes before cited are so high in silica. 

Fourth. It is evident that KgO-ZnO-SiOo compounds 
cannot fnllill the requirements of ceramists. 

GROUP in. 



Sub-diviiions into Series] 


Variation in RO 


Content 


Oxvgen 


Ratio of 


Glaze Number 


on Basis of SiOj Molec-| 


with! 


1 Kach S 


eries 


Each 


Series 


in Each Series 


ular Content [ 

i 














Series No. 


SiO, 
Content 


No. 


ZnO 


Na„0 


Series No. 


O. R. 


a-j 


1 


0.50 


a 


1.0 


00 


1 


1.0 


221-230 


2 


66 


b 


0.9 


0.1 


2 


1.1 


231-240 


3 


0.60 


c 


08 


0.2 


3 


1.2 


241-250 


4 


0.65 


d 


0.7 


0.3 


4 


1.3 


251-260 


5 


0.70 


e 


0.6 


0.4 


5 


1 4 


1 261-270 
271-280 


6 


0.75 


f 


0.5 


0.5 


6 


1.5 


7 


0.80 


g 


0.4 


0.6 


7 


1.6 


281-290 


8 


0.85 


h 


0.3 


0.7 


8 


1.7 


291-300 


9 


0.90 


1 


0.2 


0.8 


9 


1.8 


301-310 


10 


0.95 


2 


0.1 


0.9 


10 


1.9 


311-320 


11 


1.00 








11 


2.0 


321-330 



FitKihiliti/. The glazes of this group, having NagO as 
the alkaline flux, are more fusible than the KgO glazes of 
Group I. As a consequence, in this group tliere can be 
used 0.1 Eqv. more of ZnO than in Grou]) I, and yet pro- 
duce glazes sufficiently fluid to cover the whole surface of 
the vases. 

When the RO contains less than 0.3 Na^O, irrespective 
of the acid content, the glazes are infusible at cone 10. 



30 CRYSTALLINE GLAZES. 

HoluhiUty. With more than 0.6 NaaO, iu the case of 
oxygen ratio 1:1, and 0.7 NagO, iu the case of oxygen ratio 
1 : 2, tlie glazes are soluble. 

Luster. As compared with those of (Iroup I the glazes 
of this group exhibit a dimness which in many cases would 
be styled "matt." In no instances in this group was there 
a bright glossy glaze developed. 

Glazes having RO content of 

(0.3—0.4 Nai-O 
\ 0.2 — 6 ZnO 

are decidedly "matts/' irrespective of acid content. 

Craz'iiH/ ami shivcriiKj. Fine mesh crazing, as noted 
in Group I, was developed also iu this group. No shivering 
developed. 

Crijsfaliization. As shown in Plate No. IV, the glazes 
having as an RO 



0.3 — 0.5 Na20 
0.7 - 0.5 ZnO 



irrespective of acid content, develop crystalline structure 
to a notable extent. The glazes contxiining less than 
0.5 ZnO melted to clear glasses, but still having less gloss 
than the K2O glasses. 

The crystals developed in the glazes of this group, 
while needle-like, are comparatively large and synametri- 
cally arranged in radiating groups from definite centers — 
like the spokes around a hub. In the case of the potash 
glazes, the needle-like crystals are massed like a pile of 
loose pins, no systematic arrangement being observable in 
any instance. 

Owing to the consistent manner in which matt sur- 
faces and crj^stalline effects were developed in this group, 
notwithstanding the fact that they were burned in three 
installments, as were those of Group I, the statement that 
Na^O will, in glazes of this type, permit of a more exten- 
sive development of crystals than K2O seems warranted. 
It is also evident that on a commercial basis, none of the 
glazes of this group could be used for the production of 



CRYSTALLINE CJ LAZES. 



81 



either matts or crystalline glazes, owiug to the fact that 
neither are developed to a commercially noteworthy ex- 
tent. 



(JROUP IV. 



Sub-division into Series 
on Basis of SiO., Molec- 
ular Content 


Variation 
within 


in RO 
EachS 


Content 
eries 


Oxygen Ratio in 
Each Series 


Glaze Numbers 
in Each Series 


Series No. 


SiO, 
Content 


No. 


ZnO 


Na,0 


Series No. 


O. R. 


a-j 


1 


1.0 


a 


1.0 


0.0 


1 


2.0 


331-340 


2 


11 


b 


0.9 


0.1 


2 


2.2 


341-360 


3 


1.2 


^ i 


0.8 


0.2 


3 


2.4 


351-860 


4 


1.3 


d ' 


0.7 


0.3 , 


4 


2.6 


361-370 


5 


1.4 


e 


0.6 


0.4 1 


5 


2.8 


371-380 


6 


1.5 


f 


0.5 


0.5 


6 


3.0 


381-390 


7 


1.6 


g 


0.4 


0.6 1 


7 


8.2 


I 391-4 00 


8 


1.7 


h 


0.3 


0.7 


8 


3.4 


401-410 


9 


1.8 


1 


0.2 


0.8 


9 


3.6 


' 411-4'JO 


10 


1-9 : 


D 


0.1 


0.9 


10 


3.8 ' 


421-430 


11 


2.0 


i 




i 


11 


4.0 


431-440 



FnsihUitjj. The infusil>le area is the same in this, as 
in Group III. 

Soluhilitij. The soluble area is confined to those 
glazes having XaoO content of 0.9 Eqv. 

Luster. The glazes of this group have a more bril- 
liant gloss than tho.se of the preceding group. Glazes hav- 
ing an KG of 0.3 K2G, 0.7 ZnG, irrespective of oxygen ratio, 
are either so minutely crystalline as to be matts, or have 
large crystals in a glassy matrix. This production in the 
zinc-soda groups of small crystals causing a matt effect, or 
large crystals causing a characteristic crystalline effect, 
would seem to be a function of the burning, rather than of 
glaze composition. 

In all of the matured glazes of this group, those lowest 
in oxygen ratio are the dimmest. 

Crazing and shircrinfj. The same i)h(Mi(»iii(Mia of fine 
mesh crazing with the lowei- oxygen ratios, and relativelv 
coarse mesh crazing with the higher, as in fJroup IT, are 
noted in this group. 

p. & K — 4. 



32 



CRYSTALLINE GLAZES. 




30IX0 lAiniaos ox Baixo dniiz jo soixva 



CRYSTALLINE GLAZES. 33 

Cri/siaUization. The raniio in coiiipositioii of glazes 
showiui' either a developed or latent crystallization is nar- 
rower in this gronp than in Group III, and about the same 
as ill Group 11. The crystallizing tendency, however, 
within this range is more pronounced in Group IV than in 
Group IT, and, where developed, the crystals are quite 
large, the needles being in some instances, particularly in 
glaze Xo. 355, 0.4 Xa^O, O.C ZnO, 1.2 SiOo, over an inch in 
length. 

The most pronounced crystallization occurs within 
oxygen ratios 2.2 to 3.2. At oxygen ratios less than 2.2, the 
crystals are so minute as to produce matt effects, and at 
ratios higher than 3.2 the glazes, while glossy, show but 
very little tendency to crystallize, except on the shoulder 
of the vases. 

Conclusions from Groups III aiu] IV. In these two 
groups, as in the case of Groups I and III, crystals were 
developed with all oxygen ratios ranging from 1 to 4, but 
the most pronounced and best devloped crystals occurred 
with oxygen ratios 2.2 to 3.2. Within these ratios also, the 
crystallization seems to be more persistent, aii<l less a func- 
tion of the burning conditions. 

In glazes 374, 384 and 31)4, having the molecular for- 
mula 0.4 XasO, 0.6 ZnO, 1.4 to 1.6 SiOs, large clusters of 
beautiful crystals were produced that would be considered 
as commercially attractive. 

Development of crystals in the soda groups, III and 
IV, was more pronounced than in the potash groups, I and 
II. This che<ks observed facts in the i)henonienon of de- 
vitrification of glasses. "It^ has been determined that 
glasses rich in soda devitrifv faster than those rich in 
potash.'" 

'Van Hise, Treatise on Metamorpliisni, Mou. No. XLVII, U. S, Geol 
Surv. , p. 248. 



34 



CRYSTALLINE GLAZES. 



GROUP V. 



Sub-division into Series 
on Basis of SiOo Molec- 
ular Content 


Variation in RO Content 
within Each Series 


Oxygen Ratio in 
Each Series 


Glaze Numbers 
in Each Series 


Series No. 


SiO^ 
Content 


No. 


ZnO 


K,0 


Na,0 


Series 
No. 


0. R. 


a-j 


1 


0.50 


a 


1.00 


0.00 


0.00 


1 


1.0 


441-450 


2 


065 


b 


0.90 


0.06 


0.06 


2 


l.l 


451-460 


3 


0.60 


c 


80 


0.10 


0.10 


3 


1.2 


461-470 


4 


0.65 


d 


0.70 


0.15 


0.15 


4 


1.3 


471-480 


5 


0.70 


e 


0.60 


0.20 


0.20 


5 


14 


481-490 


6 


0.76 


f 


0.50 


0.25 


0.25 


6 


1.5 


491-500 


7 


0.80 


g 


0.40 


0.30 


0.30 


7 


1.6 


501-610 


8 


0.85 


h 


0.30 


0.35 


0.35 


8 


1.7 


611-520 


9 


0.90 


i 


0.20 


0.40 


0.40 


9 


1.8 


521-530 


10 


0.95 


J 


0.10 


0.45 


0.45 


10 


1.9 


631-540 


11 


1.00 










11 


2.0 


541-550 



FiisibiHty. When the KO contains less than 0.4 
KNaO, the glazes are infusible at cone 10, irrespective of 
acid content. 

