Skip to main content

Full text of "Journal of the American Chemical Society. Supporting information"

See other formats


Google 


This  is  a  digital  copy  of  a  book  that  was  preserved  for  generations  on  library  shelves  before  it  was  carefully  scanned  by  Google  as  part  of  a  project 

to  make  the  world's  books  discoverable  online. 

It  has  survived  long  enough  for  the  copyright  to  expire  and  the  book  to  enter  the  public  domain.  A  public  domain  book  is  one  that  was  never  subject 

to  copyright  or  whose  legal  copyright  term  has  expired.  Whether  a  book  is  in  the  public  domain  may  vary  country  to  country.  Public  domain  books 

are  our  gateways  to  the  past,  representing  a  wealth  of  history,  culture  and  knowledge  that's  often  difficult  to  discover. 

Marks,  notations  and  other  maiginalia  present  in  the  original  volume  will  appear  in  this  file  -  a  reminder  of  this  book's  long  journey  from  the 

publisher  to  a  library  and  finally  to  you. 

Usage  guidelines 

Google  is  proud  to  partner  with  libraries  to  digitize  public  domain  materials  and  make  them  widely  accessible.  Public  domain  books  belong  to  the 
public  and  we  are  merely  their  custodians.  Nevertheless,  this  work  is  expensive,  so  in  order  to  keep  providing  tliis  resource,  we  liave  taken  steps  to 
prevent  abuse  by  commercial  parties,  including  placing  technical  restrictions  on  automated  querying. 
We  also  ask  that  you: 

+  Make  non-commercial  use  of  the  files  We  designed  Google  Book  Search  for  use  by  individuals,  and  we  request  that  you  use  these  files  for 
personal,  non-commercial  purposes. 

+  Refrain  fivm  automated  querying  Do  not  send  automated  queries  of  any  sort  to  Google's  system:  If  you  are  conducting  research  on  machine 
translation,  optical  character  recognition  or  other  areas  where  access  to  a  large  amount  of  text  is  helpful,  please  contact  us.  We  encourage  the 
use  of  public  domain  materials  for  these  purposes  and  may  be  able  to  help. 

+  Maintain  attributionTht  GoogXt  "watermark"  you  see  on  each  file  is  essential  for  in  forming  people  about  this  project  and  helping  them  find 
additional  materials  through  Google  Book  Search.  Please  do  not  remove  it. 

+  Keep  it  legal  Whatever  your  use,  remember  that  you  are  responsible  for  ensuring  that  what  you  are  doing  is  legal.  Do  not  assume  that  just 
because  we  believe  a  book  is  in  the  public  domain  for  users  in  the  United  States,  that  the  work  is  also  in  the  public  domain  for  users  in  other 
countries.  Whether  a  book  is  still  in  copyright  varies  from  country  to  country,  and  we  can't  offer  guidance  on  whether  any  specific  use  of 
any  specific  book  is  allowed.  Please  do  not  assume  that  a  book's  appearance  in  Google  Book  Search  means  it  can  be  used  in  any  manner 
anywhere  in  the  world.  Copyright  infringement  liabili^  can  be  quite  severe. 

About  Google  Book  Search 

Google's  mission  is  to  organize  the  world's  information  and  to  make  it  universally  accessible  and  useful.   Google  Book  Search  helps  readers 
discover  the  world's  books  while  helping  authors  and  publishers  reach  new  audiences.  You  can  search  through  the  full  text  of  this  book  on  the  web 

at|http: //books  .google  .com/I 


HARVARD  UNIVERSITY 


LIBRARY  OF  THE 

BIOLOGICAL  LABORATORIES 
HARVARD  COLLEO!!  I.'P'^' 5'' 


WAY  4     194S 


THE  JOURNAL 


—OP  THK— 


AMERICAN  CHEMICAL  SOCIETY. 


VOLUME  XVIII. 


1896. 


COMMITTEE  ON  PAPERS  AND  PUBLICATIONS: 

Edward  Hart,  Editor, 
J.  H.  Long, 
Thomas  B.  Osborns. 


Photolithograph  Reproduction 
By  Permission  of  The  American  Chemical  Society 


JOHNSON  REPRINT  CORPORATION 
New  York,  N.  Y.,  U.  S.  A. 


kcff  /•-■/ 


•  ^^ 


HARVARD 

UNIVERSITY 

LIBRARY 


copyright,  1896, 

By  Edward  Hart,  John  H.  Long, 
AND  Thomas  B.  Osborne. 

Committee  on  Papers  and  Publications  of  thcT 
American  Chemical  Society. 


PHCrrOLITHOGRAPHED  BY 

THE  MURRAY  PRINTING  COMPANY 

WAKEFIELD,  MASSACHUSETTS 


Vol.  XVIII.  [August,  1896.]  No.  8. 


THE  JOURNAL 


OF  THE 


AMERICAN  CHEMICAL  SOCIETY. 


PHOTOMETRIC  METHOD  FOR  THE  QUANTITATIVE  DETER- 
niNATION  OF  LIHE  AND  SULPHURIC  ACID.' 

By  J.  I.  D.  Hinds. 

Received  May  14,  itgd. 

THE  want  of  a  rapid  method  of  determining  with  a  close 
approximation  the  amount  of  lime  and  sulphuric  acid  in 
drinking  water  led  me  to  the  study  of  the  opacity  of  fine  white 
precipitates  suspended  in  water.  I  precipitated  in  weak  solu- 
tions lime  with  ammonium  oxalate,  and  sulphuric  acid  with 
barium  chloride,  then  measured  the  height  of  a  column  of  the 
liquid  containing  the  precipitate  through  which  the  flame  of  a 
common  candle  was  just  invisible.  I  expected  only  a  rude 
approximation,  but  to  my  surprise,  I  found  that  between  cer- 
tain limits,  an  accuracy  is  attainable  equal  to  that  of  the  ordi- 
nary volumetric  methods. 

APPARATUS. 

The  only  apparatus  needed  is  a  cylinder  graduated  from  the 
bottom  in  centimeters  and  tenths.  The  cylinder  should  have  a 
plain  polished  bottom,  like  Nessl'er  cylinders,  and  should  have  a 
lip  at  the  top.  The  one  I  use  was  made  for  me  by  Eimer  and 
Amend.  It  is  four  cm.  wide  and  twenty  cm.  high.  The  gradu- 
ations runs  to  eighteen  cm.  This  cylinder,  however,  is  not 
absolutely  necessary.  A  common  beaker  may  be  used  and  the 
depth  of  the  liquid  measured  with  a  small  ruler. 

1  The  manuscript  of  this  srticle  was  sent  similtaneously  to  the  Chemical  News  and 
to  this  Jonmal.  Owing  to  the  absence  from  home  of  Professor  Hinds,  his  proof  was  de- 
layed too  long  to  allow  of  publication  of  the  article  in  the  July  issue. 


/.■-  i 


B'* 


Com* 


^  /^ 


O.OS73 
0.0570 
o.05« 


0.0570 
Itlution    of  the 
very  slightly. 
)re  dilute  with 


0.0595 
0.0589 
0.0597 
0.0591 
0.0591 


0.0S91 

0.0594 
....  O.OS93 
....  0.0587 
....  0.0590 
imber  obtained 
the  solution  by 
:  is  just  invisi- 
by  taking  the 
is  an  hyperbola 
is 


liagram.     The 
oi  per  cent,  to 


lution,  observe 


/  \  'J-  ' 


f        / 


/ 


HARVARD  UNIVERSITY 


LIBRARY  OF  THE 
BIOLOGICAL  LABORATORIES 

HARVARD  COLLEC!^  I'P^'  ^'v 


MAY  4    1949 


666  J.  I.  D.  HINDS. 

cipitate  the  whole  of  the  calcium.  The  solution  was  then  poured 
into  the  photometric  cylinder  and  the  depth  measured  as  in  the 
case  of  sulphuric  acid.  Portions  of  ten  or  twenty  cc.  of  water 
were  successively  added  and  the  depth  observed  after  each  addi- 
tion. The  results  are  given  in  the  following  table.  In  column 
I  is  the  number  of  the  solutioti ;  column  2  shows  the  per  cent, 
of  calcium  carbonate ;  columns  3,  4,  5,  6,  and  7  contain  the 
measured  depths  of  the  liquid  at  which  the  flame  became  invisi- 
ble ;  column  8  contains  the  means  of  these  depths,  and  column 
9  the  product  of  these  means  by  the  per  cents,  in  column  2,  rep- 
resented as  before  by  xy.  The  three  determinations  in  the 
fifth  series  were  made  simply  as  a  check.  Many  other  indepen- 
dent determinations  were  made  in  order  to  ascertain  whether 
there  was  a  change  of  opacity,  and  whether  the  precipitation 
would  be  different  in  the  weaker  solutions.  No  material  differ- 
ence was  found. 

Per  cent, 
calcium 
No.     carbonate.        cm.  cm.  cm.  cm.  cm.  x.  xy. 

1 0.0333  2.1  2.3  2.3        '    2.4  2.250         0.0750 

2 0.0250  2.8  2.9  2.9  2.9  2.875  0.0718 

3 o.oioo  3.5  3.6  3.5  3.5          3.525  0.0705 

4 0.0167  4.1  4.2  4.1  4.1  4.2    4.14  0.0691 

5 0.0143  4-7  4-8  4.7  4.7          4.725  0.0676 

6 0.0125  5.3  5.5  5-3  5.3          5-35  0.0669 

7 o.oiii  6.0  6.1  5.9  6.0          6.0  0.0666 

8 0.0100  6.6  6.8  6.6  6.7          6.675  0.0668 

9 0.0091  7.3  7.4  7.3  7.4  7-4    7-36  0.0670 

10 0.0083  8.0  8.0  8.0  8.1                        8.03  0.0666 

II 0.0077  ^-8  8.6  8.6  8.8                       8.7  0.0670 

12 0.0071  9.5  9.3  9.3  9.5                       9.4  0.0667 

13 0.0067  10.2  0.9  9.9  10. 1  9.9        lo.o  0.0670 

Examining  the  values  of  xy,  we  find  that  they  are  not  con- 
stant. They  diminish  rapidly  at  first,  then  more  slowly.  The 
equation  is,  therefore,  not  so  simple  as  in  the  case  of  sulphuric 
acid.  It  appears,  however,  to  be  an  hyperbola,  and  we  may 
assume  that  its  equation  has  the  form 

xy-^by^a, 

in  which  b  and  a  are  constants  whose  values  are  to  be  deter- 
mined. Substituting  the  values  of  x  and^  from  the  above  table, 


UMB  AND  SUIrPHURIC   ACID.  667 

we  obtain  thirteen  observation  equations.  The  values  of  a  and 
6  are  then  found  according  to  the  method  of  least  squares  by 
forming  and  solving  the  two  sets  of  normal  equations.  The  first 
set  will  be  the  same  as  the  observation  equations  ;  the  second 
set  is  obtained  by  multiplying  each  equation  by  its  cofficient  of 
d.     These  equations  are  as  follows : 

0.0750  +  O.Q333  d  ==  a  0.002500  +  o.ooiiii  d  =s  0.0333  a 

Q.0718  +  0.0250  b  =s  a  0.001795  4-  0.000625  d  =  0.0250  a 

0.0705  +  o.oaoo  b  =^a  0.001410  4  0.000400  d  =  0.0200  a 

0.0691  +  0.0167  6  ss  a  0.001151  -h  0.000271  b  =  0.0167  a 

0.0676  +  0.0143  b  =s  a  0.000967  +  0.000204  b  =  0.0143  a 

0.0669  +  0.0125  b  ssa  0.000836  4-  0.000156  b  ^  0.0125  a 

0.0666  +  o.olii  b  ^a  0.000740  +  0.000124  b  3B o.oiii  a 

0.0668  +  o.oioo  b  =s  a  0.000668  4*  o.oooioo  b  ss  o.oioo  a 

0.0670  +  0.0091  b^a  0.000610  +  0.000083  b  =  0.0091  a 

0.0666  -f-  0.0083  b^sza  0.000553  +  0.000069  b  ^  0.0083  « 

0.0670  +  0.0077  b-^a  0.000516  -+-  0.000059  b  =  0.0077  a 

0.0667  +  0.0071  b  ==  a  0.000474  -f-  0.000050.  b  =  0.0071  a 

0.0670  +  0.0067  ^  =  a  0.000449  +  0.000045  b  ^  0.0067  ^ 

Adding  the  equations  together,  we  have 

0.8886  +  0.1818*=  13a.    0.012668-1-0.003304*  =  0.18 i8a. 
Dividing  by  the  coefBicient  of  a  and  eliminating,  we  have 

a  =  0.0642  *  =  —  0.3 

The  required  equation  is  therefore 

xy  —  0.3  *  =  0.0642, 

or,  solving  for^ 

0.0642 


X  —o,^ 

For  the  per  cent  of  CaO  the  equation  is 

0.0360 

^= — . 

^  —  0.3 

This  is  the  equation  of  an  hyperbola  referred  to  one  of  its 
asymptotes  as  the  axis  of  x  and  to  an  axis  of  y  three-tenths  cm. 
to  the  left  of  the  other  asymptote.  The  abscissas  are  centime- 
ters and  the  ordinates  are  o.oi  per  cent,  to  the  cm.  The  curves 
are  shown  in  the  accompanying  diagram. 

As  an  example,  let  us  suppose  that  the  observed  depth  is  four 
and  seven-tenths  cm.     Subtract  0.3  and  divide  0.0642  by  the 


;.  I.  D.  HINDS. 


DUcram  i. 
remainder.  The  quotient  0.0146  is  the  per  cent,  of  calcium  car- 
bonate. Dividing  this  by  1000  we  have  14.6  parts  to  the  100,000. 
PROBABLE  ERROR. 
To  determine  the  probable  error  of  an  observation  we  may 
compare  as  before  the  numbers  fdund  by  observation  with  those ' 
computed  from  the  equation,  as  follows  : 

StrcDKth  Slrencih 

X.  us«d.  computed.  v,  i^. 

3.9       0.0350     O.oa47     0.0003  0.00000009 

3.5       0.0100  o.oaor     o.oooi     o.oooooooi 

4.1       0.0167     0.0170     0.0003 


UMK   AND  SULPHURIC  ACID. 

669 

Strength, 
used. 

strength. 

jr. 

computed. 

V. 

t4. 

4.7 

0.0143 

0.0146 

0.0003 

0.00000009 

5-35 

0.0125 

0.0127 

0.0002 

0.00000004 

6.0 

O.OI  1 1 

O.OI 12 

O.OOOI 

O.OOOOOOOI 

6.7 

O.OIOO 

O.OIOO 

0.0000 

0.00000000 

7-4 

0.0091 

0.0091 

0.0000 

0.00000000 

8.0 

G.0083 

o.oo8>; 

0.0002 

O.OOOOOOQ4 

8.7 

0.0077 

0.0077 

o.oono 

0.00000000 

9-4 

0.0071 

0.0071 

0.0000 

0.0000000(> 

lO.O 

0.0067 

0.0066 

O.OOOI 

Sum  2t^ 

0.00000901 
0.00000038 

Using  the 

same  value  for  error  as 

before,  in  which  in  this  case 

n,  the  number  of  observations,  is  12 

,  and  ^»  the  number  of  con- 

stants  in 

the  equation  is 
r=  0.6745^. 

2,  we  have 

=  0.00013  per  cent. 

0.000P0038 
12 — 2 

That  is,  the  probable  difference  between  an  observed  and  com- 
puted strength  of  a  solution  is 0.00013  percent.,  or  thirteen  parts 
in  ten  million. 

SOURCES  OF   ERROR. 

The  principal  sources  of  error  in  this  method  are  two.  In  the 
first  place  a  light  of  constant  intensity  should  be  used.  It  makes 
but  little  difference  what  the  light  is,  so  it  is  the  same  as  that 
with  which  the  constant  in  the  equation  is  determined.  I 
employed  the  flame  of  an  ordinary  candle  as  the  most  con- 
venient. A  brighter  and  steadier  light  would  give  better  results. 
Any  change  of  light  will  of  course  change  the  constants. 

The  second  source  of  error  is  the  personal  equation.  Each 
individual  can  determine  this  for  himself.  The  error  dependent 
upon  the  eye  can  be  almost  eliminated  by  using  it  in  the  usual 
way,  that  is  with  or  without  glasses. 

A9y  one  can  obtain  the  constants  for  himself  by  making  a 
few  determination  with  solutions  of  known  strength.  The  best 
strength  to  use  is  that  between  o.oi  and  0.03  per  cent.  Great 
care  must  be  used  in  measuring.  If  ten  cc.  of  a  decinormal 
solution  are  taken,  a  difference  of  one  drop  in  the  measurement 
may  make  an  error  ten  times  as  gpreat  as  that  involved  in  the 
method. 


670  HERMANN   FI^ECK.      SEPARATION  OF 

PRACTICAL  APPLICATION. 

I  have  so  far  used  the  method  and  tested  it  only  in  sanitary 
water  analysis  and  in  the  analysis  of  urine.  To  the  water  analyst 
it  will  be  of  great  value.  It  gives  the  lime  and  sulphuric  acid 
with  almost  the  accuracy  of  the  gravimetric  method.  It  is  more 
accurate  than  the  soap  test  and  is  but  slightly  affected  by  the 
presence  of  magnesium  salts. 

For  determining  the  sulphuric  acid  in  urine  I  have  found  it 
quite  satisfactory.  The  urine  has  to  be  diluted  with  nine  vol- 
umes of  water  and  then  the  color  does  not  sensibly  affect  the 
determination. 

I  see  no  reason  why  this  method  may  not  be  successfully  used 
with  all  fine  white  precipitates.  It  is  not  suitable  for  precipi- 
tates that  settle  rapidly  or  gather  quickly  into  flakes.  Whether 
colored  precipitates  may  be  determined  in  this  way  is  still  to  be 
investigated. 

I  desire  to  acknowledge  obligation  to  Professor  A.  H.  Buchanan 
for  assistance  in  determining  the  equations  and  probable  errors. 

Chemical  Laboratory,  Cumberland  University, 

Lebanon,  Tbnn. 


[Contributions  from  John  Harrison  Laboratory  op  Chemistry. 

No.  12.] 

THE    SEPARATION    OF    TRIHETHYLAMINE    FROil 

AnMONIA. 

By  Hermann  Fleck. 
Received  May  8.  XB96. 

THE  quantitative  estimation  of  trimethylamine  in  presence  of 
ammonia  is,  I  believe  not  mentioned  in  the  literature, 
although  a  number  of  publications  have  appeared  in  which  the 
detection  of  trimethylamine,  in  presence  of  ammonia,  by  means 
of  the  different  solubilities  of  their  hydrochlorides  in  absolute 
alcohol,  has  been  successfully  carried  out. 

Dessaignes*  prepared  and  analyzed  with  good  results  the  plati- 
num  double  salt  of  trimethylamine,  by  conducting  the  mixture 
of  ammonia  and  trimethylamine  vapors  into  hydrochloric  acid, 
evaporating  to  drynesst  extracting  with  absolute  alcohol,  pre- 
cipitating with  platinic  chloride  and  recrystallizing  the  precipi- 
tate formed  several  times  from  hot  water. 

I  Ann.  Ckem.  (Uebig).8i,  106. 


TRIMBTHYI«AMINB   PROM  AMMONIA. 


671 


Wicke'  adopts  the  same  method,  using,  however,  alcohol- 
ether  extract. 

Winkeles,'  in  using  this  method,  further  states  that  while 
ammonium  chloride  is  soluble  to  some  extent  in  absolute  alco- 
hol, it  is  rendered  totally  insoluble  by  the  presence  of  salts  of 
such  bases  as  trimethylamine. 

Eisenberg,'  by  a  similar  procedure,  obtained  the  platinum 
double  salt  in  crystals  of  great  purity  and  perfection. 

The  success  in  each  case  is  undoubtedly  due  to  the  fact  that 
large  quantities  of  hydrochlorides  were  used.  Winkeles,^  for 
example,  employed  the  hydrochlorides  obtained  from  twenty-six 
gallons  of  herring  brine.  Further  the  mixtures  were  very  rich 
in  trimethylamine. 

This  method  applied  to  a  substance  containing  a  low  percent- 
age of  the  latter  yielded  results,  which  clearly  show  that  tri- 
methylamine hydrochloride  does  not  render  ammonium  chloride 
insoluble  in  absolute  alcohol,  and  further  does  not  serve  as  a 
good  means  of  qualitative,  much  less  of  quantitative,  separation. 
A  portion  of  the  mixture  containing  trimethylamine  and  ammo- 
nia was  saturated  with  hydrochloric  acid,  evaporated  to  dryness 
and  extracted  several  times  with  portions  of  several  times  the 
volume  of  boiling  absolute  alcohol.  The  alcoholic  extract 
evaporated  to  dryness  gave  eighteen  per  cent,  of  supposed  tri- 
methylamine hydrochloride.  To  identify  the  latter,  the  residue 
was  taken  up  with  alcohol  and  platinic  chloride  added.  The 
precipitate  formed  was  redissolved  in  boiling  water  and  the  dif- 
ferent fractional  crystallizations,  consisting  of  octahedra,  anal- 
yzed. 

Pt  fontid. 

First  crystallization 43.6 

Last  '*  39.5 


Required  for 

(NH^ClXPtCl^.  43.84 

Corresponding  to  a 

mixture  of 


1  Ann.  Ckem.  (Liebig),  gx,  lai. 
*Ann.  Chem.  (Uebig),  93,  331. 
*Ber.  d.  ehem,  Ges.,  1880,  1669. 
4  Loc.  cit. 


2ANH,Cl),PtCl4W 

3(N(CH,),.HClVptCl„ 
which  require  39.4  per  cent.  Pt. 


672      SEPARATION  OP  TRIMBTHYLAMINB  PROM  AMMONIA. 

Intermediate  crystallizations  gave  intermediate,  gradually 
decreasing  results,  showing  that  the  isomorphous  forms  of  the  two 
salts  crystallized  together. 

Duvillier,  Buisine'  extract  the  mixed  sulphates  to  prepare  pure 
trimethylamine  from  the  technical  product.  The  suggestion  led 
to  the  use  of  the  following  method  which  yielded  satisfactory 
results. 

The  mixed  hydrochlorides  are  repeatedly  extracted  with  por- 
tions of  a  total  of  five  or  six  times  the  volume  of  boiling  absolute 
alcohol  and  the  solvent  distilled  off  in  a  three-quarter  liter  dis- 
tilling bulb.  An  excess  of  caustic  soda  is  added  to  the  residue 
and  the  gases  formed  on  boiling  driven  over  into  a  large  quan- 
tity of  water.  Litmus  is  added,  followed  by  the  exact  quantity 
of  dilute  sulphuric  acid  required  to  neutralize.  The  liquid  is 
evaporated  to  dryness  and  extracted  with  one  liter  cold  absolute 
alcohol,  in  which  trimethylamine  sulphate  dissolves,  leav- 
ing ammonium  sulphate  undissolved.  The  alcohol  is  dis- 
tilled off,  the  residue  transferred  to  a  weighed  dish,  dried  and 
weighed.  In  this  manner  32,910  grams  of  the  carefully  dried 
mixed  chlorides  gave  two  and  five-tenths  grams  trimethylamine 
sulphate,  corresponding  to  2.21  grams  hydrochloride,  or  6.71 
per  cent. 

That  the  extraction  was  complete  is  evident  from  the  total 
absence  of  the  fishy  odor  when  the  extracted  residues  are  treated 
with  alkali.  That  the  extracted  material  is  pure  is  shown  by 
the  following  analyses  of  the  octahedral  crystals  of  the  platinum 
double  salt  prepared  from  the  trimethylamine  sulphate  : 

Required  for 
[N(CH,),.HCl]tPtCl4. 
Per  cent  Pt.  Per  cent.  Pt. 

I.  0.0985  gram  gave 36.92  .... 

II.  0.3017      *•        **    37.12  36.93 

1  Ar.n.  Oum,^  (Liebig)  (s)  S3,  399. 


ZIRCONIUM  TETRAiODIDE. 

By  h.  M.  Dennis  and  a.  B.  Spbncbr. 

Received  June  9,  i8p6. 

WITH  the  exception  of  the  tetraiodide  all  of  the  normal 
halides  of  zirconium  have  been  prepared  and  described, 
the  fluoride,  chloride,  and  bromide  being  white,  crystalline,  sub- 
limable  solids. 

A  few  attempts  to  make  the  iodide  are  recorded  in  the  jour- 
nals, but  in  no  case  was  the  normal  compound,  zirconium  tetra- 
iodide, Zrl^*  obtained.  Melliss'  passed  the  vapor  of  iodine  over 
a  glowing  mixture  of  zirconia  and  carbon ;  he  also  treated  zir- 
conium tetrabromide  with  potassium  iodide,  but  in  neither  case 
did  zirconium  tetrachloride  result.  Hinsberg'  added  an  aqueous 
solution  of  barium  iodide  to  a  solution  of  zirconium  sulphate, 
filtered  off  the  barium  sulphate,  and  evaporated  the  filtrate  over 
concentrated  sulphiuic  acid.  He  obtained  a  compound  of  the 
formula  Zr,I,0„  or  Zrl  (OH),.  He  also  passed  the  vapor  of 
iodine  over  a  heated  mixture  of  zirconium  dioxide  and  carbon 
and  states  that  no  reaction  whatever  took  place.  Bailey'  states 
that  '*  zirconium*  is  acted  upon  by  chlorine  and  bromine, 
in  which,  on  gentle  heating,  it  undergoes  vivid  combustion, 
forming  the  tetrahaloid  derivatives,  and  this  is,  indeed,  a  con- 
venient method  for  obtaining  these  bodies.  The  iodide  could 
not  be  obtained." 

In  the  work  here  to  be  described,  the  authors  first  attempted 
to  prepare  zirconium  tetraiodide  by  passing  the  vapor  of  iodine 
over  heated  zirconium.  The  zirconium  first  used  was  made  by 
reducing  zirconium  dioxide  with  magnesium  powder,  the  two 
substances  being  mixed  in  the  proportion  employed  by  Winkler^ 
and  demanded  by  the  equation 

ZrO,  +  2Mg  =  Zr  +  2MgO. 

This  mixture  was  heated  in  hydrogen  in  the  usual  manner 
and  the  resulting  black  powder  was  removed  from  the  boat, 
thoroughly  ground,  and  again  heated  in  hydrogen  to  insure 

1  Ztsehr.  Chem.,  1870,  296 :  Jsb.,  /870,  328. 

^Ann.  Chtm.  (Uebig),  939,  253. 

s  Chtm  News.y  60,  8. 

*  Prepared  by  the  reduction  of  zirconia  with  maflrnesium  powder. 

S  Ber.  d.  ckem.  Ges.,  33,  2664  ;  94,  888. 


674  I'-   M.    DENNIS  AND  A.   B.   SPENCBR. 

conplete  redaction.  To  free  it  from  magnesia,  the  product  was 
treated  with  a  saturated  solution  of  ammonium  chloride.  Dur- 
ing this  treatment  a  gas  of  very  disagreeable  odor  was  evolved. 
It  is  doubtless  similar  to  that  observed  by  Winkler  at  this  point. 
The  powder  was  then  warmed  with  dilute  twelve  per  cent,  hydro- 
chloric acid  and,  after  collecting  it  on  a  filter,  it  was  washed 
with  water  containing  hydrochloric  acid,  then  with  alcohol  and 
ether,  and  finally  was  dried  in  a  current  of  hydrogen.  The 
analysis  gave 

Per  cent 

Zirconium 80.670 

Silicon 0.807 

Magnesium 0.117 

Hydrogen 0.362 

Oxygen  (diff.) 18.044 

100.000 

These  results  agree  quite  closely  with  those  obtained  by 
Winkle^,'  and  indicate  that  the  product  of  the  reduction  is  chiefly 
zirconium  monoxide  rather  than  zirconium. 

Although  the  powder  probably  contained  but  very  little  free 
zirconium,  it  was  nevertheless  heated  in  hydrogen  and  vapor  of 
iodine  was  passed  over  it.  An  examination  of  the  product  gave 
no  satisfactory  indications,  however,  that  an  iodide  of  zirconium, 
had  been  formed. 

Inasmuch  as  the  failure  to  obtain  union  between  the  zirconium 
and  iodine  might  reasonably  be  ascribed  to  absence  of  free  zir- 
conium in  the  above  product,  it  seemed  advisable,  before 
attempting  any  modification  of  the  iodine  treatment,  to  prepare 
zirconium  by  some  other  method  and  especially  by  some  pro- 
cedure in  which  the  presence  of  any  appreciable  amount  of  oxy- 
gen is  avoided.  Under  the  circumstances  the  method  of  Berze^ 
lius,*  the  reduction  of  potassium  fluozirconate  with  metallic 
potassium,  seemed  the  most  promising  and  was  therefore  em- 
ployed. 

The  potassium  fluozirconate  was  prepared  from  zircon.  The 
zircon  was  finely  ground,  sifted  through  bolting  cloth,  and 
digested  with  concentrated  hydrochloric  acid  until  the  acid  gave 

1  Ber.  d.  chem.  Ges.,  as»  2667. 
^Ann.  derPkys,  (Pogg),  4.  "7 


ZIRCONIUM   TETRAIODIDE.  675 

no  reaction  for  iron.  The  powdered  zircon,  which  was  now 
almost  perfectly  white,  was  dried  and  mixed  with  four  times  its 
weight  of  sodium  carbonate.  The  mixture  was  fused  in  an 
assay  crucible  furnace,  allowed  to  cool,  pulverized,  and  repeat- 
edly extracted  with  water.  The  residue,  consisting  of  zirconia 
and  unattacked  zircon  together  with  some  silica  and  ferric  oxide, 
-was  heated  with  concentrated  hydrochloric  acid,  evaporated  to 
dryness,  and  heated  in  an  air-bath  to  120°  to  render  the  silica 
insoluble.  The  dried  mass  was  treated  with  a  little  hydrochloric 
acid,  water  was  added,  and  the  silica  and  other  insoluble  matter 
was  filtered  off.  The  filtrate,  now  containing  zirconium  chloride 
and  some  ferric  chloride,  was  largely  diluted  with  water,  and 
ammonium  hydroxide  was  added  until  there  was  formed  a  slight 
but  permanent  precipitate  which  was  then  dissolved  by  adding 
as  little  hydrochloric  acid  as  possible.  Sulphur  dioxide  was 
then  passed  into  the  solution  until  the  liquid  smelled  strongly  of 
the  gas.  In  many  cases  a  precipitate  of  basic  zirconium  sul- 
phite formed  at  once,  but,  as  the  compound  seemed  to  be  some- 
what soluble  in  an  excess  of  sulphurous  acid,  the  solution  was 
always  boiled  for  from  ten  to  fifteen  minutes  to  insure  complete 
precipitation.  In  the  reaction  free  hydrochloric  acid  is  formed 
both  by  the  conversion  of  the  zirconium  chloride  into  the  basic 
sulphite  and  by  the  reductfon  of  the  ferric  chloride  to  the  ferrous 
salt.  As  this  acid  would  dissolve  the  zirconium  sulphite,  it  was 
partially  neutralized  by  the  addition,  from  time  to  time,  of  a  few 
drops  of  dilute  ammonium  hydroxide.  The  zirconium  precipi- 
tate not  being  wholly  free  from  iron,  it  was  dissolved  in  hydro- 
chloric acid  and  again  precipitated  with  sulphur  dioxide.  The 
pure  zirconium  basic  sulphite  thus  pbtained  was  dissolved  in 
hydrochloric  acid  and  zirconium  hydroxide  was  precipitated  by 
adding  ammonium  hydroxide.  The  well-washed  hydroxide  was 
dissolved  in  hydrofluoric  acid,  potassium  fluoride  was  added, 
and  the  resulting  potassium  fluozirconate  was  dissolved  in  hot 
water  and  recrystallized. 

The  potassium  fluozirconate  thus  prepared  was  reduced  with 
metallic  sodium,  the  operation  being  carried  out  in  a  cast-iron 
crucible.  The  crucible  is  cylindrical  in  form  with  an  internal 
diameter  of  two  inches  and  depth  of  five  inches.     The  wall  and 


676  I,.    M.    DENNIS  AND   A.   E.   SPENCER.     ' 

bottom  are  over  one  inch  in  thicknebs.  At  the  top  it  has  a 
flange  seven  inches  in  diameter  and  is  provided  with  a  cast-iron 
cover  one  inch  in  thickness,  which  can  be  firmly  fastened  to  the 
flange  by  means  of  six  one-half  inch  bolts. 

In  charging  the  crucible,  sodium  chloride,  finely  ground  and 
thoroughly  dried,  was  first  put  in  to  the  depth  of  about  an  inch 
and  a  half,  and  this  was  then  well  pounded  down  with  a  wooden 
plunger  to  compact  the  salt  and  expel  the  enclosed  air.  On  top 
of  the  salt  were  placed  alternate  layers  of  potassium  fluozircon- 
ate,  also  thoroughly  dried,  and  metallic  sodium,  these  being 
pounded  down  as  before.  The  remaining  space  in  the  crucible 
was  then  filled  with  sodium  chloride  and,  after  pounding  this 
down,  the  top  was  bolted  on  and  the  crucible  was  heated  for 
about  three  hours  with  three  triple  burners.  This  heat,  how- 
ever, was  not  suflScient  to  raise  the  crucible  to  redness. 

The  crucible  was  then  allowed  to  cool  and,  upon  opening  it, 
the  charge  was  found  to  be  so  compact  that  it  had  to  be  loosened 
with  a  chisel.  On  treating  the  mass  with  water  the  metallic 
zirconium,  together  with  a  small  amount  of  the  oxide  which  had 
formed,  settled  to  the  bottom  while  the  sodium  chloride  and 
potassium  and  sodium  fluorides  dissolved. 

The  zirconium  and  zirconium  oxide  were  separated  by  first 
floating  off  the  lighter  zirconium  with  water  and  then  digesting 
it  with  dilute  hydrochloric  acid  at  40^  until  all  of  the  oxide  had 
been  dissolved.  The  resulting  product  was  a  black,  amorphous 
powder  which,  after  washing  with  water,  alcohol,  and  then  with 
ether,  showed  no  trace  of  impurity  before  the  spectroscope 
except  a  slight  amount  of  sodium. 

Vapor  of  iodine  was  pass^  over  some  of  this  zirconium  heated 
to  dull  redness  in  a  current  of  hydrogen,  but  with  no  better  suc- 
cess than  with  the  other  sample.  We  then  concluded  to  substi- 
tute hydriodic  acid  gas  for  the  iodine.  Considerable  difficulty 
was  encountered  in  finding  a  suitable  method  of  preparing  the 
gaseous  hydriodic  acid.  That  described  by*Merz  and  Holzmann' 
was  finally  found  to  answer  admirably.  It  consists  in  passing 
dry  hydrogen  and  vapor  of  iodine  through  a  red  hot  tube  filled 

I  Ber.  d.  chem.  Ges.,  aa,  867. 


ZIRCONIUM  TETRAIODIDE.  677 

with   pumice  stone  and  freeing   the  hydriodic   acid  gas  from 
iodine  by  passing  the  gases  through  cotton. 

In  treating  the  zirconium  wiih  hydriodic  acid  gas  the  follow- 
ing apparatus  was  used. 

Iodine  was  placed  in  a  small  tubulated  flask  connected  on  one 
side  with  an  apparatus  furnishing  pure,  dry  hydrogen  and  on 
the  other  side  with  a  long  piece  of  combustion  tubing.  The 
half  of  this  tube  nearest  the  iodine  flask  was  filled  with  pieces  of 
pumice  stone  and  rested  in  a  combustion  furnace.  The  other 
half,  extending  beyond  the  combustion  furnace,  w^as  filled  with 
cotton.  The  end  of  this  tube  was  connected  with  another  com- 
bustion tube  resting  in  a  second  combustion  furnace.  The  por- 
celain boat  containing  the  zirconium  was  placed  in  this  second 
tube. 

The  hydrogen  was  first  passed  through  the  whole  apparatus 
for  several  hours  and  then  the  first  furnace  was  lighted.  When 
the  pumice  had  become  red  hot  the  flask  containing  the  iodine 
was  gently  heated.  The  tube  containing  the  zirconium  soon 
became  filled  w^ith  the  hydriodic  acid  gas,  whereupon  the  second 
furnace  was  lighted.  As  the  temperature  rose,  a  brownish-yel- 
low substance  collected  in  the  cold  end  of  the  combustion  tube, 
but  as  the  heat  became  greater  the  color  entirely  disappeared 
and  there  remained  an  amorphous  white  sublimate.  No  further 
sublimate  was  formed  until  the  tube  had  almost  reached  a  bright 
red  heat  when  there  appeared  just  beyond  the  point  where  the 
tube  was  red  hot  a  white  crystalline  sublimate,  different  in  ap- 
pearance from  that  which  first  formed.  The  gas  escaping  from 
the  end  of  the  tube  contained  hydriodic  acid,  hydrogen,  some 
iodine,  and  a  trace  of  iron,  the  last  probably  being  present  in 
traces  in  the  zircotlium  and  volatilizing  as  ferrous  iodide.  The 
tube  was  kept  at  a  bright  red  heat  for  from  three  to  four  hours. 
The  gas  was  then  turned  off  and  when  the  boat  had  cooled  con- 
siderably the  heating  of  the  iodine  fla.sk  was  discontinued.  The 
first  furnace  was  then  shut  off  and  the  whole  apparatus  was 
allowed  to  cool  in  the  current  of  hydrogen. 

The  material  in  the  boat  had  changed  from  a  black  to  a  gray- 
ish-white color,  but  a  chemical  examination  showed  that  it  con- 
tained ver>'  little  iodine.     The  amorphous  sublimate  which  first 


678  ZIRCONIUM   TETRAIODIDE. 

formed  was  found  not  to  be  zirconium  iodide  but  to  contain 
chiefly  iron  and  iodine. 

The  crystalline  sublimate  which  w^as  formed  only  at  a  red  heat 
was  next  analyzed.  These  crystals  were  found  to  be  insoluble 
in  water,  nitric  acid,  hydrochloric  acid,  aqua  regia,  and  carbon 
disulphide.  They  were  decomposed  and  dissolved  by  concen- 
trated sulphuric  acid  ;  they  were  also  decomposed,  but  not  com- 
pletely, by  concentrated  nitric  acid,  iodine  being  liberated  and  a 
white  powder,  insoluble  in  the  nitric  acid,  remaining.  This 
residue  was  soluble  in  concentrated  sulphuric  acid  and  from  this 
solution  ammonium  hydroxide  threw  down  a  white  gelatinous 
precipitate.  Upon  dissolving  this  precipitate  in  hydrochloric 
acid  and  dipping  turmeric  paper  into  the  solution,  the  orange 
color  characteristic  of  zirconium  was  obtained .  The  solution 
gave  no  reaction  for  iron. 

The  zirconium  in  the  compound  was  quantitatively  deter- 
mined by  expelling  the  iodine  by  heating  a  portion  of  the  salt 
with  a  mixture  of  sulphuric,  nitric,  and  nitrous  acids,  dissolving 
the  residue  in  concentrated  sulphuric  acid,  diluting  with  water, 
and  precipitating  the  zirconium  with  ammonium  hydroxide. 
The  precipitate  was  washed,  dried,  and  ignited,  and  the  zirco- 
nium weighed  as  the  dioxide. 

The  iodine  was  determined  by  fusing  some  of  the  compound 
with  about  five  times  its  weight  of  a  mixture  of  potassium  and 
sodium  carbonate.  The  mass  was  then  treated  with  water,  fil- 
tered, and  after  acidifying  the  filtrate  with  nitric  acid  the 
hydriodic  acid  was  precipitated  with  silver  nitrate  and  weighed 
as  silver  iodide. 

The  results  were 

Calculated  (or  Zrl4.  Pound. 

Per  cent.  Per  cent.         Per  cent.         Per  ceut. 

Zirconium 15.15  15.17  15.00  15.37 

Iodine 84-85  85.34  85.27  

The  crystals  when  examined  under  the  microscope  proved  to 
be  clear,  colorless  cubes  which  showed  no  double  refraction. 

When  heatedjor  some  hours  in  hydrogen  the  zirconium  tetra- 
iodide  becomes  black  and  iodine  and  hydriodic  acid  are  formed. 
Heated  in  the  air  the  iodide  melts  and  sublimes.  A  weighed 
amount  was  placed  in  a  porcelain  crucible,  covered  with  water, 


PHTHAUMID.  679 

and  evaporated  to  dryness.  No  change  in  weight  and  scarcely 
any  in  color  resulted  after  two  such  treatments.  This  behavior 
toward  water  is  surprising,  for  from  the  published  descriptions 
of  zirconium  tetrachloride  and  tetrabromide,  it  was  to  be 
expected  that  the  iodide  would  prove  to  be  a  hygroscopic  com- 
pound easily  decomposed  by  water.  It  seems,  however,  to 
more  nearly  resemble  the  fluoride  which  Deville  states  to  be  a 
colorless  crystalline  substance  volatile  at  a  white  heat  and  iti- 
soluble  in  water  or  acids. 

Cornell  Ukivbrsity,  Ithaca.  N.  Y. 


PHTHALiniD,* 

Br  J.  A.  Matrbws. 
Receired  June  9,  tMg6. 

A  NUMBER  of  years  ago  Prof.  C.  E.  Colby  and  Mr.  Dodge, 
of  Columbia  University  were  led  to  try  the  effect  produced 
by  heating  together,  under  pressure,  mixtures  of  (i)  fatty  acids 
and  fatty  nitrils;  (2)  fatty  acids  and  aromatic  nitrils  ;  (3)  fatty 
nitrils  and  aromatic  acids ;  and  (4)  aromatic  acids  and  aromatic 
nitrils.  The  reactions  were  carried  on  in  sealed  tubes.  The 
score  or  more  reactions  that  they  tried  were  done  at  tempera- 
tures ranging  from  235**  to  280°  C.  As  the  result  of  their  work 
they  reached  these  conclusions  regarding  what  is  likely  to  take 
place,  at  least  when  monobasic  acids  and  mononitrils  are  em- 
ployed.' 

1.  Fatty  nitrils  and  fatty  acids  give  secondary  amids. 

2.  Fatty  nitrils  and  aromatic  acids  give  fatty  acids  and  aro- 
matic nitrils. 

3.  Aromatic  nitrils  and  fatty  acids  give  mixed  secondary 
amids. 

4.  Aromatic  nitrils  and  aroniatic  acids  gave  secondary  amids, 
except  in  one  case  when  exceptionally  high  heat  was  used  (280'') 
in  which  case  the  cyanide  of  the  higher  radicle  was  formed. 

In  regard  to  dibasic  acids  and  dicyanides  not  so  much  has 
been  done.  Miller  first  tried  reactions  with  succinic  acid  and 
ethylene  cyanide.'  He  found  that  succinimid  resulted  from 
each  of  the  following  experiments  : 

1  Re«d  before  the  American  Chemical  Society.  New  York  Section,  June,  1896. 
*  Am.  Chem.J.^  13.  i8^t. 
t  This  Journal,  16, 443,  1B94. 


68o  J.   A.   MATHEWS. 

1.  Ethylene  cyanide  and  acetic  acid  heated  in  a  sealed  tube. 

2.  Acetonitril  and  succinic  acid,  and 

3.  Ethylene  cyanide  and  succinic  acid. 

Some  other  acids  in  this  series  have  been  tried.  Malonic  acid 
was  rather  imperfectly  tested.  In  every  case  the  tubes  ex- 
ploded and  malonimide  was  not  obtained  at  all. 

Seldner'  reports  parallel  results  to  those  obtained  by  Miller 
when  he  used  glutaric  acid  and  trimethylene  cyanide.  In  the 
following  trials  which  he  made  glutarimid  resulted  every 
time: 

1.  Glutaric  acid  (i  mol.)  and  acetonitril  (2  mols.). 

2.  Glutaric  nitril  (i  mol.)  and  acetic  acid  (2  mols.). 

3.  Glutaric  acid  and  glutaric  nitril,  equal  molecules. 

Until  the  writer,  at  Prof.  Colby's  suggestion,  made  the  experi- 
ments hereinafter  recorded  no  one,  to  my  knowledge,  had  applied 
these  methods  to  aromatic,  dibasic  acids.  The  results  of  the 
first  experiments  are  very  gratifying  and  I  hope  in  the  near 
future  to  try  the  reaction  with  other  dibasic,  aromatic  acids. 

Since  no  information  regarding  phthalic  nitril  could  be 
obtained  I  was  obliged  to  do  without  it.  The  experiments 
were  therefore  made  with  phthalic  acid  and  propionitril. 

Four  sealed  glass  tubes  each  containing  phthalic  acid  ( i  mol.) 
and  propionitril  (2  mols.),  plus  about  three  drops  of  acetic 
anhydride  were  heated  in  an  oven  for  varipus  lengths  of  time 
and  at  different  temperatures. 

Tube  I.  The  first  tube  was  opened  after  ten  hours  heating  at 
180°  C.  The  contents  of  the  tube  had  a  pungent  acid  odor  and 
were  treated  with  cold  dilute  potassium  carbonate  solution.  A 
residue  consisting  of  needle-like  crystals  remained.  These 
were  filtered  off,  washed  with  water,  and  dried.  The  crystals 
then  had  a  melting  point  of  228°  C.  I  immediately  suspected 
from  this  melting  point  that  phthalimid  had  been  formed  by  the 
reaction 

C.H,(COOH).  +  C,H,CN  =  C.H,(CO).NH  +  C.H.COOH. 

The  yield  of  phthalimid  in  this  experiment  was  about  sixty 
per  cent,  of  the  theoretical. 

^  Am.  Chem.J.^  17,  /^. 


PHTHALIMID.  68 1 

Tube  II.  On  heating  the  remaining  three  tubes  higher  No. 
2  broke  at  about  215^. 

Tube  III.  After  further  heating  of  eight  hours  at  200**  to  215** 
C  the  third  tube  was  opened  and  the  contents  treated  with 
potassium  carbonate  solution.  The  crystals  remaining  were  not 
so  light  colored  as  those  from  Tube  I,  and  were  so  different  in 
appearance  that  it  was  thought  some  other  reaction  had  taken 
place.  The  melting-point,  however,  was  about  the  same  as  in 
the  first  case,  viz,,  227'',     Yield  eighty-four  per  cent. 

Tube  IV.  Exploded  at  258*  C. 

Since  the  theoretical  equation  requires  only  one  molecule  of 
nitril  to  one  of  phthalic  acid  two  more  tubes  were  prepared,  each 
containing  equal  molecules  of  phthalic  acid  and  propionitril. 

Tube  V.  After  three  and  a  half  hours  at  i8o*-2oo**  C.  the  fifth 
tube  was  opened  and  treated  with  potassium  carbonate  solution 
as  before.  Residue  crystalline;  melting-point  228.5®,  yield 
eighty-eight  per  cent. 

Tube  VI.  Heated  five  and  a  half  hours  at  i8o*'-200*',  melting 
point  of  residue,  insoluble  in  cold,  dilute  potassium  carbonate 
solution,  228.3"  C.,  yield  92.5  per  cent. 

The  crystals  of  phthalimid  were  all  more  or  less  colored,  the 
color  being  darkest  in  the  case  of  the  third  tube  which  had  been 
subjected  to  long,  high  heat.  In  no  instance  was  any  outward 
pressure  noticed  on  opening  the  tubes. 

Portions  of  the  products  were  recrystallized  from  acetic  acid, 
from  alcohol,  and  from  alcohol  with  the  addition  of  ani- 
mal charcoal  to  decolorize.  The  melting  points  of  the  recr>'s- 
tallized  products  were  a  little  higher  than  before  purification, 
viz.,  230**,  229.5**  and  229.5'*,  respectively.  These  agree  very 
closely  with  the  point  given  in  Beilstein. 

Biedermann*  gives  the  melting-point  as  228°  or  229**  C. 

Michael*  gives  the  corrected  melting-point  as  233.5'*  C.  The 
decolorized  crystals  from  alcohol  form  beautiful  long  needles. 

Notwithstanding  the  close  agreement  of  the  melting-points 
obtained  with  those  given  by  the  authorities,  some  other  tests 
were   made  to  show  that  the  product  was  nothing  else  than 

1  Ber.  d.  ckem,  Ges.^  lo,  1166. 
«  Ber.  d.  ckem.  Ces..  10,  579. 


682  N.  J-  LANE.    DETERMINATION  OF  SULPHURIC  ACID. 

phthalimid.  A  portion  of  the  crystals  heated  with  potassitiin 
hydroxide  went  into  solution  with  evolution  of  ammonia .  Another 
portion  of  the  cr3'stals  were  covered  with  concentrated  ammo- 
nia and  allowed  to  stand  for  some  time.  They  were  soon  con- 
verted into  microscopic  crystals  of  phthalamid 

(=C.H,(CONH,),). 
These  crystals  were  filtered  oflF,  washed,  and  dried.  They 
melted  at  217.5^  (uncorrected)  with  an  evolution  of  ammonia, 
which  began  at  about  200^.  The  phthalamid  was  further 
proved  by  its  insolubility  in  cold  water,  alcohol,  and  ether,  and 
by  boiling  it  with  water  it  was  decomposed,  giving  off  ammonia 
and  on  cooling  phthalimid,  melting  at  230**  C,  crystallized  out. 
The  results  of  these  tests  show  conclusively  that  the  product 
is  phthalimid  and  that  when  it  is  made  by  the  action  of  equal 
molecules  of  acid  and  nitril  the  yield  is  large.  The  reaction 
works  comparatively  readily,  and  at  a  much  lower  temperature 
than  was  needed  to  affect  the  reactions  recorded  by  Colby  and 
Dodge.  It  is  highly  probable  that  with  slight  changes  of  con- 
ditions any  one  of  a  variety  of  nitrils  would  give  the  same 
result.  I  hope  to  report  further  experiments  with  phthalic  acid 
and  other  dibasic  aromatic  acids  at  a  later  day. 

Orgaivic  Labokatort,  Columbia 
University,  New  York. 

DETERMINATION  OF  SULPHURIC  ACID. 

By  N.  J.  Lane. 

Rereivcd  May  19,  itp6. 

SOME  months  ago,  before  hearing  of  the  controversy  between 
Dr.  Lunge  and  Mr.  Gladding,  some  experiments  were 
made  on  this  subject,  the  results  of  which  sustain  Mr.  Gladding's 
case.  The  determinations  were  made  on  nearly  normal  sul- 
phuric acid  to  establish  its  strength  with  the  following  results  : 

Barium  chloride  Barium  chloride 

added  suddenly.  added  by  drops. 

1.  Sulphuric  acid 5o-03  49-23 

2.  "  *'   4990  4932 

3.  "  "   50.14 

And  the  average  of  several  practically  identical  titrations  on 
C.  P.  sodium  carbonate  gave  sulphuric  acid  49.33. 

The  above  results  were  obtained  with  the  greatest  care,  and 
every  precaution  used  to  insure  accuracy.  This,  in  my  opinion, 
conclusively  proves  the  accuracy  of  Mr.  Gladding's  statements. 


NOTE  ON  THE  SOLUBILITY  OF  BISMUTH  SULPHIDE   IN 
SODIUM  SULPHIDE,  WITH  SPECIAL  REFERENCE  TO 
THE    ESTiriATION    OF  SMALL  AflOUNTS  OF 
BISnUTH  IN  ANTI-FRICTION  ALLOYS. 

By  Thomas  B.  Stxllman. 

Received  June  i6,  1896. 

THE  method  of  separation  of  lead,  copper  and  bismuth  from 
antimony,  arsenic  and  tin  by  the  use  of  sodium  sulphide 
is  quite  general.  This  is  dependent  upon  the  usually  accepted 
statement  that  the  sulphides  of  bismuth,  lead  and  copper  are 
insoluble,  and  the  sulphides  of  arsenic,  antimony  and  tin  are 
soluble  in  sodium  sulphide.  This  process  of  separation  is 
employed  in  the  analysis  of  various  alloys,  especially  of  anti- 
friction alloys,  containing  lead,  tin,  antimony,  etc. 

An  alloy,  used  for  similar  purposes,  but  containing,  in  addi- 
tion to  lead,  copper,  antimony  and  tin,  a  very  small  amount  of 
bismuth,  was  recently  submitted  to  me  for  analysis. 

After  complete  solution  of  the  alloy  in  hydrochloric  acid  with 
a  few  drops  of  nitric  acid,  the  acid  was  neutralized  with  sodium 
hydroxide,  sodium  sulphide  solution  (1.06  sp.  gr.)  added  and 
the  heat  applied  for  twenty  minutes.  The  solution  was  filtered 
and  the  filtrate  examined  for  the  antimony  and  tin  with  satisfac- 
tory results. 

The  precipitate  of  insoluble  sulphides  remaining  upon  the 
filter  was  found  to  contain  lead  and  copper,  but  no  bismuth. 
This  indicated  that  the  small  amount  of  bismuth  which  was 
present  in  the  alloy  had  gone  into  solution  in  the  sodium  sul- 
phide. 

To  prove  this  theory,  I  weighed  0.128  gram  of  pure  bismuth 
nitrate,  dissolved  it  in  twenty-five  cc.  of  water  with  a  few  drops  of 
nitric  acid,  the  clear  solution  neutralized  with  sodium  hydroxide, 
seventy-five  cc.  solution  of  sodium  sulphide  added,  and  warmed 
to  a  temperature  near  boiling  for  twenty  minutes.  The  solution 
was  filtered  from  the  bismuth  sulphide,  remaining  insoluble  in 
the  sodium  sulphide.  The  clear  filtrate  was  rendered  faintly 
acid  with  hydrochloric  acid,  when  a  brownish-black  precipitate 
immediately  formed.     This  precipitate  was  filtered,  dissolved  in 


684  SOLUBILITY   OF   BISMUTH   SULPHIDE. 

hot  nitric  acid  and  evaporated  to  dryness  and  ignited  in  a 
weighed  porcelain  crucible.  The  residue  obtained  was  0.031 
gram  of  bismuth  trioxide,  and  strongly  yellow  in  color.  It  was 
dissolved  in  a  few  drops  of  hydrochloric  acid,  and  the  three  fol- 
lowing confirmatory  tests  for  bismuth  were  made  : 

1 .  A  portion  of  the  solution  was  poured  into  a  large  amount  of 
water,  forming  immediately  a  white  precipitate  of  bismuth  oxy- 
chloride. 

2.  A  portion  was  tested  by  Schneider's  test,  the  most  delicate 
test  for  bismuth,  the  reaction  obtained  being  strong  and  charac- 
teristic. 

3.  A  portion  was  diluted  with  water,  not  enough  to  cause  pre- 
oipitation,  and  the  solution  saturated  with  hydrogen  sulphide. 
The  precipitate  formed  was  brownish-black  in  color. 

These  three  tests  are  absolutely  confirmative  of  the  presence 
of  bismuth,  and  also  show  the  absence  of  the  other  metals.  By 
thus  using  pure  bismuth  nitrate  for  this  test,  lead,  copper,  anti- 
mony and  tin  are  not  present. 

If  now  an  analyst  should  weigh  twelve  grams  of  an  alloy, 
composed  approximately  of  lead  eighty  per  cent.,  antimony  fif- 
teen per  cent. ,  tin  4.75  per  cent.,  and  bismuth  0.25  per  cent. 
('*  magnolia  metal,)''  and  sodium  sulphide  solution  be  used  for 
the  separation  of  the  tin  and  antimony  from  the  lead  and  bismuth, 
ali  of  the  bismuth  present  would  pass  into  solution  and  escape 
determination  by  the  analyst. 

No  analyst,  however,  would  use  as  much  as  twelve  grams  of 
such  an  alloy  for  analysis,  but  rather  one  or  two  grams. 

If  one  gram  be  taken  and  sodium  sulphide  used  as  above 
indicated,  three  per  cent,  of  bismuth  might  be  present  and  a// of 
it  pass  into  solution  in  the  sodium  sulphide  instead  of  remain- 
ing as  an  insoluble  sulphide  with  the  lead  sulphide. 

Dbpartmekt  op  analytical  Chemistry, 
Stevens  Institute  op  Technology. 


ON  THE  ESTIMATION  OF  SULPHUR  IN  PYRITES. 

By  G.  Lunob. 
Received  June  19.  s8g6. 

IT  has  taken  Mr.  T.  S.  Gladding  six  months  to  reply  to  my 
last  paper  on  the  above  subject.  I  will  not  take  much  more 
than  six  days  from  the  date  of  receiving  the  May  number  of  the 
Journal  of  before  dispatching  my  final  reply  to  that  gentleman. 

Mr.  Gladding  avoids  any  mention,  and  of  course  offers  no 
refutation,  of  the  charges  I  had  brought  against  him,  but  he 
again  puts  me  into  a  totally  false  light,  by  saying  that  I 
**  attempt  no  further  support  of  my  position  by  chemical  experi- 
ment." This  suppresses  the  fact  that  I  had  referred  to  my  more 
than  sufficient  experimental  proof  for  Mr.  Gladding's  and  his 
assistants'  inability  to  handle  my  process,  i^hich  has  been  in 
daily  successful  use  by  scores,  if  not  hundreds,  of  chegiists  for  a 
number  of  years  past,  and  is  that  employed  in  Presenius'  own 
laboratory,  as  I  hear  from  his  son-in-law  and  laboratory  chief. 
Dr.  Hintz.  Mr.  Gladding  now  exacts  a  further  reply  from  me, 
more  especially  on  the  strength  of  some  new  comparative  tests 
of  what  he  states  to  be  the  main  point  at  issue,  namely  the 
necessity  of  a  very  slow  addition  of  the  barium  chloride. 

I  am  convinced  that  our  readers  are  as  tired  of  this  dispute  as 
I  am,  but  as  some  of  them  might  construe  my  silence  into  the 
admission  that  Mr.  Gladding  is  right  on  this  point,  and  might 
saddle  themselves  with  a  total  unnecessary  complication  in  their 
daily  work,  I  will  not  shirk  a  further  reply,  although  I  think  it 
unnecessary  after  having  quoted  already  in  March,  1895,  eleven 
experiments  by  entirely  independent  chemists,  refuting  all  Mr. 
Gladding's  assertions. 

In  his  former  paper  Mr.  Gladding  states  that  the  error  caused 
by  the  rapid  addition  of  the  barium  chloride  solution  is  from 
two-tenths  to  three-tenths  per  cent,  of  sulphur,  and  according  to 
his  last  paper  it  is  even  one-half  per  cent.  He  appeals  to  inde- 
pendent chemists  to  settle  this  discrepancy  between  his  state- 
ments and  my  own.  I  have  taken  this  up  in  the  following  man- 
ner :  I  instructed  one  of  my  assistants,  Mr.  U.  Wegeli,  a  skilled 


686  G.  LUNGE.      SULPHUR   IN   PYRITES. 

worker,  but  entirely  ignorant  of  the  above  dispute,  to  make  a 
series  of  very  careful  tests  of  a  sample  of  pyrites,  just  arrived  for 
analysis  and  belonging  to  an  important  commercial  case.  I 
enjoined  him  to  give  me  absolutely  unvarnished  results  (which 
in  our  laboratory  it  would  not  have  been  at  all  necessary  to  say), 
and  I  told  him,  as  we  must  be  quite  sure  of  the  matter,  he  must 
not  merely  employ  all  the  ordinary  precautions,  but  also  try  both 
the  usual  quick  addition  of  the  barium  chloride  and  a  process 
recently  very  much  recommended,  namely,  the  very  slow  addi- 
tion of  the  precipitant ;  I  did  not  express  any  opinion  of  my  own 
upon  that  point,  and  left  it  entirely  for  him  to  find  out  what 
there  was  in  the  matter.  I  had  just  then  to  undertake  a  short 
journey,  and  on  my  return  he  handed  to  me  the  following 
results. 

A.  Quick  addition  \^i,  <?.,  pouring  in  the  hot  barium  chloride 
solution  in  about  ten  portions,  occupying  about  half  a  minute 
in  all,  and  stirring  the  mixture  all  the  time,  as  every  chemist 
would  do). 

I.  39.83         2.  39.65         3.  39.65  per  cent,  sulphur. 

B.  Slow  addition  from  a  burette,  one  drop  per  second  (exactly 
as  described  by  Mr.  Gladding). 

4-  3963         5-  39.69        6.  39.44  per  cent. 

This  means  :  In  No.  2  and  3  the  quick  addition  has  given 
idejitical  results  with  the  slow  addition  in  No.  4  and  5.  No.  i 
shows  a  little  more.  No.  6  a  little  less.  I  have  suppressed  noth- 
ing, and  I  give  these  results  as  well,  although  they  are  evi- 
dently not  as  reliable  as  the  other  four,  entirely  concordant, 
results  ;  but  even  if  we  admit  the  less  reliable  results  in  striking 
an  average,  we  find  a  difference  of  only  one-tenth  per  cent, 
between  the  quick  (39.71)  and  the  slow  (39.59)  process.  Such 
a  difference  is  evidently  within  the  limits  of  ordinary  experi- 
mental error. 

Zurich. 

[This  discussion  closes  with  the  present  .paper. — Ed.] 


BACTERIA  IN  HILK  SUGAR. 

Bt  Albert  R.  Lbbds. 

Receircd  June  6. 1896. 

CERTAIN  phases  of  bacteriological  investigations  command 
universal  and  profound  popular  interest,  and  any  publica- 
tion relating  to  the  connection  of  a  specific  organism  with  a 
zymotic  disease,  elicits  general  attention  and  discussion.  This 
intimate  connection  of  bacteriology  with  questions  of  life  and 
death,  has  led  many  to  regard  the  study  as  the  proper  province 
of  medical  specialists,  despite  the  first  uses  made  of  bacteriologi- 
cal methods  by  Pasteur  and  his  followers  and  to  neglect  them  as 
instruments  of  chemical  research.  But  the  morphology,  the 
classification,  the  physiology,  and  the  botany  of  the  bacteria  are 
in  such  a  rudimentary  and  unsatisfactory  condition  that  the 
most  valuable  methods  of  bacteriological  investigation  are  still 
of  a  chemical  nature.  The  preparation  of  the  culture  fluids,  the 
application  of  the  tests,  and  the  isolation  of  the  products  are 
chemical  operations,  and  the  advances  to  be  made  in  the  near 
future  are  to  be  looked  for  mainly  on  the  chemical  side  of  the 
subject.  For  this  reason  the  absence  from  the  columns  of  this 
Journal  of  papers  resting  upon  the  bacteriological  questions,  has 
been  a  matter  of  surprise  to  the  writer,  and  the  important  con- 
tributions which  have  been  herein  recently  made  by  Dr.  Schwei- 
nitz,  Dorsett,  Bennett,  Pammel,  and  Mason,  a  source  of  con- 
gratulation. Their  results  foretell  the  rich  harvest  of  the  future 
when  the  complete  quantitative  value  of  the  chemical  actions 
involved  are  known,  and  the  different  views  which  they  may  be 
expected  to  inaugurate  as  to  the  nature  of  many  bodies  now 
grouped  closely  together,  but  which  deport  themselves  very  dif- 
ferently when  bacteria  are. the  reagents  made  use  of. 

It  is  for  these  reasons  that  the  writer  desires  to  put  on  record 
the  slight  observations  which  he  has  made  during  the  course  of 
ordinary  chemical  work.  They  spring  out  of  some  anomalous 
behavior  of  specimens  of  milk  sugar,  which  were  being  examined 
for  purity.  All  the  samples  of  pulverized  milk  sugar  coming 
from  the  drug  stores,  which  he  examined,  proved  to  contain  a 
ferment  when  their  solutions  were  kept  at  the  optimum  tempera- 


688  A.  A.  BENNETT  AND   L.  A.  PLACEWAY. 

ture  for  a  sufficient  length  of  time.  The  lactic  acid  produced 
was  isolated  in  the  form  of  calcium  lactate.  This  was  not  the 
case  with  some  lactose  crystallized  in  nodular  masses  of  pris- 
matic crystals  which  had  been  obtained  originally  from  Kahl- 
baum,  and  had  been  standing  for  twenty-five  years  in  a  stoppered 
jar.  It  was  sterile.  With  the  exception  of  this  specimen,  all  the 
others  gave  an  abundant  crop  of  bacteria  when  definite  weights 
dissolved  in  sterilized  water  were  submitted  to  ordinary  gelatin- 
peptone  culture.  The  maximum  number  obtained  in  this 
medium  was  1400  colonies  per  gram  of  milk  sugar.  In  studying 
these  colonies  I  looked  more  particularly  for  the  bacillus  acidi 
lactici  and  the  other  ten  or  twelve  species,  which  are  at  the  pres- 
ent time  classified  as  the  specific  milk  bacteria,  but  without  suc- 
cess. With  a  lactose-litmus  gelatin  solution  a  still  larger  num- 
ber of  colonies  was  obtained  and  possibly  larger  search  in  this 
medium,  might  have  revealed  the  specific  milk  bacteria  of  lactic 
acid  fermentation.  But  my  immediate  object  had  been  attained, 
and  the  presence  of  bacteria  as  a  common  impurity  in  lactose, 
to  be  looked  for  and  avoided  by  the  chemist  and  the  druggist, 
sufficiently  demonstrated. 


THE  QUANTITATIVE  DETERMINATION  OF  THE  THREE 

HALOGENS,  CHLORINE,  BROMINE  AND  IODINE, 

IN  niXTURES  OF  THEIR  BINARY  COH- 

POUNDS. 

By  a.  a.  Bbnnbtt  ahd  L.  A.  Placbwily. 

Received  June  t.  i996. 

CHEMICAL  literature  contains  many  records  of  methods  for 
the  quantitative  estimation  of  the  halogen  elements,  and 
for  any  one  of  these  elements  in  the  absence  of  the  others  they 
are  as  satisfactory  as  may  be  required.  There  are  also,  it  is 
true,  many  suggestions  and  several  proposed  methods  for  the 
separation  and  estimation  of  these  elements  when  present  together 
or  when  some  two  are  found  in  the  same  mixture,  although  they 
are  generally  unsatisfactory  for  one  reason  or  another.  The 
methods  for  qualitative  determinations  as  s:iven  by  Hart  and 
by  Kebler^  inthe /(mrnai  o/AnafyHcai  CA^mistfy.sxe  thorovLghly 
satisfactory.   A  very  convenient  qualitative  method  that  is  in  use 


CHI<ORINK,    BROMINB,   AND  lODINB.  689 

in  this  laboratory  consists  in  first  using  chlorine  water,  or 
enchlor,  (made  from  potassium  chlorate  and  hydrochloric  acid), 
which  immediately  determines  the  presence  or  absence  of  iodine 
and  in  its  absence  that  of  bromine.  Carbon  bisulphide  is  used 
as  the  final  indicator.  If  iodine  is  present  more  chlorine  water 
is  added  and  the  whole  is  heated  until  the  iodine  color  is  re- 
placed by  the  light  yellow  color  due  to  bromine.  This  point  is 
easily  discerned.  If  now  one  or  more  of  these  halogens  are 
present  a  portion  of  the  original  solution  is  treated  with  concen- 
trated nitric  acid  and  boiled  until  both  of  these  elements  are 
removed.  This  solution  is  now  tested  for  chlorine  by  the  usual 
methods. 

There  are  several  methods  for  the  quantitative  estimation  of  the 
halogens  by  the  formation  of  their  silver  salts,  the  further  treat- 
ment depending  on  whether  two  or  three  of  these  elements  are 
present.  In  all  cases,  however,  much  time  is  required  for  the 
analysis  and  great  care  in  the  manipulation  of  the  precipitates. 
Sexton  says,  in  his  work  on  Quantitative  Analysis,  Third^  Edi- 
tion, that  there  is  no  known  method  by  which  the  two  acids, 
hydrogen  bromide  and  ^hydrogen  chlorine  can  be  coi^pletely 
separated.  He  recommends  their  precipitation  as  silver  salts, 
the  weighing  of  this  product  and  the  conversion  of  the  bromide 
present  into  the  chloride  by  passing  chlorine  gas  over  the  fused 
mixtures.  Frpm  the  results  the  amount  of  each  halogen  is 
determined.  Of  course  the  general  procedure  could  be  used  if 
an  iodide  were  associated  with  the  chloride  but  would  not  be 
applicable  in  case  all  three  halogens  were  present. 

Dr.  Prescott,  in  the  Journal  of  Analytical  Chemistry^  8, 
gives  an  acceptable  method  for  the  estimation  of  bromine  in  the 
presence  of  chlorine  and  calls  attention  to  several  others  that 
have  been  employed.  Fresenius  gives,  on  pages  592  to  600  in 
the  Second  American  edition  of  his  work  on  Qualitative  Analysis, 
elaborate  methods  for  determining  these  elements  in  all  possible 
mixtures  of  the  binary  compounds  of  these  elements.  They  are 
generally  difficult  of  application  and  employ  rare  reagents.  It 
may  be  said,  in  general,  that  all  methods  of  indirect  estimation 
of  the  halogens  in  mixtures  of  their  binary  compounds  are  troub- 
lesome,  although  some  of    the  recent  modifications  of  these 


692  CHIX>RINS,   BROMINE,   AND   lODINB. 

utes  time  was  used  for  each  distillation,  but  less  time  was  usually 
sufficient.  In  fact  most  of  the  halogens  were  driven  over  during 
the  first  few  minutes  of  heating  after  boiling  temperature  was 
reached.  In  case  very  great  accuracy  is  not  required  an  esti- 
mation can  be  completed  in  a  few  minutes. 

It  is  evident  that  in  all  cases  there  must  be  relatively  large 
excess  of  reagents.  When  the  distillations  were  complete  the 
iodine  set  free  in  the  receiver  was  titrated  against  decinormal 
sodium  thiosulphate.  The  titration  can  be  made  in  the  receiver 
but  it  was  found  most  convenient  to  pour  the  liquid  into  a  six- 
inch  evaporating  dish  before  estimation.  • 

The  contents  of  the  flask  are  now  removed  to  a  beaker  and 
the  excess  of  the  permanganate  reduced  by  ferrous  sulphate, 
adding  sulphuric  acid  enough  to  render  the  solution  clear.  The 
solution  was  slightly  warmed  to  hasten  the  action.  It  was  then 
cooled  and  made  .up  to  a  definite  volume  and  an  aliquot  part 
estimated  by  precipitation  with  silver  nitrate.  There  was  noth- 
ing to  prevent  the  estimation  of  the  chlorine  by  titration,  but  no 
determinations  were  made  by  that  method. 

