Industrial and Engineering Chemistry : analytical edition, Vol. 10, No. 2

IN D U S T R IA L

a n iE N G IN E E R U V G analytical e . ™

20,700 Copies of This Issue Printed

February 15, 1938

C H E M I S T R Y ^

Vol. 30, Consecutive No. 7

H a r r iso n E. H o w e , E d ito r

Vol.

No.

Pr e p a r a t i o n a n d Te s t i n g o f La t e x Co m p o u n d s . . .

... J. W. MacKay 57

P u r i f i c a t i o n o f G r a p h i t e f o r S p e c t r o c h e m i c a l An­ a l y s i s ...A. H . Staud and A. E. Ruehle 59

A n a l y s i s o f C o m m e r c i a l P h e n o t h i a z i n e U s e d a s In­ s e c t i c i d e ... L . E . Smith 60

De t e r m i n a t i o n o f Ir o nw i t h o- Ph e n a n t i i r o l i n e . . .

... W. B. Fortune with M. G. Mellon 60

Qu a n t i t a t i v e Sp e c t r o c h e m i c a l An a l y s i s . J. S . Owens 6 4

Hy d r o l y t i c Pr e c i p i t a t i o n o f Ca d m i u m Se l e n i d e f r o m Se l e n o s u l f a t e So l u t i o n s...

... E .C . Pitzer and N. E. Gordon 68

De t e r m i n a t i o n o f Am m o n i a a n d Ur e a i n Mi l k . . . ... A . E . Perkins 6 9

De t e r m i n a t i o n o f Eq u i v a l e n t Ac i d i t y a n d Ba s i c i t y o f F e r t i l i z e r s ...

. . . W. H. Pierre, Nelson Tully, and H. V. Ashburn 72

Co n f i n i n g Li q u i d s f o r Ga s An a l y s i s...

.Kenneth A. Kobe and Frank H. Kenton 76

De t e r m i n a t i o n o f Ca m p h o r i n Al c o h o l i c So l u t i o n s

.Elmer M. Plein and Charles F . Poe 78

Io d o f l u o r i d e Me t h o d f o r De t e r m i n a t i o n o f Co p p e r

William It. Crowell, Sidney H. Silver, and Alan T. Spiher 80

Ro u t i n e De t e r m i n a t i o n o fLow Ch r o m i u m..i n Al u m i­

n u m Al l o y Louis Silverman 81

Lo a d- Ve r b u s- Co m p r e s s i o n Ch a r a c t e r i s t i c s o f Ge l a­ t i n s, Fi b e r s, a n d Ot h e r Ma t e r i a l s . Erwin J. Saxl 82

Di s c o n t i n u o u s Fr a c t i o n a l Ex t r a c t i o n Ap p a r a t u s Ut i l i z i n g Re f l u x... R . E . Hersh, K. A. Varteressian, R. A. Rusk, and M. R. Fenske 86

Fu s e d Ma g n e s i a Cr u c i b l e s...

... E. P. Barrett and W. F. Holbrook 91

Co m p l e t e Me r c u r y- Pu r i f i c a t i o n Sy s t e m...

... Walter A. Carlson and L. F. Borchardt 94

Mo d e r n La b o r a t o r i e s:

Ar m c o Re s e a r c h La b o r a t o r i e s . . . Anson Hayes 97

Mi c r o c h e m i s t r y:

El e c t r i c Fu r n a c e f o r Au t o m a t i c Co m b u s t i o n i n Mi c r o e l e m e n t a r y An a l y s i s . . . L . T. Hallett 101

Mi c r o t e c h n i c o f Or g a n i c Qu a l i t a t i v e An a l y s i s ...D . Gardner Foulke and Frank Schneider 104

Qu a l i t a t i v e Se p a r a t i o n s o n Mi c r o Sc a l e . . . .

. . A. A. Benedetti-Piehler and James T. Bryant 107

Mi c r o d e t e r m i n a t i o n o f Ha l o g e n b y Co m b u s t i o n .

... L. T. Hallett 111

Mi c r o d e t e r m i n a t i o n o f Ca l c i u m . . . G . H . Ellis 112

Fu m e Tu b e f o r Mi c r o- Kj e i.d a h l Di g e s t i o n s . . .

... J. S . Blair 112

T h e A m erican C hem ical Society assum es no resp o n sib ility for th e s ta te m e n ts a n d opinions a d v an c e d b y co n trib u to rs to its pu b licatio n s.

P u b l i c a t i o n O f f ic e s E d i t o r i a l O f f ic e s R o o m 7 0 6 , M i l l s B u i l d i n g , W a s h i n g t o n , D . C .

T e l e p h o n e : N a t i o n a l 0 8 4 8 . C a b l e s J i e c h e m ( W a s h i n g t o n )

P u b lish e d b y th e A m erican C hem ical Society, P u b lica tio n Office, 20 th <&

N o rth a m p to n Sts., E a s to n , P a. E n te re d as second-class m a tte r a t th e P o st Office a t E a s to n , P a ., u n d e r th e A ct of M arch 3, 1879, as 48 tim es a y ear.

In d u s tria l E d itio n m o n th ly on th e 1 st; A n aly tical E d itio n m o n th ly on th e 1 5 th ; N ew s E d itio n on th e lp th a n d 20 th . A ccep tan ce for m ailing a t special ra te of p o stag e p ro v id ed fo r in Section 1103, A ct of O cto b er 3, 1917, a u th o r­

ized J u ly 13, 1918.

A n n u al s u b sc rip tio n ra te s: In d u s t r i a l a n d En g i n e e r i n g Ch e m i s t r t

co m p lete, $6.00; ( a ) In d u s t r i a l Ed i t i o n $3.00; ( 6 ) An a l y t i c a l Ed i t i o n

E a s t o n , P a .

A d v e r t i s i n g D e p a r t m e n t : 3 3 2 W e s t 4 2 n d S t r e e t , N e w Y o r k , N . V . T e l e p h o n e : B r y a n t 9 - 4 4 3 0

$2.50; (c) N e w s Ed i t i o n $1.50; (a) an d (6) to g e th e r, $5.00. F o reig n p ostage to co u n tries n o t in th e P a n A m erican U nion $2.40, (a) $1.20; (6) $0.60;

(c) $0.60; C a n a d ia n p o stag e o n e-th ird these ra te s. Single copies: (a) $0.75;

(6) $0.50; (c) $0.10. Special ra te s to m em bers.

