Industrial and Engineering Chemistry : analytical edition, Vol. 9, No. 3

Analytical Edition Vol. 9, No. 3

IN D U S T R IA L

andENtilNEERING

C H E M IS T R Y

Vol. 29, Consecutive No. 11

Published by the American Chemical Society Harrison E. Howe, Editor

March 15, 1937

Publication Office: Easton, Pa. . Editorial Office: Room 706, Mills Building, Washington, D. C. . Telephone: National 0848 Cable: Jiechem (Washington) . Advertising Department: 332 West 42nd Street, New York, N. Y. . Telephone: B ryant 9-4430

C O N T E N T S

18,900 Copies of This Issue Printed

Direct Determination of Eleostcaric Acid in Tung Oil . . ... ... P. S. K u 103

The Fusibility of Coal A sh...J. J. Brennan, D. F. Mitchell, F. P. Tierney, and W. C. Thompson 106

Separating Butenes from Butanes. Distillation of Azeo- tropic Mixtures with Sulfur Dioxide...

... M .P . Maluszak and F. E. Frey 111

Analysis of Sodium A c e t a t e ...

... Charles B. Hurd and William Fiedler, Jr. 116

Physical Testing Procedure for Latex S to c k s ...

.L . A . Wohler 117

Effect of Tem perature on the Consistency of Asphalts.

Viscosity-Temperature Susceptibility Coefficient as an Index...IT. G. Neviltand L. C. Krchma 119

Arearnetric Estim ation of Small Amounts of Sulfate as Barium Sulfate. . . .V . R. Damerell and P. Spremulli 123

Rare E arth Salts. Precipitation and pH Studies with the Glass Electrode . J . A . C. Bowles and H. M . Partridge 124

A Dithizone M ethod for Measurement of Small Amounts of Zinc... P. L. Hibbard 127

Refractometric Determination of F a t in Chocolate. . . . ...Joseph Stanley 132

Detection of the Elements in Organic Compounds. . . . ... Robert II. Baker and Charles Barkenbus 135 Determination of Potassium. By Means of an Aqueous

Solution of Trisodium Cobaltinitrite in the Presence of N itric Acid... L .V . Wilcox 136 Physical Properties of Asphalt. . . Seward Mason, R. J .

Loomis, S. D. Patterson, II. G. Nevitl, and L. C. Krchma 138 A Simple Auto-Bubbler P i p e t ...

... Milton Burton and Thomas W. Davis 139 Carbonate Content of Oil Lye. Clarified by Centrifuga­

tion and Long Settling...John E. S. Ila n 140 Anhydrous Sodium Carbonate as a Standard of Reference

in Acidimetry. . . .G. Frederick Smith and G. F. Croad 141 Microciiemistry :

The Living Cell. Physical Properties of and Micro­

chemical R e ac tio n s...

. . . Robert Chambers, M . J . Kopac, and C. G. Grand 143 Colorimetric Microdetermination of Cobalt and Potas­

sium ...C. P. Sideris 145 Micromethod for Determining Viscosity of Lubricating

0>ls ...Harry Levin 147 Microchemical Analysis of Pigments. Used in the Fos­

sae of the Incisions of Chinese Oracle Bones...

... A . A . Benedetti-Pichler 149

T h e A m erica n C h e m ica l S o c ie ty a ssu m es no r e sp o n sib ility for th e sta te m e n t« a n d o p in io n s a d v a n c e d b y c o n trib u to r s to it s p u b lic a tio n s .

P u b lish e d b y th e A m erica n C h em ical S o cie ty , P u b lic a tio n Office, 2 0 th <fc N o r th a m p to n S t s ., E a sto n , P a . E n tered as seco n d -cla ss m a tter a t th e P o e t- Office a t E a s to n . P a ., u n d er th e A c t o f M arch 3 , 1879, as 4 8 tim es a year.

In d u stria l E d itio n m o n th ly on th e 1st; A n a ly tic a l E d itio n m o n th ly on th e 1 5 th ; N e w s E d itio n on th e 10th an d 2 0 th . A ccep ta n ce for m ailin g a t sp e cia l ra te of p o sta g e p r o v id ed for in S ectio n 1 1 03, A c t of O ctober 3 , 1917, a u th o r­

ized J u ly 13, 1918. „ n/1 .

A n n u a l su b sc rip tio n r a tes: (a) In d u s t r i a l Ed i t i o n$5 .0 0 ; (6) An a l y t i­ c a l Ed i t i o n $ 2 .0 0 ; (c) Ne w s Ed i t i o n $1 .5 0 ; (a) and (o) to g eth er, $6 .0 0 ;

( a ) , (b ), a n d (e) c o m p le te, $7 .5 0 . F o r eig n p o sta g e t o co u n tr ie s n o t in th e P a n A m eric a n U n io n , (a) $ 1 .2 0 ; (b) $ 0 .3 0 ; (c) $ 0 .6 0 ; to C a n a d a o n e-th ird th e se r a tes. S in g le co p ie s: (a) $ 0 .7 5 ; (6) $ 0 .5 0 ; (c) $ 0 .1 0 . S p e cia l rates to m em b ers.

C la im s fo r co p ies lo s t in m a ils to b e h o n o red m u st b e re c eiv e d w ith in 60 d a y s of d a te of is s u e an d b a sed on r ea so n s o th er th a n “ m issin g fro m files."

T en d a y s a d v a n c e n o tic e o f ch a n g e o f a d dress is re q u ir ed . A d d ress C harles L . P a r so n s, B u sin ess M a n a g er , M ill« B u ild in g , W a sh in g to n , D . C

■1 IND U STR IA L AND E N G IN E E R IN G CHEM ISTRY VOL. 9, NO. 3

\

. M P P ...

' j

>

'

v*<M8p¡iSÍtF¡s

»

..

• . . —

H ^ S i a

311 ®

H ere’s the CONTROL of Plant PRODUCTION

A v it r e o u s e n a m e l j o b b i n g p l a n t u s e s t h i s H o s k in s E le c t r ic F u r n a c e t o d e t e r m i n e t h e b e h a v io r o f t h e e n a m e l a p p lie d o n t h e i r c o n ­ t r a c t w o r k . T h e r e s u l t s f r o m t h i s f u r n a c e h a v e b e e n f o u n d t o b e c o m p l e t e l y t r u s t ­ w o r t h y a s a g u id e t o e n a m e l b e h a v io r in t h e i r p r o d u c t io n f u r n a c e s . T h e m a n s h o w n h e r e s a y s h e d o e s n o t s e e h o w a b u s i n e s s lik e t h e i r s c o u ld b e r u n , w i t h o u t s u c h c o n t r o l e q u i p m e n t . . . . F o r y o u r o w n la b o r a t o r y , H o s k in s F u r n a c e s a r e j u s t a s t r u s t w o r t h y a n d u s e f u l . A ls o , y o u ’l l li k e t h e d u r a b ilit y a n d e a s e o f r e n e w a l, o f t h e i r C h r o m e l e l e ­ m e n t s . . . . F o r m o r e i n f o r m a t i o n , w e s u g ­ g e s t y o u w r it e t o y o u r d e a le r . . . . H o s k in s M a n u f a c t u r i n g C o ., D e t r o i t , M ic h .