The  following  tables  give  the  results  of  the  work  : 

Potas-  Iodine  Potas-  Bromine  Potas-  Chlorine 

sium  in  potas-  sium  in  potaa-  sium    in  potas- 

iodide      ainm  Iodine  bromide       sium     Bromine  chloride    slum     Chlorine 

taken,  iodide,  found,     taken,  bromide,    found,     taken,  chloride,    foand. 

1  0.986  0.0754  0.0745  0.198  O.T330  0.1329  0.994  0.4869  0.4829 

2  0.493  0-0377  0-0374  0.198  0.1330  0.1299  0.994  0.4869  0.4859 

3  0.493  0.0377  0.0377  0.099  0.0665  0.0658  1.988  0.9738  0.9699 

4  0.493  0.0377  0.0376  0.099  0.0665  0.0662  0.994  0.4869  0.4870 

5  0.493  0.0377  0.0378  0.099  0.0665  0.0659  0.994  0.4869  0.4858 

6  0.493  0.0377  0.0375  0.0495  0.0332  0.0328  0.994  0.4869  0.4867 

7  0.493  0.0377  0.0375  0.0495  0.0332  0.0331  0.994  0.4869  0.4857 

8  1.972  0.1508  0.1499  0.099  0.0665  0.0664  1.988  0.9738  0.9679 
9*  1.972  0.1508  0.1484  0.099  0.0665  0.0659  0.497  0.243  0.241 

10   0.493   0.0377   0.0375   0.0495    0.0332   0.0329   0.497   0.243     0.242 

The  tabular  statement  needs  no  particular  explanation.  The 
quantities  represented  are  the  amounts  in  grams  in  each  case. 

It  may  be  well  to  note  that  this  general  method  is  applicable 
for  rapid  technical  estimations  of  bromine  or  of  iodine  either  by 
themselves  or  in  case  of  mixtures  of  the  same.  Single  analyses 
can  be  readily  made  in  ten  to  fifteen  minutes. 

Iowa  Agricultural  Colleob,  amb8,  Iowa. 


ON  THE  INVERSION  OF  SUGAR  BY  SALTS.     NO.  a. 

By  J.  H.  Long. 

Reed  red  June  a^,  1896. 

IN  a  recent  paper*  I  have  shown  that  in  their  behavior  with 
cane  sugar  solutions  many  so-called  neutral  salts  closely 
resemble  weak  mineral  acids.  Salts  of  the  heavy  metals  in 
general  have  the  power  of  inverting  sugar  solutions,  and  in  some 
cases  very  rapidly,  especially  at  an  elevated  temperature.  The 
same  fact  has  been  pointed  out  for  certain  salts  by  others,  nota- 
bly by  Walker  and  Aston,*  who  determined  the  speed  of  inver- 
sion of  four  nitrates,  comparing  them  with  dilute  nitric  acid. 
This  inversion  is  due  to  the  hydrolysis  of  the  salts  in  question, 
the  hydrogen  of  the  acids  formed  being  in  all  cases,  probably, 
the  active  catalytic  agent. 

In  my  former  paper  I  gave  some  results  obtained  in  a  prelimi- 
nary investigation  on  ferrous  iodide  with  very  strong  sugar  solu- 
tions, and  in  the  present  paper  I  shall  give  the  results  obtained 
with  other  salts,  as  well  as  more  extended  tests  with  the  iodide. 

METHOD. 

In  the  experiments  before  reported  I  made  very  strong  syrups 
containing  usually  fifty  grams  of  sugar  in  loo  cc,  and  to  these 
syrups  before  final  dilution  weighed  amounts  of  the  salts  were 
added,  the  volume  being  brought  up  to  loo  cc.  with  distilled 
water.  In  the  following  series  of  tests  the  amount  of  sugar 
present  is  much  smaller,  being  in  all  cases  fifty  grams  in  250  cc. 
of  the  finished  solution.  This  solution  is  much  stronger  than  is 
usually  employed  in  inversion  experiments,  but  with  many  of 
the  salts  dissolved  weaker  sugar  solutions  could  not  be  well 
used.  The  ferrous  salts,  especially,  require  relatively  large 
amounts  of  sugar  to  hold  them  in  clear  solution,  and  as  many  of 
the  experiments  given  below  were  made  on  such  salts,  it  was 
decided  to  employ  the  same  weight  of  sugar  in  all  cases.  For 
each  experiment,  therefore,  fifty  grams  of  pure  sugar  was  dis- 
solved in  water  in  a  250  cc.  flask  by  aid  of  heat.  The  strong 
syrup  was  cooled  and  to  it  was  added  the  salt  in  the  powdered 
form  or  dissolved  in  a  little  water.      After  securing  a  complete 

i  This  Journal,  i8,  lao. 
«/.  Chem.  Soc.,  July,  1S95. 


694  J-   H.   IX>NG. 

solution  in  either  way,  it  was  diluted  to  the  mark  and  shaken  to 
mix  thoroughly. 

The  syrup  so  made  was  poured  into  small  tubes  of  thin  glass 
for  inversion.  These  tubes  held  about  twenty  cc.  and  were 
three-fourths  filled.  They  were  cleaned  for  use  by  boiling  in 
hydrochloric  acid  and  then  in  distilled  water  repeatedly.  After 
having  been  employed  for  several  series  of  tests  it  was  found 
sufficient  to  soak  them  twenty-four  hours  in  weak  acid,  and 
then  in  distilled  water,  rinsing  thoroughly  finally.  After  receiv- 
ing the  sugar  solutions  they  were  closed  with  perforated  rubber 
stoppers  holding  each  a  short  glass  tube  with  capillary  opening. 
The  tubes  were  placed  in  a  receptacle,  which  was  finally 
immersed  in  the  water  of  a  thermostat  holding  over  twent>' 
liters.  The  receptacle  for  the  tubes  consists  essentially  of  two 
copper  disks,  twenty-five  cm.  in  diameter,  soldered  six  cm.  apart 
on  a  copper  rod  as  an  axis.  The  lower  disk  is  furnished  with 
fine  perforations,  and  the  upper  one  with  larger  openings  to 
receive  the  tubes.  The  copper  axis  below  the  lower  disk  ends 
in  a  hardened  point,  resting  in  a  socket,  and  is  extended  above 
to  a  length  of  fifteen  cm.,  ending  in  a  grooved  pulley  around 
which  a  belt  passes.  Power  applied  to  this  belt  rotates  the  tube 
receptacle,  which  at  the  same  time  keeps  the  water  of  the  ther- 
mostat in  motion.  The  thermostat  itself  consists  of  a  large 
copper  oven  covered  with  asbestos  boards  on  five  sides.  The 
top  has  perforations  for  the  temperature  regulator,  thermometer 
and  rotating  axis  of  the  tube  receptacle.  A  section  of  the  top 
can  be  quickly  removed  to  take  out  tubes,  but  at  other  times 
should  be  left  closed  to  exclude  light.  The  capillary  tubes  in 
the  stoppers  closing  the  inversion  tubes  project  about  two  cm. 
above  the  water. 

With  the  apparatus  employed  it  was  possible  to  maintain  a 
constant  high  temperature  with  a  little  watching  through  ten 
hours.  A  temperature  of  85°  was  held  with  variation  of  less 
than  o.  I**  in  either  direction.  With  many  salts  the  rate  of  inver- 
sion is  exceedingly  slow  at  ordinary  temperatures,  in  fact  almost 
imperceptible.  For  convenience  in  working,  therefore,  it  was 
found  necessary  to  invert  at  a  high  temperature,   and  85°  was 


INVBRSIOK  OP  SUGAR   BY  SAINTS.  695 

chosen.  In  a  few  instances  a  slightly  higher  temperature  was 
employed,  but  the  results  obtained  are  not  included  below. 

The  reaction  between  the  sugar  and  salt  is  probably  in  most 
instances  analogous  to  that  between  sugar  and  weak  acids,  and 
the  rate  of  inversion  may  therefore  be  expressed  by  the  same 
differential  equation : 

The  integration  of  this  for  /  and  x  =  o,  together,  leads  to  the 
well  known  formula  : 

K '=^—r  nat.  log.  —1 , 

where  A  represents  the  amount  of  sugar  present  at  the  beginning 
of  the  inversion,  x  that  inverted  at  any  time,  /,  of  an  observation, 
and  ^the  ** constant**  or  **  coefficient'*  of  inversion. 

As  the  reaction  is  most  easily  followed  by  means  of  the  polar- 
istrobometer,  A  is  conveniently  measured  by  the  total  change  in 
rotation  which  is  observed  between  the  beginning  of  the  reac- 
tion and  after  complete  inversion,  x  is  measured  by  the  change 
of  rotation  from  the  beginning  up  to  the  time,  /,  of  any  observa- 
tion. For  convenience  common  logarithms  are  employed  in  all 
the  calculations  below.  As  the  sugar  solutions  were  mixed  with 
the  inverting  substances  at  a  low  temperature,  the  intervals,  /, 
could  be  reckoned  only  from  the  time  when  the  mixtures  in  the 
tubes  had  reached  the  constant  temperature  of  the  experiment. 
Preliminary  tests  were  therefore  made  to  determine  several 
points  of  practical  manipulation.  The  thermostat  was  first 
brought  to  a  temperature  of  about  87**-88**,  and  the  filled  experi- 
mental tubes  and  their  receptacle  immersed  in  it.  From  this  a 
fall  of  temperature  resulted,  because  of  the  low  temperature  of 
the  solution.  In  five  or  six  minutes  the  constant  temperature  of 
Z^  was  reached,  and  by  regulation  of  the  gas  flame  this  was 
maintained.  In  another  set  of  expervpients  it  was  found  that 
the  solutions  in  the  experimental  tubes  could  be  brought  to  a 
temperature  of  85**  from  the  room  temperature  in  four  to  six 
minutes.  It  appeared,  therefore,  that  ten  minutes  was  amply 
sufficient  time  to  allow,  after  introducing  the  tubes  into  the 


696  J.    H.    LONG. 

thermostat,  before  beginning  the  actual  observations,  and  this 
was  done  in  all  cases  in  the  experiments  given  below.  In  the 
case  of.  bodies  which  invert  but  slowly  there  is  little  objection  to 
the  loss  of  this  first  ten  minutes  of  the  reaction,  but  in  a  few 
instances  it  was  found  to  be  a  decided  drawback,  as  will  be  seen 
below. 

Usually  250  cc.  of  the  solution  was  prepared  for  experiment, 
and  this  was  filled  into  fifteen  or  sixteen  tubes,  and  put  into  the 
thermostat.  At  the  end  of  ten  minutes  a  tube  was  withdrawn 
and  cool6d  very  quickly  by  immersion  in  cold  water,  or  by  hold- 
ing it  under  a  flowing  hydrant.  The  contents  were  then  poured 
into  a  polarization  tube  and  polarized  at  the  constant  tempera- 
ture of  20"*  in  most  cases.  In  a  few  tests  made  in  warm  weather  a 
temperature  of  25^  was  maintained  in  the  dark  room  and  in  the 
water  flowing  around  the  observation  tube.  This  first  observa- 
tion gives  the  initial  rotation,  and  the  time  of  removing  the  tube 
may  be  put  as  /=  o.  Tubes  were  removed  at  different  inter- 
vals following  and  treated  in  the  same  manner.  The  results  of 
the  polarizations  were  always  very  constant  during  the  first  few 
hours  heating  in  the  thermostat,  as  was  found  by  removing  and 
polarizing  the  contents  of  three  tubes,  but  after  five  or  six  hours 
less  regular  results  were  found,  and  I  adopted  the  plan  of  taking 
the  mean  result  obtained  by  examining  two  or  three  tubes. 
With  fifteen  or  sixteen  tubes  I  made  observations  at  eight  or 
nine  intervals. 

After  polarizing  the  liquids  in  the  last  tubes  removed,  the 
contents  were  mixed,  returned  to  a  tube  and  heated  longer  to 
obtain  the  end  point  of  the  reaction,  that  is,  the  point  of  com- 
plete inversion.  The  point  found  in  this  manner  does  not 
always  agree  with  that  calculated  from  the  known  weight  of 
pure  cane  sugar  in  the  original  solution.  Even  with  dilute 
acids  the  phenomenon  of  inversion  is  not  as  simple  a  thing  as 
usually  represented.  As  shown  by  Gubbe*  and  others,  the 
specific  rotation  of  invert  sugar  depends  not  only  on  the  concen- 
tration, but  on  the  time,  temperature  and  acid  used.  Prolonged 
heating  with  salts  produces  in  many  cases,  apparently,  a  slight 

1  Bn-.  d.  chem.  Ges.,  i8,  2207. 


INVERSION   OP   SUGAR   BY   SAI.TS.  697 

decomposition  of  the  levulose,  from  which  the  negative  rotation 
of  the  invert  sugar  is  found  smaller  than  it  should  be  theoretic- 
ally. In  a  few  instances,  however,  the  negative  rotation  of  the 
invert  sugar  was  increased.  From  the  experiments  of  Gubbe  it 
may  be  calculated  that  fifty  grams  of  cane  sugar  in  250  cc. 
would  yield  a  solution  after  inversion,  which  in  a  200  mm.  tube 
should  show  a  negative  rotation  of  -—8.6°.  The  rotation  observed 
in  my  experiments  was  usually  about  — 8.3^,  but  an  accurate 
determination  was  not  always  possible,  as  some  of  the  solutions 
became  slightly  colored  before  inversion  was  quite  complete,  and 
in  other  cases  a  negative  rotation  once  observed  seemed  to  grow 
slightly  less  on  longer  heating,  making  the  exact  end  point 
somewhat  uncertain.  The  discrepancies  were  not  large  in  any 
case,  however,  and  I  decided  to  take  — 8. 3'' as  the  true  end  point 
for  the  200  mm.  tube,  and  — ^4.15*  for  the  100  mm.  tube. 

With  some  of  the  salts  examined  the  velocity  coeflBcient,  A',  is 
practically  constant,  with  others  it  increases  with  the  time, 
while  in  still  other  cases  it  decreases. 

The  sugar  used  in  all  the  experiments  was  crystallized  cut 
loaf  of  high  degree  of  purity,  and  selected  for  the  purpose. 
With  fifty  grams  in  100  cc.  it  yields  a  solution  of  almost  perfect 
clearness,  which  can  be  easily  polarized  in  a  400  mm.  tube. 
Weaker  solutions  yield,  on  inversion,  results  which  agree  per- 
fectly with  the  theoretical  requirement. 

POTASSIUM   ALUM. 

Solutions  of  this  salt  invert  very  rapidly.  A  sample  of  pure 
alum  was  crystallized  several  times  from  distilled  water  to  secure 
a  product  free  from  traces  of  uncombined  sulphuric  acid,  some- 
times present  in  the  commercial  article.  This  carefully  purified 
salt  was  used  in  all  the  inversion  tests.  In  the  tables  below,  t 
refers  to  the  time  in  minutes,  and  under  a  is  given  the  observed 
angle  of  rotation  in  degrees  and  hundredths. 


69S 


J.   H.   lOHG. 


\l,(SO.),.24H.O. 

250  cc.,  fifty  grams 

of  sugar 

+  0.617  S™n  of  alum. 

^  =  33.03* 

• 

/.                a. 

^»  • 

t     "^  A^ 

0            24.7s*' 

«  •  •  • 

•  *  •  * 

•  V  •  s 

15             ao.15 

4.58^ 

0.06483 

0.00432 

50            16.16 

8.57 

0.13045 

0-00434 

60                9-75 

14.98 

OLJ6243 

0.00437 

90                4* 

19.88 

0-39998 

O.OCH44 

lao                1.25 

23.48 

0.53891 

0.00449 

I5D                —1. 3D 

26.^5 

0.67581 

0.00449 

210            —4-88 

29.61 

0.98488 

0.00468 

270            -6.40 

51.13 

I.240I6 

OJOKHS9 

K,Al,(SO.),.24H.O. 
In  250  cc.,  fifty  grains 


/. 


2. 

N 

of  sagar-|-  1.234  gnuns  of  alum. 


^  =  32.37- 


X. 


0 

24.or 

■  *  •  • 

IS 

J  7.83 

6. 24' 

30 

12.92 

11.15 

60 

550 

18.57 

90 

0-75 

23-32 

120 

-2.48 

26,55 

150 

-4.76 

28.83 

210 

—7.00 

31.07 

270 

-7.80 

31.87 

ill*     ^ 

^'  A^x 

'   lor     ^ 

0.09300 

0.00620 

0.18339 

0.0061 « 

0.37026 

0.00612 

0.55349 

0.00615 

0.74522 

0.00621 

0.961 14 

0.00641 

1.39620 

0.00661 

I.81II7 

0.00670 

inversion  op  sugar  by  salts.  699 

Experiment  3. 

K,Al,(SOj,.24H.O.    -^ 

In  250  cc.,  fifty  grams  of  sugar  +  2.468  grams  of  alum. 

A  =  31.25. 


/. 

a. 

X. 

]     log.  / 
t              A-^x 

0 

22.95^ 

.  • .  • 

.  • .  • 

.... 

15 

T4.80 

8.15° 

0.13124 

0.00875 

30 

8.79 

14.16 

0.262 I I 

0.00873 

60 

1.07 

21.88 

0.5231 I 

0.00872 

90 

—3.03 

25.9« 

0.77304 

aoo859 

120 

-5.64 

28.59 

1.06997 

0.00891 

180 

—7.53 

3048 

1.55533 

0.00864 

340 

—8.15 

31.10 

2.31866 

0.00966 

0.00886 
Experiment  4. 

K,Al.(SO,),.24H.O.     ^. 

In  250  cc.,  fifty  grams  of  sugar  +  4.936  grams  of  alum. 

A  =  30.03. 


/. 

a. 

X, 

A—x 

»  log   ^ 
t        *  A^x 

0 

21.73'' 

a  .  •  • 

.... 

•  ■  •  • 

15 

11.23 

10.50° 

0.18686 

0.01245 

30 

4.45 

17.28 

0.37205 

0.01240 

60 

-2.87 

24.60 

0.74276 

0.01238 

90 

—5.95 

27.68 

1. 10649 

0.01230 

120 

—7-34 

29.07 

1.49529 

0.01244 

180 

-8.18 

29.91 

2.39838 

0.01332 

0.01255 
Experiment  5. 

K.Al,(SO,),.24H,0.    ^. 

In  250  cc,  fifty  grams  of  sugar  +  9.872  grams  of  alum. 

A  =  29.19. 


/. 

a. 

X. 

log.-l_. 
A-^x 

;  log. ^-^ 

/      A-^x 

0 

20.89<5 

.  .  .  • 

«  •  •  • 

• 
•  •  •  • 

10 

10.89 

10.00° 

O.18216 

0.01822 

25 

2.10 

18.79 

0.44820 

0.01793 

50 

—4.70 

25.59 

0.90893 

O.OI818 

9U 

*— 7.73 

28.62 

1.70936 

0.01899 

150 

—8.25 

29.14 

2.76626 

0.01844 

0.01835 


TOO  J.    e.    LONG. 

An  attempt  was  made  to  invert  with  a  half  normal  solution 
but  at  the  temperature  employed  the  rate  was  found  to  be  too 
rapid  for  accurate  obsen-ation. 

With  the  first  four  solutions  no  difficult}*  was  found  in  mak- 
ing accurate  polarimetric  obsen'attons  in  the  200  mm.  tube. 
The  last  solution,  however,  became  final!)'  somewhat  colored, 
and  slightly  turbid  from  precipitation  of  what  appeared  to  be 
aluminum  hydroxide.  A  portion,  heated  180  minutes,  became 
too  turbid  for  direct  reading  and  had  to  be  examined  in  the  100 
mm.  tube  after  filtration.  The  rotation  was  found  now  to  be 
— S-to",  corresponding  to  — 7.20°  for  the  200  mm,  tube,  instead 
of  — 8,25°  or  — 8.30°.  From  the  slight  concentration  due  to  the 
filtration  a  still  greater  negative  value  instead  of  a  lower  one 
should  be  expected.  We  have  here  an  illustration  of  the  fact 
referred  to  abo^-e,  ru.,  that  prolonged  heating  makes  the  end 
point  determination  somewhat  uncertain  at  times. 

It  is  interesting  to  note  the  relation  existing  between  the  con- 
centrations of  the  solutions  and  their  rates  of  inversion  in  the 
above  examples.  For  comparison  we  can  call  the  lowest  con- 
centration unity  and  arrange  them  as  follows  : 

Cone  A" 

--  I  0.00446 

■''  1  0.0063? 


^^  S  O.OIJ55 

Inspection  ot  the  table  shows  that  the  coefficient.  A",  increases 
rapidly  with  the  concentration,  but  is  not  directly  proportional 
to  it.  It  is  apparent  that  the  numbers  in  the  third  colnmn  i-ary 
approximately  as  the  square  nx>is  ot  tho^  in  the  second,  which 
is  cleariy  s^hown  in  the  next  table. 

I  coo44^  0.00116 


»t4l 


INVERSION  OF  SUGAR   BY   SAI.TS.  7OI 

The  regular  results  obtained  from  the  aluminum  salt  are 
probably  due  in  a  measure  to  the  inertness  of  the  hydroxide 
toward  sugar,  as  well  as  to  the  behavior  of  sulphuric  acid  in 
inversion.  The  bases  of  the  other  salts  examined  below  form 
combinations  with  sugar  more  or  less  readily,  not  only  with  sac- 
charose, but  also  with  the  products  of  inversion,  so  that  the  nor- 
mal results  of  the  reaction  may  be  modified  in  a  manner  difficult 
to  compute.  The  rather  rapid  rate  of  inversion  in  the  above 
points  to  a  relatively  great  degree  of  hydrolysis  in  the  alum. 
Walker  and  Aston*  found  something  similar  in  a  half  normal 
solution  of  the  nitrate,  studiedat  a  temperature  of  80".  From  their 
polarizations  a  value  of  0.0077  ^^^  ^  was  found,  and  this  was 
much  in  excess  of  the  values  found  for  other  salts  at  the  same 
time. 

FERROUS  SULPHATE. 

A  sample  of  the  purest  obtainable  sulphate  was  reccystallized 
from  water  containing  a  trace  of  sulphuric  acid,  then  dissolved 
in  distilled  water  and  precipitated  by  alcohol.  The  crystal 
meal  secured  was  washed  several  times  with  alcohol  and  dried 
by  fanning.  The  finished  product  was  bright  green  and  gave  a 
nearly  clear  solution  with  pure  water.  It  still  held  a  trace  of 
alcohol  as  disclosed  by  the  odor.  The  experimental  solutions 
were  made  by  dissolving  the  sugar  first  and  adding  to  this  s^Tup 
the  weighed  sulphate  meal.  The  mixtures  were  shaken  to  com- 
plete solution  without  application  of  heat,  and  then  poured  into 
the  tubes  for  inversion.  The  solutions  soon  became  turbid  on 
warming  and  a  minute  amount  of  flocculent  precipitate  sepa- 
rated, making  direct  polarization  impossible.  The  readings 
could  be  made  therefore  only  after  filtration,  which  was  not 
without  slight  effect  on  the  result.  The  total  amount  of  sepa- 
rated hydroxide  or  basic  salt  was  and  remained  through  the 
test,  minute. 

1  Loc.  eiL 


702 


J.   H.   LONG. 


EXPBRIMBNT    6. 

FeSO,.7H,0.     ^. 

In  250  cc,  fifty  grams  of  sugar  +  17.38  grams  of  sulphate. 

A==i  17.12. 


/. 

o 

15 
45 
75 
135 
195 
255 
375 
495 


a. 

12.97 
12.48 
11.50 
10.40 

8.43 
6.72 

5" 
2.87 
1.03 


X. 

. . .  • 

0.49 
1-47 
2-57 
4.54 
6.25 

7.76 
10.10 
11.94 


log. 


A^x 


I 


log. 


A^x 


O.OI261 

0-03899 
0.07064 

0.13382 

0.19727 

0.26222 

0.38716 

O.51917 


0.00084 
0.00086 
0.00094 
0.00099 
O.OOIOI 

0.00102 
0.00103 
0.00105 

0.00099 


Experiment  7. 

FeS0,.7H,0.     N, 

In  250  cc,  fifty  grams  of  sugar  +  34.75  grams  of  sulphate. 


/!=  17.10. 

/. 

a. 

X. 

'^'  A-x 

X  ,     A 
i    '"^A^x 

0 

12.95 

.... 

15 

12.45 

0.50 

0.01289 

0.00086 

45 

11.26 

1.69 

0.04520 

O.OOIOO 

75 

10.08 

2.87 

0.07980 

0.00106 

135 

8.07 

4.88 

0.14593 

0.00108 

195 

6.30 

6.65 

0.21388 

O.OOIIO 

255 

4.70 

8.25 

0.28606 

O.OOII2 

375 

2:25 

10.70 

0.42682 

0.00II4 

495 

0.15 

12.80 

0.59953 

O.OOI2I 

0.00107 
Other  tests  were  made  with  a  second  preparation  of  ferrous 
sulphate  from  which  the  alcohol  had  not  been  as  completely 
removed.  For  a  half  normal  solution  the  coefficient,  0.00094^ 
was  found,  and  for  a  normal  solution  the  value,  o.ooioo,  both 
results  being  but  a  trifle  lower  than  those  obtained  from  the 
pure  products.  It  is  possible  that  the  differences  may  be  due 
to  the  presence  of  the  trace  of  alcohol.      In  any  case  it  is  evi- 


INVERSION  OF  SUGAR   BY  SALTS.  703 

dent  that  with  solutions  as  strong  as  those  used  the  larger 
amount  of  sulphate  inverts  but  little  more  rapidly  than  the 
smaller. 

AMMONIUM  FERROUS  SUPHATE. 

But  one  experiment  was  made  with  this  salt,  a  very  nice  crys- 
tallized preparation  being  used. 

Experiment  8. 

(NHj,Pe(SO.),.6H.O.    4' 

In  500  cc.y  100  grams  of  sugar  +  49  grams  of  sulphate. 

A  =  17.08. 


•  •  .  • 

0.44 

0.01 134 

.  0.00066 

1.19 

0.03137 

0.00069 

2.00 

0.05409 

0.00072 

2.80 

0.07776 

0.00074 

4.33 

0.I269S 

0.00077 

563 

0.17368 

0.00077 

8.33 

0.29048 

0.00084 

10.08 

0.38739 

0.00083 

10.73 

0.42972 

0.00082 

o  12.93 

17  12.49 

45  "-74 

75  10.93 

105  io.13 

165  8.60 

225  7.30 

345  4.60 

465  2.85 

525  2.20 

0.00076 

The  coefficient  is  seen  to  be  low,  but  nearly  a  constant.  In 
this  case,  as  in  that  of  the  ferrous  sulphate,  the  mixture  became 
slightly  turbid  on  heating. 

ZINC  SULPHATE. 

It  is  practically  difficult  to  secure  a  good  preparation  of  zinc 
sulphate  crystallized  without  the  addition  of  a  trace  of  sulphuric 
acid.  In  absence  of  the  acid  crystallization  is  very  slow.  The 
preparation  used  below  was  made  from  a  chemically  pure  com- 
mercial sample,  by  crystallizing  with  a  trace  of  acid  first  and 
then  from  pure  water,  after  heating  the  solution  with  pure  zinc. 
The  final  crystallization  to  secure  fifty  grams  required  weeks  for 
its  completion.  In  my  former  paper  attention  was  called  to  the 
fact  that  Inversion  with  zinc  sulphate  is  very  slow,  which  is  well 
shown  below.     The  experiment  was  closed  when  the  sugar  was 


704  J.    H.    LONG. 

about  half  inverted,  and  as  the  coefficient  is  not  regular,  it  is 
not  possible  to  estimate  accurately  the  mean  rate  for  the  whole 
period. 

Experiment  9. 

ZnSO,.7H.O.     7' 

In  250  cc,  fifty  grams  of  sugar -f-  17.94  grams  of  the  sulphate. 

A  =  17.25. 


/. 

a. 

X. 

A 
^  A—x' 

'  lair   '* 

/  ^^"  A^x 

0 

13.10 

•  •  •  • 

15 

12.88 

0.22- 

o.oossS 

0.00037 

45 

".35 

0.75 

0.01935 

0.00043 

105 

11.34 

1.76 

0.04674 

O.O0Q44 

165 

10.40 

2.70 

0-07393 

O.OOQ45 

^5 

8.48 

4.62 

013539 

0.00048 

405 

6-51 

6.59 

0.20933 

0.00052 

525 

4.68 

8.42 

0.29083 

o.ou.>S5 

M.\NG.\XOUS  SULPHATE. 

After  several  attempts  a  salt  was  obtained  crystallized  from 
perfectly  neutral  solution.  Some  of  the  cr>-stals  were  so  irregu- 
lar in  outline  that  it  was  not  possible  to  determine  from  inspec- 
tion whether  they  contained  four  or  five  molecules  of  water. 
Determination  of  SO^  in  the  product  showed,  however,  that  a 
very  small  amount  only  of  the  latter  salt  was  present.  In  mak- 
ing the  solutions  I  assumed  for  conx-enience  that  the  compound 
had  the  formula  MnS0«.4H,0,  and  weighed  out  accordingly. 

As  I  pointed  out  in  my  former  paper,  a  solution  of  manganous 
sulphate  and  sugar  undergoes  a  peculiar  decomposition  when 
heated,  in  which  a  ver>'  fine  dark  substance  is  thrown  out  from 
solution.  The  amount  of  this  is  so  small  that  I  could  not  collect 
enough  for  tests,  in  the  work  done,  but  it  is  still  sufficient  to 
make  the  polarimeter  readings  ver>'  difficult.  All  solutions  had 
to  be  filtered  before  examination,  but  even  with  this  precaution 
the  readings  were  often  obscure. 


INVERSION   OP  SUGAR  BY  SALTS.                             705 

EXPERIMENT  10. 

MoSO,.4H.O,      ^• 

In  250  cc,  fifty  grams  of  sugar  +13.94  g^ams  of  sulphate. 

A  =  34.80. 

o               26.50°               ....  

45      26.33      0.17°  0.00213  0.000047 

75      26.15      0.35  0.00439  0.000058 

135      25.76      0.74  0.00934  0.000069 

195      25.05       1.45  0.01848  0.000095 

315      22.33      417  0.05543  0.000176 

435       19.84       6.66  0.09226  0.000212 

535       16.75       9.75  0.14277  0.000257 

Experiment  ii. 
MnS0..4H,0.    N, 

In  250  cc,  fifty  grams  of  sugar  +  27.88  grams  of  sulphate. 

^  =  34.75. 

/.                 or.                  X,  10^.;^.      -^^o^i^- 

o                26.45^               .  • . .  

15       26.25       0.20^'  0.00250  0.00017 

45       26.00       0.45  0.00566  0.00013 

75       25.75       0,70  0.00883  0.00012 

135       24.90       1.55  0.01981  o.ooois 

195       23.00       3.45  0.04541  0.00023 

315       18.20       8.25  0.1 1 770  0.00037 

435       14-45       '2.00  0.18397  0.00042 

555        9-50      16.95  0.29053  0.00052 

Experiment  1.2. 
MnS0,.4H,0.  2N, 

In  250  cc,  fifty  grams  of  sugar  +  55.76  grams  of  sulphate. 

A  =  34.42. 

/.                  a.                   X,  'o^'iih'        -r"  ^°»- T^lr' 

o                26.12^               —  

30         2^.12          1. 00  0.01280  0.00043 

90         22.80          3.32  0.04405  0.00049 

150       17.82       8.30  O.I  1984  0.00080 

220       12.33       13.79  0.22231  O.OOIOI 

338          4.60         21.52  0.42622  0.00126 

450          0.27         25.85  0.60383  0.00134 

570        —3-80         29.92  0.88360  0.00155 


706  J.   H.   LONG. 

EXPERIMBNT  13. 

MnS0,.4H,0.     3N. 

In  250  cc.,  fifty  grams  of  sugar  -|-  83.64  grams  of  sulphate. 

A  =  34.00. 

t  a  cT.  log— d —       _!_  log.  -A—^ 

O  25  •  7^        •*..          •*••  *... 

30       24.22         1.48°  0.01933  0.00064 

90      T8.76       6.94  o.%99i5  o.ooiio 

150  11.50  14.20  0.23481  0.00156 

220       4.75  20.95  0.41587  0.00189 

338  —2.69  28.39  0.78252  0.00232 

450  — 6.25  31-96  1. 22185  0.00272 

570  —8.05  33.75  2.13354  0.00363 

The  rates  of  inversion  cannot  be  directly  compared  in  the 
above  experiments  because  the  latter  were  not  carried  to  com- 
pletion. In  the  first  case  over  one-third  of  the  sugar  originally 
present  was  inverted,  in  the  second  case  almost  exactly  one-half, 
in  the  third  case  about  six-sevenths,  while  in  the  last  case  the 
inversion  was  very  nearly  complete.  By  plotting  the  results  it  is 
possible  to  determine  approximately  the  rate  of  inversion  when 
just  one-half  of  the  sugar  has  been  inverted  and  this  I  have 
done.  The  results  are  given  below,  and  show  that  the  coeffi- 
cients, K^  are  nearly  proportional  to  the  concentrations,  these 
being  referred  to  that  of  the  half-normal  solution  as  unity. 

Cone.  K. 

1  (0.00032) 

2  0.00054 
4  0.00109 
6                      0.00172 

The  first  coefficient,  0.00032,  is  uncertain  because  it  was 
found  by  a  rather  wide  extrapolation,  but  between  the  others 
there  is  fair  agreement. 

MANGANOUS  CHLORIDE. 

The  salt  used  was  purified  by  several  crystallizations  from  the 
best  obtainable  Schuchardt  product. 


invbrsion  op  sugar  by  sai«ts.  707 

Experiment  14. 
MnCl,.4H,0.    ^• 

In  250  cc.,  fifty  grams  of  sugar  +6-18  grams  of  chloride. 

A  =  35.00. 


/. 

a. 

X, 

log./* 

0 

26.70° 

•  •  •  • 

• .  ■ . 

...  a 

15 

26.60 

0.10° 

0.00124 

0.00009 

45 

26.30 

0.40 

0.00499 

O.OOOII 

75 

26.00 

0.70 

0.00878 

O.OOCI2 

135 

2525 

1.45 

0.01838 

0.00014 

255 

22.66 

4.04 

0.05327 

0.00021 

375 

18.75 

7.95 

O.I I 190 

0.00030 

The  high 

initial  rotation  here  is  ' 

very  extraordinary,  corres- 

ponding  to  a  specific 

rotation  of  66.75'. 

« 

Experiment 

15. 

MnCl,.4H,0.  -^• 

In  250  cc. 

,  fifty  grams  of  sugar  + 

12.35  grams 

of  chloride. 

A  =  34.84 

• 

0 

a. 

•26.54^ 

X, 

.... 

log.-!-. 

A^x 

•  •   •  • 

t              A^x 

.... 

15 

26.45 

0.09° 

0.001 13 

0.00008 

45 

26.16 

0.38 

0.00476 

O.OOOII 

75 

25.85 

0.69 

0.00869 

0.00012 

135 

24.52 

2.02 

0.02594 

0.00019  ' 

255 

22.26 

4.28 

0.05693 

0.00022 

375 

17.15 

9-39 

0.13639 

0.00036 

495 

13.00 

13.54 

0.21370 

0.00043 

555 

11.52 

15.02 

0.24498 

0.00044 

Experiment  16. 
MnCl,.4H,0.    N. 

In  250  cc,  fifty  grams  of  sugar  -|-  24.70  grams  of  chloride. 

A  =  34.63. 


L 

a. 

X. 

log.    " 
A^x 

■T-'^iaF 

0 

26.33° 

•  •  •  • 

•  • . . 

•  •  •  • 

15 

2)S.I2 

0.21° 

0.00264 

0.00018 

30 

25.86 

0.47 

0.00593 

0.00020 

60 

25.15 

1. 18 

0.01505 

0.00025 

120 

23.05 

3.28 

0.04321 

0.00036 

180 

20.07 

6.26 

0.08659 

9.00048 

300 

15.60 

10.73 

O.16105 

0.00054 

708  J.    H.    LONG. 

Experiment  17. 

MnCl,.4H,0.  2iV. 

In  250  cc,  fifty  grams  of  sugar  +  49.40  grams  of  chloride. 

A  =  34.18. 


/. 

a. 

X, 

'o«^-  At.- 

I  ,     A 

0 

25.88^ 

•  •  •  ■ 

•   •   •   ■ 

•  •  •  • 

15 

25.38 

0.50^ 

0.00640 

o.cx)Q43 

45 

23.91 

1.97 

0.02578 

0.00057 

75 

22.21 

3.67 

0.04933 

0.00065 

135 

18.25 

763 

O.IO97I 

0.00081 

195 

14.80 

11.08 

O.I  7016 

0.00087 

345 

5.25 

20.63 

o.<doi83 

0.00116 

No  very  plain  relation  can  be  found  connecting  these  rates  of 
inversion.  The  coeflScients  corresponding  to  the  time  of  com- 
pletion of  one-third  of  the  inversion  are  here  given. 

COQC.  IC. 

I  (0.00038) 

2  0.00041 

4  0.00055 

8  0.00088 

The  first  coefficient  had  to  be  estimated  and  is  uncertain. 

FERROUS  CHLORIDE. 

Considerable  difficulty  was  experienced  fn  preparing  a  solu- 
tion of  ferrous  chloride  devoid  of  traces  of  free  acid.  A  weighed 
excess  of  pure  iron  wire  was  covered  with  water  in  a  small  flask 
and  then  the  calculated  volume  of  titrated  hydrochloric  acid  was 
added  in  amount  just  sufficient  to  produce  the  solution  of 
required  strength.  The  mixture  was  gently  warmed  and  allowed 
to  stand  a  short  time.  Wanning  was  repeated  at  intervals 
through  several  hours,  until  the  liberation  of  hydrogen  became 
very  feeble.  The  solution  so  obtained  stood  five  days  in  the 
presence  of  the  excess  of  iron,  being  boiled  twice  in  the  interval, 
and  was  then  filtered  cold  into  the  sugar  solution,  which  was 
made  up  to  the  proper  volume  with  fresh  distilled  water. 

The  actual  strength  of  solutions  made  in  this  manner  was 
determined  by  titration  later.  The  two  following  were  almost 
exactl}'  normal  and  half-normal. 

Both  solutions  became  turbid  on  heating  and  had  to  be  fil- 


INVERSION  OP  SUGAR   BY  SALTS.  709 

tered  before  polarization  for  the  first  tests.  After  the  lapse  of 
about  two  hours  the  cloudiness  disappeared  and  the  solutions 
then  taken  from  the  thermostat  were  clear  enough  for  direct 
polarization. 

KXPBRIMBNT  18. 

Feci,     ^. 

In  250  cc,  fifty  grams  of  sugar  +  7.925  grams  of  chloride. 


A  =  16.18. 

/. 

a. 

X. 

log.  '^  . 
A'^x 

;  log.-/ 

/      A'-x 

0 

12.03° 

.... 

.... 

.... 

15 

9-47 

2.56° 

0.07507 

0.00500 

45 

7.44 

4.59 

0.145 1 7 

0.00322 

105 

5.00 

703 

0.24783 

0.00236 

165 

3.83 

8.20 

0-30725 

0.00186 

285 

1.90 

10.13 

0.42749 

0.00150 

405 

—0.50 

12.53 

0.64696 

0.00160 

525 

— 2.01 

14.04 

0.87884 

0.00167 

KXPBRIMBNT  19. 

FeCl,.  iV. 

In  250  cc,  fifty  grams  of  sugar  +  15.85  g^ams  of  chloride. 

A  =  15.71. 


/. 

a. 

X. 

log.  '^  . 
A'-x 

__  log.  — 

/      A^x 

0 

11.56'^ 

.... 

•  •  •  • 

•  •  •  • 

15 

9.40 

2.16° 

0.06424 

0.00428 

45 

6.88 

4.68 

0:15360 

0.00341 

105 

4.75 

6.81 

0.24679 

0.00235 

165 

3.56 

8.Q0 

0.30913 

0.00188 

285 

1. 15 

10.41 

0.47190 

0.00165 

405 

—1.42 

12.98 

0.76002 

0.00187 

525 

—3.25 

14.81 

I.24194 

0.00236 

These  results  are  very  surprising,  inasmuch  as  they  show  but 
little  difference  between  the  rates  for  the  two  concentrations.  In 
both  instances  the  rates  rapidly  decrease  from  the  beginning 
and  after  the  sugar  has  been  about  half  inverted  they  increase  a 
little.  I  give  next  some  results  from  solutions  which  had  not 
been  boiled  so  thoroughly,  and  which  may  have  held  a  little 
free  acid. 


yio 


J.   H.    LONG. 


Experiment  20. 
FeCl,.  o.52A^. 

In  250  cc,  fifty  grams  of  sugar  H*  8.242  grams  of  chloride 

A  =  34.00. 

A  I    ,__     ^ 


/. 

a. 

X. 

0 

25.70° 

•  •  •  • 

15 

20.00 

5.70' 

45 

12.80 

12.90 

75 

9.00 

16.70 

105 

6.19 

19.51 

165 

2.42 

23.28 

285 

—1.95 

27.65 

345 

-3.96 

29.66 

405 

—5.30 

31.00 

lOf. 


A^x 


0.07969 
0.20720 
0.29343 
0.37041 
0.50129 
0.72771 

0.89399 
1.05436 


i  A — X 


0.00531 
0.00460 
0.00391 
0.00353 
0.00304 
0.00256 
6.00258 
0.00260 


Experiment  2 1 . 
FeCl,.  0.98^. 
In  250  cc,  fifty  grams  of  sugar  -|-  15.53  grams  of  chloride. 

A  =  33.60. 

A 


t, 
O 

15 

45 

75 

105 

165 

285 

345 


a. 

25.30^ 

19.84 

14.20 

11.80 

9.88 

7.65 

1.45 

—1.50 


X. 

.... 

5.46° 
II. 10 

1350 

15.42 

17.65 

23.85 
26.80 


ioi:. 


A—x 


_i_ 


log. 


A^x 


0.07702 

O.I 74 16 

0.22314 
0.26675 
0.32358 

OS53734 
0.69383 


0.00513 
0.00387 
0.00297 
0.00254 
0.00196 
0.00x88 
0.00201 


T^he  effect  of  free  acid  is  not  apparent.     Six  other  experiments 
ere  macie  with   normal  and  half-normal  ferrous  chloride  solu- 
^2j    ^'  ^^^  results  of  which  were  very  similar  to  those  above.    In 
tjozj    ^^^  ^^e  constant  was  found  to  increase  before  the  comple- 
Th^  ^  invei^sion. 

^^^*aa    ^^^stant  for  0.00 1  A^  hydrochloric  acid  was  determined 
^^e  ^  ^J^^iison  at  the  same  temperature,  /=  85**,  and  with  the 

^^'^nt  of  sugar.     It  was  found 

A^=  0.0051, 

'^Ai  ^^  ^  FBRROUS  BROMIDE. 

^s  of  this  salt  ipvere  made  by  adding  the  proper  amouflt 


INVERSION  OF  SUGAR  BY  SALTS. 


711 


of  bromine  to  an  excess  of  iron  and  water.  A  reaction  soon 
begins  which  is  hastened  by  heat.  Finally  the  solution  is  thor- 
oughly boiled,  which  eliminates  all  free  bromine  and  leaves  the 
iron  in  the  ferrous  condition.  It  is  then  filtered  into  the  cold 
sugar  solution  and  is  ready  for  use.  A  solution  so  made  is  prac- 
tically neutral. 

Experiment  22. 
FeBr,.  0.54  N. 
In  250  cc,  fifty  grams  of  sugar  -|-  14.58  grams  of  bromide. 

A 


t. 
O 

»5 

45 

75 

105 

165 

285 
345 


a. 

23.13* 
16.76 
9.76 
6.07 
4.00 
0.68 

•-4.03 
— 5-6o 


X, 

... 

6.3/- 

13-37 
17.06 

19-13 

22.45 
27.16 

28.73 


log. 


A-'X 


-7-lOff- 


A—x 


0.09836 
0.24062 
0.33988 
0.40743 
0.54406 
0.86691 
1.06598 


0.00655 
0.00534 
0.00453 
0.00388 
0.00329 
0.00304 
0.00309 


Experiment  23. 
FeBr,.  1.04  N. 

In  250  cc.,  fifty  grams  of  sugar  +  28.08  grams  of  bromide. 

A  =  29.50. 


/. 

a. 

X. 

^^«^-  A^x 

1  ,     A 
i     '^'A^x' 

0 

21.20° 

•  •  •  » 

.... 

.... 

15 

13.60 

7.60° 

0.12938 

0.00862 

45 

6.22 

14.98 

0.30785 

0.00684 

75 

2.90 

18.30 

0.42060 

0.00561 

105 

0.75 

20.45 

O.513I7 

0.00489 

165 

— 2  70 

23.90 

0.72163 

0.00437 

285 

—6.35 

27.55 

I. 17979 

0.00414 

345 

—7.50 

28.70 

1.56673 

0.00454 

The  normal  solutions  here  invert  but  little  faster  than  the 
half«normal.  The  rates  in  both  cases  diminish  rapidly  from  the 
start,  but  after  the  middle  of  the  inversion  become  nearly  con- 
stant, as  was  observed  with  the  ferrous  chloride.  The  first  three 
of  the  solutions  taken  from  the  thermostat  had  to  be  filtered 
before  polarizing. 


712  J.    H.    LONG. 

FERROUS  IODIDE. 

A  half-normal  solution  was  made  by  mixing  15.87  grams  of 
iodine  with  an  excess  of  iron  and  water,  in  the  usual  manner. 
On  complete  disappearance  of  the  iodine  the  solution  was  boiled 
and  filtered  into  a  cold  sugar  solution.  Water  was  finally  added 
to  make  the  volume  up  to  250  cc.  The  amount  of  sugar  present 
is  not  sufficient  to  prevent  some  decomposition  on  heating,  but, 
as  in  the  other  cases  referred  to,  the  turbidity  at  first  noticed 
disappeared  after  longer  warming  in  the  thermostat.  The  first 
polarizations  were  made  after  filtering,  and  those  later  were 
made  directly. 

Experiment  24. 
Pel,  4. 

In  250  cc,  fifty  grams  of  sugar  +  19.37  grams  of  iodide. 

A  =  32.03. 

f\  2^7  "s  ....  ....  ...• 

15     17.45      6.28'^     0.09478      0.00632 
30     13.57     10.16      O.I657I     0.00552 

45       11.62       1 2. 1 1       0.20627       0.00458 

60  9.73  14.00  0.24956  0.00416 

90  7.50  16.23  0.30690  0.00341 

150  4.40  19.33  0.40176  0.00268 

270  0.90  22.83  0.54177  *  0.00200 

390  —2.80  26.53  0.7^520  0.00196 

In  my  former  paper  a  preliminary"  experiment  with  ferrous 
iodide  was  described  in  which  the  coefficient  appeared  to  be 
nearly  constant  and  much  smaller  than  here.  The  experiments 
are,  however,  not  comparable,  as  in  the  former  case  the  sugar 
solution  was  very  strong,  containing,  in  250  cc,  125  grams  of 
sugar.  In  such  a  solution  the  degree  of  dissociation  of  the 
iodide  would  be  necessarily  very  different  from  that  in  a  weaker 
solution.  In  the  strong  solution  no  separation  of  ferrous  hydrox- 
ide or  other  compound  appears,  even  on  warming.  A  strong  syrup 
is  much  more  stable  than  a  weak  one,  and  the  lower  rate  of 
inversion  may  be  thus  easily  accounted  for. 


INVERSION   OF   SUGAR   BY  SAWS.  713 

CADMIUM    CHtORIDB. 

One  solution  of  cadmium  chloride  was  tested  as  to  its  invert- 
ing power.  It  was  made  with  a  salt  purified  by  several  crj's- 
tallizations  at  a  low  temperature,  free  from  uncombined  acid. 

KxPBRiMBNT  25. 

CdCl,.  o.94iV. 

In  250  cc,  fifty  grams  of  sugar  -j-  42.958  grams  of  chloride. 

^  =  29.71. 


/. 

a. 

X. 

_-«.  log. 

0 

21.41° 

•  •  •  • 

• 

15 

12.80 

8.61° 

0.14862 

0.00990 

30 

6.59 

14.82 

0.30001 

O.OIOOi) 

60 

—1.33 

22.74 

0.62967 

0.01049 

90 

-4.89 

26.30 

0.94015 

0.01044 

150 

—7.70 

29.11 

1.69475 

O.OII29 

The  rate  of  inversion  is  about  as  rapid  as  with  0.Q02N  hydro- 
chloric acid  at  the  same  temperature  and  same  sugar  concentra- 
tion. 

LEAD  NITRATE.  . 

A  single  test  was  made  with  a  solution  containing  lead 
nitrate.  The  salt  was  recrystallized  from  a  pure  Schuchardt 
specimen  and  was  weighed  in  proper  amount  directly. 

Experiment  26. 
Pb(NO.)..  ^- 

In  250  cc,  fifty  grams  of  sugar  +  20.65  grams  of  nitrate. 

A  =  33.70- 

o  25.40°  ....  

15               22.86               2.54^             0.03403  0.00227 

45               17*38               8.02              0.1 1803  0.00262 

75               12.10              i3'3o               0.21800  0.00284 

135                 2.28              23.12               0.50314  0.00372 

195              —3.63              29.03               0,85831  0.00440 

345               —7-70               33-IO                1.74948  0.00507 

The  coefficient  here  is  found  to  increase  very  rapidly,  as  was 

noticed  by  Walker  and  Aston  in  their  experiments/  which  were 

^Loc.  cil. 


714  J-   H.    LONG. 

carried  out  with  a  half  normal  nitrate  solution  at  So^,  but  with  a 
weaker  sugar  solution.  The  mean  value  they  give  from  the 
results  of  polarization  at  three  intervals  is  0.00159,  but  the  inver- 
sion was  not  carried  nearly  to  completeness,  as  in  the  above 
case. 

The  experiments  given  show  in  a  marked  manner  the  extreme 
variations  in  the  value  and  constancy  of  the  inversion  coefficient 
and  the  data  obtained  may  be  roughly  tabulated  as  follows  : 

Potassium  alum  K  constant. 

Ferrous  sulphate  '*  increases  slowly. 

Ammonium  ferrous  sulphate  *' 
Zinc  sulphate 


Cadmium  chloride 

Manganous  sulphate  '*  *'  rapidly. 

Manganous  chloride  **'  *'  •  " 

Lead  nitrate  "  *' 

Ferrous  chloride  "  decreases  rapidly. 

Ferrous  bromide  "  **  ** 

Ferrous  iodide  "  "  ** 

In  the  cases  of  the  last  three  salts  the  values  of  K  decrease 
very  rapidly  at  the  beginning  of  the  heating,  but  become  nearly 
constant  later,  finally,  in  fact,  appearing  to  increase  a  little. 
This  behavior  seems  to  bear  some  relation  to  the  stability  of  the 
salts  in  aqueous  or  weak  saccharine  solution.  As  was  mentioned 
these  ferrous  halogen  solutions  became  turbid  in  the  thermostat, 
and  the  first  three  or  four  portions  withdrawn  in  each  case  for 
polarization  had  to  be  filtered.  Later,  the  liquids  became  per- 
fectly clear  under  the  influence  of  longer  heating. 

During  the  turbid  stage  of  the  reaction,  owing  to  the  tempo- 
rary separation  of  a  trace  of  base  in  insoluble  form,  the  amount 
of  free  acid  present  would  be  relatively  increased,  and  would 
therefore  greatly  accelerate  the  speed  erf  inversion.  With  the 
clearing  of  the  solutions  on  longer  heating  the  normal  hydroly- 
sis only  would  obtain  and  then  the  reaction  should  approach  in 
•regularity  that  due  to  the  presence  of  a  small  constant  amount  of 
mineral  acid. 

It  was  mentioned  that  the  solutions  with  ferrous  sulphate  and 
ferrous  ammonium  sulphate  became  likewise  turbid  on  heating. 
But  here   the  very   slight  opalescence  persisted  through  the 


INVERSION  OF  SUGAR  BY  SALTS.  715 

whole  time  of  heating,  and  was  perhaps  greater  at  the  end  of 
the  reaction  than  at  the  beginning.  Other  experiments  also 
show  in  this  respect  a  marked  difference  between  ferrous  sul- 
phate and  chloride.  In  my  former  paper  I  referred  to  solutions 
of  these  salts  which  had  been  used  qualitatively.  Portions  of 
these  solutions  that  had  not  been  heated  are  still  in  existence. 
After  standing  eight  months  in  the  light  I  find  that  the  chloiide 
is  practically  clear,  while  the  sulphate  has  become  much  changed. 
The  bottle  contains  a  decided  flocculent  precipitate.  My  former 
experiments  with  a  strong  solution  seemed  to  indicate  that  at  a 
temperature  of  loo"*  the  first  slight  precipitate  which  forms  dis- 
appears, but  this  is  not  true  of  the  weaker  solutions  at  85^. 

The  slight  precipitate  of  ferrous  chloride  and  other  halogen 
compounds  being  temporary,  while  that  of,  ferrous  sulphate  is 
apparently  permanent,  we  should  expect  just  such  irregularities 
in  the  speed  of  inversion,  as  the  experiments  actually  show.  A 
solution  of  manganous  sulphate  with  sugar  becomes  also  slightly 
decomposed  on  heating,  and  the  decomposition  increases  with 
the  time  and  temperature.  At  a  temperature  of  100°  a  solution 
of  fifty  grams  of  sugar  and  ten  grams  of  the  sulphate  in  100  cc. 
becomes  so  dark  that  an  exact  polarization  is  not  possible,  even 
after  filtering.  The  solution  in  the  present  case  is  much  less 
concentrated,  but  the  precipitate  is  still  marked  and  its  forma- 
tion is  undoubtedly  attended  by  the  separation  of  a  little  free 
acid.  We  should  therefore  expect  an  acceleration  in  the  rate  of 
inversion  as  before. 

These  considerations  do  not  aid  us  in  explaining,  however, 
the  increase  in  K  for  manganous  chloride,  cadmium  chloride  or 
lead  nitrate.  The  solutions  with  these  salts  are  clear  and  remain 
so  throughout  the  reaction.  In  the  case  of  manganous  chloride 
it  must  be  remembered  that  an  almost  complete  loss  of  color  fol- 
lows after  heating.  The  pink  fades,  and  in  a  few  hours  at  the 
temperature  of  the  thermostat  becomes  imperceptible  in  a  small 
volume  of  the  liquid.  The  color  is  not  restored  by  cooling.  We 
have  here  evidently  a  reaction  in  which  a  change  takes  place  in 
the  form  of  combination  of  the  manganese,  with  a  necessary 
alteration  in  the  degree  of  dissociation  of  the  salt. 


7l6  INVERSION  OF  SUGAR   BY   SALTS. 

It  is  true,  as  already  said,  that  most  of  the  bases  under  con- 
sideration form  compounds  with  the  sugars,  so  that  we  should 
expect  from  this  cause  a  slight  disturbance  at  least  in  the  appar- 
ent rate  of  inversion.  Too  little  is  known  of  the  optical  proper- 
ties of  these  saccharose,  dextrose  and  levulose  metallic  compounds 
to  say  just  what  effect  they  would  have  on  the  rotation,  but  that 
they  have  some  action  is  suggested  by  the  results  of  some  of  the 
polarizations  to  determine  the  end  point  in  the  inversion.  This 
was  usually  found  a  little  below  the  theoretical,  — 8.6°  for  a  200 
mm.  tube,  but  in  several  cases  it  was  found  above  after  pro- 
longed heating.  This  was  also  true  of  a  solution  of  sugar  with 
manganous  chloride,  which  stood  exposed  to  the  light  several 
months. 

It  must  be  remembered  also  that  solutions  of  dextrose  are 
easily  oxidized,  and  those  of  levulose  much  more  so.  The  dark 
color  often  seen  near  the  end  of  the  reaction,  points  to  such  a 
decomposition. 

It  will  be  recognized  that  a  determination  of  the  hydrol3'sis  of 
many  of  the  heav}*^  metallic  salts  cannot  be  measured  with  great 
accuracy,  because  of  these  several  disturbing  influences,  but  a 
comparison  of  some  little  value  in  the  above  cases  may  be  made 
by  considering  the  results  obtained  at  the  beginning  of  the  reac- 
tions in  which  the  coefficient  is  an  increasing  one,  and  near  the 
end  of  the  reaction  in  cases  where  it  decreased  and  then  became 
nearly  constant.  By  taking  the  mean  of  the  first  two  values  in 
the  one  case,  and  of  the  last  two  in  the  other,  we  obtain  the 
second  column  of  the  table  below  as  the  most  probable  values  of 
the  coefficient  for  half- normal  solutions. 

In  the  third  column  is  given  a  calculation  of  the  extent  of 
hydrolysis  of  the  salts,  expressed  in  per  cents,  of  total  salt  pres- 
ent, and  based  on  a  comparison  with  hydrochloric  acid  acting  in 
o.ooi  normal  solution  at  the  same  temperature  on  same  amount 
of  sugar.  This  comparison  is  at  best  a  rough  one,  assuming  as 
it  does  complete  hydrolysis  of  the  acid,  and  weglecting  the  effect 
of  the  excess  of  undecomposed  salts  on  the  rate  of  inversion. 


DBTERMINATION   OF   IRON   AND   ALUMINA.  717 

Salt  bydrolyzed  in 
K.  p«r  cent. 

Lead  nitrate 0.00244  0.096 

Manganous  chloride     •.  0.00095  0.035 

Manganous  sulphate 0.00052  0.020 

Ferrous  sulphate 0.00085  0.033 

Ferrous  ammonium  sulphate 0.00068  0.026 

Zinc  sulphate 0.00040  0.016 

Ferrous  chloride 0.00164  0.063 

Ferrous  bromide  (0.54^) 0.00300  o.  109 

Ferrous  iodide •0.00198  0.078 

Potassium  aluminum  sulphate,  —  0.01835  i<440 

Cadmium  chloride  0.94 A^ o.oiooo  2.080 

The  amount  of  hydrolysis  is  small  in  all  cases  except  those  of 
the  alum  and  cadmium  chloride. 

My  thanks  are  due  to  Mr.  S.  R.  Macy  for  much  assistance  in 
the  experimental  work  of  the  above. 

K0RTHWB8TRRX  University, 
Chicago. 


DBTERMINATION    OF    IRON    OXIDE    AND    ALUHINA    IN 
PHOSPHATE  ROCK  BY  THE  AMMONIUM  ACETATE 

METHOD. 


By  Thomas  S.  Gladding. 

Received  Jua»  30,  1896. 


THE  oldest  method  of  separating  alumina  and  iron  phos- 
phates from  lime  phosphate  is,  probably,  the  ammonium 
acetate  method.  This  has  been  severely  criticised,  and  just  at 
present  seems  to  be  under  condemnation.  The  following  inves- 
tigation has  convinced  the  writer  that,  when  properly  carried 
out,  not  only  does  the  method  give  an  accurate  separation  of 
iron  and  alumina  from  lime  phosphate,  but  also  gives  a  neu- 
tral phosphate  of  uniform  composition  from  which  the  iron  oxide 
and  alumina  present  may  be  accurately  estimated. 
'  In  brief,  the  method  used  is  this.  If  a  weakly  acid  solution 
of  phosphates  of  iron  and  alumina  together  with  a  large  amount 
of  calcium  phosphate  be  slowly  poured  into  a  strong  solution  of 
ammonium  acetate  made  acid  with  acetic  acid,  the  iron  and 
alumina  are  precipitated  as  phosphates,  upon  digestion  for  a 
short  time  at  a  gentle  heat.     This  precipitate,   however,  con- 


7X8  THOMAS  S.  GLADDING.      IRON  AND  ALUMINA 

tains  more  or  less  calcium  phosphate,  which  is  removed  by  sev- 
eral reprecipitations.     I  shall  demonstrate  by  experiment : 

First,  That  upon  continued  reprecipitations  of  iron  and 
alumina  as  phosphates  in  this  manner,  there  is  no  appreciable 
diminution  of  the  quantity  of  either  finally  obtained,  provided 
there  always  be  a  large  excess  of  phosphoric  acid  present. 

A  standard  solution  was  made  by  dissolving  twenty  grams  of 
ammonia  alum  (C.  P.)  in  distilled  water.  This  was  slightly 
acidified  with  hydrochloric  acid,  in  order  to  prevent  the  alumina 
from  separating  on  standing,  and  diluted  to  one  liter.  This 
solution,  upon  being  standardized,  was  found  to  contain  the 
theoretical  amount  of  alumina,  that  is, 

Ten  cc.  r=  0.0225  grams  A1,0,. 

One  precipitation,  in  the  manner  described  above,  of  the 
alumina  in  ten  cc.  gave 

AlsOs.PaOft,  found.  Al,Ot. 

1  0.0545  0.0298 

2  0.0549  0.0229 

3  0.0546  0.0228 

4  0.0540  0.0226 

5  0.0545  0.0228 

Three  successive  precipitations,  in  which  one  gram  of  ammo- 
nium phosphate  was  added  before  each  precipitation,  gave 

AlaOs.P,0».  Al,Os. 

1  0.0550  0.0230 

2  0.0547  0.0229 

3  0.0544  0.0227 

Five  successive  precipitations  were  also  tried  under  the  same 
conditions,  with  the  following  results  : 

AlfOi-P^Oft.  Al,Oa. 

1  0.0536  0.0224 

2  0.0530  0.0222 

When,  however,  the  excess  of  phosphoric  acid  was  omitted 
before  the  reprecipitations,  there  was  a  loss  of  alumina. 

An  iron  solution  was  made  by  dissolving  C.  P.  iron  wire  in 
hydrochloric  acid  and  oxidizing  it  with  nitric  acid.  When  care- 
fully standardized  it  was  found  that 

Ten  cc.  =  0.0296  Fe,0,. 


BY  THE  AMMONIUM  ACETATE  METHOD.        719 

Three  successive  precipitations,  adding  one  gram  ammonium 
phosphate  before  each,  gave 


Pe,0a.Pi05. 

Fc.Oj. 

I 

0.0545 

0.0289 

2 

0,0550 

0.0291 

3 

0.0548 

0.0290 

Five  successive  precipitations,  in  the  same  way,  gave 

PesOs.PtO».  PetOg. 

1  0.0550  0.0291 

2  0.0560  0.0297 

Second,  That  upon  three  successive  precipitations  in  the  pres- 
ence of  a  large  amount  of  calcium  phosphate,  as  is  the  case  in 
the  analysis  of  rock  phosphate,  the  precipitate  of  the  phos- 
phates of  iron  and  alumina  is  sufficiently  pure  to  be  taken  as 
such.  Of  the  standard  solutions,  five  cc.  of  each  would  together 
give  a  precipitate  of  combined  phosphates  about  equal  to  that 
usually  found  in  one  gram  of  phosphate  rock.  The  mixture  so 
analyzed  was  made  up  as  follows : 

Five  cc.  alumina  solution  =  0.01125  A1,0,. 
Five  cc.  iron  solution  =  0.01480  Fe^O,. 
0.7000  grams  calcium  phosphate. 
This  was  given  three  precipitations,  the  excess  of  phosphoric 
acid  being  supplied  before  the  second  and  third  precipitations. 

Phosphates  obtained.  Al^Oi  obtained.  P^sOa  obtained. 

1  0.0552  0.OII5  0.0146 

2  0.0540  O.OIIO  0.0146 

3  0.0537  0.0109  0.0146 

4  0.0536  0.0109  0.0146 

The  iron  oxide  was  determined  by  volumetric  method  in  the 
ignited  precipitate  and  the  alumina  by  subsequent  calculation. 

In  addition  twenty  cc.  alumina  solution  containing  0.0450 
grams  A1,0„  together  with  0.700  grams  calcium  phosphate,  were 
given  three  successive  precipitations  in  the  same  way  with  the 
following  results : 

AlsOs.Pips  obtained.  AlaO.  obtained. 

Grams.  Grams. 

1  0.1092  0.0456 

2  0.1074  0.0449 

In  order  to  prove  that  the  aluminum  phosphate  precipitated 
was  the  normal  phosphate,  the  ignited  precipitates  were  fused, 
and  the  phosphoric  acid  in  them  estimated. 


720  THOMAS  S.    GI.ADDING. 

A.I,Oa.PtO».  P,0»obUined.        AlfOs  by  diff.         Al,Oa  by  calc. 

1  0,0538  0.0313  0.022$  0.0225 

2  0.0533  0.0312  0.0221  0.0223 

The  phosphate  of  alumina  is  multiplied  by  the  factor  0.418  to 
obtain  the  alumina. 

Therefore,  in  determining  iron  oxide  and  alumina  in  phos- 
phate rocks  proceed  as  follows : 

Pour  grams  of  the  finely  ground  sample,  previously  freed 
by  a  magnet  from  any  metallic  iron  derived  from  the  iron  mortar 
used  in  grinding  the  sample,  are  digested  for  half  an  hour,  at  a 
temperature  just  below  the  boiling-point,  with  about  thirty  cc. 
dilute  hydrochloric  acid  (i-i).  This  will  prevent  the  solution 
of  any  pyrites  if  present.  Filter  and  wash  thoroughly  into  a  200 
cc.  flask,  add  a  little  nitric  acid,  and  boil  to  oxidize  the  iron, 
cool,  and  fill  to  mark  with  water.  Take  two  portions,  fifty  cc. 
=  one  gram,  twenty-five  cc.  =  one-half  gram,  and  proceed  with 
each  as  follows : 

Almost  neutralize  the  solutions  with  strong  ammonium 
hydroxide  until  the  precipitate  formed  dissolves  with  difficulty, 
and  thoroughly  cool  by  placing  the  beaker  in  a  dish  of  cold 
water.  The  neutralization  is  then  completed  by  carefully  adding 
dilute  ammonium  hydroxide  until  the  precipitate  remains  per- 
manent, then  just  dissolve  by  adding  dilute  hydrochloric  acid, 
drop  by  drop,  stirring  well.  Have  ready  in  another  beaker  a 
mixture  of  fifteen' cc.  of  a  strong  solution  of  ammonium  acetate 
(made  by  neutralizing  thirty  per  cent,  acetic  acid  with  strong 
ammonium  hydroxide)  and  five  cc.  of  acetic  acid.  Carefully 
pour  the  cold  faintly  acid  solution  of  phosphates  in  a  fine  stream 
into  this  mixture,  stirring  all  the  while.  Digest  at  60"*  C.  from 
one-half  hour  to  one  hour,  until  the  supernatant  liquid  is  clear 
and  the  flocculent  precipitate  is  well  settled  to  the  bottom. 

Filter  and  wash  the  precipitate  once  with  a  ten  per  cent, 
ammonium  acetate  solution,  merely  rinsing  out  the  beaker  in 
which  the  precipitation  was  made.  Dissolve  the  precipitate 
from  the  paper  into  the  same  beaker  with  a  few  cubic  centime- 
ters of  hot  dilute  hydrochloric  acid  (1-4).  Use  as  little  acid  as 
possible  in  order  to  keep  the  bulk  of  the  solution  small.  Add 
one  gram  of  ammonium  phosphate,  neutralize  with  ammonium 


IRON  AND   ALUMINA.  72 1 

hydroxide  and  add  hydrochloric  acid  until  the  precipitate  just 
dissolves  as  before  and  pour  into  a  mixture  of  fifteen  cc.  ammo- 
nium acetate  solution  and  five  cc.  acetic  acid.  Digest  at  do"*  C. 
for  one-half  to  one  hour  and  filter,  and  wash  once  with  the  ten 
per  cenf .  ammonium  acetate  solution.  Redissolve  and  repeat  the 
precipitation,  being  careful  to  again  add  one  gram  of  ammonium 
phosphate  to  the  solution,  in  order  that  there  be  a  sufficient 
excess  of  phosphorus  pentoxide  to  precipitate  all  the  alumina  as 
a  neutral  phosphate.  Wash  the  precipitate  three  times  with 
dilute  ammonium  acetate  solution. 

Take  the  filter,  while  wet,  from  the  funnel  and  ignite  in  a 
tared  platinum  capsule,  using  a  very  low  flame  until  the  filter 
paper  is  thoroughly  charred.  The  heat  is  increased  gradually 
until  the  paper  is  completely  consumed ,  and  finally  the  blast  lamp  is 
used  for  a  minute.  Weigh  as  combined  phosphates  of  iron  and 
alumina.  The  iron  is  determined  volumetrically  in  the  solution 
of  the  weighed  precipitates.  The  iron  oxide  present  in  the  rock 
is  also  determined  separately  by  volumetric  process,  preferably 
the  bichromate  method,  in  a  solution  of  five  grams  of  the  rock  in 
dilute  hydrochloric  acid  (i-i),  reducing  all  iron  to  protoxide 
and  titrating  with  bichromate. 

The  ignited  precipitate  from  one  of  the  duplicate  precipitations 
may,  if  desired,  be  dissolved  and  subjected  to  a  fourth  precipita- 
tion and  the  filtrate  tested  for  lime  by  adding  ammonium  oxalate 
and  heating.  My  thanks  are  due  to  our  assistant,  Thomas  Brown, 
Jr.,  for  valuable  aid  in  the  above  analytical  determinations. 

I^BORATORY  OF  STILLWBLL  <^  GLADDING, 

New  York  City. 


A  NEW  METHOD  FOR  THE  ESTIHATION  OF  IRON  OXIDE 
AND  ALUMINA  IN  PHOSPHATE  ROCK. 

By  Thomas  S.  Gx^dding. 

Received  June  90.  1896. 