^olubiUty. With more than 0.6 KNaO in the case of 
oxygen ratio 1, and 0.7 KNaO in case of oxygen ratio of 2, 
the glazes are soluble. This confirms the results noted in 
Group I. 

Luster. All the matured glazes are matts, except 
those havinu' the molecular formula 



0.4 ZnO 
0.3 K2O 
0.3 Na^O 



!■ 



3 to 2.0 SiOs 



These glazes are clear bright glasses without the least 
sign of crystallization. 

Crazing iuid sJtirrriiif/. Crazing in fine meshes, but 
no shivering are noted in tliis group. 

CrjjstaUkatioii. The essential difference between the 
crystals in the zinc-potash and zinc-soda glazes in the al- 
most total absence of symmetrical grouping in the former, 
and almost perfect symmetry in the latter. In this group, 
where the potash and soda are present in equal equivalents, 
these two crystallizing habits seem to operate simultane- 
ously, but independently of one another, causing a matt 



CRYSTALLINE GLAZES. 



35 



surface. Under an ordinary magnifying glass there may 
he no ci-ystalline structure noted, hut when examined un- 
der a higher power ghiss, the crystals are observed to be 
irregularly arranged, i. e., here a symmetrical grouping 
and there a heterogeneons mixture. The force tending to 
develop the characteristic crystals of the zinc-soda mixture 
seems, however, to be the stronger, for wherever the crys- 
tals are developed sufficiently to be distinguished by the 
naked eye, they are arranged in symmetrical groups simi- 
lar to those described in the soda glazes. 

GROUP ^^. 



Sub-division into Series 
on Basis of SiO^ Molec- 
ular Content 


Variation in RO Content 
within Each Series 


Oxygen Ratio in 
Each Series 


Glaze Numbers 
in Each Series 


Series No. 


SiO, 
Content 


No. 


ZnO 


K„o 


Na^O 


Series 
No. 


O. R. 


a-j 


1 


1.0 


a 


1.00 


0.00 


0.00 


1 


2.0 


551-560 


2 


1.1 


b 


0.90 


0.05 


0.05 


2 


2.2 


561-570 


3 


1.2 


c 


0.80 


0.10 


0.10 


3 


2.4 


571-580 


4 


1.3 


d 


0.70 


15 


0.15 


4 


2.6 


581-690 


5 


1.4 


e 


0.60 


0.20 


20 


5 


2.8 


591-600 


6 


1.5 


f 


0.50 


0.25 


0.25 


6 


3.0 


6U1-610 


7 


1.6 


g 


40 


0.30 


0.80 


7 


3.2 


611-620 


8 


1.7 


h 


30 


0.35 


0.35 


8 


3.4 


621-630 


9 


1.8 


1 


0.20 


0.40 


0.40 


9 


3.6 


631-640 


10 


1.9 


J 


10 


0.45 


0.45 


10 


3.8 


641-650 


11 


2.0 










11 


4.0 


651-660 



Fiisihilitu, ^oluhilitij. The fusible and soluble areas 
of this group correspond exactly with those of Group II. 

Lufifer. Most of the matured glazes of this group 
exhibit a surface-that is neither as glossy as the potash, nor 
as dim as the soda glazes. Mattness is most lu-onounced 
with oxygen ratios ranging from 2.0 to 2.0. 

Craziufi and shirrrinf/. All the matured glazes of this 
group are crazed, the craze mesh being coarsei' wilh each 
increase in acid content. 

Criistcillizatiou. Considering the most promising re- 
sults shown in Groups III and IV, it is indeed very puzz- 



36 



CRYSTALLINE GLAZES. 




93xmd oa N\ saoixo lAinioos qnv wmssviod ox aaivo oniz do soixva 



;sg.£ 



o N in 



CKYSrAl,LINE CLAZKP. 37 

liiij; to note that the ghizes of this ••roup show so little 
tendency to crystallize. No large clusters of crystals were 
developed. 

('o)icUisions from Groups V and \\. The crystalline 
areas in (ironi)s V and VI are shown in Plate No. V, on 
page 3<J. 

It is impossible to state whj' there was so little crys- 
tallization in these double alkali mixtures. The same 
treatment throughout was given to these groups as to the 
former, so that the limited extent of crystallization cannot 
be charged wholly to the conditions of tiring or cooling. 
The only analogous phenomena that the writers have per- 
sonally seen is the attempt by Mr. Jones to obtain in our 
laboratories a crystalline mass by supersaturation of a 
mixture of several salts in solution, as reported by him in 
his paper on efflorescence. Supersaturation of a single 
salt produced definite crystals, but with a mixture of salts 
the crystalline nature, if it i)ossessed any, could not be de- 
termined. It would seem far-fetched to attempt the ex- 
planation of the dearth of crystal development in these 
groups on any such basis, but the facts in these two experi- 
ments, the one in ordinary aqueous solution and the other 
in an igneous silicate solution, coincide so closely that we 
cannot do otherwise than give the involved idea some 
consideration. 

OXini: OF MANOANESi: AS A CUYSTALLIZIX(; ACJKNT. 

The following manganese groui>s arc similar in «-on- 
struction to those of the i)receding zinc gi«.ups. They were 
planned and nearly all made up before the rc^-alts obtained 
in the zinc groui)s had been studied. In this we were fortu- 
nate, for b}' blindly planning these groups, much that is of 
interest, as will be seen, would have otherwise not been 
brought out. 



-38 



CRYSTALLINE GLAZES. 



GROUP XI. 



Sub-division into Series 


Variation in RO 


Content 


Oxygen Ratio of 


Glaze Number 


on Basis of SiO, IVIolec- 


within Each S 


eries 


Each Series 


in Each Series 


ular Content 












Series No. 


SiOs 
Content 


No. 


MnO. 


K„0 


Series No. 


O. R 


a-j 


1 


0.50 


a 


1.0 


0.0 


1 


1.0 


1101-1110 


2 


0.B5 


b 


0.9 


0.1 


2 


1,1 


1111-1120 


3 


0.60 


c 


0.8 


0.2 


3 


1.2 


1121-1130 


4 


0.65 


d 


0.7 


0.3 


4 


1.3 


1131-1140 


5 


0.70 


e 


0.6 


0.4 


5 


1.4 


1141-1160 


6 


0.75 


f 


0.5 


0.5 


6 


1.6 


1151-1160 


7 


0.80 


^ 


0.4 


0.6 


7 


1.6 


1161-1170 


8 


0.85 


h 


0.3 


0.7 


8 


1.7 


1171-1180 


9 


0.90 


i 


0.2 


0.8 


9 


1.8 


1181-1190 


10 


0.95 


j 


0.1 


0.9 


10 


1.9 


l!91-12uO 


11 


1.00 








11 


2.0 


1201-1210 



Fiisihiliti/. The fluxiiio power of mangauese mani- 
fested itself ill member a of each series where the glazes, 
while not smooth, have flowed sufficiently to cover the bot- 
tom of the saucer. On comparing the members a of this 
group with those of Groups T and III, it would seem as 
though MnO is a more powerful flux than ZnO. The re- 
maining members of this group fuse to a smooth coating at 
cone 10. 

Sohibiliti/. AVitli a potash content of 0.4 or more, and 
oxygen ratio of 0.5, and also with 0.7 or more KoO and an 
oxygen ratio of 1.0, the glazes are soluble in water. 

Crazing and shiverinf/. There is practically no craz- 
ing or shivering noted in this group. 

Li(stei\ There are no bright glazes in this grou]). All 
are dark brown, but streaked with purple in those of 
higher, and light "flesh'- brown in those of lower alkaline 
content. 

CrysiaUization. In Plate VI is shown the crystalline 
area of this group. 

The character of cry.stals, i. e., their habits of growth, 
vary considerably in a seemingly erratic manner. For in- 
stance, in glaze 1144, 11G3, 1164, 1172, and 1173, the crys- 
tals are projected from the surface in globular form, while 



CRYSTALLINE GLAZES. 



ill the other iiistauces, the (•rA>!tals are concentrically de- 
veloped and lie wholly within the ])lane of the glaze matrix. 
In the case of the globular crystals, the greatest develop- 
ment is near the base of the vase, while the non-globular 
forms are developed mostly at the top of the vase. 

Under a magnifying glass, the crystals in both in- 
stances are similar in ueneral character. 



GROUP xn. 



Sub-division into Series 
on Basis of SiO., Molec- 
ular Content 


1 

j Variation in RO Content 
1 within Each Series 

1 


Oxygen Ratio of 
Each Ratio 


Glaze Numbers 
in Each Series 


Series No. 


SiO, 
Content 


No. 