HOSKINS

Electric

FURNACES

The W ire th a t M a d e E le c tr ic a l H e a t P o s s ib le

MARCH 15, 1937 ANALYTICAL ED ITIO N 11

The dominant position in the ceramic industry held by Ingram-Richardson is the result of careful scientific research of enamel materials and products. In their new modern laboratories Hevi Duty furnaces are used

in "Porcelfrit" investigations.

H E V I D U T Y E L E C T R I C C O M P A N Y

HEAT TREATING FURNACES ELECTRIC EXCLUSIVELY

M I L W A U K E E , W I S C O N S I N

12 IND U STR IA L AND E N G IN E E R IN G CHEM ISTRY VOL. 9, NO. 3

N E W M O D E L

B U R E T T E

^{a n d}

T H E R M O M E T E R READERS

A. H. T . CO. S P E C I F I C A T I O N

ARTHUR H. THOMAS COMPANY

R E T A I L W H O L E S A L E E X P O R T

LABORATORY APPARATUS AND REAGENTS

W EST WASHINGTON SQUARE PHILADELPHIA, U.S.A.

Cable Address, “Balance,” Philadelphia

BURETTE MENISCUS AND THERMOMETER READERS, A. H . T . Co. Specification.

A new model, smaller in size and lighter in weight than the original model, and provided with an adjust­

able hinge which makes it possible to increase or diminish the distance between the clamping arms.

Construction of the holder is rigid, but light, being of aluminum finished in dull black.

W ith focusing lens, improved type of background and quick-acting spring clamp as in the original model. W ith the cotter pin inserted in the outside hole of the adjustable hinge, tubes of from 6 to 2 2 mm diameter can be accommodated; with the cotter pin inserted in the inside hole, tubes from 5 to 14 mm diameter can be accommodated.

The Reading Attachm ent w ith 4-inch focus lens is intended for observing the position of the meniscus of an aqueous liquid in a burette of regular size, which is sharply defined as a curved black line against a white field. The depth of focus of the lens is sufficient to permit superposition of lines on the front of a tube with those on the back—on burettes with all-round graduations—avoiding parallax. Observations can be made with either reflected or transm itted light and the position of the meniscus a t a rnercury-gas or mercury-water interface can be accurately observed.

The Reader for Burettes of small diameter consists of a 2-inch focus lens of the same diameter as th a t of 4-inch focus b u t mounted in a shorter tube. I t is suitable for Burettes of approximately ⁸ mm diameter such, for example, as our No.

2458 Koch Automatic, No. 8205 Folin Sugar Titration, and No. 8205-A Micro Burettes with filling tube at the side, but is not intended for use with Automatic Micro or other Burettes with filling tube directly behind the graduated tube.

The Thermometer Reading A ttachm ent contains a 2 '/s inch focus lens and is so mounted th a t reflections from the surface of the outside tube of an enclosed scale thermometer such, for example, as a Beckmann thermometer, are reduced to a minimum.

This greatly increases ease and accuracy in reading a therm ometer scale graduated in small intervals.

All three Reading A ttachm ents can be used interchangeably in the Holder.

2 5 0 1 . B u r e tte M e n is c u s R e a d e r , A . H . T . Co. S p e c ific a tio n , a s a b o v e d e scrib e d , o o m p lete w ith re a d in g a t t a c h m e n t w ith 4 -in o h fo c u s le n s for u se

w ith regular s iz e b u r e t t e s ... 10.00

Code Word... ... Bilsm

2 5 0 1 -A . B u r e tte M e n is c u s R e a d e r , A . H . T . Co. S p e c ific a tio n , as a b o v e b u t w ith re a d in g a tta c h m e n t w ith 2 -in ch fo c u s le n s for u se w ith sm a ll b u r e tte s — b u t n o t w ith a u to m a tic m ioro b u r e tte s w ith fillin g tu b e in r e a r ... 10.00

Code Word... Bilwt

9 6 3 2 -A . T h e r m o m e te r R e a d e r , A* H . T . C o. S p e c ific a tio n , as d eso rib ed a b o v e b u t w ith re a d in g a tta o h m e n t w ith 2 ^ - i n c h le n s m o u n te d s p e c ia lly for u se w ith an en c lo se d so a le th e r m o m e te r ... 1 0 .00

Code W ord... O vjk i

F ie ld of v ie w on b u r e tte w ith a ll-ro u n d g r a d u a ­ tio n s, w ith e y e in cor­

r e c t p o sitio n , i.e . no p a ra lla x on cr itic a l lin e

A — B

R e p la c e m e n t P a rts

2 S 0 2 -B . B u r e tte M e n is c u s R e a d in g A tta c h m e n t, o n ly , w ith 4 -in o h fo c u s le n s C ode for u se w ith re gular s iz e b u r e tte s ; fo r in te r c h a n g e a b le u se in th e W ord h o ld er of a b o v e R e a d e r s ... 4 .0 0 B im g k 2 5 0 2 -C . D itto , b u t w ith 2 -in c h fo cu s le n s for u se w ith s m a ll b u r e t t e s ... 4 .0 0 B im ia 2 5 0 2 -D . T h e r m o m e te r R e a d in g A tta c h m e n t, o n ly , w ith 2 H - in c h fo c u s le n s

for en clo se d s c a le th er m o m e te r s; fo r in te r c h a n g e a b le u se in th e

h o ld er of a b o v e R e a d e r s ... 4 .0 0 B im la 2 5 0 2 - H . H o ld e r , o n l y ... 6-00 B im o t

A c o p y o f n e w p a m p h l e t E E -8 4 , w i t h m o r e d e ta ile d d e s c r ip ti o n th a n a b o v e a n d w i t h d ir e c tio n s f o r u s e , s e n t u p o n r e q u e s t.

9632-A

F ie ld o f v ie w in B e c k ­ m a n n T h e rm o m ete r

IN D U S T R IA L

a n d E N G IA T E E R IN G

C H E M IS T R Y

Harrison E. Howe, Editor

ANALYTICAL EDITION

D irect D eterm ination o f E leostearic A cid in T ung Oil

I*. S. KU

C h em ica l R esearch L aboratory, G overn m en t T e stin g B u reau o f H an k ow , H an k ow , C h in a

M

E T H O D S h ith e rto proposed for estim ating eleostearic acid in tu n g oil are m ostly based upon calculations from u n satu ratio n values. T h u s in K aufm ann's m ethod (6) eleo­

stearic glyceride is calculated from th e bromine num ber and the thiocyanogen num ber; in T om s’ m ethod (17), from iodine num bers by W ijs’ m ethod, and b y his bromine vapor m ethods; and in B olton and W illiams’ m ethod (2), from the so-called “ instantaneous iodine num ber.” All these methods are indirect and hence th e ir valid ity is lim ited by th e faults of the m ethods for halogen values and also b y th e weak assump­

tions involved in th eir calculations. In stead of the calcula­

tion m ethods, Bolton and W illiams (8), however, suggested an o th er m ethod to determ ine eleostearic glyceride by estim at­

ing th e “ polym erizable m a tte r” of th e oil (1). B u t the as­

sum ption on which this m ethod was based seems to be incon­

sisten t w ith Rhodes and W elz’s theories of head gelatinization of tu n g oil (11), an d experim entally G ardner (3) showed this m ethod to be n o t dependable unless modified.