THE  method  for  the  separation  of  alumina  from  phosphate 
of  lime  by  three  successive  precipitations  with  ammonium 
acetate  is  tedious,  though  accurate  if  proper  precautions  be  taken, 
as  shown  in  the  preceding  paper  on  this  subject. 
The  following  modification  suggested   itself  as  saving  both 


722  THOMAS  S.  GLADDING.      A  NEW  MBTHOD 

time  and  labor.  This  modification  consists  of  the  separation  of 
alumina  from  calcium  phosphate  and  iron  by  means  of  its  solu- 
bility in  an  excess  of  caustic  potash.  To  demonstrate  the  accu- 
racy of  this  method,  a  solution  of  ammonia  alum,  twenty  grams 
in  a  liter,  was  used  as  in  the  previous  experiments,  ten  cc.  con- 
taining 0.0225  grams  A1,0,.  The  caustic  potash  solution  was 
made  by  dissolving  500  grams  of  caustic  potash  in  distilled  water 
and  diluting  to  one  liter.  Chemically  pure  caustic  potash,  puri- 
fied by  barium,  was  used  and  was  carefully  tested  for  alumina, 
as  much  so-called  chemically  pure  potash  contains  an  apprecia- 
ble amount  of  alumina. 

To  a  solution  of  mixed  phosphates  of  alumina,  iron,  and  lime 
were  added  fifteen  cc.  of  the  C.  P.  potash  solution.  The  mixture 
was  digested  for  an  hour  at  a  temperature  of  70^  C,  with  occa- 
sional stirring.  It  was  then  filtered,  the  filtrate  neutralized 
with  hydrochloric  acid,  and  the  alumina  was  precipitated  as 
a  phosphate  with  ammonium  acetate  as  described  in  my  ammo- 
nium acetate  method. 

Ten  cc.  standard  alumina  solution  -|-  0.030  gram  iron  oxide  + 
0.500  gram  calcium  phosphate  gave 


AltOa.PfOg  found. 

AltO,. 

Grams. 

Grams. 

I 

0.0538 

0.0225 

2 

0.0542 

0.0227 

3 

0.0543 

0.0227 

4 

0.0543 

0.0227 

Comparative  tests  were  made  on  phosphate  rocks  between  this 
method  by  solution  in  C.  P.  potash  and  by  three  successive  pre- 
cipitations with  ammonium  acetate. 

By  new  potash  method.  By  acetate  method. 

AlfOi  found.  A1«0|  found. 

Per  cent.  Per  cent. 

1  1.05  1.03 

2  1. 19  Z.16 

3  1.86  I.61 

4  1.07  0.99 

5  1.88  1.98 

These  results  show  the  accuracy  of  this  method,  both  in 
obtaining  a  known  amount  of  alumina  and  in  showing  close 
agreement  with  results  by  the  acetate  method. 


FOR   IRON  AND   ALUMINA.  723 

This  method  has  been  in  use  in  our  laboratory  for  over  a  year. 
A  reprint  of  an  article  by  M.  Henri  Lasne'  has  just  been 
received,  giving  a  method  for  the  separation  of  alumina  from 
phosphates  of  iron  and  lime  very  similar  to  this.  M.  Lasne  uses 
caustic  soda  instead  of  potash  and  precipitates  his  aluminum 
phosphate  with  ammonium  hyposulphite  instead  of  ammonium 
acetate.  I  have  made  a  few  comparative  tests  by  my  method 
and  that  of  M.  Lasne  and  find  closely  agreeing  results. 

Using  ten  cc.  standard  alumina  solution  ^-  0.500  grams  cal- 
cium phosphate  I  found 

By  my  method.  By  Lasne's  method. 

Al«0|.PtO(.  AlfOg.  Al,Oa.PtO«.  Al«Oa. 

Grams.        Grams.        Grams.        Grams. 

I       0.0542      o.o2ao      0.0540      0.0226 

2         0.0538         0.0225         0.0533         0.0223 

In  the  analysis  of  a  phosphate  rock  I  found 

By  mr  method.  By  Lasne's  method. 

Al«Ot  found.  Al|Ot  found. 

Per  cent.  Per  cent. 

1-75  1-73 

1.80 

The  detailed  method  used  in  my  work  is  as  follows :  Treat 
the  finely  ground  rock  phosphate  with  a  magnet  to  remove  any 
metallic  iron  derived  from  the  iron  mortar  used  in  the  prepara- 
tion of  the  sample.  Dissolve  four  grams  of  the  rock  in  thirty  cc. 
dilate  hydrochloric  acid  (i-i),  heating  just  below  the  boiling- 
point  for  half  an  hour.  This  prevents  the  solution  of  pyrites. 
Filter  into  a  200  cc.  flask,  add  a  few  drops  of  nitric  acid,  and 
boil  to  oxidize  the  iron,  cool  and  dilute  to  mark.  Take  fifty  cc. 
containing  one  gram  of  rock  and  run  into  twenty  cc.  of  the  solu- 
tion of  C.  P.  caustic  potash.  Digest  for  an  hour  at  70°  C,  stir- 
ring occasionally.  Let  the  precipitate  settle  and  filter  on  a  large 
paper,  first  decanting  the  supernatant  liquid  on  the  paper  and 
finally  washing  on  the  precipitate.  Wash  two  or  three  times 
with  hot  water. 

To  the  filtrate  add  one  gram  of  ammonium  phosphate,  acidify 
with  hydrochloric  acid,  add  ammonia  until  a  permanent  precipi- 
tate is  formed  and  dilute  hydrochloric  acid,  drop  by  drop,  until 
it  is  just  dissolved.     Add  a  mixture  of  fifteen  cc.  neutral  ammo- 

1  Prom  the  BuUeHn  de  la  SocUU  ckimique  de  Paris,  [3]  15,  xi8. 1896. 


724  CLARENCE   L.   SPBYERS. 

nium  acetate  solution  and  five  cc.  acetic  acid  (thirty  per  cent.) 
and  digest  for  half  an  hour  at  70**  C,  by  which  time  the  precipi- 
tation is  complete. 

Filter,  washing  five  or  six  times  with  hot  ammonium  acetate 
solution  (ten  per  cent.)i  stirring  up  the  precipitate  with  the  jet 
each  time.  Ignite  with  a  low  flame  until  the  paper  is  charred, 
increase  the  heat,  and,  when  the  paper  is  completely  consumed, 
blast  for  a  minute.  The  precipitate  is  the  normal  aluminum 
phosphate  and  its  weight  multiplied  by  the  factor  0.418  gives 
the  A1,0,. 

The  iron  oxide  is  determined  volumetrically,  preferably  by 
the  bichromate  method,  in  a  solution  of  the  precipitate  of  iron 
oxide  and  calcium  phosphate  thrown  down  by  the  caustic  potash. 
It  is  also  determined  separately,  by  the  same  method,  in  a  solu- 
tion of  five  grams  of  the  rock  in  dilute  hydrochloric  acid  ( i-i ) . 

My  thanks  are  due  to  Mr.  H.  E.  Cutts,  A.M.,  for  valuable 
assistance  in  the  above  investigation. 

Laboratory  of  Stillwell  &  Gladding. 
New  York  City. 


SOME  THOUGHTS  ABOUT  LIQUIDS. 

fiY  Clarence  L.  Speyerb. 

Received  Tune  3.  1896. 

CONSIDER  an  empty  closed  space.  Imagine  a  quantity  of 
liquid  put  into  it,  enough  to  fill  the  space  with  vapor  and 
leave  some  liquid  over.  A  portion  of  the  liquid  changes  into 
vapor  and  passes  into  the  previously  empty  space  above  the 
liquid  and  continues  doin'g  so  until  the  pressure  of  the  vapor 
reaches  a  certain  value,  when  the  vaporization  ceases. 

The  usual  way  of  explaining  this  vaporization  starts  out  by 
assuming  that,  with  the  exception  of  the  surface,  the  liquid  is 
perfectly  homogeneous  in  a  physical  sense.  That  is,  there  is  not 
a  single  particle  of  the  liquid  which  for  any  appreciable  length 
of  time  is  different  from  any  other  particle,  but  of  course, 
spaces  between  the  particles  of  liquid  are  recognized.  At  the 
surface  of  the  liquid,  however,  a  distinction  is  to  be  made.  For 
outside  the  surface,  the  activities  are  different  from  those  within 
the  surface,  otherwise  there  would  be  no  boundary.     So  that  the 


SOMB  THOUGHTS  ABOUT  UQUIDS.  725 

particles  at  the  surface  are  subjected  to  activities  that  are  differ- 
ent in  different  directions,  and  consequently  the  particles  ft 
situated  will  behave  differently  from  those  particles  entirely 
within  the  liquid. 

In  van  der  WaaPs  theory  the  mutual  attractions  of  the  parti- 
cles of  the  liquid  are  considered  as  the  restraining  force  to  keep 
the  particles  more  or  less  together.  This  assumed  force  must  be 
very  great — a  good  many  hundred  atmospheres.  Inside  the 
liquid,  below  the  surface,  the  attraction  is  equal  in  all  direc- 
tions, but  at  the  surface  it  acts  only  in  one  direction,  inwards, 
normal  to  the  surface.  Now,  although  the  force  restraining  the 
particles  of  liquid  from  separating  is  so  great,  yet  the  theory  of 
common  acceptance  assumes  that  some  particles  do  break  away 
from  the  mass  of  the  liquid  and  form  vapor.  The  liquid  is  said 
to  evaporate.  It  is  hard  to  accept  this  view  of  the  case,  particu- 
larly as  electrical  results  point  towards  an  exceedingly  quiet 
condition  of  affairs  within  the  body  of  liquids. 

Still  admitting  that  the  particle  does  break  away  from  this 
attraction,  it  cannot  do  so  without  an  abundant  supply  of  energy, 
which  must  be  accounted  for.  It  does  not  seem  right  to  find 
this  energy  in  the  heat  of  vaporization,  for  a  particle  of  liquid 
will  voluntarily  take  heat  energy  from  the  liquid  to  do  this  work, 
and  so  go  off  as  a  particle  of  vapor  at  the  sacrifice  of  the  energy 
of  the  liquid. 

It  is  not  possible  to  prevent  a  liquid  from  vaporizing  by  refus- 
ing to  give  it  heat ;  it  will  take  the  required  heat  from  the  rest 
of  the  liquid.  In  other  words,  the  condition  of  the  liquid  state 
strongly  favors  vaporization. 

The  common  theory  tries  to  get  over  this  difl&culty  by  claim- 
ing that  the  particle  which  gets  away,  gets  away  by  virtue  of  an 
inherent  kinetic  energy  greater  than  the  attractic  energy  of  the 
particies  of  liquid,  that  is,  greater  than  the  force  denoted  by  van 

n  "^ 

der  Waal  by  A'=  ~y,  and  that  it  possesses  this  excess  of  kinetic 

energy  in  the  body  of  the  liquid,  before  it  got  away,  and  that  it 
got  away  only  by  virtue  of  this  excess  of  kinetic  energy.  Simi- 
larly with  all  particles  in  the  liquid  having  a  kinetic  energy 


should  exl>«<:'   ,  ^„.  Uq»*  ■  \^iat««"  \«it.     ""     „e«>i  ° 
^y  and  so  '« 


SOME   THOUGHTS  ABOUT   LIQUIDS.  727 

due  to  the  escape  of  the  particles  with  great  kinetic  energy  jrom 
the  liquid.  But  all  of  the  lost  kinetic  energy  cannot  be  absorbed 
here  in  the  liquid,  some  must  also  go  into  the  vapor  particles. 
It  may  take  the  form  of  heat,  as  we  have  suggested  in  the  case 
of  the  liquid,  but  then  we  have  to  assume  that  the  kinetic  energy, 
while  within  the  liquid,  of  the  particles  that  escape  is  of  just 
such  a  value  that,  after  they  have  all  got  out  of  the  liquid,  the 
diminution  in  their  mean  kinetic  energy,  due  to  the  attraction  of 
the  liquid  plus  this  correction,  brings  the  kinetic  energy  left  to 
them  to  the  mean  kinetic  energy  of  the  liquid,  which  is  absurd. 

Nor  does  the  attractive  energy  seem  to  be  stored  up  as  poten- 
tial energy,  as  in  the  case  of  a  stone  raised  above  the  surface  of 
the  earth,  for  there  is  no  evidence  at  all  that  a  vapor  particle 
tends  toward  the  liquid  as  the  stone  does  toward  the  earth. 
When  the  partkle  gets  out  of  the  liquid  it  seems  to  be  utterly 
indifferent  to  the  liquid. 

Of  course  the  mutual  attraction  that  all  bodies  have  for  each 
other  is  left  out  of  account. 

Nor  is  there  any  sign  of  electrical  action,  at  least  if  the  ex- 
periments made  up  to  the  present  time  are  conclusive. 

There  are  then  a  good  many  very  serious  objections  to  the 
present  theory  of  vaporization. 

First,  in  accounting  for  the  escape  of  the  vapor. 

Second,  in  accounting  for  the  temperature  of  the  vapor. 

Third,  in  accounting  for  the  kinetic  energy  lost  by  the  particle 
in  getting  through  the  surface  of  the  liquid  and  beyond  the 
sphere  of  action  of  the  liquid  particles. 

Let  us  now  turn  our  attention  to  another  view  of  the  case. 
Consider  a  liquid  which  has  no  vapor-tension  of  its  own,  a  non- 
volatile liquid,  but  which  can  dissolve  gases.  The  liquid  and 
gas  are  supposed  to  act  according  to  Henry's  law,'  that  is,  the 
ratio  of  concentration  of  the  gas  in  the  liquid  part  and  in  the 
gaseous  part  is  to  be  constant,  or  in  other  words,  the  quantity 
of  gas  dissolved  by  the  liquid  is  to  be  proportional  to  the  pres- 
sure on  the  gas. 

In  such  a  system  there  are  three  constituents  to  be  considered. 
The  gas  in  the  gaseous  state,  the  gas  in  solution,  and  the  sol- 
vent. 


72S  CLARENCE   L.    SPEYEES. 

The  state  of  the  dissolved  gas  is  not  positively  known,  but  in 
all  probability  it  is  in  a  state  corresponding  to  a  gas  under  high 
pressure  for  these  reasons.  In  the  first  place,  it  is  hard  to  see 
how  a  substance  like  nitrogen,  for  example,  could  be  in  the 
liquid  state  in  a  solution  of  moderate  concentration.  Great 
pressure  is  required  to  liquefy  it  even  when  the  temperature  is 
far  below  the  ordinary  temperature,  and  at  the  ordinary  tem- 
perature it  has  hitherto  been  found  impossible  to  liquefy  nitro- 
gen, no  matter  how  great  the  pressure.  Still  it  would  be  con- 
sistent with  the  ordinarily  accepted  theory  to  claim  that  the 
attraction  of  the  particles  of  solvent  could  overcome  the  great 
internal  energy  of  the  gas  particles  and  bind  them  down  to  a 
lesser  activity  and  produce  the  liquid  state.  But  on  the  other 
hand,  modern  investigation  has  very  plainly  shown  that  dis- 
solved substances  have  a  gaseous  nature  ;  the  particles  of  the 
dissolved  body  are  free  to  assert  their  physical  individuality. 
That  is  to  say,  the  solvent  is  to  be  considered  rather  as  a  medium 
through  which  the  dissolved  body  can  be  put  under  certain  con- 
ditions, the  conditions  varying  to  some  extentwith  each  solvent, 
but  all  solvents  having  the  common  action  of  aUowing  a  sort  of 
gasification  of  the  substance  dissolved,  In  general  the  solvent 
is  not  to  be  considered  as  a  substance  which  unites  with  the  dis- 
solved body,  forming  a  new  compound.  For  example,  consider 
anhydrous  calcium  chloride.  When  this  is  treated  with  water 
there  is  strong  evidence  of  combination  of  the  two  to  form  cal- 
cium chloride  hydrate.  If  the  quantity  of  water  is  properly 
adjusted  the  whole  of  it  combines  with  the  calcium  chloride, 
forming  a  crystallized  hydrate.  If  this  crystalline  substance  is 
treated  with  more  water,  solution  begins  and  during  this  process, 
which  is  the  real  solution,  there  is  no  sign  of  chemical  action. 
It  is  true,  some  scientists,  particularly  those  of  the  English 
school,  have  denied  this  and  have  claimed  to  find  strong  evi- 
dence of  a  chemical  action  during  the  process  of  solution,  but  so 
far  all  such  claims  have  turned  out  to  be  mere  opinions  based 
upon  very  doubtful  measurements. 

So  we  are  to  look  upon  solution  as  being  a  change  in  which 
the  dissolved  body  is  gasified.  Sometimes  a  further  change, 
electrolytic  dissociation,   takes  place,  but  that  is  outside  the 


SOMB  THOUGHTS  ABOUT  LIQUIDS.  729 

scope  of  this  article.  It  is  in  best  accordance  with  what  we  know 
about  other  bodies  to  assume  that  the  dissolved  nitrogen  is  in  the 
form  of  a  gas,  and  to  recognize  two  states  in  the  solution,  the 
gaseous  state  of  the  substance  in  solution  and  the  liquid  stat^  of 
the  solvent. 

Let  us  now  pass  on  to  a  liquid  which  gives  off  vapor.  The 
purpose  of  this  article  is  to  justify  the  view  that  this  vapor 
behaves  toward  the  liquid  just  as  the  nitrogen  did  toward  its 
solvent  in  the  previous  case,  of  course,  with  the  obvious  limita- 
tions due  to  identity  in  the  composition  of  vapor  and  liquid. 

The  boundary  dividing  vapor  from  liquid  is  commonly  sup- 
posed to  be  at  the  surface  of  the  liquid,  although  the  possibility 
of  a  differentiation  occurring  inside  the  liquid  does  not  seem  to 
be  denied,  for  so  far  as  could  be  found  out  by  the  writer,  the 
question  of  such  a  possibility  has  never  been  raised. 

The  tendency  for  a  liquid  to  vaporize  and  the  pressure  of  its 
saturated  vapor  is  evidently  a  function  of  temperature  only. 
There  seems  to  be  no  reason,  therefore,  why  the  fluid  should 
not  separate  into  vapor  and  liquid  within  the  surface  of  the 
liquid.  That  it  is  possible  for  vapor  to  be  there  follows  from 
what  we  know  about  the  gaseous  nature  of  the  substances  in 
solution.  It  is  rather  odd  that  this  view  of  the  case  was  not 
adopted  at  the  outset  by  chemists. 

According  to  this  view,  when  we  heat  a  liquid  we  increase  the 
energy  of  translatory  motion,  we  increase  its  temperature.  But 
besides  this  we  cause  a  separation  of  some  of  the  liquid  particles 
from  the  body  of  the  liquid,  bringing  them  into  a  state  of  free- 
dom, such  that  they  can  behave  just  as  the  particles  of  any 
other  substance  would  do  in  the  same  solvent.  This  of  course 
will  consume  considerable  energy.  These  free  particles  of  vapor 
in  the  liqiiid  we  shall  call  dissolved  vapor  particles.  So  that  on 
heating  in  liquid  we  produce  dissolved  vapor  and  raise  the  tem- 
perature of  the  whole  fluid ;  possibly  we  do  more,  but  at  any 
rate  we  do  these  two  things.  Now  by  Clausius'  theory  of  the 
true  specific  heat,  the  heat  required  to  raise  only  the  tempera- 
ture of  a  unit  mass  of  substance  one  degree,  should  be  the  same 
whatever  the  state  of  the  substance  may  be,  and  the  value  of  the 
true  specific  heat  should  be  the  value  of  the  specific  heat  when 


730  CLARENCE   L.    SPEYERS. 

the  substance  is  in  such  a  state  that  the  heat  added  can  only- 
change  its  temperature  and  not  do  any  other  internal  work, 
namely  when  the  substance  is  in  a  state  of  gas.  So  if  we  sub- 
tract from  the  specific  heat  of  the  liquid  the  specific  heat  of  the 
gas,  the  remainder  should  be  the  heat  consumed  in  other  inter- 
nal work,  and  if  no  other  internal  work  is  done  than  the  rise 
in  temperature  and  production  of  dissolved  vapor,  we  should  get 
the  heat  required  to  change  some  of  the  liquid  into  dissolved 
vapor.  The  quantity  changed  into  vapor  however  is  so  far 
unknown. 

The  dissolved  vapor  is  supposed  to  be  produced  until  its 
energy  balances  the  energy  of  the  liquid  part. 

Suppose,  for  example,  we  heat  one  gram  of  water  one  degree 
in  a  closed  vessel  which  does  not  allow  it  to  give  off  gaseous 
vapor.  The  heat  required  will  be  about  one  calorie,  depending 
upon  the  initial  temperature;  one  calorie  is  near  enough  for  our 
purpose.  A  part  of  the  heat  goes  to  increase  the  translatory 
motion  and  is  the  true  specific  heat ;  but  another  part,  perhaps 
the  whole  of  the  remainder,  we  claim  goes  to  produce  dissolved 
vapor.  Subtracting  the  true  specific  heat  of  water,  namely  the 
specific  heat  of  water  vapor  at  a  high  temperature  =  0.4776,  we 
have  left  0.5224  as  the  heat  required  to  change  a  certain  un- 
known quantity  of  water  into  dissolved  water  vapor,  provided 
that  no  internal  work  is  done  but  the  two  kinds  we  have  con- 
sidered. We  shall  assume  this  to  be  true  until  there  is  evi- 
dence of  a  more  complex  change. 

Now  suppose  a  space  be  made  over  the  liquid,  to  let  a  certain 
quantity,  say  one  per  cent.,  be  changed  into  gaseous  vapor.  It 
is  of  course  evident,  if  the  theory  be  at  all  tenable,  that  the  vapor 
arising  from  the  liquid  comes  from  the  dissolved  vapor  and  bears 
to  the  dissolved  vapor  the  same  relation  that  the  nitrogen  did  to 
the  dissolved  nitrogen.  Comparatively  little  heat  should  be 
required  in  this  process,  for  most  of  the  change  has  been  effected 
in  the  body  of  the  liquid.  Whatever  is  required  here  should  be 
looked  upon  as  the  true  heat  of  vaporization ;  that  which  is 
usually  so  called  we  are  to  consider  as  including  the  beat 
required  to  change  a  unit  mass  of  liquid  into  dissolved  vapor  as 
well  as  the  heat  required  to  vaporize  the  unit  mass  of  dissolved 


SOME  THOUGHTS  ABOUT  LIQUIDS.  73 1 

vapor.  The  two  quantities  should  evidently  be  kept  carefully 
separated. 

Let  us  now  proceed  to  determine  the  concentration  of  the  dis- 
solved water  vapor.  As  the  dissolved  water  vapor  is  supposed 
to  be  like  a  dissolved  gas,  Henry's  law  should  give  us  some  aid 
in  finding  the  quantity.  We  might  assume,  in  the  first  place, 
that  the  relative  vapor  density  of  a  liquid  at  two  different  tem- 
peratures gives  the  relative  osmotic  pressures  of  the  dissolved 
vapor  at  those  temperatures,  were  it  not  for  the  uncertainty  as 
to  tow  the  temperature  affects  the  pressure  of  the  vapor  and  the 
osmotic  pressure  of  the  dissolved  vapor.  It  would  not  do  to 
assume  that  each  was  affected  in  the  same  degree  by  a  change 
in  temperature.  But  our  theory  does  allow  us  to  claim  in 
the  c£se  of  a  given  liquid  at  a  constant  temperature  that  two  dif- 
ferent vapor-tensions  will  correspond  to  two  different  concentra- 
tions of  the  dissolved  vapor  by  Henr>''s  law,  and  that  the  rela- 
tive vtpor-tensions  are  as  the  relative  concentrations  of  the  dis- 
solved vapor.  Now  we  can  change  at  will,  within  quite  a  wide 
range,  the  vapor  tension  of  a  liquid  without  changing  its  tem- 
perature and  without  introducing  any  complications. 

To  understand  this  let  us  refer  back  to  the  original  conception 
of  the  dissolved  vapor.  If  we  have  liquid  water  in  a  vessel  with 
any  number  of  gases  under  moderate  pressure,  the  partial  pres- 
sure of  the  saturated  water  vapor  will  be  ver}'  nearly  the  same 
as  if  it  alone  were  present  in  the  space  containing  the  gases. 
So  when  we  dissolve  a  substance  in  water  it  would  seem  as  if 
we  might  argue  that  the  osmotic  pressure  of  the  dissolved  sub- 
stance should  not  affect  the  pressure  of  the  dissolved  water  vapor. 
However  the  conditions  in  the  two  cases  are  very  different.  In 
the  first  case  there  is  abundant  space  for  the  water  vapor  so 
that  all  that  is  necessary  is  time  for  the  concentration  of  the 
water  vapor  to  reach  the  same  value  no  matter  how  many  gases 
may  be  present,  provided  of  course  that  the  total  pressure  be 
not  very  great.  When  however  the  total  pressure  is  great,  the 
vapor-tension  of  the  liquid  diminishes  very  much.  This  is  just 
the  condition^that  holds  in  a  liquid.  The  volume  available  for 
a  dissolved  substance  is  very  small,  and  so  anything  put  into  this 
space  will  very  materially  lessen  the  space  available  for  the  dis- 


732  CLARENCE   L.   SPEYERS. 

solved  vapor,  particularly  as  the  quantities  used  in  solutions  are 
generally  very  much  greater  than  those  used  in  the  gaseous  state. 
Suppose  we  have  n  gram-molecules  of  a  substance  whose 
molecules  do  not  dissociate  on  dissolving,  say  sugar,  and  dissolve 
it  in  water.  Let  y  be  the  number  of  gram-molecules  of  dissolved 
vapor  after  the  n  gram-molecules  of  substance  have  been  dis- 
solved, then  the  total  number  of  gram-molecules  present  in  solution 
will  be  y-i-n,  and  the  relative  number  of  gram-molecules  of  sub- 
stance dissolved  to  total  number  of  gram-molecules  in  solution  is 


v-|-  n  ' 

Now  Jet  J  be  the  concentration  of  the  dissolved  vapor  when 
alone  in  the  liquid,  and  f  its  concentration  after  the  new  sub- 
stance has  been  added,  in  this  case  the  sugar,  j — /  will  be  the 
decrease  in  the  concentration  of  the  dissolved  water  vapor  due 
the  addition  of  the  n  gram-molecules  of  sugar,  and  since  a  gram- 
molecule  of  all  substances  occupies  the  same  volume,  the 
decrease  in  concentration/ — /  will  be  the  same  whate\er  the 
substance  dissolved  may  be,  provided  the  same  number  of  gram- 
molecules  be  taken  in  each  case,  or  the  decrease  in  concentration 
of  the  dissolved  vapor  is  proportional  to  the  number  of  gram- 
molecules  dissolved  in  a  certain  fixed  volume  of  solution.  If  the 
temperature  is  constant  the  concentration  of  the  dissolved  water 
vapor  cannot  rise  above  the  value  y,  which  it  has  when  only  dis- 
solved vapor  is  present ;  when  we  trj*^  to  get  above  this  value  the 
dissolved  vapor  turns  to  liquid  water.  Hence  the  number  of 
gram-molecules  in  a  unit  volume  must  be  fixed,  if  the  tempera- 
ture is  constant,  that  is 

y-^-n  =  constant. 
We  have,  therefore, 


J  y  +  n 

where  a  is  a  constant. 

J — f  can  be  calculated  by  van't  Hoff's  law,  and  n  is  known, 
but  the  other  quantities  are  not,  so  neither  /  or  y  can  be  calcula 
ted  from  this  equation. 


SOMB  THOUGHTS  ABOUT   LIQUIDS.  733 

There  is  however  another  relation  that  can  be  deduced. 

The  concentration  of  the  dissolved  vapor  is  measured  b}-  its 
osmotic  pressure. 

Let  n^  ^,  be  respectively  osmotic  pressure  and  osmotic  volume 
of  the  dissolved  vapor,  when  it  alone  is  present ;  n\  ^\  the  cor- 
responding quantities  when  a  substance  is  in  solution  \  p,v,  the 
pressure  and  volume  of  the  vapor  in  contact  with  the  pure  sol- 
vent ;  p\  v\  the  corresponding  quantities  when  a  substance  is 
in  solution. 

Consider  an  isothermal  reversible  cycle  composed  of  the  fol- 
lowing parts. 

*  By  means  of  a  semipermeable  diaphragm  let  a  gram^molecule 
of  dissolved  vapor  pass  from  the  pure  solvent,  the  work  will  be 

—  n^  =  —RT, 

Let  the  g^am-molecule  of  vapor  expand  until  it  has  a  pres- 
sure of  n  the  work  will  be 


—  I  ^n  =  —R  Tl  —,. 


Let  it  then  pass  into  the  solution ;  the  work  will  be 

-f.;r'^'=:  +  /?7'. 

Let  X  gram-molecules  pass  out  of  the  solution  in  the  form  of 
vapor ;  the  work  will  be 

—  xp'v'=.—xRT, 

w^here  x  denotes  the  number  of  gram- molecules  of  gaseous  vapor 
necessary  to  make  one  gram-molecule  of  dissolved  vapor. 

Let  the  x  gram-molecules  of  vapor  be  compressed  until  the 
pressure  equals  p  ;  the  work  will  be 

J^x^  vdp=^'xRTl{,, 

Let  the  x  gram-molecules  be  driven  into  the  pure  solvent ; 
the  work  will  be 

'\'Xpv=^xRT. 


734  CLARENCE  L.   SPEYERS. 

Thus  the  cycle  .is  completed.  The  quantity  of  solution  is 
supposed  to  be  so  large  that  the  addition  and  removal  of  the 
quantity  of  the  solvent  used  in  the  cycle  will  have  no  appreciable 
effect  upon  the  concentration  of  the  solution. 

The  sum  of  the  changes  of  energy  must  be  zero,  so 

--RT  —  RTl^  +   RT—  xRT-h  xRTl-^  +  xRT  =  o; 

n  P 

We  shall  assume  that  x  equals  i  ;  there  is  no  good  reason  for 
thinking  otherwise,  and  the  simplicity  in  the  structure  of  dis- 
solved bodies  favors  this  assumption. 

From  the  theory  we  have 


J       ^ 

We  have  therefore  from  i  through  3  and  2, 


(3> 


/'      p—p'  n  ,  . 


but  from  experiment, 

p-p'_       n 


(5) 


p  N+  n 

where  A^  is  the  number  of  gram-molecules  of  liquid  in  which  n 
gram- molecules  of  substance  have  been  dissolved. 
Hence, 

n  n 

^—T =    Art  (6) 

Now  as  equation  (6)  is  true  for  any  small  value  of  n  it  will 
be  true  for  a  value  so  small  in  comparison  with  v  and  N,  that  it 
may  be  neglected,  and  so 

an  _   n 

or, 

«  =  ^  (7> 


SOME   THOUGHTS  ABOUT   LIQUIDS.  735 

Substituting  in  (6)  we  have 


y       n  n 


N    y+n      N+n* 
or, 

y=iJV  (8) 

That  is,  the  concentration  of  the  dissolved  vapor  is  the  same 
as  the  concentration  of  the  liquid,  or  in  other  words,  all  the  sol- 
vent is  to  be  considered  as  dissolved  vapor. 

This  is  very  interesting,  for  it  is  in  effect  the  same  conclusion 
that  van  der  Waals  reached  in  his  celebrated  treatise,  though 
he  pursued  a  very  different  method. 

It  would  seem  from  this  result  that  matters  were  left  in  about 
the  same  state  that  they  were  in  at  the  outset;  that  the  view  of 
dissolved  vapor  was  no  better  than  the  old  view,  which  claimed 
that  the  change  into  vapor  took  place  only  on  the  surface  of  the 
liquid.     But  we  have  really  gained  several  things. 

In  the  first  place  we  have  found  that  reasoning  from  the 
analogy  that  a  dissolved  gas  and  the  same  gas  in  contact  with 
the  solvent  bears  to  the  liquid  and  its  vapor  we  got  to  the  idea 
of  dissolved  vapor  and  from  that  to  a  result  in  agreement  with 
a  much  older  theory. 

Secondly,  we  have  found  that  a  liquid  is  to  be  looked  upon  as  a 
condensed  gas,  not  simply  condensed  in  the  sense  that  it  is  a  mat- 
ter compressed  into  smaller  space,  but  condensed  in  the  sense 
that  the  gaseous  activity,  pressure,  is  carried  into  the  liquid  con- 
dition, and  we  are  to  treat  a  liquid  as  we  would  a  gas. 

Thirdly,  it  follows  from  this  view  that  a  substance  dissolved 
is  simply  brought  into  the  same  condition  that  the  liquid  is  in, 
and  consequently  should  have  the  same  property  of  exerting  an 
osmotic  pi'essure  that  the  liquid  has. 

Finally,  what  causes  the  condensed  gaseous  state  ?  Until  this 
is  answered  the  problem  of  liquid  and  gas  is  essentially  unsolved. 
That  it  is  due  to  an  attraction  between  the  molecules,  is  hardly 
possible,  as  we  have  seen  at  the  beginning  of  this  paper.  Indeed 
so  soon  as  we  begin  to  reflect  upon  the  complications  that  are 
introduced  the  moment  the  ideas  of  molecule  and  attraction  are 
brought  into  an  investigation,  and  these  complications  are  all 


736  SOME  THOUGHTS  ABOUT   LIQUIDS. 

the  time  increasing  instead  of  diminishing,  the  more  natural 
and  simple  appears  the  view  of  Ostwald  that  we  shall  find  a 
better  solution  of  such  problems  in  energy  alone,  matter  being 
only  a  collection  of  energies  in  space. 

Now  as  to  the  value  of  the  osmotic  pressure  in  some  liquids. 

In  looo  cc.  of  water  there  are 

looo  ,      . 

=  55-55  gram-molecules. 


i8 

Every  gram-molecule  at  25**  C.  (=298®  absolute  temperature) 
in  1000  cc.  has  a  pressure  of 

22222  298  , 

. .  0.70  m. 

1000    273        ' 

Hence  for  the  55.55  gram- molecules  of  water  we  have 

22222   298         ^  1000  ^         , 

n  = . — ^.  0.76 — r— =  1024  meters  of  mercury. 

1000    273       '18 

In  1000  cc.  methyl  alcohol  there  are 

i^.  0.79  gram-molecules. 

and  hence  for  methyl  alcohol  we  have 

22222  298         ,  1000 

n  = .-^--.  0.76 .  0.79  =  455  m. 

1000  273       *       32  '^       ^^"^ 

In  1000  cc.  ethyl  alcohol  there  are 

1000  ,       - 

— 2~'  0.79  gram- molecules, 
46 

and  hence  for  ethyl  alcohol  we  have 

22222   298         ,  1000 

«'  = . — ^-.  0.76 — 7—.  0.79  =  316  m. 

1000  273      46 

In  1000  cc.  propyl  alcohol  there  are 

i^.  o,,o,„„.^,.u.«, 

and  hence  for  propyl  alcohol  we  have 

22222   298         ,  1000        o 

n  ■=:  .-^-.  0.76 — r-.  0.80  =  249  m. 

1000    273  46 


VOLUMETRIC   ESTIMATION  OP  LEAD.  737 

In  looo  cc.  chloroform  there  are 

lOOO  -        , 

.   1.52  gram-molecules, 

and  hence  for  chloroform  we  have 

22222   298         ^  1000 

n  = .— ^-.  0.76 .  1.52  =  235  m. 

xooo    273  119 

In  1000  cc.  toluene  there  are 

'"^  .  0.88  gram-molecules, 


92 
and  hence  for  toluene  we  have 


22222   298         -  1000        „„  ^ 

n  ■=. . — ^^.  0.76 .  0.88  =  176  m. 

1000    273  92 


RUTGRRS  COLX.BGB, 


[Contributions  from  the  Laboratories  of  the  School  of  Mining, 

Kingston,  Ontario.] 

VOLUriETRIC  ESTiriATION  OF  LEAD. 

By  Fred.  J.  Pope. 
Received  May  at,  1896.      « 

aUITE  frequently  of  late,  the  attention  of  readers  of  chemi- 
cal journals  has  been  directed  to  various  methods'  for 
estimaFing  lead  volumetrically .  But,  while  some  of  these  methods 
are  superior  to  any  previously  made  public,  yet,  for  none  of 
them  is  that  degree  of  accuracy  claimed  which  is  so  essential  in 
a  reliable  quantitative  operation.  The  chief  objection  to  all  of 
these  methods  is  the  use  of  an  outside  indicator.  However,  by 
using  an  inside  indicator  and  modifying  slightly  the  usual 
preliminary  steps  (necessary  for  the  conversion  of  the  ore  into 
the  sulphate)  results  have  been  obtained  by  the  writer  which 
are  quite  satisfactory. 

The  operation  may  be  briefly  outlined  as  follows  :  The  lead  is 
first  converted  into  lead  sulphate,  then  into  lead  acetate.  Excess 
of  standard  potassium  bichromate  is  added,  which  precipitates 
the  lead  as  lead  chromate.  The  unused  potassium  bichromate 
is  reduced  by  excess  of  standard  arsenious  acid,  and  this  latter 

IThis  Journal,  17,  90Z ;  Engineering  and  Mining  Journal^  July  7,  1894. 


73^  FRED.  J.   FOPB. 

titrated  with  iodine  solution,  using  starch  paste  as  an  indicator. 

PREPARATION  AND  STANDARDIZING  OP  SOLUTIONS. 

Taking  tenth  normal  solution  of  iodine  as  the  standard,  4.995 
grams  of  arsenious  add  per  liter  and  4.763  grams  of  potassium 
bichromate  per  liter  give  standard  solutions  of  equivalent  value 
per  equal  volumes. 

Iodine. — 12.7  g^ms  are  dissolved  in  concentrated  potassium 
iodide  solution  and  made  up  to  one  liter. 

Arsenious  Acid. — Dissolve  4.95  grams  in  twenty  or  thirty  cc. 
of  saturated,  filtered  solution  of  sodium  carbonate,  gently  warm- 
ing. If  too  strong  heat  is  applied  the  arsenious  acid  cakes  and 
dissolves  with  difficulty. 

By  means  of  a  burette  accurately  measure  ten  to  fifteen  cc.  of 
arsenious  acid  solution,  running  it  into  a  large  porcelain  dish. 
Acidify  faintly  with  sulphuric  acid,  add  fifty  cc.  saturated  solu- 
tion of  pure  sodium  bicarbonate,  add  starch  paste  and  titrate 
with  the  iodine. 

Potassium  Bichromate. — Weigh  out  approximately  five  grams, 
dissolve  and  make  up  to  one  liter.  Remove  twenty-five  cc.  to  a 
porcelain  dish,  add  fifty  cc.  of  the  standard  arsenious  acid  and 
proceed  with  titration  as  already  indicated. 

Note. — Since  all  commercial  sodium  bicarbonate  will  decolor- 
ize more  or  less  iodine,  it  is  well  in  neutralizing  to  get  the  neu- 
tral point  exactly.  When  this  is  attained,  add  fifty  cc.  sodium 
bicarbonate  and  deduct  its  iodine  value  from  the  quantity  con- 
sumed. 

The  Operation  in  Detail. — Take  from  three  to  seven  grams  of 
ore,  according  to  its  richness  in  lead.  Place  this  in  a  deep 
three-inch  porcelain  dish,  thoroughly  moisten  it  with  water, 
cover  the  dish  with  a  watch-glass  and  for  each  g^am  of  ore  used 
add  four  to  five  cc.  of  a  previously  prepared  mixture  of  two  parts 
by  volume  of  sulphuric  acid,  three  parts  by  volume  of  nitric 
acid  and  one  part  by  volume  of  water.  When  the  reaction, 
which  first  results,  diminishes,  evaporate  as  nearly  to  dryness  as 
is  possible  without  spurting.  Cool,  fill  the  dish  with  cold  water, 
stir  well  and  allow  to  settle  for  two  or  three  minutes.  Filter  and 
wash  with  cold  water  until  most  of  the  acid  is  removed.  Convey  the 


VOLUMETRIC   ESTIMATION   OF  LEAD.  739 

filter  paper  with  the  precipitate  to  a  300  to  400  cc.  beaker  or 
Erlenmeyer  flask  and  neutralize  any  remaining  acid  with  dilute 
ammonia.  To  the  porcelain  dish  add  ten  to  fifteen  cc.  strong 
ammonium  acetate,  made  decidedly  acid  with  acetic  acid.  Add 
an  equal  volume  of  water  and  boil  for  two  or  three  minutes, 
washing  the  sides  of  dish  so  as  to  remove  any  remaining  lead  sul- 
phate. This  solution  is  then  added  to  the  flask  containing  the 
precipitate  and  the  whole  boiled  from  seven  to  ten  minutes  with 
frequent  stirring.  Cool,  neutralize  with  ammonia,  add  100  cc. 
of  standard  potassium  bichromate,  stirring  well.  Filter  into  a 
half  liter  measuring  flask,  moistening  the  filter  paper  with 
dilute  ammonia  or  ammonium  acetate.  Wash  precipitate  as 
much  as  is  possible  in  the  flask,  using  hot  water.  The  filtrate 
make  up  to  the  mark,  and  for  titrating  remove  100  cc.  to  a  large 
one  and  one-half  liter  porcelain  basin.  Add  ten  to  twenty  cc. 
(or  less  if  ore  is  rich  in  lead)  of  standard  arsenious  acid.  Make 
decidedly  acid  with  forty  per  cent,  sulphuric  acid  and  stir  until 
the  yellow  color  disappears  or  the  liquid  has  a  greenish  tinge.  A 
large  excess  of  sulphuric  acid  is  to  be  avoided.  Neutralize  with 
saturated  solution  of  sodium  bicarbonate  and  then  add  an  excess 
of  fifty  cc.  If  the  solution  has  a  deep  greenish  tinge  dilute  it  with 
distilled  water.  Finally  add  starch  paste  and  titrate  with  stand- 
ard iodine  solution. 

As  a  test  of  the  accuracy  of  method,  five  portions  of  pure  lead 
sulphate  were  acted  upon  and  the  following  results  obtained  : 

Grams  taken.  Grams  found. 

1.0  1.000568 

1.1  1.099375 

1.2  1.200467 

1.3  1.300673 

1.4  I.39957I 

With  a  specimen  of  galena  containing  quartz  and  calcium  car- 
bonate, the  writer  obtained  the  following  percentages  : 

Grams  taken.  Per  cent,  lead  found. 
0.7    '  81.89 

0.7  81.96 

0.71  S1.94 

0.68  81.90 


740  SULPHIDES   IN   CALCIUM   CARBIDE. 

As  a  test  of  the  method  in  the  hands  of  inexperienced  opera- 
tors, it  was  outlined  and  explained  to  four  junior  students,  who 
with  the  galena  ore  already  mentioned,  obtained  the  following 
results : 

Grams  taken.  Per  cent,  lead  found. 

R.  H  =  0.7  81.86 

G.  E.  R.  =  0.7  81.78 

S.  D.  I  '  =  °-7  82.00 

12  =  0.85  81.89 

G.  E.  S  =  0.7  81.95 

With  another  ore  containing  five  per  cent,  of  copper,  twenty- 
six  per  cent,  of  iron,  quartz  and  gypsum,  one  of  the  students 
obtained  the  following  results : 

Grams  taken.  Per  cent,  lead  found. 
3.0  15.89 

3.5  16.01 

4.0  15.97 


ESTIMATION  OF  SULPHIDES  IN  CALCIUil  CARBIDE. 

Bt  Prbd.  J.  Pope. 

Received  May  31,  1896. 

A  WEIGHED  quantity  of  calcium  carbide  was  conveyed  to 
a  dry  Erlenmeyer  flask  provided  with  a  stop-cock  funnel 
and  a  delivery  tube,  which  latter  led  to  a  ten  ounce  wash  bottle, 
this  in  turn  being  connected  with  a  smaller  one.  The  wash  bot- 
tles contained  150  cc.  lead  acetate  of  known  strength  (about 
tenth  normal) .  By  means  of  a  stop-cock  water  was  carefully  added 
until  there  was  no  further  evolution  of  acetylene.  On  the  reac- 
tion ceasing,  twenty-five  to  forty  cc.  sulphuric  acid  (1:3)  was 
run  into  the  flask  and  the  whole  gently  boiled,  the  liberated 
hydrogen  sulphide  passing  into  the  wash  bottles  and  precip- 
itating the  lead  as  lead  sulphide.  When  the  reaction  had 
ceased  the  flask  and  liquid  was  washed  free  of  hydrogen 
sulphide  by  a  current  of  air  and  the  contents  of  wash 
bottles  filtered.  The  filtrate  containing  unconsumed  lead  ace- 
tate was  made  up  to  a  half  liter.  To  100  cc.  of  this  solution 
were  added  standard  potassium  bichromate,  arsenious acid,  etc., 
(as  indicated  in  preceding  article)  and  total  amount  of  uncon- 
sumed lead  acetate  estimated.      The  difference   between   this 


OIL   IN   BOILBR   SCALB.  741 

amount  and  the  quantity  of  lead  acetate  started  with  gave 
amount  precipitated  by  the  hydrogen  sulphide  from  which  the 
sulphur  existing  as  sulphide  was  calculated. 

Grams  calcium  carbide  taken.  Per  cent,  sulphur  found. 

2.4492  3-37 

3- "34  3-57 

No  attempt  was  made  to  check  the  application  fi  the  method. 
It  is  obvious  that  the  impure  calcium  carbide  may  have 
evolved  other  products  capable  of  removing  lead  from  the  solu- 
tion. It  is  the  writer's  intention  to  investigate  this  and  other 
points  connected  with  this  method. 


NOTE  ON  THE  PRESENCE  OP  OIL  IN  BOILER  5CALE.> 

By  CRAaLBS  A.  DoaBMUft. 
Received  June  9.  iBg6. 

IT  is  difficult  to  remove  cylinder  oils,  whether  pure  mineral  or 
mixtures  of  mineral  and  animal  from  condensed  exhaust 
steam.  The  practice  of  recovering  steam  either  for  the  prepara- 
tion of  distilled  water  or  for  boiler  feed  water  is  now  so  general 
that  opportunities  for  observing  the  troubles  attending  the  pro- 
cedure are  not  wanting. 

This  sample  of  water  was  obtained  by  melting  the  **core  "  of 
cakes  of  artificial  ice.  The  sediment  is  fine»  flocculent  and  of 
red  color.  When  removed  from  the  water  and  dried  it  is  pul- 
verulent. There  is  very  slight  evidence  of  oil  in  the  dry  mass, 
the  moist  sediment  does  not  appear  oily.  The  large  proportion 
of  oil  extracted  by  ether  shows  how  inefficient  the  filters  were  in 
purifying  the  condensed  steam.  Yet  ver>' great  pains  were  taken 
at  the  ice  plant  to  secure  pure  distilled  water,  and  there  was  no 
visible  oiliness  in  the  water  as  it  flowed  to  the  freezing  cans. 
Here  however  the  corrosive  action  of  the  distilled  water  on  the 
galvanized  iron  produced  a  mass  of  iron  and  zinc  hydrates  which 
in  being  pushed  to  the  centre  by  the  gradual  formation  of  ice 
gathered  the  oil  and  carried  it  to  the  core. 

Another  specimen  is  one  obtained  from  a  steamboat  trafficing 
on  the  Hudson  river  and  using  salt  or  brackish  water  in  the 
surface  condensers.    The  boilers  were  said  to  be  foul  with  masses 

1  Read  before  the  New  York  Section.  June  5th.  1886. 


742  OIL  IN   BOILBR  SCALE. 

of  oil  coating  the  sides  and  tubes.  Having  determined  the 
presence  of  the  salts  of  sea  water  in  the  boiler,  due  to  leaky 
condensers,  a  treatment  was  suggested  which  caused  a  fine  pre- 
cipitate. This  precipitate  gathered  the  oil  in  masses  easily 
brushed  from  the  crown  sheets.  When  this  mass  is  treated  with 
ether  a  dry  powder  remains  and  oil  dissolves. 

A  third  specimen  sent  for  examination  from  a  large  plant  in 
Chicago,  evaporating  2500  gallons  of  filtered  river  water  and 
25,000  condenser  water  every  twenty-four  hours.  Lubricating 
oil,  mineral  with  ten  per  cent,  animal,  is  freely  used,  and  the 
fine  clay  in  the  water  has  together  with  some  incrusting  in- 
gredients, caused  the  oil  to  form  into  balls. 

The  next  two  specimens  are  in  striking  contrast  to  the  fore- 
going. This  is  light  colored,  one-quarter  inch  thick,  has  a  layer 
of  dense  nature  near  what  must  have  been  the  heated  surface  on 
which  the  scale  formed  while  the  bulk  of  the  incrustation  is 
fibrous.  The  incrustation  consists  of  calcium  carbonate  and 
sulphate,  with  which  is  intermingled  clay  and  organic  matter, 
the  latter  partly  oil. 

The  general  appearance  of  the  next  sample  is  quite  different. 
The  incrustation  is  in  thin  sheets  about  three-sixteenths  inch 
thick,  of  light  slate  color,  and  made  up  of  alternating  la^'ers  of 
deposit  of  varying  hardness.  The  ingredients  are  again  calcium 
carbonate  and  sulphate  and  clay,  while  there  is  much  organic 
matter.  This  can  be  separated  from  the  mineral  in  great  part 
by  a  little  acid.  The  presence  of  oil  is  then  noticeable.  The 
boiler  of  this  plant  is  fed  wnth  Lake  Michigan  water  and  con- 
denser water.  The  latter  goes  directly  to  the  hot  well  of  twenty 
barrels  capacity.  While  there  are  no  oil  filters  the  boiler  is 
provided  with  a  skimmer,  which  draws  off  floating  materials 
from  just  below  the  water  line.  The  lubricating  oil  used  is 
mineral  with  fifty  per  cent,  animal. 

Notwithstanding  the  skimmer,  the  scale  has  formed  and  baked 
into  a  hard  mass.  It  is  highly  non-conducting.  It  can  beheld 
by  the  fingers  quite  near  to  where  a  portion  is  heated  in  a  Bun- 
sen  flame,  the  heat  of  which  distils  out  and  ignites  the  oil.  A 
few  pieces  of  this  scale  heated  in  an  improvised  retort  made 
from  a  test  tube  yield  quite  a  gas  flame.     The  presence  of  oil 


AMINES  IN  THE  JUICE  OP  SUGAR  CANE.  743 

to  the  extent  of  from  twenty  to  fifty  per  cent,  in  the  deposits 
and  scale  of  marine  boilers  filled  with  fresh  water,  any  loss 
being  made  up  from  the  exhaust  or  from  sea  water  has  been 
fully  set  forth  by  I<ewes/  who  also  gives  the  causes  thereof  and 
remedies  therefor.  He  also  alludes  to  the  possibilities  of  this 
type  of  scale  forming  in  stationary  boilers. 

The  specimens  presented  serve  to  illustrate  the  importance  of 
critically  examining  the  nature  of  the  ''organic  matter'^  of 
incrustations,  the  statement  '*loss  of  ignition"  being  far  too 
general. 


[COMTRIBUTBD  FROM  THE  LA-BORATORY  OP  THE  LOUISIANA  EXPERIMENT 

Station  and  Sugar  Schooi,.] 
OCCURRENCE  OP  THE  AfllNES  IN  THE  JUICE  OP  SUGAR 

CANE. 

By  J.  L.  Bebson. 
Received  June  15,  X896. 

THE  presence  of  amines  in  the  products  of  the  sugar  beet 
has  long  been  known,  but  until  this  sugar  season  they 
have  not  been  known  to  exist  in  the  juices  of  sugar  cane.  Last 
December, while  working  with  the  precipitate  formed  by  the  addi- 
tion of  lime  water  to  cane  juice,  it  was  noticed  that  the  product 
dried  at  about  no''  C.  had  a  fishy  odor.  Upon  heating  some  of 
this  in  a  test  tube  over  a  burner,  an  alkaline  vapor  was  given 
off  which  had  a  fishy  ammoniacal  odor.  So  about  300  grams  of 
the  dried  substance  was  gradually  heated  in  a  hard  glass  retort 
upon  a  sand  bath  until  an  almost  complete  destructive  distilla- 
tion was  effected.  The  products  evolved  were  passed  through  a 
condenser  and  then  through  a  series  of  [)  tubes,  each  of  which 
was  kept  at  a  temperature  a  little  below  the  boiling-points  of 
each  of  the  principal  amines.  A  solid  collected  in  the  condenser 
tube,  and  an  illuminating  gas  escaped  from  the  last  (j  tube,  which 
was  kept  at  — 10**  C.  These  products  were  not  examined.  There 
collected  in  the  first  receptacle  about  twenty  cc.  of  an  acid 
liquid.  This  was  made  alkaline  with  caustic  soda  and  dis- 
tilled. The  products  as  before  were  passed  through  the  series 
of  tubes  maintained  at  the  different  temperatures,  when  there 

1  Chew,  News^  63, 181. 


744      EXTRACTION  APPARATUS  FOR  FOOD-STUPP  ANALYSIS- 

collected  in  the  first,  along  with  some  water,  about  five  cc.  of 
clear  liquid,  which  was  strongly  alkaline,  had  a  pungent  fishy 
odor,  combined  with  hydrochloric  acid,  and  otherwise  manifested 
the  general  properties  of  the  amines.  An  attempt  was  made  to 
further  purify  it  by  freeing  it  from  the  water,  but  the  amount 
was  too  small  to  bring  to  a  definite  boiling-point.  The  remain- 
ing liquid  was  neutralized  with  hydrochloric  acid,  and  slowly 
evaporated  down,  whereupon  a  few  crystals,  slightly  colored  and 
deliquescent,  were  obtained.  The  quantity  was  too  small  to 
admit  of  an  elementary  analysis,  so  it  was  not  possible  to  say 
whether  the  product  was  a  single  amine  or  a  mixture  of  amines. 
The  filter  cake,  the  refuse  from  the  clarification  of  cane  juice, 
gave  the  same  odor  and  alkaline  vapor  upon  heating.  It  was 
my  aim  to  subject  several  pounds  of  the  filter  cake  to  the  same 
treatment  in  order  to  fully  clear  up  the  question,  if  possible,  but 
the  amount  of  other  work  required  of  me  prevented.  The  clear- 
ing up  of  the  matter  is  of  the  greatest  scientific  and  practical  inter- 
est to  the  sugar  industry,  as  it  will  doubtless  throw  light  upon 
the  nature  both  of  the  amido  and  albuminous  bodies  of  the  cane 
juice.  I  write  the  account  of  the  work  with  the  hope  that  some 
chemist  may  be  induced  to  continue  the  work,  as  the  writer  will 
discontinue  sugar  work. 


[Contributed  from  the  Laboratory  op  the  Louisiana  Ezpbrxmbnt 

Station  and  Sugar  School.] 

A  SIMPLE  AND  CONVENIENT  EXTRACTION  APPARATUS 

FOR  FOOD-STUFF  ANALYSIS. 

Bt  J.  L.  Bkbson. 

Received  Tans  15,  1896. 

THE  apparatus  shown  in  the  accompanying  illustration  I 
have  adapted  from  the  Johnston  extractor,  for  the  general 
use  of  the  average  student  in  the  laboratory  aiming  at  simplicity, 
greater  compactness,  convenience,  rapidity  of  operation,  and  accu- 
racy. The  extraction  tube  -£",  which  is  rather  short,  is  provided 
as  usual  with  a  perforated  platinum  disk  fused  into  the  bottom, 
and  in  addition  with  a  specially  devised  funnel  stopper  of  ground 
glass,  by  means  of  which  the  weighed  sample  can  be  rapidly 


NBW  BOOKS. 


745 


and  effectively  dried  to  constant  weight  in  a  current  of  dry 
hydrogen  or  other  inactive  gas  for  the  es- 
timation of  the  moisture,  and  at  the  same 
time  preparing  the  sample  for  extraction. 
Rubber  caps  are  placed  over  the  two 
ends  of  the  tube  during  the  cooling  and 
weighing.  For  the  extraction  of  the 
sample,  the  tube  displaced  in  a  Stutzer 
tube  S  as  shown  in  the  figure,  which  id 
connected  as  usual  with  an  ether  flask 
below,  and  by  means  of  either  a  cork  or 
mercury  joint  with  a  short  bulb  conden- 
ser above.  The  funnel  stopper,  placed 
as  shown,  directs  the  returning  drops  of 
the  liquid  upon  the  center  of  the  sam- 
ple, and  especially  it  prevents  the  loss  of 
the  sample  by  spattering.  This  is  a 
source  of  objection  to  all  forms  of  open 
extractors.  Owing  to  the  very  small 
percentage  of  fats  or  ether  extracts  in 
most  food  stuffs  a  small  loss  of  the  sam- 
ple from  this  cause  makes  a  very  large 
analytical  error  in  the  work,  whether  es- 
timated from  loss  of  the  sample  orgain  in  weightoftheetherflask. 
During  two  years  use  in  this  laboratory  we  have  obtained  with 
the  apparatus  very  concordant  results  between  duplicate  analyses, 
and  would  commend  it  for  the  use  of  students  especially.  By 
means  of  a  seven  mm.  glass  tube,  six  tubes  and  samples  are  dried  in 
a  current  of  hydrogen  at  a  time  in  a  water-oven.  The  whole 
apparatus  may  be  had  of  Max  Kaehler  and  Martini,  Berlin. 


NEW  BOOKS. 

Chbicistry  por  Enginbbrs  and  Manttfacturbrs.  By  Bbrtram  Bi^oukt 
AND  A.  G.  Bu)XAM.  In  two  Yolumes.  VoIvUMB  I,  Chbmistry  op  Bn- 
GiNBBRiNG,  Buii«DiNG  AND  Mbtali^urgy.  8vo.  244  pp.  London : 
Charles  Griffin  &  Co.,  Ltd.    Philadelphia  :  J.  B.  Lippincott  Co. 

This  work  is  a  compilation  of  material  intended  to  cover  the 
chief  branches  of  chemical  industry.    The  first  volume  deals  in 


74^  NEW   BOOKS. 

the  first  part  with  the  chemistry  of  building  materials,  fuel, 
steam  making  and  lubrication.  The  second  part  is  entirely  de- 
voted to  metallurgy. 

The  scope  of  the  work  necessitates  condensation,  yet  tbe 
reader  will  be  impressed  at  times  with  the  meagemess  of  descrip- 
tion ,  especially  as  the  treatment  of  other  subjects  seems  dispropor- 
tionately extended.  An  appearance  of  unevenness  in  treatment 
is  thus  given,  which  might  have  been  avoided. 

Books  of  this  class  are  more  difficult  to  write  as  the  limits  of 
technical  Science  are  widened  and  there  is  room  for  much  judg- 
ment in  holding  a  proper  balance  between  the  necessities  of  the 
reader  and  the  restricted  space  of  a  hand  book  or  text  book .  While 
this  book  will  be  very  serviceable  to  the  large  class  of  engineers 
and  manufacturers  for  whom  it  is  especially  written,  and  even 
to  the  student  of  industrial  chemistry,  it  can  hardly  be  of  much 
interest  to  **  the  expert  in  any  one  of  the  branches  touched 
upon"  {vide  preface).  The  touch  is  entirely  too  light  as  a  rule 
for  those  who  seek  extended  information.  The  entire  absence 
of  references,  also,  deprives  the  work  of  much  of  the  usefulness 
it  might  have  had  for  professional  readers  in  subjects  not  strictly 
their  own. 

The  illustrations  are  good  as  far  as  they  go,  but  are  much  less 
freely  supplied  than  the  nature  of  the  book  requires. 

The  subjects  of  gaseous  fuel  and  water  for  steam  making 
are  well  and  clearly  treated.  Of  boiler  cpmpositions  the  authors 
justly  say  that  '*  none  should  be  used  without  a  knowledge  both 
of  its  composition  and  of  that  of  the  water  to  be  treated, '  *  and  that, 
* '  all  are  sold  at  prices  bearing  but  a  remote  relation  to  their 
intrinsic  values.'*  As  to  the  preservation  of  iron  by  paint,  the 
statement  that  red  lead  paint  is  the  best  will  hardly  meet  un- 
qualified assent. 

The  treatment  of  the  metallurgy  of  iron  is  very  full,  and 
contains  a  good  though  brief  discussion  of  the  influence  of 
foreign  elements  on  the  quality  of  iron.  The  statement  that  **the 
chief  gold-producing  countries  are  Australia,  America  (Cal- 
ifornia), and  Russia'*  is  more  compact  than  edifying.  Electro- 
metallurgical  processes  are  given  in  treating  of  many  of  the 
metals.     The  commercial  production  of  aluminum  is  described 


BOOKS  RECEIVED.  747 

briefly  but  no  allusion  is  made  to  the  part  which  the  United 

States  liave  played  in  the  development  of  this  industry,  nor  do  the 

names  of  Cowles  or  Hall  appear  in  the  text. 

The  second  volume  will  treat  of  the  chief  chemical  industries 

other  than  those  referred  to. 

A.  A.  Breneman. 

Laboratory  Experiments  in  General  Chemistry.    By  Charles  R. 

Sanger,  A.M.,  Ph.D.     Paper.    St.  Louis.     Published  bv  the  Author. 

1896. 
Experiments  in  General  Chemistry  and  Qualitative  Analysis.    By 

Charles  R.  Sanger,  A.M.,  Ph.D.     Paper.    St.  Louis.    Published  by 

the  Author.     1896. 

These  two  little  books  by  Professor  Sanger  contain  well  se- 
lected collections  of  experiments  for  beginners  in  chemistry.  The 
first  collection  was  prepared  for  students  in  a  general  college 
course,  while  the  second  collection  appears  to  have  been  arranged 
for  students  beginning  a  medical  course.  In  the  first  collection 
for  college  students  there  is  evidence  that  the  author  had  in  mind 
the  needs  of  those  who  spend  but  part  of  a  year  in  the  labora- 
tory. What  the  student  is  told  to  do  is  clearly  indicated  and  his 
attention  is  directed  at  every  step  to  the  important  points  in  the 
reactions  considered.  The  experimental  course  offered  to  med- 
ical students  is  not  as  extended  as  the  present  writer  would  like 
to  see,  but  is  as  full  as  this  class  of  students  is  supposed  to  need, 
and  has,  besides,  the  advantage  of  systematic  arrangement. 

' J.  H.  Long. 

BOOKS  RECEIVED. 

Eighth  Annual  Report  of  the  Kentucky  Agricultural  Experiment  Sta- 
tion of  the  State  College  of  Kentucky,  for  the  year  1895.  Lexington,  Ky. 
Ixvi,  150  pp. 

North  Carolina  Weather  during  the  Year  1895.  North  Carolina  Agri- 
cultural Experiment  Station,  Raleigh,  N.  C.     lii,  264  pp. 

Bulletin  No.  122.  Cost  of  Nitrogen,  Phosphoric  Acid  and  Potash. 
Proper  Use  of  Tables  of  Analysis  of  Fertilizers.  Connecticut  Agricultu- 
ral Experiment  Station,  New  Haven,  Conn.     16  pp. 

Reduction  of  Nitrates  by  Bacteria  and  Consequent  Loss  of  Nitrogen. 
By  Ellen  H.  Richards  and  George  William  Rolfe.  20  pp.  Reprinted 
from  the  Technology  Quarterly,  Vol.  IX,  No.  i,  March,  i89i5. 

Nitro-Explosives.   A  Practical  Treatise.   By  P.  Gerald  Sanford,  F.  I.  C, 


74^  OBTITART  NOTE. 

F.  C  S.    LondoD :  Croabj,  Lockvwid  &  Sod.     1896L    zii,  270  pp.    Price, 


EmbAlmiag  mod  Embalming  Fluids,  with  the  BibliogiaphT  of  Em- 
balming. Thesis  hj  Charles  W.  McCnrdj.  Sc.D.,  Ph.D.  Wooster.  Ohio : 
The  Herald  Pnblishiiig  Co.    April.  1896L    84  pp. 

Bnlletio  No.  64.  Analysis  of  Commercial  Fertilizers.  Keotnckj  Ag- 
ricnltnral  EiEperiment  Station  of  the  State  of  Kectockj.  Lexington,  K j. 
JBI7, 1896.     16  pp. 


OBITUARY  NOTE. 

Peter  Coluer,  Ph.D.,  was  bom  in  Chittenango,  New  York, 
August  17,  1835.  He  graduated  A.B.  at  Yale  College  in  iS6i 
and  later  Ph.D.  He  also  graduated  at  the  Sheffield  Scientific 
School,  and  was  for  some  time  an  assistant  and  instructor  in  that 
School.  From  1S67  to  1S77,  he  was  Professor  of  Ph\-sics  and 
Chemistry  in  the  University  of  Vermont,  and  also  Professor  of 
Chemistry  in  the  Medical  Department  of  that  University,  and 
for  some  time  Dean  of  the  Medical  Faculty.  In  1S73  he  was  ap* 
pointed  one  of  seven  scientific  commissioners  to  the  Vienna 
Exposition,  by  President  Grant.  From  1S77  to  1S82  he  was 
Chief  Chemist  to  the  Department  of  .Agriculture  of  the  United 
States,  at  Washington.  During  his  omcial  term,  he  gave  very 
great  attention  to  the  problems  of  cultivating  sorghum  and 
manufacturing  sugar  from  it.  From  1SS2  to  1S35  he  still  re- 
mained in  Washington,  engaged  in  chemical  pursuits  and  writ- 
ing for  scientific  and  agricultural  publications.  From  1S87  to 
I S95  he  was  Director  of  the  New  York  State  Experiment  Station 
at  Geneva,  New  York,  where  he  instituted  much  experimental 
work  especially  upon  fertilLzers  and  dairy  problems.  He  had  a 
wide  acquaintance  with  scientific  men,  and  himself  possessed 
great  energy  and  force.  Illness  obliged  him  to  resign  his  posi- 
tion last  year  and  he  came  to  Ann  Arbor  last  December.     He 

died  on  June  29. 

A.  B.  Prescott- 


ERRATA. 

Oa  pazc  651.  Ja!y  aamber,  15th  lice  frcrm  top,  for  159.000  rtraJ  166,000. 
On  pa.^e  653,  7th  line  frozn  bcttozn.  for  159  000  riraJ  166.000. 
On  page  654.  4th  line  from  top,  for  156. 5 19.5  r^J  116,519.5. 


Vol.  XVIII.  [September,  1896.3  No.  9. 


THE  JOURNAX 


OF  THE 


AMERICAN  CHEMICAL  SOCIETY. 


THE  DETERMINATION  OF  REDUCING  5UQARS  IN  TERHS 

OF  CUPRIC  OXIDE. 

By  George  Depren. 

Received  July  9.  1896. 

IT  is  now  approximately  fifty  years  since  alkaline  metallic 
solutions  were  first  used  in  determining  quantitatively  the 
various  reducing  sugars.  During  this  period  of  time  many 
investigators  have  worked  on  the  subject,  and  much  has  been 
done  towards  perfecting  the  method  of  analysis,  so  that  to-day 
there  are  several  admirable  processes  in  use  for  the  exact  esti- 
mation of  these  carbohydrates. 

The  quantitative  methods  of  determining  reducing  sugars 
may  be  divided  into  two  main  classes  :  those  based  upon  the 
volumetric  plan,  and  those  which  depend  on  a  gra\nmetric  esti- 
mation of  the  precipitated  cuprous  oxide. 

Of  the  first  class  many  processes  have  been  suggested  which 
have  met  with  more  or  less  success.  The  volumetric  methods 
are  mainly  used  for  factory  control  work ,  where  the  progress  of 
some  processes  require  a  rapid  and  fairly  accurate  idea  of  the  stage 
of  manufacture.  In  expert  hands  the  volumetric  methods  are 
capable  of  giving  excellent  and  concordant  results,  and  are, 
therefore,  used  in  the  laboratories  of  many  consulting  chemists, 
and  even  in  scientific  institutions. 

The  main  objections  to  the  use  of  the  volumetric  methods  are 
that  each  freshly  prepared  quantity  of  Fehling  solution  requires 
accurate  standardization  against  the  same  kind  of  pure  sugar  as 
that  which  is  undergoing  analysis.     Different  dilutions  and  the 


75^  GBORGB  DBPRBK. 

* 

time  of  boiling  affect  the  results  materially.  The  exact  deter- 
mination of  the  "end  point"  also  requires  considerable  practice 
and  skill. 

On  the  other  hand,  the  Fehling  liquor  used  in  the  gravimetric 
processes  need  not  be  made  up  as  accurately  as  is  required  for 
volumetric  work.  The  gravimetric  methods,  however,  ordi- 
narily require  more  time.  A  possible  loss  of  cuprous  oxide  by 
filtration,  and  an  incomplete  oxidation  to  the  higher  oxide  are 
also  potent  factors,  though  where  the  requisite  degree  of  care  is 
exercised  these  need  not  cause  anxiety.  The  same  statement 
regarding  dilution  and  time  of  boiling  holds  true  with  as  much 
force  in  gravimetric  as  in  volumetric  work. 

The  gravimetric  methods  are  generally  employed  for  scientific 
and  accurate  analytical  work.  Here  the  processes  are  compara- 
tively few,  all  depending  upon  the  oxidation  of  the  total 
sugar  present  in  an  excess  of  the  alkaline  copper  solution. 

The  tables  in  use  for  the  determination  of  reducing  sugars  are 
mainly  constructed  in  terms  of  metallic  copper.  As  the 
amount  of  metal  precipitated  per  gram  of  carbohydrate  is  not  a 
constant  for  all  dilutions  of  any  sugar,  specially  constructed 
tables  are  generally  employed.  Several  such  tables  have  been 
prepared,  as  for  instance  Allihn's  table  of  reduced  copper  for 
dextrose,  Wein's  table  for  maltose,  and  Soxhlet's  table  for  lac- 
tose, etc. 

Various  modifications  of  the  alkaline  copper  solutions  are  used 
for  the  determination  of  the  different  sugars,  each  requiring 
special  treatment.  Therefore  a  chemist  in  determining  the 
amount  of  malt  sugar  in,  say  beer,  must,  if  he  uses  Wein's  table 
for  maltose,  follow  exactly  Wein's  method  for  the  estimation  of 
that  sugar. 

Where  a  varietj-  of  work  is  carried  on  in  a  laboratory,  it  is 
therefore  necessary  to  have  several  different  Fehling  solutions  on 
hand  for  each  special  kind  of  determination.  If  all  the  tables 
for  the  estimation  of  the  different  carbohydrates  could  have  been 
prepared  for  use  under  uniform  conditions,  the  existing  state  of 
affairs  would  be  much  simplified. 