MnO 


K,o 


Series No, 


O. R. 


a-j 


1 


1.0 


a 


1.0 


0.0 


1 


2.0 


1211-1220 


2 


1.1 


b 


0.9 


0.1 


2 


2.2 


1221-1230 


3 


1.2 


c 


0.8 


0.2 


3 


24 


1231-1240 


4 


1.3 


d 


0.7 


0.3 


4 


2.6 


' 1241-1250 


o 


1.4 


e 


0.6 


04 


5 


2.8 


1251-1260 


6 


1.6 


f 


0.5 


0.5 


6 


30 


1261-1270 


7 


1.6 


g 


0.4 


0.6 


7 


3.2 


1271-1280 


8 


1.7 


h 


0.3 


0.7 


8 


3.4 


1281-1290 


9 


1.8 


j i 


0.2 


0.8 


9 


3.6 


12!tl-130O 


10 


1.9 


j 


0.1 


0.9 


! 10 


38 


1301-1310 


11 


2.0 








! 11 


4.0 


1311-1320 



Fu^ihUitij. All glazes having an oxygen ratio 1 : 2 are 
fusible at cone 10. When the oxygen ratio has been in- 
creased to 1 : 4.0 the glazes containing less than 0.3 equiva- 
lents of K^O are infusible. 

^olKhilitij. All glazes of this group containing more 
than 0.7 efiuivalents of KoO are soluble. 

Luster. While no "matts" occur in this group, none 
of the glazes containing more ^InO or less K^O than 0.5 
e(iuivalent.s, can be said to be l)right glazes. 

Notwithstanding the fact that with low K^O we have 
a high ]MnO content, there are no evidences in this or the 
preceding group of any sejiaiating out of the manganese 
forming the metallic lustrous effect that would api)ear in 
a raw lead glaze, if more than 0.2 equivalents of ^InO were 
used. If anything, the maximum density of color in these 
groups is produced with JIO, ().'> MuO, ()..j K.O. 



40 



CRYSTALLINE GLAZES. 




•S-BXHTJ oa Nl 3aiXO l^jniSSViOd ox 301X0 393NV9NVIAI dO 901XVH. 






CRYSTALLINE GLAZK8. 41 

Crystallization. In Plate No. VI is shown the crystal- 
line area of this group. 

The crystals in glazes having an oxygen ratio of 2.0 — 
2,6 are very consistent in their manner of growth, and, in 
repeated burns, in their extent of development. Data is 
not at hand from which to judge of the intluence of varying 
heat treatment and tire gases on the development of crystals 
in this acid range. So pronounced is the crystallizing ten- 
dency in this acid range, that in nearly every instance, the 
crystals have projected from the surface of the vase in 
globular forms. 

With higher acid content, the crystals are all within 
the plane of the glaze matrix, and assume beautiful purple 
concentric forms. 

Color. Rich purple tints are developed in the follow- 
ing glazes : 

Groups XIII and XIV. 

Groups XIII and XIV were similar in construction to 
the two preceding groups, ditt'ering from them only in the 
use of Na.O in place of KoO. 

The general features of Groups XIII and XIV are 
similar to those of the corresponding Iv.O groups. Tiieir 
< lystalline areas are represented in Plate No. VII. 

The habit of the crystals developing into the nodular 
forms is not so pronounced in the Xa^O glazes as in those 
of the KoO group, but occurs in the same relative glaze 
composition in these as in the K^O groups. Crystals, when 
lying wholly within the same plane as that of the glaze 
matrix, are more fan-like, and the spaces l)etween the radial 
extension are webbed or filled in moie than is the case with 
the KoO glazes. 

Purple tints occur in glazes of the same relative com- 
position in these groups as in the ])otash gronjts, but are 
much liiihter in tint. 



42 



CRYSTALLINK GLAZES. 




S3xmj OH NJi aaixo lAimaos ox aaixo 3S3NV9nvia) jo soavti 



^"^ 
o^ 



Ein 



3s •£ 



CRYSTALLINE GLAZKS. 43 

Under favorable tiriiij;- conditions, <j;lazes with 0.5 Eqv. 
,MnO and liiij;li acid content, assume a permanent soft black 
velvet-like texture that is very strikin*;-. The writers are 
unable to explain this phenomenon, owinj;- to the fact that 
its development has been proven not to be wholly a func- 
tion of either the composition or tiring conditions. In in- 
tensity of color and softness of texture, it surpasses the 
most perfect black tile, made by the use of recovered man- 
uanese oxide, that the writers have ever seen. Vases ex- 
hibiting this beautiful black etfect are in the majority of 
cases entirely free from definite crystals of notable size, 
and yet there is occasionally a vase having this smooth 
black texture as a background out from which stand purple 
crystals V^ of an inch in diameter. 
l<liiiinn(ir}/ of Groups XI, XII, XII, XIV. 

1. In both the potash and soda groups, the most in- 
tense crystallizing development occurs between oxygen 
ratios 1.8 and 2.6, or acid content of 0.9 to 1.3. 

2. The crystals developed in the KoO, ]MnO, SiOg 
matrix are purple in color, and when not globular assume 
circular forms. In these circular clusters, light colored 
spiral lines, which are more or less regular in their spiral 
growth, produce a damask effect that is truly beautiful. 
In IMate XXI, page 67, is shown the globular growth of 
crystals in glaze Xo. 1173, having the formula 

and in Plate XX, page 66, the spiral growth in glaze Xo. 
1155, having the formula 

0.6 MnO 1^7"; tun 
4K.O }0-75SiO, 

In Plate XXII, page 68, are shown crystals that are 
just beginning to ])ass from the flat to the gloluilar form. 
This is developed in glaze Xo. 1235, having the formula 

0.6 MnO/ , oq:.) 

3. In contrast to the puri)le crystals of the K.O mix- 
tures, the soda crystals stand out as white, or at least light- 



44 CRYSTALLINE GLAZES. 

tinted forms ou the browu and black baekgrouuds. lu 
these, the concentric or spiral rings are entirely absent. 
The crystals, even when globular, have the circular crystals 
arranged in radial clusters that, with the shading from 
light to dark tints and the silky luster, look very much like 
pansy Idossoms. 

4. The soda crystals, unlike any of those in KoO 
glazes, are in some instances indistinct, and show merely 
as a lighter tinted spot below the surface of the glaze, pro- 
ducing a mottled effect that is superb. As these hidden 
crystals develop, light-tinted concentric rings form around 
the incipient crystals, increasing in both size and intensity, 
until when the crystal appears at the surface, the concen- 
tric rings are no longer apparent and the radial pansy-like 
clusters of the acicular crystals appear. This transition 
from the hidden incipient crystals, to a large well-devel- 
oped cluster of crystals, is displayed on a single vase in a 
large nund)er of instances, and particularly in those glazes 
that have an oxygen ratio of 2.G to 3.0. 

5. With 0.2 eqv. of manganese, the purple tint pro- 
duced is much lighter and less rich in the soda mixtures 
than in the potash. 

6. For variety of crystallizing phenomena, and var- 
iety of tints developed, the soda mixtures surpass those 
containing potash. 

7. As shown by contrasting Plates VI and VII, the 
crystallizing areas in the potash-manganese glazes are as 
large as are those of the soda manganese, but the soda- 
manganese crystals are more persistent in their growth. 



CBYSTALLINK OLAZKS. 



46 



(iKOLF XV. 



Sub-division Into Scries 
on Basis of SiO., Molec- 
ular Content 


Variation in RO Content 
within Each Series 


; 1 

Oxygen Ratio of 
Each Series 


Cflaze Numbers 
in Each Series 


Series No. 


SiO, 
Content 


No. 


MnO 


K,0 


Na,0 


i Series 
No. 


O. R. 


a-j 


1 


0.50 


a 


1.00 


0.00 


0.00 


1 


1.0 


1541-1560 


2 


0.55 


b 


0.90 


0.06 


0.05 


o 


1.1 


1561-1660 


3 


0.60 


c 


0.80 


0.10 


0.10 


3 


1.2 


1561-1670 


4 


0.65 


d 


0.70 


0.15 


0.15 


4 


1.3 


1571-1580 


5 


0.70 


e 


0.60 


0.20 


0.20 


i 5 


1.4 


1.581-1690 


6 


0.75 


f 


0..=iO 


0.25 


2o 


, 6 


1.5 


1591-1600 


7 


80 


g 


O.-IO 


0.80 


0.80 


( 


1.6 


1601-1610 


8 


0.85 


h 


0.80 


0..S5 


0.86 


8 


1.7 


1611-1820 


9 


0.90 


1 


0.20 


0.4i» 


0.40 


9 


18 


1621-1630 


10 


0.96 


3 


0.10 


0.45 


0.45 


10 


1.9 


1631-1640 


11 


1.00 










11 


2.0 


1641-1650 



In solubility, fusibility, luster and freedom from craz- 
ing and shiverinji-, the group corresponds with Group XT. 

Crjjstanizatiou. Globular crystals, ^yhile developed 
Avith the higher oxygen ratios and manganese content as in 
Group XI, are not so pronounced and are more confined to 
definite areas. In Group XI these globular crystals occur 
irregularly at any oxygen ratio, while in this group, they 
occur only at the oxygen ratios of 1.0 and 2. On the whole, 
the growth of nodular crystals is less i)ronounced in this 
than in Group XI. In this respect Group XV resembles 
Group XIII. The crystalline area and the glazes in which 
there is a development of good crystals are in<licated in 
Plate No. VIII. 