I t seems th a t no d irect m ethod is y e t available for determ in­

ing eleostearic acid itself or its glyceride in tung oil. As the drying pro p erty of tu n g oil is chiefly due to its eleostearic acid content, it appears th a t a direct determ ination of it would give a b e tte r com mercial evaluation of the oil th a n any con­

sideration of constants or variables (3).

Separation o f Eleostearic Acid

Two ways to separate eleostearic acid from th e other fa tty acids had been tried ; by precipitation w ith catalysts and by suitable solvents. I n th e m ethods usually employed for de­

term ining tu n g oil in varnishlike products, such as those of G oldsm ith (4), Scheiber (12), von Reibnitz (10), and M arcus- son (8), th e oil is separated o u t through precipitation w ith cata­

lysts such as halogens, nitrites, stannic chlo­

rides, etc. Such precipi­

ta tio n r e a c t i o n s are probably due to poly­

m erization, and it ap­

pears possible then, to utilize these reactions for the determ ination of eleostearic acid in tung oil. C om parative tests on these c a t a l y s t s showed th a t a solution of iodine in petroleum

ether is m ost suitable because it gives a slow precipitation and a noncolloidal precipitate.

From 0.5 to 1.0 gram of pure dried eleostearic acid (pre­

pared in this laboratory) was dissolved in 15 cc. of petroleum ether. The solution was ice-cooled an d 5 cc. of cooled iodine reagent were added. I t was allowed to sta n d in a cool place until precipitation was complete. T he precipitates were fil­

tered on a Gooch crucible, washed free from iodine, dried a t 30° C. in a vacuum oven, and then weighed. In a te s t experi­

m ent three duplicates using th e same sam ple of eleostearic acid gave the following percentages of precipitates: 83.03, 73.96, and S0.31. T he inherent errors of th e various proce­

dures are too great to give concordant results.

Since th e results of th e precipitation m ethod did n o t justify fu rth er study, th e au th o r turned to separation by suitable solvents. Before atte m p ts were m ade to separate fa tty acids, separation of th eir lead and m agnesium soaps w ith both ether and petroleum ether as solvents had been tried. The results were unsatisfactory and hence the o ther w ay of sepa­

ration was resorted to.

U nfortunately the literatu re gives only a few d a ta on solu­

bilities of eleostearic and oleic acids. T he d a ta collected from th e literatu re and those obtained in this la boratory are given together in T able I.

A glance a t T able I will show th a t th e solvents listed would n o t perm it v ery sharp separation. A stu d y of the available solubility d a ta of tung-oil acids in aqueous alcohol solutions, however, reveals th a t th e separation of eleostearic acid from th e o ther acids can be accomplished in alco­

hol solutions of proper stren g th a t low tem peratures. T he solubilities of these acids in aqueous alcohol solutions of different strengths a t 0° C. were carefully investigated. T he

results are given in Table I I an d p lotted in Figure 1.

F i g u r e 1 s h o w s clearly th a t th e com­

plete s e p a r a t i o n of s t e a r i c and palm itic acids from eleostearic acid is i m p o s s i b l e . T he solubility differ­

ence betw een oleic and e l e o s t e a r i c acids is greatest as th e percent­

age of alcohol increases

Ta b l e I . So l u b i l i t i e s o f t h e Fa t t y Ac i d s o p Tu n g Oi l

S o lv en t A ceto n e A m y l a c e ta te C arb on d isu lfid e C arb on tetra ch lo rid e C hloroform E th e r (a b so lu te) E t h y l a c e ta te N itro b en zen e T o lu en e P e tro leu m eth er

S tea ric A cid 4 .7 3 a t 2 5 ° ° 1 1 .1 9 a t 2 5 °*

1 9 .2 0 a t 2 5 ° a 1 0 .2 5 a t 2 5 0a 1 5 .5 4 a t 2 5 ° ° 2 0 .0 4 a t 2 5 ° °

7 . 3 6 a t 2 5 ° ° 0 .0 6 7 a t 0 ° ° 1 3 .6 1 a t 2 5 ° ° V ery so lu b le &

P a lm itic A cid S o lu b le * 1 6 .6 a t 2 5 ° c V ery so lu b le*

S olu b le*

V ery so lu b le*

3 2 . 8 a t 2 5 ° c 1 0 .7 a t 2 5 ° c 0 . 1 4 a t 0 ° « S olu b le*

S olu b le*

a -E le o ste a r ic A cid V ery so lu b le*

V ery so lu b le*

V ery so lu b le*

V ery so lu b le*

V er y so lu b le*

V ery so lu b le*

V ery so lu b le*

3 . 7 5 a t 0 ° c V e ry so lu b le*

V ery so lu b le*

« S o lu b ilities in gram s per 100 g ram s o l so lu tio n .

6 R esu lts of a u th o r’s in v estig a tio n . O thers w ere q u o te d from S eid e ll (1 4 ).

c S o lu b ilities in gram s per 100 gram s of s o lv e n t.

O leic A cid * V ery so lu b le V ery so lu b le V ery s o lu b le V ery so lu b le V ery s o lu b le V ery s o lu b le V ery so lu b le S o lu b le V e r y so lu b le V ery so lu b le

103

104 INDUSTRIAL AND E N G IN E E R IN G CHEM ISTRY VOL. 9, NO. 3 to 76 per ccnt by volum e or

68.98 per cent by weight. T his particu lar stren g th of alcohol is, therefore, chosen for th e separation.

D e v e lo p m e n t o f t h e M e th o d

As early as 1915 Schum ann (13) described a m ethod to de­

t e r m i n e th e "eleom argaric acid” in tu n g oil by m eans of a n alcohol solution which was found also to contain ab o u t 76 per cent of alcohol b y volume.

B u t u n fo rtu n ately th e original procedure gave too few details and th e results of m ore th a n tw en ty trials m ade by the au th o r showed th a t it was far from being a q u a n t i t a t i v e m ethod.