C laim s for copies lost in m ails to be ho n o red m u st be received w ith in 60 d ay s of d a te of issue a n d b ased on reasons o th e r th a n “ m issing from files.”

T en d a y s ’ ad v an ce n o tice of ch an g e of ad d ress is re q u ire d . A ddress C harles L. P arso n s, B usiness M an a e er, M ills B uilding, W ash in g to n , D . C ..

l r. S. A.

4 IN D U ST R IA L A N D E N G IN E E R IN G CHEM ISTRY VOL. 10, NO. 2

ACID CITRIC A . R.

LOW LEAD

Especially Designed for

L E A D D E T E R M I N A T I O N

...NEW

ST. LOUIS CHEMICAL WORKS ^{NEW YORK}

CHICAGO PHILADELPHIA

W it h a fixed m ax im u m lim it o f 0.00005% lead, th is n ew C itric Acid has been especially designed to facilitate accurate m easurem ent o f m icro q u an tities o f lead by th e delicate d ith izo n e m eth o d and also for th e separation as sulfide.*

N o t only in analyzing foods and biological m a­

terials, b u t w h erev er lead c o n te n t m u st be deter­

m ined by these m ethods, Acid C itric A. R. L o w Lead is an invaluable reagent. I t is being m an u ­ factured at th e u rg en t request o f ag ricu ltu ral chem ists, food chem ists, and others, w hose w o rk requires th e u tm o st accuracy.

M a x i m u m Limits o f I m p u r it ie s M e e t s A . C . S . S p e c ific a tio n s

I g n it io n R e s id u e ... 0 .0 2 0 % I n s o lu b le M a tte r ... 0 .0 0 5 % Ir o n (F e )...0 .0 0 0 5 % Lead ( P b ) ...0 .0 0 0 0 5 % O th e r H eavy M e t a ls ...T o p a ss test O x a la te (CsO .,)... 0 .0 5 % P h o s p h a te ( P O4) ... 0 .0 0 1 % S ulfate (S O .,)...0 .0 0 2 % T artrate (Q H o O s)... . . .0 .2 %

Send for the M allin ck ro d t catalog of analytical reagents and la b o ra to ry chemicals. All are m anu­

factured to predetermined standards of purity.

* Official and Tentative Methods of Analysis, Association of Official Agricultural Chemists (Washington, D . C.), 4th Edition, 193J, Chapter X X I X .

FEBR U A R Y 15, 1938 ANALYTICAL E D IT IO N 5

T ’ke eiAeatia# '¿acton jpn ciepeaciabriiitty in a iabonatony. cfnyiacf. o a ea -u a ijp n m tempenatune c5iAtnibu.ti.oa in tke u&abCe ckamben Apace-Un ia- keneat in. tke cie&icpojj tke C eaco-D e-K kotiaiky Oi?ea&. 'Tempenatuneii ane ke?ci eoa&taat by tke c?epeacfab£e D eK kotia& ky. d iim etaM ic H'kenmonecfuiaton, u>kick neipaacita to ckaac^eA oj; oae-jpuntk degnee Ceatic^nacie.

^fkeie ooen& ane macfe ia tu?o ^L-g.e^—tke m oit popuictn beiac^ Tlo. 9 5 1 0 0 untk 2 8 2 icjuane iacke^ o | ¿kei^ Apace -pnieeci at $ 1 4 0 .0 0 .

g g N T R ^ .S'OIENT I FlCj C,QMPANY>

S C I E N T I F I C F f c \!fr 5 L A B O R A T O R Y I NSTRUMENTS y L ljv if f A P P A R A T U S N e w Y o rk • Boston • C H I C A G O • Toronto • L o s A n g e le s

Merck Reagent

PERCHLORIC

. ■•V - , •-

IN D U ST R IA L A N D E N G IN E E R IN G C H EM ISTRY VOL. 10, NO. 2

M E R C K & CO. INC. H anu^aciurlnrj ^(o/iemiMá R A H W A Y , N. J.

.?/„■ C H E M I S T R Y

o /tÁ e A U T O M O B I L E

film*

S w ift and sile n t—responding to every need for surging p ow er to flatten h ills and turn m inutes in to m ile s— th e modern car takes in its stride a hundred demands w h ic h w o u ld h ave taxed th e top perform ance o f o n ly a few years ago.

Y et, for th e secret o f its m echanical p erfection, w e m ust g o far beyond the m ost carefully coordinated assem bly lin e. V irtu a lly every part o f to d a y ’s au tom ob ile is m ade o f m aterials as carefully designed for a specific job as th e part itself. T h is " d esig n in g ” o f special m aterials to ex a ctin g specifica­

tio n s is the actual startin g p o in t o f your car’s performance. W h ether the m aterial be an a llo y o f steel or oth er m etal, a rubber or sy n th e tic product, chem ical reagents in the hands o f sk ille d techn icians played a v ita l part in its d evelopm ent.

For w o rk o f th is ex a ctin g nature, research ch em ists rely on M erck R eagent C hem icals. E xperience has sh o w n th at their p u rity and d ep en d ab ility are essen tial for uniform results, b o th in th e lab oratory and under th e co n d i­

tion s o f actual p roduction. A ca ta lo g w ill be m ailed on request.

FEB R U A R Y 15, 1938 ANALYTICAL E D IT IO N 9

Here is the b ook y o u h a v e n e e d e d for a lo n g tim e. The first com plete d iscu ssio n of the fun- dam entcd principles u n d erlyin g m o d e m G la ss Electrode pH M eters . . . written in sim p le u n d erstan d ab le la n g u a g e , it con tain s facts a n d figures e sse n tia l to the intelligent selectio n a n d u se of this n e w a n d pow erful tool.

Y ou n e e d this important inform ation before y o u b u y a sin g le p ie c e of pH eguipm ent. pH practice h a s d e v e lo p e d so recen tly a n d rap­

id ly it's n atu rally subject to m isunderstanding an d exploitation. This authoritative b ook w ill arm y o u w ith facts so that y o u w ill k n ow y o u are gettin g the right eq u ip m en t for your parti­

cular requirem ents.

A real revelation in pH theory a n d practice, in m a n y resp ects revolutionary in its a d v a n c e tea ch in g s, r e v ea lin g the findings from y e a r s of research n e v er before p u b lish ed . This b ook pictures an d d escrib es the full lin e of C ole­

m an pH Electrom eters, E lectrodes a n d A c­

cesso r ies. A source b ook y o u w ill w e a r out w ith referen ce an d of v er y real im m ed iate interest to e v e r y pH operator a n d th o se interested in pH equipm ent.