In  order  to  supply  this  need.  I  have  constructed  such  tables, 
using  a  method  which  I  have  employed  for  some  time,  in  deter- 


DBTBRMINATION  OP  RBDUCING  SUGARS.  75 1 

mining  reducing  sugars.  This  method,  proposed  by  O'  Sullivan' 
in  1876,  is  used  to  some  extent  in  England,  but  as  it  seems  to  be 
not  generally  known,  I  here  give  the  procedure  in  detail : 

To  fifteen  cc.  of  the  copper  sulphate  solution,  prepared  as  given 
below,  are  added  fifteen  cc.  of  the  alkaline  tartrate  solution,  in  an 
Brienmeyer  flask  having  a  capacity  of  from  250-300  cc.  The 
mixture  is  diluted  with  fifty  cc.  of  freshly  boiled  distilled  water 
and  placed  in  a  boiling  water  bath  for  five  minutes.  Twenty  to 
twenty-five  cc.  accurately  measured  from  a  calibrated  burette  of 
an  approximately  one-half  per  cent,  solution  of  the  sugar  to  be 
analyzed,  are  then  run  into  the  hot  Fehling  liquor  and  the 
whole  kept  in  the  boiling  water  bath  for  from  twelve  to  fifteen 
minutes.  The  flask  is  then  removed  from  the  bath  and  the  pre- 
cipitated cuprous  oxide  is  filtered  as  rapidly  as  possible,  either 
through  filter  paper  or  asbestos  in  a  Soxhlet's  tube  or  porcelain 
Gooch  crucible,  and  washed  with  boiling  distilled  water  until  the 
wash  water  no  longer  reacts  alkaline.  It  is  ignited  and  weighed 
as  cupric  oxide  as  described  below.  The  corresponding  amounts 
of  dextrose,  maltose  or  lactose  are  ascertained  by  reference  to 
the  tables  at  the  end  of  this  article.  It  should  be  noted  that  the 
above  directions  must  be  closely  followed.  The  volume  of  the 
Fehling  liquor  and  the  added  sugar  solution  should  be  from 
ioa-105  cc. 

The  Fehling  solution  used  is  made  up  according  to  Soxhlet's 
formula,  with  a  very  slight  modification.  69.278  grams  of  pure 
crystallized  copper  sulphate,  pulverized  and  dried  between  filter 
paper,  are  dissolved  in  distilled  water.  It  is  advantageous  to 
add  one  cc.  of  strong  sulphuric  acid  to  this,  as  recommended  by 
Sutton.'  The  whole  is  then  made  up  to  one  liter  with  distilled 
water  and  kept  in  a  separate  bottle.  The  alkaline  tartrate  solu- 
tion is  made  by  dissolving  356  grams  of  crystalline  Rochelle  salt 
and  100  grams  of  sodium  hydroxide  in  distilled  water  and  mak- 
ing up  to  one  liter.     This  is  also  kept  in  a  separate  bottle. 

Two  methods  of  filtration  of  the  precipitated  cuprous  oxide 
and  further  treatment  are  generally  adopted.  In  the  first  double 
"  washed "  filter  paper  is  used ;  in  the  other  the  precipitate  is 

ly.  CAem.  Soc..  a,  ijo,  /Sjd, 

s  Sutton  :  Fourth  edition,  (1883),  256. 


752  GEORGE   DEFREX. 

retained  by  a  layer  of  asbestos.  After  washing  the  precipitate 
on  the  filter  paper  as  above  described,  it  is  dried  in  the  usual 
manner  and  ignited  in  a  previously  weighed  porcelain  crucible » 
taking  care  to  bum  the  filter  paper  cautiously,  heating  for  fif- 
teen minutes  to  a  red  heat,  cooling  the  crucible  over  sulphuric 
acid  in  a  desiccator  and  weighing  as  cupric  oxide.  Additional 
treatment  with  nitric  acid  has  been  found  of  no  practical  advan- 
tage, the  results  by  direct  ignition  being  very  exact,  providing 
the  filter  paper  is  slowly  burned.  The  chief  objection  to  the 
employment  of  filter  paper  to  retain  the  precipitated  cuprous 

« 

oxide,  is  that  some  of  the  finely  divided  particles  are  liable  to 
go  through,  thus  causing  low  results. 

As  a  substitute  for  paper  carefully  selected  asbestos  is  often 
used  for  filtering  purposes.  To  insure  a  layer  of  asbestos  which 
shall  be  kept  at  constant  weight  under  the  action  of  hot  Fehling 
liquor,  it  is  advantageous  to  boll  the  mineral  with  nitric  acid 
(1.05-i.iosp.  gr.  >  for  a  short  time,  washing  the  acid  out  with 
hot  water,  and  then  boiling  with  a  twenty-five  per  cent,  solution 
of  sodium  hydroxide.  This  is  also  washed  out  with  hot  water. 
Reboiling  with  the  above  reagents  as  before  diminishes  the 
liability  of  leaving  any  soluble  portions.  As  thus  prepared  the 
filtering  material  may  be  kept  indefinitely  under  water  in  a  wide- 
mouthed  bottle  ready  lor  use. 

The  objections  of  some  chemists'  to  the  employment  of  asbes- 
tos on  the  ground  that  it  loses  weight  on  using,  does  not  seem 
to  hold,  when  it  is  prepared  as  above.  A  sample  boiled  as  stated 
with  acid  and  alkali  three  times,  lost  only  two-tenths  milligram 
when  two  "blanks  "  of  hot  dilute  Fehling  solution,  as  used  in 
the  process  above  described,  were  passed  through  the  mineral  in 
a  t>orceIain  Gooch  crucible. 

For  use,  a  layer  of  asbestos,  about  one  centimeter  in  thick- 
ness, is  placed  in  a  porcelain  Cooch  crucible,  to  retain  the  finely 
tiivuied  precipitate,  which  is  filtered  by  means  of  suction,  in  the 
u^l2al  manner.  The  crucible  containing  the  cuprous  oxide  is 
the.i  dropped  into  a  triangular  frame,  made  ot  platinum  wire, 
sa>penced  within  an  iron  radiator,  or  shell,  heated  to  redness. 
Th:**  quietly  ai^vi  thoroughly  drio  the  asbestos  without  cracking 


DBTBRMINATION  OF  REDUCING  SUGARS.  753 

I 

the  crucible.  After  about  five  minutes  the  crucible  is  trans- 
ferred by  means  of  a  pair  of  nippers  to  a  red  hot  platinum  cruci- 
ble and  heated  for  about  fifteen  minutes.  It  is  then  quickly 
transferred  to  a  desiccator  near  at  hand  to  prevent  cracking, 
allowed  to  cool  and  weighed.  As  cupric  oxide  is  somewhat 
hygroscopic,  it  is  advantageous  to  weigh  quickly  and  to  keep 
the  balance  case  as  dry  as  possible.  Prolonged  heating  in  the 
iron  radiator  would  have  changed  the  cuprous  oxide  to  the 
cupric  state.  The  advantage  of  transferring  the  porcelain  cru- 
cible to  a  red-hot  platinum  crucible,  is  that  the  oxidation  is 
quickly  completed,  as  a  much  higher  temperature  is  available. 

If  pressed  for  time,  another  determination  can  be  made  in  the 
same  crucible  without  cleaning  it.  As  a  rule,  it  is,  however, 
advisable  to  wash  out  the  cupric  oxide  by  means  of  hot  nitric 
acid  ( 1. 05-1. 10  sp.  gr.)  and  then  with  hot  water.  The  crucible 
is  then  heated,  cooled  and  weighed  as  before.  It  must  necessarily 
be  weighed,  because  this  treatment  with  hot  nitric  acid  dissolves 
some  of  the  asbestos. 

If  preferred,  a  Soxhlet's  tube  may  be  used  to  retain  the  pre- 
cipitated cuprous  oxide.  As  a  porcelain  Gooch  crucible  possessed 
obvious  advantages  over  this  apparatus,  I  have  used  it  in  all  my 
determinations  with  success. 

The  cupric  reducing  powers  of  dextrose,  maltose  aud  lactose 
were  determined  by  the  method  given  above.  A  table  for  invert 
sugar  was  not  constructed  because  most  invert  sugar  determina- 
tions are  made  by  double  polarization  in  a  saccharimeter. 

DEXTROSE. 

The  * 'cupric  reducing  power'*  of  dextrose  was  first  determined. 

This  is  defined  as  **the  amount  of  cupric  oxide  which  100  parts 

100  IV 
reduce."  *     This  may  be  represented  by jz — ,  in  which  W^is 

the  weight  of  cupric  oxide  obtained  by  the  given  weight  of  any 
sugar,  and  D  the  weight  of  cupric  oxide  formed  by  an  equal 
weight  of  dextrose.*  Hence,  if  the  amount  of  cupric  oxide 
formed  by  one  gram  of  dextrose  be  known,  the  amount  of  cupric 
oxide  reduced  by  one  gram  of  any  other  substance,  calculated 

I/.  Chem.  Soc.,  2,  130. 1876. 
ay.  CA€m.  Soc.,  Trans.,  606, 1879. 


754  GEORGE  DEPREN. 

upon  this  number  as  a  percentage,  will  represent  the  cupric  oxide 
reducing  power  of  the  substance,  which  we  denote  by  the  sym- 
bol A". 

The  amount  of  cupric  oxide  has  been  determined  by  O' Sulli- 
van* to  be  2.205  grams  per  gram  dextrose.  The  factor  for  dex- 
trose in  terms  of  cupric  oxide  is,  therefore,  the  reciprocal  of 
2.205  or  0.4535.  This  value,  0.4535,  ^as  assumed  to  be  a  con- 
stant for  all  amounts  of  dextrose  when  used  with  Fehling's  solu- 
tion in  the  manner  indicated.  As  such  it  was  a  very  convenient 
quantity,  it  being  only  necessary  to  obtain  the  weight  of  cupric 
oxide  formed  by  the  action  of  a  dextrose  solution,  multiply  this 
by  0.4535,  and  the  amount  of  dextrose  corresponding  was 
obtained.  No  tables  are  needed  if  this  assumption  be  true. 
Consequently  the  determination  of  dextrose  was  indeed  a  very 
simple  one. 

On  an  extended  investigation  of  this  subject,  using  various 
amounts  of  dextrose  on  the  same  volume  of  Fehling  liquor  in 
each  determination,  I  find  that  the  value  2.205,  above  given  as 
representing  the  quantity  of  cupric  oxide  obtained  by  the  action 
of  one  gram  of  dextrose,  is  not  as  was  heretofore  assumed,  a 
constant  for  all  weights  of  dextrose  taken,  the  amount  varying 
from  2.27  grams  cupric  oxide  per  gram  dextrose  for  small  quan- 
tities of  sugar,  to  2.22  grams  cupric  oxide  for  the  largest  amount 
of  dextrose  permissible.  AUihn,'  boiling  his  sugar  solutions  with 
the  Fehling  liquor  and  reducing  the  cuprous  oxide  to  copper, 
obtained  analogous  varying  results. 

The  purity  of  the  dextrose  used  was  first  determined,  dextrose 
anhydride  being  employed.  10.008  grams  of  anhydrous  dextrose 
were  disssolved  in  distilled  water  and  the  solution  boiled  to  pre- 
vent  birotation.  It  was  then  transferred  to  a  flask,  the  capacity 
of  which  at  15.5**  C.  was  100.08  cc,  thus  giving  a  solution  which 
contained  o.ioo  gram  dextrose  anhydride  per  cc. 

The  specific  gravity  of  the  above  solution  at  15.5*  was  deter- 
mined in  the  usual  manner  by  means  of  a  picnometer  with  ther- 
mometer attached. 

Capacity  picnometer  (at  15.5^)  =  55-ao55  cc. 
Dextrose  solution  (at  15.5^)  =  57*5083  grams. 

I  Loc.  cit. 

ij.praki,  OUm.,  (a).  M.63. 


DETBRMINAl'ION   OF   REDUCING  SUGARS.  755 

On  calculating  from  these  values  we  find  the  specific  gravity 
of  a  dextrose  solution  containing  ten  grams  dextrose  in  lOO  cc. 
to  be  1.03809  at  15.5**. 

The  specific  rotatory  power  was  determined  by  the  usual 
method,  a  Schmidt  and  Haensch  saccharimeter  being  used  in 
polarizing  the  dextrose  solution.  The  polarizations  were  carried 
out  in  a  200  millimeter  tube  at  20**.  To  change  from  the  read- 
ings of  a  saccharimeter  to  the  rotary  degrees,  it  is  necessary  to 
multiply  the  reading  observ^ed  by  0.344,  as  shown  by  Reinbach.* 
I  have  verified  this  value  with  concordant  results,  a  Laurent 
polariscope  being  used  for  comparison.  The  rotation  of  the  above 
solution  was  30.7  divisions.     This  gives  by  means  of  the  usual 

av 
formula  — [a]5  =  -^ —a specific  rotatory  power  of  52.8  ,  which 

is  in  accordance  with  that  obtained  by  other  observers.'     The 
dextrose  used  was  consequently  pure. 

For  the  determination  with  Fehling  liquor,  twenty-five  cc.  of 
the  dextrose  solution  at  15.5°  were  accurately  measured  from  a 
calibrated  burette  and  made  up  to  500  cc.  with  distilled  water  at 
the  same  temperature.  This  consequently  gave  a  solution,  each 
cubic  centimeter  of  which  contained  five  milligrams  dextrose. 
Various  quantities  of  this  were  then  taken  to  ascertain  the  cupric 
reducing  power  of  dextrose.  The  results  in  detail  are  given 
below.  In  each  case  the  combined  volumes  of  the  Fehling 
liquor  and  the  sugar  solution  were  made  up  to  105  cc.  as 
described  above. 


Milligrrams 

Cupric  oxide 
obtained. 

Cupric  oxide 

Dextrose 

Mean  dextrose 

dextrose. 

per  grram  dextrose. 

equivalent. 

equivalent. 

I2i 
I2i 

0.0283 
0.0285 

2.264 
2.280 

0.4416) 

0.4386  y 

0.4401 

25 
25 

0.0569 
0.0565 

2.276 
2.260 

0.4393  \ 
0.4425  i 

0.4419 

50 

O.I  129 

2.258 

0.4429  \ 
0.4452  i 

0.4440 

50 

O.II23 

2.246 

62i 

0.1407 

2.251 

0.4443  \ 
0.4454  ^ 

0.4449 

62i 

0.1403 

2.245 

75 

0.1683 

2.244 

0.4457  I 
0.4467  f 

0.4462 

75 

0.1679 

2.239 

^T^¥ 

1  Ber.  d.  chem.  Ges.^  87,  2282. 

<  Pribram :  Monat.f,  Chem.,  9,  399;  Landolt :  Ber.  d.  chem.  Ges.,  si,  191. 


MilUfframs 
dextrose. 

Cupric  oxi 
obtained 

lOO 

0.2233 

■ 

lOO 

0.2227 

"5 

0.2776 

"5 

0.2782 

"5 

0.2770 

"5 

0.2774 

125 

0.2777 

140 

0.3105 

140 

0.3100 

756  GEORGE   DEPREN. 

Cupric  oxide  Dextrose  Mean  dextrose 

per  gram  dextrose,      equivalent.  equivalent. 

2.233  0.4478 1  „.^3 

2.227  0.4489   » 

2.221  0.4503 
2.225  0.4493 
2.216  0.4512  |-  0.4503 
2.219  0.4506 

2.222  0.4500  J 
2.218  0.4508)  ^^5„ 
2.215  0.4515  J 

The  foregoing  values  of  the  amounts  of  cupric  oxide  per  gram 
dextrose  are  given  graphically  in  cur\'e  A,  Plot  I,  and  the  dex- 
trose equivalents  of  this  in  A,  Plot  II. 

From  this  we  get  the  amount  of  dextrose  corresponding  to  a 
given  weight  of  copper  oxide  by  means  of  the  formula : 

D  =  (0.4400  -f-  0.000037  ^)  ^» 
in  which  D  is  the  amount  of  dextrose,  and   W  the  weight  of 
cupric  oxide. 

The  dextrose  table  given  in  this  article  is  based  on  this  for- 
mula, the  values  of  W^  varying  from  30  to  320. 

MALTOSE. 

The  cupric  reducing  power  of  dextrose  is  given  as  100. 
Using  this  as  a  basis,  the  reducing  force  of  maltose,  as  given  by 
O'Sullivan,*  is  65.  Brown  and  Heron*  place  the  value  some- 
what lower,  claiming  that  61  is  more  exact.  The  results  which 
I  have  obtained  agree  very  well  with  this  latter  number. 

In  the  case  of  maltose,  as  with  dextrose,  it  was  found  that  the 
amount  of  cupric  oxide  obtained  per  gram  of  sugar  was  not  a 
constant.  The  cupric  reducing  power  of  various  amounts  of 
maltose  was,  however,  found  to  be  almost  exactly  a  constant 
when  referred  to  the  cupric  oxide  from  equal  weights  of  dex- 
trose. That  is,  calling  the  reducing  power  of  dextrose  100  for 
different  aliquot  parts  of  that  sugar,  the  cupric  reducing  power 
of  maltose  referred  to  this  standard  was  always  61. 

The  specific  gravity  of  maltose  was  determined  in  the  usual 
manner.  9.7558  grams  maltose  anhydride  were  dissolved  in  dis- 
tilled water  to  100.08  cc.  at  i5-5°- 

1  Loc.  a'l. 

«y.  CAem.  Soc.,  1879,  Trans.,  619. 


DETBBHINATION   OP   REDUCING   SUGARS. 


GEORGE   DBPRBI4. 


DBTERMINATION  OP   RBDUCING  SUGARS. 


759 


Maltose  solution  at  15.5*^  =  57*3049  grams. 

On  calculating  this  we  find  the  specific  gravity  of  the  above 
solution  to  be  1.03803.  For  a  solution  containing  ten  grams 
maltose  anhydride  in  100  cc.  it  would  consequently  be  1.03900 
at  15.5**. 

The  specific  rotatory  power  was  determined  as  usual.  The  rota- 
tion of  the  above  solution  at  20*^,  in  a  200  millimeter  tube,  was 
74.4  divisions  on  the  saccharimeter  scale.  This  gives  [flf]?  = 
I36.6'. 

As  maltose  anhydride  is  somewhat  difficult  to  prepare,  the 
solutions  used  to  determine  the  cupric  reducing  powers  were 
made  up  to  approximately  ten  per  cent,  from  the  maltose 
hydrate.  The  specific  gravity  of  the  solutions  was  then  deter- 
mined. Subtracting  from  this  value  i  .00000— the  specific  gravity 
of  water — and  dividing  the  remainder  by  0.00390,  we  get  the 
amount  of  maltose  anhydride  in  100  cc.  of  solution. 

Maltose  solution  at  15.5**  =  57.2511  grams, 
which  gives  a  specific  gravity  of  1.03754,  or  9.501  grams  mal- 
tose anhydride  in  100  cc. 

The  solution  for  Pehling  determinations  was  made  in  the 
same  manner  as  the  dextrose  solutions  above.  Each  cubic  cen- 
timeter of  the  diluted  maltose  solution  therefore  contained  4.75 
milligrams  maltose  anhydride. 

Milllfirnims 
maltiMe. 

23-75 

2375 

47-5 


cupric  oxide  Cupric  oxide  per 
obtained. 


47.5 

71.25 
7125 

95.0 

950 
118.75 

118.75 

142.5. 

142.5 
190.0 

190.0 

237.5 
237.5 


0.0329 
0.0327 
0.0656 
0.0654 
0.0983 

0.0979 
0.1304 
0.1300 
0.1623 
O.1619 
0.1940 

0.1934 
0.2572 
0.2566 
0.3198 

0.3193 


Strain  maltose. 
.386 


377 
381 

377 
380 

374 
373 
369 
370 

367 
361 
357 
353 
350 

347 
345 


Maltose 
equivalent. 

0.7218  ) 
0.7263  i 

0.7243 1 
0.7263  / 

O.72A7  > 
0.7278  ) 
0.7286 1 
0.7308  i 
0.7302  1 
0.7336  J 
0.7345  \ 

0.7369  r 
0.7284) 
0.7406  J 

0.7429  \ 
0.7437 ' 


Mean  maltose 
equivalent 

0.7240 


0.7253 


0.7263 


0.7297 


0.7319 


0.7354 


0.7395 


0.7433 


760  GEORGE   DEFREN. 

The  maltose  equivalent  in  terms  of  copper  oxide  is  shown  in 
B,  Plot  II.  From  this  we  get  the  amount  of  maltose  corres- 
ponding to  a  given  weight  of  cupric  oxide  by  the  formula  : 

J/=  (0.7215  +0.000061    W)  W, 

in  which  Af  is  the  weight  of  maltose,  and  IV  the  amount  of 
cupric  oxide  obtained.  It  will  be  seen  that  these  values  make 
the  cupric  reducing  power  of  maltose  0.61  that  of  dextrose. 

LACTOSE. 

Lactose  was  investigated  in  the  same  manner  as  the  preced- 
ing. 10.008  grams  lactose  anhydride  were  dissolved  in  distilled 
water,  boiled,  and  made  up  to  100.08  cc.  at  15.5^. 

The  above  solution,  polarized  in  a  200  millimeter  tube  at  20", 
gave  a  rotation  of  30.7  divisions.  This  gives  the  specific 
rotary  power  of  lactose  of  52.8". 

The  amounts  of  cupric  oxide  found  by  the  reduction  of  known 
weights  of  lactose  were  determined  as  in  the  previous  cases  with 
the  following  results  : 


Milligrams 

Cupric  oxide 
obtained. 

Cupric  oxide  per 

Lactose 

Mean  lactose 

lactose. 

gram  lactose. 

equivalents. 

equivalents. 

20 

0.0319 

1-595 

0.6269  ) 

0.6289 

20 

0.0317 

1.585 

0.6308  i 

^ 

50 

0.0798 

1.596 

0.6266  \ 

0.6274 

50 

0.0796 

1.592 

0.6282  i 

/  ~ 

75 

O.I  188 

1.584 

0.6313  \ 
.0.6334  * 

0.6323 

75 

O.I 184 

1-579 

100 

0.1577 

1.577 

0.6340  1 
0.6369  i 

0.6355 

100 

0.1570 

1.570 

125 

0.1955 

1.564 

0.6395  \ 

0.6^79 

125 

0.1964 

1. 561 

0.6363  i 

%j  §  f 

150 

0.2345 

1-563 

0.6397  I 
0.6410  > 

0.6404 

150 

0.2340 

1.560 

175 
175 

0.2729 
0.2724 

1.56c 
1-557 

0.6412) 
0.6424  / 

0.6418 

200 

O.3II2 

1-556 

0.6425  \ 

0.6430 

200 

0.3107 

1.553 

0.6436  i 

^^ 

The  cupric  oxide  values  per  gram  lactose  are  presented  graph- 
ically in  curve  C,  Plot  I,  while  the  reciprocals  of  these  quanti- 
ties are  shown  in  C,  Plot  II.  For  this  latter  the  amount  of 
lactose  corresponding  to  the  weight  of  cupric  oxide  obtained  is 
determined  by  the  following  : 


DETERMINATION   OF    REDUCING   SUGARS. 


761 


L  =  (0.6270+  0.000053  W)  W, 

in  which  L  is  the  lactose,  and  W  the  amount  of  copper  oxide. 
The  acompanying  table  for  lactose  is  constructed  on  this  basis. 

It  will  be  seen  from  the  above  results  that  the  amount  of 
cupric  oxide  produced  by  the  action  of  one  gram  of  reducing 
carbohydrate  on  Fehling  liquor,  in  the  manner  described,  is  not 
a  constant  for  all  dilutions. 

The  cupric  reducing  power  of  maltose  is  0.61  that  of  dextrose. 

The  following  tables  for  the  determination  of  the  reducing 
sugars  in  terms  of  cupric  oxide  are  based  on  the  analytical  results 
presented  above,  and  can  be  used  in  the  process  outlined  in  the 
same  manner  as  any  other  table  for  the  same  purpose  : 


P«rts 

Parts 
dextrose. 

Parts 
maltose. 

Parts 
lactose. 

Parts 

copper 

oxide. 

Parts 
dextrose. 

Parts 
maltose. 

Parts 

lactose. 

30 

13.2 

21.7 

18.8 

57 

25.1 

41.3 

35.9 

31 

13-7 

22.4 

19.5 

58 

25.5 

42.1 

36.5 

32 

14. 1 

231 

20.1 

59 

26.0 

42.8 

37.1 

33 

14.6 

23.9 

20.7 

60 

26.4 

43.5 

37.8 

34 

15.0 

24.6 

21.4 

61 

26.9 

44.3 

38.4 

35 

15.4 

253 

22.0 

62 

27.3 

45.0 

39.0 

36 

159 

26.1 

22.6 

63 

27.8 

45.7 

39.7 

37 

16.3 

26.8 

23-3 

64 

28.2 

46.5 

40.3 

38 

16.8 

27.5 

23.9 

65 

28.7 

47-2 

40.9 

39 

17.2 

28.3 

24.5 

66 

29.1 

47-9 

41.6 

40 

17.6 

29.0 

25.2 

67 

29.5 

48.6 

42.2 

41 

18. 1 

29.7 

25.8 

68 

30.0 

49.4 

42.8 

42 

18.5 

30.5 

26.4 

69 

30.4 

50.1 

43-5 

43 

19.0 

31.2 

27.1 

70 

30.9 

50.8 

44.1 

44 

19.4 

31.9 

27.7 

71 

31.3 

51.6 

44.7 

45 

19.9 

32.7 

28.3 

72 

31.8 

52.3 

45.4 

46 

20.3 

33.4 

29.0 

73 

32.2 

53.0 

46.0 

47 

20.7 

34.1 

29.6 

74 

32.6 

53.8 

46.6 

48 

21.2 

34.8 

30.2 

75 

331 

54.5 

47.3 

49 

21.6 

35.5 

30.8 

76 

33.5 

55.2 

47.9 

50 

22.1 

36.2 

31.5 

77 

34-0 

56.0 

48.5 

51 

22.5 

37.0 

32.1 

78 

34.4 

56.7 

49.2 

52 

23.0 

37-7 

32.7 

79 

34.9 

57-4 

49.8 

53 

23.4 

38.4 

33.3 

80 

35.4 

58.1 

50.5 

54 

238 

39-2 

34.0 

81 

35.9 

58.9 

51. 1 

55 

24.2 

39.9 

34-6 

82 

36.3 

59-6 

51.7 

56 

24.7 

40.5 

35-2 

83 

36.8 

60..^ 

52.4 

762 


OBORGB  DBPRBN. 


Parts 

Parts 

S2fl?.' 

Parts 

Parts 

Parts 

SSC:^ 

Parts 

Parts 

Parts 

dextrose. 

maltose. 

lactose. 

dextxx»e. 

maltose. 

lactose. 

84 

37-2 

61.1 

53.0 

127 

56.5 

92.5 

80.4 

85 

37.7 

61.8 

53.6 

128 

569 

93-3 

8x.i 

86 

38.1 

62.5 

54.3 

129 

57.3 

94.0 

81.7 

87 

38.5 

63.3 

54.9 

130 

57.8 

94.8 

82.4 

88 

39.0 

64.0 

55-5 

131 

58.2 

95.5 

83.0 

89 

39.4 

64-7 

56.2 

132 

58.7 

96.2 

83.6 

90 

399 

65.5 

56.8 

133 

59.1 

97.0 

84.2 

91 

40.3 

66.2 

57.4 

134 

59-6 

97-7 

84.9 

92 

40.8 

66.9 

58.1 

135 

60.0 

98.4 

85.5 

93 

41.2 

677 

58.7 

136 

«o.5 

99.2 

86.1 

94 

41.7 

68.4 

59-3 

137 

60.9 

99.9 

86.8 

95 

42.1 

69.1 

60.0 

138 

61.3 

100.7 

87.4 

96 

42.5 

69.9 

606 

139 

61.8 

101.4 

88.1 

97 

430 

70.6 

61.2 

140 

62.2 

102. 1 

88.7 

98 

43.4 

71.3 

61.9 

141 

62.7 

102.8 

89.3 

99 

43.9 

72.1 

62.5 

142 

63.1 

103.5 

90.0 

100 

44.4 

72.8 

63.2 

143 

63.6 

104.3 

90.6 

lOI 

44.8 

73.5 

63.8 

144 

64.0 

105.0 

91-3 

102 

45-3 

74.3 

64.4 

145    . 

64.5 

105.8 

91.9 

103 

45-7 

75.0 

65.1 

146 

64.9 

106.5 

92.6 

104 

46.2 

75-7 

65.7 

147 

65.4 

107.2 

93.^ 

105 

46.6 

76.5 

66.3 

148 

65.8 

108.0 

93.9 

106 

47.0 

77-2 

67.0 

149 

66.3 

108.7 

945 

107 

47.5 

77.9 

67.6 

150 

66.8 

109.5 

95-2 

108 

48.0 

78.7 

68.2 

I5» 

67.3 

1 10.2 

95.8 

109 

48.4 

79-4 

68.9 

15^ 

67.7 

III.O 

96.5 

no 

48.9 

80.1 

69.5 

153 

68.3 

111.7 

97.1 

III 

49.3 

80.9 

70.1 

154 

68.7 

112^ 

97.8 

112 

49.8 

81.6 

70.8 

155 

69.2 

1 13.2 

98.4 

"3 

50.2 

82.3 

71.4 

156 

69.6 

113.9 

99.1 

114 

50-7 

83.1 

72.0 

157 

70.0 

1 14. 7 

99-7 

"5 

5«  I 

83.8 

72.7 

158 

70.5 

115-4 

100.4 

116 

51.6 

84.5 

73-3 

159 

70.9 

1 16. 1 

10 1. 0 

"7 

52.0 

85.2 

74.0 

160 

7'.3 

1 16.9 

101.7 

118 

52.4 

85.9 

74.6 

161 

71.8 

1 17.6 

102.3 

119 

52.9 

86.6 

752 

162 

72.3 

1 18.4 

103.0 

I20 

53.3 

87.4 

75.9 

163 

72.7 

119.1 

103.6 

121 

53.8 

88.1 

76.6 

164 

73.2 

1 19.9 

104.3 

122 

54-2 

88.9 

77-2 

165 

73.6 

120.6 

104.9 

123 

54.7 

89.6 

77.9 

166 

74.1 

121.4 

105.6 

124 

551 

90.3 

78.5 

167 

74.5 

122. 1 

106.2 

"5 

556 

91. 1 

79.1 

168 

74-9 

122.9 

106.9 

126 

56.0 

91.8 

798 

169 

75-4 

123.6 

"*7-5 

DBTBRHINATION  OP  KBDUCING  SUGARS. 


763 


Parts 

Parts            Parts 
dextrose,      maltose. 

Parts 
lactose. 

Parte 

Parte           ] 
dextrose,     mt 

'arte 
tltose. 

Parts 
lactose. 

170 

75.8           ] 

[24.4 

108.2 

213 

95.3           3 

156.3 

136.0 

171 

76.3           ] 

[25.1 

108.8 

214 

95.8           ] 

157. 1 

136.7 

172 

76.8           ] 

[25.8 

109.5 

215 

96.3           3 

157.8 

137.3 

173 

77.3           3 

[26.6 

1 10. 1 

2X6 

96.7           1 

[58.6 

138.0 

174 

77.7           ' 

127.3 

1 10.8 

217 

97.2           3 

1593 

138.6 

175 

78.2           ] 

[28.1 

III.4 

218 

97.6         : 

[60.0 

139.3 

176 

78.6         : 

[28.8 

1 12.0 

219 

98.1         : 

[60.8 

139.9 

177 

79.1         ] 

129.5 

II2.6 

220 

98.6         ] 

161.5 

140.6 

178 

79.5 

^30-3 

"3.3 

221 

99.0         : 

162.3 

I41.2 

179 

80.0         ] 

131-0 

113.9 

222 

99.5         : 

[63.0 

141 .9 

180 

80.4         ] 

t3i.8 

114.6 

223 

99.9 

163.7 

142.5 

181 

80.8         ] 

132.5 

115. 2 

224 

100.4         ^ 

164.5 

143.2 

182 

81.3         ] 

t33.2 

I15.8 

225 

100.9         ^ 

165.3 

143.8 

183 

81.8         ] 

134.0 

1x6.5 

226 

101.3         : 

[66.0 

144.5 

184 

82.2         ] 

134.7 

117.1 

227 

101.8         : 

[66.8 

145. 1 

185 

82.7         ] 

^35.5 

1 1 7.8 

228 

102.2         ] 

167.5 

145.8 

]86 

83.1          ] 

136.2 

1 18.4 

229 

102.7         ^ 

[68.3 

146.4 

187 

83.5         3 

136.9 

1 19. 1 

230 

103. 1 

169. 1 

147.0 

188 

84.0         ] 

137.7 

1 19.7 

231 

103.6         ] 

[69.8 

147.7 

189 

84.4         ^ 

138.4 

120.4 

232 

104.0         ] 

[70.6 

148.3 

190 

84.9         J 

f39.i 

I2X.O 

233 

104.5         ^ 

I7«.3 

149.0 

191 

85.4         3 

^399 

121. 7 

234 

105.0         ] 

[72.1 

149.6 

192 

85.9         3 

[40.6 

122.3 

235 

105.4         ] 

[72.8 

150.3 

193 

86.3         1 

[41.4 

123.0 

236 

105.9 

'73.6 

150.9 

194 

86.8          ] 

[42.1 

123.6 

237 

106.3         ] 

174.3 

151. 6 

195 

87.2          1 

[42.8 

124.3 

238 

106.8         : 

175-1 

152.2 

196 

87.7          J 

143.6 

124.9 

239 

107.2         : 

175.8 

152.9 

197 

88.1          1 

144.3 

X25.6 

240 

107.7         ] 

[76.6 

153.5 

198 

88.6          ] 

[45.1 

126.2 

241 

108. 1         ] 

177.3 

154.2 

199 

89.0          ] 

145.8 

126.9 

242 

108.6         ] 

[78.1 

154.8 

200 

89.5          3 

[46.6 

127.5 

243 

109.0         ] 

[78.8 

155.5 

201 

89.9          ^ 

147-3 

128.2 

244 

109.5         ^ 

179.6 

156.1 

202 

90.4          ] 

[48.1 

128.8 

245 

109.9         1 

180.3 

156.8 

203 

90.8          ] 

[48.8 

129.5 

246 

110.4         ] 

[81. 1 

157.4 

204 

91-3          J 

[49.6 

130. 1 

247 

I  X0.9         ] 

[81.8 

158.1 

205 

91.7          3 

150.3 

130.8 

248 

111.3         J 

[82.6 

158.7 

206 

92.2          ] 

[51. 1 

UI.5 

249    • 

111.8         ] 

183.3 

159.4 

207 

92.6          1 

[51.8 

I32.I 

250 

112.3         1 

184.1 

160.0 

208 

93.1             3 

:52.5 

132.8 

251 

112.7         ] 

[84.8 

160.7 

209 

93-5          J 

153.3 

133.4 

252 

113.2         : 

185.5 

161 .3 

210 

94.0          ] 

1 54.1 

134.1 

253 

"3.7         3 

186.3 

162.0 

211 

94.4          i 

t54.8 

134.7 

254 

114.1         ] 

[87.1 

162.6 

^12 

94.9          3 

155.6 

135.4 

255 

1 14.6 

[87.8 

163.3 

764 


GEORGE   DEFREN. 


Parts. 

copper 

oxide. 

256 

257 
258 

259 
260 

261 

262 

263 
264 
265 
266 
267 
268 
269 
270 
271 
272 

273 

274 

275 
276 

277 
278 

279 
280 

281 

282 

283 
284 
285 
286 

287 
288 


Parts 
dextrose. 

II50 

"5-5 

1 16.0 

1 16.4 
116.9 

II7-3 
117.8 

118.3 
118.7 
1 19. 2 
119.6 

1 20. 1 
120.6 

121. 0 
121.4 
121. 9 
122.4 
122.8 

123.3 
123.7 

124.2  , 
124.6 

125. 1 
125.6 
1 26. 1 
126.5 
127.0 
127.4 
127.9 
12S.3 
128.8 

129-3 
129.7 


Parts 
lualtose. 

188.6 

189.3 
190. 1 

190.8 

191.6 

192.4 

193- 1 
193.9 
194.6 

1954 
196. 1 

196.9 

^1977 
198.4 

199.2 

199.9 
200.7 

201.5 

202.2 

203.0 

203.7 
204.5 
205.2 

206.0 
206.8 

207.5 
208.3 

209.0 

209.8 

210.5 

211.3 

212. 1 

212.8 


Parts 
lactose. 

163.9 
164.6 

165.2 

165.9 
166.5 

.167.2 

167.8 

168.5 
169. 1 
169.8 

170.4 
171. 1 
171. 7 
172.4 
173.0 

173.7 
174.4 
175.0 

175.7 
176.3 

177.0 

177.6 

178.3 
178.9 

179.6 
180.2 
180.9 
181.5 
182.2 
182.9 
183.6 
184.2 
184.9 


Parts. 

copper 

oxide. 

289 

290 

291 

292 

293 
294 

295 
296 

297 

298 

299 

300 

.301 
302 

303 
304 

305 
306 

307 
308 

309 
310 

311 
312 

313 
3M 
315 
316 

317 
318 

319 
320 


Parts 
dextrose. 


30.2 
30.6 

31. 1 

31.5 
32.0 

32.5 
33-0 
33.4 
33-9 
34.3 
34.8 

35-3 

35.7 
36.2 

36.6 

37.1 
37.6 
38.0 

38.5 
38.9 
39.4 
39-9 
40.3 
40.8 
41.2 

41.7 
42.2 

42.6 

431 
43-6 
44.0 

44.5 


Parts 
maltose. 

213.6 

214.3 
215.I 

215.9 

216.6 

217.4 
218.2 
218.9 

219.7 
220.4 

221.2 

221.9 

222.7 

223.5 
224.2 

225.0 

225.8 

226.5 

227.3 

228.1 

228.8 

229.6 

230.4 

231. 1 

231-9 
232.7 
2334 
234.2 

234.9 
235.7 
236.5 
237- 2 


Parts 
lactose 

185.6 
186.2 
186.9 
187.6 
188.2 
188.9 

189.5 
190.2 

190.8 

191.5 
192. 1 

192.8 

193.4 

194. 1 

X94.7 

195.3 
196.0 

196.6 

197.3 

197.9 
198.6 

199.3 
199.9 

200.6 

201.3 

202.0 

202.6 

203.3 

203.9 
204.6 

2053 
205.9 


SUPPLEMENTARY  T.ABLE  FOR  GLUCOSE  ANALYSIS. 

The  amounts  of  cupric  oxide  given  above  are  those  obtained 
by  the  use  of  absolute  weights  of  sugar.  The  tables  are  con- 
structed on  this  basis.  In  the  case  of  a^iixed  product,  like  com- 
mercial glucose,  which  may  be  considered  made  up  of  the  sim- 
ple bodies,  dextrin,  maltose,  and  dextrose,  it  is  far  more-  con- 
venient to  determine  the  total  carbohydrates  present  in  solution 
by  means  of  the  specific  gravity  than  by  drying  the  glucose  and 


DETERMINATION   OF   REDUCING  SUGARS.  765 

obtaining  in  this  way  the  total  solids.  For  this  purpose  an 
arbitrary  value  is  taken  which  shall  represent  the  influence  of 
one  gram  of  a  mixture  of  the  three  substances  above  mentioned 
on  the  specific  gravity  if  dissolved  to  100  cc.  in  distilled  water. 
Brown  and  Heron'  claim  that  this  influence  on  the  specific 
gravity  of  one  gram  starch  conversion  product  in  100  cc.  is 
0.00386.  This  value  has  been  determined  to  be  correct  for 
solutions  of  cane  sugar »  and  is  much  used  for  glucose  work. 

As  above  mentioned  the  specific  gravity  of  a  dextrose  solution 
containing  ten  grams  dextrose  anhydride  in  100  cc.  is  1.03809  at 
15.5*.  To  determine  the  cupric  reducing  power  of  a  substance 
using  the  value  3.86  as  a  divisor,  it  therefore  becomes  necessary 
to  change  the  figures  given  in  the  tables  to  conform  to  this  new 
factor,  that  is,  the  dextrose  equivalents  must  be  multiplied  by 


.AA  Aiaa   l;\;^ij 

1  vi\/u^  ivri 

\.vfuv^uiv:ii 

N.^      \Jl.       l.Si\\ 

.A^U^^      111 

following 

table : 

obtained. 

Dextrose 
equivalent. 

obtained. 

Dextrose 
equivalent. 

Copper 
oxide 
obtained. 

Dextrose 
equivalent. 

5 

0.4461 

no 

0.4500 

215 

0.4540 

10 

0.4463 

"5 

0.4502 

220 

0.4542 

15 

0.4465 

120 

0.4504 

225 

0.4543 

20 

0.4467 

125 

0.4506 

230 

0.4545 

25 

0.4468 

130 

0-4508 

235 

0.4547 

30 

0.4470 

135 

0.4510 

240 

0.4549 

35 

0.4472 

140 

0.4512 

245 

0.4551 

•40 

0.4474 

145 

0.4513 

250 

0.4553 

45 

0.4476 

150 

0.4515 

255 

0.4555 

50 

O.447S 

155 

0.4517 

260 

0.4557 

55 

0.4480 

160 

0.4519 

265 

0.4558 

60 

0.4482 

165 

0.4521 

270 

0.4560 

65 

0.4484 

170 

0.4523 

275 

0.4562 

70 

0.4485 

17s 

0.4525 

280 

0.4564 

75 

0.4487 

180 

0.4527 

285 

0.4566 

80 

0.4489 

185 

0.4528 

290 

0.4568 

85 

0.4491 

190 

0  4530 

29s 

0.4570 

90 

0.4493 

195 

0.4532 

300 

04572 

9S 

0.4495 

200 

0.4534 

305 

0-4574 

100 

0.4497 

205 

0.4536 

310 

0.4576 

105 

0.4498 

210 

0.4538 

315 

0.4578 

I  Loc.  cit. 

320 

0.4580 

766  JAMBS  OTIS  HANDY. 

Thus  a  solution  containing  loo  milligrams  of  mixed  carbo- 
hydrates, using  the  factor  0.00386,  if  it  formed  20d  milligrams 
cupric  oxide  by  reduction  of  the  Fehling  solution  in  the  manner 
above  described,  would  have  a  cupric  reducing  power,  or  K^^k 
of  90.68. 

MASSACMUSSTTS  IltaTITUTB  OP  Tbchnologt, 

Boston.  M Asa. 


ALUMINUn  ANALYSIS. 

B^*  Jambs  Otis  Hakot. 


ALTHOUGH  the  aluminum  industry  is  not  a  large  one 
in  the  sense  that  the  iron  industry  is,  it  is  growing 
ver>*  rapidly*  The  output  of  the  United  States  in  1894  was 
550^000  pounds,  and  in  1895  it  was  about  850,000  pounds.  The 
Pittsburg  Reduction  Company,  with  works  at  New  Kensington, 
near  Pittsburg,  Pa.,  and  at  Niagara  Falls,  N.  Y.,  is  the  only 
American  producer  of  aluminum.  The  material  is  made  by  the 
electroh^sis,  in  carbon-lined  pots,  of  alumina  dissolved  in  a  fused 
bath  of  fluorides^  The  product  of  each  pot  is  ladled  out  at  inter- 
vals and  is  graded  according  to  the  analyses  of  the  Pittsburgh 
T^s^ing  Laboratory.  Limited.  Some  of  the  aluminum  is  sold  as 
it  i$  made  and  some  is  alIo3red  to  modify  its  physical  properties. 
Allv^ys  of  aluminum  with  three  per  cent,  nickel,  or  with  three  to 
seven  per  cent,  copper,  or  similar  amounts  of  xinc  are  very  ose- 
tul  on  account  of  increased  strength  with  only  slightly  incseased 
spec^Sc  gravity.  The  aluminum  at  present  produced  with  the 
5st  vves  av^ilabue  contains  troa 

«»  to  cv;  g  rer  cent,  o:  jLl:im:aurs, 

o.  t  to  0.0^  per  cent,  o:  silkoa    corsbined  aad  crarbitic^. 

o.  ^o  to  o  o  rer  cent-  o:  cocrer, 

o  >?  to  0.0  per  cent  oc  iron. 

O—  *     iT"i.ie    ■^*'*;^'****7^    CO— tlilSS    ^'i -^^A~~.<r ^r  ^r* 

.  •.  *  » 

«      .^^..    —.»««..    >>...-vV  <^  Jl^K.^    .rv-.^       ■  .fcfc>^i 


AI^UMINUM   ANAI<YSIS.  767 

nium,  with  aluminum  ;  alumimim  solders,  containing  tin,  zinc, 
and  phosphorus ;  aluminum  hydrate,  bauxite,  and  electrode 
carbons ;  hydrofluoric  acid  and  fluorides. 

akai^ysis  op  commercial  aluminum.     (95  to  99.9  pbr  cent. 

pure). 

In  the  analysis  of  aluminum  we  are  offered  a  choice  of  sol- 
vents. 

Solubility  of  Aluminum :  Hydrochloric  acid  of  thirty-three 
per  cent.,  (/.  ^.,  one  part  of  hydrochloric  acid  of  1.2  sp.  gr.  to 
two  parts  water)  is  a  rapid  solvent. 

Sulphuric  acid  of  twenty-five  per  cent,  dissolves  aluminum 
completely  on  long  boiling. 

Nitric  acid  of  one  and  two-tenths  specific  gravity  dissolves 
aluminum  on  prolonged  boiling. 

Acid  mixture  :  A  qiixture  of  the  three  acids  which  we  term 
'  •  Acid  Mixture* '  is  made  of 

100  cc.  nitric  acid  of  1.42  sp.  gr. 

300  cc.  hydrochloric  acid  of  1.20  sp.  gr. 

600  cc.  sulphuric  acid  of  twenty-five  per  cent. 

It  is  a  very  useful  solvent  for  aluminum  because  it  supplies 
oxidizing  conditions  during  solution  and  so  prevents  loss  of 
combined  silicon  as  hydride.  The  sulphuric  acid  content  of  the 
acid  mixture  furnishes  a  means  of  rapidly  dehydrating  the  silica. 

Sodium  hydroxide  solution  of  thirty-three  per  cent,  is  a  useful 
solvent  when  it  is  desired  to  separate  the  metallic  impurities 
from  the  bulk  of  the  aluminum  at  once.  Weaker  solutions  do 
not  work  as  quickly  or  completely.  Solution  succeeds  best  in 
an  Erlenmeyer  flask. 

Fifteen  cc.  of  the  sodium  hydroxide  solution  suffice  for  one 
g^am  of  aluminum. 

Commercial  soda  lye  may  be  used  if  dissolved  and  filtered 
through  asbestos. 

OTHER  REAGENTS  AND  STANDARD  SOLUTIONS  USED   IN  AI,UMI- 

NUM   ANALYSIS. 

Sodium  carbonate,  chemically  pure. 

Soda  ash  :  *'  Solvay"  soda  ash,  a  saturated  solution  in  water, 
filtered. 


768  JAMES  OTIS  HANDT. 

Powdered  zinc  :  Practically  free  from  iron  and  copper. 

Fifteen  per  cent,  nitric  wash  :  (Fifteen  parts  1.42  nitric  acid 
to  eighty*fi\-e  parts  water). 

Standard  potassium  permanganate  :  5.76  grams  in  two  liters. 
One  cc.  equals  0.005  grams  iron. 

Standard  potassium  cyanide:  Forty-five  grams  in  two  liters. 
One  cc.  is  made  to  equal  0.005  gram  coppper. 

SPBCIAL  APPARATUS. 

Two  narrow  glass  tubes,  graduated  roughly,  one  to  hold  one 
gram  of  powdered  zinc  and  the  other  one  gram  of  chemically 
pure  sodium  carbonate. 

The  evaporating  dishes  which  are  used  are,  preferably,  about 
four  and  a  half  inches  in  diameter,  and  are  covered  with  five- 
inch  glasses. 

The  Erlenmeyer  flasks  are  of  about  twelve  ounce  capacity  and 
furnished  with  porcelain  crucible  covers. 

THH  METHOD. 

Determination,  of  Silicon^  Iron^  and  Copper  in  Comwumai 
Aluminunt. — One  gram  of  aluminum  drillings  is  dissolved  in  a 
four  and  a  half  inch  evaporating  dish  in  thirty  cc.  of  *'acid 
mixture."  If  the  drillings  are  thin  it  is  best  to  add  only  fifteen 
cc.  at  first.  Placing  cold  water  on  the  cover  glass  sometimes 
prevents  loss  from  too  energetic  foaming*.  Solution  having^ 
been  completed  by  warming  slightly,  evaporate  quickly  to  strong^ 
fumes  of  sulphuric  acid  and  continue  heating  lor  five  minutes. 
Experience  will  show  on  what  parts  of  the  hot  plate  these  solu- 
tions can  be  evaporated  without  spattering  at  the  time  when 
aluminum  sulphate  begins  to  cr>*stallize  out.  Remove  the 
dishes  from  the  plate  by  means  of  an  iron  fork,  and  in  a  fe^v 
moments  add  to  the  contents  of  each  seventy-five  to  100  cc.  of 
water  and  ten  cc.  of  twenty-five  per  cent,  sulphuric  acid,  break 
up  the  residue  in  each  dish  with  a  short,  heavy  glass  rod,  and 
place  the  dishes  back  on  the  hot  plate.  Boil  until  all  aluminum 
sulphate  dissolves.  Add  to  each  dish  one  gram  of  metallic  zinc 
powder,  measured.  Be  careful  to  pour  the  zinc  into  the  middle 
of  the  liquid.     If  it  touches  the  sides  of  the  dish  it  sometimes 


AtUMlKUM   ANAI^YSIS.  769 

becomes  firmly  fixed  there.  Keep  the  dish  contents  at6o®*to  70® 
C.  until  the  zinc  has  dissolved,  leaving  the  iron  reduced  and 
the  copper  precipitated.  Filter  and  wash  well  with  hot  water. 
Cool,  titrate  the  filtrates  with  standard  potassium  permanganate. 
One  cc.  equals  0.50  per  cent,  iron  when  one  gram  of  aluminum 
has  been  used.  Placing  new  receivers  under  the  funnels,  treat 
each  residue  twice  with  hot  fifteen  per  cent,  nitric  acid  wash. 
Wash  out  with  water,  and  in  the  solutions  thus  obtained,  titrate 
the  copper  with  standard  potassium  cyanide,  after  adding  satu- 
rated soda  ash  solution  until  the  precipitated  copper  carbonate 
redissolves.  The  end  point  of  the  titration  is  very  satisfactory. 
Th^  cyanide  solution  should  be  standardized  with  copper  of 
known  purity  in  about  the  amount  usually  found,  z^>.,  0.005  to 
o.oio  gram.  The  residue  of  silicon  and  silica  are  burned  off 
in  numbered  crucibles  and  each  fused  with  one  gram  of  chemic- 
ally pure  sodium  carbonate  (measured).  The  crucible  con- 
taining the  fused  mass  is  placed  in  fifteen  cc.  of  water  in  the  por- 
celain dish  originally  used,  and  twenty-five  cc.  of  twenty-five 
per  cent,  sulphuric  acid  are  added.  Solution  takes  place  quickly 
without  separation  of  silica,  and  after  rinsing  out  and  removing 
the  crucible,  the  solution  is  evaporated  to  five  minutes  fuming 
on  the  hot  plate.  After  cooling  add  seventy-five  to  100  cc.  of 
water  and  boil  to  disintegrate  the  silica.  Filter  and  wash  well 
with  water.  Bum  off  and  weigh  silica  and  crucible,  treat  with 
hydrofluoric  acid  and  a  drop  of  sulphuric  acid  if  impurity  is  sus- 
pected. Evaporate,  ignite,  and  weigh  again.  Loss  equals 
silica ;  calculate  to  silicon. 

DeUrminaHan  of  Crystalline  {Graphitic)  Silicon  in  Aluminum. 
— ^Dissolve  one  gram  of  aluminum  in  thirty  cc.  of  thirty-three 
per  cent,  hydrochloric  acid  (two  parts  of  water  to  one  of  hydro- 
chloric acid)  in  a  platinum  dish  ;  add  about  two  cc.  of  hydro- 
fluoric acid,  stir,  and  filter  at  once  through  a  No.  o  nine  cm. 
filter,  contained  in  a  funnel  which  has  been  thinly  coated  with 
paraffin.  Wash  with  water  and  burn  off  in  a  platinum  crucible. 
Fuse  with  one  gram  of  sodium  carbonate,  cool  in  fifteen  cc.  of 
water  in  a  four  and  a  half  inch  evaporating  dish.  Add  twenty 
cc.  of  twenty-five  per  cent,  sulphuric  acid.  Rinse  out  the  cru- 
cible.   Evaporate  to  fumes,  cool,  add  seventy-five  cc.  of  water. 


770  JAMES  OTIS  HANDY. 

boil  up  and  filter  o£F  the  silica.      Wash,  ignite,  and  weigh. 
Calculate  to  silicon. 

The  determinations  of  silicon,  copper,  and  iron  are  the  every 
day  methods  of  grading-  aluminum.  It  is  recog^zed  that 
sodium  and  carbon  occasionally  exist  in  aluminum,  and  they  are 
determined  by  methods  described.  In  certain  samples  it  is 
desirable  to  know  the  approximate  percentage  of  graphitic  and 
combined  silicon.  These  determinations  are  also  described. 
We  determine  nitrogen,  if  present,  by  a  special  method. 

DETERMINATION  OF  SODIUM  IN  ALUMINUM. 

One  gram  of  drillings  is  dissolved  in  a  porcelain  evaporating 
dish  in  fifty  cc.  of  1.3  sp.  gr.  nitric  acid  and  sufficient  hydro- 
chloric acid  to  effect  solution.  Boil  down  until  all  hydrochloric 
acid  has  been  removed.  Rinse  the  solution  into  a  large  plati- 
num dish  and  evaporate  to  dryness.  Heat  over  a  good  Bunsen 
burner  until  nitric  oxide  fumes  cease  to  be  evolved.  Grind  the 
residue  finely.  Mix  it  by  grinding  with  one  gram  of  chemically 
pure  ammonium  chloride  and  eight  g^msof  chemically  pure  cal- 
cium carbonate.  Heat  the  mixture  in  a  large  covered  platinum 
crucible.  For  the  first  fifteen  minutes  have  a  Bunsen  burner 
flame  just  touching  the  bottom  of  the  crucible,  and  for  the  next 
forty-five  minutes,  have  the  whole  crucible  heated  bright  red  by 
a  full  Bunsen  burner  fiame.  Cool,  and  treat  the  residue  with 
hot,  distilled  water  until  it  becomes  just  friable  under  pressure. 
Avoid  adding  an  excess  of  water  beyond  that  necessary  to  make 
the  sintered  mass  just  friable.  Grind  it  in  a  wedgewood  mor- 
tar and  rinse  out  with  hot  distilled  water.  Filter,  rejecting  the 
well  washed  residue,  and  treat  the  filtrate  at  the  room  tempera- 
ture with  saturated  ammonium  carbonate  solution  in  slight 
excess.  Stir  very  thoroughly.  The  precipitated  calcfum  carbon- 
ate is  at  first  fiocculent,  but  on  standing  for  about  ten  minutes, 
it  becomes  crystalline.  Filter  into  a  platinum  dish  ;  reject  the 
residue  and  evaporate  the  solution  on  the  water-bath  to  dryness. 
Heat  carefully  to  dull  redness  to  expel  ammonium  salts.  Dis- 
solve the  residue  in  a  little  water  and  add  a  few  drops  of  ammo- 
nium carbonate  solution.  If  this  produces  a  precipitate,  add 
sufficient  ammonium  carbonate  solution  to  precipitate  all  of  the 


ALUMINUM   ANALYSIS.  77 1 

remaining  lime.  Stir  well,  wait  ten  minutes,  filter,  evaporate  to 
dryness,  heat  to  dull  redness,  and  weigh  sodium  chloride. 
Deduct  any  sodium  chloride  found  in  a  blank  determination, 
using  acids,  etc.,  as  above,  and  finally  eight  grams  of  calcium 
carbonate  and  one  gram  of  ammonium  chloride. 

NaCl  X  0.39316  =  Na. 

Care  should  be  taken  when  heating  up  the  residue  of  sodium 
chloride,  etc.,  after  evaporating  on  the  water-bath.  If  the  plati- 
num dish  an4  contents  are  heated  for  a  few  minutes  on  sheet 
asbestos  on  the  hot  plate  before  placing  over  the  lamp,  spatter- 
ing may  be  avoided.  Sodium  is  generally  absent  from  alumi- 
num, but  it  has  been  found  in  amounts  as  high  as  0.20  per  cent. 
and  is  considered  a  cause  of  the  occasional  deterioration  of  the 
metal  in  water. 

DETERMINATION  OF  CARBON  IN  ALUMINUM.   (  MOI^SAN'S  METHOD 

MODIFIED.) 

Triturate  two  grams  of  drillings  in  a  Wedgewood  mortar  with 
ten  to  fifteen  grams  of  mercuric  chloride,  powdered  and  dissolved, 
or  partly  dissolved,  in  about  fifteen  cc.  of  water.  Reaction  takes 
place  rapidly  and  a  heavy  gray  residue  is  left.  Persistent  tritu- 
ration removes  the  last  particles  of  metallic  aluminum.  Evapo- 
rate on  the  water-bath  to  dryness.  The  dry  residue  is  heated  in 
a  current  of  pure  hydrogen  to  expel  mercuric  compounds.  The 
remaining  material  is  then  placed  in  a  boat  in  a  combustion  tube 
and  burned  off  as  in  carbon  determination  in  steel.  The  carbon 
dioxide  is  caught  ^s  barium  carbonate,  and  the  excess  of  barium 
hydroxide  determined  by  means  of  standard  acid.  We  are  work- 
ing on  a  more  generally  applicable  method  for  carbon  in  alumi- 
num. 

DETERMINATION  OF  NITROGEN  IN  ALUMINUM. 

Aluminum,  when  overheated  in  re- melting,  is  believed  to  have 
the  property  of  combining  with  nitrogen.  The  metal  becomes 
weaker.  Moissan's  method  for  determining  nitrogen  in  alumi- 
num may  be  found  in  Compt.  Rend,,  119,  12.  Nitrogen  thus 
absorbed  would  undoubtedly  exist  as  nitride  of  aluminum  and 
solution  of    sodium    hydroxide    with    subsequent    distillation 


772  JAMES  OTIS   HANDY. 

would  seem  to  be  the  best  method  of  procedure.     We  are  work- 
ing up  this  method. 

DETERMINATION  OF  ALUMINUM  IN  METALLIC  ALUMINUM. 

Dissolve  one  gram  of  metal  in  thirty  cc.  of  thirty-three  per 
cent,  hydrochloric  acid  in  a  porcelain  dish  and  evaporated  cau- 
tiously to  complete  dr>'ness.  Redissolve,  by  boiling,  with  ten 
cc.  of  concentrated  hydrochloric  acid  and  seventv-five  cc.  of 
water.  Wash  into  a  twelve  ounce  beaker;  dilute  to  250  cc. 
and  pass  hydrogen  sulphide  until  saturated.  Filter  into  a  beaker 
and  boil  off  hydrogen  sulphide.  Oxidize  by  adding  one  cc.  of 
concentrated  nitric  acid  and  continuing  to  boil  for  ten  minutes. 
Cool  and  make  the  solution  up  to  500  cc.  '  Remove  fifty  cc.  of 
the  solution,  and  having  diluted  to  250  cc.  and  heated  to  boil- 
ing, add  ammonium  hydroxide  in  slight  excess  and  boil  well  for 
twenty  minutes.  Let  settle  ;  filter,  and  wash  thoroughly  with 
hot  water.  It  is  necessary  to  wash  the  precipitate  off  from  the 
filter,  break  it  up,  and  wash  it  back  again.  Finally  bum  off  in 
a  thin-walled  platinum  crucible,  igniting  most-  intensely,  and 
weighing  the  instant  the  crucible  and  content  are  cool.  We 
have  found  that  alumina  is  one  of  the  most  difficult  oxides  to 
dehydrate  completely,  and  when  dehydrated  it  absorbs  atmos- 
pheric moisture  even  more  rapidly  than  calcium  oxide  does. 
Moissan  prefers  to  precipitate  aluminum  by  ammonium  sul- 
phide. Having  prepared  a  solution  in  hydrochloric  acid,  he 
takes  an  amount  equal  to  0.15  gram  of  aluminum,  neutralizes  it 
in  the  cold  with  ammonia,  and  precipitates  it  by  ammonium  sul- 
phide, which  has  been  recently  prepared.  He  then  digests  for 
one  hour,  filters,  washes  with  hot  water,  ignites  and  weighs. 

ANALYSIS   OF   ALLOYS  OF    ALUMINUM   WITH  SMALLER  AMOUNTS 

OF  OTHER  METALS. 

Copper  Alloys. — Three  to  thirty  per  cent,  copper,  and  no  zinc 
or  nickel. 

Dissolve  one-half  gram  or  one  gram  in  fifteen  cc,  of  thirty- 
three  per  cent,  sodium  hydroxide  solution  in  an  Erlenmeyer 
flask  of  twelve  ounce  capacity.  If  the  flask  is  covered  and  set 
in  a  warm  place,  solution  is  complete  in  a  few  minutes,   even  if 


ALUMINUM   ANALYSIS.  773 

the  drillings  are  quite  coarse.  Dilute  to  thirty  cc.  with  hot 
water  and  filter  through  a  coarse,  lintless  filter  paper  (Whitall, 
Tatum  &  Co.'s  five  inch).  Wash  well  with  hot  water.  Dis- 
solve residue,  atter  washing  it  off  the  filter  paper  into  a  twelve 
ounce  beaker,  by  warming  with  five  cc.  of  concentrated  nitric  acid. 
Cool,  add  saturated  commercial  sodium  carbonate  solution  until 
re-solution  occurs.  Titrate  with  standard  potassium  cyanide 
solution  to  the  disappearance  of  the  b\ue  color.  Standardize  the 
cyanide  for  about  the  same  amount  of  copper. 

For  commercial  reasons,  twenty  per  cent,  alloys  are  made  in 
the  reduction  pots,  and  these  alloys  are  subsequently  used  for 
making  copper  allo3's  of  low  percentage. 

DETEKMINATION  OF  NICKEL  IN  ALUMINUM  ALLOYS. 

The  three  per  cent,  nickel  alloy  is  now  used.  The  addition  of 
three  per  cent,  of  nickel  increases  the  tensile  strength  of  alumi- 
num by  several  thousand  pounds  per  square  inch. 

One  gram  of  drillings  is  dissolved  in  fifteen  cc.  of  thirty-three 
per  cent,  sodium  hydroxide  solution  in  a  twelve  ounce  Erlen- 
meyer  flask.  Dilute  to  fifty  cc.  and  filter  through  a  five-inch 
lintless  paper,  washing  the  residue  thoroughly  with  hot  water. 
Rinse  the  residue  back  into  the  flask  and  add  three  to  five  cc.  of 
concentrated  nitric  acid,  and  a  drop  of  concentrated  hydrochloric 
acid.  Boil,  and  when  dissolved,  cool,  and  make  up  to  250  cc. 
In  100  cc.  determine  the  copper  by  neutralizing  with  ammonia, 
adding  two  cc.  of  cot^centrated  hydrochloric  acid,  warming  and 
passing  hydrogen  sulphide.  Filter  and  wash  with  ammonium  sul- 
phide. Bum  it  off  carefully  in  a  porcelain  crucible,  and  having 
weighed,  dissolve  in  five  cc.  of  concentrated  nitric  acid.  Then 
dilute  to  twenty  cc,  add  excess  of  sodium  carbonate  solution 
and  titrate  with  standard  potassium  cyanide.  Boil  the  filtrate 
from  the  cupric  sulphide,  oxidize  with  one  cc.  of  nitric  acid,  and 
precipitate  with  ammonium  hydroxide.  Do  not  boil,  but  digest 
for  a  few  minutes  just  below  the  boiling  point.  Filter,  wash,  re- 
dissolve  in  hot  fifteen  per  cent,  nitric  acid  wash.  Dilute  to  150 
cc.  and  again  precipitate  with  excess  of  ammonium  hydroxide, 
being  careful  to  avoid  boiling  or  prolonged  digestion.  Filter 
and  wash.     Bum  off  and  weigh  ferric  oxide,  etc.     In  a  second 


774  JAMES  OTIS  HANDY. 

loo  cc.  of  the  main  solution,  precipitate  nickel  hydroxide,  cnpric 
oxide,  ferric  hydroxide,  etc.,  by  thirty-three  per  cent,  chemically 
pure  sodium  hydroxide  solution,  added  in  slight  excess  to  the 
boiling  solution.  Boil  for  fifteen  minutes,  filter,  and  wash  most 
thoroughly  with  hot  water.  Bum  off  and  weigh  nickel  oxide, 
cnpric  oxide  and  ferric  oxide.  Deduct  cupric  oxide  and  ferric 
oxide  already  found.     Calculate  nickel  oxide  to  metallic  nickeL 

ANALYSIS  OP  ALUMINT7M-MANGANESB  ALLOYS. 

DeierminaHon  of  Manganese. — Place  one  gram  of  drillings  in. 
a  twelve  ounce  beaker.  Add  thirty  cc.  of  thirty-three  per  cent, 
hydrochloric  acid  (one  part  of  concentrated  hydrochloric  acid  to 
two  of  water).  When  dissolved,  add  twenty-five  cc.  of  nitric  acid 
( 1 .42 ) ,  and  boil  down  to  ten  cc.  Add  fifty  cc.  of  colorless  nitric 
acid  (1.42)  and  boil.  Precipitate  the  manganese  with  pow- 
dered potassium  chlorate,  added  cautiously,  and  proceed  as 
described  under  manganese  in  steel  by  Williams'  method.* 

ANALYSIS  OF  CHROMIUM-ALUMINUM   ALLOY. 

Determination  0/ Chromium, — Dissolve  one  gram  in  a  twelve 
ounce  beaker  in  thirty  cc.  of  thirty-three  per  cent,  hydrochloric 
acid,  and  when  dissolved  add  fifty  cc.  of  sulphuric  acid  (1.84), 
and  evaporate  carefully  until  fumes  of  sulphur  trioxide  escape* 
Cool,  add  sixty  cc.  of  water  and  boil.  After  five  minutes,  if  all 
aluminum  sulphate  has  been  dissolved,  add  powdered  potassium 
permanganate  until  the  solution  has  a  distinct  pink  color.  Boil 
until  the  excess  of  potassium  permanganate  is  decomposed.  Filter 
through  washed  asbestos  and  determine  the  chromium  in  the 
filtrate  as  in  chrome  steel.' 

ANALYSIS  OF  TUNGSTEN- ALUMINUM  ALLOY. 

Determination  of  Tungsten, — Dissolve  one  gram  in  thirty- 
three  per  cent,  hydrochloric  acid  in  a  four  and  a  half  inch  evap- 
orating dish.  Add  thirty  cc.  of  nitric  acid  (1 .42)  and  evaporate  to- 
dryness.  Redissolve  in  thirty  cc.  of  hydrochloric  acid  (1.20), 
dilute  to  about  ninety  cc.,  and  boil  for  two  hoars.  Filter  and 
wash  thoroughly.  Bum  off  and  weigh  Si  +  SiO,  +  WO,  + 
crucible.     Treat  with  three  drops  of  twenty-five  per  cent,  sul- 


1  BUir's  ''Chemical  Analysis  of  I 

s  Galbraith's  MeUiod.    See  BUir's  '*  Chemical  Aaal3rsis  of  I 


AI^UMINUM   ANAI^YSIS.  775 

phuric  acid  and  about  two  cc.  of  hydrochloric  acid.  Evaporate 
carefully  over  an  Argand  burner,  re-ignite,  and  weigh  crucible 
and  silicon  and  tungstic  oxide.  Fuse  with  one  gram  of  sodium 
carbonate,  cool,  place  in  dish,  and  add  fifteen  cc.  of  water  and 
twenty  cc.  of  twenty-five  per  cent,  sulphuric  acid,  remove  cruci- 
ble and  evaporate  until  white  fumes  escape.  Cool,  redissolve 
in  about  fifty  cc.  of  water.  Filter,  wash,  ignite,  and  weigh 
silica  (from  silicon),  tungstic  oxide,  and  crucible.  Treat  with 
sulphuric  acid  and  hydrofluoric  acid,  evaporate,  ignite,  and 
reweigh.  Loss  equals  silica.  Calculate  to  silicon  and  add  to 
the  weight  of  silica  lost  by  treatment  of  first  insoluble  residue. 
Deduct  this  sum  from  the  weight  of  silicon,  silica,  and  tungstic 
oxide  first  found  and  the  remainder  equals  tungstic  oxide. 
Calculate  to  tungsten. 

ANALYSIS  OF  AI^UMINUM-TITANIUM  ALLOY. 

Determination  of  Titanium, — Two  grams  of  the  alloy  in  a 
twelve  ounce  Erlenmeyer  flask  are  dissolved  by  addition  of  fifty 
cc.  of  ten  per  cent,  potash  solution.  Dilute  with  distilled  water 
to  about  125  cc,  boil  up,  and  filter  as  quickly  as  possible. 
Wash  ten  times  with  boiling  water.  Bum  off  the  residue  in  a 
porcelain  crucible,  crush  it  in  a  Wedgwood  mortar,  fuse  in  a 
large  platinum  crucible  with  ten  grams  of  potassium  bisulphate. 
Conduct  the  fusion  exactly  as  follows  :  Choose  a  good  Bunsen 
burner,  and  protect  it  from  draught  by  a  sheet-iron  chimney  ; 
make  the  flame  one  and  a  half  inches  long,  and  place  the 
triangle  carrying  the  upright  crucible  just  at  the  point  of  the 
flame.  Increase  the  heat  gradually  until  in  ten  minutes  the 
lower  fourth  of  the  crucible  is  red  hot.  Allow  it  to  remain  at 
this  temperature  ten  minutes,  moving  the  lid  slightly  to  one  side 
every  two  minutes,  and  giving  the  crucible  (held  firmly  in  the 
tongs)  a  gentle  rotating  movement,  then  turn  up  the  light  until 
the  flame  reaches  the  top  of  the  crucible  and  envelopes  it.  Five 
minutes  of  this  treatment  melts  down  any  potassium  bisulphate, 
etc.,  which  have  risen  on  the  sides.  The  flame  is  lowered  and  the 
lower  fourth  heated  for  ten  minutes  longer.  Cool,  dissolve  in 
about  200  cc.  of  water ;  filter,  rejecting  the  residue,  if  ignition 
and  treatment  with  hydrofluoric  acid  show  it  to  be  only  silica. 