Based upon th e same princi­

ple, a new procedure was then worked out. A 2-gram sample was saponified w ith 2 0 cc. of

5 per cent alcoholic potassium hydroxide an d then acidi­

fied w ith 25 cc. of ² N sulfuric acid. T he solidified fa tty

Ta b l e I I . So l u b i l i t i e s o f Tu n g- Oi l Ac i d s i n Aq u e o u s Al c o h o l a t 0 ° C .

Ph e n o l p h t h a l e i n In d i c a­ t o r. One gram of phenolphthal- ein d i s s o l v e d in ^{1 0 0} cc. of neutralized 95 per cent alcohol.

P r o c e d u r e

Fi g u r e 1. So l u b i l i t i e si n Aq u e o u s Al c o h o l So l u t i o n s

A lco h o l b y V o lu m e a t

(G ram s per 100 g r a m s of sa tu r a te d s o lu tio n )

O leic a -E le o ste a r ic S tea r ic P a lm itic

1 5 .5 6 ° C . A cid A cid A cid A cid

%

5 4 .8 4 0 .2 0 0 .0 2 3 0 .0 4 8 0 .0 2 7

5 9 .6 1 0 .4 2 0 .0 4 7 0 .0 7 0 0 .0 4 1

7 0 .1 9 1 .8 2 0 .2 6 7 0 .1 1 3 0 .0 5 4

7 1 .8 7 2 .4 4

7 3 .9 3 4 .0 6

...

7 5 .0 7 7 .0 6

7 6 .0 5 2 8 .0 0 0 Í 6 Í 4 o ‘.2 2 0 . ÍÓ5

8 0 .3 3 0 .9 8 3 0 .3 7 0 .1 6 5

9 0 .1 4 3 .2 8 0 .8 8 0 .4 4 9

acid was dissolved in 25 cc. of 76 per cent alcohol (by volume) and th e solution was allowed to cool for crystallization.

T he following conditions were found to be optim um : (a) saponification w ith 90 per cent alcoholic potassium hydroxide;

(b) a tem p eratu re of 41° ± 1° C. for acidification w ith a d u ratio n of 1 0 m inutes; (c) stren g th of alcohol for crystalliza­

tion, 76.0 to 76.1 per cent b y volum e; (d) tem p eratu re of cooling b ath , ⁰° C .; (e) washing tire acid once w ith 60 per cen t alcohol.

R eagents

Po t a s s i u m Hy d r o x i d e. A 5 per cent solution of potassium hydroxide in 90 per cent alcohol (by volume) containing 5 grams of E. Merck’s reagent potassium hydroxide per 10 cc. of alcohol.

The alcohol used for this purpose should be previously treated with potassium hydroxide and silver nitrate, allowed to stand for a few days, and then distilled and diluted. Solutions made in this way will not readily turn yellow on standing.

Di l u t e Su l f u r i c Ac i d. A 2 N solution made by diluting to 300 cc., 16 cc. of the c. p. acid of specific gravity 1.84.

Al c o h o l f o r Cr y s t a l l i z a t i o n. About SO volumes of puri­

fied alcohol (95 per cent), less if absolute alcohol is used, are diluted with water to 100 volumes. The volume percentage of alcohol a t 15.56° C . is determined by the pycnometer method, and the strength adjusted so th a t it is between 76.00 and 76.10 per cent. I t is advisable to make up a large volume of solution a t sta rt and check its alcohol strength from time to time.

Al c o h o l f o r Wa s h i n g. A 60 per cent alcohol solution made by diluting with water 64 volumes of purified alcohol to 100 vol­

umes and adjusting the strength to 60 per cent alcohol by volume.

St a n d a r d So d i u m Hy d r o x i d e. A solution of 0.2 AT sodium hydroxide standardized against c. p . benzoic acid or standard sulfuric acid, using phenolphthalein as indicator.

Saponify from 2.00 to 2.05 grams of oil with 2 0 cc. of alco­

holic potassium hydroxide in an Erlenmeyer flask by refluxing on a water bath for ¹ hour. Allow the flask to stand for a few minutes in a water b ath kept a t 41° ± 1° C. Then add slowly 25 cc. of dilute sulfuric acid and maintain the tem perature inside and outside the flask a t 41° *=

1° for ju st ^{1 0} minutes, rotating the flask occasionally. Then take out the flask, add 25 cc. of water, and cool in ice water until the fa tty acids are solidified to a cake. Filter by decantation on a small filter paper about 7 cm.

in diameter, applying suction if necessary. Then, keeping most of the cake in the flask, wash the flask a n d f i l t e r p a p e r thoroughly until free from sul­

fate ion. Carefully tu rn the flask with the cake in it upside down and drain as dry as possible. Press the filter paper dry between sheets of absorbent paper. Then dry the flask and the filter paper a t room tem perature (about 20° C.) in a vacuum desiccator under a vacuum of 1 0 to 2 0 mm. of mercury for just 30 minutes. There should be now no water visible on its wall or on the surface of the cake. Open the desiccator and dis­

solve the fatty acids in the flask and those on the filter paper in exactly 25 cc. of 76 per cent alcohol by pouring the warmed alcohol through the filter paper into the flask. Gently warm the flask, if necessary, to dissolve any undissolved cake, then cork, and immerse the flask in a bath of ice and water. P u t the bath in the specially constructed refrigeration box, or an electric re­

frigerator, maintained between —3° and 0° C. and allow to stand overnight. The tem perature of the bath next morning should be between 0° and 1° C.

F it up a jacketed Büchner funnel 5 to ⁶ cm. (2 to 2.5 inches) in diam eter and surround the funnel with ice and water. Filter

Ta b l e I I I . An a l y t i c a l Co n s t a n t s“ o f Tu n g Oi l

S p e cific g r a v ity , 1 5 .5 ° C . A cid v a lu e

S a p o n ifica tio n v a lu e R e fr a c tiv e in d e x , 2 5 ° C Io d in e n u m b er (W ijs)*

W o r s ta ll’s teat

° A n a ly se s d o n e a cc o rd in g to th e m eth o d s o f th is b u r ea u , w h ich a re e s ­ s e n tia lly th e sa m e as A . S . T . M . m e th o d s.

b V a lu es co rrected to sta n d a rd c o n d itio n s p ro p o sed b y H o , W a n , a n d W en (5 ).