This DeLuxe Edition is strictly lim ited. A re­

quest o n your b u sin e ss letterh ead w ill in su re y o u of a free c o p y . . . if y o u a ct at o n ce.

COLEMAN ELECTRIC COMPANY 310 MADISON ST. • MAYWOOD, ILLINOIS

YOUR PROBLEMS WITH

This Revolutionary NEW BOOK!

AUTHENTIC...A REAL REVELATION

10 IN D U ST R IA L A N D E N G IN E E R IN G C H EM ISTRY VOL. 10, N O . 2

A .H .T . CO. S P E C IF IC A T IO N

B U R ETTE CLAM P AND S U P P O R T

N IC K E L S IL V E R C L A M P , ON CO O RS P O R C E L A IN BASE W IT H A L U M IN U M ROD

3225.

B u re tte C lam p, A .H .T. Co. S pecification, of N ick el Silver

3225.

Show ing m eth o d of quick a tta c h m e n t a n d rem oval of B u re tte by fin g er p re ssu re

B U R E T T E S U P P O R T , A .H .T . Co. S p e cific a tio n . C o n sistin g o f C lam p o f n ic k e l silv e r o f a n ew d esig n a n d S u p p o r t w ith re cta n g u la r b a se o f so lid C oors p orcela in .

C lam p. S ta m p e d from n ick el silv e r s h e e t, w ith rein fo rcin g ribs to p ro v id e s ta b ility . So sh a p e d as to h old th e b u r ette s firm ly w ith o u t o b scu rin g t h e g ra d u a tio n s a t a n y p o in t. J a w s w ill ta k e b u r e tte s— or o th e r tu b in g — from 10 m m to 18 m m d ia m eter, i.e ., b u r e tte s from 10 m l to 100 m l c a p a c ity . C la m p in g a rr a n g e m en t p ro v id es firm a tta c h m e n t to v e r tic a l su p p o rt rod s from 10 m m to 13 m m d ia m e te r, w ith c o n v e n ie n t a d ju s ta b ility for h e ig h t. J a w s a re c o v e re d w ith ru b b er tu b in g o f a n e w c o m p o sitio n w ith o u t b lo o m an d o f in crea sed a g in g p rop erties. A lth o u g h n ick el silv e r d o es n o t corrod e, it te n d s to d iscolor a fte r ex p osu re to la b o ra to ry fu m es. T h e n ick el a n d la cq u er fin ish is p ro v id ed th erefo re— in a d d itio n to th e to o l fin ish — for reason s o f p er m a n e n t a p p ea ra n ce ra th er th a n for u tility .

Support. C o n sists o f a re cta n g u la r b ase o f so lid C oors p o rcelain 14 in ch es lo n g X 6 in c h e s w id e, w ith fou r ru b ber fee t, a n d w ith rod in ce n te r o f a lu m in u m a llo y w ith “ A lu m ilite ” finish, 24 in ch es h ig h X J/j-inch d ia m eter. C oors p o rcelain w a s ch o se n for th e b a se b e ca u se o f its su p e r io r ity in color, in re sis­

ta n c e to h e a t an d m ec h a n ica l sh o ck , an d in s t a b ilit y to w a rd th e rea g en ts in c o m m o n la b o r a to r y use.

Code

2515. Burette Support, A.H.T. Go. Specification, as above described, consisting of No. 9340-A Sup- P ric e W ord

port and No. 3225 Clamp of tool finish nickel silver. Without Burettes shown in illustration... 4.50 Biqtc 2515-A. D itto, but with nickel plated and lacquered Clamp No. 3225-A... 4.75 Biqky

3225. B u r e t t e C la m p , A .H .T . C o. S p e c if ic a tio n , for tw o b u re tte s , as above d escribed. Of tool finish nickel s ilv e r... 1.00 Cuvok 3225-A , D i t t o , b u t nickel p la te d an d la c q u e re d ... ; ... 1*25 Cuvre 9340-A . S u p p o r t , A .H .T . C o . S p e c if ic a tio n , as a b o v e d escribed, w ith re c ta n g u la r base of solid C oors p o rcelain a n d ro d of

a lu m in u m ... 3.50 Olors

10% discount in lots of 10 , 1 5% " ” » ” 5 0 20% ” ” ” ” IOO\

one catalogue number only

C opy o f p a m p h le t E E -83, w ith m ore d e ta ile d d escrip tio n , an d listin g o f accessories, in clu din g B u re tte R eader, s e n t u p o n req u est.

ARTHUR H. T H O M A S COM PANY

R E T A IL — W H O LE SA LE— E X P O R T

LA B O R A TO R Y APPARATUS A ND REAGENTS

W E S T W A S H IN G T O N SQ UARE P H IL A D E L P H IA , U. S. A.

C a b le A d d ress, “ B a la n c e ,” P h ila d e lp h ia

INDUSTRIAL *».. ENGINEERING CHEMISTRY

A N A L Y T IC A L E D IT IO N ♦ H a r r is o n E . H o w e , E d ito r

P reparation and T esting o f Latex C om pounds

J. W. MacKAY, M onsanto C hem ical C om pany, Rubber Service Laboratories D ivision, N itro, W. Va.

W

ciety for Testing M ate­

rials has standardized the con­

ditions for the testing of crude rubber (1) compounds, there are no such a c c e p t e d standards fo r t e s t i n g latex compounds.

Wohler (7) outlined the first constructive ideas on this sub­

ject. The object of this paper

is to continue the work started by Wohler (7) and to suggest simple methods which will give reproducible results. Since latex is a variable material in itself and presents numerous difficulties in testing, it is necessary to employ certain precau­

tions in the testing procedure which will be outlined below.

Ta b l e I. Ty p i c a l La t e x Fo r m u l a

D ry B asis W et B asis R u b b er as 60% cen trifu g ed la te x 1 0 0 .0 166.6

Zinc oxide as a 50% dispersion 1 .0 2 .0

Sulfur as a 5 0 % dispersion 1 .5 3 .0

P ip erid in e cyclopentam efchylene d ith io c a rb a m a te

(P ip -P ip ) 0 .5 0 .5

W ater to 54% to ta l solid c o n te n t . . . 1 8 .9

In Table I is given a typical latex formula which is em­

ployed throughout this paper with the one exception noted below. The zinc oxide and sulfur are added as aqueous dis­

persions, and the piperidine cyclopentamethylene dithiocar­

bamate as an aqueous solution. This formula was chosen on account of its simplicity and to eliminate as many variables as possible.