^^6  JAMBS  OTIS   HANDY. 

If  it  contains  anything  more,  fuse  with  four  grams  of  potassium 
bisulphate  again.  The  filtrate  contains  all  the  titanic  oxide 
and  the  ferric  oxide.  Add  ammonia  until  a  slight  perma- 
nent precipitate  is  formed,  then  add  dilute  sulphuric  acid  from  a 
pipette  or  burette  until  this  precipitate  just  redissolves.  Finally 
add  one  cc.  more  of  twenty-five  per  cent,  sulphuric  acid  and 
dilute  to  300  cc.  If  the  solution  is  high  in  iron  (which  will  be 
indicated  by  its  distinct  yellow  color)  sulphur  dioxide  gas 
must  be  run  into  it  until  it  is  decolorized  and  smells  strongly 
of  sulphur  dioxide,  but  if  the  solution  is  nearly  colorless,  indi- 
cating a  low  percentage  of  iron,  only  sulphur  dioxide  water 
need  be  added  for  the  reduction.  Boil  well  for  one  hour,  adding 
water  saturated  with  sulphur  dioxide  occasionally.  Filter  o£F 
the  titanic  oxide  through  double  filters  and  wash  well  with  hot 
water.  Bum  off  and  weigh  as  titanic  oxide.  If  the  precipitate 
is  yellow,  indicating  the  presence  of  iron,  it  may  be  fused  with 
one  gram  of  potassium  bisulphate,  the  fusion  dissolved  in  ten  cc. 
of  dilute  sulphuric  acid,  and  the  iron  determined  in  this  solu- 
tion by  reducing  with  one  gram  of  zinc,  and  titrating  with  per- 
manganate. This  is  not  often  necessary.  Calculate  titanic 
oxide  to  titanium.     TiO,  X  0.6  =  Ti. 

DETERMIN.\TION    OP    ZINC    IN   ZINC- ALUMINUM  AI.LOT.      FIRST 

METHOD. 

Dissolve  one  gram  in  thirt>'  cc.  of  thirty-three  per  cent,  hydro- 
chloric acid  in  a  twelve  ounce  beaker.  Dilute  to  200  cc.  and 
heat  nearly  to  boiling.  Pass  hydrogen  sulphide  till  all  copper 
is  precipitated.  Filter  and  boil  off  hydrogen  sulphide,  oxidize 
with  one  cc.  nitric  acid  by  boiling  ten  minutes.  Add  sodium 
hydroxide  solution  until  neutral,  then  make  barely  acid  with 
hydrochloric  acid,  and  stir  until  the  aluminum  hydroxide  all 
dissolves.  Add  ten  grams  of  sodium  acetate  and  500  cc.  of 
water,  boil  up,  and  filter  at  once.  Dissolve  the  washed  precipi- 
tate in  hydrochloric  acid  and  repeat  the  acetate  separation. 
Heat  the  united  filtrates  to  boiling  and  pass  hydrogen  sulphide. 
Filter  off  the  zinc  sulphide  on  double  filters,  wash  thoroaghly 
with  hot  water.  Bum  off  in  a  porcelain  crucible,  and  weigh  zinc 
oxide.     Calculate  to  zinc.     This  method  may  be  used  when 


ALUMINUM   ANALYSIS.  777 

only  a  small  quantity  of  the  sample  is  available,  but  when  this 
is  not  the  case,  it  is  better  to  use  the  method  given  below. 

DETERMINATION    OF    ZINC    IN    ZINC-ALUMINUM    ALLOYS. 

SECOND  METHOD. 

Dissolve  one  gram  of  drillings  in  thirty-three  per  cent,  sodium 
hydroxide  solution  in  a  twelve  ounce  Erlenmeyer  flask.  Filter 
as  soon  as  dissolved  through  a  four  inch  lintless  filter  paper. 
Wash  thoroughly  with 'hot  water.  Rinse  the  residue  of  zinc, 
iron,  copper,  silicon,  etc. ,  back  into  the  flask.  This  may  require 
25  cc.  of  water.  Add  .five  cc.  of  hydrochloric  acid  and  boil. 
Dilute  to  150  cc.  with  hot  water  and  pass  hydrogen  sulphide. 
Filter  and  boil  off  hydrogen  sulphide,  reoxidize  by  adding  one 
cc.  nitric  acid  and  boiling  ten  minutes.  Add  sodium  hydroxide 
till  neutral,  then  add  dilute  hydrochloric  acid  till  just  acid,  and 
then  ten  grams  of  sodium  acetate,  and  300  cc.  of  boiling  water, 
and  boil  for  five  minutes.  Wash  well.  If  the  precipitate  is  small, 
resolution  and  re  precipitation  are  not  necessary.  Pass  hydro- 
gen sulphide  through  the  filtrate.  Filter  off  zinc  sulphide 
through  double  filters.  Wash  well.  Ignite  in  a  porcelain  cru- 
cible, heating  finally  over  the  blast,  to  zinc  oxide.  ZnO  X 
0.8032  =  Zn. 

ANALYSIS  OF  ALUMINUM  SOLDERS. 

Determination  of  Tin,  Phosphorus^  and  Zinc, — Aluminum  sold- 
ers generally  contain  phosphor-tin,  and  zinc.  As  presented  for 
analysis,  they  usually  consist  of  a  soldered  joint,  from  which 
the  solder  must  be  scraped  and  analyzed.  The  analysis,  there- 
fore, involves  a  separation  of  the  elements  aluminum,  zinc,  tin, 
and  phosphorus.  It  is  a  difficult  matter  to  determine  whether 
aluminum  was  a  constituent  of  the  solder  when  only  a  soldered 
joint  is  available  for  examination.  It  is  best  to  dissolve  all 
adhering  aluminum  from  the  pieces  chosen  for  analysis  by 
treatment  with  thirty-three  per  cent,  sodium  hydroxide  solution 
after  which  the  residue  is  filtered  off,  dried,  and  weighed  out 
for  analysis.  Dissolve  or  decompose  three-tenths  to  five-tenths 
gram  in  a  twelve-ounce  beaker  by  means  of  twenty  cc.  of  nitric 
acid  (1.42).  If  necessary,  five  cc.  of  hydrochloric  acid  (1.2) 
may  be  used  to  effect  complete  decomposition.      Evaporate  to 


778  JAMBS  OTIS  HANDY. 

complete  dryness  on  a  hot  plate.  Cool,  add  twenty-five  cc.  of 
nitric  acid  (1.13),  and  boil  thoroughly.  Filter.  The  residue 
contains  all  of  the  tin,  most  of  the  phosphdrus,  and  possibly 
some  zinc.  Bum  it  off  in  a  porcelain  crucible  and,  after  pul- 
verizing the  residue  in  an  agate  mortar,  mix  it  with  two  grams 
of  sodium  carbonate  and  two  grams  of  sulphur,  fuse  it  in  a  cov- 
ered porcelain  crucible  over  a  Bunsen  burner  for  about  half 
an  hour.  Give  it  three  minutes  of  gentle  blast  flame  at  the  last. 
Cool,  boil  .out  with  150  cc.  of  water  in  a  twelve-ounce  covered 
beaker.  Filter  and  wash.  Extract  any  possible  zinc  sulphide, 
etc.,  from  the  residue,  by  dissolving  in  nitric  acid,  boiling  off 
hydrogen  sulphide,  and  adding  this  to  the  first  filtrate 
obtained  after  evaporating  to  dryness  with  nitric  acid.  The 
sodium  sulphide  solution  contains  the  tin  and  phosphorus.  Add 
to  it  hydrochloric  acid  until  just  acid.  Warm  slightly  and  pass 
hydrogen  sulphide.  Filter  off  stannous  sulphide  and  wash 
thoroughly  with  hot  water.  Bum  off  in  a  porcelain  crucible 
and  weigh  stannic  oxide.  Calculate  to  metallic  tin.  The 
filtrate  from  the  stannous  sulphide  is  boiled  to  expel  hydrogen 
sulphide  and  then  oxidized  by  adding  two  cc.  of  nitric  acid  and 
boiling  for  fifteen  minutes  more.  Filter  off  any  sulphur  which 
separates,  and  in  this  filtrate,  which  should  amount  to  only 
about  100  cc,  precipitate  the  phosphorus  by  adding  pure 
sodium  hydroxide  solution  till  alkaline,  then  nitric  acid  till  dis- 
tinctly acid,  heating  to  85®  C,  and  adding  fifty  cc.  of  filtered 
molybdate  solution.  Stir  or  shake  well  for  five  minutes,  filter 
on  a  weighed  filter  paper,  and  after  washing  with  one  per  cent, 
nitric  acid  wash,  dry  at  100*  C.  and  weigh.  Yellow  precipitate 
multiplied  by  0.0163  equals  phosphorus.  The  nitric  acid  solu- 
tion obtained  after  evaporating  the  first  solution  to  dryness,  etc.^ 
is  now  neutralized  with  sodium  hydroxide  solution,  and  then 
made  just  acid  with  hydrochloric  acid.  Ten  grams  of  sodium 
acetate  are  now  added,  and  300  cc.  of  water  (hot) .  Boil  up  for  five 
minutes,  then  filter  and  wash.  If  the  precipitate  is  of  consider- 
able size,  it  is  probable  that  aluminum  was  a  constituent  of  the 
solder.  Redissolve  it  in  a  little  hydrochloric  acid,  neutralize, 
acidify,  and  make  a  basic  acetate  separation  as  before.  Precipi- 
tate the  zinc  in  the  acetate  solutions  by  hydrogen  sulphide. 


ALUMINUM   ANALYSIS.  779 

Filter,  wash,  ignite  in  a  porcelain  crucible,  and  weigh  as  zinc 
oxide.  Calculate  to  metallic  zinc.  Dissolve  the  precipitate  of 
aluminum  acetate  in  hydrochloric  acid,  dilute  to  250  cc,  and 
precipitate  with  ammonia.  After  filtering,  washing,  igniting, 
and  weighing  as  alumina,  calculate  to  metallic  aluminum. 
Solders  containing  lead  are  sometimes  met  with.  In  such  cases, 
evaporate  the  nitric  acid  filtrate  from  the  metastannic  acid  to 
small  bulk,  add  twenty-five  cc.  of  twenty-five  per  cent,  sul- 
phuric acid,  and  evaporate  until  white  fumes  escape.  Cool,  add 
100  cc.  of  water,  stir,  and  let  stand  for  an  hour  in  a  warm  place. 
Filter  and  wash  with  water  containing  five  per  cent,  of  sulphuric 
acid .  Bum  off  in  a  porcelain  crucible  at  a  low  temperature.  Reoxi- 
dize  any  reduced  lead  oxide  and  restore  its  sulphur  trioxide  by 
adding  a  few  drops  of  nitric  acid  and  sulphuric  acid  and  evapo- 
rating.  Finally  weigh  lead  sulphate.   Calculate  to  metallic  lead. 

Zinc  is  determined  in  the  lead  sulphate  filtrate. 

« 

ANALYSIS  OF   ALUMINA. 

Alumina  is  made  from  bauxite  or  cryolite.  It  is  usually  pur- 
chased in  the  hydrated  form. 

HYDRATED   ALUMINA. 

Hydrated  alumina  is  analyzed  for  water,  silica  and  sodium 
carbonate. 

Water, — Ignite  one  gram  in  a  closely  covered  crucible,  at  first 
gently  and  then  intensely  for  fifteen  minutes  over  the  strongest 
blast.  The  loss  on  ignition  includes  water  and  the  carbon 
dioxide  of  the  sodium  carbonate.  Calculate  the  carbon  dioxide 
from  the  sodium  oxide  found  and  deduct  it  from  the  loss  on 
ignition. 

Silica. — Hydrated  alumina  is  soluble  in  sulphuric  acid  of  42^ 
B.  The  silica,  however,  is  left  undissolved.  42**  B.  sul- 
phuric acid  is  made  by  mixing  900  cc.  of  concentrated  sul- 
phuric acid  with  1290  cc.  of  water.  Five  grams  of  hydrated 
alumina  are  treated  with  twenty-five  cc,  of  42®  B.  sul- 
phuric acid  and  heated  until  the  alumina  appears  to  be  all  dis- 
solved. Dilute  to  100  cc.  and  boil.  Filter,  wash,  ignite  and 
fuse  the  residue  with  one  gram  of  potassium  bisulphate  and 
cool.     Dissolve  in  water,  filter,  wash,  ignite,  and  weigh  in  cru- 


7^0  JAMES  OTIS   HANDY. 

cible,  treat  with  sulphuric  acid  and  hydrofluoric  acid,  evaporate, 
ignite  and  weigh  again.     Loss  equals  silica. 

Soda. — The  method  for  the  determination  of  soda  is  the  same 
in  calcined  and  hydrated  alumina.  The  method  is  that  of  J.  L- 
Smith,  and  is  described  under  **  Sodium  in  Aluminum."  Cal- 
culate sodium  chloride  to  sodium  carbonate,  if  the  sample  is 
hydrated,  and  to  sodium  oxide  if  the  sample  is  calcined  alumina. 

CALCINED   ALUMINA. 

Water  and  soda  are  determined  as  in  hydrated  alumina. 

Silica, — Fuse  one  gram  of  the  finely  ground  alumina  with  ten 
grams  of  potassium  bisulphate.  If  this  does  not  make  a  clear 
fusion  add  two  g^ams  of  bisulphate  and  heat  up  again.  Dissolve 
the  fusion  when  cool  in  water  and  filter.  Bum  off  the  insoluble 
residue.  Fuse  it  with  one  gram  of  sodium  carbonate  and  cool 
in  fifteen  cc.  of  water  in  a  four  and  a  half  inch  evaporating  dish. 
Add  twenty-five  cc.  of  twenty-five  per  cent,  sulphuric  acid. 
When  all  soluble  matter  has  dissolved,  remove  the  crucible  and 
evaporate  down  until  sulphuric  acid  kimes  escape.  Cool,  dilute 
with  water,  boil,  filter,  ignite,  and  weigh  silica  plus  crucible, 
treat  with  sulphuric  and  hydrofluoric  acids,  and  weigh  again. 
Loss  equals  silica. 

ANALYSIS  OF   BAUXITE. 
Method  adopted,  May,  1895. 

No  unusual  apparatus  or  reagents  are  required. 

One  and  five- tenths  grams  of  very  finely  ground  bauxite  (pre- 
viously dried  at  loo**  C.  and  bottled),  is  taken  for  analysis. 
Weigh  into  a  five  inch  porcelain  evaporating  dish  and  dissolve 
in  fifty  cc.  of  acid  mixture.  This  mixture  is  the  same  as  that 
used  for  aluminum  analysis.  Boil  the  solution  down  until  fumes 
escape  and  keep  the  residue  fuming  strongly  for  about  fifteen  min- 
utes. Cool,  add  100  cc.  of  water,  stir  and  then  boil  for  ten  minutes. 
Filter,  wash  well  with  water,  receiving  the  filtrate  in  a  beaker 
of  about  300  cc.  capacity.  The  filtrate  and  washings  should 
amount  to  about  175  cc.  Bum  off  the  insoluble  residue  (which 
consists  chiefly  of  silica,  with  a  little  titanic  acid,  oxide  of  iron, 
and  alumina)  and  weigh  it  in  the  crucible,  add  three  drops  of 
twenty-five  per  cent,  sulphuric  acid  and  about  five  cc.  of  hydro- 


ALUMINUM  ANALYSIS.  781 

fluoric  acid  and  evaporate  slowly  to  dryness.  Ignite  very 
strongly  and  weigh.  The  loss  in  weight  equals  silica.  Add  to 
the  residue  in  the  crucible  one  gram  of  potassium  bisulphate  and 
fuse  quickly  and  thoroughly  over  a  Bunsen  burner,  cool  and  place 
the  crucible  in  the  beaker  containing  the  main  sulphuric  acid 
solution.  The  small  residue  from  this  fusion  will  be  silica,  and 
is  to  be  added  to  the  silica  already  found.  Having  obtained  the 
sulphate  solution  containing  all  the  alumina,  ferric  oxide  and 
titanic  oxide,  make  it  up  to  550  cc.  and  mix.  Then  fifty  cc. 
will  equal  three-tenths  gram  bauxite.  Take  fifty  cc.  and  dilute 
to  300  cc.  Add  two  cc.  of  concentrated  hydrochloric  acid  and 
ammonia  in  slight  excess,  boil  for  five  minutes,  let  the  precipi- 
tate settle,  filter  and  wash  very  thoroughly  with  hot  water.  Test 
the  filtrate  for  further  alumina  by  boiling.  Bum  off  the  filter 
paper  and  ignite  the  precipitate  very  strongly  after  crushing  all 
the  lumps  of  alumina.  Weigh  alumina,  ferric  oxide  and  titanic 
oxide. 

Titanic  Acid, — Take  100  cc.  of  the  original  sulphate  solution 
(six-tenths  gram),  add  ammonia  until  a  slight  permanent  pre- 
cipitate is  formed,  then  add  sulphuric  acid  from  a  pipette  or 
burette  until  this  precipitate  just  redissolves.  Finally  add  one 
cc.  more  of  twenty-five  per  cent,  sulphuric  acid  and  dilute  to 
400  cc.  If  the  bauxite  is  high  in  iron  (which  will  be  indicated 
by  the  distinct  yellow  color  of  this  solution)  sulphur  dioxide  gas 
must  be  run  into  it  until  it  is  decolorized  and  smells  strongly  of 
sulphur  dioxide,  but  if  the  solution  is  nearly  colorless,  indicat- 
ing a  low  percentage  of  iron,  only  sulphur  dioxide  water  need 
be  used  for  the  reduction.  Boil  well  for  one  hour,  adding  water 
saturated  with  sulphur  dioxide  occasionally.  Filter  off  the 
titanic  oxide  through  double  filters  and  wash  well  with  hot  water. 
If  the  precipitate  is  yellow,  indicating  the  presence  of  iron,  it 
can  be  fused  with  one  gram  of  potassium  bisulphate,  the  fusion 
dissolved  in  water,  and  the  iron  determined  in  this  solution  by 
reducing  with  zinc  and  titrating  with  permanganate.  This  is 
not  often  necessary. 

Oxide  of  Iran, — Take  fifty  cc.  of  the  sulphate  solution,  add  ten 
cc.  of  dilute  sulphuric  acid  and  one  gram  of  granulated  zinc, 


782  CHARLBS  GI«ASBR. 

and  set  the  beaker  in  a  warm  place.  When  reduced,  filter  and 
titrate  the  iron  with  standard  potassium  permanganate.  More 
zinc  is  used  for  bauxites  high  in  iron. 

METHOD  FOR  IRON   DBTBRMINATION,  USING  A    LARGBR   QUAN- 
TITY OP  BAUXITB.      (APPI,ICABI,B  TO  PURBST  ORBS). 

Place  a  half  gram  of  the  finely  powdered  ore  in  a  large  plati- 
num crucible  and  add  three  cc.  of  twenty-five  per  cent,  sulphuric 
acid  and  five  cc.  of  hydrochloric  acid,  and  evaporate  very  slowly 
to  fumes ;  drive  oS  the  excess  of  sulphuric  acid  by  heat,  boil  out 
the  residue  with  water  and  add  ten  cc.  of  dilute  sulphuric  acid. 
Remove  the  crucible  and  reduce  with  zinc,  as  above,  and  titrate. 

WdUr  and  Organic  Matter. — Ignite  three-tenths  gram,  cau- 
tiously at  first  and  finally  very  strongly  in  a  covered  crucible. 
The  loss  of.  weight  equals  water  and  organic  matter. 


ESTIHATION  OF  THORIA.    CHEMICAL  ANALYSIS  OF 

nONAZITE  SAND. 

By  Charles  Glaser. 

Received  July  9, 1896. 

SINCE  the  introduction  of  the  Auer-Welsbach  light,  the  com- 
mercial importance  of  monazite  sand  has  grown  greatly, 
and  chemists  are  now  asked  to  determine  the  percentage  of  true 
monazite,  and  especially  that  of  thoria,  in  samples  of  the  sand. 
This  has  heretofore  been  accomplished  indirectly  ;  the  quantities 
of  iron,  titanium  and  silica  were  determined  and  the  remainder  of 
the  material  calculated  as  monazite.  A  sample  treated  in  this 
manner  gave  the  following  results  : 

Per  cent 

Iron  oxide 3.50 

Titanic  acid 4.67 

Silica 6.40 

Monazite,  by  difference 85*43 


100.00 

The  sample  contained  18.38  per  cent,  phosphoric  acid,  which 
calculated  as  cerium  phosphate  (factor  3.32)  equals  61.10  per 
cent. 

From  analyses  printed  in  Dana's  Mineralogy,  it  was  inferred 


ESTIMATION  OF  THORIA.  783 

that  after  elimination  of  nitile  and  silica,  the  remainder  would 
be  found  to  consist  chiefly  of  phosphates  of  the  cerium  gropp,  but 
this  is  not  true. 

For  the  determination  of  the  actual  composition  of  the  mona- 
zite  sand  in  question,  it  was  decided  to  attempt  an  estimation  of 
each  of  its  components,  by  means  of  methods  to  be  found  in  the 
available  literature.  As  chief  sources  of  information,  Graham- 
Otto's  Chemistry  and  Crookes'  Select  Methods  in  Chemical 
Analysis  were  used ;  due  regard  was  also  given  to  the  work 
which  has  appeared  in  the  chemical  journals  of  recent  years.  I 
was  not  able,  however,  to  make  an  exhaustive  examination  of 
the  literature. 

It  became  evident  that  no  reliable  method  could  be  worked 
out  until  examination  had  been  made  of  all  the  work  which  had 
been  done  in  the  field,  and  it  seemed  necessary  to  investigate 
the  whole  question.  In  the  following  statements  of  preliminary 
experiments  a  large  portion  of  analytical  data  has  been  omitted, 
because  otherwise  this  paper  would  have  been  bulky.  Only  the 
outlines  of  a  general  plan  of  procedure  will  therefore  be  given. 

So  far  as  possible,  it  was  my  intention  to  examine  all  the 
methods  proposed  for  estimation  of  thoria,  but  in  one  notable 
instance  this  could  not  be  done.  In  Volume  XVI  of  the  Ameri- 
can Chemical  Journal,  L.  M.  Dennis  and  F.  L.  Kortright 
describe  a  method  for  estimation  of  thoria  by  means  of  potassium 
hydronitride,  KN,.  An  attempt  to  work  by  the  method  proved 
a  failure  in  my  hands,  partly  because  of  a  mishap  while  prepar- 
ing the  reagent,  only  enough  of  which  was  saved  for  a  single 
qualitative  reaction ;  but  chiefly  because  Mr.  Dennis  declined, 
when  requested,  to  give  me  further  information.  He  replied 
that  he  was  not  then  at  liberty  to  detail  his  experience,  *'  as  the 
potassium  hydronitride  process  is  more  than  an  analytical  one. 
It  is  a  commercial  process  for  the  preparation  of  pure  thoria, 
which  is,  I  think,  unequalled  by  any  of  the  methods  employed 
by  the  Welsbach  chemists,  Shapleigh  included.  Some  of  them 
have  tried  to  use  the  method  and  have  failed.  I  think  I  know 
why  they  failed.  But  I  do  not  think  it  quite  fair  for  them  to 
ask  me  to  help  them  out  of  their  difficulties." 

Although  the  publication  was  made  in  a  scientific  journal,  it 


784  CHARLBS  GLASBR. 

seems  to  have  been  but  a  partial  statement.  For  which  reason 
criticism  is  invited  and  the  value  of  the  work  is  thrown  some- 
what in  doubt.     No  further  attempt  was  made  to  follow  it  out. 

By  means  of  fusion  with  alkali  carbonates,  an  attempt  was 
made  to  separate  monazite  sand  into  two  parts.  According  to 
Wohler  all  titanic  acid  ought  to  become  soluble  provided  the 
fusion  is  made  at  a  su£Bciently  high  temperature.  Therefore  a 
blowpipe  was  used.  In  later  work  I  employed  the  highest  tem- 
peratures afforded  by  a  muffle,  and  for  as  many  as  two  hours. 
But  at  no  time  was  more  than  a  fraction  of  the  titanic  acid  ren- 
dered soluble  in  water.  Moreover,  Wohler's  directions  to  pour 
the  fusion  upon  an  iron  plate,  and  afterwards  to  powder  it,  are 
not  practicable  because  of  loss  likely  to  ensue.  It  was  found 
best  to  let  the  fusion  soak  in  water  over  night,  sometimes  even 
for  several  days,  or  until  perfect  disintegration  resulted.  But 
such  a  procedure  may  have  decreased  the  solubility  of  titanic 
acid  in  water.  Phosphoric  acid  and  alumina  (and  also  silica  to 
a  large  extent)  were  completely  dissolved  out  of  the  fused  mass. 
The  portion  insoluble  in  water  was  rendered  soluble  by  the  well 
known  treatment  with  strong  sulphuric  acid,  and  also  by  fusion 
with  acid  potassium  sulphate.  The  solution  thus  obtained, 
after  being  freed  from  silica,  was  boiled  to  separate  titanic  acid» 
from  four  to  seven  hours  during  the  first  experiment.  Later, 
after  addition  of  sodium  sulphite,  this  was  accompanied  by  satu- 
rating with  hydrogen  sulphide,  first  in  the  hot  and  then  in  the 
cooled  solution.     This  method  is  preferable  to  the  first. 

After  separation  of  titanic  acid  and  the  metals  of  the  fifth 
group,  various  methods  were  tried  for  separation  of  thoria  from 
the  other  earths.  It  was  found  that  the  solution  must  not  be 
strongly  acid  when  treated  with  ammonium  oxalate  for  precipi- 
tation of  thoria  and  the  metals  of  the  cerium  group,  or  traces  of 
thoria  will  remain  in  solution.  It  is  best  to  nearly  neutralize 
with  ammonia,  and  to  precipitate  in  boiling  solution. 

During  the  earlier  experiments  some  difficulty  was  found  in 
keeping  in  solution  all  of  the  zirconia,  which  is.  accomplished 
only  by  a  large  excess  of  the  reagent,  while  yttria  and  glucina 
readily  form  soluble  double  salts.  Under  these  conditions  oxa- 
lates of  the  cerium  metals  precipitate  immediately,  while  thorium 


ESTIMATION   OF  THORIA.  785 

oxalate  separates  upon  cooling.  Attempts  to  separate  thorium 
oxalate  from  oxalates  of  the  metals  of  the  cerium  group  by  fil- 
tration of  the  hot  solution,  gave  unsatisfactory  results.  The 
oxalates  will  pass  through  the  filter  for  a  long  time.  Bumping 
of  the  liquid  made  it  impracticable  to  keep  it  boiling  until  the 
entire  precipitate  became  crystalline.  But  if  large  quantities  of 
thoria  are  to  be  separated  from  small  ones  of  the  other  oxalates 
the  method  works  wdl. 

After  the  insoluble  oxalates  were  separated  by  filtration  and 
were  washed  with  water,  they  were  converted  into  oxides  by 
heating  and  were  redissolved  as  sulphates.  In  this  strongly  con- 
centrated solution,  made  nearly  neutral  by  ammonia,  an  attempt 
was  made  to  separate  thoria  from  the  other  metals  by  boiling 
with  sodium  hyposulphite.  In  no  instance  was  a  complete  sep- 
aration effected,  but  such  portions  as  were  obtained  proved  to 
be  quite  pure.  The  single  exception  was  that  in  which  the 
whole  of  the  cerium  was  precipitated,  for  reasons  not  ascer- 
tained. Attempts  were  made  to  free  thoria  from  most  of  the 
cerium  by  fractional  precipitation  with  weak  ammonia,  but  no 
considerable  advantage  was  gained  thereby,  since  tepeatedly  the 
second  fraction  showed  traces  of  thorium. 

To  determine  the  solubility  or  insolubility  of  the  different  sub- 
stances left  in  the  insoluble  residue  from  fusions,  such  residue 
was  treated  with  dilute  hydrochloric  acid  both  cold  and  hot. 
The  solution  was  found  to  contain  all  the  iron  and  titanium,  the 
larger  part  of  the  silica,  and  about  one-half  of  the  earths  present ; 
these  consisted  of  relatively  large  portions  of  zirconia  and  glu- 
cina.  Thoria  seems  not  to  enter  into  solution,  but  is  left  with 
the  remainder  of  the  earths. 

An  attempt  was  made  to  separate  thorium  oxalate  from  the 
mixed  precipitated  oxalates,  by  boiling  with  ammonium  oxalate. 
Such  boiling,  filtering  and  crystallizing  yielded  oxalates,  which 
after  ignition,  corresponded  to  2.29  per  cent,  of  oxides.  The 
earths  were,  however,  of  a  deep  orange  color,  and  contained 
both  cerium  and  zirconia.  The  latter  was  present  because  an 
insufficient  quantity  of  ammonium  oxalate  had  been  used  in  the 
first  precipitation.  In  the  oxalates  of  the  cerium  metals  found 
insoluble  in  the  above  treatment,  the  presence  of  thoria  could  be 


786  CHARI^BS  GI^ASBR. 

distinctly  proven  by  means  of  sodium  hyposulphite,  for  which 
reason  the  work  proved  unsatisfactory. 

To  facilitate  a  comparison  of  the  more  important  reactions  of 
the  elements  herein  studied,  the  table  on  the  next  page  has  been 
prepared  partly  from  their  known  behavior,  and  partly  from  the 
results  obtained  during  this  investigation. 

With  the  view  of  obtaining  further  knowledge  of  the  behavior 
of  thoria,  fragments  of  Welsbach  mantles  were  subjected  to 
analysis.  They  weighed  0.6591  gram,  which,  after  ignition, 
fell  to  0.6552  gram.  Prolonged  treatment  with  boiling  sul- 
phuric acid  left  a  residue  of  0.0883  gram,  which  became  solu- 
ble in  water  after  fusion  with  acid  potassium  sulphate.  The  solu- 
tions thus  obtained  were  examined  by  the  same  method,  but 
separately,  as  follows :  After  neutralizing  with  ammonia  the 
greater  part  of  the  free  acid,  the  solutions  were  heated  to  boiling 
and  hot  solution  of  ammonium  oxalate  was  added. 

In  solution  I  a  precipitate  appeared,  but  dissolved  rapidly 
upon  addition  of  more  of  the  reagent. 

In  solution  II  a  slight  turbidity  appeared,  there  was  no  pre- 
cipitate, and  it  soon  became  perfectly  clear. 

Upon  cooling,  solution  I  yielded  a  moderate  quantity  of  a 
crystalline  deposit,  while  solution  II  gave  a  copious  one.  Both 
precipitates  were  collected  on  one  filter,  washed,  ignited,  and 
weighed.     They  yielded  o.  1 124  grams  of  thoria. 

The  filtrate  from  I  gave  a  copious  precipitate  with  ammonia, 
while  that  from  II  gave  only  a  slight  one  :  both  of  these  were 
washed  on  one  filter,  redissolved  in  dilute  hydrochloric  acid,  and 
again  precipitated  by  ammonia.  An  excess  of  ammonium  car- 
bonate entirely  dissolved  the  precipitate.  Potassium  hydroxide 
gave  a  precipitate  not  soluble  in  excess  of  the  precipitant,  indi- 
cating zirconia,  the  weight  of  which  was  0.5580  gram.  An 
attempt  to  purify  it  from  occluded  alkali,  by  again  precipitating 
with  ammonia,  failed  through  an  accident,  in  which  part  of  the 
material  was  lost.  Calculating  by  difference,  the  weight  of  zir- 
conia ought  to  have  been  0.5428  gram.  Both  precipitates  were 
pure  white. 

Therefore,  this  analysis  afforded  the  following  composition  of 
the  mantles :  thoria  17.15  per  cent.,  zirconia  82.85  per  cent. 


I 

5 

1     III 

1 

1 

► 

1! 

Is 
'1 

d 

1?      f? 

■    d                      B 

1 

I 

'  8 

i 

11 

hS  Ii  ,s 

1 

I 

■^s 

'if  11 

1 

> 

\ 

1    1 

1  1 

5 
1 

1 
1 

lis 

1 

1 

% 

hi 

a 

I    I 

1 

r 

ill? 

i 

z 

1 

-1  \Mi 

c 

i 

ift 

1 

3 

1 

5      % 

?      1 

5 
1 

1 

1 

1 
1    - 

1 

Us! 

Irf 

2 

■5 

1 

X 

0 

1   1 
1   1 

7 
8 

K»    - 

: 

F 

i!n 

z 

:» 

s 

788  CHARLES  GLASBK. 

The  separation  of  the  two  earths  was  effected  without  diffi- 
culty and  the  thoria  was  used  in  the  following  experiments : 

0.0487  gram  was  weighed,  dissolved,  and  mixed  with  the 
solution  of  cerium  metals  from  a  previous  experiment.  The 
solution  was  nearly  neutralized  with  ammonia,  heated  to  boil- 
ing, a  hot  solution  of  ammoni'am  oxalate  added,  and  the  mix- 
ture allowed  to  cool.  The  precipitate  was  caught  on  a  filter  and 
washed  with  cold  water,  extracted  in  boiling  ammonium  oxalate 
solution,  caught  on  a  filter,  and  washed  hot :  the  filtrate  was 
allowed  to  cool  (precipitate  i).  The  residue  was  macerated  in 
a  hot  solution  of  ammonium  acetate,  filtered  (residue  A),  and 
filtrate  examined  for  thoria,  as  follows  :  hydrochloric  acid  was 
added  to  separate  thoria  as  oxalate,  which  fell  in  part  only 
and  the  remainder  was  obtained  by  sodium  hydroxide  (precipi- 
tate 2).  Both  these  precipitates  afforded  batapart  of  the  thoria 
originally  weighed,  the  greater  part  being  held  yet  with  the 
cerium  metals.     The  method  had  failed. 

The  residue  (A)  upon  the  filter  was  reduced  to  oxide  and  dis- 
solved as  sulphate.  After  neutralizing  with  ammonia,  the 
liquid  was  heated  to  boiling,  and  there  was  added  an  excess  of 
ammonium  oxalate  with  some  ammonium  acetate  :  after  filter- 
ing, the  filtrate  was  treated  with  sodium  hydroxide  (precipi- 
tate 3). 

The  precipitates,  thus  obtained  in  three  fractions,  were  ignited 
and  found  to  weigh  0.0774  gram,  showing  that  the  thoria  was 
very  impure.  The  grayish  mass  was  fused  with  acid  potassium 
sulphate,  and  unfortunately,  a  small  fraction  of  the  fused 
mass  was  lost.  However,  from  the  saved  portion  a  pure  thona, 
weighing  0.0402  gram,  was  obtained. 

In  the  next  experiment,  0.0343  gram  of  thoria  and  0.1004 
gram  of  impure  cerium  oxide  were  dissolved  as  sulphates,  and 
aium  oxalate  and  acetate,  as  for  precipitate 
precipitating  the  filtrate  with  ammonia  there 
o  of  impure  thoria,  which,  after  purification, 
im.     Cerium  oxide  reco\-ered  weighed  0.0935 

Mention  to  what  has  been  observed  frequently 
iments.     If  thorium  oxalate,  held  in  solution 


ESTIMATION   OF  THORIA.  789 

by  ammonium  acetate,  be  precipitated  by  ammonia,  the  earth  so 
obtained,  when  washed  with  the  greatest  care  and  redissolved 
in  a  mineral  acid,  cannot  from  an  almost  neutral  solution  be 
again  completely  precipitated  by  ammonium  oxalate ;  even  if 
the  earth  had  been  ignited  after  re-solution.  It  will  also  be  found 
that  a  considerable  increase  has  occurred  in  its  solubility  in 
liquids  containing  much  potassium  or  ammonium  sulphate. 
When  enough  thoria  has  been  collected,  it  is  my  intention  to 
iurther  examine  this  peculiar  behavior. 

SYSTEMATIC  METHOD  OP  ANALYSIS. 

From  the  analytical  data  given,  the  following  method  has 
been  deduced  : 

It  is  essential  that  the  mineral  be  divided  to  the  greatest  pos- 
sible degree.  Prolonged  powdering  in  an  agate  mortar  is  indis- 
pensable. Solution  is  effected  either  by  prolonged  heating  with 
strong  sulphuric  acid,  or  by  fusion  with  acid  potassium  sul- 
phate. In  the  latter  case,  the  cooled  mass  is  warmed  with  so 
much  sulphuric  acid  that  the  product,  after  cooling,  may  be 
poured  from  the  crucible.  The  first  method  takes  more  time 
than  the  second,  but  it  introduces  less  of  the  objectionable 
potassium  salts.  It  is  advisable  to  fuse  only  those  portions 
which  are  insoluble  in  sulphuric  acid. 

For  estimation  of  silica  the  sulphuric  acid  treatment  is  pref- 
erable, in  which  case  it  is  best  to  evaporate  once  on  a  sand- 
bath  to  dryness  to  render  silica  insoluble,  and  then  to  add  fresh 
sulphuric  acid.  The  resulting  mixture  should  be  added  slowly 
to  ice  cold  water,  which  dissolves  the  mass  excepting  silica  and 
tantalic  acid,  with  possibly  traces  of  titanic  acid,  thoria,  and 
zirconia.  After  filtering,  the  residue  should  be  ignited  and 
weighed.  Silica  is  eliminated  by  repeated  treatment  with 
hydrofluoric  acid.  Any  residue  remaining  should  be  moistened 
with  sulphuric  acid,  to  convert  the  fluorides  of  the  earths  into 
sulphates,  which,  after  ignition  at  a  high  temperature,  are 
weighed  as  oxides,  and  silica  determined  by  the  loss  in  weight. 
The  residue  of  tantalic  acid,  with  possibly  traces  of  the  bodies 
mentioned  above,  is  treated  with  sulphuric  acid  and  hydrofluoric 
acid.    Tantalic  acid  remains  insoluble,  and  may  be  filtered  off 


79^  CBARLBS  GIASBR. 

and  weighed.  The  matter  soluble  may  be  added  to  tbe  main 
solution. 

The  original  solution  is  saturated  with  hydrogen  stilpbtde, 
first  at  boiling  and  then  at  the  ordinaiy  temperature.  Titanic 
acid  is  precipitated,  together  with  metals  of  the  fifth  group. 
That  sodium  sulphite  assists  in  the  precipitation  of  titanic  acid 
has  not  been  verified  in  my  work. 

When  completely  settled,  the  liquid  is  filtered  and  the  filtrate 
boiled  to  expel  hydrogen  sulphide.  Any  free  acid  may  be 
nearly  neutralized  with  ammonia  ;  to  the  boiling  liquid  is  added 
an  excess  of  a  boiling  solution  of  ammonium  oxalate,  as  much 
as  loo  cc.  of  the  cold  saturated  solution  for  two  grams  of  mona- 
zite  sand.  The  excess  necessarily  must  be  large.  The  mixture 
is  then  permitted  to  cool,  best  for  an  entire  night.  The  solu- 
tion will  contain  phosphoric  acid,  the  oxides  of  iron,  manganese, 
aluminum,  glucinum,  yttrium,  zirconium,  and  calcium.  In  the 
precipitate  will  be  found  thoria  and  the  oxides  of  the  cerium 
group. 

If  the  bodies  in  solution  are  to  be  estimated,  add  ammonia  to 
precipitate  the  metals  as  phosphates.  Filter  and  wash  thor- 
oughly, preserve  the  filtrate  for  estimation  of  phosphoric  add 
and  alumina.  Ignite  the  precipitate  and  fuse  it  with  mixed  car- 
bonates of  potassium  and  sodium.  The  fused  mass  is  exhausted 
with  hot  water,  filtered,  and  the  residue  well  washed  with  hot 
water.  The  filtrate  is  added  to  that  containing  phosphoric  acid 
and  alumina. 

The  remaining  oxides  and  carbonates  are  dissolved  in  sulphuric 
acid  and  precipitated  with  ammonia.  Lime  is  estimated  in  the 
filtrate  therefrom. 

When  an  attempt  is  now  made  to  dissolve  the  precipitated  hy- 
droxides on  the  filter  by  dilute  hydrochloric  acid,  it  sometimes 
occurs  that  zirconia  in  part  remains.  Therefore  it  is  best,  after 
this  operation,  to  incinerate  the  filter.  Then  neutralize  the  so- 
'   ■■         '*  aia  as  far  as  practicable.     Pour  this  slowly, 

ing,  into  a  mixture  of  ammonium  carbonate 
Iphide,  prepared  as  follows  :  To  a  solutiofi  of 
late  more  than  enough  to  neutralize  the  free 
above  indicated,  and  to  hold  in  solution  the 


ESTIMATION   OF   THORIA.  79 1 

earths  to  be  dealt  with,  add  enough  of  ammonium  sulphide  (usually 
a  few  cc.)  to  precipitate  the  metals  of  the  fourth  group.  The 
latter  will  be  precipitated,  while  zirconia,  yttria,  and  glucinum 
remain  in  solution.  Iron  and  manganese  may  be  determined  by 
the  usual  methods. 

If  the  carbonate  solution  be  boiled  for  one  hour  the  earths  are 
completely  precipitated.  They  may  be  caught  on  a  filter  and 
dissolved  in  hydrochloric  acid  ;  or  the  carbonate  solution  may 
be  treated  directly  with  that  acid,  carbon  dioxide  expelled  by 
boiling,  the  solution  cooled  and  treated  with  an  excess  of  sodium 
hydroxide.  Zirconium  and  yttria  are  completely  precipitated 
while  glucina  remains  dissolved :  to  precipitate  this,  boil  the 
solution  one  hour. 

To  separate  zirconia  from  yttria,  dissolve  the  hydroxides  in 
hydrochloric  acid,  warm,  then  saturate  the  solution  with  sodium 
sulphate.  When  cold,  zirconia  separates  in  the  well-known 
manner.   From  the  filtrate  ammonia  separates  yttria. 

As  the  earths  are  apt  to  occlude  alkali  salts,  it  is  best  to  dis- 
solve and  again  precipitate  them  (with  ammonia)  before  they 
are  ignited  and  weighed. 

Separation  of  the  precipitated  oxalates  of  thoria  and  of  the 
cerium  group  is  accomplished  as  follows :  The  oxalates  are 
reduced  to  oxides  by  ignition,  then  converted  into  sulphates, 
the  greater  part  of  the  free  acid  neutralized  with  ammonia,  the 
solution  boiled,  and  boiling  ammonium  oxalate  added  in  excess. 
After  a  short  time  (as  soon  as  oxalates  of  the  cerium  metals 
have  formed  but  before  the  liquid  has  cooled) ,  a  few  cc.  of  solu- 
tion of  ammonium  acetate  are  added.  When  cold,  the  entire 
cerium  group  is  precipitated  as  oxalates,  while  thoria  remains 
in  solution.  After  prolonged  standing,  best  over  night,  the 
insoluble  oxalates  are  removed  by  filtration  ;  in  the  filtrate,  pre- 
cipitate thoria  with  ammonia  in  excess,  filter,  ignite,  and 
weigh. 

Separation  of  cerium  from  lanthanum  and  didymium  is  easily 
accomplished  by  the  well  known  method.  Pass  a  current  of 
chlorine  through  the  liquid  containing  the  hydroxides.  Which 
have  been  freshly  precipitated  by  a  fixed  alkali. 

Separation  of  lanthanum  from  didymium  was  not  attempted. 


792  ESTIMATION  OP  THORIA. 

An  analysis  of  the  monazite  sand  used  in  my  work,  made  as 

indicated  in  the  foregoing  notes,  gave  results  as  follows  : 

Titanic  acid 4.67 

Silica 6.40 

Phosphorns  pentoxide ^ 18.38 

Lead trace 

Alumina 1.62 

Lime 1.20 

Cerium  oxide  (CeO) 32.93 

Lanthanum  and  didymium  oxides 7.93 

Thoria 1.43 

Ferric.oxide 7.81 

Zirconia  and  yttria i3-9o 

Glucina 1.25 

Tantalic  acid 0.66 

Not  determined 1.72 

100.00 
Titanic  acid  and  silica  was  determined  in  a  separate  portion. 
The  determination  of  tantalic  acid  was  only  approximate, 
since  a  part  of  it  is  dissolved  by  fusion  with  acid  potassium  sul- 
phate, and  thus  escapes  weighing.  As  several  such  fusions  were 
made,  it  is  probable  that  the  greater  part  of  the  matter  '  'not  deter- 
mined" ought  to  be  reckoned  as  tantalic  acid.  The  quantity 
stated  was  an  average  of  three  determinations  (minus  or  plus 
0.05)  from  the  residue  of  repeated  fusions. 

Through  the  courtesy  of  Mr.  H.  B.  C.  Nitze,  of  the  Geologi- 
cal Survey  of  North  Carolina,  I  have  received  a  number  of  sam- 
ples of  monazite  sand  mined  at  various  localities  in  that  state. 
Two  of  these  had  been  prepared  by  a  new  process  and  were 
found  to  be  practically  free  from  rutile  and  garnets.  They  were 
excellent  material  for  my  methods  of  analysis,  and  they  gave 
results  as  follows : 

Anai«vsis  of  a  Coarse  Monazite  Sand  prom  Shelby,  NoRTrf  Caro- 
lina. 

Silica 3.20 

Titanic  acid 0.61 

Cerium  metals  as  CeO ^3*^ 

Phosphorus  pentoxide 28. 16 

Thona 2.32 

Zirconia,  glucina,  yttria 1.52 

Manganese trace 

No  iron,  alumina,  or  lime 0.00 

99.61 

The  color  of  this  sand  was  honey-yellow. 


PRECIPITATION   OP  BARIUM   SUI.PHATE.  793 

Analysis  op  k  Pinb  Monazitb  Sand  prom  Bbllbwood,  North  Caro- 
lina. 

Silica 1.45 

Titanic  acid 1.40 

Cerium  metals  as  CeO < . . .  59.09 

Phosphorus  pentozide 26.05 

Tbona 1. 10 

Zirconia»  glucina,  yttria 2.68 

Tantalic  acid 6.39 

Iron  and  manganese  oxides 0.65 

Alumina 0.15 

9905 
The  color  of  this  sand  was  honey-yellow. 

Laboratory  op  Lbhkann  Sl  Gx^abbr, 
Saltimorr. 


[Contributions  prom   Metallurgical   ItAboratory  op  thb  Ohio 

Statk  University,  Columbus,  Ohio.] 

THE  EFFECT  OF  AN  EXCESS  OF  REAGENT  IN  THE  PRE- 
CIPITATION OF  BARIUM  SULPHATE. 

'  By  C.  W.  F0UI.X. 

Received  July  6,  tBfb. 

*  <T7XCESS  of  reagent*'  is  a  term  often  used  by  writers  in 
L/  quantitative  chemistry,  and  the  necessity  in  any  given 
case  for  adding  more  of  a  precipitating  reagent  than  is  just  suffici- 
ent for  complete  reaction  is  well  known  to  analysts ;  but  what  con- 
stitutes such  excess,  whether  it  differs  for  different  salts,  whether 
its  effect  is  counteracted  by  the  presence  in  the  solution  of  other 
bodies  not  taking  part  in  the  reaction,  or  whether  the  effect  of 
such  bodies  may  be  counteracted  by  the  addition  of  a  greater 
amount  of  precipitant,  etc.,  etc.,  are  questions,  the  answers  to 
which  are  difficult  to  find  in  chemical  literature. 

With  a  view  to  answer,  in  part  at  least,  these  questions,  the 
following  work  on  the  precipitation  of  barium  sulphate  was 
undertaken. 

A  preliminary  experiment,  which  perhaps  is  worth  noting, 
was  first  tried : 

A  solution  of  140  cc.  water  and  five  cc.  concentrated  hydro- 
chloric acid  was  heated  nearly  to  boiling  and  0.1984  gram  pure 
recently  ignited  barium  sulphate  was  added.  This  was  then 
stirred  up  and  set  aside  for  one  hour,  when  it  was  filtered  and 


794  C.   W.   FOULK. 

the  barium  sulphate  washed  well  with  hot  water.  The  filter 
and  the  contents  were  then  ignited  and  weighed,  when  it  was 
found  that  ten  milligrams  of  the  sulphate  had  been  dissolved. 
The  filtrate  was  now  divided,  and  to  one-half  some  sulphuric 
acid  was  added,  and  to  the  other  some  barium  chloride  solution. 
A  precipitate  of  barium  sulphate  was  produced  in  both  cases. 

Standard  solutions  of  sulphuric  acid  and  barium  chloride  were 
now  prepared.  These  were  standardized  by  precipitation  from 
pure  water  solutions. 

The  sulphuric  acid  used  in  this  work  was  the  chemically  pure 
acid  of  the  laboratory,  tested  for  the  ordinary  impurities. 

The  barium  chloride  was  recrystallized  from  the  chemically 
pure  salt. 

The  hydrochloric  acid  was  the  chemically  pure  acid  of  the 
laboratory  tested  for  sulphuric  acid. 

The  graduated  ware  was  calibrated  and  found  to  be  good. 

All  the  precipitates  of  barium  sulphate  were  ignited  by  fold- 
ing up  the  moist  filter,  putting  into  a  platinum  crucible,  '*  pre- 
cipitate end"  up  and  so  adjusting  the  flame  that  the  paper 
would  be  charred  away  "without  letting  the  crucible  become  red 
hot.  Finally  the  heat  was  raised  and  the  ignition  finished.  No 
lid  was  used  on  the  crucible.  By  following  this  plan  no  reduc- 
tion to  sulphide  need  be  feared. 

A  number  of  the  precipitates  were  moistened  with  sulphuric 
acid  and  ignited.     No  change  was  noticed. 

In  the  course  of  the  work  the  following  solutions  were  made  : 

Sui^PHURic  Acid  Solution. 
Solution  A. 

cc.  Barium  sulphate. 

1.  20 0.1978 

2.  20 0.1975 

3.  20 0.1970 

4.  20 0.1978 

Average o.  1970 

Solution  B, 
cc.  Barium  sulphate. 

1.  50 0-3277 

2.  50 0.3271 

3-  50 0.3279 

Average 0.3275 


PRECIPITATION  OP  BARIUM  SULPHATE.  795 

Solution  C. 
cc.  Barium  sulphate. 

I-     5 0.1944 

2.    5 o.  1940 

Average*. 0.1942 

Solution  D. 

cc.  Barium  sulphate. 

1.  25 0.1544 

2.  25 0.1534 

3-    25 0.1543 

4.  25 0.1538 

5.  25 0.1535 

6.     25 0.1546 

7-     25 0.1539 

Average 0.1542 

Rejecting  Nos.  2  and  5. 

Barium  Chu»ridb  Solutions. 
Solution  A, 

cc.  Barium  sulphate. 

1.  20 O.I181 

2.  20 O.1812 

3.  20 o.  181 1 

4.  20 0.1792 

5.  20 0.1802 

Average o.  1805 

Rejecting  No.  4. 

Solution  B. 
cc.  Barium  sulphate. 

1.  50 0.1985 

2.  50 0.1980 

3.  50 0.1086 

4.  50 01985 

Average # 0.1984 

Solution  C, 
cc.  Barium  sulphate. 

1.  10 0.4004 

2.  10 0.4002 

3.  10 0.4006 

Average 0.4004 

Solution  D, 

cc.  Barium  sulphate. 

1.  10 0.3998 

2.  10 0.3994 

3.  10 0.3996 

Average 0.3996 


79^  C.   W.   FOULK. 

Note. — The  apparent  discrepancies  in  some  of  the  above  aver- 
ages are  to  be  explained  by  the  fact  that  before  beginning  the 
work  the  burette  used  had  been  very  carefully  calibrated,  and 
the  averages  were  calculated  to  correct  number  of  cubic  centi- 
meters from  the  readings  as  given  on  the  burette.  In  the  course 
of  the  work  this  refinement  was  found  to  be  wholly  unnecessary 
and  was  therefore  disregarded. 

The  equation  of  solutions  of  sulphuric  acid  and  of  barium 
chloride  is:  Twenty  cc.  barium  chloride  solution  =  21.8  cc. 
sulphuric  acid.  That  is,  when  mixed  in  these  proportions 
they  will,  theoretically,  mutually  precipitate  each  other  and 
give  0.1970  gram  barium  sulphate. 

The  effect  of  bringing  these  two  solutions  together  in  this  pro- 
portion was  first  tried.  The  barium  chloride  solution  plus  water 
to  make  the  whole  volume  up  to  140  cc.  was  heated  to  boiling 
and  the  sulphuric  acid  run  in  from  the  burette. 

Barium  snlphate.  Error. 

1.  20CC.  BaCl,yH-2i.8cc.  HjSOi^ 0.1966  —0.0004 

2.  20  **        *•              •*     **          "       0.1973  40.0003 

3.  20  •*        '•              "     '*          ••       0.1979  +0.0009 

Solutions  of  BaCl,^  and  H,SO«,  when  brought  together  in 
their  molecular  proportions,  weighed  as  follows : 

Barium  salptaate.         Error. 

1.  50  CC.  BaCl^-h  30.2  cc.  HjSO^^ 0.1979  —0.0005 

2.  50  ••         *•  "     •*  "       0.1976  — OlOOOS 

These  had  stood  twenty-two  hours  before  filtration,  and  the 
results,  while  not  very  close,  show  at  least  that  in  water  solu- 
tions precipitation  is  practically  complete  without  the  presence 
of  an  excess  of  reagent. 

A  series  of  precipitations  was  now  made  in  order  to  determine 
the  ettect  of  \-ar>'ing  quantities  of  hydrochloric  acid  upon  the 
precipitation  when  the  two  reagents  were  brought  together  in 
their  molecular  proportions. 

The  barium  chloride  solution,  water  to  mabe  the  volume  np 
to  14c*  cc.,  and  the  hydrochloric  acid  were  heated  to  boiling  and 
the  sulphuric  acid  run  in  cold  from  the  burette. 

The  same  quantities  of  barium  chloride  and  sulphuric  acid 
were  usevi  as  aV\-e.     The  time  of  standing  before  filtration  is 


PRECIPITATION  OP  BARIUM  SULPHATE.  797 

marked  over  each  set.     Three  precipitations  were  made  with 
each  portion  of  the  hydrochloric  acid. 

Sbribs  I. 

1.  a.  3*  4- 

Five  cc.  Ten  cc.  Fifteen  cc.  Twenty  cc. 

hydrochloric  hydrochloric  hydrochloric  hydrochloric 

acid.  acid.  acid.    .  acid. 

Twenty-five  Twenty-nine  Thirty-three  Forty-four 

hours.  hours.  hours.  hours. 

Barium  sulphate.  Barium  sulphate.  Barium  sulphate.  Barium  sulphate. 

I P-I90^  0.1879  0.1837  0.1875 

2 0.1902  0.1870  0.1844  0.1863 

3 0.1904  O.1881  0.1838  0.1873 

It  was  thought  that  after  standing  twenty-four  hours  precipita- 
tion would  be  complete  and  a  longer  time  would  have  no  effect. 
The  results  of  series  No.  4  seem  to  show  differently,  however. 
Accordingly  another  series  was  run  in  which  the  time  of  stand- 
ing was  regulated.  Otherwise  the  precipitations  were  made  as 
above. 

These  stood  twenty- three  hours  before  filtration. 

Shribs  II. 
I.  2.  3. 

Five  cc.  Ten  cc.  Fifteen  cc. 

hydrochloric  hydrochloric  hydrochloric 

acid.  acid.  acid. 

Barium  sulphate.  Barium  sulphate.  Barium  sulphate. 

I 0.1902  0.1870  0.1852 

2 0.1884  6.1854  0.1849 

3 0.1904  0.1846  0.1827 

4.  5-  6. 

Twenty  cc.  Twenty-five  cc.  Thirty  cc. 

hydrochloric  hydrochloric  hydrochloric 

acid.  acid.  acid. 

Barium  sulphate.  Barium  sulphate.  Barium  sulphate. 

I 0.1832  0.1822*  0.1766 

2 0.1885  0.1793  0.1833 

3 0.1850  0.1789  0.1733 

The  above  results  show  three  things  :  (i)  That  less  barium 
sulphate  is  precipitated  in  the  presence  of  larger  amounts  of 
hydrochloric  acid,  but  this  solubility  is  not  proportional  to  the 
amount  of  hydrochloric  acid.  (2)  That  the  greatest  variation 
of  results  takes  place  in  the  presence  of  the  larger  amounts  of 
acid.  In  other  words  parallel  precipitations  don't  '*  check.*' 
(3)  A  much  longer  time  is  required  to  reach  the  maximum  of 
precipitation  in  the  presence  of  the  larger  amounts  of  hydro- 
chloric acid.     See  No.  4,  Series  I. 


798  c.  w.  FOULE. 

The  efiect  of  a  small  excess  of  sulphuric  acid  was  dqw  tried. 
Three  solutions  each  containing  fifty  cc.  barium  chloride  B, 
sixty  cc.  water  and  twenty  cc.  hydrochloric  add  were  heated  to 
boiling  and  the  amounts  of  sulphuric  acid  B,  indicated  below, 
were  run  in  from  a  burette. 

These  stood  twenty-four  hours  and  weighed  as  follows : 
Series  III. 

Barium  lulpfaatf.      Smr. 

1.  50  CC.  BaCl,  B  +  31.3  cc.  H,SO,  f «  t  cc.  exceu  =  0.1839  — «-oi45 
a.  50  "  "  +33.3  "  "  —a  "  "  —  0.1881  —0.0103 
3.    50  "  .      "         +  33a  "        "         =3  "        "      —  o-'97i       —0.0013 

The  filtrates  from  the  above  gave  no  further  precipitate  on 
standing  several  days. 

Another  series  was  run,  using  five  cc.  hydrochloric  acid  in- 
stead  of  twenty  cc,  but  conducted  otherwise  in  the  same  man- 
ner except  that  they  stood  from  Friday  to  the  following  Monday 
and  undoubtedly  the  maximum  of  precipitation  was  reached. 
Sbriks  IV. 

Barinm  ■nlpbatt      Bnvr. 

I.    50  CC  BaCl)  B  -H  31.3  cc.  H,SOt  B  =  \tx.  excess  =  0.1951  0.0033 

a.    50  "      "      '•  +33.3  "       "      "  =3  "       "       =0.1963  0.003I 

3.    50  •'      "      "  4-33-3  "       "      "  =3  "       "       =0.1964  0.0030 

It  was  now  decided  to  use  lai^er  amounts  of  sulphuric  acid 
in  excess,  but  in  order  to  hurry  matters  along,  cut  down  the 
time  of  standing  before  filtration. 

In  the  following  series,  accordingly,  the  barium  sulphate  was 
filtered  off  after  standing  three  hours.  The  whole  volume  of 
solution  in  each  case  was  150  cc. 

Series  V. 


.^^B 

f^ 

ih 

u 

"g^'p" 

t  Barium 

Srrar. 

50 

SO 

'5 
15 

35' 
40.3 
45-3 

SO.  a 

S 
30 

o.i«8 
0.1590 
0.1688 
0..7fa 

0.05« 
0.0394 
0.0^ 

lowing  the  larger  amounts  of  sulphuric 
■  be  noted  that  30.2  cc.  sulphuric  acid  in 

PRECIPITATION  OF  BARIUM   SULPHATE.  799 

three  hours  did  not  bring  down  so  large  a  precipitation  as  31.2 
cc.  sulphuric  acid  did  in  twenty-four  hours  though  in  the  pres- 
ence of  a  larger  portion  of  hydrochloric  acid.     See  Series  III. 

In  order  to  get  comparative  results  the  various  conditions  of 
the  precipitation  had  to  be  more  carefully  regulated.  The 
above  results  show  this  very  plainly. 

Accordingly,  the  following  problem  was  set :  How  great  an 
excess  of  sulphuric  acid  is  required  to  precipitate  completely  as 
sulphate,  the  barium  from  fifty  cc.  of  barium  chloride  B,  in  the 
presence  of  five  cc.  hydrochloric  acid  in  one  hour,  the  whole 
volume  of  solution,  after  adding  the  sulphuric  acid,  to  be  150CC.  ? 

Instead  of  adding  a  certain  number  of  cc.  in  excess  the  sul- 
phuric acid  was  now  measured  in  equivalents,  30.2  cc.  the  exact 
amount  to  precipitate  fifty  cc.  barium  chloride  was  called  one 
equivalent  and  different  multiples  of  it  were  taken. 

The  barium  chloride,  water,  and  hydrochloric  acid  were 
heated  on  the  water-bath  and  the  sulphuric  acid  run  in  cold 
from  the  burette. 


Series  VI. 

Hydro. 
Barium    chloric 
chloride^,   acid. 

Sul- 
phuric 
acid  B. 

Eqivalents 
sul- 
phuric 
acid. 

Barium 
sulphate. 

Error. 

cc.            cc. 

I 

50            5 

37.8 

1.25 

0.1564 

0.0420 

2 

50            5 

45-3 

1.50 

0.1624 

0.0360 

3 

50            5 

52.8 

1.75 

O.I7S4 

0.0200 

4 

50            5 

60.4 

2.00 

O.1S57 

0.0127 

5 

50            5 

68.9 

2.25 

0.1842 

0.0142 

The  fact  that  No.  5  was  lower  than  No.  4  was  referred  to  the 
lowering  of  temperature  produced  by  the  addition  of  the  sixty- 
eight  cc.  cold  sulphuric  acid. 

The  following  plan  was  now  adopted  : 

The  sulphuric  acid  was  measured  out  into  beakers  and  also 
heated  on  the  water-bath.  It  was  then  added  to  the  barium 
chloride  solution,  the  beakers  being  washed  out  three  times  with 
hot  water,  using  about  four  or  five  cc.  each  time  and  the  wash- 
ings also  added. 


8oo 


C.   W, 

FOULK. 

Sbribs  vn 

.    (Continued  from  above. ) 

Hydro- 
Barium   chloric 
chloride  J?,  acid. 

Sul. 
phuric 
acid  B. 

BquivalenU 

sulphuric 

acid. 

Barium 
sulphate. 

Srror. 

cc. 

cc. 

6 

50 

5 

67.9 

2.25 

0.193 1 

--O.OQ53 

7 

50 

5 

75.5 

2.50 

0.1935 

—0.0049 

8 

50 

5 

83.0 

2.7s 

0.1956 

—0.0028 

9 

50 

5 

90.5 

3.00 

0.1963 

— 0.0021 

lO 

50 

5 

39-4 

4.00 

0.196 I 

—0.0023 

II 

50 

5 

49.2 

5.00 

0.1962 

—0.0022 

Note. — The  last  two  results  were  obtained  with  a  stronger 
sulphuric  acid  solution,  which  was  run  in  cold. 

A  rapid  increase  is  seen  with  the  first  additions  of  sulphuric 
acid,  the  difference  becoming  less  as  the  sulphuric  acid  increases. 

Another  peculiarity  was  also  seen  in  each  one  of  these  series. 
Although  the  solutions  had  been  well  stirred  on  bringing  the 
reagents  together,  had  settled  clear  in  a  few  minutes,  and  the 
supernatant  liquid  had  remained  clear,  yet  in  running  through, 
the  filter  the  filtrates  soon  became  cloudy  and  a  copious  pre- 
cipitate of  barium  sulphate  settled  out. 

This  could  be  due  only  to  the  agitation  produced  by  running 
through  the  filter.  Later  an  experiment  was  tried  on  this  point. 
Fifty  cc.  of  barium  chloride  solution,  0.0992  barium  sul- 
phate -f-  five  cc.  hydrochloric  acid  and  water  to  make  the  total 
volume  up  to  150  cc,  was  heated  in  a  flask  and  two  equivalents 
of  sulphuric  acid  added.  This  was  then  shaken  for  ten  minutes, 
allowed  to  settle  for  fifty  minutes,  and  then  the  precipitate  was 
filtered  off  and  weighed. 

It  gave  barium  sulphate  0.1979,  a  minus  error  of  0.0013  as 
against  an  error  of  — 0.0127  in  Series  VI,  with  two  equivalents. 

It  seems  that  in  the  presence  of  hydrochloric  acid  unless  there 
is  a  sufficient  amount  of  sulphuric  acid  present  to  effect  com- 
plete precipitation,  a  delicate  balance  is  formed  which  is  affected 
by  a  difference  in  time  of  standing,  in  temperature,  and  amount 
of  agitation  on  stirring.  To  avoid  adding  so  large  a  volume  of 
sulphuric  acid  solution  **  C"  was  prepared. 

Series  VIII  was  now  run.  Both  solutions  wer^  heated  on  the 
water-bath  and  brought  together  as  described  above.     Solution 


PRBCIPITATION  OP   BARIUM   SUI^PHATB.  8oi 

in  each  case  was  stirred  one  and  one-half  minutes  and  allowed 
to  settle  one  hour. 


Series  VIII. 

Hydro- 
Barium    chloric 
chloride^,   acid. 

Snl-     Equivalents 
phuric    sulphuric 
acid  C.        acid. 

Barium 
sulphate. 

Error. 

cc.           cc. 

cc. 

I 

50             5 

20.4             4 

O.I971 

—0.0013 

2 

50            5 

255           5 

0.1978 

—0.0006 

3 

50            5 

30.6           6 

O.I981 

—0.0003 

4  • 

50            5 

35-7           7 

0.1980 

—0.0004 

5 

50            5 

40.8           8 

0.1984 

0.0000 

6 

50            5 

45-9            9 

0.1985 

+0.0001 

7 

50'         5 

51.0          10 

0.1984 

0.0000 

8 

50            5 

56.1          II 

0.1985 

+O.OOOI 

At  last  the  proper  excess  to  effect  complete  precipitation  un- 
der the  conditions  described  above  had  been  found.  Seven  or 
eight  times  the  theoretical  amount  seems  necessary.  It  is  to  be 
noted  that  the  change  is  extremely  slow  when  near  the  critical 
point. 

A  short  series  was  precipitated  and  weighed,  using  other 
solutions,  the  equation  of  which  was  as  follows  : 

Fifty  cc.  BaCl,,  ;r  =  i  cc  ±:  H,SO,  D  =  0.1992  BaSO,. 


Series  IX. 

Barium 
chloride  x. 

Hydro, 
chloric 
.    acid. 

Sul. 
phuric 
acid  D. 

Equiva- 
lents sul- 
phuric acid. 

Barium 
sulphate. 

Error. 

cc. 

cc. 

cc. 

I 

50 

5 

3 

3 

0.1957 

--O.OO35 

2 

50 

5 

4 

4 

0.1992 

0.0000 

3 

50 

5 

5 

5 

0.1983 

—0.0009 

3 

50 

5 

5 

5 

0.1992 

0.0000 

In  this  series  the  sulphuric  acid  was  run  in  cold. 

The  maximum  amount  of  precipitate  seems  to  be  reached  here 
with  less  sulphuric  acid  than  when  a  more  dilute  solution  was 
used.  The  same  is  true  of  the  precipitations  made  in  the  pres- 
ence of  ten  cc.  hydrochloric  acid. 

Series  X  was  now  run,  the  precipitations  being  made  in  ex- 
actly the  same  manner  as  those  of  Series  VIII,  except  that  ten 
cc.  of  hydrochloric  acid  was  put  into  the  solutions  instead  of 
five  cc. 


8o2 


C.    W. 

FOUI.K. 

Series  X. 

Bariam 
chloride  B. 

Hydro- 
chloric 
acid. 

Sol- 
phuric 
acidC 

Bqaiva- 

lenU  lal- 

pharic 

acid. 

Bariam 
sulphate. 

BiTor. 

cc. 

cc. 

I 

SO 

10 

30.6 

6 

0.1964 

^-0.0020 

2 

50 

10 

35-7 

7 

0.1974 

— O.OOIO 

3 

50 

10 

40.8 

8 

O.1981 

—0.0003 

4 

50 

10 

45.9 

9 

0.1982 

^-0.0002 

3 

50 

10 

40.8 

8 

O.X970 

— O-OOI4 

4 

50 

10 

45-9 

9 

0.1982 

^-0.0002 

In  the  presence  of  ten  cc.  hydrochloric  acid  then,  a  somewhat 
greater  excess  of  sulphuric  acid  is  required  than  with  five  cc. 
hydrochloric  acid. 

A  short  series  with  the  stronger  solution  gave 


Barium 
chloride  J?. 

cc. 

1  50 

2  50 

3  50 
3  50 


Series  XI. 

Eqniva> 
Hydro>       Sul-     lenta  snl- 
chloric    phuric    phuric  Barium 

acid.       acidZ).    acid.  sulphate. 

cc. 

10  5  5  0.1975 

10  6  6  0*1982 

10  7  7  0.1992 

10  7  7  0.1991 


Srror. 

0017 
0010 
O'OOOO 

0001 


Series  XII.  was  conducted  exactly  as  Nos.  VIII.  and  X.,  ex- 
cepting that  fifteen  cc.  hydrochloric  acid  was  used. 


Hydro- 
Barium     cnloric 
chloride  B.    acid. 


Series  XII. 

Eqniva- 
Sul-         lents  sul- 
phuric       phuric 
acid  C.  acid. 


I 

2 

■ 

3 
4 

5 
6 

7 


cc. 

50 

50 

50 

50 

50 

50 

50 


cc. 

15 
15 
15 
15 
15 
15 
15 


30.6 

35-7 
40.8 

45-9 
510 
56.1 

61.2 


6 

7 
8 

9 
10 

II 

12 


Barium 
sulphate. 

0.1957 

0.1955 
0.1965 

0.1973 
0.1972 

0.1984 
0.1983 


Error. 

0027 

0029 

0019 

.0011 

.0012 

0.0000 

.0001 


The  point  to  be  noted  in  this  series  is  that  more  sulphuric 
acid  is  required  in  the  presence  of  the  larger  amount  of  hydro- 
chloric acid. 

The  other  side  of  the  question  was  now  taken  up,  namely,  the 
precipitation  of  sulphuric  acid  with  an  excess  of  barium  chlo- 
ride in  the  presence  of  hydrochloric  acid. 


PRECIPITATION  OF  BARIUM   SULPHATE.  803 

A  new  difficulty  at  once  presented  itself.  The  old  trouble  in 
filtering  barium  sulphate  was  experienced.  When  a  small 
amount  of  hydrochloric  acid  was  present  it  was  found  utterly 
impossible  to  do  it  under  the  conditions  which  had  previously 
been  followed. 