In general character, the crystals resemble more those 
of the soda-manganese crystals, but have a little more in- 
definite character. In many instances, and this is not con- 
fined to any given area, the crystals are nearly devoi<l of 
color, i. e. they are very light. In ninny otiiers, there seems 
to be a mixture of white and purjdc crystals infcrfci-ing 
with one another in development in a way that results in a 
most striking variegated fawn-colored ctTcct. Tliis, how- 



46 



CRYSTALLI^fE GLAZES. 



ever, occurs only on a portion of a vase, and often a vase, 
as for example in glaze No. 1646, 

(1.5 MnO \, ^o-p. 

a large area of this variegated or mottled fawn effect will 
be fringed by small radial fan-like crystals resembling in 
detail those occurring in the soda-manganese glazes. In 
glaze No. 1636, 

0.6 MnO 



0,5 KNaO 



0.96 Si02 



the crystallizing tendency results in blurred, but more or 
less round, white spots scattered over a brown background. 
The greater number and the more pronounced crystals 
occur in glazes having as an RO 

S 0.5 MnO 
{ 0.5 KNaO 

GROUP 'SXl. 



Sub-division into Series 


V'ariation in 


RO Content 


Oxygen Ratio of 


Glaze Numbers 


on Basis of SiO™ Molec- 
ular Content 


within Each Series 




Each Series 


in Each Series 


Series No. 


SiO„ 
Content 


No. 


MnO 


K,0 


Na^O 


Scries 

No. 


O. R. 


a-j 


1 


1.0 


a 


0.00 


000 


0.00 


1 


2.0 


1661-1660 


2 


1.1 


i b 


0.90 


0.05 


0.05 


2 


2.2 


1661-1670 


3 


1.2 


c 


0.80 


0.10 


0.10 


3 


2.4 


1671-1680 


4 


1.3 


d 


0.70 


0.15 


0.15 


4 


2.6 


1681-1690 


5 


1.4 


e 


0.60 


0.20 


0.20 


5 


2.8 


1691-1700 


6 


1.5 


f 


0.50 


0.25 


0.25 


6 


B.O 


1701-1710 


7 


1.6 


g 


40 


0.30 


0.30 


7 


3.2 


1711-1720 


8 


1.7 


h 


0.30 


0.35 


0.35 


8 


3.4 


1721-1730 


9 


1.8 


i 


0.20 


0.40 


0.40 


9 


36 


1731-1740 


10 


1.9 


D 


0.10 


0.45 


0.45 


10 


3.8 


1741-1750 


11 


2.0 










11 


4.0 


1761-1760 



The fusibility and solubility area of this group have 
the same relative extent and position as in groups XI and 
XIV. The color is dark brown or black, except in the case 
of low manganese and high silica. 

CrystaUization. The occurrence and extent of crystal- 
lization developed in this group is shown in Plate No. 



("KYSTALMNK GLAZES. 47 

YIII. The cixxtals are not as well developed as in «»TOups 
XI and XIV, where only one alkali is nsed. In fact, the 
combat between the crystallizing forces tending to develop 
the two types characteristic of the K^O glazes on the one 
hand and the Xa^O glazes on the other, has resulted in that 
dense black velvety finish before noted. On some of the 
vases, the characteristic crystals of the KoO glazes are de- 
veloped, and on <»thers those of the Na^O, but on none do 
the two types occur togethei-. 

('ouci ttxUnis OH Hit Maiif/ainsc (I'loiips XI, XVI. 

First. While rich ettects are produced with KgO as 
the only alkali, and with a combination of the two alkalies, 
the most extensive develoiuuent and the most beautiful ef- 
fects are obtained with a simple AInO, Xa^O, SiOo mixture. 

Second. Manganese-alkali-silica combinations have 
in these experiments proven to better support the develop- 
ment of crystals than the zinc-alkali-silica combinations. 

TJiird. For variety of eftects and richness in play of 
dark and light colors, the soda mixtures prove to be the 
best. 

Fonrtlt. For rich purples, the potash mixtures have 
proven to give the best results. 

Fifth. The develojunent of what has been termed 
^'globular'" crystals will be of special interest in further 
.study of this problem, in that they can be easily detached 
and analyzed. The writers have in a speculative way, as- 
sumed various compositions for these globular crystals, on 
the assumption that they are similar in composition to 
some of nature's manganese crystals like rhodonite 
(MnSiO.j), in an attempt to ascertain whether these man- 
ganese glazes in many cases have not been carried to toa 
high a temperature to ])roduc(' the best results. On every 
hypothetical assumption, the non-crystalline jjortion or 
these glazes proved to be similar in make-up to many glazes: 
reported as maturing at cone 1 or less. Tlose adherence to» 
the original ])lan of l)urning all glazes at cone 10, however, 
prevented the following of these suggestions at the lower 

~P. & K.-5. 



48 



CRYSTALLINE GLAZES. 




S3aiyo i/Nniao9 anv lAjnissviod oi. 3aivo 3S3njvonvin jo soixva 



0& 



CRYSTALLINE ULAZHSi 49 

Iieats. After detenu iiiiii«;- the coinpositioii of the globular 
crystals, there will no doubt be disclosed a very fruitful 
field for further research aloug- these lines. 

Si.ith. The use of niauganese as a crystallizing agent, 
so far as the writers can learn, has never been exploited in 
this country before. Keferences have been made in various 
places to the possibility of its u.se, from the fact that the 
Japanese have claimed it was used in the production of 
their magnilicent purple crystals. Being to the writers an 
untried crystallizing agent, no further studies were pro- 
jected at this time, but for future studies it is planned to 
try various zinc-manganese-alkali-titanium and silica com- 
binations, which, it is believed, will be productive of many 
new and highly interesting ])henomena in crystalline 
glazes. 

TITANIC OXIDE AS A CRYSTALLIZING AOKNT. 

In most of the reported researches on crystalline 
glazes, considerable stress has been laid upon the strong 
influence of titanium in the development of crystals. In 
view of the erratic manner in which titanium behaves as a 
flux in clays and pottery bodies, and in view of the widely 
divergent re])orts as to the manner in which crystals grow 
under its influence, it was thought well to study the effect 
of additions of titanium in several equivalent amounts at 
different oxygen ratios, using as a basis the zinc-alkali- 
silica mixtures of Groups I — VI. 

If titanium had the power ascribed to it, eithei- of aug- 
menting the crystallization of other snl»stantes, or by en- 
tering into combination with the crystallizing ingredients, 
of producing greater constancy in liabit of the crystals, the 
writers thought that its true fnnctions in these connections 
would surely be revealed in the scheme presented below. 

These gTon])s were planned, and a few of them Itni-ned, 
))efore the writers had oi)i)(U-(nniJy lo stndy the results 
obtained from Groups I to VI inclusive. A more detailed 
scheme was therefore executed than would have perhaps 
seemed justifiable. Foreseeing this possibility, ;\11 gi-onps 



60 CRY.STAI.I>INK GLAZES. 

of the lower ox.vii,eii ratios were not made up. Coiisideriug 
the surprising" results obtained, however, it is considered 
fortunate that the titanium groups were carried out in the 
detail shown in table IV, page 51. 

Group XVII. 

All glazes of this group containing more than O.-i 
equivalents of KNaO are soluble, irrespective of the nature 
of acid content. 

All glazes of this group containing less than 0.3 K^O 
are infusible. 

Acicular crystals in the ground mass are visible to the 
naked eye, (iven in those mixtures which are too infusible 
to flow, and cover the entire surface of the vase. In fact, it 
would seem in many instances, as though the whole mass 
crystallized in its entirety, thus preventing its becoming 
sullftciently fluid to attain a smooth surface. No commer- 
cially attractive crystals were developed in this group. 

(iioiili XIX. 

In tliis group, all glazes containing mor(? than 0.4 
equivalents of KNaO are soluble. 

In this, as well as in (Jroup XVII, TiOo does not ap- 
pear to operate as a flux. In fact, those glazes in which 
silica is the only acid present, are more fusible than any in 
which TiOo has been added at the expense of some of 
the SiOo. 

In l*late Xo. IX is shown the soluble, fusible and 
crystalline areas of this group. 

The crystals in this group are a trifle more developed, 
and fewer in number. In fact, on the vertical faces of the 
vases, the crystals may be said to be fascicularly grouped, 
each group assuming its individual plane and direction of 
development. On the shoulders of many of the vases in- 
cluded in the crystalline area shown in Plate No. TX, the 
crystals are arranged in radial clusters. Notwithstanding 
the fact that the crystals of this group have definite habits, 
thev are so small that thev mav be called miero-crvstals. 