P r e p a r a tio n o f L a te x T e s t F ilm s

The preparation of smooth uniform latex test films is nec­

essary to obtain reliable results. It is generally customary to prepare the test films by: drying on glass trays in an oven at 45° C. {S), on unglazed tile at room temperature (5), or on glass trays at room temperature (7). These methods were not used here because the glass trays are difficult to clean;

the use of the oven is impractical when comparing a series of compounds and may also cause prevulcanization at 45° C.;

and the unglazed tile will remove water-solubles and will not produce a smooth surface on the film adjacent to the tile.

Experiments proved that ordinary plane glass surfaces were satisfactory and much easier to prepare, the surface tension of any latex compound being sufficient to prevent its flowing over the edges, if the film is not too thick and there are no large chips around the edges of the glass surface.

Therefore, ordinary window-glass plates 25.4 X 33 cm. (10 X 14 inches) are employed. These glass plates are placed on racks (Figure 1), leveled by means of screws, and protected from air

currents by sheets of heavy single­

nap cotton sheeting supported on a wood framework, forming a hood or chamber. A similar method has been suggested by Flint and Naunton (¿).

The compound shown in Table I is allowed to stand about 1 hour to permit the escape of air bubbles, the foam is removed, and the com­

pound is then stirred carefully to prevent possible stratification.

Wohler’s (7) use of burets was not followed because of their inconvenience and possibility of stratifica­

tion; 250-cc. beakers were substituted. Two layers of fine cheese­

cloth (about 36 threads per inch, 14 threads per cm.) are fitted into the beakers, about 225 cc. of the compound are poured into this cheesecloth, and the cheesecloth is removed by withdrawing upward and over the side of the beaker, thus removing foreign material and breaking up all residual air bubbles. The 225 cc.

of latex compound are poured on the glass plates, giving a dry film 1.52 mm. (0.060 inch) thick. Care must be taken to avoid all air currents during the pouring and first 8 hours of drying.

During this first drying period the relative humidity must be maintained at 60 to 70 per cent and the temperature must not be permitted to go below 21.2° C. (70° F.); after 8 hours the films have set sufficiently (although they may contain 15 per cent moisture) so as not to be affected by external conditions.

The chamber is then opened for the free circulation of air to com­

plete the drying process. In approximately another 8 hours the films have dried to a uniform color, which may be taken as an end point, and are removed from the glass plates and suspended to dry at room temperature.

Fi g u r e 1

The elimination of all air currents is essential from the time of pouring the film to the end of the setting period; otherwise a w in k led surface m il be obtained. Relative humidities below 60 per cent cause the formation of a surface skin over the wet film, resulting in deep surface cracks. Tempera­

tures below 21.2° C. (70° F.) require a prolonged setting period and may result in the formation of wrinkles.

The absolute elimination of moisture in the uncured film is necessary to prevent retardation in air cures, but the presence A s im p lifie d m e th o d for o b ta in in g u n i­

fo r m a n d r e p r o d u c ib le p h y s ic a l te s t d a ta o n la te x c o m p o u n d s is d e sc r ib e d . T h e c o n ­ trol o f te m p e r a tu r e a n d h u m id it y fr o m th e p o u r in g o f th e film s to t h e u lt im a t e te s t in g o f th e t e s t str ip is n e c e ssa r y to p r o d u c e

ih e s e r e s u lts .

58 IN D U ST R IA L A N D E N G IN E E R IN G C HEM ISTRY VOL. 10, NO. 2 of small percentages of moisture in the uncured film does not

affect hot water cures. The drying period before cure should not be greater than 48 hours, as shown in Table II.

T a b l e H . T e s t s om U n c u r e d F i l m D ry in g a t 26.6° C . (80° F .) 2 D ay s 5 D ay s 9 D ay s

K g ./ L b ./ K g ./ L b ./ K g ./ L b ./

aq. cm. sq. in . sq. cm. sq. in . sq. cm. sq.in.

M odulus at. 500% 2 0 .4 290 2 6 .7 380 3 2 .3 460

M odulus a t 700% 6 2 .5 890 7 8 .8 1100 10 3 .3 1470 T ensile a t b re a k 1 2 8 .0 1820 151 .3 2150 2 2 5 .0 3200

E lo n g atio n a t b re a k , % 870 860 880

The dried films are each divided into six sections of 7.35 X 13.97 cm. (3.25 X 5.5 inches) in order to obtain a range of cures, and the balance of the film is kept to determine the degree of prevulcanization at the time of testing. The curing of these films can be carried out at any desired temperature or time. The usual procedure with this compound is to carry out cures for 0, 5, 10, 15, 20, 30, and 50 minutes in water at 100° C., or for 0, 10, 20, 30, 50, 90, and 130 minutes in air at 85° C. The films cured in water at 100° C. must be dried at room temperature for about 24 hours to allow them to shrink to a minimum thickness and the hot air-cured films are placed in a desiccator for about 24 hours to prevent their adsorbing moisture from the atmosphere.

The preparation of the latex test pieces is carried out in the manner outlined by the American Society for Testing M a­

terials (1). The die for the preparation of the test pieces m ust have perfect cutting edges (4), giving a clean cut; other­

wise short breaks and variable tensiles and elongations will be obtained. This is more evident in essentially pure-gum stocks than in stocks containing even 5 per cent filler on the rubber.

In order to accomplish this, one die must be used only for pure-gum latex stocks, and this die should be sharpened fre­

quently with a fine stone or razor hone.

The conditioning of the test pieces is markedly more im­

portant when testing latex compounds than when testing crude rubber compounds. Wohler (7) recognized the ab­

solute necessity of controlling the humidity during the con­

ditioning for testing of latex test pieces. Figure 2 shows a photograph of the device used for controlling the humidity for conditioning the test strips.

Fi g u r e 2

Holes are punched in the end of the test pieces which are sus-

E

( 1 . 5 inches) deep containing concentrated sulfuric acid (100 per cent) fill the bottom as completely as possible, while air is circu­

lated by the 15.24-cm. (6-inch) fan driven by a motor at approxi­

mately 400 r. p. m. The cabinet is constructed of a wood frame rabbeted to nt 6.35-mm. (0.25-inch) waterproofed Masonite board, the joints being sealed with several coats of shellac. The

W A T E R C U R E S

1002

A I R C U R E S

% H u m i d i t y 1007.