Various  experiments  were  made  to  avoid  this  trouble  and  at 
last  the  following  scheme  was  adopted : 

The  volume  was  kept  at  150  cc.  as  in  the  other  work.  The 
sulphuric  acid,  water,  and  hydrochloric  acid  were  heated  on  the 
water-bath  and  the  barium  chloride  solution,  also  hot,  was  added 
drop  by  drop  with  constant  stirring.  The  beakers  were  then  set 
back  on  the  battf  and  the  solutions  stirred  at  intervals  for  thirty 
minutes.  They  were  then  set  off  and  stirred  at  intervals  again 
until  cold. 

Just  before  pouring  upon  the  filter  the  precipitate*  was  stirred 
up  and  the  filter  filled  several  times.  At  first  a  small  portion 
ran  through,  but  this  was  poured  back,  and,  generally,  the  rest 
could  be  filtered  without  trouble. 

A  series  was  run  according  to  this  description,  except  that  the 
volume  was  250  cc. 

The  exact  time  of  standing  before  filtering  was  not  noted  in 
this  case.     It  was  probably  about  four  or  five  hours. 


Sbribs  XIII. 

Sul- 
phuric 
acid^. 

Hydro- 
chloric 
acid. 

Eqivalents 
Barium      barium 
chloride  C.  chloride. 

Barium 
sulphate. 

Error. 

cc. 

cc. 

cc. 

I 

30.2 

10 

14.9 

3 

0.1967 

—0.0017 

2 

30.2 

10 

19.8 

4 

0-1959 

—0.0025 

3 

30.2 

xo 

14.9 

5 

0.1965 

^.0019 

There  was  nothing  satisfactory  to  be  derived  from  this  series, 
so  Series  XIV  was  run.  The  volume  here  was  kept  down  to 
150  cc. 


Series  XIV. 

Sul- 
phuric 
acid^. 

Hydro- 
chloric 
acid. 

Houivalents 
Barium     barium 
chloride  C.  chloride. 

Barium 
sulphate. 

Error. 

cc. 

cc. 

cc. 

I 

30.2 

10 

14.9 

3 

0.1975 

—0.0009 

2 

30.2 

10 

19.8 

4 

0.1994 

-f  O.OOIO 

3 

30.2 

10 

24.9 

5 

0.1983 

— O.OOOI 

4 

30.2 

10 

29.8 

6 

0.2005 

-H>.002I 

804  C.   W.    FOULK. 

Two  of  these  precipitates  weigh  heavier  than  theory  demands. 
This  could  come  only  from  contamination  with  barium  chloride. 
To  test  this  No's  3  and  4  were  transferred  to  beakers,  boiled  up 
with  about  seventy-five  cc.  of  water  and  again  filtered,  ignited, 
and  weighed. 

They  then  gave 


3.  0.1980  barium  sulphate  =  — o.cxx34  error. 

4.  0.1987      **  **         as  +0.0003     " 

Another  series  was  run  in  exactly  the  same  manner,  except 
that  more  care  was  taken  in  washing.  Each  precipitate  was 
washed  with  boiling  water  until  the  filtrate  no^  longer  reacted 
with  silver  nitrate. 

Sbriks  XV. 


Sul- 
ptauric 
acid^. 

Hsrdro- 

chloric 

acid. 

Barium 
chlofide 

Bquivalents 

Darium 
C  chloride. 

Barium 
sulphate. 

Brror. 

cc. 

cc. 

I 

30.2 

10 

9-9 

2 

0.1965 

—0.0019 

2 

30.2 

10 

14.9 

3 

0.1982 

— O.OOOI 

3 

30.2 

10 

19.8 

4 

0.1975 

—0.0008 

4 

30.2 

10 

24.9 

5 

0.1988 

+0.0004 

5 

30.2 

10 

29.8 

6 

0.1994 

+0.0010 

No*s  4  and  5,  on  being  boiled  up  with  water  and  reweighed, 
gave 

3.  o.  198 1  =  —0.0003  error. 

4.  0-1975  =  —0.0009     '* 

At  its  best,  however,  this  method  of  working  was  unsatisfac- 
tory. The  precipitate  seemed  always  on  the  point  of  running 
through  the  filter  and  indeed  traces  generally  did  go  through. 

The  following  scheme  was  accordingly  tried  in  Series  XVI. 

The  volume  was  kept  as  before  at  150  cc,  but  thirty  cc.  of 
hydrochloric  acid  instead  of  ten  was  put  into  each  solution. 
The  precipitates  were  not  stirred  up  after  being  thrown  down. 
In  presence  of  this  large  excess  of  acid  the  precipitates  soon 
became  coarse  and  crystalline  and  settled  rapidly.  No  troubU 
whatever  was  experienced  in  filtering  them. 


PRBCIPITATION  OP  BARIUM   SULPHATE.  805 

Sbriss  XVI.    (These  had  stood  about  four  hours. ) 


Sul- 
phuric 
acid^. 

Hydro- 
chloric 
acid. 

BquiTalents 
Barium       Darium 
chloride  C.  chloride. 

Barium 
sulphate. 

Brror. 

cc. 

cc. 

I 

20 

30 

4.6 

1.5 

0.0947- 

—0.0289 

2 

20 

30 

6.2 

2.0 

0.0997 

—0.0239 

3 

20 

30 

9.3 

3.0 

0.III4 

—0.0122 

4 

ao 

30 

12.4 

4.0 

O.I 169 

—0.0067 

5 

20 

30 

15.5 

5.0 

0.1207 

—0.0029 

6 

20 

30 

18.6 

6.0 

O.I  192 

—0.0044 

These  filtrates,  on  standing  over  night,  all  showed  further 
precipitates  of  barium  sulphate.  The  series  was  .accordingly 
continued,  this  time  the  solutions  standing  about  seven  hours 
before  being  filtered. 

The  precipitates  were  crystalline  and  filtered  easily  and 
rapidly  and  the  filtrates,  on  further  standing,  showed  no  traces  of 
barium  sulphate. 

In  spite,  however,  of  the  greatest  care  in  washing,  it  was  im- 
possible to  get  rid  of  the  occluded  barium  chloride  before  igni- 
tion. 

Sbribs  VII. 


Sul- 
phuric 
acid£. 

Hydro- 
chloric 
acid. 

Barium 

chloride 

C. 

BquivVs 

Barium 

chloride. 

Barium 
sulphate. 

Error. 

I 

20 

30 

18.6 

6 

0.1258 

-I-0.0022 

2 

20 

30 

21.7 

7 

0.1252 

4-0.0016 

3 

20 

30 

24.8 

8 

0.1268 

+0.0032 

4 

20 

30 

27.9 

9 

0.1260 

-f"O.0O24 

Nos.  2,  3  and  4  were  boiled  up  with  water,  re  washed,  ignited 
and  weighed. 

Barium  sulphate.        Error. 

2  0.1236  0.0000      Filtrate  reacted  strongly  with  silver  nitrate. 

3  0.1238  +0.0002 

4  0.1253  +O.OOI7  *'  *•        slightly     " 

No.  4  boiled  up  the  second  time  0.1237  BaSO^  =  +  0.0001 
error.  The  filtrate  in  this  case  reacted  strongly  with  silver 
nitrate. 

To  test  this  boiling  up  process  No.  4  was  treated  the  third 
time.  This  time  the  precipitate  weighed  0.1235  and  the  filtrate 
did  not  react  with  silver  nitrate. 

A  last  experiment  was  made  to  determine  the  effect  of  barium 


8o6  C.   W.   POULK. 

chloride  upon  the  direct  solubility  of  barium  sulphate  in  hydro- 
chloric acid. 

0.1248  {^am  barium  sulphate  was  put  in  120  cc.  water  and 
thirty  cc.  hydrochloric  acid  and  beaker  marked  **^." 

o.  1228  gram  barium  sulphate  was  weighed  into  another  beaker 
with  105  cc.  water,  thirty  cc.  hydrochloric  acid  and  fifteen  cc. 
barium  chloride  C.     This  beaker  was  marked  ''^.'* 

Both  were  heated  on  the  water- bath  with  frequent  stirring  and 
then  stood  over  night. 

On  being  filtered  and  weighed, 

**  w4  "  gave  —0.1 106  bftrium  snlphate  =  0.0142  loss. 
•«^»»    •«     —0.1228      "  *•       s  0.000      •• 


To  further  test  precipitate  '*^  "  it  was  boiled  up  with  water 
as  those  of  Series  XVII,  and  re-weighed.  It  lost  by  this  oper- 
ation 0.0002,  which  is  practically  nothing. 

Prom  the  results  obtained  in  this  investigation  the  following 
conclusions  seem  justified : 

(i)  In  the  precipitation  of  a  barium  salt  with  sulphuric  acid 
in  the  presence  of  hydrochloric  acid,  a  ^ery  large  excess  of  sul- 
phuric acid  is  required. 

(2)  This  excess  should  be  greater  the  greater  the  amount  of 
hydrochloric  acid  present  in  the  solution. 

(3)  It  should  be  greater  the  shorter  the  time  of  standing 
before  filtration.  In  fact  a  very  great  excess  seems  to  effect 
immediate  precipitation.' 

(4)  The  greater  the  excess  of  sulphuric  acid  the  less  stirring 
seems  necessary  to  bring  down  the  precipitate  in  a  given  time. 

(  5  )  While  barium  sulphate  obtained  by  precipitating  a  barium 
salt  with  sulphuric  acid  in  the  presence  of  hydrochloric  acid  is 
coarse,  crystalline  and  easily  filtered/  that  obtained  by  precipi- 
tating sulphuric  acid  with  a  barium  salt  in  the  presence  of 
hydrochloric  acid  is  fine  and  much  disposed  to  run  through  the 
filter  unless  special  precautions  are  taken. 

(6)  In  general  a  large  excess  of  barium  chloride  is  required 
to  completely  precipitate  the  sulphuric  acid  in  the  presence  of 
hydrochloric  acid. 


PRECIPITATION  OF  BARIUM   SULPHATE.  807 

(7)  As  the  hydrochloric  acid  increases  the  amount  of  barium 
chloride  should  also  be  increased. 

(8)  The  greater  the  amount  of  hydrochloric  acid  present  the 
coarser  and  more  crystalline  in  character  is  the  precipitated 
barium  sulphate.  In  precipitating  in  the  presence  of  large 
amounts  of  hydrochloric  acid  the  solution  should  be  quite  con- 
centrated. 

(9)  The  barium  sulphate  so  obtained,  will,  however,  be  con- 
taminated with  adhering  barium  chloride,  and  no  amount  of 
washing  before  ignition  can  entirely  free  it  from  this  occluded 
chloride.  If,  after  ignition,  the  precipitate  be  boiled  up  with 
water,  again  washed,  ignited  and  weighed,  and  this  process  be 
continued  until  a  constant  weight  is  obtained,  the  sulphate  may 
be  entirely  freed  from  the  barium  salt. 

Some  subsequent  work  in  this  line  has  shown  that  heavy  pre- 
cipitates sometimes  require  three  or  four  treatments  before  a  con- 
stant weight  is  obtained. 

(10)  Both  in  the  precipitation  of  barium  with  sulphuric  acid 
and  of  sulphuric  acid  with  barium,  very  concordant  results  may 
.be  obtained  if  the  conditions  under  which  the  precipitations  are 
made  are  similar,  but  these  results  may  be  quite  far  from  cor- 
rect. A  note  of  this  commonplace  occurrence  in  analytical 
work  is  made  here,  because  by  following  the  usual  method  of 
testing  the  filtrate  for  an  excess  of  the  precipitating  reagent,  a 
strong  reaction  might  be  obtained  and  yet  not  more  than  ninety 
per  cent,  of  the  original  precipitate  be  down. 

In  conclusion  I  wish  to  express  my  thanks  to  Professor  N.  W. 
Lord  for  helpful  suggestions  during  the  course  of  this  work. 

Discussion.' — ^T.  S.  Gladding:  I  have  already  shown  (see 
this  Journal,  16,  398;  17*  181,  397,  772;  18,  446)  that  correct 
results  may  be  obtained  if  the  barium  chloride  solution  be  added 
drop  by  drop  instead  of  all  at  once.  This  is  confirmed  by  I^ane 
(18,  682)  and  is  now  virtually  admitted,  by  Lunge,  who  precipi- 
tates (18,686)  by  *' quick  additions  (1.  r,  pouring  in  the  hot 
barium  chloride  solution  in  about  ten  portions,  occupying  about 
half  a  minute  in  all,  and  stirring  the  mixture  all  the  time,  as 
every  chemist  would  do.)** 

1  Bttfialo  McctiniTi  Aug.,  1896. 


THE  ACTUAL  ACCURACY  OF  CHEHICAL  ANALYSIS.* 

BV  PRBOSftIC  P.  DSWBV. 
Received  July  14.  it96. 

THE  subject  of  this  paper  does  not  embrace  the  consideration 
of  ways  and  means  for  the  increase  of  analytical  accu- 
racy»  or  the  question,  what  can  be  or  should  be  attained  in  that 
direction.  I  desire  simply  to  call  attention  to  the  degree  of 
accuracy  exhibited  in  actual  ever>'  day  practice.  In  estimating 
this,  little  weight  will  be  given  to  the  evidence  afforded  by  the 
agreement  of  duplicate  or  multiple  determinations  by  the  same 
chemist ;  for  I  am  convinced  that  such  agreement  is  a  delusion 
and  a  snare.  Nor  will  special  importance  be  attached  to  the 
agreement  of  two  or  even  three  analysts  in  special  cases,  or  to 
the  agreement  between  two  methods  practiced  by  the  same 
analyst.  I  propose  to  compare  the  results  obtained  by  several 
chemists,  working  upon  the  same  sample  and  by  various 
methods,  in  order  to  exhibit,  as  I  have  said,  the  actual  condi- 
tion of  practice. 

The  available  material  for  illustrating  this  phase  of  the  ques- 
tion is  unfortunately  scanty  ;  but  something  has  been  done  ; 
and  I  hope,  by  calling  attention  to  some  of  the  work  in  this  line, 
to  stimulate  further  work  in  the  same  direction  by  inducing 
others  to  prepare  suitable  samples  and  submit  them  to  various 
chemists  who  are  competent  and  willing  to  make  the  necessary 
determinations  and  fully  describe  the  methods  they  employ. 

I  draw  most  of  my  illustrations  from  the  * '  Transactions  of  the 
American  Institute  of  Mining  Engineers,''  the  ** Proceedings  of 
the  Association  of  Official  Agricultural  Chemists,"  and  from 
personal  experience. 

MANGANESE   IN   STEEL. 

In  May,  1881,  Mr.  William  Kent  presented  a  paper  to  the 
American  Institute  of  Mining  Engineers  entitled  *'  Manganese 
Determinations  in  Steel,"  *  in  which  he  gave  twenty-four  deter- 
minations of  manganese,  made  by  ten  different  chemists, 
employing  two  main  methods,  on  samples  from  a  plate  of  steel. 

1  Read  before  the  Washingtoa  Section  of  the  American  Chemical  Society.  May  X4tli, 
1896.  and  published  jointly  with  the  American  Institute  of  Mining  Engineers. 
3  Trans.  A.  I.  M.  E-.  xo.  loi. 


ACCURACY   OP  CHEMICAL   ANALYSIS.  809 

These  results  presented  the  remarkable  range  of  from  1.14  to 
^•3<^3  percent.,  and  one  chemist  reported  results  ranging  from 
1. 14  to  0.434  per  cent. 

A  portion  of  this  variation  was  undoubtedly  due  to  variations 
in  the  sample,  since  the  same  sample  was  not  used  throughout 
by  the  different  chemists. 

Throwing  out  the  anomalous  result  of  1.14  per  cent,  we  have 
twenty-three  determinations  running  from  0.619  P^r  cent,  to 
0.303  per  cent.,  with  an  average  of  0.415  per  cent.  Thus  show- 
ing that  at  that  time  the  determination  of  manganese  in  steel, 
when  only  about  four- tenths  per  cent,  was  probably  present, 
might  exhibit  an  extreme  variation  between  the  highest  and  the 
lowest  results  of  about  three-tenths  per  cent.,  or  seventy-five  per 
cent,  of  the  amount  of  manganese  present. 

These  results  were  certainly  very  discouraging  ;  but  if  they 
did  nothing  else  they  served  to  call  attention  to  the  very  unsat- 
isfactory character  of  the  determination  of  manganese  in  steel  at 
that  time. 

I  do  not  recall  any  recent  symposium  on  the  determination  of 
manganese  in  this  class  of  material,  but  in  1886  Capt.  A.  £. 
Hunt,*  in  giving  a  measure  of  the  accuracy  of  the  colorimetric 
method,  speaks  of  a  variation  of  0.02  per  cent,  in  steels  contain- 
ing 0.15  to  one  and  five-tenth's  per  cent,  of  manganese 
as  •' sufficiently  accurate  for  all  practical  w^ork,*'  thus  clearly 
intimating  that  the  current  results  of  analj'sis  by  other  methods 
were  at  least  as  good.  This  degree  of  accuracy,  if  attained  by 
different  chemists  upon  the  same  sample,  must  be  considered  a 
satisfactory  advance  over  the  results  reported  by  Mr.  Kent. 

Early  in  1883  Mr.  G.  C.  Stone  began  a  series  of  contributions 
on  the  **  Determination  of  Manganese  in  Spiegel.'**  In  his  first 
paper  he  reported  thirteen  determinations  by  five  chemists,  all 
working  upon  the  same  ** works'*  sample,  showing  from  15.49 
to  13.83  per  cent.,  and  also  twenty-six  determinations  by  ten 
chemists,  all  working  upon  a  sample  of  the  same  spiegel,  pre- 
pared with  especial  care  jointly  by  Mr.  Stone  and  one  of  the 
other  chemists,  showing  from  14.56  to  10.36  percent.  But  some 
of  the  low  results  were  obtained  by  experimental  methods. 

1  Trans.  A.  I.  M.  E.,  I5t  104. 

STrans.  A.  I.  M .  K..  11, 323  :  12,  295  and  514. 


8lO  PRBDERICK   P.    DEWEY. 

In  the  fall  of  1883  Mr.  Stone  reported  twenty  additional 
determinations  by  five  other  chemists,  ranging  from  14.20  to 
10.76  per  cent. ;  the  extremes  being  reported  by  the  same  chemist 
when  working  by  different  methods,  his  favorite  method  giving 
from  13.84  to  13.65  per  cent.,  and  three  low  results,  less  than 
eleven  per  cent.,  being  obtained  by  the  Williams'  method.  In 
this  connection  Mr.  Stone  presented  an  interesting  table,  divid- 
ing the  methods  used  into  four  classes  and  the  results  into  three 
classes,  giving  respectively,  below  thirteen  per  cent.,  between 
thirteen  and  fourteen  per  cent.,  and  above  fourteen  per  cent. 

In  the  spring  of  1884  Mr.  Stone  reported  twenty -seven  new 
results,  nineteen  by  four  new  chemists,  and  eight  by  one  pre- 
viously reported,  whose  new  results  were  obtained  by  several 
methods. 

We  have  thus  seventy-three  determinations  by  nineteen  differ- 
ent chemists.  Of  these  two  are  thrown  out  on  account  of  the 
method  used,  and  eleven  **  because  the  chemists  were  not 
entirely  satisfied  with  them,"  leaving  sixty  determinations  by 
eighteen  chemists,  using  twelve  methods. 

These  sixty  results  range  from  14.47  to  12.60  per  cent.,  and 
average  13.39  per  cent .  Leaving  out  eight  determinations  by  one 
method  which  is  considered  to  give  low  results,  the  lowest 
determination  becomes  12.92  per  cent,  and  the  average  13.48 
percent.,  showing  an  extreme  variation  of  1.45  per  cent,  of 
manganese  between  the  highest  and  lowest  results,  and  showing 
only  forty-four  per  cent,  of  the  results  within  two-tenths  per  cent, 
of  the  average. 

In  the  discussion  of  Mr.  Stone's  second  paper,  Mr.  J.  B. 
Mackintosh'  presented  an  analysis  of  Mr.  Stone's  first  forty-six 
results,  retaining  the  results  by  the  Williams'  method,  from 
which  he  argued  that  the  evidence  pointed  to  12.956  per  cent,  as 
the  true  content  of  manganese  in  this  Spiegel.  If  this  is  the 
case,  then  there  is  a  very  decided  tendency  to  get  too  high 
resultb  in  this  class  of  work. 

Taken  as  a  whole,  this  investigation  would  seem  to  show  that 
variations  of  five-tenths  per  cent,  in  the  determination  of  man- 
ganese in  this  grade  (ten  to  fifteen  per  cent,  manganese)  of 

I  Trans.  A.  I.  M.  B..  ta,  300. 


ACCURACY   OP   CHEMICAI,  ANALYSIS.  8ll 

Spiegel  are  to  be  expected,  and  much  wider  variations  may  be 
found. 

PHOSPHORUS   IN   PIG   IRON. 

Early  in  the  8o*s,  Messrs.  Potter  and  Riggs,  of  St.  Louis, 
Mo.,  sent  out  a  sample  of  pig-iron  for  the  determination  of 
phosphorus. 

This  examination  yielded  twenty-six  results,  by  eleven  chem- 
ists, using  five  methods,  ranging  from  0.181  to  0.141  per  cent., 
and  averaging  0.160  per  cent,  and  showing  an  extreme  varia- 
tion of  0.040  per  cent.  The  maximum  variation  reported  by 
any  one  chemist  was  0.017  P^r  cent.,  while  three  reported  dupli- 
cated agreeing  with  o.ooi  per  cent.  These  results  have  never 
been  published.  One  of  the  chemists  discovered  arsenic  in  the 
sample,  which  would  account  for  some  of  the  variation  in  the 
series.  His  determinations  in  duplicate  were  0.151  and  0.152 
per  cent. 

In  February,  1882,  Mr.  F.  E.  Bachman  presented  a  paper  to 
the  American  Institute  of  Mining  Engineers,*  in  which  he 
reported  forty-four  results  by  eighteen  chemists,  using  four 
methods,  ranging  from  0.165  to  0.096  per  cent,  and  averaging 
0.143  per  cent.  The  extreme  variation  was  0.069  P^r  cent. 
The  maximum  variation  reported  by  any  one  chemist  on  straight 
duplicates  was  o.oi  per  cent.,  and  the  minimum  0.0004  P^r  cent. 
Experimental  determinations  by  Mr.  Bachman,  using  different 
processes,  yielded  variations  amounting  to  0.043  per  cent. 

At  the  Atlanta  meeting  in  October,  1895,  Mr.  Geo.  Thackray 
presented  a  paper,  entitled  ' '  A  Comparison  of  Recent  Phos- 
phorus Determinations  in  Steel.*"  He  first  gives  a  table  of 
determinations  of  phosphorus  by  two  chemists  on  eight  samples 
ranging  from  0.033  to  0.012  per  cent.,  one  chemist  uniformly 
getting  high  results.  One  chemist  found  from  0.080  to  0.074 
per  cent.,  and  the  other  0.1 16  to  0.088  per  cent  in  these  steels. 
These  results  were  manifestly  unsatisfactory. 

A  second  table  shows  results  by  three  chemists,  the  buyer's, 
the  seller's  and  an  arbitrator.  By  the  arbitrator's  determinations 
these  steels  carried  from  0.080  to  0.063  P^>^  cent,  of  phosphorus. 

1  Trans.  A.  I.  M.  £.,  xo,  33a. 
f  Trans.  A.  I.  M.  B.,  as,  370. 


8l2  FREDERICK   P.  DEWEY. 

The  maximum  difference  in  any  set  of  three  results  "was 
0.017  percent.,  and  the  minimum  0.005  per  cent. 

These  results  were  obtained  in  the  settlement  of  sales.  As  a 
result  of  the  discussion  which  accompanied  the  matter,  two  sam- 
ples of  steel  were  prepared  and  sent  to  various  chemists.  A 
fourth  table  gives  thirty-six  results  obtained  from  twenty-three 
chemists,  using  twenty-nine  methods  on  one  steel,  showing  re* 
suits  averaging  0.0496  per  cent.,  and  ranging  from  0.055  too.045 
percent.,  an  extreme  variation  of  only  o.oio  per  cent.  Any- 
individual  result  was  practically  within  0.005  per  cent,  of  the 
average. 

On  the  second  sample  thirty-eight  results  were  reported  avtrag^- 
ing  0.0835  per  cent.,  and  ranging  from  0.091  to  0.076  per  cent., 
an  extreme  variation  of  0.015  per  cent. 

My  own  results  on  these  steels  are  not  given,  as  they  were  not 
reported  in  time ;  but  they  add  two  more  results  by  one  more 
chemist  in  each  case,  and  the  results  fall  within  the  limits. 

These  results  must  be  regarded  as  highly  satis&u:tory,  and 
show  that  here,  at  least,  is  one  determination  that  can  be  made 
by  many  chemists,  working  in  different  ways,  and  yet  with 
results  agreeing  ver>-  closely  together.  While  it  may  not  be 
necessary*  to  determine  many  things  as  closely  as  phosphorus  in 
steel,  yet  it  would  be  highly  satisfactory  if  we  could  do  so ;  and 
this  is  a  good  standard  of  excellence  for  us  to  aim  at. 

PHOSPHORIC   ACID. 

As  compared  with  the  accuracy  secured  in  the  determination 
of  phosphorus  in  steel,  the  1S94  report  of  the  Association  of 
Official  Agricultural  Chemists,*  shows  that  on  one  sample  thirty- 
nine  determinations  of  insoluble  phosphoric  acid  by  eighteen 
chemists,  working  by  the  oflBcial  method,  showed  results  rang- 
ing from  0.45  to  0.03  per  cent.,  with  an  average  of  0.27  per 
cent.,  the  extreme  variation  being  0.42  per  cent.,  or  over  one 
and  one-half  times  the  average  determination. 

By  another  method,  on  the  same  sample,  thirty-six  determina- 
tions by  nineteen  chemists  showed  results  varying  from  0.34  to 

1  PToce«^.ai;s  of  :hc  HTevrnth  Annua!  Cocveotion  of  the  Associatioa  of  HIKi  ial  Acii- 
cii!tQraI  Chrcu<:5.  Ancust  73.  24.  ^  *'>S4  Bu;:mn  4k.v  f.  S^  Dgprinif  ■?  of  A^rtcwltBrT. 
Division  of  Chemistry,  p.  "^ 


ACCURACY  OF   CHEMICAL  ANAI^YSIS.  813 

0.04  per  cent.,  with  an  average  of  0.19  per  cent.,  the  extreme 
variation  being  0.30  per  cent.,  or  over  one  and  one-half  times 
the  average. 

We  have  thus  seventy-five  determinations  by  nineteen  chem- 
ists working  by  two  methods,  showing  results  ranging  from  0.45 
to  0.03  per  cent.,  with  an  average  of  0.233  P^r  cent.,  the 
extreme  variation  being  0.42  per  cent.,  or  nearly  twice  the 
average  determination. 

On  another  sample  thirty-three  determinations  by  seventeen 
chemists  working  by  the  official  method,  showed  results  rang- 
ing from  3.85  to  2.24  per  cent.,  with  an  average  of  2.82  per 
cent.,  the  extreme  variation  being  1.61  percent.,  or  considerably 
more  than  one-half  of  the  average. 

By  another  method,  on  the  same  sample,  tliirty-five  determi- 
nations by  seventeen  chemists  showed  results  ranging  from 
3.49  to  2.18  per  cent.,  with  an  average  of  2.83  per  cent.,  the 
extreme  variation  being  1.31  per  cent.,  or  nearly  one-half  the 
average. 

Summing  up  again,  we  have  sixty-eight  determinations  by 
eighteen  chemists  working  by  two  methods,  showing  results 
ranging  from  3.85  to  2.18  percent.,  with  an  average  of  2.82  per 
cent.,  the  extreme  variation  being  1.67  per  cent. 

The  same  report*  shows  that  on  one  sample  the  results  of 
twenty-nine  determinations  of  citrate  soluble  phosphoric  acid  by 
fourteen  chemists,  by  the  direct  method  of  Ross,  varied  from 
2.47  to  1.04  per  cent.,  with  an  average  of  1.52  per  cent.,  the 
extreme  variation  being  1.43  per  cent.,  or  nearly  equal  to  the 
average  of  all  the  determinations. 

On  the  same  sample,  by  the  official  method,  the  results  of 
twenty-three  determinations  by  fourteen  chemists  ranged  from 
2.26  to  1. 18  percent.,  with  an  average  of  1.46  per  cent.,  the 
extreme  variation  being  1.08  per  cent.,  or  over  two-thirds  of 
the  average  determination. 

Summing  up.  we  have  fifty-two  determinations  by  fourteen 
chemists  working  by  two  methods,  ranging  from  2.47  to  1.04 
percent.,  and  averaging  1.49  per  cent.,  the  extreme  variation 
being  1.43  per  cent.,  or  nearly  equal  to  the  average. 

1  Ibid,^  P-  72. 


8 14  FREDERIC   P.    DEWEY. 

On  another  sample  thirty-six  determinations  by  fifteen  chem- 
ists by  the  direct  method  of  Ross,  range  from  3.29  to  1.87  per 
cent.,  with  an  average  of  2.36  per  cent.,  the  extreme  variation 
being  1.42  per  cent.,  or  considerably  over  one-half  of  the  average 
determination. 

On  the  same  sample,  twenty-four  determinations  by  fifteen 
chemists,  ranged  from  3.40  to  2.08  per  cent.,  with  an  average  of 
2.60  percent.,  the  extreme  variation  being  1.32  per  cent.,  or  a 
little  over  one-half  of  the  average  determination. 

Summing  up,  we  have  sixt}'  determinations  by  fifteen  chemists 
working  by  two  methods,  ranging  from  3.40  to  2.08  per  cent., 
and  averaging  2.44  per  cent.,  the  extreme  variation  being  1.32 
per  cent.,  or  a  little  over  one-half  of  the  average  determinations. 

In  the  determination  of  the  total  phosphoric  acid,'  fort>'-five 
determinations,  by  eighteen  chemists,  ranged  from  20.67  to  19.74 
per  cent.,  with  an  average  of  20.09  per  cent.,  the  extreme  varia- 
tion being  0.93  per  cent.  By  a  volumetric  method,  thirty  deter- 
minatiqns,  by  eleven  chemists,  ranged  from  20.60  to  19.83  per 
cent.,  with  an  average  of  20.14  per  cent.,  the  extreme  v*ariation 
being  0.77  per  cent.  By  another  volumetric  method,  twenty-one 
determinations  by  ten  chemists,  ranged  from  20.45  ^o  19-27  per 
cent.,  with  an  average  of  19.96  per  cent.,  the  extreme  variation 
being  1.18  per  cent. 

Combining  these  results,  we  have  ninety-six  determinations 
by  eighteen  chemists  working  by  three  methods,  ranging  from 
20.67  to  19.27  per  cent.,  with  an  average  of  20.08  per  cent.,  the 
extreme  variation  being  1.40  per  cent. 

Similarly,  on  another  sample,  we  have  120  determinations,  by 
twenty-two  chemists,  working  by  the  same  three  methods,  rang- 
ing from  1.8. 15  to  16.25  per  cent.,  with  an  average  of  17.26  per 
cent.,  the  extreme  variation  being  1.90. 

Again,  on  another  sample,  we  have  ninety-six  determinations 
by  twenty-one  chemists,  working  by  the  same  three  methods, 
ranging  from  2.85  to  2.20  per  cent.,  with  an  average  of  2.50  per 
cent.,  the  extreme  variation  being  0.65  per  cent. 

COPPER. 

At  the  August  meeting  of  the  A.  I.  M.  E.,  in  1882,  Mr.  W. 

I  Ibtd.,  pp.  Si,  S2.  53. 


ACCURACY  OF   CHEMICAL  ANALYSIS.  815 

E.  C.  Eustis  presented  a  paper  entitled ''  Comparison  of  Various 
Methods  of  Copper  Analysis.*'*  For  the  purpose  of  this  com- 
parison a  ver>'  complex  sample  was  made  up,  containing  sul- 
phides, oxides  and  metallic  copper,  a  silicate,  sulphides  of  iron 
and  zinc,  arsenic  and  nickel.  The  paper  reports  forty-five 
determinations  by  seventeen  chemists,  using  some  eight  methods. 
The  results  showed  a  wide  variation,  ranging  from  53.34  to 
43.92  per  cent,  and  averaging  47.75  per  cent.  On  throwing  out 
a  set  of  six  results  from  one  concern,  all  of  which  were  more 
than  two  per  cent,  and  two  of  them  nearly  five  per  cent,  above 
the  nearest  other  result,  as  being  manifestly  too  high,  and  two 
results  by  one  chemist  and  one  method,  which  were  more  than 
two  per  cent,  below  the  nearest  other  result,  the  series  still 
rang^es  from  48.72  to  46.24  per  cent.,  with  an  average  of  47.23 
per  cent.,  and  a  maximum  variation  of  2.48  per  cent.,  which 
cannot  be  considered  very  satisfactory. 

The  same  paper  reported  seventeen  determina^ions  by  seven 
chemists  on  borings  of  pig  copper.  These  ranged  from  91.07  to 
98.17  per  cent,  and  averaged  94-25  per  cent.  On  throwing  out 
two  results  that  were  nearly  three  per  cent,  higher  than  the 
nearest  other  result,  and  four  that  were  over  three  per  cent. 
below  the  nearest  other  result,  the  series  ranges  from  94.91  to 
94.38  per  cent,  with  an  average  of  94.^9  per  cent.  The  extreme 
variation  of  only  0.53  per  cent,  must  be  regarded  as  very  good 
work,  especially  when  we  consider  the  character  of  the  material. 
At  the  Florida  meeting  in  March,  1895,  the  results  of  a  sym- 
posium on  copper  and  copper  matte,  initiated  bj'  Dr.  A.  R. 
l,edoux,  of  New  York  City,  were  presented.*  Eight  chemists 
reported  the  copper  in  the  matte,  some  in  duplicate  or  more,  as 
determined  by  electrolysis,  as  ranging  from  55.17  to  54.50  per 
cent,  and  averaging  54.91  per  cent.  The  extreme  variation  was 
only  0.67  per  cent ;  and  this  must  be  regarded  as  satisfactory, 
and  very  much  better  than  the  results  on  Mr.  Eustis  complex 
mixture. 

-  Six  chemists  reported  results  by  the  cyanide  method,  ranging 
from  54.8  to  50.55  per  cent,  all  but  one  of  the  results  being  below 

1  Trnii!».  A.  I.  M.  K.,  it.  120. 

2  Trans.  A.  I.  M.  K..  as*  ^5°  <^"d  io<^- 


8l6  FREDERIC   P.    DEWEY. 

the  lowest  electrolytic  result.     These  cannot  be  regarded  as  sat- 
isfactory. 

A  plate  of  copper  made  from  melted  anodes  was  drilled  and 
six  chemists  reported  the  copper  in  the  drillings,  as  found  by 
the  electrolytic  method,  as  ranging  from  98.46  to  97.04  per  cent., 
and  averaging  97.67  per  cent,  with  a  maximum  difference  of 
1.42  per  cent.  These  results  are  not  as  good  as  those  previously 
reported  by  Mr.  Eustis. 

GOLD  AND  SILVER  IN  COPPER  MATERIALS. 

The  symposium  above  referred  to  was  undertaken  primarily 
to  test  methods  of  assaying  copper  material  for  gold  and  silver. 
Fourteen  chemists  reported  the  silver  by  scorification  assay, 
some  entirely  uncorrected,  some  partially  corrected,  and  some 
corrected  for  both  loss  in  slag  and  cupel  and  presence  of  copper 
in  the  silver  button.  The  averaged  results  ranged  from  135.38 
to  122.88  ounces  per  ton  and  averaged  128.86  ounces  per  ton  ; 
the  extreme  variation  being  12.5  ounces  per  ton,  or  nine  and 
seven-tenths  per  cent,  of  the  average. 

Nine  chemists  reported  ten  results  by  combined  wet  and  scor- 
ification methods,  a  few  of  them  corrected  for  slag  and  cupel 
absorption.  The  averaged  results  ranged  from  130,68  to  123.03 
and  averaged  127.25  ounces  per  ton.  The  extreme  variation 
was  seven  and  six-tenths  ounces  per  ton,  or  5.97  per  cent,  of  the 
average  determination. 

One  chemist  reported  123.6  ounces  per  ton  by  crucible  method. 

Another  reported  126.2  ounces  per  ton  by  combined  wet  and 
crucible  method,  corrected  for  slag  and  cupel. 

Summing  up,  we  have  twenty-six  results  by  twenty  chemists 
working  by  two  main  methods,  but  both  of  them  modified  in 
various  ways,  and  two  methods,  each  by  a  single  chemist,  vary- 
ing from  135.38  to  122.88  and  averaging  127.94  ounces  per  ton. 
The  extreme  variation  was  12.5  ounces  per  ton,  or  9.77  per  cent, 
of  the  average  determination. 

In  the  case  of  the  silver  assay  of  the  copper  borings,  nine 
chemists  reported  by  the  scorification  method,  with  and  without 
corrections.  The  averaged  results  varied  from  164.35  to  154.40, 
and  averaged  159.36  ounces  per  ton.  The  extreme  variation 
was  9.95  ounces  per  ton,  or  6.24  per  cent,  of  the  average. 


ACCURACY  OP  CHEMICAL  ANALYSIS.  817 

Fifteen  chemists  reported  sixteen  results  by  combined  wet  and 
scorification  methods,  with  and  without  corrections.  The  aver- 
aged results  varied  from  161.40  to  148.50  and  averaged  156.48 
ounces  per  ton.  The  extreme  variation  was.  13. 9  ounces  per 
ton,  or  8.88  per  cent,  of  the  average.  A  single  chemist  reported 
161.35  ounces  per  ton  by  combined  wet  and  crucible  process, 
corrected  for  slag  and  cupel. 

Summing  up,  we  have  twenty-six  determinations  by  twenty 
chemists  working  by  three  methods,  ranging  from  164.35  to 
148.5  and  averaging  157.67  ounces  per  ton.  The  extreme  varia- 
tion was  15.85  ounces  per  ton,  or  10.05  per  cent,  of  the  average 
determination. 

Twenty  chemists  working  by  the  four  methods  reported  twen- 
ty-six results  on  the  gold  in  the  matte  varying  from  2.41  to  1.85 
and  averaging  2.245  ounces  per  ton.  The  extreme  variation 
was  0.56  ounce  per  ton,  or  24.94  P^^  cent,  of  the  average. 

On  the  gold  in  the  copper  borings  twenty  chemists  working 
by  two  main  methods,  each  one  variously  modified,  and  the  com- 
bined wet  and  crucible  method  by  a  single  chemist,  reported 
twenty-six  results  varying  from  0.501  to  0.205  ^^<^  averaging 
0.307  ounce  per  ton.  The  extreme  variation  was  0.296  ounce 
per  ton,  or  96.4  per  cent,  of  the  average  determination. 

POTASH. 

In  the  determination  of  potash  the  1894  report  of  the  Associa- 
tion of  Official  Agricultural  Chemists'  gives  six  determinations 
of  potassium  chloride  by  six  chemists  by  one  method,  ranging 
from  97.79  to  99.32  per  cent,  with  an  average  of  98.56  per  cent, 
the  extreme  variation  being  1.53  per  cent.  By  another  method 
on  the  same  sample  seven  determinations  by  seven  chemists 
range  from  97.21  to  98.86  per  cent.,  averaging  98.16  per  cent. 
Combining  these  results  we  have  thirteen  results  by  seven 
chemists,  by  two  methods,  ranging  from  97.21  to  99.32  per  cent. 
and  averaging  98.35  per  cent.,  the  extreme  variation  being  2. 11 
per  cent. 

This  report  contains  also  a  table  of  results  on  soil  analyses' 
which  I  quote  entire. 

I  page  M. 
«  paffc  41. 


8i8 


ACCURACY  OF  CHEMICAL  ANALYSIS. 


1 

E 


O 

o  c;5 


1 

V 

s 


CB 

a 
o 


«*5 


fooo  j*«  poo  «o^ 


r>.i>i  rO^  •* 


5 


s^OOOQOOO 


1  _  _   _ 

v  too  6  6  «*>^ 6  6 

a.  «^ 


u   M  ^oB  M  o  M  ^  O 


tx»00  0>OsO0f« 


g 

> 


a   O^  •- 00  ^'J  t>i  M^  O  ■** 


2  >o  O  O  O  ^nO  O  i^ 


5    **- 


CT3 
—  V 

o  s 


N  M  o  ONr«-tx»  v«  O 


V 


t' ».  c  u 


-iQ 

o 


.^  ^ 


g  o\  r>.*5  ^  »o  ^  ►-  w  »o  N 

g  ^  6  6  6  ^^ 6  6  6  6 


3 


"S    fO 


2  ^St^  8  8  8  8> 


»N.O  o  o 


^  r».0  "^  O  O 


e 

km 

> 
< 


S  00  ^  -^  -^  "OvO  •«*•  r*  «  "* 
u  >0  O  O  O  rO\0  O  O  O  O 


O  3 


^  fO  ro  •-  ff*  O^nO  p*  r*  w^ 


a; 

pi* 


^  U  o  S  X  < 


r^-^:. 


i 

■a 


t^ 


g 

s 


•*)0Q  O^  M  O^  9^  O  O 


I 


s 


V 


Q  »o»fi«po  a»o 


^  «  e«  «  O 


^o  o  o 


f)  lOO  9^    • 


bo 


U    0>0^  '^00  M  lOiO 

u  i^  6  6  6  ^^  6  6 


ba 

c 

> 


xj    ^  rO  ^  «*>  t^  t** -^  O 


O  OnoO  r>»ioi^OM^ 


V 


1 

I; 

S 

"3 
5 

p 


••  >0  fO  t>i  w 


OS  O^  Os>0 


9, 


-a    0^ 


**5  f*i  w  fn>0  eo  ^  —  •*> 


9^6  6  o  <i*o6  o^o 

1^ 


be 


"5    lOQ  »0 


W   lO 


^  O  O  lOO  "♦  O 
f<i  ro  ^  lO  lO  C«  <S 
»/^  •O  ^  lO  »*5  r«   ro 


e*  oo 


o  o 


t^O  o  o  o 


be 


> 
< 


S    0c*<<**'^i^*^00i>iQO^ 

r  M  e«  t^Qo  to  o\  •«  t*  a«^ 


15 


O  C  O   rOVO   O   O^O   C 


a-o 

—  CI 

o  s 


^  O  Ov00>0>0  •-iW^O 


c  • 


0) 

o 


•«   P-4  /-•  >rt 


O  ' 


c 

4) 


O 


O 


^  :><  u  S  Oi  <  pu  X ;?;  x 


CI 

B 

«> 
«» 

o 

O 


REVIEW 


ON  THE  DEVELOPMENT  OF  SMOKELESS  POWDER.' 

By  Charles  K.  Munroe. 

To  intelligently  present  a  sketch  of  what  has  been  done  in  the 
development  of  smokeless  powder,  it  is  necessary  to  first 
briefly  review  the  history  of  black  gunpowder.  Although 
the  place  and  date  of  its  origin  and  the  name  of  its  inventor 
are  yet  open  to  dispute,  it  is  generally  accepted  that 
it  was  employed  as  a  propellent  in  cannon  at  the  battle  of 
Cr&y  in  1346,  and  in  small  arms  for  some  time  prior  to  this 
date,  and  that  it  then  consisted  of  a  mixture  of  niter,  charcoal 
and  sulphur.  Considering  the  existing  state  of  chemistry,  it  is 
fair  to  infer  that  the  making  of  gunpowder,  like  the  manufac- 
ture of  guns,  was  for  long  an  empiric  art,  and  that,  notwith- 
standing that  Tartaglia,  Galileo,  Newton,  puygens,  and  many 
others  speculated  upon  and  discussed  the  effects  which  gun- 
powder produced  upon  projectiles ;  that  granulating  was  em- 
ployed in  1445 ;  that  Cellini  had  observed  the  necessity  of 
adapting  the  grains  to  the  piece  ;  that  sizing  was  practised  in 
France  in  1525,  and  that  Hawksbee  had  in  1702  measured  the 
volume  of  gas  resulting  from  a  known  volume  of  gunpowder, 
the  science  of  gunnery  had  no  existence  until  Robins  devised  the 
ballistic  pendulum  by  which  he  measured  the  velocity  of  pro- 
jectiles and  with  which  he  obtained  the  experimental  data  upon 
which  his  **  New  Principles  of  Gijnnery,*' printed  in  1742,  was 
founded.  The  science  of  exterior  ballistic  was  materially 
improved  when  Hutton,  in  1778,  extended  Robins'  principle  to 
the  use  of  the  gun  as  the  pendulum  also,  for  it  became  then  pos- 
sible to  not  only  measure  the  velocity  of  the  projectile,  but  the 
energy  involved  in  the  reaction,  and  this  method  was  employed 
for  larger  an4  larger  calibers  until  it  reached  its  practical  limit 
in  the  ver>'  elaborate  and  precise  series  of  experiments  made  at 
the  arsenal  in  this  city  (Washington)  from  1842  to  1847,  by 
Major  Mordecai,  who  succeeded  in  swinging  cannon  weighing 
about  7,700  pounds  and  throwing  32-pound  balls  ;  but  this 
necessitated  the  use  of  a  pendulum  weighing  over  9,300 
pounds,  the  center  of  gravity  of  which  was  over  fourteen  feet 
below  the  axis  of  suspension.  The  weight  and  length  of  the 
pendulum  increases  so  rapidly  with  the  increase  of  the  projectile 
that  to  determine  by  this  method  the  velocity  of  the  projectile 
from  a  100- ton  gun,  would  require  towers  like  those  from  which 
the  Brooklyn  bridge  is  suspended,  between  which  to  swing  the 
pendulum. 

1  Presidential  addrees  delivered  before  the  Washing^ton  Section  of  the  American 
Chemical  Society.  Feb.  21.  1896. 


820  REVIEW. 

Opportunely  as  this  limit  was  approached,  Dr.  Joseph  Henry 
announced,  in  1843,  his  invention  of  a  method  for  the  determi- 
nation of  velocities  by  interposing  screens,  which  were  electri- 
cally connected  with  chronographs,  in  the  path  of  a  projectile 
and  at  definitely  determined  distances  from  the  gun,  and  this 
method,  which  while  possessing  the  merit  of  great  simplicity,  is 
at  the  same  time  very  precise  and  capable  of  being  used  for 
determining  the  velocities  of  projectiles  from -guns  of  every  cali- 
bre, is  now  universally  employed  with  chronographs,  such  as  the 
Bouleng6,  Schultz-Deprez,  and  Mahieu,  while  the  principle  has 
been  extended  by  Captain  Noble  to  the  study  of  interior  ballis- 
tics, in  his  very  ingenious  chronoscope  by  which  the  velocity  of 
the  projectile  can  be  determined  at  frequent  intervals,  even  when 
it  is  moving  through  the  bore  of  the  gun. 

The  ability  to  measure  the  velocities  which  it  produced  led  to 
active  investigations  into  the  properties  of  gunpowder  and 
resulted  in  the  experiments  of  Lavoisier,  between  1777  and  1778 
on  the  deflagration  of  powder,  of  Berthollet,  on  the  best  propor- 
tions for  mixing  the  ingredients,  of  Gay  Lussac,  on  the  refining 
of  niter,  of  Violette,  on  the  production,  composition  and  pro- 
perties of  charcoal,  of  Gay  Lussac  and  Chevreul,  of  Bunsen  and 
Schischkoff,  of  Linck,  and  of  K&royli,  on  the  composition  and 
volume  of  the  products  of  the  combustion  of  this  substance,  and 
of  many  other  experimenters  on  the  effects  resulting  from  differ- 
ences in  the  density,  hardness,  size  of  grain  and  other  physical 
characteristics  of  the  explosive.  But  notwithstanding  the  great 
advance  made  through  the  invention  of  methods  by  which  to 
measure  the  velocity  of  the  projectile  and  the  recoil  of  the 
piece,  the  science  of  gunnery  was  still  incomplete  without  an 
accurate  knowledge  of  what  was  going  on  within  the  chamber 
and  particularly  what  pressures  were  produced  and  how  this 
pressure  was  distributed  within  the  gun  before  the  projectile 
left  its  seat  and  while  it  was  traveling  through  the  chase  ;  yet, 
although  direct  experimental  determinations  of  the  pressure 
exerted  by  fired  gunpowder  were  made  by  Count  Rumford  in 
1797  in  a  somewhat  rude  device,  and  numerous  indirect  estima- 
tions were  deduced  from  the  observations  of  Robins  on  the  vol- 
ume of  the  gases  produced  by  its  combustion  and  from  the  more 
precise  and  detailed  researches  of  Bunsen  and  Schischkoff  and  the 
other  experimenters  previously  referred  to,  no  practical  means 
were  at  command  by  which  to  piake  direct  measurements 
of  the  pressure  developed  within  the  gun  itself  until  Captain 
Rodman,  in  1857,  invented  the  pressure  gauge,  described  in  his 
**  Reports  of  Experiments,'*  published  in  Boston,  in  1861.  which 
in  common  with  several  modifications  of  it,  such  as  the  Noble 


RBVIBW.  821 

crusher  gauge  and  the  Woodbridge  spiral  gauge,  came  into  gen 
eral  use  in  all  experimental  firing  and  in  the  proving  of  guns 
and  powders. 

In  estimating  the  pressure  developed  by  powder  from  the  data 
obtained  in  their  chemical  analyses  of  its  products,  Bunsen  and 
Schiachkoff  proceeded  on  the  assumption  that  Piobert's  conclu- 
sion from  his  experiments,  viz. ,  *  *  that  the  rate  of  combustion  of 
powder  is  not  affected  to  any  sensible  degree  by  h%at  or  pres- 
sure/' was  correct;  but  their  conclusions  having  been  ques- 
tioned by  many  authorities,  among  them  by  Vignotti,  in  1861, 
and  by  Craig  about  the  same  time,  who  showed  that  the  pro- 
ducts of  combustion  differs  with  the  pressure,  and  their  physical 
data  by  F.  A.  P.  Barnard,  who  submitted  them  to  a  rein- 
vestigation in  1863,  and  arrived  at  a  widely  different  result ;  and 
they  having  also  failed  of  verification  by  the  pressure  gauge,  the 
matter  was  again  experimentally  attacke4  by  Noble  and  Abel, 
who  employed  as  a  firing  chamber  a  hermetically  closed  steel 
cylinder  sufficiently  strong  to  resist  rupture  by  the  explosion  of 
a  charge  of  powder  which  completely  filled  it  (such  as  Dr. 
Woodbridge  had  previously  used  at  the  Washington  navy  yard 
in  1856),  in  which  pressure  gauges  were  enclosed,  and  they  fired 
the  charge  by  the  electric  method  invented  by  Dr.  Robert  Hare 
in  1832.  In  addition  the  apparatus  was  so  contrived  that  the 
gaseous  and  solid  products  could  be  collected,  measured  and 
analyzed  at  will. 

With  this  they  found  that  whjen  powder  is  fired  in  a  confined 
space  the  products  of  combustion  are  about  fifty-seven  per  cent, 
by  weight  of  ultimately  solid  matter,  and  forty-three  of  gases, 
which  at  o^  C.  and*  760  mm.,  o(^cupy  about  280  times  the  volume 
of  the  original  powder.  That  the  temperature  of  explosion  is 
about  2,200''  C,  and  the  tension  of  the  products,  when  the  pow- 
der entirely  fills  the  space  in  which  it  is  fired,  is  about  6,400 
atmospheres,  or  forty-two  tons  per  square  inch. 

When  fired  in  the  bore  of  the  gun  it  was  shown  that  the  work 
on  the  projectile  is  effected  by  the  elastic  force  resident  in  the 
permanent  gases,  but  the  reduction  of  temperature,  due  to  the 
expansion  of  the  permanent  gases,  is  in  a  great  measure  com- 
pensated for  by  the  heat  stored  up  in  the  liquid  residue.  The 
total  theoretical  work  of  gunpowder  when  expanded  indefinitely 
(as  for  instance  in  a  gun  of  infinite  length)  was  deduced  from 
the  data  which  they  accumulated  as  about  486  foot  tons  per 
pound  of  powder. 

They  further  ascertained  that  the  fine  grain  powders  furnish 
decidedly  smaller  portions  of  gaseous  products  than  large  grain 
or  cannon  powders  ;  that  the  variations  in  the  '  composition  of 


822  REVIEW. 

the  products  of  explosion,  in  a  closed  vessel,  furnished  by  one 
and  the  same  powder,  under  different  conditions  as  regards  pres- 
sure, and  by  two  powders  of  similar  composition,  under  the  same 
conditions  of  pressure,  are  so  considerable  that  no  chemical 
expression  can  be  given  for  the  metamorphosis  of  a  gunpowder 
of  a  normal  composition,  and  that  the  proportions  of  the  several 
constituents  of  the  solid  residue  are  quite  as  much  affected  by 
slight  accidental  conditions  of  explosion  of  one  and  the  same 
powder  in  different  experiments  as  by  decided  differences  in  the 
composition  or  in  the  size  of  the  grain. 

The  subsequent  researches  of  Berthelot  and  Vieille,  and  of 
Sarrau  and  Vieille  showed  that  gunpowder  was  not  singular  in 
that  its  combustion  products  varied  with  the  variations  in  the 
conditions  prevailing  in  the  firing  chamber,  but  that  this  same 
rule  held  for  gun  cotton,  picrates  and  other  explosives,  also,  and 
that  consequently  the  chemical  reaction  taking  place  and  the 
physical  phenomena  attending  them  were  changed  with  these 
varying  conditions,  and  more  particularly  with  variations  in  the 
densit}'  of  loading. 

Before  the  invention  of  the  instruments  of  precision  above 
alluded  to,  guns  were  constructed  largely  on  principles  deduced 
from  obser\'ations  of  exterior  phenomena,  and  powder  was  manu- 
factured largely  by  rule  of  thumb.  With  the  ability  to  determine 
quantitatively  their  behavior,  each  has  been  studied  in  a  scientific 
manner  and  improved  by  rational  methods. 

By  their  use  the  real  importance  of  uniformity  in  chemical  and 
physical  composition  was  demonstrated  for  the  powder,  and  the 
means  by  which  to  **  prove  powder"  before  issue  were  supplied, 
while  rational  blending,  by  which  to  minimize  the  irregularities 
incident  to  the  best  commercial  processes  was  made  possible. 
At  the  same  time  greater  uniformity  in  granulation  was  secured ; 
the  best  form  of  grain  was  developed  for  g^eat  guns  through  the 
pebble  to  the  mammoth,  disk,  pellet,  sphere,  cylinder,  hollow 
cylinder,  hexagon  and  cube  to  the  hexagonal  prism,  with  one 
canal,  which  is  now  generally  adopted,  and  which  is  a  modified 
form  of  the  grain  invented  by  Rodman  ;  the  size  of  the  grain 
best  adapted  for  a  given  gun  was  ascertained,  and  the  size  rose 
trom  one-sixth  of  an  inch,  as  used  in  the  15-inch  S.  B.,  to 
fhe  hexagonal  prism  one  inch  in  height  by  1.36  inches  in  diame- 
ter ;  the  density  of  the  grain  rose  from  1.60  to  1.86  ;  the  effect 
of  prearranged  variations  of  density  in  grain's,  as  proposed  by 
Doremus  and  carried  out  in  the  Fossano  powder,  was  deter- 
mined ;  and  the  important  part  which  moisture  played  in  the 
reactions  going  on  in  the  chamber  with  the  necessity  for  introduc- 
ing it  into  the  grain  in  definite  quantities  and  retaining  it  there 


REVIEW.  823 

wicbin  very  narrow  limits  was  discovered.  In  fact  these  methods 
of  inspection  have  become  so  precise  and  the  powder  specifica- 
tions so  severe  that  the  manufacture  of  military  gunpowder  is 
now  a  most  difficult  art,  and  the  maker  must  not  only  watch 
the  barometer  and  thermometer  and  hygrometer  to  determine 
his  action  at  each  step  of  his  process,  but  according  to  one 
authority,  he  must  "vary  his  treatment  with  each  passing 
cloud/*  and  notwithstanding  all  precautions,  it  is  no  uncommon 
thing  for  the  best  makers  to  have  their  product  rejected  at  the 
proving  ground. 

Besides  these  improvements  in  black  gunpowder,  which  have 
resulted  from  our  ability  to  accurately  gauge  its  performance, 
these  instruments  have  shown  us  that  it  is  possible  to  avail  our- 
selves of  the  energy  stored  up  in  underburned  charcoal  or  carbo- 
hydrates if  we  but  modify  the  brusqueness  incident  to  mixtures 
containing  them  by  adopting  the  proper  size,  form,  hardness 
and  density  for  the  grain,  and  this  has  resulted  in  the  cocoa  or 
brown  prismatic  powders  which  have  come  into  very  extented 
use  since  1880. 

The  valuable  properties  of  the  compressed  powder  were  then 
applied  for  use  in  small  calibers  and  enabled  Hebler  to  realize  a 
marked  increase  in  efficiency  for  his  rifles,  and  in  these  forms 
the  limit  of  efficiency  of  gunpowder  appeared  to  be  reached. 

But  while  this  was  being  accomplished,  progress  was  being 
rapidly  made  along  other  lines  which  we  will  briefly  point  out. 

Among  the  other  inventions  in  gunnery  which  preceded  the 
invention  of  smokeless  powder,  and  made  its  use  possible  or 
essential,  we  may  mention  the  introduction  of  rifling,  by  which 
greater  accuracy  of  fire  and  a  higher  velocity  and  penetrating 
efiFect  is  obtained,  and  which,  while  invented  by  Gaspard  Zoll- 
ner,  of  Vienna,  in  1480,  did  not  come  into  vogue  until  1850,  or 
general  use  until  much  later.  Breech  loading,  which  was 
known  among  the  Chinese  as  early  as  13 13,  but  which  has  prac- 
tically been  developed  since  1863,  our  civil  war  having  been 
fought  chiefly  with  muzzle  loaders.  Percussion  caps,  invented 
by  Joseph  Egg,  in  18 18,  and  adopted,  with  the  nipple,  in  France 
in  1838.  Self  obturating  metallic  ammunition,  which  depended 
on  the  preceding  invention,  and  which  we  owe  apparently  to  Flo- 
bert,  who  introduced  it  for  use,  with  a  quick  powder,  in  his  parlor 
rifle  in  1845,  though  it  did  not  come  into  use  for  larger  caliber 
for  some  years  later,  and  then  only  after  the  discovery  of  a  metal 
having  the  necessary  ductility  and  strength  from  which  to 
strike  the  shells  and  the  perfecting  of  machinery  for  their  eco- 
nomic and  rapid  production.  Magazine  rifles  and  machine 
guns,  the  earlier  practical  forms  of  the  latter  being  the  weapon 


824  RBVIBW. 

exhibited  by  Dr.  Gatling  in  1867,  and  the  French  mitraiUettse, 
and  which  have  now  developed  into  the  automatic  machine  guns, 
such  as  Maxim,  Colt,  Hotchkiss,  and  others  possessing  an 
almost  incredible  rapidity  of  discharge.  Rapid  fire  large  caliber 
guns,  which,  like  the  foregoing,  depend  for  their  development  on 
the  prior  invention  of  the  breech  mechanism,  and  the  metallic 
ammunition  and  which  have  reached  calibers  of  six-inch  diame- 
ter and  throw  lOO-pound  shot  at  the  rate  of  six  per  minute,  with 
a  velocity  of  over  2,000  feet  per  second.  Breech-loading,  built- 
up  steel  rifles,  which,  while  embodying  the  ideas  of  a  gun  of  equal 
strength,  as  announced  by  Professor  Treadwell,  in  1843,  the 
mechanical  devices  of  Chambers  patented  in  1849,  and  the  prin- 
ciples of  initial  tension,  as  expounded  in  Rodman's  publication 
of  the  same  year,  have  been  developed,  at  least  in  this  country, 
only  since  the  appointment  of  the  Gun  Foundry  Board  by  Sec- 
retary Chandler,  and  whose  manufacture  was  then  rendered  pos- 
sible only  through  the  perfection  which  our  machine  tools  had 
attained  and  the  improvements  achieved  in  the  metallurgy  of 
steel.  Small  caliber  rifles,  with  steel  or  german-silver  mantled 
bullets,  which  are  sighted  for  about  two  miles,  and  whose  projec- 
tile will  pierce  six  men,  standing  one  behind  the  other  in  close 
order,  at  1,000  yards.  And  finally  to  the  invention  of  range- 
finders  or  telemeters,  through  which  by  trigonometric  or 
mechanical  methods,  the  position  of  the  far  distant  targets  now 
in  range  of  new  weapons  may  be  located  with  precision. 

For  it  is  evident  that  to  use  these  precise  and  powerful  weap- 
ons and  instruments,  with  the  accuracy  and  rapidity  they 
are  capable  of,  the  atmosphere  must  remain  clear,  and  the 
piece  must  remain  clean,  while  at  the  same  time  the  highest 
attainable  velocity  must  be  imparted  to  the  projectile  without 
an  undue  strain  being  brought  upon  the  gun.  Yet  we  have 
seen  that  Noble  and  Abel  found  that  military  gunpowder 
gives  off,  on  combustion,  fifty-seven  per  cent,  by  weight  of 
ultimately  solid  matter  which  is  either  thrown  into  the 
atmosphere  to  produce  smoke  or  left  as  a  residue  to  foul  the 
bore.  How  considerable  this  smoke  producing  capacity  of  gun- 
powder is  maybe  estimated  if  we  take  a  Gatling  firing  1200 
rounds  of  small  arm  ammunition  per  minute  (and  this  by  no 
means  expresses  the  highest  attainable  speed  to-day)  and 
assume  that  all  the  solid  matter  is  driven  out  the  gun,  when  we 
shall  find  that  each  minute  six  and  six-tenth  pounds  of  finely 
divided  solid  matter  will  be  projected  into  the  atmosphere.  Add 
to  this,  in  a  general  engagement,  the  smoke  from  the  great  g^ns, 
which,  as  with  the  iio-ton  gun,  can  project  528  pounds  of  this 
solid  product  at  each  discharge,  and  that  coming  from  the  rapid 


REVIEW.  825 

fire,  and  magazine  rifles,  and  it  is  obvious  that  unless  a  favora- 
ble breeze  is  blowing  or  other  favorable  atmospheric  conditions 
prevail,  the  force  or  ship  will  soon  be  enveloped  in  an  opaque 
cloud  of  smoke  and  be  at  the  mercy  of  an  invisible  foe.  It  is,  I 
repeat,  conditions  such  as  these  which  have  rendered  smoke- 
less powder,  of  good  ballistic  qualities,  a  great  desideration, 
if  not  an  absolute  necessity. 

While  the  development  of  the  projectile,  the  musket,  the 
machine  gun«  and  ordnance ;  the  perfection  in  the  composi- 
tions, forms,  and  manufacture  of  gunpowder ;  and  the  inven- 
tion of  the  instruments  and  devices  for  gauging  and  controlling 
their  performance  was  going  on,  chemists  were  engaged  in  add- 
ing their  contributions  to  the  fund  of  human  knowledge  in  the 
field  of  explosives.  In  1788  Hausmann  discovered  **  picric 
acid,**  in  1800  Howard  discovered  mercuric  fulminate,  in  1845 
Schonbein  discovered  gun  cotton,  in  1845  Sobrero  discovered 
nitroglycerin,  in  1875  Nobel  invented  explosive  gelatine,  and 
in  the  meantime,  or  subsequently,  numerous  allied  nitro-substi- 
tution  compounds,  nitric  ethers  and  diazo-bodies,  less  generally 
known  than  those  above  enumerated,  were  produced,  and  iden- 
tified, and  shown  to  possess  explosive  properties. 

The  earlier  experimental  tests  of  these  bodies  proved  that  not 
only  were  some  of  them  more  powerful  or  more  violent  explo- 
sives than  gunpowder,  but  that  no  smoke  accompanied  their 
explosion,  since  the  products  of  their  explosive  decomposition 
were  gases  or  vapors  at  the  prevailing  temperatures  and  efforts 
were  put  forth  soon  after  their  discoverj'^  to  adapt  them  for  use 
as  propellents.  These,  together  with  various  organic  solids, 
and  liquids  to  ser\'e  as  solvents  and  hardening  agents  and  am- 
monium and  barium  nitrates  to  serve  as  oxidizing  aj^^ents  were 
know^n  and  at  hand. 

The  earliest  experiment  with  smokeless  powder  was  probably 
that  made  by  Howard,  in  1800,  when  he  tested  the  properties  of 
his  newly  discovered  mercuric  fulminate  and  found  that  though 
this  violent  agent  produced  little  smoke,  imparted  a  low  velocity 
to  the  projectile  and  but  a  slight  recoil  to  th«  piece,  it  burst  the 
chamber,  and  demonstrated  its  unfitness  to  compete  wnth  gunpow*- 
der  as  a  ballistic  agent.  Nevertheless  this  substance  has  since 
found  a  limited  use,  when  mixed  with  solid  diluents  which  act 
as  restrainers,  in  ammunition  for  parlor  rifles,  and  it  is  notice- 
able that  when  firing  this  ammunition  there  is  little  smoke  and 
a  scarcely  audible  report  attending  the  discharge. 

In  1806  Grindel  carried  out  a  somewhat  extended  series  of 
experiments  with  a  view  to  substituting  ammonium  nitrate  for 
potassium   nitrate  as  the  oxidizing  agent  in  gunpowder  mix- 


826  RBVIBW. 

tures  but  the  deliquescent  character  of  the  ammonium  salt  ren- 
dered the  powder  made  with  it  useless  under  the  then  existing 
conditions,  and  has  proven  a  formidable  obstacle  to  its  use  in 
many  of  the  attempts  subsequently  made.  The  fact,  however, 
that  the  products  of  its  combustion,  at  the  prevailing  tempera- 
ture, are  wholly  gaseous  rendered  it  a  tempting  material  to 
inventors  of  smokeless  powders  and  it  has  been  more  recently 
used,  among  others,  by  F.  Gaens,  who,  in  1885,  patented,  in 
Germany,  his  so-called  **  Amide  Powder,'*  produced  by  mixing 
eighty  parts  of  ammonium  nitrate  and  loi  parts  of  potassium 
nitrate,  with  forty  parts  of  charcoal.  He  claimed  that  this  mix- 
ture was  not  hygroscopic  and  was  practically  smokeless,  and  he 
held  that  by  the  reaction  consequent  on  the  ignition,  a  potass- 
amine  was  formed  which  was  both  volatile  and  explosive. 
Whatever  the  nature  of  the  reaction,  it  appears  from  the  reports 
that  an  ammonium  nitrate  powder  was  produced  about  this  time 
in  Germany  and  later  in  England,  under  the  name  of  Chilworth 
Special,  which  possessed  remarkable  ballistic  properties  and 
3'ielded  comparatively  little  smoke,  which  speedily  dispersed, 
and  which  bore  exposure  very  well  until  the  humidit)'  of  the 
atmosphere  approached  saturation. 

It  is  possible  that  the  ammonium  nitrate  used  may  have  been 
produced  by  Benker's  process,  in  which  the  salt  is  formed  by 
metathesis  from  solutions  of  sodium  nitrate  and  ammonium  sul- 
phate exposed  to  a  temperature  of  — 15*",  or  below,  for  it  is 
claimed  that  the  ammonium  nitrate  which  crystallizes  out  under 
these  circumstances  is  of  extraordinar>'  purity  and  not  at  all 
hygroscopic. 

It  would  appear  that  though  these  ammonium  nitrate  powders 
are  slightly  hygroscopic,  they  may  retain  their  good  qualities 
for  long  times  in  the  hermetically  sealed  cases  used  in  fixed  am- 
munition up  to  the  six-inch  rapid  fire  gun,  but  that  we  know  that 
the  small  amount  of  water  necessarily  present  produces  marked 
changes  during  long  periods  of  storage  »yith  var>'ing  tempera- 
tures and  that  the  ammoniacal  salts  attack  the  copper  of  the 
shells.  Besides,  too,  we  must  remember  that  ammonium  nitrate 
in  common  with  other  ammonium  salts  gives  off  ammonia,  when 
heated  or  exposed  to  the  air,  and  becomes  acid  so  that  we  are 
debarred  from  using  it  in  the  presence  of  any  bodies  affected  b)' 
the  acid. 

The  next  step  toward  the  development  of  our  modern  smoke- 
less powder  was  taken  when,  soon  after  the  discovery  of  guncot- 
ton,  in  1845,  attempts  were  made  to  use  this  material  as  a  pro- 
pellent. These  experiments  were  made  in  Gennany,  France, 
and  England,  and  a  very  extended  series  were  carried  on  by 


REVIEW.  827 

Major  Mordecai,  at  the  Washington  Arsenal,  but  the  material, 
owing  to  its  form  and  the  imperfection  in  its  manufacture, 
proved  too  brisant  and  too  irregular  in  its  action,  and  so  unsta- 
ble on  keeping  as  to  undergo  decomposition  in  storage.  The 
material  having  been  proved  to  possess  many  valuable  qualities 
was  not  wholly  abandoned,  but  it  continued  to  be  the  subject  of 
study  by  many  chemists  until  in  1862,  it  seeming  that  Baron  von 
Lenck  had  so  perfected  the  methods  for  its  manufacture  and 
purification  as  to  ensure  stability  and  uniformity  of  composition. 
Austria  adopted  it  as  a' propellent  and  supplied  thirty  howitzer 
batteries  with  g^ncotton  cartridges. 