O -J 





cS 


■^ 








> 


tx 


S3 


•— 1 


:i-i 


■r 


w 


=^ 




hJ 


X 


r-i 


pq 


(B 




<1 


Oi 


5S 


H 


X 


^ 




-C 


1 



0; S 



J 

B 
O 

E 

V 

J} 
6 

V 

o 

c 

c 
o 

"> 

Q 


c 

V 

o 
o 
U 

O 
Pi 

"o 




■ON 


jsqui})^ 


1 




OiOOlOOiOO»00»0 

d o d o d o o o d o 


lii 


OlOOlOOiOOiOOiO 


oooooooooc 








o 

c 
N 


O0»J0t-2i0^C0(Ni-H 


r-<OCOOOO©00 




s 


3IJ3S OJUI 


UOISIAIQ 


r-i(Mec^iO«ot-oooso — 


o 

- 

i Oi 

V 
Bt 
>. 
X 

O 
o 

M 

i' c 
o 

a 

! I 

i o 

i 2 

1 ^ 

1 "^ 
o 

'> 

(5 


Ji 
<J 
a 

V 

c 

c 
o 

•o 
'o 
a 

_c 

5 

.2 

> 
n 


a 

3 
O 

O 


> 


T 

X 


6 

'Si 




6 


— icooooooodc 


a 

3 
O 


> 


•J 

1 

X 

o 


q 


xaso-<(r)w^»oeDr-QO 


OO — -i'^^ — ^^r-<-H 


6 


005XI^:DiO'*c«3C<I-hO 


— oooooooooo 


a 

3 

o 


> 

X 

X 


T 

d 


c' 

c75 


O O O O " ^' ^ -H —' rn' -^' 


o' 


005GOt-CD>OTa3S<Ir-iO 


«ooooooo©oo 


a. 
3 
C 


X 
X 


7 

c 


6 
6 


•^>ocot-xo50^!riM'# 


OOOOOO^^— ^" 


—dooooooooo 


3. 
3 
O 


X 
X 


X 

d 


o 

''Si 


OOOOOOOO — .-H^ 




-iodocoooooo 


a 

o 
O 


X 
X 


30 

7 

X 

6 


O 

o 


Q05Xt-ceiO-t<0r3(M^O 


oosododddoo 

— . CI CO •^ iC «o 1 - X OJ o 


ooooooooooo 


a 

3 
O 

5 


X 
X 


1 
d 




00^<Cl(NerjTt<-^iOCOt^ 


d d d d o cr o c o d d 




' o o o o d d c d o d d 


a 

3 

o 

u 


X 


7 

d 




OiOOiCOiCOiOO»OQ 

ooi-H — c<io^eocct"^S 
d o o d o' d d o o d d 


d" 


iQiOOiOOiOOiOOiCO 

1 d o o o d d d o o d o 



51 



62 



CRYSTALLINE GLAZES. 



TRANS AM CER.SOC VOL ix 
" GROUP XIX 



PURDY AND KREHBIEL, PLATE IX 
OXYGEN RAT\0 1.1.4^ 




7 0.63 56 49 04-2 35 26 2I O \A 07 

EOUIN/A-UEMTS OF TlTArsiC ACIO 

Chart showing results of crystalline glaze experiments, using zinc 
and the alkalies as RO fluxes, and titanic acid as part or all of the acid 
portion. Oxygen Ratio of all glazes on this chart 1:1.4. 



CRYHTAI.LINE IJLAZES. 68 

The jilnzes that liiivc the most promising results are 
those liavinii tlie molecular composition of 

0.6 ZuU I \ U. 21 too. -49 8102 
0.4 KNaO ( ) 0.49 to 0.21 TiOz 

(Iroup XXI. 

In this iiroup, maximum fusibility is attained when 
the ratio of Si(\. t«» TiO. is 4: 1. (Haze No. 2283, having 
the formula 

0.8 ZuO 1 ( 0.72 SiOz 
2KNaO/ 10.13 Ti02 

has fully matured; in fact, fused sufficiently to have run 
off the vase to a considerable extent. 

On vertical faces the crystals are fascicular, but do 
not intertwine or overlap one another. On horizontal 
faces, however, the crystals are more ueedlelike and so 
thickly interlaced that the mass has a dull rough appear- 
ance. 

Crystals of fair size are develo}>ed as shown on Plate 
X. In plates XXIX, XXX and XXXI, pages 75, 76 
and 77, are shown the character of these crystals. 

While it is evident that a small addition of TiOo ren- 
ders the zinc-alkali-silica mixtures having low oxygen 
ratios nu)re fusibh'. the TiO^, or possibly the titanate that 
is formed, does not renmin in solution, but is merely sus- 
pended in the glassy matrix. In fact, many of the glazes 
of Group XXI resemble, in eftect, raw tin glazes. 

There is no evidence in the three titanic oxide groups 
.so far studied, that the titanic oxide enters into combina- 
tion with the crystallizing compounds. On the contrary, 
the crystals develojx-d in glaze 2306, shown in plate XXXI, 
and which c<»ntains no TiO^, are typical of the crystals de- 
veloped in all of the groups which contain TiOo. Further, 
crystals seem to develop only in the traiispai-ent glossy 
patches, and not in those portions which are opaque. A 
mottled effect, i. e. brown splotches, on a white or clear 
background, is so <liaracteristic of the titauiferous 
glaze.s, that it seems jilausible that the titanic acid mole- 



64 



CRYSTALLINE GLAZK8. 



TRANS. AM CER 30C. VOL IX 
GROUP xxt 



PURDV AND KRCHBIEL. PLATE X 
OXVGEN RATIO l;l.8 




09 081 072 063 0.54 045 036 02T 

EQUiyAV-ENTS OF TITANIC ACIP 



Chart showing results of crystalline glaze experiments, using zinc 
and the alkalies as RO fluxes, and titanic acid partly or wholly substi- 
tuting silica in the acid portion. Oxygen ratio of all glazes in this 
chart 1:1.8. Circles indicate position of glazes furnishing good crystals. 



ORVSTALr.INK OLAZRS. 66 

cules have clustereil togetlier iu small groups. These 
small jiTotips aro seldom, if ever, crTstalline, and appear to 
have nothing in lommoii with the clear matrix that sur- 
louiids them. 

That TiO^ conduces to crystallizatiou is shown by the 
fact that the <ilazcs which contain the largest molecular 
equivalent of TiO^ are a mass of minute crystals. As silica 
decreases, and titanic acid increases in equivalent amounts, 
the matrix assumes the characteristic dark brown tint of 
the titanium compounds. These dark brown masses are 
devoid of gloss, and yet when KXaO amounts to 0.7 ecpiiva- 
lents, the mass becomes sufficiently fluid at cone 10 to over- 
flow the saucer on which the vase is placed. Whether this 
fluidity is due to fusion of the wdiole nmss into a molten 
umtrix, of to the fusion of simply a zinc-alkaline silicate, 
by which the titanium compounds are carried in mechan- 
ical suspension, is not known. It is i)lain, however, that 
the minute crystals which are formed in these dark brow'n 
masses, and are devoid of color, resemble those developed 
in the glazes which are free from titanium. 

Plate No. X shows the general results obtained in 
Group XXI. 

GroKj) XXIII. 

In general characteristics, this group resenddes XXI. 
The crystals are larger and, for tlie flrst time, have passed 
from the fascicular to clustered growths. 

In plates XXXII, XXXIII, XXX IV and XXXV, 
pages 78, 71), 80 and 81, are shown characteristic spec- 
imens of the crystals developed in this group. 

As the titanic acid content increases from to 0.2 
ecjuivalents, crystals increase in size and nundjer. lint as 
the titanic acid is increased beyond 0.2 eijuivalents, al- 
though the ci'vstals continue to increase in nund»er, they 
decrease rapidly in size until, as shown in Plate XI, when 
the titanic acid content amounts to 0.6 equivalents the 
crystals are microliti<' in character, and the mass so thor- 
oughlv crvstallized as to be devoid of uloss. When the 



66 



CRYSTALLINE GLAZES. 



TRANS. AM. CER.SOC VOL \X. 
GROUP XXlll 



PURDV AND KREHBIEL. PLATE XI 
OXYGEN RATIO 1:2 4 




EQUIVALENTS OF TlTAMlC ACID 

Chart showing results of crystalline glaze experiments, using zinc 
and the alkalies as RO fluxes, and titanic acid partly substituting silica 
in the acid portion. Oxygen ratio of all glazes on this chart 1:2.4. 
Circles indicate position of glazes furnishing good crystals. 



CKYSTALI.INE GLAZKS. 67 

titanic acid has iucreased to 0.9 equivalents, the whole 
mass has crystallized to a dull porphyritic mass. 

a roup XXIV. 

The essential features of this group are shown in 
l»late XII. 

In this group, the true role of TiOg is better shown 
than in any of the preceding. The sharp line of demarka- 
tion between the area in which crystals are developed in 
clusters, and the area in which the crystals, while more 
numerous, pass from fascicular habits of growth to inde- 
pendent thread-like units, is considered as being due en- 
tirely to the influence of increasing content of TiOg, and 
not to accidental factors such as firing conditions, etc. 

Plate XXXVII, page 83, shows crystals that devel- 
oped in glaze Xo. 2(>13 which has tlie molecular com])()si- 
tion of 

0.8 ZnO (^02 Ti02 
U.2KNaOi '( 1.2Si02 

In no two glazes were the crystal clusters identical. 
This diversified manner of development in zinc-titanium 
glazes has been mentioned by other experimentors, but we 
are still unable to explain the phenomenon. 

In the introductory remarks, reference was made to 
the fact that crystals have been developed in the cold under 
the influences of vapor, from window glass that was to all 
appearances a homogeneous mass. The writers are una- 
ware of any experiments designed to show the possible 
diversity in character of crystals developed under the in- 
fluence of different vapors, Init it seems plausible (hat such 
a diversity would be probable. At all events, the fact that 
ceramists have laid considerable stress upon maintaining 
oxidizing conditions when Iniriiing crystalline glazes, 
shows that considerable im])<)rtance has been attached to 
the composition of the vapors in which the crystals are 
developed. Further, some writers insist that the character 
of the vapor fumes should be controlled by heavily glazing 



58 



CRYSTALLINE GLAZES. 