O 3 5 6 5 IOO

% H u m i d i t y

% H u m id it y

1002

door is fitted like a refrigerator door on soft-rubber-tubing gaskets.

In order to study the effect of humidity on latex test strips, eight films were dried as above. Four films were cured in water at 100° C. and four in air at 85° C., over the standard range of cures. Three test strips from each cure were suspended for 48 hours in desiccators containing vari­

ous concentrations of sulfuric acid (5) so as to give humidities of 0,35, 65, and 100 per cent, respectively (6), at 21.2° C. (70° F.) before testing. The test pieces were removed from the des­

iccator in quantities which could be broken on the test machine in less than 30 minutes, which exposure has shown no effect on physical tests and has shown a saving in time when compared with removing the test pieces individually.

The physical test results on the conditioned test strips, as affected by the various humidities, are shown in Figure 3.

These data were obtained by taking the average of the tensile results, and the average of the 700 per cent modulus results over the complete range of cures at each humidity. The val­

ues at 0 per cent humidity, being the highest ( 7 ) , are taken as 100 per cent, and the values obtained for the higher humidi­

ties are expressed as a certain per cent of the 0 per cent hu­

midity values. Similar results are obtained by taking the values at optimum cure or any time of cure, since the effect of humidity is similar throughout the entire range of cures.

In the water cures the values obtained for 35, 65, and 100 per cent humidity are equal but below the 0 per cent humidity value (Figure 3, /I), while the air cures (Figure 3, B) show a decrease in properties in the order of ascending humidity.

This result probably is due to the extraction of water- adsorbing materials during the water cure.

The differences in the effect of humidity between the air and the water cures were checked by the following method:

O 3 5 6 5 to o

% H u m id ! t y

F i g u r e 3 . T e s t Re s u l t s

7. H u m i d i t y Tcnslfe a t B re a h

M odulus o f 700 TZ E/onyet ton 1007.

FEBR UA RY 15, 1938 ANALYTICAL ED ITIO N 59 A sample of 60 per cent latex which had been purified by

triple centrifuging (2) was compared with the regular 60 per cent latex. This purified latex was very slow in curing with Pip-Pip acceleration in the standard formula; so this test was carried out using the formula in Table III, keeping the quan­

tity of dispersing agents added comparable to that given in Table I by decreasing the sulfur.

Ta b l e III. La t e x Fo r m u l a

D ry B asis W et B asis R u b b e r as 6 0 % cen trifu g ed latex 1 0 0 .0 166 .6

Zinc oxide as a 50% dispersion 1 .0 2 .0

Su lfu r as a 5 0 % dispersion 0 .7 5 1 .5

P ip erid in e c y clo p en tam eth y len e d ith io c a rb a ra ate

(P ip -P ip ) 0 .5 0 .5

D i(b e n z o th ia zy l th io l) d im e th y l u re a 0 .5 1 .0 W a te r to 54% to ta l solid c o n te n t . . . 1 8 .9

Eight test films were prepared from each compound fol­

lowing the procedure given above. Four of each were cured in water at 100° C. and in air at 82° C. over a range of cures approximating those carried out on the formula given in Table I. Test strips for the complete range of cures were conditioned as above and tested. The effect of humidity on the physical tests is shown in Figure 3, C, D, E, F. The regular 60 per cent latex shows the same general trend as the standard test formula in both water (Figure 3, C) and air cures (Figure 3, D). The purified 60 per cent latex shows much less effect of humidity on physical properties of the cured test strips and practically no difference between the water (Figure 3, E) and air cures (Figure 3, F). This is be­

cause the purification process has removed the greater por­

tion of the water-solubles present in the latex. Latex puri­

fied by this centrifuging process is known to have a nitrogen con­

tent below 0.10 per cent and to be very much less hygroscopic.

C o n c lu s io n s

Films should be prepared in an atmosphere free from air currents at 60 to 75 per cent relative humidity and at a tem­

perature of 21.2° to 29.2° C. (70° to 85° F.) during the first 8 hours of the drying period.

These films should be suspended for complete drying for an additional 48 hours at 0 per cent hum idity and 10° to 15.5° C.

(50° to 60° F.) before curing.

The water-cured films should be dried 3 hours in the open room followed by about 20 hours at 0 per cent humidity and 10° to 15.5° C. (50° to 60° F.) before the preparation of test pieces.

The air cures are kept in a desiccator for about 24 hours at 0 per cent humidity and 10° to 15.5° C. (50° to 60° F.) be­

fore the preparation of test pieces.

The die for preparing test pieces must have perfect cutting edges.

Test pieces should be conditioned 48 hours at 0 per cent humidity and 21.1 ° C. (70° F.) before testing to obtain maxi­

mum tensile and modulus results.

The effect of humidity on physical test will be greater on normal latex, Revertex, and stabilized latex because of their increased water-soluble constituents which increase water adsorption. Good averages are obtained from three of four test strips broken, but for accurate work the average of the best four of six test strips should be taken.

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

The author wishes to thank the Monsanto Chemical Com­

pany for permission to publish this information, and the various members of the staff for helpful suggestions in its preparation.

L ite r a tu r e C ited

(1) Am. Soc. T estin g M aterials, D412-35T, p. 7 (October, 1935).

(2) Bogg3 and B lake, I n d . E n o . C h e m ., 28, 1198 (1936).

(3) D u P o n t L ab o rato ry R ep o rt 181 (Decem ber 4, 1934).

(4) F lin t and N aunton, T ram . In st. Rubber Ind., 12, 367 (1937).

(5) Vanderbilt News, 4, No. 5, 20 (1934).

(6) W ilson, J . I n d . E n o . C h e m ., 13, 326 (1921).

(7) W ohler, Ibid., Anal. Ed., 9, 117 (1937).

Re c e i v e d S ep tem b er 29, 1937. P rese n ted before tb e D iv isio n of R u b b er C h em istry a t th e 9 4 th M eetin g of th e A m erican C hem ical Society, R ochester N . Y., S ep tem b er 6 to 10, 1937.

P u rifica tio n o f G raph ite fo r S p e ctr o ch e m ic a l A n alysis

A. H. STAUD1 A N D A. E. RUEHLE, B ell Telephone Laboratories, New York, N. Y.

G

cially but its impurities are of such a nature that its use in spectrochemical work is somewhat limited. For prac­

tically all qualitative and many quantitative analyses, it is highly important that the impurities commonly present in this material be entirely eliminated or reduced to the merest traces. A purer but much more expensive grade is also ob­

tainable, but this is less uniform with regard to its impurities than the product obtained by the method described here.