This  is  the  first  instance  in  which  a  really  smokeless  powder 
was  employed  on  any  but  an  experimental  scale  and  this  pow- 
der foreshadowed  in  its  composition  and  many  of  its  character- 
istics, the  best  modem  powders  of  the  smokeless  class.  The 
guncotton  as  then  made  retained  the  fibrous  condition  of  the 
original  cotton  and  in  the  Austrian  cartridges  it  was  spun  into 
thread  and  woven  into  circular  webs  like  lamp  wicks,  or 
braided,  or  wound  on  wooden  or  paper  bobbins,  and  so  arranged 
in  the  piece  as  to  secure  the  desired  air  spacing  as  well  as  to 
insure  ignition  from  the  front.  As  thus  used,  it  was  claimed  to 
be  uninjured  by  dampness ;  to  require  a  charge  of  but  one- 
fourth  to  one-third  of  that  of  the  powder  previously  employed  ; 
to  be  capable  of  being  regulated  so  as  to  produce  widely  var^'- 
ing  effects  at  will ;  to  leave  no  residue  to  foul  the  piece ;  and  to 
produce  no  smoke,  while  the  gases  evolved  were  less  injurious 
to  both  the  piece  and  men  serving  it  than  those  of  gunpowder. 
At  the  same  time  it  produced  less  heating  effect  on  the  gun. 

Unfortunately,  about  this  time,  the  factory  at  Hirtenberg, 
where  the  guncotton  was  made,  blew  up  for  some  undiscovered 
cause,  and  accidents  having  occurred  with  the  guns,  the  use  of 
guncotton  was  abandoned  by  the  Austrians. 

Its  fate  seemed  now  to  be  sealed,  but  such  was  not  the  case, 
for  the  scene  of  action  then  passed  to  England,  where  Abel  not 
long  after  succeeded  in  effecting  a  more  complete  purification  of 
the  body  by  pulping  it  prior  to  the  final  washing  processes,  thus 
cutting  the  tubular  fiber  into  short  lengths  and  rendering  it  pos- 
sible to  remove  the  last  traces  of  acid  retained  within  the  tubes 
by  capillarity  and  which  had  been  the  occasion  of  its  decomposi- 
tion with  time.  Having  thus  obtained  his  pulped,  purified  gun- 
cotton  he  compressed  it  into  such  forms  as  was  desired,  and  in 
1867  and  1868  he  obtained  with  it  some  very  promising  results 
when  used  with  field  guns.  But  although  comparatively  small 
charges  often  gave  high  velocities  of  projection  without  any 
indications  of  injury  to  the  gun,  the  uniform  fulfillment  of  the 


828  RBVIEW. 

conditions  essential  to  safety  proved  then  to  be  beyond  control, 
and  the  military  authorities  not  being,  at  that  time,  alive  to  the 
advantages  that  might  accrue  from  the  employment  of  a  smoke- 
less explosive  in  artillery,  experiments  were  discontinued  not 
to  be  resumed  for  nearly  twenty  years,  and  use  was  found  for 
compressed  guncotton  in  military  and  naval  mining  and  espe- 
cially in  filling  torpedoes,  where  it  has  been  found  the  most 
efficient  and  satisfactory  explosive  thus  far  applied  to  this  pur- 
pose. 

But  sportsmen,  to  meet  whose  wants  and  wishes  many  note- 
worthy improvements  have  been  made  in  the  arts,  did  appreciate 
the  value,  to  marksmen,  of  smokelessness  combined  with  high 
velocities  and  absence  of  fouling,  and  the  progress  made  during 
the  succeeding  twenty  years  in  the  adaptation  of  organic  nitrates 
to  use  as  propellents  was  under  their  patronage  and  in  response 
to  their  demands,  and  naturally,  the  first  object  sought  was  to 
so  restrain  the  violence  of  the  explosive  that  rupturing  explo- 
sions, such  as  had  occurred,  could  not  be  induced  under  the 
conditions  in  which  the  powder  was  to  be  used. 

One  of  the  first  to  realize  a  considerable  degree  of  success  was 
Captain  Schultze,  of  the  German  artillery,  who  made  a  powder 
from  well  purified  and  partly  nitrated  wood.  For  this  pur- 
pose he  sawed  the  wood  into  sheets  about  one-sixteenth  of  an 
inch  in  thickness,  which  were  passed  through  a  machine  that 
punched  out  discs  or  grains  of  uniform  size.  The  grains  were 
then  deprived  of  their  resinous  matter  by  being  boiled  in  sodium 
carbonate,  washed,  steamed,  and  then  bleached  with  chloride  of 
lime,  when  finally,  after  drying,  the  cellulose  was  nitrated  in  an 
acid  mixture,  such  as  is  used  for  making  guncotton.  The 
nitrated  wood  was  then  steeped  in  a  solution  of  potassium  and 
barium  nitrates,  and  when  dry  the  powder  was  finished.  By 
this  means  a  nitrocellulose  was  produced  which  was  diluted 
with  unconverted  cellulose  and  metallic  nitrates,  which  were  so 
intimately  mingled  that  a  fairly  even  rate  of  combusition  was 
obtained  though  abnormal  results  were  not  wholly  avoided. 

The  advantage  of  using  nitrates  and  combustible  organic  sub- 
stances as  diluents  was  soon  recognized  ;  and,  as  a  consequence, 
many  powders  of  this  nature  were  devised,  some  thirty  of  them 
having  been  produced  and  many  of  these  put  on  the  market,  in 
which  we  find  that  potassium,  sodium  and  barium  nitrates,  and 
potassium  chlorate  were  used  as  oxidizing  agents  and  sugar, 
cellulose,  charcoal,  sulphur,  starch,  dextrin,  gums,  resins,  and 
paraffine  as  combustible  diluents  and  cementing  agents.  All, 
however,  approximated  black  gunpowder,  as  regards  physical 


REVIEW.  829 

Structure  and  none  attained  to  complete  success  as  regards  uni- 
formity of  fire  and  reliability  of  pressure. 

In  1882  Messrs.  Reid  and  Johnson  patented  the  process  for 
making  £.  C.  powder,  in  which  the  pulped  nitrocellulose  and 
nitrates  was  agglomerated  into  grains  by  revolving  the  mois- 
tened mass  in  barrels,  drying  the  grains,  moistening  with  ether  to 
harden  them,  and  then  coloring  them  with  aurine. 

About  1885  Messrs.  Johnson  and  Borland  produced  the  J.  B. 
powder,  in  which  a  new  idea,  as  regards  powder  manufacture, 
was  introduced,  though  it  had  been  used  elsewhere  for  many 
years.  The  inventors  mixed  nitro  cotton  with  barium  nitrate 
and  with  or  without  charcoal  or  torrefied  starch  and  granulated 
the  mixture  in  a  revolving  drum,  while  the  water  was  admitted 
in  a  fine  spray.  When  granulated  the  grains  were  dried  and 
then  moistened  with  a  solution  of  camphor  in  petroleum  spirit, 
and  after  a  time  heated  in  a  water  jacketed  vessel  to  evaporate 
the  benzine,  and  the  bulk  of  the  camphor.  By  this  treatment 
the  grains  were  hardened  and  rendered  more  slowly  inflamable. 

As  this  method  of  treatment  resembles  in  some  particulars  that 
followed  in  the  production  of  celluloid,  though  it  differs  in 
details,  and  as  several  of  the  smokeless  powders  are  made  by 
methods  which  are  adapted  from  this  art,  you  will  pardon  me  if 
I  briefly  describe  it. 

Celluloid  is  made  from  that  form  of  cellulose  nitrate  known  as 
nitro-cotton  or  soluble  guncotton,  and  which  is  produced  by 
immersing  unsized  and  uncalendered  tissue  paper  for  a  short 
time  in  a  comparatively  weak  acid,  both  being  kept  at  a  mode- 
rately high  temperature.  This  nitro-cotton  is  pulped  in  a  rag 
engine,  dried  and  moistened  with  camphor  spirits.  If  a  con- 
siderable portion  of  camphor  spirits  be  added,  and  the  mixture 
be  allowed  to  stand  for  awhile,  the  mass  becomes  converted  into 
a  soft  translucent  amber  gum  ;  with  more  of  the  spirit  the  nitro- 
cotton  will  be  completely  dissolved  ;  but  as  carried  out,  the  pro- 
portion of  spirit  added  is  insufficient  to  produce  a  very  apparent 
change. 

The  mixture  is  now  taken  to  incorporating  rolls  or  '  ^grinders, ' ' 
(as  they  are  called  in  the  caoutchouc  industry),  where  it  is  inti- 
mately mixed  and  well  pressed  ;  when  the  particles  cohere  and 
the  whole  becomes  converted  into  a  plastic,  translucent  homo- 
geneous mass  which  behaves  like  India  rubber  and  resembles  it 
superficially  in  every  particular  but  color.  After  incorporation, 
by  cutting  the  length  of  the  roll,  the  mass  may  be  stripped  off 
in  one  continuous,  coherent  sheet,  which  on  exposure  to  the 
atmosphere,  through  which  the  spirit  and  camphor  are  volatil- 
ized, hardens  to  a  hornlike  mass. 


830  REVIEW.  • 

■ 

In  the  manufacture  of  a  smokeless  powder  by  this  means,  it  is 
customary  to  mix  with  the  nitro-cotton  or  mixed  cellulose 
nitrates,  a  small  proportion  of  other  nitrates  in  order  to  effect 
complete  combustion  and  a  restrainer  to  assist  in  bringing  the 
rate  of  combustion  within  normal  limits;  and  this  mixing  is 
easily  effected  on  the  incorporating  rolls.  Barium  nitrate  is  the 
salt  whibh  is  perhaps  most  largely  used,  and  it  is  preferred 
because  it  is  very  permanent,  contains  a  fair  proportion  of 
available  oxygen  which  it  j'ields  with  comparative  readiness, 
and  possibly  because  the  carbonate  which  is  formed  by  the  com- 
bustion has  so  high  a  specific  gravity  that  it  settles  with  consid- 
erable speed. 

Other  solvents  besides  camphor  spirits  are  employed  when  the 
higher  cellulose  nitrates  are  used  in  the  manufacture  of  the 
powder.  Thus  Engel  takes  a  cellulose  nitrate  prepared  from 
wood,  while  Glaser  employs  that  prepared  from  paper  or  card- 
board and  treats  it,  when  dry,  with  ethyl  acetate  or  acetone,  the 
action  of  the  solvent  being  aided  by  mechanical  kneading  in  a 
suitable  vessel  until  a  viscid  paste  or  gelatinous  mass  is 
obtained  with  which  the  barium  nitrate  and  a  h5-drocarbon, 
such  sis  naphthalene,  is  incorporated.  The  mass  is  then 
formed  into  any  desired  shape  and  the  solvent  is  allowed  to 
evaporate  or  is  distilled  off  by  any  suitable  means  when  the  pow- 
der is  left  as  a  dense  horny  material,  with  a  glassy  fracture, 
which  can  be  readily  granulated. 

The  first  military  smokeless  powder  of  the  modem  class  was 
made  in  France  in  1886  by  Vieille,  and  is  said  to  have  been 
compounded  of  cellulose  nitrates  mixed  with  picric  acid,  but  it 
was  soon  abandoned  m  favor  of  the  Poudre  B.,  which  consisted 
of  cellulose  nitrates  alone,  or  Poudre  B.  N.,  which  consisted  of 
these  nitrates  mixed  with  barium  nitrate  and  potassium  nitrate 
as  oxidants,  and  sodium  carbonate  as  a  neutralizer.  Both  these 
mixtures  were  condensed  and  hardened  to  a  celluloid-like  mass 
by  means  of  a  solvent  like  ether-alcohol,  ethyl  acetate  or  ace- 
tone. 

Excellent  ballistic  results  have  been  reported  from  France  as 
being  obtained  with  these  powders,  and  they  have  been  adopted 
by  the  French  government.  At  the  same  time  similar  mixed 
cellulose  nitrate  powders  have  been  produced  and  used  in  Ger- 
many, Austro-Hungary  and  Switzerland  ;  the  Weteren,  Trois- 
dorf  and  Von  Forster  powders  being  of  this  class.  Notwith- 
standing that  these  have  so  long  been  known,  our  government 
has,  with  regal  graciousness,  recently  granted  a  patent  to  two 
of  its  officers  for  a  powder  of  this  composition. 

These  are  made  by  mixing  the  ingredients  together  with  the 


REVIEW.  S3 1 

solvent  in  a  kneading  machine  of  the  Werner  and  Pfleiderer 
class,  in  batches  of  one  to  two  hundred  weight,  until  it  is  con- 
verted into  a  dough,  when  it  is  incorporated  and  the  solvent 
partly  driven  off  by  putting  on  the  grinding  rolls,  by  which 
means  it  is  also  formed  into  continuous  sheets,  whose  thickness 
is  fixed  by  the  set  of  the  rolls.  It  is  preferable  where  thick 
masses  are  desired  to  first  roll  into  thin  sheets  so  as  to  evaporate 
the  solvent  as  completely  as  possible  from  the  gelatinized  mass, 
and  then  by  piling  the  thin  sheets  on  one  another,  weld  them 
together  b}'  running  them  through  the  rolls.  They  are  then 
granulated  by  passing  them  under  a  set  of  revolving  circular 
knives  which  cut  them  first  into  strips  and  then .  into  rectangles 
of  the  desired  size  and  shape.  These  powders  are  dense,  hard 
and  hornlike  in  appearance. 

Following  Vieille  by  about  two  years,*  Nobel  invented  ballis- 
tite,  which  practicall}'  is  a  modified  explosive  gelatine,  differing 
from  it  only  in  that  while  the  gelatine  consists  of  ninety-three 
per  cent,  of  nitroglycerin,  and  seven  per  cent,  of  nitro-cotton, 
ballastite  contains  about  forty  per  cent,  of  nitro-cotton  and  one 
to  two  per  cent,  of  anilin  or  diphenylamin,  which  is  added  to 
the  nitroglycerol  nitro-cotton  mixture  as  a  neutralizing  agent 
to  ensure  stability.  At  first  the  solution  of  the  gun-cotton  and 
gelatinization  of  the  mixture  was  effected  by  means  of  camphor 
and  later  by  means  of  benzene,  but  it  is  now  produced  under 
the  English  patent  of  Lundholm  and  Sayer  of  1889.  They  dis- 
covered that  while  dry  nitro-cotton  is  but  slightly  soluble  in 
nitroglycerin  even  at  moderately  high  temperatures,  when  mixed 
^'ith  warm  water  and  stirred  up  by  compressed  air,  gelatiniza- 
tion sets  in  and  solution  may  be  completed  by  pressing  out  the 
water  and  working  in  the  grinder.  Flexible,  transparent  rub- 
ber-like sheets  are  formed,  which  may  be  cut  into  flakes  in  cut- 
ting machines  of  the  usual  type,  or  in  pastry  cutters,  or  may  be 
squirted  through  spaghetti  machines,  as  is  done  in  It'ily,  where 
these  cords  or  threads  of  ballistite  are  known  as  **  Filite." 

It  is  curious  to  note  how  many  of  the  machines  devised  for 
bread  making,  pastry  cutting  and  macaroni  forming,  have  been 
employed  in  the  manufacture  of  smokeless  powder. 

In  1889  Sir  Frederick  Abel  and  Professor  James  Dewar 
secured  their  patents  on  cordite,  which  like  ballistite,  contains 
nitroglycerin  and  cellulose  nitrate,  but  whereas  ballistite  is 
made  from  nitro-cotton  alone,  cordite  is  made  from  **  gun-cot- 
ton *'  containing  from  ten  to  twelve  per  cent,  of  nitro-cotton,  to 
TThicli  is  added  a  little  tannin,  dextrin  or  vaseline  to  ser\''e  as  a 
restrainer.     The  gelatinization  is  effected  by  means  of  acetone, 

1  Engliflh  PMent.  Januar>'  31.  iv>S. 


832  REVIEW. 

the  mixture  being  kneaded  to  a  dough  in  a  water-jacketed 
kneading  machine,compacted  in  a  mould  in  a  preliminary  press, 
and  the  mould  transferred  to  a  spaghetti  machine,  where  the 
explosive  is  squirted  into  cords.  As  these  cords  issue,  they  are 
reeled  on  bobbins,  which  are  placed  in  the  drying  house  to  drive 
off  the  acetone.  Whe^  this  is  completed  the  product  of  ten 
pressings  is  wound  from  ten  one-strand  reels  on  to  one  ten-strand 
reel  and  then  the  cordite  on  six  ten-strand  reels  is  wound  on  one 
drum,  making  a  cord  of  sixty  strands,  which  in  short  lengths 
forms  the  thirty  and  one-half  grains  charge  for  the  magazine 
rifle.  For  the  higher  calibers  the  cords  are  cut  in  lengths  as 
they  issue  from  the  press,  dried  and  made  up  into  bundles. 
Cordite  is  an  elastic  rubber-like  mass  with  a  light  to  dark  brown 
color. 

Analogous  to  these  in  composition,  in  that  they  consist  of 
nitroglycerin  with  cellulose  nitrates,  are  many  powders,  such 
as  amberite,  Maxim's  powder,  Leonard's  powder,  P.  P.  G., 
Peyton's  powder,  German  smokeless  powder  and  others,  and 
they  differ  in  but  slight  particulars.  Thus  Curtis  and  Andr6 
blend  different  cellulose  nitrates  before  incorporation  so  as  to 
secure  a  definite  nitrogen  content,  and  then  cement  by  ether- 
alcohol  ;  Maxim  restrains  his  powder  with  castor  oil ;  Leonard 
restrains  his  with  l3xopodium,  and  adds  urea  crystals  as  a  neu- 
tralizer  ;  Walke  claims  to  make  P.  P.  G.  from  a  nitro-cellulose, 
which  is  not  gun-cotton,  and  so  on. 

The  employment  of  nitro  substitution  compounds  as  bases  for 
smokeless  powders  has  been  comparatively  limited.  Over  twenty 
years  ago  Designolle  invented  powders  made  by  mixing  potas- 
sium picrate,  potassium  nitrate  and  charcoal  in  various  propor- 
tions. Borlinetto  produced  them  from  picric  acid,  sodium 
nitrate  and  potassium  dichromate.  Abel  and  Brug^re  from 
ammonium  picrate,  potassium  nitrate  and  charcoal,  and  more 
recently  Nobel  from  ammonium  picrate,  barium  nitrate  and 
charcoal.  Within  a  few  years  past  a  powder  has  been  manu- 
factured in  this  country  and  put  upon  the  market  as  a  sporting 
powder,  which  was  composed  of  ammonium  picrate,  potassium 
picrate,  and  ammonium  dichromate,  but  I  understand  it  has 
given  such  irregular  and  abnormal  pressures  that  its  manufac- 
ture has  been  discontinued. 

While  these  powders  may  have  been  smok€-weak  as  compared 
with  gunpowder,  it  is  difficult  to  understand  how,  in  the  pres- 
ence of  such  amounts  of  metallic  radicles,  they  could  have  been 
smokeless.  A  powder,  however,  which  is  made  by  Hermann 
Gttttler,  by  dissolving  nitro-lignin  in  molten  dinitro-toluene  and 
which  he  calls  Plastomentite,  may  well  possess  this  property^ 


REVIEW.  833 

and  it  is  reported  to  have  given  good  ballistic  results  at  the 
Bucharest  tests  of  1893. 

The  powder  called  Gelbite,  and  invented  by  Dr.  Stephen  H. 
£mmens»  was  also  smokeless.  This  was  made  by  an  ingenious 
process  in  which  paper  in  strips  was  nitrated  to  a  moderate  degree 
of  nitration,  then  fumed  with  ammonia  to  neutralize  the  acid, 
and  then  treated  with  picric  acid  to  neutralize  the  ammonia  and 
form  ammonium  picrate.  These  strips  were  then  rolled  up  into 
rolls  as  charges,  but  as  might  have  been  foreseen  from  a  study 
of  the  behavior  of  gunpowder  in  guns  and  the  study  of  the  his- 
tory of  gun-cotton,  this  powder  was  too  brusque  in  action  and 
has  been  abandoned. 

I  began  my  own  experiments  with  smokeless  powder  manu- 
facture in  1889.  At  this  time  the  remarkable  results  published 
from  France,  and  the  announcement  that  that  country  had 
adopted  a  smokeless  powder,  had  produced  their  desired  strate- 
gic effect.  All  her  rivals  were  seeking  to  be  equally  well 
equipped  and  were  hastening  to  adopt  a  powder  even  before  its 
qualities  were  thoroughly  proven.  The  newspapers  contained 
remarkable  accounts  of  their  performances  and  alleged  descrip- 
tions of  their  methods  of  production,  which  while  interesting  as 
news  and  conveying  valuable  suggestions,  could  not  be  relied 
upon  as  to  accuracy  in  details. 

At  the  outset,  being  familiar  with  the  impossibility  of  secur- 
ing absolute  uniformity  and  constancy  of  composition  in  phy- 
sical mixtures  like  gunpowder,  and  realizing  how  important  this 
feature  was  with  our  precise  modern  weapons,  and  when  employ- 
ing an  explosive  possessing  great  energy,  I  determined  to 
attempt  to  produce  a  powder  which  should  consist  of  a  single 
substance  in  a  state  of  chemical  purity.  This  was  a  thing 
which  I  had  not  known  of  having  been  done,  nor  have  I  yet  learned 
that  any  one  else  has  attempted  it.  Among  the  bodies  at  com- 
mand, the  nitric  ethers  seemed  most  available,  and  of  these  cel- 
lulose nitrate  seemed  for  many  reasons  the  most  promising. 

There  are.  as  you  are  aware,  several  of  these  nitrates  (authori- 
ties differ  as  to  the  number)  which  differ  in  their  action  towards 
solvents,  though  all  except  the  most  highly  nitrated  are  soluble 
in  methyl  alcohol.  In  the  commercial  production  of  cellulose 
nitrate  certainly,  and  so  far  as  I  have  observed  under  all  cir- 
cumstances, when  nitrating  cellulose  the  product  is  a  mixture 
of  different  cellulose  nitrates.  Even  in  the  perfected  Abel  pro- 
cess for  making  military  gun-cotton,  as  carried  out  at  the  Royal 
Gun  Powder  Factory,  at  Waltham  Abbey,  according  to  Gutt- 
manu\  the  product  contains  as  a  rule,  from  ten  to  twelve  per  cent. 
of  nitro-cotton. 

1  Manafacture  of  Bzplosires.  a,  259,  1&95. 


834  REVIEW. 

Consequently  I  began  by  purifying  my  dried  pulped  military- 
gun-cotton,  which  was  done  by  extracting  it  with  hot  methyl 
alcohol  in  a  continuous  extractor,  and  when  this  was  completed 
the  insoluble  cellulose  nitrate  was  again  exposed  in  the  drying 
room.  The  highly  nitrated  cellulose  was  then  mixed  with  a. 
quantity  of  mono-nitro-benzene,  which  scarcely  affected  its  ap- 
pearance and  did  not  alter  its  powdered  form.  The  powder  was 
then  incorporated  upon  a  grinder  by  which  it  was  colloidized 
and  converted  into  a  dark  translucent  mass  resembling  India, 
rubber.  The  sheet  was  now  stripped  off  and  cut  up  into  flat 
grains  or  strips,  or  it  was  pressed  through  a  spaghetti  machine 
and  formed  into  cords,  either  solid  or  perforated,  of  the  desired 
dimensions,  which  were  cut  into  grains.  Then  the  granu- 
lated explosive  was  immersed  in  water,  boiling  under  the 
atmospheric  pressure,  by  which  the  nitro-benzene  was  carried 
off  and  the  cellulose  nitrate  was  indurated  so  that  the  mass 
became  light  yellow  to  gra\',  and  as  dense  and  hard  as  ivory, 
and  it  was  by  this  physical  change  in  state,  which  could  be 
varied  within  limits  by  the  press  that  I  modified  the  material 
from  a  brisant  rupturing  explosive  to  a  slow  burning  propellent. 

This  is  the  powder  which  I  styled  indurite,  and  which  has 
been  popularly  known  as  the  Naval  Smokeless  Powder. 

I  was  satisfied  that  I  was  justified  in  starting  on  this  new- 
practice  in  powder-making  when  I  found,  on  examination  of  the 
samples  of  foreign  military  powders*  which  later  began  to  reach 
me  officially,  that  they  were  heterogeneous  mixtures  as  the  old 
gunpowder  is  and  that  they  contained  matter  which  was  volatile 
at  ordinary  temperatures,  and  when  I  learned  that  the  nitro- 
glycerin powders  cracked  from  freezing. 

I  was  still  more  satisfied  when  I  learned  the  results  of  the 
proving  tests  which  were  all  made  except  the  chemical  stability 
and  breaking  down  tests  by  naval  officers  detailed  for  this  pur- 
pose at  the  Proving  Ground  and  elsewhere,  and  who  had  no  pre- 
judice in  its  favor.  All  of  the  numerous  publications  which 
have  appeared  about  it  have  issued  from  headquarters,  and  I 
present  the  matter  myself  here  for  the  first  time. 

I  have  appended  the  data  from  these  trials  to  this  address 
where,  on  inspection  it  will  be  seen,  that  after  development,  the 
powder  in  use,  in  successive  rounds,  gave  remarkably  regular 
pressures  and  uniform  velocities.  I  was  informed  by  the 
Chief  of  the  Bureau  before  the  firing  trials,  recorded  in  the 
tables  began,  that  if  I  could  produce  a  powder  giving  2,000 
feet  initial  velocity  and  but  fifteen  tons  pressure,  it  would  be 
a  complete  success.  Inspection  of  the  tables  show  that  this  was 
more  than  realized  and  that  in  two  successive  rounds  in  the 

1  Table  I. 


REVIEW.  835 

six-inch  rapid  fire  gun,  using  twenty-six  pounds  of  my  pow- 
der and  a  100  pound  projectile,  the  pressures  were  13.96  and 
13-93  tons,  and  the  velocities  2,469  and  2,456  feet  per  second 
respectively,  while  according  to  the  Report  of  the  Secretary 
of  the  Navy,  1892,  page  26,  **  the  powder  manufactured  for  use 
in  the  six-inch  rapid  fire  guns  was  stored  at  Indian  Head 
proving  ground,  through  a  period  of  six  months,  covering  a 
hot  summer,  and  at  the  end  of  the  time  showed  no  change 
in  a  firing  test.*' 

On  page  25  Secretary  Tracy  says,  * '  It  became  apparent  to 
the  department  early  in  this  administration  that  unless  it  was 
content  to  fall  behind  the  standard  of  military  and  naval  progress 
abroad  in  respect  to  powder,  it  must  take  some  steps  to  develop 
and  to  provide  for  the  manufacture  in  this  country  of  the  new 
smokeless  powder,  from  which  extraordinary  results  had  been 
obtained  in  Europe.  With  this  object  negotiations  were  at  first 
attempted  looking  to  the  acquisition  of  the  secret  of  its  composi- 
tion and  manufacture.  Finding  itself  unable  to  accomplish  this, 
the  Department  turned  its  attention  to  the  development  of  a 
similar  product  from  independent  investigation.  The  history  of 
these  investigations  and  of  the  successful  work  performed  in 
this  direction  at  the  torpedo  station  has  been  recited  in  previous 
reports.  It  is  a  gratifying  fact  to  be  able  to  show  that  what  we 
could  not  obtain  through  the  assistance  of  others,  we  succeeded 
in  accomplishing  ourselves,  and  that  the  results  are  considera- 
bly in  advance  of  those  hitherto  attained  in  foreign  countries." 

From  this  survey  we  see  that  all  of  the  smokeless  powders 
that  have  met  with  acceptance  and  proved  of  value  as  ballistic 
agents  with  the  exception  of  Indurite  are  mixtures  of  one  or 
more  of  the  cellulose  nitrates,  or  mixtures  of  these  bodies,  with 
nitroglycerin  or  some  other  oxidizing  agent,  like  barium 
nitrate,  and  a  restrainer  or  with  a  nitro  substitution  compound 
and  that  all  have  been  condensed  or  hardened  into  a  rubber-like 
or  celluloid-like  form,  by  which,  even  under  the  high  pressures 
which  obtain  in  the  gun,  they  are  expected  to  undergo  combus- 
tion only  and  that  at  a  moderate  and  regular  rate. 

In  thus  condensing  the  material,  and  in  determining  the 
best  form  of  grain,  it  wull  be  observed  that  we  have  been 
guided  by  the  experience  gained  in  the  compression  of 
g^unpowder,  and  we  have  been  able  to  effect  this  as  we  have 
by  the  experience  gained  in  the  development  of  celluloid,  and 
we  have  been  able  to  manipulate  our  product  and  shape  it  into 
grains  only  by  adopting  the  methods  and  machines  developed  in 
the  manufacture  of  food,  while  we  have  been  able  to  test  our 
product  and  check  our  results  and  thus  ensure  a  more  rapid  and 


836  RBVISW. 

certain  advance  by  the  constant  use  of  the  pressure  gauge  and 
velocimeter.  In  my  opinion,  if  these  resources  had  not  been  at 
command  and  available  the  smokeless  powder  Industry  wotild 
not  yet  exist. 

From  what  has  been  said  it  may  properly  be  inferred  that  we 
seek  in  these  new  powders  all  the  virtues  of  the  old  gunpowder 
with  the  addition  that  the  new  powder  shall  be  smokeless,  impart 
higher  velocities  while  producing  no  greater  pressures  and  that 
less  of  it  shall  be  required  to  do  the  work.  These  requirements 
may  be  summed  up  as  follows  : 

The  conditions  that  a  smokeless  powder  suitable  for  a  propel- 
lent should  fulfill  are  : 

1.  That  it  shall  be  physically  and  chemically  uniform  in  com- 
position. 

2 .  That  it  shall  be  stable  and  permanent  under  the  varying  con- 
ditions of  temperature  and  humidity  incident  to  service  storage 
and  use  for  all  time. 

3.  That  it  shall  be  sufficiently  rigid  to  resist  deformation  in 
transportation  and  handling. 

4.  That  it  shall  produce  a  higher  or  as  high  a  velocity  with 
as  low  a  pressure  as  the  service  charge  of  black  powder  for  a 
given  piece. 

5.  That  it  shall  be  incapable  of  undergoing  a  detonating 
explosion. 

6.  That  the  products  of  its  combustion  shall  be  nearly  if  not 
quite  gaseous  so  that  there  shall  be  no  residue  from  it  and  little 
or  no  smoke. 

7.  That  it  shall  produce  no  noxious  or  irrespirable  gases  or 
vapors. 

8.  That  it  shall  not  unduly  erode  the  piece  by  developing  an 
excessive  temperature. 

9.  That  it  shall  be  as  safe  as  gunpowder  in  handling  and 
loading. 

10.  That  it  shall  be  no  more  than  ordinarily  dangerous  to 
manufacture. 

Most  of  these  requirements  have  been  satisfied  in  several  of  the 
powders,  but  time  alone  can  determine  the  question  of  absolute 
stability  and  especially  as  the  comparison  is  instituted  with  gun- 
powder which  has  oeen  under  observation  for  over  500  years. 

We  can  and  do  apply  tests  whose  results  give  us  some  confi- 
dence as  I  did  when  I  exposed  Indurite  wrapped  in  felt  in  an 
iron  vessel  to  a  temperature  of  208°  F.  for  six  hours  without  its 
undergoing  change,  and  again  at  a  temperature  of  212°  F.  for 
twenty  hours  before  any  change  was  observed,  and  again  to  5** 
F.  without  its  being  affected. 


REVIEW.  837 

In  fact  from  the  outset  I  have  advised  the  application  of 
most  rigid  tests  and  drew  up  the  following  scheme  for  the  Navy 
Department  in  July,  1891,  by  which  to  test  Indurite. 

**  The  most  important  requisite  of  powder,  after  passing  the 
proof  test,  is  that  it  shall  retain  its  characteristics  under  all  the 
conditions  of  storage  or  transportation  which  may  obtain  in  the 
service  or  that,  if  any  change  does  take  place,  it  shall  not  cause 
the  powder  to  develop  under  the  **  proof*'  conditions  any  greater 
pressure  than  it  did  at  the  time  of  proving,  and  that  such  falling 
off  in  velocity  as  may  result  from  this  change  in  the  powder 
shall  not  be  relatively  greater  than  that  which  obtains  for  ser- 
vice black  powder,  and  shall  be  uniform  for  the  same  conditions 
of  exposure. 

"In  providing  for  this  test  I  would  first  prove  a  ten  pound 
lot  to  determine  the  maximum  weight  that  will  come  within  the 
limits  fixed  for  pressure  and  velocity,  and  then  I  would  load  1000 
Winchester  30.1  cal.  and  1000  Mannlicher  shell  with  a  charge 
some  grains  (say  five)  less  than  the  maximum,  so  as  to  be 
doubly  safe  in  case  the  pressure  should  become  increased  through 
the  treatment  to  which  the  powder  is  subjected. 

**The  loading  should  be  done  with  extreme  care  by  skilled 
workmen  in  an  especially  clean  and  uniformly  heated  and  dried 
room.  The  charges  should  be  weighed  on  chemical  balances 
and  with  all  the  precautions  surrounding  an  analytical  opera- 
tion. The  balls  should  be  weighed  and  gauged,  and  the  shell 
should  be  gauged  so  as  to  secure  as  nearly  absolute  uniformity 
as  possible,  while  the  caps  and  priming  (if  used)  and  wads 
should  be  identical  for  each  shell  of  each  1000  lot. 

**These  being  prepared,  I  would  pack  these  ball  cartridges  pre- 
ciselj'  as  if  ready  for  issue  to  the  service,  and  then  I  would  store 
385  Winchester's  and  385  Mannlicher's  in  the  regular  magazine 
at  the  Naval  Torpedo  Station,  and  the  same  number  of  the  same 
kind  in  the  regular  magazine  at  the  Naval  Ordnance  Proving 
Ground.  I  would  then  draw  from  the  magazine  at  the  Torpedo 
Station  tri'enty-five  Winchester's  and  25  Maunlicher's  and  fire 
them,  using  the  muskets  and  measuring  instruments  which  are 
to  be  used  throughout  the  trials,  and  I  would  repeat  this  trial 
every  month  for  three  years,  firing  ten  rounds  of  each  form  of 
ammunition  and  using  the  same  muskets  and  instruments 
throughout.  At  the  same  time  I  would  have  an  identical  set  of 
tests  made  at  the  Proving  Ground,  the  same  precautions  being 
taken  there  regarding  the  instruments  and  tools.  Throughout 
the  tests  a  close  watch  should  be  kept  on  the  magazines  by 
means  of  maximum  and  minimum  thermometers  so  that  if  abnor- 
mal results  are  obtained  in  firing  it  may  be  known  whether  or 


838  REVIEW. 

not  any  abnormal  conditions  have  obtained  in  tbe  magazine. 
This  series  of  tests  will  consume  1540  rounds.  It  would,  in  my 
judgment,  be  of  much  value  to  store  with  these  cartridges  and 
fire  with  them  an  equal  number  of  charges  of  standard  service 
black  powder,  to  be  used  as  a  standard  for  reference  by  which 
any  error  in  the  observations,  or  defects  in  the  instruments  may 
be  detected. 

'*I  would  take  eighty  rounds  of  the  Winchester*s  and  eighty  of 
the  Mannlicher's  and  place  them  in  an  oven  heated  to  140^  F. 
or  thereabouts.  At  the  end  of  one  month  twenty  of  each  are  to 
be  drawn  out  and  this  to  be  repeated  each  month  for  four 
months.  One  half  of  each  form  should  be  proved  at  the  Torpedo 
Station  and  the  other  half  at  the  Proving  Ground. 

**I  would  take  eighty  rounds  of  the  Winchester's  and  eighty  of 
the  Mannlicher's  and  subject  them  for  two  weeks  to  the  freez- 
ing temperature,  then  for  two  weeks  to  a  temperature  of  about 
140"^  F.,  and  then  draw  twenty  of  each,  and  this  should  be 
continued  until  the  last  forty  drawn  out  have  been  exposed  for 
eight  weeks  to  freezing  and  eight  weeks  to  the  high  tempera- 
ture. The  firing  trials  with  these  should  be  made  as  with  pre- 
ceding ones. 

"The  remaining  shell  should  be  stored  in  the  regular  magazine 
to  be  used  in  any  test  case  which  may  arise  or  in  any  manner 
suggested  by  the  results  obtained  in  the  tests  described  above. 

"In  the  meantime  tests  could  be  made  with  the  hand  cutS.  P. 
for  the  capacity  of  the  powder  to  resist  crumbling  and  dusting 
during  transportation  and  the  tendency  of  the  fixed  ammunition 
to  explode  en  masse  by  the  impact  of  projectiles,  or  by  the 
explosion  of  a  single  cartridge  in  the  midst  of  a  box  filled  with 
them.  The  first  can  be  effected  by  taking  a  pound  or  a  kilo- 
gram of  carefully  sifted  powder,  placing  in  a  copper  vessel  which 
it  only  partly  fills,  and  attaching  it  to  a  shaft  so  that  it  will  be 
continually  and  violently  shaken,  and  allowing  this  to  go  on 
every  working  day  for  a  week.  The  powder  can  then  be  sifted, 
using  the  same  mesh  as  before,  the  weight  of  the  dust  found  and 
the  percentage  of  dusting  for  the  given  circumstances  deter- 
mined. 

"In  the  trials  for  tendency  to  explode  ^;«wawtf  fifty  or  forty -five 
caliber  ammunition  can  be  used  and  the  weights  of  charges  need 
not  be  very  precise,  but  the  ammunition  should  be  packed  in, 
as  nearly  as  possible,  the  same  way  as  would  obtain  in  service 
practice." 

We  have  seen  that  the  development  of  smokeless  powder  has 
been  rendered  necessary  by  the  improvement  in  the  gun.  It 
now  appears  that  in  consequence  of  the  possession  of  the  powder 


REVIEW.  839 

we  must  further  improve  the  gun  for  we  cannot  in  our  present 
guns  utilize  all  the  energy  now  available.  Experiments  look- 
ing to  this  have  been  going  on  in  France,  where  in  a  Canet  ten 
cm.  gun  of  eighty  calibers,  withachargeof  12. 35poundsof  powder 
and  a  projectile  weighing  28.66  pounds  there  was  obtained  the 
extraordinary  muzzle  velocity  of  3366  feet  per  second,  while  the 
maximum  pressure  was  18.91  tons  per  square  inch.  Longridge, 
an  English  authority,  deprecates  the  lengthening  of  the  gun  as 
it  becomes  too  unwieldy  and  he  advocates  utilizing  the  energy 
of  the  gun  by  strengthening  it  so  it  will  endure  greater  pressures 
and  then  using  larger  charges.  He  points  out  that  if  this  Canet 
gun  were  reduced  to  forty-five  calibers,  and  strengthened,  we 
could  obtain  from  it  the  same  enormous  muzzle  velocity  by 
increasing  the  charge  to  thirteen  and  a  half  pounds,  though  the 
pressure  would  rise  to  twenty-five  tons  per  square  inch. 

What  the  result  will  be  where  authorities  of  standing 
disagree  is  impossible  to  foresee,  but  the  fact  is  demonstrated 
that  the  powder  is  now  more  highly  developed  than  the  gun, 
and  that  while  seeking  for  smokelessness,  we  have  secured  a 
propellent  which  is  capable  of  producing  much  higher  velocities 
than  gunpowder,  with  all  the  additional  advantages  of  flat  tra- 
jectory, increased  danger  area,  greater  accuracy,  and  greater 
range  which  follow  as  consequences. 


•J 
PS 

«< 


f6 
X 
Q 

> 

O 

p 

O 

M 

< 
> 

o 

z 
o 

cr, 

O 
0;. 

O 

o 


•9)mduo 


'fnniBH 


[voojvio 


mnipos 


'9fiO(nu93 


•aipviOA 


•9:|BooqjV3 
mnipos 


•91BJJXK 

tunnsB^Oti 


•a^BJiiK 
mnuBa 


'aoYIOdanO 


-ooi)03 


•aiiBK 


V 


^  fo  3*  »o   . 

•        •        •        •       • 

lO  to  lO  ^ 


g  »^  o 

•      •••••••••••     ^^  (^  00      ••••'••• 

..ooo 


OS  «   ^ 

•  •  • 

M    M     O 


4^     V 

u    u 


4; 
g 


u    O 


asoo«ooooiOfo(^*         .os»    •    .oo»oh.w^^>o 

«  * 

o  00  00   ^  ^  w?   ^  ***         •••••*••■••• 

fOiOM(^(^>OC>>*^ •       •       ■       •      • 

.^...l*"/.       •       ■       •       •       •       • •       ■ 

OOONOOO^C^OOOOO^*  •      *      I      I      I      I      I      •      •      \      I 

^uSiO»-00    »0^»O^OSO    <^     '*5I>.00    ^-^W     tN.     •     o^^ 

,,«..»T.  •••••••  ••••"•• 

^^fO«55•53.?5r>^c^w^eoO^^^^«^00«cl     •  «« 

^osioiii«oo   t^M   •^©^•^^fiPB   O   «oop   iM   Q  ^»ooo^   •- 
00t^0c*5fOw>S00^00or«*»-«^ciO0    ^»o«^ 

^oo^oo^«*-•-«c*c•««o«w^o•o^*ooo^^oo^*^r>. 


OS 


rO 


OS    OS    <JN 

j'^ «"  oo"  oo" 


>    u 


2a    •   S.^  o  ^qX'-»^cc'^«<«5 


«    »0  ro 


W  ^'   O^  «    .o  ^ 
"-» <i  ^     «    .    .    _ 


9     4> 


*      *      ~      *      *    rf     ij     '1     ^ 


o 

o 


s 


O 


V 


9 
O 


CO  CO 

CO       4^ 


es    CO 


SS  8>>-^8« 


pu  pu 
^  CO 

c  s 


a<0 


»*•   . 


Cm  •*i 

(X  o 


s    e    s    u     .  "^  '•^ 
u  *u  'C  CO  s   «r  «r 


4>    o 


C 
-   09 

O 


4>     O 


^  E-H 


u    u 

O    O 
O  U 


8s 

Os    4^     OS 

to  ^  o 

S  •§  X 

(S  #Q  oi 


8 

M 

•I 


9 


i 


g 

I 


•anqdwo 


-ii|xaxC|o 


tnmnH 


fvoaiviio 


mnfpos 


■MOinno 


•»m»lOA 


mnjpoe 


lunpva 


'ooiQOdnno 


4»    •    p 


•v«w 


^  t^  »n  O  Q 

«  •  ■         •  • 

«  r*  r*)  »^  ^ 

^  ^  ^  J*  * 


•  •       •      •      • 

•  •      •      •       • 

u 

^» 

t>  S\ 

•  •     •     •     *   -rt  o 
oS 

s « 

'••••■ 
•       •       ■       •        • 


•     o\  CO    ro  ^  >0  O 


V  C 

V  B  {£3 

5         '^      ^       .    .    .    . 


o   ^ 


8 


2  S* 


c  2 


•         •        •       ^      -J        •      -? 

M    rO   O^  00    0>   O   ^ 

M       M       M        M       fl|       M 


e*4  O^  «*)  O^^O 

fO  «o  to  1^    ^• 

,      •  •  •      • 

o  o  o  o  o 


e*5  e*)    I     if>  K. 


S 

m 

O 


f>   lO 


•       •       •       • 
M>    fO    »0    « 


0\  00    o 


00 


o  o 


M       M 


V9 

9  •- 

I" 

00 


:   :  9?5 


^P.'^. 


^   M   00     M   00 

^    f*  00   00  « 
lO   lO  ^ 


^0 

IN 

00 

• 

N 


•      •  « 

00    ov        ^ 


»^  o^'S  o6 

S:^  bo  «-" 
»A  ^  '^ 


o^  00  S  '^ 

8  00  Sd  '^ 
W   •-     lO 


So    **•  T^cioOoO    ''^S  cB    fo 

w    OS  9^  %n  t^  't  ^^  u  y6 

•••••••  "^ 

r* _-: 

(•^•■•••*  00 

s5    ^  O    O  >0  >0  00 

•         •         •         •         •         ,         •  • 

c«  «o  Q  ^  to  00  :: 


8.8^-' 


.   «   J?^ 


•*   -*  »^  ^     . 


►    > 

2  Z 


•>    ■>    • 


00  oc 


>  > 


^  OS    u 

g  B  E 

ij    cS    ej 


OS 


M  fo  0   »o  »i  «  1^ 

r^  ^  O^  w   O^  w  *> 

•  •  •     •     •^     »  • 

f^  fO  «     «     «     S  M 


P^   9S 


j?^  OS  i>  if  a 


51  « 
OS  0> 


CO  CO 


OS 

o 


2    M    V 

(ii  n  < 


tc  i5  ^  ^ 

«  ••  a  s 

hfi   00 

??*        Art  «k               "^ 

-<  <  g 


o 


o 
6 


u 
H 

1! 

o 

PQ 
I 


i! 

a 

ft) 

1 

O 

a* 

a 
*5 


! 
I 

> 


d 

-a 

I 

a 
2 

9 


a 

& 

a 

3 

9 


■« 

^ 


1 


i 
a 

3 


a 
a 


842 


REVIEW. 


< 


^?  f 


:S 


O 
twj     net OO     •     •     •w*';         nMioo0m«^0r)0r^<O«x-O 

o 


« 


9) 


S  <  s  s 

>  -^  2  •'  -  :  s 

O  ^    be  .  "g    .    6. 

Jl  -2^                 —  •*                       —                       ^ 

5  5'*'  *»                       "^                      t 

S5  ^  S              *"'  *                       w                       w 


CR      "i^bc       8       ...be       s..b«       c.bfl 

W    O      ^    «  -     S  «  -     =  «  .     S       -5 

> 

< 

<       <  (aiBjS  day  -^tia)  6  x   K 


X 


Dec 


A  ^     000C.00Q00)Air:>O000000og0CQQO00O0»'^O 


•a 

0. 


o 

c/: 

c; 

o 

X 


^3 


«; 


Ji 


•o 

flu 

r^ 

« 

•           .           • 

••        an 

«         •         • 

•           •          • 

«         <• 

Hotchkii 

• 

^ 

^ 

0 

6 

Z 

2; 

Ml 

o 

0 

• 

0 

• 

o 

X 

2 

W 

2: 

flu 

• 

as : 

-    a. 

•J3 

CO 

s 

tfi 

IT) 

C     55  25     as*:    a: g 

fe  *  «  ■ 

flu             flu      fi^   "^  a<             Gu 

'ii                              'j:      CO   S  tfi             QQ 


RBVIBW. 


843 


.^5  -  ►. 


If  lg|lfll5tnil1lfll§«lllltll?IIS 


^1 


•         «         ■ 


•*!  •O   fO  W  •'J   r*?  'O 


•5  X  "N      -68 


•n|»jB  dug -^IQ  n  x  N 


I 


s 


R 


0< 


.3 

.a 
u 


00 


S. 


■ 

80 


0) 


o 


I 

09 


844 


REVIEW. 


.il 


ooio^M<eMOiooo«oegoootf}«ooo»4ioo^iO(^io«ot« 


O    ••    rt  s  ^ 

M      M     M      M     ■« 


^  : 


««««*o««^«o««««*o«o««^«««o«« 


^^^n^t^mtt^n^ 


ispmod  paofvaji 

dug  |o  tmvij  fie 

"I  .X  'K  mnviJ  8  Xninp^  dJjvqa  JBrnmua 

I 


8 

a 

K 


I 


II 

0)  M 


I 


I 


I 


I 

■ 

0) 


is 


o  o  d 

zzz 

pit  0<  »« 

0  tt  tri 


RBVISW. 


845 


R 


If. 


»» 


•       •••••• 


SH^''9Q*'>*^OdrooooBeaoi'>««<o**5t<iMtf»ciin09^« 

■^       SA         mMMM  MMMMMtaM'N  SmM  MM  MMM  MMM  M 


it   z   : 


*      i      * 

»     4      « 


Q^>  MMMMMMM 


.1 


•4 


< 


Q 
SB 

M 

fib 

o 


V 

i 

E 


X«       «       «       •       «       «       • 
•       «       «       •       a       •       • 


B  > 


* JV9 J  )B  pa^f oJBi 


t 


vi.rjo.'jgj'y^poo   g   g  y  g 


6 


■•    V    •    ^  •    «    •    •    * 

s 

3 


• 

^« 

u 

t 

Kl 

£ 

t  s 

% 

• 

1 

>: 

• 

&i 

M 

Si 

M 
1        V            .     •       •        •        < 

. . . !. 

M 
^^ 

M 

fc   *     -     -      - 

Si 

•      «      •       V)  « 

.       .       «        M    •       < 

Z 
*    SB  -    - 

5 

:    :    X    3 

|- 

0 
SB 

^ 

"  ■ 

o: 

o; 

• 

• 

■ 

0< 

(C 

if 

« 

■ 

• 

n 

9 

846  REVIEW. 


I 


Q 

z 

o 

o       .     . 
•&  .0 


^    ft  8  'S)^  5  f»   ej  <0    <♦  y 

•4MV4W  »•  •«>«MM>4««»« 


I 


II. £ 

s  tl 

o   ^  I 

<  G 

> 
< 

Z 


■««  10    CO 


^ 


g 


5    I  :    :    1   :    •-    -    •    r  "j  = 


3  ^ 


a 


d  § 

^                          .^:  r    :    :    :    :    J    " 

s a  s 

•                                       •  «* 

(0                     (fl  tfi 


ACIDITY  OF  niLK  INCREA5E0  BY  BORACIC  ACID. 

By  B.  H.    PARRIIfOTON. 
Receired  July  n,  1S96. 

WHILE  making  some  investigations  with  milk  preserva- 
tives, the  writer  noticed  that  sweet  milk  in  which  a 
small  quantity  of  boracic  acid  (preservaline)  was  dissolved « 
required  what  appeared  to  be  an  abnormally  large  quantity  of 
one-tenth  normal  alkali  to  neutralize  it  and  much  more  than 
water  in  which  the  same  amount  of  **  preservaline  *'  was  dis- 
solved. One-half  gram  of  preservaline  was  dissolved  in  500  cc. 
water,  and  twenty  cc.  of  this  solution  required  one  cc.  of  one- 
tenth  normal  alkali  to  produce  the  pink  color,  when  phenol- 
phthalein  was  used  as  an  indicator  in  titrating.  Before  adding 
preservaline  the  water  had  a  neutral  reaction. 

One-half  gram  preservaline  was  dissolved  in  500  cc.  of  sweet 
milk,  and  twenty  cc.  of  it  required  eight  cc.  of  one-tenth  normal 
alkali  to  give  the  pink  color,  although  before  adding  the  preserv- 
aline twenty  cc.  of  this  same  milk  gave  the  pink  color  with  onl}^ 
four  cc.  of  one-tenth  normal  alkali. 

The  same  amount  of  preservaline  increased  the  acidity  of  a 
given  quantity  of  milk  four  times  as  much  as  it  did  the  acidity 
of  the  water. 

The  writer  is  unable  to  explain  this  reaction,  but  it  gives  a 
simple  means  of  detecting  preservaline  or  boracic  acid  in  milk, 
as  normal  milk  will  smell  or  taste  sour  when  it  contains  as  much 
natural  acidity  as  is  represented  by  eight  cc.  of  one-tenth  normal 
alkali  to  twenty  cc.  milk.  This  represents  0.36  per  cent,  lactic 
acid,  and  it  can  be  safely  stated  that  milk  which  contains  over 
three-tenths  per  cent,  lactic  acid  and  neither  tastes  or  smells 
sour,  has  been  adulterated  with  some  preservative,  probably 
boracic  acid. 

Dairy  School, 
Ukivbrsity  op  WiacoifsiN. 


848  BOOKS   RECEIVED. 

CORRESPONDENCE. 


United  States  Department  of  Agriculture, 

Division  op  Chemistry, 

Washington,  D.  C,  July  23,  1896. 

Editor  Journal  of  the  American  Chemical  Society,'  Easion^  Pa,: 

Dear  Sir  : — A  majority  of  the  Executive  Committee  has 
decided  to  call  the  annual  meeting  of  the  Association  of  Official 
Agricultural  Chemists  for  Nov.  6,  7  and  9,  1896.  These 
dates  immediately  precede  the  meetings  of  the  Association  of 
Agricultural  Colleges  and  Experiment  Stations,  which  will  con- 
vene in  Washington,  Nov.  10.  The  session  will  be  held,  as 
heretofore,  in  the  lecture  hall  of  the  National  Museum,  at 
Washington,  D,  C. 

The  Ebbitt  House  offers  to  the  Association  entertainment  at 
the  rate  of  93.00  per  day  and  the  free  use  of  its  parlors  for  com- 
mittee meetings,  if  desired.  The  Ebbitt  House  is  on  the  cor- 
ner of  **  F  "  and  14th  Sts.,  and  can  be  reached  from  all  stations 
by  the  **F"  St.  or  Avenue  cars. 

Respectfully, 

H.  W.  Wiley. 

Secretary  A.  O.  A.  C. 


BOOKS  RECEIVED. 

Bulletin  No.  42.  Second  Series.  Horticulture.  Louisiana  State  Ex- 
periment Station,  Baton  Rouge,  La.     1896.    44  pp. 

Bulletin  No.  123.  Examination  of  Pood  Products  sold  in  Connecticut. 
Connecticut  Agricultural  Experiment  Station,  New  Haven,  Conn.  Jul^-, 
1896.     79  pp. 

Transactions  of  the  American  Institute  of  Mining  Engineers.  Vol. 
XXV.  February  to  October,  1895,  inclusive.  New  York  City :  Published 
by  the  Institute.     1896.    xvii,  1068  pp. 

Part  III.  Geology  and  Agriculture.  A  preliminary  report  upon  the 
Florida  parishes  of  East  Louisiana  and  the  Blufif.  Prairie  and  Hill  Lands 
of  Southwest  Louisiana.  By  VV.  W.  Clendeuin,  A.M.,  M.S.,  Geologist. 
Louisiana  Experiment  Station,  Baton  Rouge,  La.    96  pp. 

ERRATA. 

On  page  667,  August  number,  12th  line  from  bottom,  in  equation,  for 
o.^b  read  o.^y. 
On  page  668,  2nd  line  from  top,  for  Dividing  read  Multiplying, 


Vol.  XVIII.  [October,  1896.]  No.  10. 


THE  JOURNAL 


OF  THE 


AMERICAN  CHEMICAL  SOCIETY. 


SOriE  EXTENSIONS  OF  THE  PLASTER  OF  PARI5  nETHOD 

IN  BLOWPIPE  ANALYSIS. 

By  W.  w.  Andrews. 

Received  August  7,  i8s)6. 

IN  the  years  1883  and  1884  two  papers  were  published  by  Dr. 
Eugene  Haanel,  of  Victoria  College,  Cobourg,  Ontario,  now 
of  Syracuse  University,  in  the  Proceedings  of  the  Royal  Society 
of  Canada,  in  which  he  described  the  brilliant  results  he  was 
able  to  obtain  in  the  production  of  the  Bunsen  iodide  films  on 
the  blowpipe  support  then  proposed  for  the  first  time ;  namely, 
thin  tablets  of  plaster  01  Paris  made  by  casting  sheets  three-six- 
teenths of  an  inch  thick  on  panes  of  glass  and  scratching  them, 
before  hardening,  with  ruled  lines,  so  that  when  set  they  would 
readily  break  into  oblongs  measuring  two  and  one-half  by  one 
and  one-quarter  inches.  The  pure  white  and  highly  polished 
surface  of  these  tablets  and  its  great  power  of  condensing  heated 
gases  and  exhibiting  the  true  colors,  their  cheapness,  thermal 
and  hygroscopic  properties  of  the  tablets,  the  ease  with  which 
they  may  be  prepared  and  carried,  and  the  excellence  of  the 
results  when  the  sublimed  iodides,  bromides,  oxides  and  sul- 
phides are  deposited  as  coatings  upon  them,  make  them  an  ideal 
form  of  support  in  blowpipe  work. 

A  small  pit  is  made  at  one  end  of  the  tablet  somewhat  larger 
than  a  pin's  head,  and  in  this  the  ore  to  be  tested  is  heated. 
The  oxide  coatings  are  produced  by  heating  the  substance  per 
se^  the  bromides  by  adding  to  the  substance  a  drop  of  fuming 
hydrobromic  acid,  and  the  iodides  by  adding  a  strong  solution 


850  W.   W.   ANDREWS.      PLASTER   OF   PARIS 

of  hydriodic  acid  (made  by  dissolving  five  ounces  of  metallic 
iodine  in  seven  ounces  of  water,  by  passing  a  steady  stream 
of  hydrogen  sulphide  through  the  solution  while  the  iodine  is 
slowly  added).  All  who  have  experimented  with  this  solution 
will  be  ready  to  admit  that  it  yields  superb  results,  but  thougli 
easily  renewable  when  one  is  near  a  hydrogen  sulphide  gener- 
ator it  is  very  unstable,  takes  a  long  time  to  prepare  and  is 
troublesome  to  carry. 

In  1890  Mr.  F.  A.  Bowman  read  a  paper  before  the  Nova 
Scotia  Natural  History  Society,  in  which  was  described  a  search, 
for  a  solid  reagent  to  replace  the  hydrogen  iodide  solution.  He 
found  that  potassium  hydrogen  sulphide  or  any  alkaline  sul- 
phate, which  does  not  yield  a  coating  of  its  own,  mixed  with 
potassium  iodide  would  do  very  well .  He  also  found  that  microcos- 
niic  salt  and  potassium  iodide  gave  good  results.  This  mixture 
i  s  a  favorite  one  with  some  blowpipe  experts .  Tin  is  the  only  metal 
in  the  three  series  of  the  periodic  table,  beginning  with  copper, 
silver  and  gold,  which  does  not  yield  a  characteristic  coating- 
with  this  reagent. 

The  writer  has  not  been  able  to  find  whether  there  have  been 
any  other  reagents  besides  these  seriously  proposed.  Plaster  of 
Paris  as  a  support  is  mentioned  in  Moses  and  Parsons'  late  work, 
as  an  alternative  to  charcoal.  This  is,  as  far  as  known  to  the 
writer,  the  only  standard  work,  in  which  the  colors  of  the  films 
on  the  tablets  are  described. 

In  the  rapid  development  of  other  methods  in  chemical  work 
the  blowpipe  has  fallen  largely  into  disuse,  and  for  many  years, 
besides  the  work  outlined  above  and  that  of  Col.  Ross  and  some 
valuable  tests  for  individual  elements  proposed  by  Chapman, 
little  or  no  advance  has  been  made.  There  are  two  possible 
lines  of  future  progress  in  blowpiping,  one  in  the  direction  of 
increased  power  and  simplicity,  so  as  to  make  the  method  more 
valuable  for  the  field  work  of  the  mineralogist,  geologist  and 
prospector,  and  the  other  in  the  direction  of  increased  range 
and  delicacy  until  the  dry  way  tests  rival  the  delicacy  and  dis- 
tinctiveness of  the  wet  tests,  as  they  surpass  them  in  expedi- 
tiousness.  It  may  not  be  amiss,  therefore,  to  call  attention 
to  the  instrument  of  Plattner  and  Berzelius,  which,  in  its  mod- 


METHOD   IN  BI^OWPIPK  ANAI^YSIS.  85 1 

em  form  as  the  hot-blast  blowpipe  and  with  the  new  support 
and  the  new  reagents  and  reactions  now  known  to  chemistry, 
is  an  instrument  surpassed  by  the  electric  furnace  only. 

The  cleanliness  of  the  method  here  described,  as  compared  with 
the  charcoal  method  and  the  quickness  with  which  sure  results 
can  be  obtained  with  very  small  amounts,  should  call  the  blow- 
pipe back  to  the  table  of  the  chemist  for  preliminary  and  con- 
firmatory tests,  to  class  work  as  an  accompaniment  of  the  wet 
methods,  and  to  the  lecture  table  for  the  purposes  of  illustration. 
It  is  possible  to  detect  five  or  six  metals  in  presence  of  each 
other  on  one  tablet.  Many  of  the  coatings  are  permanent  and 
are  all  renewable  on  reheating  with  addition  of  a  drop  of  the 
reagent,  so  that  a  set  of  tablets  carefully  labelled  with  a  pencil 
forms  a  permanent  record  of  a  set  of  experiments.  The  value 
of  this  to  the  practical  chemist  and  to  the  student  need  not 
be  emphasized.  It  may  be  noted  that  blowpiping  is  so  much  of 
an  art  that  new  methods  are  seldom  well  enough  practiced,  by 
those  who  have  become  skillful  in  other  methods,  to  reveal  their 
value. 

The  extensions  of  the  plaster  of  Paris  method  here  proposed 
are  :  A  set  of  new  reagents,  which  yield  some  new  reactions 
which  are  of  value  in  detecting  elements  in  the  presence  of  each 
other,  notably  gold  and  copper  in  very  small  amounts  in  the 
presence  of  all  elements  so  far  experimented  with  ;  arsenic,  tin 
and  antimony  in  presence  of  each  other;  sulphur  in  the  presence 
of  selenium  and  tellurium,  and  chlorine,  bromine  and  iodine  in 
the  presence  of  each  other  ;  a  new  set  of  film  tests  which  are 
found  to  be  of  great  delicacy  (the  limits  of  delicacy  are  now 
being  measured,  it  being  found  that  gold,  one  part  in  one  mil- 
lion, and  copper,  one  part  in  four  millions,  are  easily  detectable) ; 
a  change  in  the  composition  of  the  tablets  which  does  away  with  the 
necessity  for  using  platinum  wire  in  the  production  of  the  colored 
glasses  with  borax  and  metaphosphoric  acid,  these  being  formed 
on  the  tablets  with  a  decided  gain  in  facility  and  deb'cacy;  and 
lastly  several  new  methods  of  handling  the  tablets  themselves. 

It  is  evident  that  solid  reagents  will  always  be  the  more  con- 
venient to  carry  afield,  but,  in  the  laboratory,  liquids  are  to  be 
preferred,  since  they  are  more  readily  applied,  and  when  the 


852  W.   W.   ANDREWS.      PLASTER  OF  PARIS 

assay  is  heated »  tbe  reagent,  which  has  soaked  into  the  tablet, 
is  fed  steadily  toward  the  hot  portion  of  the  tablet,  so  that  the 
heated  assay  is  constantly  enveloped  in  the  vapor  of  the  reagent. 
For  over  two  years  the  writer  has  used  with  satisfaction  the 
following  reagents,  which  have  been  selected  from  a  score  of 
experimental  ones.  They  are  stable  and  almost  odorless,  can 
be  carried  to  the  field  in  a  solid  form  and  so  used  if  need  be, 
while  a  few  seconds  suffice  to  prepare  them  in  liquid  form  if  it  be 
desired  so  to  use  them. 

The  chief  reagent  is  a  saturated  solution  of  iodine  in  a  strong 
solution  of  potassium  thiocyanate  in  water.  The  solution  takes 
place  almost  instantly  and  with  great  absorption  of  heat.  The 
bottleful  now  in  use  has  been  in  use  for  over  two  years,  a  little  of 
one  or  other  of  the  ingredients  being  added  from  time  to  time  as 
seemed  to  be  required.  Exact  proportions  are  not  necessary  to 
the  efficiency  of  the  reagent.  It  can  be  prepared  on  the  field 
from  the  solid  chemicals  at  a  moment's  notice.  The  brilliancy 
of  the  iodide  films  produced  with  this  solution  are  not  one  whit 
behind  those  possible  with  the  pure  solution  of  hydriodic  acid. 
Its  coatings  tend  to  form  in  definite  bands  of  color.  The  spheres 
of  desposition  of  the  iodide  and  the  oxy-iodide  are  sometimes  very 
well  defined.  Some  striking  and  important  variations  are  pro- 
duced by  the  presence  of  the  potassium  thiocyanate,  for  example, 
with  molybdenum,  osmium,  iridium,  tin,  antimony,  lead,  bis- 
muth, cadmium  and  mercury. 

Dr.  Haanel  showed  in  his  second  paper  that  by  means  of 
hydrobromic  acid,  copper  and  iron  could  be  detected  at  one  oper- 
ation in  the  presence  of  each  other  and  in  the  presence  of  nickel 
and  cobalt  and  any  other  flux-coloring  substances.  Instead  of 
the  fuming  acid  with  its  dangerous  properties,  a  mixture  in 
molecular  proportions  of  powdered  potassium  bromide  and  meta- 
phosphoric  acid,  or  potassium  hydrogen  phosphate  or  sulphate 
may  be  used.  This,  suggested  by  Bowman's  work,  suggests 
further  a  set  of  solid  reagents,  made  by  using  potassium  chlo- 
ride, potassium  fluoride  and  potassium  iodide  with  metaphos- 
phoric  acid,  and  these  form  a  valuable  set  for  special  tests. 
They  have  the  advantage  of  yielding  at  once  the  colored  flux 
and  the  coatings  produced  by  any  volatile  matter  in  the  assay. 


METHOD   IN   BLOWPIPE   ANALYSIS.  853 

When  heated,  the  reaction  represented  by  the'following  general 
equation  takes  place :  KX+  HPO,  =  KPO,  +  HX 

These  two  reagents,  the  iodine  solution  and  the  bromide 
mixture,  suffice  for  the  production  of  coatings.  The  following 
which  are  used  to  differentiate  them,  are  dropped  upon  the 
oxide  or  iodide  film  and  colored  spots  are  produced,  or  the  color 
is  discharged  to  white  (technically,  wiped),  or  the  coating  dis- 
appears through  solution  and  absorption  by  the  tablet. 

Dr.  Haanel  used  ammonium  hydroxide  and  yellow  ammonium 
sulphide  for  the  purpose  of  testing  the  solubility  of  the  films  and 
to  produce  the  sulphide  spots.  Both  of  these  are  troublesome  to 
carry,  and  the  latter  is  objectionable  on  account  of  its  intolera- 
ble odor,  its  instability  and  the  fact  that  for  its  renewal  the 
hydrogen  sulphide  generator  is  required.  It  has  been  found  that 
a  solution  of  potassium  sulphide,  strong  enough  to  show  a  clear 
amber  color,  made  by  dissolving  the  solid  potassium  sulphide  in 
water,  or  by  boiling  a  strong  solution  of  potassium  hydroxide 
with  an  excess  of  flowers  of  sulphur  till  the  solution  assumes  a 
blackish  color,  which  on  cooling  will  be  amber  yellow,  fulfils  all 
the  required  conditions.  If  through  the  action  of  light  it  is 
decomposed,  all  that  is  necessary  for  its  renewal  is  to  boil  the 
solution  and  perhaps  add  a  little  sulphur.  We  therefore  have  a 
reagent  which  can  be  carried  as  a  solid,  can  be  renewed  any- 
where, is  as  efficient  as  the  ammonium  sulphide  solution  and  is 
almost  odorless. 

In  the  place  of  ammonium  hydroxide  a  solution  of  potassium 
cyanide  is  used,  made  a  little  more  stable  by  the  addition  of  a 
little  ammonium  or  potassium  hydroxide.  Besides  these  the 
common  acids  of  the  laboratory  are  useful  and  a  solution  of 
potassium  thiocyanate. 

The  potassium  thiocyanate  solution  is  used  in  two  ways.  It 
is  either  dropped  on  the  coating  to  test  its  solubility  and  to  note 
the  colors  produced  after  heating,  or  it  is  dropped  on  the  tablet 
before  the  coating  is  deposited,  and  then  the  hot  vapors  sweep- 
ing over  the  moist  spot,  give  with  some  metals  characteristic 
reactions. 

Those  coatings  which  are  pure  white  and  therefore  invisible 
on  the  white  tablet,  are  examined  on  a  tablet  which  has  been 


854  W.   W.   ANDREWS.      PLASTBR  OF  PARIS 

smoked  in  a  flame,  or  on  one  streaked  up  the  middle  by  means 
of  a  glass  rod  which  has  been  dipped  in  a  solution  of  boric  and 
metaphosphoric  acids  mixed  with  lampblack  or  bone  charcoal. 
In  this  way  the  coatings  may  be  viewed  on  a  white  and  on  a 
black  surface  at  the  same  time. 

In  order  that  the  colored  fluxes  may  be  made  on  the  tablets,  the 
latter  must  be  made  more  resistant  to  the  dissolving  effect  of  the 
metaphosphoric  acid  and  the  alkali  in  the  borax.  If  one  teaspoonful 
of  boric  acid  be  added  to  each  quart  of  the  water  used  in  mak- 
ing the  tablets,  they  will  be  found  to  be  denser  and  to  have  the 
necessary  quality.  Borax  can  be  fused  on  them  without  gath- 
ering any  impurities  from  the  plaster  and  if  metaphosphoric  acid 
be  substituted  for  phosphor  salt,  we  have  a  flux  which  will  spread 
upon  the  tablet  and  exhibit  the  colors  of  all  degrees  of  satura- 
tion at  the  same  time.  This  reagent,  first  proposed  by  Ross, 
who  described  its  reactions,  is  preferable  to  microcosmic  salt, 
since,  as  it  contains  no  volatile  matter  and  melts  readily  to  a  clear 
glass,  it  will  show  by  effervescence  the  presence  of  water  or  car- 
bon dioxide,  or  other  gas  in  a  mineral.  With  cobalt  it  yields  a 
fine  violet  when  cold,  which  becomes  blue  on  the  addition  of 
any  of  the  alkali  metals,  for  which  therefore  it  furnishes  a  ready 
test.  The  only  objection  to  this  reagent  is  the  tendency  of  the 
sticks  to  deliquesce,  but  a  piece  can  be  kept  in  a  corked  test- 
tube,*  which  can  be  readily  dried  over  the  flame,  if  dampness 
should  gather.  In  dry  weather  it  causes  no  trouble.  Its  sol- 
vent power  is  very  great  and  the  colors  are  fine.  Ross  asserts 
that  silica  and  zirconia  are  the  onl}'  oxides  which  are  not 
soluble  in  this  flux.  The  whole  operation  may  be  completed  in 
the  time  usually  required  to  form  the  bead  in  the  platinum  wire 
loop  and  the  volatile  oxide  films  will  be  found  on  the  tablet  above  the 
glass,  where  they  may  be  tested  with  potassium  sulphide  and 
the  other  reagents.  One  operation,  -therefore,  suffices  for  the 
determination  of  the  volatile  acid  elements,  the  volatile  metal  or 
metals,  and  flux-coloring  metal.  Metaphosphoric  acid  well  re- 
places   potassium    hydrogen    sulphate    in    the    operation    as 

1 1n  this  laboratory  each  student  is  supplied  with  a  set  of  very  small  dipping  tubes 
and  a  wooden  block  into  which  holes  are  bored  for  the  reception  of  a  set  of  test-tubes 
closed  with  paraffined  corks,  to  hold  the  reagents. 