TRANS. AM CER. SOC VOL. IX PURDY AND KREHBIEL, PLATE XII 

GROUP X.XIV OXYGEN RATIO 1:2.8 




OS 07 06 05 04 03 

EQUIVALECMTS OF TITAMIC ACID 



Chart showing results of crystalline glaze experiments, using zinc 
and the alkalies as RO fluxes, and titanic acid partly substituting silica 
in acid portion. Oxygen ratio of all glazes in this chart 1:2.8. Circles 
indicate position of glazes furnishing good crystals. 



t'KYSTAI.MNE (JI.A/ES. 69 

tlic sides of tlie sa«i<;ei- with tlie same kind of <;laze as that 
on the ware. The writers, however, have not been able to 
(►bserve that the crystals developed in a tightly luted and 
lieavily ulazed saiiuer were any more consistent in their 
liabits of growtli than those developed in u poorly glazed, 
nnluted sagger. 

It is noted that in this groiij), the most consistent de- 
velopment of large crystals occurs when the RO contains 
0.8 e(iuivalents of ZnO. In all groups so far studied, the 
large clusters were most consistently developed when the 
Zn() content was 0.5 equivalents. Further, in the case of 
the preceding titanic acid groups, the crystals attained 
their best development when tlie ratio of TiO^ to Si02 was 
] : 4, but in this group, the oxygen ratio of which is 2.8, they 
attain their best development when the ratio of TiOs to 
SiO^ is 0.5 to 0.1) or 1 to 1.9. That there is some significance 
in the fact that as the titanic acid increases, tlie zinc must 
Ite increased in order to permit of extensive growth of crys- 
tals, cannot be (luestioned. Can it be possible that zinc 
oxide, which, as was demonstrated in Groups I to VI, has 
a i)ronounced crystallizing tendency, must be increased in 
order to overcome the tendency of titanic acid to develo]) 
micro-cr^^stals? This is indeed an imi)ortant consideration, 
but one which the writers feel unable to discuss from data 
at hand. 

(iroiip XXV. 

The oxygen ratio of this grouj) (3.2) is iiigiier than 
that with which, in the zinc and manganese groups, there 
was attained maximum development of crystals. In this 
respect, the results ol)taine(l in this group check those ob- 
tained in the analogous zinc and mangane.se glazes. Con- 
sidering the fact that the several glazes having this oxygen 
ratio were not burncMl at one time, and that, no matter what 
the crystallizing agent emjiloyed happened to be, all agreed 
in having a very limited tendency to develop either large 
individual or extensive groups of crystals, it must be con- 



CRYSTALLINE GLAZES. 



TRANS AM CER SOC VOL JX 
GROUP XXV 



PURDY AND kREHBIEL. PLATE XIII 
OXYGEN RATIO r.3a 




EQUIVALEINTS Or XlXAfMIC ACID 



Chart showing results of crystalline glaze experiments, using zinc 
and the alkalies as RO fluxes, and titanic acid partly substituting silica 
in the acid portion. Oxygen ratio of all glazes in this chart 1:3.2. 
Circles indicate position of glazes furnishing good crystals. Triangles 
indicate glazes which made acicular crystals on vertical surfaces and 
large crystals on horizontal surfaces. 



ceded that the results are not accidental but rather are 
characteristic of glazes of this type and oxygen ratio. 

In Plate XIII are shown the salient features developed 
in this group. 

ft roup XXVI. 

The results obtained in this group are so similar to 
those obtained in the preceding group that they do not 
require elaboration. 

In Plate XIV are shown the general results. 



CRYSTALMNK GLAZES. 



61 



TRANS AM CER SOC VOU »X 
GROUP X)(,VI 



PURDY AND KREHBIEL. PLATE XIV 
OXYGE^4 RATIO 1.3 6 




O 9 7 6 5 4 O 3 

EQUIVALENTS Of^ TlT/O^Nic ACID 



Chart showing results of ciTstalline glaze experiments, using zinc 
and the alkalies as RO fluxes, and titanic acid partly substituting silica 
in the acid portion. Oxygen ratio of all glazes on this chart 1:3.6. 
Circles indicate position of glazes furnishing good crystals. 



Groui) XXVII. 

To this ^Toii]) beloiijjj all of the coininercial ziiu-titanic! 
a<id glaze.s citod in Table I. Notwitlistandiuo the fact that 
these eonimercial glazes have an oxygen ratio of 4.0, the 
writers (li<l not feel justified in enlai-uing the scope of these 
investigations by including ratios higher than 4.0. In the 
first place, they felt that the scope as presented, was per- 
haps larger than the time and facilities at hand wouM war- 
rant, and would rather .sacrifice breadth of sc<»[>e than 
thoroughness in detail. In the second place, it seemed in- 
<onsistent to expect greater crystallizing tendencies in 
mixtures of high acid content, when all of our exi)erience 



62 



CRYSTALLINE GLAZES. 



iu the production of matt glazes, which are crypto,crystal- 
line, indicated that it was in glazes of low acid content that 
crystals would be easiest developed. For these reasons, the 
results of this group are studied with added interest. 

Larger individual crystals and larger clusters were 
developed in this group than in the two preceding groups. 
In Plate XV the results are shown diagramatically. In 
plate XXXVIII, page 84, are shown the crystals obtained 
in glaze No. 2944, which has the molecular composition of 



0.6 ZuO / 
0.4 KNaO \ 



3 TiOa 



f 3 Ti( 
tl.7Si( 



TRANS AM CER SOC. VOL.IX 
GROUP XXVI I 



PURDY AND KREHBIEL, PLATE XV 
OXYGEN RATIO r. A 




10 09 00 0.7 06 05 O.A 03 OZ 

EOUIVAUE^4TS OF TITANIC ACID 

Chart showing results of crystalline glaze experiments, using zinc 
and alkaline as RO fluxes, and titanic acid partly replacing silica in 
the acid portion. Oxygen ratio of all glazes on this chart 1 : 4.0. Circles 
indicate position of .glazes furnishing good crystals. 



CKVSTAKLINK GLAZES. 63 

Trans. Am. Cer. Soc. Vol. IX. Pnrdy.Krehbiel, Plate XVII. 




Glaze No. 34tJ, having the formula 
0.5 Na„0 



1 .10 SiO, 
0..5 ZnO 

Bein^ (tIuzp f. Series 2 (Jroup IV. 



P. & K.— <; 



64 crystallinp: glazes. 

Trans. Am. Cer. Soc. Vol. IX. Purdy-Kvehbiel, Plate XVIIL 




Glaze No. 384, having the formula 
0.3 Na.O 



0.7 ZnO ^ 



1.5 SiO: 



I 



Being Glaze d, Series 6, Group IV. 



OKYSTALMNK (JLAZES. 65 

Trans. Am. Cer. Soc, Vol. IX. Pmdy-Krehbiel, Plate XIX. 




Glaze No. ;J94, having the formula 

0.3 Na.O 

■ l.G SiO. 
0.7 ZnO 



Bi'iiif; Glaze d, Series 7, (irouj) IV 



66 CRVSTALLIXE GLAZES. 

Trans. Am. Cer. Soc, Vol. IX. Purdy-Kiehbiel, Plate XX, 




Glaze No. 115.5, having the formula 
0.4 K„0 



0.6 MnO 



0.75 SiO, 



Being Glaze e, Series 6, Group XI, 



("RYSTALI.INK GLAZES. 



67 



Trans. Am. Cer. Soc, Vol. IX. 



Purdy-Krehbiel. Plate XXI. 







^^K ^^^.^IH^^^I 







Glaze No. llT.j, having the formula 

0.2 K.O 1 

- 0.85 SiOj. 
0.8 MnO S 

Being Glaze c, Serie.s S, GiDup XI. 



68 



CKYSTALLINE GLA/ES 



Trans. Am. Cer. Soc, Vol. IX. 



Purdy-Krehbiel, Plate XXII. 




Glaze No. 1235, having the formula 
0.4 K.O ^ 

0.6 MnO f 

Being (xlaze e, Serie.s 3, Gronp XII. 



1.20 SiO.. 



CRYSTALLINE GLAZES. 69 

Trans. Am. Cev. Soc, Vol. IX. Purdy-Krehbiel, Plate XX[II. 




Glaze No. 1412, having the foiniiila 

0.1 Na.O ) 

I 0.9.J SiO.. 
0.9 MnO ) 

Boiiifi (xlaze b, Scries 1(), (Jroiip XIII. 



70 CRYSTALLINE CLAZES 

Trans. Am. Cer. Soc. Vol. IX. Purdy-Krehbiel, Plate XXIV. 



i 




k 


1 












] 


S^^HPfi 










t 





Glaze No. 1414, having the I'onnula 

U.3 Na.O ) 

I 0.9.5 SiO.. 
0.7 MnO J 

Being- Glaze d, Series 10, Group XIII. 



CKYSTALLINK tJLAZKS. 71 

Trans. Am. Cer. Soc, Vol. IX. Purdy-Krehbiel, Plate XXV. 