i M e e T u n g S h ia o M e e T u n g T s a i T u n g

0 .9 4 1 5 0 .9 4 1 5 0 .9 4 1 6

1 .5 3 . 6 1 .2

1 9 1 .9 1 9 1 .8 1 9 2 .0

1 .5 1 8 1 1 .5 1 8 3 1 .5 1 7 8

1 6 8 .7 1 7 0 .1 1 6 8 .2

5 ' 28* 5 ' 28" 5 ' 42*

Ta b l e IV. Pr e c i s i o n OF THE Me t h o d S h ia o M e e T u n g

D e v ia -

T a M e e T u n g D e v ia -

T s a i T u n g D e v ia -

E le o - tio n E le o - tio n E lecn tio n

E x p e r i­ stea r ic from stea ric from s te a ric from

m en t acid, a rith ­ a cid , a r ith ­ a cid , a r ith ­

N u m ­ u n c o r­ m etica l u n c o r ­ m e tic a l u ncor­ m etica l

ber re cted m ea n re cte d m ea n re c ted m ea n

% % % % % %

1 7 6 . 0 0 .0 2 7 6 .5 0 . 1 0 7 5 .9 0 . 1 0

2 7 7 .0 0 .0 7 7 6 .4 0 .0 6 7 5 . 8 0 .0 1

3 7 7 . 0 0 .0 5 7 6 .5 0 .1 3 7 5 . 8 0 .0 7

4 7 6 .9 0 .0 6 7 6 .3 0 . 0 7 7 5 .6 0 .1 2

5 7 7 .1 0 .1 3 7 6 .2 0 .1 4 7 5 .9 0 .1 1

6 7 6 .9 0 .1 0 7 6 .5 0 . 1 0 7 5 . 8 0 . 0 6

7 7 7 .0 0 .0 0 7 6 .3 0 . 1 2 7 5 .7 0 . 0 8

8 A rith -

7 6 .9 Av,

0 .1 2 . 0 .0 7

7 6 .3 0 .0 9

A v . 0 . 1 0

7 5 .6 0 .1 3

A v . 0 .0 9 m etica l

m ea n 7 7 . 0 A v era g e d e v ia ­

t io n /a r it h ­ m e tic a l m ean

7 6 . 4 7 5 . 8

0 . 9 p a rt per 1000

1 . 3 p arts per 1000

1 .3 p arts per 1000

M ARCH 15, 1937 ANALYTICAL ED ITION 105 the creamlike crystals on the cooled funnel and suck the eleo-

stearic acid dry with a vacuum pump. Wash first the flask and then the acid with 15 cc. of 60 per cent alcohol which has been previously cooled overnight in the refrigeration box. Suck again with pump until no liquid drops. Dissolve the crystals in warmed redistilled alcohol and wash the top and tip of the funnel thoroughly, using a total volume of about 65 cc. of alcohol.

T itrate the acid with 0.2 N sodium hydroxide, using phenol- phthalein as indicator. ¹ cc. of 0.200 N sodium hydroxide = 0.05566 gram of eleostearic acid.

P r e c is io n

T hree genuine and pure specimens of tung oil know n as

“ T a M ee T u n g ,” “ Shiao M ee T u n g ,” and “ Tsai T ung” in native term s were used to te st the precision of the proposed method. These specimens were expressed from seeds of Aleurites ford ii bought in Szechuan in Jan u ary , 1935. T heir analytical constants are given in T able III.

A series of eight determ inations was m ade on each specimen, following stric tly th e procedure described above. No re­

su lt was rejected unless the difference between the doubtful determ ination and the arithm etical m ean of the other de­

term inations was four tim es or m ore th e average deviation of th e o ther determ inations, because it can be shown from the theory of p robability th a t such a determ ination is 99 per cent th e resu lt of accidental errors (9). The results given in T able IV show th e concordance of checks by careful workers. T he precision of th e m ethod was found to be from 0.9 to 1.5 p arts per 1000. Unnecessary exposure to air, w arm th, and light of b oth th e crude and crystallized eleostearic acid should be avoided as m uch as possible.

Experience shows th a t it is advisable n o t to carry out this procedure in places where th e tem perature is above 25° C.

C o n s t a n t E rror o f t h e M e th o d

The acid crystals finally obtained in this m ethod were found to have usually a m elting point of 46-47.5° C. (m. p.

of pure a-eleostearic acid, 48.0°) and a W ijs iodine number in 1-hour con tact of ab o u t 175-183. These constants defi­

nitely show th a t the crystals are m ainly a-eleostearic acid b u t slightly ad u lterated w ith palm itic and stearic acids and traces of oleic acid, as is to be expected from solubility con­

siderations. I t is evident th a t th e presence of palmitic, stearic, and oleic acids will raise th e result, while the eleo­

stearic acid which is inevitably present in the m other liquor and in the washings will cause a lowering in the result. The sum of these errors m akes the constant error of this method.

To te st this error, control experiments were made in which m ixtures of pure eleostearic acid and other tung-oil acids were employed, b u t for reasons m entioned above th e author did n o t use the halogen m ethods to te st th e proposed method.

The pure a-eleostearic acid used was prepared by a modified procedure of Thom as and Thomson (16) w ithout th e use of carbon dioxide.

Reflux 75 grams of pure tung oil for 1 hour with 300 cc. of 10 per cent potassium hydroxide in 90 per cent alcohol. Decompose the soap with 400 to 450 cc. of 3Ar sulfuric acid warmed to 45° C. and keep the mixture a t 40° C. for ^{1 0} minutes, stirring occasionally. Add 200 cc. of cold water and cool in ice until the acids solidify to a cake. Decant off the acid and wash the cake until free from sulfate ion. Drain and then dissolve in 300 to 350 cc. of warmed 90 per cent alcohol, and allow to stand over­

night a t 0° C. Filter in an ice-cooled funnel and wash three times with 30 to 35 cc. of cooled 90 per cent alcohol. Recrystallize twice in 150 and 130 cc.

of 90 per cent alcohol, using each time a shorter period of cooling. For use in control experiments, a portion of crystals may be now taken out, dried in vacuum, and recrystallized twice in 76 per cent alcohol.

Solutions of eleostearic acid in 76 per cent alcohol were then prepared w ith these crystals an d their strengths ad­

justed to proper values by titra tio n w ith 0.2 N sodium hy­

droxide. F or solubility determ inations the crystals after being recrystallized thrice in 90 per cent alcohol were sub­

jected to a fourth recrystallization in 90 per cent alcohol and then dried a t ab o u t 20° C. under high vacuum . T he final products from either procedure were found to have a m elt­

ing point of 48.0° C.

Six kinds of solutions of eleostearic acid were freshly pre­

pared and standardized, containing from 1.276 to 1.554 gram s of this acid in each 25.0 cc. These correspond to from 65 per cent to 78 per cent of eleostearic acid in 2 gram s of oil, which will nearly cover th e range of percentages in ordinary tung oil. According to v an Loon an d Steger (7, 15) tu n g oil contains ab o u t 13.6 per cent of oleic acid, 3.7 per cent of palm itic acid, and 1.2 per cent of stearic acid.

To im itate this com position approxim ately on th e basis of a 2-gram sample, 0.27 gram of oleic acid, 0.074 gram of palm itic acid, and 0.024 gram of stearic acid were mixed and dissolved by w arm ing below 40° C. in 25-cc. portions of th e six sta n d ard solutions of eleostearic acid and th e n th e m ixtures were allowed to stan d overnight a t 0° C. T he eleostearic acid in each m ixture was determ ined b y th e procedure pro­

posed for tung oil. By analysis of six m ixtures, th e co nstant error of this m ethod for a 2.00- to 2.05-gram sam ple was found to be from 20.6 to 25.0 p arts per 1000 and th e per­

centage of eleostearic acid found experim entally was lower th a n the theoretical value by approxim ately ^{1 . 6} per cent.