Ta b l e I .

O riginal Im p u ritie s A lum inum B oron C alcium C opper Iro n M agnesium

Im p u r i t i e s i n Gr a p h i t e b e f o r e a n d a f t e r Su l f u r i c Ac i d Tr e a t m e n t

A fter T re a tm e n t N one U n changed T race

N one to fa in t trace N one to fa in t trace T race

O riginal Im p u ritie s M anganese Silicon Silver Sodium T ita n iu m V an ad iu m

A fter T r e a tm e n t N one U nchanged N one T race N one N one

Standen and Kovach (1) have described two methods of purification which they have found to be satisfactory. Both methods, however, appear to be somewhat more involved and time-consuming than the procedure described in this paper.

1 P re s e n t address, 101 M aiden Lane, N ew Y o rk , N . Y .

The authors have found sulfuric acid more effective in the washing than nitric acid, hydrochloric acid, or aqua regia.

Attem pts to remove silicon with hydrofluoric acid were un­

successful.

P r o c e d u r e

Cut the graphite into the desired form for use as electrodes.

Heat in a silica dish over an oxy-gas burner to redness. Cool and place in a flask fitted with a reflux condenser. Cover the elec­

trodes with 1 to 1 sulfuric acid and boil on a hot plate for at least 24 hours. Wash by decantation with distilled water until the water is no longer acid to litmus, then boil 15 minutes in a fresh portion of water. Again wash by decantation and repeat the boiling. Continue the alternate washing and boiling until acid is no longer extracted. Usually four such operations will suffice. Transfer to the silica dish and heat to bright redness, allowing the flame to play directly on the electrodes. Cool and store in capped bottles until used.

Table I shows the effectiveness of the treatment.

L ite r a tu r e C ited

(1) S tan d en an d K ovach, Proc. A m . Soc. Testing M aterials, 35, 79 (1935).

Re c e i v e d D ecem ber 28, 1937.

A nalysis o f C om m ercial P h en o th ia zin e Used as an Insecticide

L. E. SMITH, Bureau o f E ntom ology and Plant Q uarantine, U. S. D epartm ent o f A griculture, W ashington, ü . C.

I

tance, a great number of organic compounds have been pre­

pared and tested. One compound, phenothiazine, has been found especially toxic to newly hatched codling moth larvae under laboratory conditions. It has also given a high degree of control of this insect in-field tests in the Northwest, but because of certain practical difficulties it is not yet in com­

mercial use.

The method for the preparation of phenothiazine on a com­

mercial basis consists in heating 1 mole of diphenylamine with 2 atoms of sulfur at about 180° C., using iodine as a catalyst.

The reaction is practically quantitative, and for use as an insecticide no purification of the product is necessary.

However, a dark green compound, which is insoluble in anhydrous ethyl ether, is formed in varying quantities. When tested against certain species of insects, this compound has been found to be relatively nontoxic.1 Its chemical nature has not been fully determined, but it appears to be isomeric with, or a polymer of, phenothiazine.

Calculated for CnHjNS: C, 72.36; H, 4.52 Found: C, 71.41; H, 4.31

The insolubility of this green material in anhydrous ethyl ether is utilized in analyzing commercial phenothiazine. A

1 T hese resu lts will ap p ea r as a scientific n o te in th e Jo urnal o f Economic Entomology.

weighed amount of the compound is placed in a tared Soxhlet thimble and extracted with ether in the usual manner. The residue, which consists of the green material, is then deter­

mined from the increase in weight of the dried thimble.

Little or no unchanged diphenylamine has been found in samples of commercial phenothiazine. Diphenylamine is pre­

cipitated almost quantitatively, however, from an anhydrous ethyl ether solution by means of dry hydrogen chloride. This precipitate can be filtered on a tared Gooch crucible, washed with anhydrous ethyl ether, dried, and weighed. B y this procedure, 97.2 per cent recovery, as the hydrochloride, was obtained from an ether solution containing a known quantity of diphenylamine.

Six samples of phenothiazine that were used in various field tests the past season were submitted to the Division of Insecticide Investigations for analysis. All material was purchased from the same manufacturer. The results are as follows:

Sam ple Insoluble Sam ple Inso lu b le

No. in Ether® No. in Ether®

% %

1 1 .1 4 4 1 .3 4

2 1 .43 5 1.0 9

3 1.21 6 1 .2 3

° T h e analyses were c arried o u t b y M iss R u th C ap e n of th e D ivision of Insecticide In v e stig a tio n s.

Re c e i v e d D ecem b er 2, 1937.

D eterm in ation o f Iron w ith o-P henanthroline

A Spectropliotometric Study

W. B. FORTUNE W I T H M. G. MELLON, Purdue University, Lafayette, Ind.

I

scription of the properties of o-phenanthroline, along with the method of preparation (1). Say well and Cunningham re­

ported recently that one can make a quantitative determina­

tion of iron in small concentrations in various fruit juices and other products by comparing the color of the o-phenanthroline complex with that of a series of standards (5). Hummel and Willard (1 A ) have applied the method to biological materials.

The purpose of the present paper is to present a critical study of various factors which m ay affect the formation of the colored complex, including a study of the effect of varying concentrations of fifty-five ions liable to be encountered in routine analysis.

A p p a r a tu s a n d M e th o d s

In the determination of iron in fruit products, as carried out by Saywell and Cunningham (5), the color comparisons were madę in graduated test tubes and in a colorimeter. The develop­

ment of the photoelectric spectrophotometer, described by Michaelson and Liebhafsky (4), has provided a means of detect­

ing very small color changes with a much higher degree of pre­

cision and accuracy than is possible with visual methods. A General Electric recording instrument was used in all trans- mittancy measurements in this work. The cells were 1.00 cm.

thick, the “blank” in the reference beam of light being filled with a solution containing the same amount of hydroxylamine hydro­

chloride and o-phenanthroline as was used with the iron in the other cell.

All pH measurements were made with a “universal" potenti­

ometer and glass electrode, as described by Mellon (2).