Glaze No. \i22. having the foiniula 
0.1 Na.O I 



0.9 MnO 



, 1.(10 SiO.. 



J'ciii}; (Haze li. Serifs II (Jroui) XIII. 



72 CRYSTALLINE GLAZES. 

Trans. Am. Cer. Soc, Vol. IX. Purdy-Krehljiel. Plate XXVI. 



Glaze No. 142o, having the formula 
0.2 Na..O ) 

I 1. 00 SiO„ 
0.8 MnO ) 

Being- Glaze c, Serie.s 1 1 , Group XIII. 



CKVSTAM.INK (il.AZE.S. 73 

Trans. Am. Cer. Soc, Vol. IX. Puidy-Krehbiel, Plate XXVII. 




Glaze Xo. 1442, having tho lorniula 
0.1 Na.O ) 

'- 1: 111 SiO.. 
O.rt MnO ) 

Bfiii;,' Glazc.Ji, Series 2, Grou]) XIV 



74 CRYSTALLINE GLAZES. 

Trans. Am. Cer. Soc, Vol. IX. Purdy-Krehbiel. Plate XXVIII. 




Glaze No. 1G35, having ihe formula 
0.4 KNaO ) 

[0.95 SiOs 
0.6 MnO ) 

Being' Glaze e, Series 10, Group XY 



CRVSTALLIXK (JLAZES. 75 

Trans. Am. Cer. Soc. Vol. IX. Purdy-Krehbiel, Plate XXIX. 




Glaze No. 2286, having the formula 

0..50 KNaO ^0.81 SiO, | 

0.50 ZnO i 0.09 TiO., J 

Beinfj Glaze f, Serie.s 9, Group XXI. 



76 CRYSTALLIXE GLAZES. 

Trans. Am. Cev. Soc, Vol. IX. Purdy-Krehbiel, Plate XXX. 




Glaze No. 2294, having the formula 

0.3 KNaO ^0.81 SiO, 



0.7 ZnO 



0.09 TiO. 



Being Glaze d, Series 10, Group XXI. 

Blisters chie to use of too mnch gnm of tragacantli. 



CRVSTALLINK (JI.AZKS. 77 

Trans. Am. Cer. Soc. Vol. IX. Puidy-Kiehhit'l, Plate XXXI. 




Glaze No. 2:jIm;, having the formula 

0.50 KNaO ■» 0.90 SiO. 



).50 KNaO 1 0.90 SiO, ^ 
1.50 ZnO j 0.00 TiO, S 



Beiii^' (ilazc f. Series 11, (irou]) XXI. 



78 CRYSTALLINE GLAZKS. 

Trans. Am. Cer. Soc, Vol. IX. Purdy-Krehbiel, Plate XXXII. 




Glaze No. 2505, having the formula 

0.4 KNaO "» 1.0 SiO, 



. G ZnO ) . 2 TiOs ) 

Being Glaze e, Series 9, Group XXIII. 



CKYSTAI-LINK <iI,A/KS. 79 

Trans. Am. Cer. Soc. Vol. IX. Puidy-Krehbiel, Plate XXXIII. 




Glaze No. 25U{J, having the loiiiiiila 

0.7, KNaO I I .11 SiO, ^ 
(i.r, ZnO ^ <i.2 TiO, ^ 

P>f'iiij; dhi'/A' f. Seric.'^ 9, (xron]) XXIII. 

A.C. H.— 2H. 



80 CRYSTALLINE GLAZES. 

Trans. Am. Cer. Soc. Vol. JX. Purdy.Krehbiel, Plate XXXIV. 




Glaze No. 2515, having the formula 

0.4 KNaO I 1.10 SiOj j 

0.6 ZnO i 0.10 TiO; \ 

Being Glaze e, Series 10, Group XXIII. 



CRVS'I AIJ.INK GI.AZKS. 81 

Trans. Am. Cer. Soc. Vol. IX. Pmdy-Krehhiel, Plate XXXV. 




Glazf No. 2')\{'>. liaviiiK tliu loiiiiula s 

6 KNaO \ 1 10 SiOv ^ 

r, ZiiO ) 10 Ti().j ( 

Btnuiz (ihi/A f, Scri.-s 10, (in>up XXIII. 



82 CRYSTALLINE GLAZES. 

Trans. Am. Cer. Soc. Vol. IX. Purdy-Krehbiel, Plate XXXVI. 




Glaze No. 2584, having the formula 

0.3 KNaO -» 0.9 SIO, 



0.7 ZnO J 0.5 TiO, 

Being Glaze d, Series 6, Groiap XXIV. 



(;kvstalhnk (ji.azks. 83 

Trans. Am. Or. Sor.. Vol. IX. Purdy-Krehbiel. Plate XXXVH. 




Glaze No. ^tJlD, having the lorniula 

0.2 KNaO ^1.2 SiO., ^ 

0.8 ZnO WJ.2 TiO, ) 

licint; (Uazf c, Series 9, (Jronj) XXIV. 



84 CBYSTAI.T^lNK (JLAZKS. 

Trans. Am Cer. Soc, Vol. IX. Purdy-Krehbiel, Plate XXXVIII. 




Glaze No. 2941, having the formula 

0.3 KNaO 1 1.8 SiO, 



0.7 ZnO ) 0.2 TiO., 

Being Glaze d, Series 9, Group XXVIl. 



CKYS'l'AM.INK tH.AZES. 85 

Trans. Am. Cer. Soc. Vol. IX. Purd.v-KrPhhiol. Plate XXXIX. 




Glaze No. 2954, having the fornmla 

0.:'. KNaO \ 1.9 SiO,. ^ 

0.7 ZnO ) 0.1 TiO, N 

BeiuK Glaze d. Series 10, (Jroup XXVII. 



86 CRYSTALLINE GLAZES. 

Trans. Am. Cer. Soc. Vol. IX. Purdy-Krehbiel, Plate XL. 




I 



A mixed glaze having approximately the formula 
0.33 KNaO ^1.24 SiO, ^ 

0.G6 ZnO ) ti.1'9 TiO, ^ 



CRYSTALLINE GLAZES. 87 

Summari/ of titanic acid groups. In Plate XVT are 
shown TiOo contents with which commercially noteworthy 
(1 ystals were obtained at the various oxygen ratios. The 
writers are unable to explain the contour of the curve, and 
tlie possible significance of the upward trend from O.K. 
3.6 to 4. is obvious. 

While TiOo has proven to be a powerful crystallizing 
agent, its tendency has been toward the production of the 
micro, fascicular and acicular forms. In small amounts 
(0.4 and less) TiOo seems to initiate the development of 
crystals after the type of those formed in the ZnO, KNaO, 
SiOo matrices, but there is no evidence of its becoming a 
constituent part of the crystallized substance. 

GENERAL CONCLUSIONS. 

The result of this investigation seems to throw some 
light on the five questions propounded in the early portion 
of this article, as follows : 

1st. Of the two alkalies, potash and soda, the latter 
in every case was the most conducive to the development 
of crystals. The crystals formed in the soda matrices were 
not only the largest, but had the most pleasing habits of 
growth and grouping. 

2nd. The proportion of zinc and alkali most con- 
ducive to the development of crystallization in every case, 
save that of the glazes high in titanic acid, was, 

0.3 ZnO I .^ ( 0.6 ZnO 
0.7 KNaO( ^" \0.4 KnaO 

3rd. Manganese oxide had by far the greatest crys- 
tallizing tendency, producing not only large crystals but 
also crystals of surprisingly varied habits of growth. Zinc 
oxide iiad a tendency to produce large crystals in local 
areas, as though the crystallizing substance had seggre- 
gated. Titanic acid, on the other liand, produced crystals 
that were small but evenly distributed throughout the 
mass. 

4th. The character and shape of crystals induced by 



o 



a. 



;C O 



E 










Z 


|3 


iC 


'^ 


S^l 




ui 


^ 


:c 


«D 












'^ 




I- 


cc 


S 


V 


u 


lO 


T-" 














e 










3 

25 


il 


■* 


-^ 


n 







o 


;d 


•* 












te 


cu « 


o 


iC 


'if 




O 


IC 


1— t 


sc 


V 

E 










, 








3 


— u 






51 


z 


rs 









— u 



X -^ X 
^-t iC •f 

|~ — I T-l 



o — 



lO — — ' 





-f 


-f 


CO 


•* 


M 


a. u 


r^ 


lO 


t-- 


U 


"^U 


:s 


•— ' 


-^ 


^ 










g 










z 




•* 


■"■ 


•"' 





rf 


X 


X 


i" = 


O 


X 


p. 


u 


■x; 


"^ 


(M 


u 


so 


^^ 


_i 


S3 













Per 
Cents. 


<* 

O 


«o 
so 


o 
«o 


J5 

s 




lO 


N 


s^ 




, 








3 

z 






C^ 


"" 




















c 










o 














-a 










E 


V 






















Z 


c 




0! 


eg 
t3 








:;i 




,9 



03 

88 



:c 



e 



^ a; 



S K 



l« CO t- 00 



crt 


03 






M 


br 


TJ 


■a 


O 


o 


O 


o 


c; 


O 



Ul 




(3 




• ■-- 




O 

o 


c 








o 


o 


o 


■M 


St 


n 


o 








cc 


a; 


O 



'S: « J5 



> -a 



a tc s _, 

a — -kJ o 

o ^ s o 

o u PQ O 



i-H C<l CO •* 



f'KYSTAIvLINK (JLAZKS. 