Hence a correction of 1.6 per cent should be added to th e experim ental values. However, this correction is b y no means very accurate, since th e tru e percentages of oleic, palmitic, and stearic acids in tu n g oil have n o t y e t been accurately determ ined.

A p p lic a tio n o f M e th o d t o C o m m e r c ia l S a m p le s As this m ethod was developed and tested on genuine oils only, it is interesting th a t when applied to com mercial samples it was found in general equally satisfactory. F or use in te st experim ents, several sam ples w ith good analytical d ata were chosen from the exported oils which th is bureau had examined during th e period from M ay to N ovem ber, 1935. In T able V there are given percentages of eleostearic acid and the analytical constants of th e oils. I t is im p o rta n t to note the consistent relationship between these percentages

Ta b l e V. Ap p l i c a t i o n t o Co m m e r c i a l Sa m p l e s Oil

N o .

R efra ctiv e Ind ex a t

2 5 ° C.

Io d in e N o . (W ijs)

A cid V a lu e

S a p o n ifi­

c a tio n V a lu e

W o ra ta ll’s T e s t

E le o ste a ric A cid (C o rrected )

%

M ean

A v era g e D e v ia tio n fro m M e a n

2918 1 .5 1 6 9 1 6 6 .8 2 .9 1 9 1 .3 6 ' 3 2* 7 6 .1 4

7 6 .2 4 7 6 .1 6 7 5 .9 6

7 6 .1 3 0 .0 8

2926 1 .5 1 7 2 1 6 6 .7 3 . 2 1 9 1 .8 6 ' 25* 7 6 .5 5

7 6 .2 9 7 6 .8 0 7 6 .5 1

7 6 .5 6 0 .1 4

2983 1 .5 1 7 3 1 6 7 .1 2 . 6 1 9 1 .5 6 ' 23" 7 6 .7 0

7 6 .7 8 7 6 .9 5 7 6 .2 6

7 6 .6 7 0 .2 1

3030 1 .5 1 6 5 1 6 7 .8 2 . 9 1 9 1 .3 6 ' 52* 7 5 .5 1

7 5 .6 3 7 5 .3 3 7 5 .4 3

7 5 .4 8 0 . 1 0

3032 1 .5 1 6 6 1 6 7 .9 2 .4 1 9 2 .0 6 ' 52* 7 5 .7 8

7 5 .6 5 7 5 .8 0 7 5 .4 6

7 5 .6 7 0 .1 2

3263 1 .5 1 7 6 1 6 7 .8 1 .3 1 9 0 .8 6 ' 23* 7 7 .1 9

7 7 .1 1 7 7 .1 3 7 6 .8 7

7 7 .0 8 0 .1 0

3284 1 .5 1 7 9 1 6 7 .6 1 .4 1 9 1 .7 6 ' 17* 7 7 .5 5

7 7 .6 7 7 7 .4 3 7 7 .3 3

7 7 .5 0 0 . 1 2

106 IND U STR IA L AND E N G IN E E R IN G CHEM ISTRY VOL. 9, NO. 3 and th e analytical constants, p articularly th e refractive

indices, as this clearly shows the valid ity of th e m ethod.

A p p lic a tio n to A d u lte r a t e d T u n g O il Before being applied to ad u lterated samples, th e m ethod had been tested on different vegetable oils com monly used as ad u lteran ts in tu n g oil. I t was found th a t under th e sam e tre a tm e n t these oils yielded less th a n ^{1 0} per cent of residue insoluble in 76 per cent alcohol, except Tapeseed and cottonseed oils which gave 28.6 per cent an d 29.2 per cent, respectively. However, th e m elting poin t and iodine num ber of the crystals of pure tung oil are sufficient to show th e id e n tity of a-eleostearic acid, while th e residues of other vegetable oils, including rapeseed and cottonseed oils, all behaved like m ixtures upon heating, were incapable of being dried, had no definite m elting points, b u t generally had very low iodine num bers. All these facts serve to dis­

tinguish clearly th e fairly p u re crystals of tu n g oil from the im pure residues of o ther vegetable oils.

T he applicability of this m ethod to ad u lterated tu n g oil was tested on m ixtures of pure tu n g oil w ith 2.5 per cent of six common ad u lteran ts. T he eleostearic acids content of these m ixtures were determ ined b y the proposed m ethod.

T h ey were found to be 1.70 to 1.98 per cent lower th a n th a t of th e pure sam ple and b y calculation this lowering would be ab o u t 1.3 per cent to 1.58 per cent for m ixtures containing 2.0 per cent of ad u lteran t. In th e te st experim ents here described only m ixtures w ith 2.5 per cent of a d u lteran t were studied, for a fu rth e r stu d y on ad u lteratio n problems h ad been planned. However, the results th u s far obtained (Table V I) show th a t this m ethod is applicable to ad u lterated tung oil and th a t b y m eans of it it is possible to d etec t ad u lter­

an ts in tu n g oil ad u lterated w ith 2 . 0 per cent or more of o ther vegetable oils. T he precision of th e m ethod will be reduced when applied to ad u lterated samples.

S u m m a r y

F rom th e stu d y of solubilities of th e f a tty acids of tung oil, a m ethod based upon th e separation of eleostearic acid in 76 per cent alcohol a t 0 ° C . has been developed for its direct determ ination in tung oil.

T o m ake th e m ethod q u an tita tiv e , th e control of its variable factors was investigated in detail. T he precision and co n stan t error of this m ethod were estim ated and the necessary correction was ascertained.

T he m ethod has been found satisfactory w ith commercial samples on th e m a rk e t as well as w ith genuine and pure samples. T he percentages of eleostearic acid are generally consistent w ith th e im p o rta n t analytical constants which usually serve to indicate the q uality of this oil.

Ta b l e VI. Ap p l i c a t i o n o f Me t h o d t o Ad u l t e r a t e d Ttjnq

Oi l

Eleo^

s te a ric E le o -

M ix - R efr a c tiv e A cid o f stea r ic

tu re C o m p o sitio n In d e x P u re A cid of L ow er-

N o . (b y P a r ts) a t 2 5 ° C . T u n g Oil

%

M ix tu r e

% in g

% 10 P u re tu n g oil

S tillin g ia oil

0 7 . 5 2 . 5

1 .5 1 6 9 7 7 .3 » 7 5 .3 7 1 .9 8

20 P u re tu n g oil P o p p y se e d o il

9 7 .5 2 . 5

1 .5 1 6 8 7 7 .3 5 7 5 .4 3 1 .9 2

30 P u re tu n g oil S o y b e a n oil

9 7 .5

2 . 5 1 .5 1 6 8 7 7 .3 5 7 5 .4 7 1 .8 8

40 P u re tu n g oil R a p e se e d oil

9 7 .5 2 . 5

1 .5 1 6 8 7 7 .3 5 7 5 .5 5 1 .8 0

50 P u re tu n g oil C o tto n se e d oil

9 7 .5 2 . 5

1 .5 1 6 8 7 7 .3 5 7 5 .6 5 1 .7 0 60 P u re tu n g o il

S esa m e o il

9 7 . 5 2 . 5

1 .5 1 6 7 7 7 .3 5 7 5 .4 2 1 .9 3

T he m ethod is also applicable to ad u lterated tu n g oil and m ay be used to d etec t 2 per cent or m ore of com mon ad u l­

te ra n ts in tu n g oil.