The standard solution of iron was prepared by dissolving electrolytic iron wire in dilute hydrochloric, nitric, or perchloric acid. The solutions were then diluted to volumes such that 1.00 ml. contained 0.100 mg. of iron. A 0.10 per cent solution of o-phenanthroline was prepared by dissolving the monohydrate in doubly distilled, iron-free water. Saywell and Cunningham used an ethanol solution but the authors found that the reagent dis­

solved readily in water heated to about 80° C. It is important that the o-phenanthroline monohydrate be free from impurities.

Certain contamination, at least, is evidenced by a pink coloration of the crystalline material, and a lowering of the melting point, stated by Smith (6‘) to be 99° to 100° C. A 10 per cent solution of hydroxylamine hydrochloride, used as a reducing agent for the iron, was prepared by dissolving the c. P . reagent in doubly distilled water. Solutions used in the determination of inter­

fering cations were prepared from the chloride or nitrate salts of the metals; the anion solutions were prepared from the sodium or potassium salts.

In making up all colorimetric solutions used in this study, the following procedure was adopted: The required amount of the standard iron solution was measured out; 1.0 ml. of the

FEBR U A R Y 15, 1938 A NALYTICAL E D ITIO N 61 hydroxylamine hydrochloride was added to reduce the ferric

iron; the solution was diluted to approximately 75 ml.; an excess of o-phenanthroline solution was added, 5.0 ml. being used with iron concentrations up to 4.0 p. p. m. and 10.0 ml. with higher concentrations; and the resulting solution was diluted to 100 ml.

The pH value was then determined and any desired adjustment in acidity was made. The volumes of acid or base required for this adjustment never amounted to more than 0.1 ml. of 6 N hydrochloric acid or G Ar ammonium hydroxide, thus keeping possible error from dilution sources below 0.1 per cent. Solutions of possible interfering ions were added before the o-phenanthro­

line color was developed.

Calculations of apparent error due to interference by ions were made from transmittancy measurements by means of

Ci

Beer’s law, T l — T f ' , where C2 is the concentration of iron in the standard solution and G'i is the calculated concentration after the possible interfering ion has been added. These cal­

culations were made by means of a special color slide rule.

Q _ Q

The error was calculated thus: —^ — 1 X 100 = per cent

"apparent error.” An apparent error of 2 per cent was arbitrarily set as the limit of negligible interference in the case of the added ions. Since visual methods of color comparison often have a precision not less than 5 per cent, it was necessary to set a lower figure in order to provide for other possible factors.

T h e C olor R e a c tio n

The colored complex ion formed between the o-phenanthro- line and the ferrous ion has been postulated by Blau (1) to be composed of three molecules of o-phenanthroline and one ferrous ion. The intensity of the color produced is determined by the amount of iron when there is an excess of the o-phen­

anthroline reagent present. It is this direct relationship be­

tween the iron and the intensity of the color which permits the use of the method.

The transmittancy curves for varying amounts of iron are shown in Figure 1. The peak of the absorption band is located at 508 mu, with a secondary band shown at 474 m/i.

Six parts per million of iron was found to be the maximum concentration which could be used with a 1.00-cm. cell. A concentration of 0.10 p. p. m. gave a minimum transmittancy of 95.5 per cent at this thickness. It would be possible to use a less or more highly concentrated solution in a color com­

parator where very thick or very thin layers of solution might be used for comparative purposes.

A study of the minimum amount of a 0.10 per cent solution of o-phenanthroline required to produce the maximum color revealed that 6.0 ml. of the reagent were required for every 5 p. p. m. of iron present. Any amount less than this did not produce complete development of the color. This study was made by keeping the amount of iron constant and varying the total amount of o-phenanthroline added.

Transmittancy curves for solutions wherein the color was fully developed and the pH varied from 2.0 to 9.0 were exactly superimposed upon one another, showing that over this pH range the color was not dependent on the pH of the solution.

The conformity of the colored solution to Beer’s law was tested over the range from 0.10 to 6.00 p. p. m. of iron, the limits found applicable for a 1.00-cm. cell, by plotting 2 + log T against concentration. A straight line indicated very close conformity. Further tests were carried out by deter­

mining transmittancy curves on solutions in 1.00-cm. cells, diluting the solutions to exactly twice the original volume and again determining the transmittancy in 2.00-cm. cells.

The two curves were superimposed on each other.

Accelerated, as well as ordinary, fading tests agreed well with Saywell and Cunningham’s conclusions that the color is stable for at least 15 days. On one test, the solutions were

sealed in test tubes and exposed to varying degrees of light over a period of 11 days. The transmittancy curves on the original solutions and the same solutions after exposure w'ere identical. In making the accelerated fading tests, the solu­

tions were placed in glass-stoppered Pyrex bottles of about 60- ml. capacity after the transmittancy curves wrere determined.

These bottles were then placed in an air thermostat at 30° C.

at a distance of 60 cm. from a mercury arc lamp for a period of 100 hours. Redetermination of the transmittancy curves indicated no fading. These solutions were then set aside for 150 days, when the transmittancy curves again agreed well within experimental error.

R e d u c in g A g e n ts

Following the original work of Saywell and Cunningham, the authors used a 10 per cent solution of hydroxylamine hydrochloride as the reducing agent in all work on interfering ions and in concentration tests. Because of the relatively high cost of the reductant, it was decided to try certain other agents. A 10 per cent aqueous solution of sodium sulfite was tried with very unsatisfactory results. In slightly acid solu­

tion a brown colored complex wras formed. The presence of the brow'n coloration proved to be a serious interference when the o-phenanthroline color was developed, causing errors varying from 4.5 to 37.5 per cent.

Sodium and potassium formates (analytical reagent) were found to be free from iron in appreciable quantities. How­

ever, determinations using these materials as reducing agents showed appreciable error. According to Mellor (8), this is probably due to the formation of a complex between the ferric ion and the formic acid foimed in the solution.

Formaldehyde was used, but again a complex was formed between the iron and the reagent, causing appreciable error.

The hydroxylamine hydrochloride solution was found to be the m ost satisfactory for the reduction of the iron. The

62 IN D U ST R IA L A N D E N G IN E E R IN G CHEM ISTRY VOL. 10, NO. 2 relatively small amounts used and the speed with- which the.

iron was reduced are sufficient to warrant the use of this

reagent. A

E ffe c t o f I o n s o n C olor D e v e lo p e d

Studies on the possible interference of ions were made in all cases on a solution containing 2.00 p. p. m. of iron. The iron was reduced with 1.0 ml. of a 10 per cent solution of hydroxylamine hydrochloride, the ion in question was intro­

duced, the o-phenanthroline was added, and the solution was diluted to the mark. After mixing, the pH was adjusted with ammonium hydroxide or hydrochloric acid to a point within the range of applicable pH values shown in Tables I and II, and the transmittancy curves were determined. These curves were then compared fo th the standard for the same amount of iron without the added ions, and the per cent

“apparent error” was determined as stated above.