89 




M 

•a 



aipv DiNvxu JO sxN3-ivAin03 











o; 


o 


CM Tf 


cc 




o 


-a 








•* 


09 










V 


















■M 


E = 


X > 






o 


o 


1 1 


1 




1 


■a' a 

■|| 
o *-■ 


wt^ 






o^ 




1 1 


1 

05 




1 

CO 








o 


d 


d d 


d 




tH 








05 


ec 


o o 


o 




o 










'^l 


x> 


lO lO 


1^ 




■^ 


i! 


o& 






o 


d 




d 




d 


c S 


H« 






-M 


1 


o o 


1 

o 




1 

o 


Si: 

— 5 








o 




















CD 


CD 30 


35 




X 








Tf 


d 


d d 


O 




d 


^'s 






o 


1 


1 1 


1 




1 


o o 








CO 


'* eo 


JO 




CO 


6 0. 
"" o 










d 


d d 


d 




d 










t^ 


;o t- 


t^ 


■c 


t^ 




o >■ 






CD 


d 


d d 


d 


d 


d 


5-^ 








O 


■* 


1 


Jh 


1 

CO 


1 


S 










o 


d d 


d 


d 


d 


c_>.^ 






















































w rt rt "O 


















Jill 


















of g 
3mm 
ve c 
level 


0; 
O 




2 
S 


z* 


O CO 


CO 


CO 


t- 








a 












-2j= b " 


















P U u V 


















=:5S s 


















^ s« 
















1 
1 


1 


o 


1-5 


an 


"3 


0^ ^ a 


jiip 




^2 c> 
f J a - 

2 aX 
^ *- ^ 

= --^Xi 

a-2> ® 
s >x s 

:^ a o 


of crystalline 




> 

3 


3-^ 
aj > 

a a 


-3 

is 


t^ a a' 
^ "C -£ ^ d 

2 OJ X 11 

-§ ^ Sa 


-S-^d^s 

a^v- o a 

J3 C"-^ > " 


> 
X 
X 


2 jd 

B 


1 








'I' o a c oj 
t- ^ a; oj s-i 
^^^co«S 

a a odi£ 
:::: o o „ s! 

*^ <l3 ^ >:a. ►*» 

5> a . cu > 


p 
Sc 


o 




C 
cS 


- OS 
— > 
^ t<-i 

a ^ 


it 

3 C 

.2 
'S 

QQ 

eg 


Sao ??^ 
^ c -5 


1 


9 3 — 


>j 




-r 




io 


io 




^o 


z^-§ 






■*! 




I- -5^ oj 


^ ce (c 




-a eg 03 


.H 


(D 


o 
o 


d 










^ 


'X 


-u 


1 


1 03 •— ' CC 


a x'-' « c 




xi-H ac s 


S^SB^ 




g 


1 


CO 


=« OJ^lS 


=3 =^tH ~ 

"- i d = ^ 




ao a *- 


c-^ 


u 




•^ 


d 


o ao c 






Ph 




d 




O 1 - 


o 1- 




o — 


u o 

tL'Z 


o 




■<* 


X 


^ 00 


CM 


CD 


o 


5« 


'"' 




'^ 


r-H 


CM d 


M 


CO 


■>* 


a 

3 
O 

5 


d 


l-H 
1— 1 

> 

X 








— > 


> 

X 
X 


> 
X 
X 


> 
X 
X 



4 

I 



90 



CRYSTALLINE GLAZES. 91 

luangaiiese can not be adequately described. The globular 
growths, that uuder a magnifying glass were handsome, 
produced en mass an effect that was anything but pleasing. 
The concentric growtli, with the delicate spiral lines in 
lighter tints, were superb, and the later crystals producing 
variegated fawn-colored effects were very beautiful. 

5th. The limits of oxygen ratio permissible in the 
crystalline glazes were not established. The results, how- 
ever, do seem to warrant the statement that an oxygen 
ratio of 1 : 2.8 is better suited to the development of crys- 
tals than that of 1 : 4 as used in practice toda}-. 

Schott^ found similar conditions in his study of tiie 
Jena glass, are shown in Table No. V, on page 88. 

A study of these results confirms our findings in that 
Schott's glasses also crystallized more freely at oxygen 
ratios well below 1 : 4, and also that glasses in which the 
RO is composed of 1 soda to 1 lime crystallize more freely 
than those in which the soda is increased at the expense of 
the. alkaline earth. 

DISCUSSION. 

The Chair: I would like to ask whether, after getting 
these results, an attempt was made to reproduce any of 
them using the same firing conditions and the same glaze 
compositions. 

Mr. Furdy: Yes, and we found considerable trouble 
in getting consistent results in different burns. In a few 
cases, we secured duplicate results. We generally purpose- 
ly varied the heat treatment, so as to produce in one case a 
clear gloss; in another, large crystals, and in another, a 
perfectly matt surface. 

The Chair: Was that variation in the heat treatment, 
or in the glaze mixture? 

Mr. Pnrdij: The variation was in the method of cool- 
ing, in all cases. This is described in the paper, but in my 
anxiety not to burden you by reading the entire paper, I 

'Wagner, Chemical Teohnology, Crooke's Traualatiou, p. 694. 



92 CRYSTALLINE GLAZES. 

gave this rather incomplete synopsis, ofP-hand. In that 
Avay, I omitted to mention many points which should be 
brought out in a proper discussion of the subject. 

The Vhair: If there is to be a commercial future for 
this type of glaze, it must be shown to be in a measure un- 
der control. 

Mr. Punlij : That is a subject which will be taken up 
at the University of Illinois the next semester. We have 
covered the field in this hurried way, and now, selecting 
tlie better crystalline glazes, we are going to try different 
heat treatments and see if they will remain constant. 

TJit Chair: When you get the cr3^stalline formation 
under varying heat treatments, are the crystals always the 
same, i. e. of the same order? 

Mr. Piirdy: Yes, sir. 

Mr. Gates: And always different. In other words, 
when you get one, you know you will never get another 
just like it, for the performance is so different under differ- 
ent cooling conditions. The liquid consistency must be 
just right, and just as long as it is right the crystals will 
continue to grow, their lines of crystallization shooting or 
growing outward from each crystal center until the one 
touches the other and the whole surface is covered. I also 
think the density of the liquid has to do with the fineness 
of the crj^stallization. 

A very interesting study on this subject is to take the 
old experiment of mixing stale beer and sulphate of mag- 
nesium and applying to glass. As the mixture dries or 
whenever it dries to the proper consistency, crystallization 
will commence, and as the lines of this crystallization ex- 
tend out, if new liquid is carefully fed on, — just ahead of 
the lines of crystallization — they can be continued in 
their growth until they become enormous. This I think 
interesting as showing that they do grow, although of 
course impossible to duplicate in a glaze. Still I think the 
same principle applies in that the glaze must reach the 
proper fluidity — neither too great or too little — and must 
be held right there during the crystallizing period. 



CRYSTALLINE GLAZES. 98 

The Chair: Are these differences merely a question 
of size? 

Mr. Gates : No. Tliere is a rtifferenco of arrangeinent 
of the crystals in <>roups, etc. 

TJtc Chair: I have heard of crystals beinr^- dcscribiMl 
as cannibals; that is, as the process of crystallization goes 
on, the bio- crystals eat the little crystals, jnst as big fish 
eat the little ones. 

Mr. Purd}/: A constancy of resnlt is found in all of 
the titanic acid groups, the needle-like crystals all being 
in the same areas of the curve sheets, and inasmuch as it 
required two burns to cover each group, we think the uni- 
formity of results there speaks well for a possible uniform- 
ity in the development of crystals from burn to burn. We 
did not have the same uniformity where zinc was used 
alone, but with zinc and titanium used together, we have 
greater uniformity. 

Mr. Stover: Did you try to make a series, using lime 
as the crystallizing oxide? 

Mr. Purdy : No. 

Mr. Stover: The reason that I ask is that I got one 
of the most beautiful crystalline effects that T ever saw. 
when firing a piece of cement rock, from up near Phillip.s- 
burg. A party sent down a piece of the rock, which T put 
in the kiln, firing it to about cone ten, or just al)out enough 
to melt it, and it produced beautiful crystals. It was 
thought to be a rock whose composition agrees pretty 
closely with the formula used in making Portland cement. 

Mr. Purdy: Perhaps it was not sufficiently empha- 
sized in those parts of the ])aper which I read, but many of 
these glazes that crystallized best, were of the same compo- 
sition as ordinary glass, except that in the place of lime we 
used oxide of zinc ; and in other places, instead of all the 
acid components being silica, \ve used some titanic acid. 
It is interesting to know that the glazes which behaved best 
on the ware, were very similar in composition and crystal- 
izing tenden("ies to ordinary glass. 



Ti 






^^/ ■<_^- 






4^.'J 



r^,j^^^^'.'7a^i^-ir*. 



fff-j> 



r 



UNIVERSITY OF ILLINOIS-URBANA 



3 0112 052567101 



^.^^ 



^ ;-fr*- ' 



KT*^^ 






*i •. '" ':'■ 



1*;^