A c k n o w le d g m e n t

T he au th o r is g reatly indebted to K . H o and C. S. W an for their valuable advice w hich has been very helpful to this investigation.

L ite r a tu r e C ite d

(1) Bolton, E. R., and Williams, K. A., Analyst, 51, 335 (1926).

(2) Ibid., 55, 300-7 (1930).

(3) Gardner, H. A., “ Physical and Chcmical Examination of Paints, Varnishes, Lacquers and Colors,” 6th ed., p. 660, Washing­

ton, Institute of Paint and Varnish Research, 1933.

(4) Goldsmith, J. N., J . Oil Colour Assoc., 9, 342-51 (1926).

(5) Ho, Wan, and Wen, I n d . Eng. Chem., Anal. Ed., 7, 96-101 (1935).

(6) Kaufmann, H. P., Ber., 59B, 1390-7 (1926).

(7) Loon, J. van, Farben-Ztg., 35, 1769 (1930).

(8) Marcusson, J., Z. angew. Chem., 38, 780-2 (1925).

(9) Mellor, J. W., “ Higher M athematics for Students of Chemistry and Physics," p. 533, 1909.

(10) Reibnitz, von, Farben-Ztg., 34, 266-8 (1929).

(11) Rhodes, F. H., and Welz, C . J., I n d . E n g . Chem., 19, 68-73 (1929).

(12) Scheiber, J., Farbe u. Lack, 1930, 513-14, 524-5.

(13) Schumann, C. L.. J. I n d . E n g . C h e m ., 8, 5-15 (1916).

(14) Seidell, A., “ Solubilities of Inorganic and Organic Compounds,”

New York, D . Van Nostrand Co., 1928.

(15) Steger, A., and Loon, J. van, J . Soc. Chem. Ind., 47, 361T (1928).

(16) Thomas, A. W., and Thomson, J. C., J. Am. Chem. Soc., 56, 898 (1934).

(17) Toms, H., Atudy.it, 53, 69-77 (1928).

Re c e i v e d S e p tem b e r 11, 1 9 36. P u b lish e d w ith th e p er m issio n of C . Y W a n g , co m m issio n er of th e b u rea u .

T he F u sib ility o f Coal A sh

J . J . B R E N N A N , D . F . M IT C H E LL , F . P . T IE R N E Y , AND W . C. TH O M PSO N N o r th e r n S ta te s P o w er C o m p a n y , M in n e a p o lis , M in n .

T

H E gas-furnace m ethod for th e determ ination of th e fusibility of coal ash w as adopted as a sta n d a rd by th e A. S. T . M . in 1924 (2). T his m ethod requires molded cones of the ash to be subjected to increasing tem peratures. T he fusing point usually reported is th e softening tem p eratu re (S) recorded w hen th e ash cone has fused down to a condi­

tion as shown b y either cone 2 o r cone 3 (Figure I). T he sta n d a rd m ethod perm its a tolerance of 54° F. for the same lab o rato ry an d 90° F. for different laboratories.

Because of th e difficulty of coordinating th e softening

tem p eratu re determ ined b y th e sta n d a rd m ethod a n d clinker form ation, th e tim e required to m ake th e test, th e tolerance perm itted in duplicate determ inations, an d th e difference in th e results obtained by different laboratories (1,6), th e pres­

e n t investigation was undertaken.

Ten samples of coal ash, the softening temperatures of which had been reported by the Bureau of Mines in 1930 (3), were formed into cones and photographed to show their condition a t various temperatures. The ten samples of ash were divided into two series: five with low softening tem perature (series A) and five with high softening temperatures (series B). A plaque bearing

MARCH 15, 1937 ANALYTICAL ED ITIO N 107 series A cones, which had been heated to 1800° F. in accordance

with the standard method, was removed from the furnace and photographed. Another similar plaque was removed a t 1850° F.

and photographed, another a t 1900° F., and so on for each 50° F.

up to 2500° F. The photographs of these plaques are assembled in Figure 2. The cones of series B were treated in the same manner as series A cones for a tem perature range of 2300° to 2850° F., as shown in Figure 3.

Fi g u r e 1

I t is possible to decide on th e softening points of these ashes by referring to th e photographs and to the definition of the softening point, as explained previously in connection w ith F igure ¹. I t is u n fo rtu n ate th a t th e sta n d ard m ethod p er­

m its the recording of the softening tem perature when the cone has fused down to th e condition as shown by either cone ² or cone 3 (S). These cones are evidently n o t in th e same condi­

tion in regard to fusion, since they differ in height and shape.

The au th o rs’ in te rp re tatio n s of the softening points, guided by th e condition of cone 2, F igure 1, are shown in Table I w ith the softening tem p eratu res of these same ashes as de­

term ined an d published by th e B ureau of M ines (9).

W ith the exception of th e softening points on ash ^{1 1} and ash 14, th e results as shown by th e photographs and as given by th e B ureau of M ines are w ithin th e tolerances perm itted

j i 1 i 7 $ II 14 19

A

\ * . i i i v l U l J H .

7 9 11 14 19

1B50°F 7 9 11 14 19

19 00° F

7 9 1114 19

19S0°F

. 1 n . « ’à _

7 9 11 14 19

2OOO0F

7 9 11 14 19

2 050°F 7 9 II 14 19

2100°F

- 2 3 _ c> a f t __ f l û O . 7 9 11 14 19

2150°F

7 9 II 14 19

2 200°F

7 9 II 14 19

2250°F

7 9 11 14 19

2300°F 2350°F^{9 11 14}

14

2400°F

. .

14

2 4 5 0 °F

14

2 500°F

14

2550°F

Fi g u r e 2

a

^{A M}

3 5 8 12 17

B

UM

^{A .}

3 5 8 12 17

2300°F

I ^{i n . juLkU}

3 5 8 12 17

2350°F

i i i i j m u

3 5 8 12 17

2450°F

A &

à

a 3 5 8 12 17

2600°F

3 5 8 12 17

27SO°F

3 5 8 12 17

2SOO°F

3 5 8 12 17

26 5 0 °F

3 5 8 12 17

2SOO°F

Fi g u r e 3

3 5 8 12 17

24 0 0 °F

û ü U i à i 3 5 8 lY T f

2 5 5 O 0F

3 5 8 12 17

2 7OO0F

3 5 8 12 17

285Q°F

Ta b l e I. Co m p a r i s o n o f Ga s- Fu r n a c e So f t e n i n g Te m p e r a t u r e s o f Co a l As h

(A s sh o w n b y p h o to g r a p h s an d as g iv e n b y th e B u rea u o. M in e s)

Series

F ro m p h o to ­ B y B u reau

A sh N o. graphs o f M in e s D iffere n ce

0 F. o

F%

o j?'