It,w a s decided that, with 2,00 p. p. m. of iron present, 500 p. p. m. of ion (250 times the concentration of iron in the solution) should be sufficient' to test for interference. If this amount does not interfere, it is very improbable that the ion will interfere in any quantity.

E f f e c t o f C a t i o n s . Of the large number of transrhittancy curves obtained for cations, only certain representative ones are shown in Figure 2. They illustrate typical types of inter­

ference.

Table I gives the maximum concentration of ions and the applicable pH ranges which may be used without error in the determination of 2.00 p. p. m. of iron.

T a b l e I. E f f e c t o f C a t i o n s M ax im u m M ax im u m

Io n A lum inum A m m onium A n tim o n y A rsenic A rsenic B ariu m BerylLium B ism u th C ad m iu m C alcium C hrom ium C o b alt C opper LeadL ith iu m

M agnesium M anganese M ercu ry M ercu ry M o ly b d en u m N ickel P o tassiu m S ilver Sodium S tro n tiu m T h o riu m T in T in T u n g ste n U ra n iu m ZincZ irconium

A dded as AlCla NH<C1 SbCla AstO*

As îOj BaCla Be(NOa)i Bi(NO«)a Cd(NO«)t C a(N O j)i Cri(SO<)*

C o(N O j)j C u(N O j)*

P b(C jH ,O i)*

L iC l M g(N O a)i MnbO*

H gC lj Hg*(NOa)j

(NHO*M070i4

N i(N O a)t KC1 AgNOa N aC l Sr(NOa)*

Th(N O j)*

HjSnCl«

HjSnCl*

Na,WO<

UOj(CjH»Oi)j Zn(N Oa)i Zr(NOa)*

C oncen­

tra tio n , In te r­

ference, A pplicable p H R an g e p. p. m. % F e

5 0 0 .0 N one 2 . 0 - 3 .0 2 5 0 .0 1 .4 2 . 0 - 5 .0 5 0 0 .0 N one 2 . 0 - 9 .0 3 0 .0 N one 3 . 0 - 9 .0 5 0 0 .0 N one 3 . 0 - 9 .0 5 0 0 .0 N one 3 . 0 - 9 .0 5 0 0 .0 N one 3 . 0 - 9 .0 5 0 0 .0 1 .3 3 . 0 - 5 .5

N one

5 0 .0 i ! o ° 3 . Ô -9.0 5 0 0 .0 N one 2 . 0 - 9 .0

20.0 N one 2 . 0 - 9 .0

10.0 1 .5 3 . 0 - 5 .0

10.0 N one 2 . 5 - 4 .0

5 0 0 .0 N one 2 . 0 - 9 .0 5 0 0 .0 N one 2 . 0 - 9 .0 5 0 0 .0 N o n e 2 . 0 - 9 .0 5 0 0 .0 N one 2 . 0 - 9 .0

1.0 N one 2 . 0 - 9 .0

10.0 N o n e 3 . 2 - 9 .0

100.0 N one 5 . 5 - 9 .0

2.0 N one 2 . 5 - 9 .0

1000.0 N one 2 . 0 - 9 .0 N one

1000.0 N o n e 2. Ô -9 .0 5 0 0 .0 N one 2 . 0 - 9 .0 2 5 0 .0 1 .5 2 . 0 - 9 .0

20.0 N one 3 . 0 - 6 .0

5 0 .0 N one 2 .5

10.0 N one 2.0 - 6 .0

20.0 N one 2 . 0 - 3 .0

10.0 N egligible 2 . 5 - 9 .0

100.0 N one 2.0 - 6 .0

10.0 N one 2 .0 - 9 .0

5 0 .0 1.8 2 . 0 - 9 .0

100.0 2.2 3 . 0 - 9 .0

° 15.0 m l. of o -p h en an th ro lin e in excess of th e original a m o u n t added.

Iro n re d u ced w ith h y d ro x y la m in e h y d ro ch lo rid e in all te sts.

Potassium and sodium showed no interference when present in quantities as high as 1000 p. p. m. The following cations may be present in concentrations as high as 500 p. p. m. over the applicable pH range without interference: ammonium, arsenic (as arsenate or arsenite), barium, calcium, lead, lithium, magnesium, manganese (as manganous ion), and strontium.

Chromic ion changed the hue of the solution by absorbing somewhat in the red and violet. There was no apparent

/V a y e L e n g t h

error in the minimum of the transmittancy curve with con­

centrations up to 50 p. p. m. However, at this concentration, the hue was quite different. At 20 p. p. m., there was very little change in hue and the solutions could be compared visually.

Bismuth and silver m ust be completely absent from the solution because of the formation of precipitates. In each case, the ion causing the precipitate came from the o-phe­

nanthroline solution. A t concentrations up to 30 p. p. m. of antimonous ion, no interference was noted. Concentrations above this precipitated insoluble basic salts. Beryllium showed no interference in concentrations up to 50 p. p. m.

when the pH was kept between 3.0 and 5.5. Above pH 5.5 the hydroxide formed, and below 3.0 a complex was formed w ith the o-phenanthroline. Cadmium, mercuric, and zinc ions also formed precipitates with the o-phenanthroline. With small amounts of these ions it was possible to prevent ap­

preciable interference by adding excess o-phenanthroline.

W ien 15 ml. excess of reagent were added to 50 p. p. m.

of cadmium ion, the interference dropped to 1.0 per cent.

Ten parts per million of zinc ion appeared to be the limiting concentration without appreciable error. One part per million was set as the maximum concentration of mercuric ion which could be present.

pH was a very important factor in the interference of molybdenum, present as molybdate ion. At a pH of 4.0, a milky solution resulted with as little as 10 p. p. m. of molybdenum. Thirty parts per million caused serious error at a pH of 4.5, but did not interfere at a pH of 5.0. A t pH values above 5.5,100 p. p. m. of molybdenum could be present without appreciable interference.

Nickel ion produced a change in the hue of the solution.

Apparently a complex was formed with the ferrous ion, the nickel ion, and the o-phenanthroline solution. A t all wave lengths below 540 m/x the transmittancy was higher than normal, in proportion to the amount of nickel present,