7 20 2 5 2055 30

9 2275 22 7 5 0

11 2300 21 9 5 105

14 2350 22 0 0 150

19 1975 2040 65

3 2600 25 9 5 5

5 2625 2615 10

8 2575 2605 30

12 2725 27 0 5 20

17 2 6 5 0 26 6 5 15

under th e A. S. T . M . sta n d a rd m ethod. T here is very close agreem ent between th e softening points as shown by th e photographs and as determ ined b y th e B ureau of M ines for th e higher fusing ash represented by series B.

A ttention should be called to th e condition of cone 11.

In m aking a fusion te st of this ash by th e sta n d a rd m ethod, it can be seen from Figure 2 th a t th e condition defined as th e softening poin t persists from 2150° to 2300° F . T his appears to be due to peculiar characteristics of this ash, as

■will be shown below.

T he value of ash-fusion te sts depends n o t only on th e accu­

racy of th e m ethod em ployed b u t also on the degree to which the results can be correlated w ith th e behavior of the coal under furnace conditions. H eavy clinkers have been encountered thro u g h o u t th e ash-softening tem peratures from 2200° to 2815° F. and no clinker trouble has been ex­

perienced w ith certain coals having ash-softening tem pera­

tures as low as 2388° F. (5). T here appears to be a general relationship between th e softening tem p eratu re and clinker- ing in the furnace (4), y e t i t is practically impossible to pre­

dict th e behavior of ash in coals having softening tem pera­

tures in the neighborhood of 2400° to 2600° F . I n m any cases clinkering characteristics are n o t represented by the

108 INDUSTRIAL AND E N G IN E E R IN G CHEM ISTRY YOL. 9, NO. 3

Fi g u r e 4

m ere figures for the initial, the softening, an d th e fluid tem ­ peratures.

Since th e sta n d a rd m ethod is tim e-consum ing and does n o t lend itself to routine w ork w here ash-fusion tem peratures are required in m inim um tim e, experim ents were m ade w ith th e more rap id D e G raaf electric fusing m ethod.

The apparatus, shown in Figure 4, consists of a platinum strip clamped in a horizontal position, enclosed in a metal cover con­

taining a glass window through which the condition of the ash on the strip may be observed by means of a magnifier and through which the radiation pyrometer is focused on the strip so as to indicate the tem perature. The platinum strip is heated by passing an electric current through it and the rate of heating is controlled by mechanical movement of contacts on the rheostat.

Tubes a t the base of the instrum ent permit the introduction of an oxidizing, neutral, or reducing atmosphere within the heating chamber.

In m aking th e first experim ents w ith th e a p p a ratu s the instructions furnished by th e m anufacturer (7) were followed, b u t th e results were n o t reliable. T here was considerable

□ □ □ □ □ □

2 0 0 0 * F 2 0 5 0 *F . 2 1 0 0 * F . 2 1 5 0 *F . 2 2 0 0 " F 2 2 5 0 " F

□ □ □ □ □

2 3 0 0 0 F. 2 3 5 0 ° r . 2 4 0 0 ° F 2 4 - 5 0 T 2 5 0 0 ° F 2 5 5 0 * r

2 6 0 0 T 2 6 5 0 °F. 2 7 0 0 T 2 7 5 0 ° f . 2 8 0 0 ° F

F i g u r e ^

confusion in regard to th e m eaning of the in itial and th e final points determ ined by th e D e G raaf a p p a ratu s w hen com pared w ith the initial deform ation, th e soften­

ing, and th e fluid tem peratures b y th e gas- furnace m ethod.

In 1930 th e B ureau of M ines (9) reported the results of an investigation of th e De G raaf a p p a ratu s in which four laboratories using it m ade determ inations of th e initial and th e final points of nineteen samples of coal ash w hich were furnished b y th e Bureau of M ines. T he results obtained in the four laboratories showed such decided discrepancies, n o t only in th e results w ith th e D e G raaf a p p a ratu s b u t also betw een these results and those of th e B ureau of M ines, determ ined by th e gas-furnace m ethod, th a t th e D e G raaf ap p a ra tu s was n o t considered satisfactory for th e d eterm ination of ash fusibility. O ther tests were m ade (8) in various laboratories w hich indicated the possibility of th e usefulness of th e m ethod, provided a modified procedure could be developed to correct some of th e variables encountered in its use.

I t was decided n o t to abandon th e m ethod an d fu rth e r experim ents were carried on a fte r m aking certain changes, particularly w ith regard to the m anner in w hich th e ash was applied to th e p la tin u m strip and th e atm osphere supplied to th e furnace. A sm all b u t definite am o u n t of finely ground ash was thoroughly m ixed w ith a definite q u a n tity of liquid (preferably glycerol) and a sm all globule of this m ixture was transferred to th e p latinum strip by m eans of a ball point.

T he glycerol was vaporized a t a low tem perature, leaving a uniform layer of ash in in tim ate co n tac t w ith th e platinum . N itrogen, under a definite pressure, was supplied to th e h ea t­

ing cham ber. W ith these changes it was possible to ob tain reproducible results w ithin a reasonable tolerance. These results were n o t significant in term s of the ash-fusion tem ­ peratures b y th e A. S. T . M . m ethod because th e points of fusibility determ ined w ith th e D e G raaf ap p a ra tu s did n o t coincide w ith th e softening te m p eratu re by th e gas-furnace m ethod. N evertheless, th e altered m ethod proved to be . of great value in grading cu rren t shipm ents of coal as to th e clinkering an d slag-forming characteristics of th e ash. B oth th e speed of m aking the te sts and th e reproducibility of the results contributed to its value for th is purpose.

In m aking a large num ber of te sts th e observer, in addition to determ ining th e in itial and th e final points, noted progres­

sive changes in th e condition of th e ash a t various tem pera­

tures between these points. These changes have different characteristics for different coals an d th e y have considerable bearing on th e behavior of th e ash w hen th e coals are burned

□ □ □ □ □ □

! 2 0 0 0 * F 2 0 5 0 ° F 2 I 0 0 * F 2 lS O *T . 2 2 0 0 « F 2 2 5 0 * F

□ □ □ □ □ □

2 3 C O T ? J W T i 4 0 0 * r 2 4 5 0 - f Z 5 0 0 » f 2 5 5 0 T .

Fi g u r e 6 Fi g u r e 7