METHOD   IN   BLOWPIPE   ANALYSIS.  855 

described  in  most  text-books  for  the  detection  of  carbon  mon- 
oxide, carbon  dioxide,  iron,  chlorine,  bromine,  iodine,  nitrogen 
tetroxide,  chlorine  tetroxide, sulphur  dioxide,  h5'drogen sulphide, 
hydrocyanic  acid  and  acetic  acid. 

DESCRIPTIVE    LIST    OF    REACTIONS    OBTAINABLE    ON    THE 

TABLETS. 

Copper/^rj^  yields  with  diflBculty  a  coating  of  volatilized  metal. 
With  the  iodine  solution  it  yields  a  white  iodide  coating  and  an 
emerald  g^een  flame.  The  iodide  treated  with  a  drop  of  potas- 
sium sulphide  gives  with  gentle  heat  a  blackish  gray,  which  is 
removed  by  greater  heat.  Potassium  cyanide  and  nitric  acid  dis- 
solve the  sulphide;  hydrochloric  and  sulphuric  acids  have  no  effect 
till  heated  and  then  they  remove  the  spot.  Potassium  thiocyanate 
applied  to  the  coating  has  no  effect  till  heated,  when  a  gray  spot 
is  shown.  Any  part  of  the  coating  touched  with  the  tip  of  the 
flame  shows  the  emerald  green  flame  (Haanel).  Metaphos- 
phoric  acid  glass  is  greenish-blue  when  hot  and  a  fine  robin's 
egg  blue  when  cold.  Metaphosphoric  acid  and  potassium  bro- 
mide yield  a  splendid  reddish  violet  coating  of  copper  bromide 
(compare  osmium) .  The  bromide  plus  potassium  sulphide  shows 
a  brown,  which  if  heated  turns  blackish  and  then  green,  not 
affected  by  sulphuric  acid,  but  immediately  destroyed  bj'  a  drop 
of  nitric  acid. 

Copper  plus  metaphosphoric  acid  and  potassium  chloride 
yields  a  yellow  brown  cupric  chloride,  which,  if  treated  with  a 
drop  of  potassium  thiocyanate,  gives  a  black  ring,  which,  if 
heated,  becomes  a  black  spot.  If,  before  the  assay  is  heated, 
a  drop  of  nitric  acid  be  placed  one-half  inch  from  the  assa}^  and 
a  drop  of  potassium  thiocyanate  be  placed  above  that,  on  heat- 
ing a  fine  and  very  volatile  blue-black  coating  is  deposited  far 
up  the  tablet.  This  blue-black  is  not  affected  b)'  acetic  acid,  is 
wiped  off  by  sulphuric  acid  slowly,  and  immediately  by  hydro- 
chloric and  nitric  acid.  The  formula  of  this  compound  will  be 
determined  if  some  method  be  found,  by  which  it  may  be  col- 
lected in  quantity.     (See  chlorine.) 

Silver  gives /^rj^  a  pinkish  gray  coating,  which  touched  by 
the  blowpipe  flame  (flamed)  becomes  mottled  brown.     Reduced 


856  W.   W.    ANDREWS.      PLASTER   OF   PARIS 

globules  are  often  shown.  Metaphosphoric  acid  yields  the  same 
coating  and  a  pearl-like  glass.  The  iodine  solution  yields  a  pale 
yellow,  paler  when  cold,  and  around  the  assay  forms  a  black, 
which  does  not  fuse  into  the  tablet  (compare  lead).  Flaming 
with  oxidizing  flame  yields  a  mottled  brown  anywhere  on  the 
tablet.  This  is  a  very  delicate  test  and  as  all  qther  coatings  are 
volatile,  the  flame  drives  them  off  and  leaves  the  silver  oxide. 
Potassium  sulphide  produces  a  spotted  blackish  brown,  proba- 
bly potassium  silver  sulphide,  the  analogue  of  ammonium  silver 
oxide,  for  if  treated  with  a  drop  of  potassium  cyanide  it  immedi- 
ately disappears,  but  if  it  be  first  heated,  the  potassium  cyanide 
has  no  effect.  If  only  one-half  of  the  sulphide  spot  be  touched 
with  the  tip  of  the  flame  and  then  the  potassium  cyanide  be  ap- 
plied, the  untouched  portion  will  disappear  while  the  other  half 
will  remain.  Potassium  thiocyanate  on  the  iodide  wipes  it  ofi; 
when  heated  the  spot  turns  black,  which  is  not  wiped  oS  by 
potassium  cyanide. 

Gold  is  slightly  volatile /^r  se  and  more  so  if  a  solution  of 
iodine  in  potassium  iodide  be  used  as  a  reagent,  and  the  result  is 
a  fine  rose-colored  film  of  the  metal.  If  potassium  thiocyanate 
be  present,  no  volatility  is  noticed.  Gold  and  the  other  ele- 
ments which  respond  to  the  new  tests  will  be  the  subject  of 
another  paper. 

Zinc  per  se  yields  a  white  coating,  not  very  volatile  and  lumin- 
ous yellow  when  hot.  Potassium  sulphide  and  potassium  thio- 
cyanate produce  no  visible  change  on  zinc  films.  The  iodide 
film  is  a  white,  which  treated  in  any  part  with  cobalt  nitrate 
solution  yields  the  well-known  zincate  of  cobalt,  which  is  quickly 
decomposed  by  a  drop  of  nitric  acid  (compare  tin).  This  reac- 
tion obtained  in  this  way  is  decisive  for  zinc,  as  aluminum  and 
silicon  do  not  volatilize  and  are  therefore  not  present  in  the  coat- 
ing. In  the  metaphosphoric  acid  glass,  zinc  causes  flashes  of 
light  and  detonations  (Chapman).  Metallic  zinc  sometimes 
yields  per  se  a  black  sublimate  along  with  tlye  white  oxide 
(compare  arsenic.) 

Cadmium /^r  se  yields  one  of  the  most  beautiful  of  the  oxide 
films,  which  consists  of  a  rich  brown  with  black  farther  away 
and  somewhat  iridescent  near  the  assav.     Acetic  acid  does  not 


METHOD  IN   BI^OWPIPE   ANALYSIS.  857 

afiect  it ;  potassium  cyanide  dissolves  it  at  once  (compare  cad- 
mium sulphide).  Potassium  sulphide  and  potassium  thiocya- 
nate  3rield  a  scarlet  when  hot,  and  bright  yellow,  cold.  This  cad- 
mium sulphide  is  not  affected  by  potassium  cyanide,  is  quickl}' 
destroyed  by  nitric  acid,  less  readily  by  hydrochloric  acid, 
immediately  by  acetic  acid  (compare  cadmium  oxide) ,  and  is 
not  affected  by  sulphuric  acid  (compare  copper). 

The  iodide  coating  is  white  with  well-defined  borders,  which 
is  easily  distinguished  in  the  presence  of  other  white  coatings 
by  the  per  se  and  sulphide  reactions.  In  the  assay  and  near  it 
the  sulphide  reaction  will  be  seen  caused  by  the  potassium  thio- 
cyanate  in  the  iodine  solution  (see  sulphur).  In  metaphos- 
phoric  acid  cadmium  acts  like  zinc  and  yields  at  the  same  time 
its  oxide  coating  beyond  the  glass. 

Mercury  gives  per  se  a  very  volatile  film  of  mercur}*  snow, 
which,  with  a  feather,  may  be  swept  into  a  globule.  It  is  not 
affected  by  the  other  reagents. 

The  iodide  coating  is  a  splendid  combination  of  scarlet,  yellow, 
and  velvety  green.  This  is  caused  by  the  mixing  of  the  green 
mercurous  iodide  with  the  scarlet  and  yellow  forms  of  the  mer- 
curic iodide.  The  reactions  of  each  kind  of  iodide  may  be 
obtained  on  the  one  tablet.  The  green  and  the  scarlet  are  the 
stable  forms  into  which  the  coating  changes  on  standing.  A 
drop  of  the  reagent  or  some  more  of  the  vapor  blown  across  the 
coating  changes  all  into  the  scarlet  form.  With  mercurous 
iodide,  sulphuric  acid  gives  a  yellow  spot  (mercurous  sulphate). 
Potassium  hydroxide  gives  a  black  ;  so  does  ammonium  hydrox- 
ide, (iodomercurosamine,  NH,Hg,I),and  potassium  sulphide. 
With  the  mercuric  iodide,  sulphuric  acid  increases  the  amount 
of  the  scarlet,  potassium  hydroxide  yields  a  white,  as  does  ammo- 
nium hydroxide  (iodomercurosamine)  and  potassium  sulphide, 
yield  a  white  spot,  quickly  turning  black.  The  sulphide  spot, 
strange  to  say,  is  partiall}*^  dissolved  in  nitric  and  hydrochloric 
acid,  while  sulphuric  acid  turns  it  brownish.  Potassium  cyanide 
yields  a  black  and  potassium  thiocyanate  a  dark  spot,and  if 
heated  both  are  wholly  volatilized  (compare  lead,  bismuth,  and 
silver).   Water  has  no  effect  on  this  coating  (compare  lead),  nor 


858  W.    W.   ANDREWS.      PILASTER  OF   PARIS 

have  hydrochloric,  nitric,  or  sulphuric  acids.      By  the  last  the 
coating  is  not  readily  wetted. 

Gallium  has  not  been  experimented  with.  Indium  yields  a 
pale  yellow  iodide  coating  and  a  blue  flame. 

Thallium  per  se  yields  a  feathery  brown  with  white  farther 
away  and  a  green  flame  (compare  arsenic  and  tellurium) .  Potas- 
siu  1  sulphide  gives  a  terra  cotta  brown  spot  with  a  black  ring. 
Potassium  cyanide  and  potassium  thiocyanate  have  no  effect  upon 
it.  The  iodide  film  is  an  ^^'g  yellow  with  a  purple  black  veil  farther 
away.  Potassium  sulphide  gives  a  rich  brown  which  potassium 
cyanide  darkens.  Hydrochloric  acid  discharges  it  slowly  and 
yellow  is  left  (compare  bismuth  and  tellurium).  Potassium 
thiocyanate  has  no  effect  on  the  yellow  or  the  black  till  heated, 
when  it  yields  a  white  (compare  bismuth,  tellurium,  tin,  and 
lead).  Potassium  cyanide  dissolves  the  black  but  has  no  effect 
on  the  yellow.  Sulphuric  acid  has  no  effect.  A  drop  of  the 
reagent  on  the  coating  heated  shows  a  spreading  black  and  an 
orange  ring. 

Carbon  yields  a  sooty  coating,  which  comes  better  if  sulphuric 
acid  or  metaphosphoric  acid  be  used  upon  the  assay.  In  the 
case  of  the  carbonates,  boric  oxide  or  metaphosphoric  acid  yield  an 
odorless  effervescence  (Ross,  Chapman) .  Organic  acids  blacken 
the  tablet  when  heated. 

Silicon.  An  interesting  reaction  given  by  the  silicates,  espe- 
cially the  hydrous  forms,  is  being  investigated.  Chapman  dis- 
solves a  silicate  in  boric  oxide  and  then  precipitates  the  silica 
by  adding  metaphosphoric  acid. 

Germanium  will  give  a  light  yellow  iodide  film,  but  none  has 
been  on  hand  to  experiment  with. 

Tin  gives  a  slightly  volatile  coating,  showing  a  trace  of  brown 
when  hot.  Potassium  thiocyanate,  if  dropped  on  the  oxide  and 
strongly  heated  gives  a  pale  yellowish  green,  infusible  (compare 
lead) .  The  slight  volatility  of  tin  oxide  suggests  a  scale  of  vol- 
atility, of  great  use  in  describing  the  formation  of  the  films  on 
the  tablets.  The  scale  runs  in  the  order  of  increasing  volatility  : 
tin,  zinc,  cadmium,  and  mercury.  Anything  less  volatile 
than  tin  might  be  classed  as  non-volatile. 

The  iodine  solution  yields  a  yellow,  reddish  brown  when  hot, 


METHOD  IN   BLOWPIPE  ANALYSIS.  859 

the  brown  fading  instantly.  Potassium  sulphide  yields  a  black 
with  a  brown  edge,  which  darkens  on  heating.  Potassium  cyanide 
discharges  the  color,  which  turns  black  on  heating,  and  when 
strongly  heated  shows  the  pale  yellowish  green  (stannous  thiocy- 
anate,  Sn(SCN),;  compare  lead,  bismuth,  arsenic,  mercury  and 
zinc).   Water  decomposes  the  film  with  formation  of  oxy-iodide. 

Cobalt  nitrate  gives  the  bluish  g^een,  which  is  not  so  readily 
attacked  by  nitric  acid  as  the  zinc  green. 

Antimony  tri-  or  pentachloride  yields  with  all  tin  salts  a  fine 
purplish  blue-black  coating,  stable  in  the  presence  of  acids. 
Potassium  thiocyanate  decomposes  it  when  heated  and  forms 
the  pale  green. 

These  tests  with  iodine,  antimony  trichloride,  and  with 
potassium  thiocyanate  remove  tin  from  the  list  o^metals  determin- 
able with  difficulty  before  the  blowpipe.  They  can  be  depended 
on  through  a  wide  range  of  mixtures. 

Lead  yields  per  se  a  white  and  yellow  ;  reddish  brown  when 
hot.  All  lead  salts  fuse  into  the  tablet  with  the  formation  of 
lead  plumbate,  one  of  the  constituents  of  glass.  Potassium  sul- 
phide produces  a  brownish  black,  with  reddish  brown  ring. 

The  iodine  solution  gives  a  film  which  is  chrome  yellow,  with 
a  band  of  fainter  yellow  farther  away  (oxy-iodide?),  and  the 
assay  is  black.  Potassium  sulphide  yields  a  spot  with  the 
reddish  brown  edge.  Hydrochloric  acid  destroys  the  edge  at 
once.  Nitric  acid  wipes  the  spot  off  slowly,  and  sulphuric  acid 
destroys  the  black  and  restores  the  yellow.  The  very  volatile 
paler  yellow  on  the  outer  edges  is  turned  to  a  brighter 
color  by  the  same  treatment  (compare  mercury).  Potassium 
cyanide  produces  a  slight  paleing  of  the  sulphide  color.  Potas- 
sium thiocyanate  on  the  iodide  film  gives  a  black  ring,  which 
heated  becomes  a  black  spot  (compare  bismuth).  Water  wipes 
off  the  coating   (compare  mercury,  arsenic  and  silver). 

The  bromide  film  made  by  using  potassium  bromide  and 
potassium  hydrogen  sulphate  presents  some  interesting  difiFer- 
ences.  It  is  white  with  a  trace  of  yellow,  the  yellow  fusing  into 
the  tablet.  Potassium  sulphide  gives  a  spot,  greenish  for  a  moment 
and  then  black,  on  which  potassium  cyanide  and  potassium 
thiocyanate  have  no  effect,  but  is  partly  destroyed  by  hydrochloric 


86o  W.    W.    ANDREWS.      PLASTER  OF   PARIS 

acid,  niorerapidly  by  nitric  acid,  and  completely  by  sulphuric  acid. 
Potassium  thiocyanate,  placed  on  the  sulphide  and  heated,  gives  a 
black  ring  ;  with  greater  heat,  a  yellow,  and  still  greater  heat,  a 
greenish  gray  ring.  Potassium  cyanide  on  the  iodide  film  has  no 
effect  till  heated  ;  then  a  white.  Potassium  thiocyanate  has  no 
effect  on  iodide  till  heated  ;  then  a  yellowish  spot  appears  (com- 
pare tin).  The  sulphide  heated  becomes  grayish  black,  on 
which  nitric  acid  and  the  other  acids  have  no  effect  (compare 
copper) . 

It  is  a  good  illustration  of  Camelley'slaw of  colorthatin general 
the  bromide  film  of  any  metal  resembles  the  iodide  film  of  an 
element  either  in  a  higher  series  in  its  own  family  or  in  the  same 
series,  in  another  family  toward  the  left  in  the  natural  classifi- 
cation. Thus  the  bismuth  bromide  film  resembles  the  iodide 
film  of  antimony  and  lead.  Lead  bromide  resembles  tin  and 
thallium  iodide.  Thallium  resembles  mercury,  and  meroury 
resembles  silver  in  the  same  way. 

Nitrogen  with  metaphosphoric  acid  in  the  nitrates  yields  an 
effervescence  with  the  fumes,  odor  and  reactions  of  nitrogen 
tetroxide  farther  up  the  tablet,  and  in  the  cyanides,  the  odor  of 
hydrocyanic  acid.  Nitrates  with  carbonaceous  matter  j'ield 
ammonia,  which  will  cause  white  fumes  to  rise  from  a  spot  on 
the  tablet  moistened  with  hydrochloric  acid.  Ross  reports  that 
any  nitrogen  compound  with  boric  oxide  yields  a  tough  trans- 
parent bead,  and  with  metaphosphoric  acid,  purple  in  the  redu- 
cing fiame  with  manganese  dioxide. 

Vanadium  gives  with  metaphosphoric  acid  a  pale  yellow  in  the 
oxidizing  fiame,  and  in  the  reducing  flame  a  green.     (Ross). 

Phosphorus.  A  great  desideratum  in  blowpipe  analysis  is  a 
good  test  for  this  element  and  the  phosphates. 

Arsenic  yields  per  se  a  brownish  black  w^ith  a  white  film  falling 
farther  away  with  odor  of  garlic  and  blue  flame  (compare 
thallium  and  tellurium) .  The  iodide  coat  is  white  and  pale  yel- 
low; the  assay  wholly  volatile.  Potassium'  sulphide,  with  a 
drop  of  hydrochloric  acid,  forms  the  yellow  sulphide,  little 
affected  by  acids.  If  oxalic  acid  be  applied  to  a  sulphide  spot  and 
then  hydrochloric  acid,  no  effect  is  noticeable.  The  yellow  will 
show  up  still  better  next  day  (compare  antimony).     If  a  drop 


METHOD   IN   BLOWPIPE   ANALYSIS.  86 1 

of  potassium  thiocyanate  be  placed  on  the  tablet  about  one  inch 
above  the  assay,  and  between  them  a  drop  of  nitric  acid  and  the 
arsenical  vapor  be  blown  over  them  from  the  assay,  there  will 
generally  be  formed  in  the  edge  of  the  potassium  thiocyanate 
spot  a  bright  bluish  green  of  unknown  composition.  All  com- 
mon acids  except  acetic  destroy  it.  It  shows  well  in  the  pres- 
ence of  salts  of  tin  and  antimony.  When  it  does  appear  it  is 
decisive  for  arsenic.  This  iodide  film  exhibits  a  very  marked 
repulsive  power  for  water,  probably  due  to  the  arsenic  oxide 
which  forms  with  it.  Potassium  iodide  with  metaphosphoric 
acid  yields  more  of  the  yellow  than  does  the  iodine  solution. 

Antimony  per  se  yields  a  white  and  yellow  band  and  white 
fumes.  Potassium  sulphide  yields  on  this  an  orange  brown, 
which  is  quickly  destroyed  by  a  drop  of  nitric  acid. 

The  iodide  film  is  a  fine  orange  yellow  far  away  with  yellow 
nearer  the  assay  and  abundant  white  fumes.  Potassium  sul- 
phide yields,  especially  when  heated,  an  orange  red  with  a  rich 
brown  and  then  a  black  beyond  the  spot.  Hydrochloric  acid 
slightly  heated  destroys  it ;  nitric  acid  destroys  it  instantly  ;  so 
also  does  its  vapor.  Potassium  thiocyanate  wipes  the  coating, 
but  heated  it  yields  a  fine  brown,  which  is  permanent  when 
exposed  for  months.  Potassium  cyanide  wipes  the  coat.  The 
orange  yellow  sulphide  spot,  produced  on  the  iodide  film  ob- 
tained with  potassium  iodide  and  metaphosphoric  acid,  is  not 
so  susceptible  to  the  action  of  nitric  acid  and  is  more  rapidly 
destroyed  by  hydrochloric  acid  than  the  one  described  above. 

If  arsenic  be  present  with  antimony,  there  will  be  shown 
inside  the  yellowish  orange  of  the  iodide  film,  a  fine  peachy  pink, 
which  is  hard  to  wet.  Stannic  chloride  yields  with  antimony  in 
most  combinations  a  purplish  blue-black,  which  is  remarkably 
stable  (Haanel).  It  is  now  being  collected  in  quantity,  with  a 
view  to  the  determination  of  its  formula.  It  will  be  seen  that 
with  the  blue  with  potassium  thiocyanate,  the  rose  pink,  and  the 
reactions  of  the  sulphide  with  hydrochloric,  nitric  and  oxalic 
acids,  the  presence  of  arsenic  can  be  easily  demonstrated  in  the 
presence  of  antimony  and,  as  far  as  experiment  has  gone,  in  the 
presence  of  any  other  substances. 

Bismuth  yields /^r  se  a  yellow  ring  near  the  assay  and  often  a 


862  W.    W.    ANDRBWS.      PLASTER  OF  PARIS 

brittle  globule.  Potassium  sulphide  gives  on  the  white  oxide  a 
brownish  black  which  nitric  acid  destroys  and  on  which  hydro- 
chloric acid  has  little  effect  till  heated,  when  it  removes  it  com- 
pletely. Sulphuric  acid  has  no  effect.  Potassium  thiocyanate 
on  the  oxide  produces  a  yellow  ring,  and  heated  a  yellow  spot 
turning  black.  (It  is  to  be  boted  that  potassium  thiocyanate 
itself  when  heated  or  treated  with  strong  acids,  shows  on  the 
tablet  a  fine  yellow,  which  further  heating  renders  colorless.) 

The  iodide  film  is  a  splendid  combination  of  chocolate  black, 
crimson  and  yellow,  the  assay  turning  black.  Potassium  sul- 
phide forms  a  chocolate  black,  soluble  in  nitric  acid  and  not 
effected  by  sulphuric  acid.  The  latter  acid  on  the  iodide  film 
produces  a  black  and  a  dull  red  edge.  This  is  probably  the  sul- 
phide formed  by  the  reduction  of  the  acid  by  the  decomposition 
products  of  the  potassium  thiocyanate,  which  fall  with  the 
iodides.  It  has  been  noticed,  however,  to  happen  with  no  other 
metal  than  bismuth.  This  reaction  is  very  useful  in  detecting 
small  quantities  of  bismuth  in  the  presence  of  other  metals  giv- 
ing dark  colored  films  (compare  tellurium).  Potassium  thio- 
cyanate on  the  iodide  wipes  it  off,  forming  a  yellow  ring,  but 
when  heated  it  forms  a  black  spot  with  a  brown  ring.  Potassium 
cyanide  also  wipes  the  iodide,  but  when  heated  forms  a  dark  g^y 
spot.  Glacial  acetic  acid  wipes  off  the  yellow  and  the  crimson, 
but  has  no  effect  on  the  chocolate  iodide. 

Sulphur.  In  looking  for  a  better  test  for  sulphur  than  the 
ordinary  one  with  soda  and  a  piece  of  silver,  the  stability  at  high 
temperatures  and  the  two  brilliant  and  characteristic  colors  of 
cadmium  sulphide  attracted  attention,  and  the  fact  that  it  is 
easily  formed  in  the  presence  of  potassium  cyanide.  To  a  solu- 
tion of  cadmium  bromide,  potassium  cyanide  was  added  till  pre- 
cipitation took  place  and  then  the  solution  of  the  precipitate  as 
potassium  cadmium  cyanide.  This,  dropped  on  a  fragment  of  the 
sulphide  and  heated,  will  show  on  the  tablet  near  the  assay  a  bril- 
liant scarlet  when  hot,  and  bright  yellow  when  cold.  This  is  not 
affected  by  potassium  cyanide  (compare  cadmium  oxide).  One 
great  advantage  of  this  is  that  selenium  and  tellurium  do  not 
yield  anything  which  can  be  confounded  with  these  colors, 
selenium  giving  a  grayish  brown  and   tellurium  a  yellowish 


MBTHOD  IN  BLOWPIPE  ANALYSIS.  863 

brown.  Sulphates  may  be  reduced  by  potassium  cyanide,  orby 
glycerol.  A  sulphide  or  sulphate  fused  with  potassium  cyanide 
will,  if  touched  with  a  drop  of  ferric  chloride,  show  in  the  tablet 
the  pinkish  red  of  ferric  thiocyanate.  The  sulphur  in  the  tab- 
let causes  no  trouble. 

Selenium  and  tellurium  are  further  differentiated  from  sulphur 
by  their  characteristic  films,  which  are  tests  of  great  delicacy. 
Twenty-seven  varieties  of  complex  sulphides,  such  as  boumonite, 
tetrahedrite,  stannite,  etc.,  and  all  of  the  common  sulphides  and 
sulphates,  were  found  to  respond  to  this  test  at  once. 

Selenium  yields/^  se  with  characteristic  odor  and-flame  a  fine 
reddish  brown,  almost  pure  red  on  the  outer  edges  and  black  on 
the  inner  edges  near  the  assay.  Potassium  cyanide  wipes  it  off, 
while  potassium  thiocyanate  has  no  effect,  except  that,  if  it  be 
heated,  a  very  stable  red  compound  is  formed  (KSeCN  ?). 

The  iodide  film  forms  in  color  very  similar  to  the  per  se  coat, 
bat  more  volatile.  Potassium  sulphide  yields  a  yellow.  Potas- 
sium cyanide  wipes  the  iodide  film  off  instantly,  and  therefore 
will  reveal  the  presence  of  any  other  element  not  so  affected, 
whose  film  might  be  hidden  by  the  pronounced  hues  of  the  sele- 
nium film.  Potassium  thiocyanate  has  no  effect,  while  it  and 
heat  wipe  off  most  other  coatings,  and  therefore  will  reveal  the 
presence  of  selenium  in  obscuring  associations,  such  as  lead. 
Sulphuric  acid  shows  a  slight  tendency  to  make  this  coating 
darker  (compare  bismuth). 

Tellurium  gives  ^^r  5^  with  flame  and  odor  a  brownish  black 
with  a  white  film  falling  nearer  the  assay  (compare  arsenic). 
Sulphuric  acid-,  if  gently  heated,  shows  an  effervescent  pink  of 
tellurium  sulphate.  Acetic  acid  wipes  off  this  coat  (compare 
cadmium).  So  do  the  potassium  cyanide  and  ammonia  fumes. 
The  iodide  film  is  brownish  and  purplish  black,  less  brown  than 
ih^perse  coat.  Potassium  cyanide  wipes  it  off  in  the  cold. 
Potassium  thiocyanate  has  no  effect  on  the  purple  (compare 
thallium),  and  slightly  dissolves  the  brown,  and  if  nitric  acid 
be  added  a  yellow  appears.  Potassium  sulphide  darkens  the 
coating  a  little.     Sulphuric  acid  acts  as  on  per  se  film. 

Chromium  yields  an  assay  which  is  dark  green  when  hot  and 


864  W.   W.   ANDREWS.      PLASTER   OF   PARIS 

a  fine  g^een  on  cooling.     This  test  can  be  made  very  delicate. 
Metaphosphoric  acid  gives  similar  colors. 

Molybdenum  yields /^r  se,  and  especially  by  flaming »  an  ultra- 
marine coating.  The  oxide  film,  which  forms  when  the  iodine 
solution  is  used,  comes  better  by  flaming  of  the  film  and  in  pres- 
ence of  vapors  of  sulphuric  acid.  A  potassium  thiocyanatespot, 
over  which  the  vapors  from  the  assay  have  swept,  exhibits  a 
splendid  hyacinthine  pink.  Metaphosphoric  and  sulphuric  acid 
vapors  aid  its  formation.  It  is  probably  molybdenum  thiocya- 
nate  (Mo(SCN)J.  If  potassium  thiocyanate  be  added  to  the 
assay  this  color  will  spread  all  around  the  edges  of  the  blue, 
extending  to  a  distance  of  two  inches  from  the  assay.  This  very 
delicate  reaction  is  of  special  interest,  from  the  fact  that  it  shows 
that  part  of  the  potassium  thiocyanate,  or  at  least  the  radical 
thiocyanogen  travels  undecomposed  that  distance  over  the  tablet 
and  that  all  these  films  are  formed  in  the  presence  of  moist  potas- 
sium thiocyanate  or  thiocyanogen  vapors,  which  will  account  for 
the  behavior  of  some  of  the  films.  This  pink  is  decolorized  by 
ammonia,  not  restored  by  nitric  acid.  Sulphuric  acid  dropped 
on  the  tablet  will  form  a  blue  ring  (MoSOJ.  Metaphosphoric 
acid  yields  blue  or  bluish  green  glasses  according  to  the  degree 
of  saturation  (Ross). 

Tungsten  and  uranium  in  metaphosphoric  in  the  reducing 
flame  yield,  the  former  a  blue  and  the  latter  a  green  glass  ( Ross) . 

Fluorine.  If  a  fluoride  be  mixed  with  phosphoric  acid  and  a 
piece  of  glass  be  laid  on  the  tablet  about  two  cm.  away  from  the 
assay,  a  fine  etched  semicircle  will  show  itself  after  the  heating 
of  the  assay.  The  radius  of  the  semicircle  is  about  three  cm. 
long. 

Manganese  yields  with  metaphosphoric  acid  a  glass,  which 
is  \dolet  hot  and  cold,  colorless  in  the  reducing  flame,  and 
turning  green  on  the  addition  of  an  excess  of  soda  (Chapman, 

Ross). 

Chlorine .  Chlorides,  bromides  and  iodides  of  the  alkali  metals 
yield  per  se  white  coatings,  which  may  be  distinguished  from 
other  white  coatings  by  their  flames  and  by  the  action  of  a  small 
quantity  of  the  coating  scraped  together  and  mixed  with  the 


METHOD   IN   BI.OWPIPE  ANALYSIS.  865 

metaphosphoric  acid  cobalt  glass,  which  will  remain  blue  on 
cooling. 

A  compound  of  chlorine  if  mixed  with  metaphosphoric  acid 
and  heated,  in  the  reducing  flame  (if  oxy  salt),  will  cause  white 
fumes  to  rise  from  a  spot  moistened  with  ammonia  situated 
about  two  cm.  above  the  assay.  If  a  copper  salt  be  present  in 
the  glass  or  near  it,  so  that  copper  chloride  vapors  are  formed 
and  these  are  allowed  to  sweep  over  a  spot  of  nitric  acid  and  then 
over  one  of  potassium  thiocyanate,  near  the  assay  a  yellowish 
brown  coating  of  cupric  chloride  will  form  with  an  azure  blue 
flame,  and  beyond  the  potassium  thiocyanate  spot  a  fine  blue- 
black,  very  volatile  (see  copper). 

A  bromide  with  potassium  cyanide  added  to  it  and  the  fused 
mass  laid  upon  a  copper  glass  and  a  drop  of  nitric  acid  added,  a 
fine  red  will  show  itself.  Bromides  with  metaphosphoric  acid 
saturated  with  copper,  upon  blowing,  yield  a  fine  and  very  vola- 
tile reddish  violet  coating.  If  a  bismuth  salt  be  exposed  to  the 
hot  vapors,  it  will  yield  a  yellow  coating.  The  spot  on  the 
tablet  moistened  with  starch  paste,  not  too  near  the  assay,  will 
turn  yellow. 

Similarly  treated  iodine  compounds  yield  violet  vapors,  a  vio- 
let in  the  glass  appearing  with  effervescence,  and  with  copper 
salt  they  yield  a. white  coating,  with  bismuth  scarlet  and  choco- 
late, and  with  starch  a  bluish  black. 

Iron  gives  an  iodide  film  too  delicate  in  color  to  show  up  well, 
either  on  the  white  or  the  black  surface.  Its  presence  can  be 
shown  by  a  red  coloration  after  blowing  hydrochloric  acid  vapors 
over  the  tablet,  to  turn  all  ferrous  compounds  into  ferric,  and 
then  adding  a  drop  of  potassium  thiocyanate  to  the  coating.  It 
is  difficult  to  obtain  plaster  of  Paris  sufficiently  pure  not  to  give 
this  reaction  for  iron.  Such  reaction  can,  however,  be  readily 
distinguished  from  that  given  by  an  assay.  Metaphosphoric 
acid  gives  a  luminous  yellow  when  hot,  which  is  perfectly  color- 
less when  cold.  A  drop  of  acid  on  this  to  produce  ferric  com- 
pounds, followed  by  a  drop  of  potassium  thiocyanate,  will  show 
the  red  of  ferric  thiocyanate,  which  is  decolorized  by  phosphoric 
add,  but  not  by  hydrochloric  acid.  Made  in  this  way  this  test 
is  not  too  delicate  to  show  the  iron  of  composition.     An  assay  of 


866  W.   W.   ANDREWS.      PLASTER  OF   PARIS 

iron  treated  with  a  drop  of  sulphuric  acid  and  heated  will  show 
on  the  tablet  a  film  of  Venetian  red. 

Cobalt  yields  a  glass  blue  hot,  and  violet  cold  ;  permanently 
blue  if  alkali  be  present.  Boron  trioxide  acts  similarly.  With 
the  iodine  solution  a  spot  around  the  assay  turns  pink,  then  deep 
blue  on  heating,  and  then  black. 

Nickel  with  boron  trioxide  separates  as  green  fragments, 
which  may  be  gathered  by  solution  of  the  glass  in  water,  and 
then  the  separated  nickel  (as  any  nickel  compound)  will  yield 
in  metaphosphoric  acid,  a  reddish  brown  when  hot  and  amber 
yellow  when  cold  (Ross). 

Palladium  gives  a  dull  blue-black  film  with  the  iodine  solu- 
tion, which  is  very  characteristic.     The  assay  turns  dull  black. 

Osmium  yields  ^^r  se  a  greenish  black.  The  iodide  film  is  a 
combination  of  olive  green,  dove  and  slate  colors,  with  red 
appearing  around  the  lower  edges.  The  edge  of  the  coating 
nearest  the  assay  shows  greenish  brown  and  the  assay  itself  will 
be  closely  surrounded  with  an  iridescent  black  film.  Potassium 
sulphide  turns  the  coating  somewhat  darker,  which  heated, 
becomes  a  brownish  film,  which  is  wiped  off  by  hydrochloric  and 
nitric  acids  and  not  affected  by  sulphuric  acid  and  potassium 
cyanide.  On  the  iodide  films  sulphuric  acid  has  no  effect; 
potassium  thiocyanate  has  none  till  heated  and  then  it  turns 
brown.  Hydrochloric  and  nitric  acids  remove  the  film.  Potas- 
sium thiocyanate  dropped  on  the  tablet  over  an  inch  from  the 
assay  before  the  coating  is  deposited,  will,  when  the  vapors 
sweep  over  it,  turn  to  a  fine  brick  red,  destroyed  by  potassium 
cyanide  and  the  acids. 

Potassium  bromide  and  potassium  hydrogen  sulphate  give  a 
pinkish  brown  (compare  copper) .  Potassium  sulphide  produces 
a  gray  not  affected,  which  turns  darker  on  being  heated, 
destroyed  by  acids,  and  not  affected  by  potassium  cyanide. 

Indium  yields  with  the  iodide  solution  an  indistinct  brownish 
yellow  coating  and  a  potassium  thiocyanate  spot  which  in  tint 
resembles  the  molybdenum  spot,  but  it  is  covered  with  dots  of 
darker  pink. 

Platinum  gives  an  infusible  gray  film.  Ruthenium  and  rho- 
dium are  being  investigated. 


METHOD   IN   BLOWPIPE  ANALYSIS.  867 

All  these  reactions  have  been  obtained  from  a  large  number 
of  the  compounds  of  each  element  except  in  the  cases  of  osmium, 
indium  and  iridium.  The  writer  will  be  glad  to  hear  of  any 
cases  in  which  they  fail  and  to  receive  specimens  of  combina- 
tions which  cannot  be  unlocked  by  this  method.  One  gram 
weight  of  any  allby  is  sufficient.  The  next  work  to  be  under- 
taken is  to  exhaustively  determine  the  lowest  percentage  of 
any  metal  which  can  be  determined  with  certainty  in  the  pres- 
ence of  one,  two,  or  any  number  of  other  metals,  to  describe 
the  characteristic  effect  that  one  metal  has  on  the  coating 
yielded  by  another  when  they  are  deposited  together  and  to 
determine  the  value  of  each  metal  as  an  interfering  element. 

COVERED  TABLETS. 

The  tablets  are  easily  cut  with  a  knife  and  therefore  they  can 
be  used  in  various  ways.  Open  tube  work  can  be  performed  on 
a  tablet,  if  a  groove  be  cut  lengthwise  of  a  tablet  and  laid  upon 
another,  groove  down.  A  small  pit  for  the  assay  is  cut  in  the 
lower  one  about  one  centimeter  from  the  end.  The  groove  is 
cut  so  that  its  narrowest  part  is  just  above  the  assay  pit,  and  from 
that  point  to  the  lower  end  it  flares  into  a  half  funnel  form  and 
into  this  the  flame  is  blown.  By  regulating  the  size  of  the 
groove  at  its  narrowest  part  the  amount  of  air  which  will  flow 
over  the  assay  may  be  regulated.  This  method  is  of  great  use 
when  very  small  quantities  of  precipitates  are  to  be  tested.  For 
instance,  five-tenths  mg.  of  arsenious  oxide  gave  in  one  experi- 
ment a  narrow  coating  one-half  inch  long  on  each  tablet.  This 
gives  ample  opportunity  for  making  confirmatory  tests.  Various 
reagents  may  be  placed  along  the  groove  to  be  acted  on  by  the 
vapors,  gold  leaf  for  mercury,  potassium  cadmium  cyanide  and  lead 
acetate  for  hydrogen  sulphide  fumes,  starch,  bismuth  and  anti- 
mony solutions  for  iodine,  copper  sulphate  for  chlorine,  etc. 

If  a  coating  be  made,  or  a  small  piece  of  volatile  salt  be 
placed  in  a  small  pit  in  the  tablet  and  a  thin  tablet  be  placed 
over  it,  it  is  found  that  if  potassium  sulphide,  or  potassium  thio- 
cyanate  be  dropped  on  the  upper  tablet  and  the  flame  be  directed 
upon  the  drop,  they  will  pass  through  the  tablet  and  reactions 
will  take  place  away  from  the  air.  After  a  few  seconds  blowing 
the  upper  tablet  will  be  found  to  be  floating  on  a  layer  of  hot 


868        PLASTER  OF  PARIS  METHOD  IN  BLOWPIPE  ANALYSIS. 

gas,  which  flows  between  the  two  smooth  surfaces.  Tin  and 
arsenic,  and  other  substances  easily  oxidizing  in  the  air,  form 
their  sulphides  very  readily  under  these  conditions.  Potassium 
thiocyanate  forms  sulphides.  It  is  in  this  way  possible,  by- 
using  ammonium  hydroxide  or  hydrochloric  acid,  to  form  the 
sulphides  in  the  presence  of  moist  acid  or  alkaline  vapors. 

Other  methods  of  using  the  tablets  will  be  described  later. 

In  teaching  research  methods,  the  plasterof  Paris  method  is  one 
of  the  finest  instruments  to  use  with  beginners.  In  the  course 
of  an  hour  a  student  will  have  been  able  to  make  from  twenty  to 
forty  different  tests  and  without  any  delay  in  preparing  solu- 
tions, or  in  waiting  for  filtration  to  take  place,  he  will  have  pro- 
duced the  oxide,  sulphide,  chloride,  bromide,  and  iodide  of  a 
given  metal,  and  will  have  noted  their  colors,  manner  of  deposi- 
tion, volatility,  solubility  in  several  reagents,  and  the  behavior 
of  the  assay  itself  at  high  temperatures  and  will  have  ransacked 
his  vocabulary  to  find  terms  to  describe  the  phenomena  in  his 
written  notes.  His  skill  in  manipulation  and  his  powers  of 
observation  are  kept  in  liveliest  exercise  and  his  independence 
developed,  for  it  is  quite  possible  to  give  each  student  in  a  large 
class  his  own  problem.  In  no  other  laboratory  work  do  the 
compelled  acts  of  judgment  follow  each  other  asrapidly.  There 
are  many  problems  which  may  be  set  requiring  reference  to 
standard  chemical  literature,  and  many  simple  and  some  very 
diflBcult  equations  of  reactions  to  be  written. 

Not  the  least  valuable  consideration  from  an  educational 
standpoint,  is  the  aesthetic  quality  of  the  work.  All  the  coat- 
ings are  symmetrical  in  form  and  beautiful  in  shading,  and 
many  of  them  in  brilliancy  of  hue  and  in  delicacy  of  shading, 
rival  the  most  splendid  colors  of  flowers.  This  gives  added 
interest  to  the  work  and  is  of  great  value  since  adult  students  are 
so  frequently  found  to  be  greatly  deficient  in  the  color-sense,  as 
children  are  not.  There  has  not  been  opportunity  to  compare 
the  shades  of  these  films  with  the  descriptions  given  in  the 
Standard  Dictionary.  When  this  has  been  done,  exact  training 
can  be  given  in  color  language  also. 

Apology  is  offered  for  publishing  the  results  of  this  research 
at  this  stage,  when  so  many  unsolved  problems  stand  along  its 


HYDROLYSIS  OF  STARCH   BY  ACIDS.  869 

path,  but   this    much  is  given  in  order    that    the    practical 
value  of  these  reactions  and  methods  may  be  put  to  the  test. 

UXIVmSITT  OF  MT.  ALUSOlf  COLLEOB, 
SACKVILLB,  N.  B. 


AN  ANALYTICAL  INVESTIQATION  OF  THE  HYDROLYSIS 

OP  STARCH  BY  ACIDS. 

By  Gbo.  W.  Rolfb  axtd  Gbo.  DBPHBif . 
Received  July  3.  sSp6b 

FEW  problems  of  commercial  analysis  have  been  so  compli- 
cated and  so  discouraging  as  that  of  the  determination  of 
the  components  of  starch  conversion  products.  The  well-known 
schemes  of  commercial  analysis  of  worts  and  similar  products  of 
the  action  of  diastase  are  based  on  the  assumption  that  but  two 
simple  compounds  are  formed  from  the  starch — maltose  and 
dextrin.  In  the  case  of  glucose  syrups  and  starch  sugars, 
which  are  the  results  of  acid  hydrolysis,  it  is  known  that  the 
reaction  proceeds  farther  as  dextrose  is  formed  from  the  maltose 
and  dextrin. 

Musculus  and  Gruber*  decided  that  these  reactions  went  on 
together  so  that  except  at  the  very  beginning  or  final  stage  of 
hydrolysis  all  of  these  compounds  must  be  present  in  solution. 

The  analysis  of  acid-converted  starch  products  must  therefore 
take  into  consideration  the  presence  of  the  third  compound, 


Much  doubt»  however,  has  been  thrown  on  the  accuracy  of 
such  analyses,  as  during  the  past  twenty  years  the  researches  of 
O'SuUivan,  Brown,  Heron,  Morris,  Bondonneau,  Herzfeld,  Mus- 
culus, Bruckner,  Fischer,  and  other  distinguished  investigators, 
have  shown  that  not  only  the  simple  compounds  referred  to  can 
be  isolated  from  starch  products  but  also  many  others  of  quite' 
distinct  optical  and  chemical  properties.  Space  will. not  permit 
a  review  of  this  work,  which  is  in  many  points  conflicting.  The 
recent  conclusion  of  Lintner  and  Diill  is  that  the  following  com- 
pounds result  from  hydrolysis  :' 

1  Bull.  Soc.  Chim.,  s,  30. 

S  Ser.  d.  chtm,  Ges.,  si,  i52»-i53i. 


870  GBO.   W.   ROLPE  AND  GBO.   DSPRBN. 

Hydrolysis  with  oulic  scid.  With  disstsse. 

Amylodextrin  Amylodeztrin 

Erythrodeztrin  I  Brythrodeztrin  I 

lla  

Up  

Achroodeztrin  I  Achroodeztrin  I 

II  "  II 

Isomaltose  Isomaltose 

Deztrose  Maltose 

Others,  as  Brown  and  Morris,'  deny  the  existence  of  the  iso- 
maltose of  Fischer  and  Lintner  and  Diill,  and  mention  another 
compound,  maltodextrin,  an  intermediate  between  dextrin  and 
maltose. 

In  1885  Brown  and  Morris'  discovered  the  remarkable  laif7 
that  at  any  stage  of  the  conversion  of  starch  by  diastase,  the 
total  product,  in  its  optical  properties  and  relation  to  Fehling^ 
solution,  behaved  exactly  as  if  made  up  of  two  components  only, 
maltose  and  dextrin,  so  that  it  was  possible  by  taking  the  rota- 
tory power  to  calculate  at  once  the  cupric  reducing  power  if  the 
total  carbohydrates  were  known.  This  law  indicated  that,  how- 
ever complicated  the  bodies  isolated,  they  could  be  considered 
as  existing  in  solution  as  two  simple  compounds,  and  did  much 
to  establish  the  validity  of  the  principles  of  the  usual  commer- 
cial analyses  of  beer-worts  and  similar  products. 

The  method  of  analysis  of  glucose  syrups  and  starch  sugars 
implies  the  assumption  of  a  similar  law,  but  the  proof  that  this 
law  actually  exists  under  varying  conditions  of  hydrolysis  appa- 
rently has  not  been  worked  out.* 

Our  investigations  have  been  made,  first,  to  determine 
whether  there  was  any  simple  constant  relation  between  the 
optical  rotation  and  the  cupric  reducing  powers  of  starch  pro- 
ducts hydrolyzed  under  different  conditions ;  and,  secondly, 
whether  any  laws  could  be  found  affecting  the  three  simple 
bodies  assumed  to  be  formed  and  determined  by  the  usual  methods 
of  analysis. 

Incidentally  we  have  collected  some  data  as  to  the  speed  of 
hydrolysis,  influence  of  carbohydrates  on  specific  gravity  of 

I  J.  Chem,  Soc.,  No.  393.  Aug.,  1893. 
*Ann.  Chem.  (Uebigr).  331,  131. 

•  A  very  complete  bibliography  of  the  original  pnblications  on  Uie  carbohydntet  is 
in  rollem*s  Handbuch  der  Kohlenhydrate^  Vol.  x,  i88S^  331-3^;  Vol.  XI,  lifs^  36^-396. 


HYDROLYSIS  OP  STAKCH  BY  ACIDS.  S71 

solutions,  and  some  looking  to  the  adoption  of  a  more  rapid  and 
accurate  mf^tliod  of  determining  cupric  reducing  power  by  Feb- 
ling  solution. 

Tbe  latter  data  are  included  in  a  separate  paper.  The  work 
on  specific  gravities  is  not  yet  sufficiently  complete  for  publica- 
tion. 

An  autoclave  of  the  usual  construction  was  modified  in  the 
following  manner  :  The  thermometer  tube  was  taken  out  and  in 
its  place  was  attached  a  specially  constructed  valve,  by  means 
of  which  liquor  cooking  in  a  beaker  in  the  interior  could  be 
removed  at  any  time  during  the  progress  of  the  experiment. 
This  superheated  liquor  was  prevented  from  vaporizing  by  pass- 
ing through  a  condenser.  Excessive  condensation  into  the 
beaker  was  prevented  in  large  part  by  a  well  fitting  lead  cap. 
The  illustration  sufficiently  explains  the  apparatus. 


In  most  of  the  work  about  100  grams  of  a  good  quality  of 
commercial  com  starch'  was  mixed  with  a  liter  of  water  con- 

I  AB  aMl j^*  of  thli  atanb  by  tbc  anal  conmercUl  nctbod*  can  : 

SUrcb B^is 

(Ml 0.14 

A*b 0.11 

AlbumlBoid 0^ 

Water ■ "o-" 


872  GEO.   W.   ROLFE   AND   GEO.    DEFREN. 

taining  the  hydrolyzing  acid.  Samples  of  from  fifty  to  sevent5'-five 
cc.  of  the  liquor  were  removed  at  different  stages  of  the  conver- 
sion and  immediately  shaken  up  with  a  few  grams  of  marble 
dust.  Two  drops  of  tenth  normal  sodium  hydroxide  solution 
were  then  added  to  the  sample,  which  was  cooled  and  filtered. 
This  method  of  neutralization,  except  in  cases  of  very  low  con- 
verted samples,  gave  an  absolutely  clear  filtrate,  the  filtration 
being  exceedingly  rapid,  and  the  removal  of  the  albuminoids 
being  practically  complete.  Low-converted  products  often  re- 
quired to  be  heated  with  aluminum  hydroxide  before  filtering. 
The  samples  were  tested  as  follows  : 

( 1 )  For  specific  gravity  by  Westphal  Balance,  corrected  to  a 
temperature  of  15.5®  C. 

(2)  Specific  rotatory  power  (  [«]d)  by  a  Schmidt  and  Haensch 
half-shade  saccharimeter. 

(3)  Cupric  reducing  power  by  means  of  Fehling  solution. 
Total  Solids — Total  solids  were  calculated  from  the  specific 

gravity  of  the  solution  by  the  factor  6.00386,  which  was  taken 
to  represent  the  influence  of  one  gram  of  the  mixed  carbohy- 
drates in  100  cc.  of  solution.  Corrections  were  made  when 
necessary  for  the  influence  of  other  substances  in  solution,  not 
carbohydrates.  This  factor  386  is  practically  that  of  Balling 
and  Brix  and  has  been  found  exact  for  approximately  ten  per 
cent,  solutions  of  cane  sugar,  and  the  balance  of  evidence  seems 
to  be  that  it  is  correct  for  starch  products. 

We  have  made  several  determinations  of  this  factor  by  drying 
ten  cc.  of  solution  on  rolls  of  dried  paper  at  a  temperature  of 
100-105®  C.  Our  results  point  to  the  constancy  of  this  factor 
386  even  in  solutions  of  low  rotatory  power,  but  are  not  yet 
complete  enough  to  establish  the  value  for  all  rotations. 

Therefore,  in  this  work  we  have  adopted  the  expedient  used 
by  Brown  and  Morris,  and  others,  and  calculated  all  optical  and 
copper  reduction  constants  on  the  assumption  that  all  three  car- 
bohydrates in  solution  affect  the  specific  gravity  like  cane  sugar 
when  the  concentration  is  approximately  ten  per  cent.  Even  if 
subsequent  investigations  show  that  this  view  is  not  exactly  cor- 
rect, the  relative  values  of  the  constant  will  not  be  appreciably 
affected  nor  the  truth  of  the  laws  as  set  forth. 


HYDROLYSIS  OF   STARCH   BY   ACIDS.  873 

To  illustrate  this  method  of  calculation  of  constants  we  give 
the  following  from  our  own  determinations : 

Ten  grams  of  dextrose  dissolved  in  100  cc.  of  water  gave  a 
rotation  of  30.70®  on  the  Schmidt  and  Haensch  saccharimeter. 
This  gives  [ajp  as  52.8.*  As  the  increase  in  specific  gravity 
per  gram  of  crystallized  dextrose  in  100  cc.  is  0.00381,  [^jojedis 

53.5. 

9.751  grams  of  crystallized  maltose  anhydride  in  100  cc.  of 
water  gave  a  rotation  of  76.40.  This  gives  an  absolute  specific 
rotatory  power  of  136.6.  The  specific  gravity  factor  of  maltose 
being  0.00390,  [«]d386  is  135.2°.  No  exact  figure  is  known  for 
the  influence  of  crj-stallized  dextrin  on  the  specific  gravity  of 
its  solution.  O^Sullivan  gives  0.00385,  and  the  balance  of  evi- 
dence seems  to  favor  this.     Hence  195  is  probably  correct  for 

In  like  manner  the  values  for  K  have  been  reduced  to  a  dex- 
trose with  the  factor  386. 

Specific  Rotatory  Power, — All  readings  were  made  as  nearly  as 
possible  at  a  temperature  of  20**  C.  in  200  mm.  tubes,  the  mean 
of  several  readings  being  taken.  Corrections  for  zero-error  were 
made  frequently,  and  the  instrument  was  carefully  screened  by 
glass  plates  from  the  heat  of  the.  lamps.  Comparisons  were 
made  with  a  Laurent  polariscope  to  determine  the  value  of  the 
division  in  terms  of  angular  degrees  for  sodium  light,  the  accu- 
racy of  the  quartz  wedges  having  being  verified  previously. 
With  standard  quartz  plates  the  usual  factor  0.346  was  obtained, 
but  solutions  of  commercial  glucose  of  approximately  ten  per 
cent,  gave  the  figure  0.344,  which  agrees  with  the  recent  work 
of  Rimbach'  and  other  investigators.  We  have  taken,  there- 
fore, the  latter  factor  in  our  calculation. 

I  Precautioas  agrainst  bi-rotation  were  taken  in  both  examples  cited, 
s  Brown  and  Heam:  Ann.  Chem.  (Liebig),  X99,  190-243. 
t  Ber.  d.  chem.  Ges.,  97.  aaSz. 


874 


GBO.   W.   ROLPB  AND  GBO.    DBPRBN. 


TABIvB  A. 

Comparison  op  Schmidt  and  Habnsch  Halp-Shadb  Saccharimbtbr 

WITH  THAT  OP  LaURBNT  PoLARISCOPB  RBADING  IN  AUfGXJULK 

Dbgrbbs. 


Test 

(t 

Quartz  A  • 


Reading. 

20—22) 

.   62.965 

•  62.800 

•  62.970 
.   62.836 

Glucose  A-..  77.510 

B...  76.355 
B  . . .  76.355 

c...  76.535 

D-..  76.110 
(t  =  25) 
Hydroliz-")  «  «-  -,i 

products  J  ^'  ^'^ 


8.  and  H.  taccharimeter. 
(Using  batpwing  burner  and  lens.) 

Zero     Corrected 
error. 


Laurent  polariscope. 
(Sodium  flame.) 

Zero       Corrected 
reading.      Reading,     error. 


reading.      Pactor. 


II 


II 


II 


i< 


II 


II 


II 


0.300 
0.150 
0.290 
0.130 
0.277 
0.150 
0.150 
0.150 
0.130 


00 
00 


62.665 
62.650 
62.680 
62.706 

77.233 
76.205 

76.205 

76.385 
75.980 


92.73 
24.84 


21°  40' 
21°  40' 
21°  40.2' 
21°  40.7' 

26°  35' 
26°  15.3^ 

26°  14' 
26°  18' 
26°  10.3' 


3»°  56' 
8°  32' 


o 

o 

0.6' 

C.6' 

o 

o 

.0 

o 

0.6' 

-I' 
—I' 


21.666° 
21.666° 
21.660° 
21.666° 
26.582° 
26.254° 
26.233° 
26.300° 
26.162° 


31.95? 
8.55" 


0.3457 
0.3458 
0.3458 
0.3455 
0.3442 
0.3445 
0.3442 

0.3443 
0.3443 


0.3445 
0.3442 


Cupric  Reducing  Power. — Our  method  is  practically  that  of 
0'Sullivan»  first  published  in  1876.  The  copper  is  weighed  as 
the  oxide.  We  have  found  this  method  exact  and  rapid.  An 
anal3rtical  investigation  of  this  process  has  been  made  by  one  of 
us  and  given  in  detail  in  a  separate  paper. 

Plotted  Results, — ^To  show  the  relationship  of  the  copper-re- 
ducing power,  and  the  specific  rotator}"  power  of  the  products 
formed  during  the  progress  of  the  hydrolysis  of  the  starch,  we 
have  plotted  our  results,  taking  as  abscissae  the  decreasing 
values  of  the  rotatory  power,  from  the  amy lodextrin  stage  (195"*) 
to  that  of  dextrose  ([^]dsk  =  53*5^) i  ^^^  as  ordinates  the  cupric 
reducing  power  (K^  )  taking  that  of  an  equivalent  weight  of 
dextrose  as  100.'    [See  Plate  A.] 

1  Using  Welsbach  burner. 
SData  given  in  Table  B. 


•«t-y 


•98t  °[»] 


*appEO  idddoD 


< 

< 
Q 

< 


s 


X 
u 

o 

M 

09 
O 

u 

Q 

s 


'S-Si  *jjl  'ds 


'39  13)9  AL 


•93  )anoiny 


•pnni< 


*ajns«Md 

3p91|d90linV 


*q9i«)v  fvnuJ 


'SnpiooD 


'dfdiBvt  )0  'OM 


'vnn 


HYDROLYSIS  OF  STARCH   BY  ACIDS.  875 

^  r<>00  On  O  M  ro  ^  10  to^O  <<^0D  00  ^^  On  On  On  O^  •-•  *0^  00  Ot  M  fO^  F^ 

ddodooddddddodddddo'dddo'oodddo 

I/)  r^  fO  Q  tr>\0  ^       O  00  lO  O^^  O  lO  Q 

r^  OnnO  iO«'>OOv)mOm  On^  OsnO  M9^r«.M  M  o  ONt^tor«.^^0  M  M 

«  QNf^O^wvp  'f^woo  ^»*i»n>ooo  f*?  dN^r>.fn  eo^  ON»i^  o^r^r»*oo  vo  6 
^  Osr^i/sONF^trt^  CI  M  o  Onqo  r^^ODNO  lOioior^M  O  t>>Ne^  ^  m  on 

»o  o  I***  nS  *8  «  s  "oS  Soo  ^  ^  ON «  't i*>  Rno  ip  8  3^  »*>  ^5  ONN?  ^ 

MMv«c4000***^MM»ii«MMMV«M(<«c«MO»««"r<nOMMM 

ddddddddddddddddddddddddddddd 


O  (OiOONtOQO  »N>vO00^O 


rN.«^ioONO  O  O  «o«  r^-  »o 


ON  tf)  ^  «0  "OOO  00  NO  ^  lO  **  5h  ^  «*iOO  wn  ^  fO  «*)«*>  O  »>»vO  ^        OvOO  vO  wn 


00  ONrs. 

•       •       • 


•^c8 


to 


y^  ^  I**  ©k  O  00  lO  »^00  0000  ro«*>^r».ON^«^o  ^ -^  O  O  »^  O^  7*  p  ^ 
ON  t«»NO  ONfi^MMMMMMMMMMMMO  O^OO  00  00  r^  t>>00  t>>  t>> 
oooooooooooooooooooooooooooo 


I     I 


«e 


I 


8 


«R 


§>  §N 


8 


e5 


8 


u 


X 


f« 


N 


Oiow^iOiOiCioO  CO  Q  Q  OO  iCO  tcQ  toQtO»/NQ  QQ  m  «nr<  m 


M   M   M   iM   C4   M   M 


«  rO^iTkM  M  f*>^  ION©  «^00  ON  O  «  «  fO  ^  *nyo  ^  «  »o^io«  eo^io 


s 


e*i 


876 


G£0.   W.    ROLFB  AND  GEO.    DBPRBN. 


'9S^X 


•99f  ^[»] 


-appKO  J9<XdoD 


< 

< 

M 


u 
pj 

«< 

O 

CO 

o 
pj 
o 
> 


JSaipvaj  *qd3«s 


'SSi  uJS  'ds 


*oo  ^a)Bii 


■oo^anomy 


< 


•ponas 


'9in«s9Jd 

3)i»qd80U)V 


'qojiv^t  tmuJB 


*J8aiiIoo3 


'dldnxv*  |o  'OK 


'anH 


DO  •*>  fO  fO  •^^0  fOOO  ^  ^  O  ^  On  »r 
^^  OstsiS)  O^r^*Aa*0  woo  •^  w! 


o  d  b  b  6  o  o  o  o  6  6  o  6  b  o  6  o  o  o  o  b  6  d  d  d  d  d  d  d 


CO  <^  r««  M  ^%o  en  u^  ^O  w^  O^vO  l««  O  ^  00  lOvO  O  *^  v)>0  vO  00  M  r^  u^  <« 
t«»vO  10  CO  »-•  9^>0  ro  CI  CsaO  00  l><>0  lO  ^  O  OvOO  t«»00  iOi-«  ChF^I^u^c^^ 


00 


•o 


ooooooooodooooooooooooooooo 


O  O 


£  9^M  o  too  ON'^^  o  loo^M  M  ior<*o  ro^  MOOf«oooOMc«aooN 


$;ss 


vo  10  ^  P«oo  r^  V)  o  oxoo  00  ^o  t^^  ^  w)  ^  O  00  ^  CO  cooo  r^^ 


IN. 


ff>  ^  ^  0^^0  t^  —  00  « 


«e 


R 


10 


«C 


o 

lO 


s> 


o 


S^ 


O 


o 


8 


8 


u 


X  X 


« 


10 


u 

» 


« 


.8 


o 

U 


o 

s 


.8       >8 


O  CO  cO>6  ON«  «  ^»/»t^M  «  fO^U)r«»0\»-i  «  COM  co^cOiO^cSnOS 

MM  M  MMM  MMM 


>0  «^«  cO'^iOM  w  (O^M  «  cO'^iOm  m  cO^V)m  («  ro^mM  t«  com 


10 


00 


Oi 


8 


'9i^Ji 


'98^   *"[»] 


'9p|so  Jdddoo 


anipvstJ  -M30B8 


3 


X 
CI 

0« 


CO 

o 

> 
o 

PC 

o 


-S'Si  -aJB'ds 


-39  aa)Bj|^ 


'39  innon  Y 


•pnra 


-aunMaad 

3U9tldtOIIIW 


'CI9JV9S  B««aJ 


'J8a!?[oo3 
S9)nu{H 


-»1<Iin««  ;o  -ox 


•nnn 


HYDROLYSIS  OP  STARCH   BY  ACIDS. 


877 


W  On  f*:0O  OnvD  OnxO  QQ  O  d\^r*rO«^0 
ro  «"  ro^  10  >0  to  N  OnoO  vO  vO  ci  O  00  I> 


t«Hf0iO'n-*25S'»O 


t^  ^  ONNO    10  I')  VNVd 


iOnO  pi  W  "H  «o  »0 
ICOO  to  MOO  10  U^ 


M  M  H 
«  «  « 


H  M  H 
CI  «  « 


X  H  H  »< 

«   N   W   « 


X  X  M  M 
w  w  «  « 


so  O  CI  M  o\  r>>  r*NO  o  for*»Ovpopi  «  rhnoo  t>.  rcoo  w  ^  ^  ^o  10  »o  f>  ^ 

•   •••■•*■•••••••••••••■••*••   •   •   • 

r>.>o  ao  «o  •*>  «o  o^oo  >o  lo  ^  ^  00  r^NO  10  on>o  rh  ro  »^no  ^  f*5  »ooo  0  ^  fo  »o 


ON  10  O   M  so  VO  ^00  N  00  to  r^vO  00Q*o000Q0^^»O  »Q  M  M   fp  j^-OO 

iAio^rN.r^O  *•  -^loio^o ^vB  r*0  cs  »*>eo«Oforo«  p*  pi  f^oo  on o 00  r^ 


I 


8 


8 


8 


8 


cS 


o 

lO 


o 
»o 


I   1% 


o 


o 


c5 


BZe 


0 

6 

^■4 

0 

rf\ 

^■4 

^■* 

v^  ^^ 

9^ 

u 

U3 

0 

u 

oa 

u 

s 

X 

X 

s 

re 

XX 

X 

NO 


10  "PQ 


PI 


10 
NO 


PI 


10 
NO 


eOM 


^ 


NO  NONO  NO 


0   fONO  r*  ON  ^  ON  »*500  M  »OnO  on  W  »0  ^  «   r»5  ^vO  M  M  CI  M  onoo  pi  on  w 

M  MNHP4  (i^M  MCIM 

MC*)^»OMP«f»5i^u^Mp|fO^»OMWe0^iOMp|fOM«HPirONfO 


PI 


»o 
PI 


NO 

PI 


PI 


^ 


878  GEO.   W.   SOLFE  AND  CEO.   DEFREN. 


HYDROLYSIS  OP  STARCH  BY   ACIDS.  879 

The  results  point  to  the  remarkable  fact  that  the  cupric  reduc- 
ing power  of  the  total  product  bears  a  constant  relation  to  the 
specific  rotatory  power,  even  when  the  starch  is  hydrolyzed  under 
widely  varying  conditions.  Hence,  given  the  one,  the  value  of  the 
other  can  be  calculated.  To  a  rotation  of  about  90'',  the  plotted 
r^ults  outline  with  extraordinary  exactness  the  arc  of  a  circle, 
the  equation  of  which  is 

•^*  +  y*  +  46&r  —  646^  +  1580  =  o, 
which  exactly  intercepts  the  **  zero'*  and  **  hundred**  points  at 
195  and  53.5,  respectively.  The  upper  part  of  the  curve  is  not 
so  well  defined,  the  results  showing  more  discrepancy  at  the 
high  conversion  stages.  This  may  be  due  to  some  decomposi- 
tion and  the  formation  of  **  reversion**  products  as  stated  by 
Wohl,*  Maercker,  Ost,  and  others.  Wohl's  figures  show  the 
maximum  amount  of  dextrose  possible  to  be  92.7  per  cent,  of 
the  theoretical  quantity.  Others  give  ninety-six  to  ninety-seven 
per  cent.,  the  missing  dextrose  being  supposed  to  be  converted 
into  dextrin-like  bodies  identical  with  those  variously  described 
as  •*gallisin,*'  **  isomaltose,'*  etc.  We  have  experimented  but 
little  along  this  line,  having  made  but  one  hydrolysis  with  this 
special  object,  using  y^  hydrochloric  acid  at  four  atmospheres 
pressure,  with  the  following  results  : 

Time  of  cooking.  L^Jd* 

60  minutes  55-24 

90       **  5309 

120       **  53.40 

150       "  54.42 

While  several  of  our  own  results  at  the  low  rotations  show  a 
cupric  reducing  power  of  only  about  ninety-six  per  cent,  of  that 
of  pure  dextrose,  we  do  not  think  that  we  are  justified  in  arriving 
at  any  definite  conclusion  with  the  data  at  hand. 

That  the  solutions  begin  to  color  considerably  at  rotations 
beyond  90*^  is,  moreover,  a  strong  indication  of  such  decomposi- 
tion. On  the  other  hand,  this  accounts  for  much  of  the  dis- 
crepancy of  the  ploi  at  this  part  of  the  curve,  as  it  is  exceedingly 
difficult  to  get  accurate  readings  on  the  saccharimeter  of  these 
highly  colored  solutions.      Obviously,  too,  slight  errors  in  the 

iBer.  d.  ckem.  Gts.,  33,  aioi. 


880  GEO.   W.   SOLPE   AND  GEO.   DEFREN. 

readings  afiect  the  calculations  of  the  rotatory  power  the  most  at 
these  lowest  rotations. 

Quite  as  noteworthy  are  the  curves'  plotted  by  taking  the 
values  of  maltose,  dextrin,  and  dextrose  as  computed  for  every 


PCffCrNT^  or  CAIfBOHYDRATCS 

Plate  B. 
five  degrees  of  rotation  from  the  values  of  /C,  as  given  by  this 
curve. 

In  this  work  we  have  figured  constants  for  solids  estimated 
from  the  specific  gravities  of  solutions  by  the  factor,  386,  and 
calculated  percentages  by  the  well-known  equations : 

5-+OT+rf=I.OO 

^  +  o.6iwi  =  A" 

Where  g-  is  per  cent,  dextrose, 

m  is  per  cent,  maltose, 

and  d  is  per  cent,  dextrin. 

27. 82 

>  Sn  platt  B. 


HYDROLYSIS  OP  STARCH   BY  ACIDS.  88l 

HxamiDing  these  curves  we  see  that  the  dextrin  starting  from 
the  maximum  of  loo  per  cent,  gradually  falls  to  zero  near  the 
rotation  corresponding  to  dextrose,  while  the  maltose  gradually 
rises,  reaches  a  maximum  percentage  of  44.1  at  about  129° 
rotation,  corresponding  to  the  usual  state  of  conversion  of  com- 
mercial glucose,  and  then  falls,  disappearing  at  53.5^.  The 
dextrose,  on  the  contrary,  steadily  mounts  to  100  per  cent.  It 
will  be  noted,  too,  that  at  the  point  of  maximum  maltose  the  dex- 
trin and  dextrose,  as  shown  by  the  intersection  of  the  curves, 
are  present  in  equal  quantity. 

Tests  with  phenylhydrazin  acetate  show  the  presence  of  the 
dextrose  distinctly  at  about  185°,  and  we  had  hoped  to  prove  the 
gradual  rise  of  the  dextrose  percentage  by  means  of  the  dex- 
trosazon.  While  copious  precipitates  of  this  beautiful  com- 
pound were  obtained,  any  attempt  of  ours  to  isolate  it  in  any- 
thing like  quantitative  amounts  proved  a  failure,  even  in  solu- 
tions containing  a  known  amount  of  pure  dextrose.  We  hope 
to  take  this* up  more  fully  in  a  later  investigation. 

We  have  also  calculated  a  table  (Table  C)  from  the  curves 
giving  the  value  of  maltose,  dextrose,  and  dextrin  within  one- 
tenth  per  cent,  for  successive  stages  of  acid  hydrolysis  repre- 
sented by  each  degree  of  rotation  between  195  and  53.5.  This 
table,  calculated  for  the  factor  386,  makes  no  allowance  for  pos- 
sible decomposition  of  high-converted  products. 

TABLE  C. 

Cai^ulatbd  Values  op  Cupric  Reducing  Powers  and  Parts  op  Mal- 
tose, Dextrose  and  Dextrin  per  Unit  op  Carbohydrate  por 
Bach  Degree  op  Rotation  op  a  Normally  Hydro- 

lyzbd  Starch  Solxttion. 


C«]i%... 

-^is.* 

^.1.- 

£^%%%' 

^.•f 

195  > 

0.000 

0.000 

0.000 

I.OOO 

194 

0.01 1 

0.017 

O.OOI 

0.982 

193 

0.022 

0.033 

O.OOI 

0.966 

19a 

0.032 

0.048 

0.002 

0.950 

191 

0.04.1 

0.063 

0.002 

0-935 

190 

0.051 

0.079 

0.003 

0.918 

189 

0.061 

0.094 

0.004 

0.902 

188 

0.071 

O.I  10 

0.005 

0.885 

187 

0.081 

0.123 

0.007 

0.870 

882  GBO.   W.   ROLPB  AND  GBO.   DBFRBN. 


[«]y.M. 

^f' 

^tM- 

^ISC* 

^•••* 

i86 

0.090 

0.135 

0.009 

0.856 

185 

o.zoo 

0.147 

0.0x0 

0.843 

184 

0.109 

0.160 

0.013 

0.827 

183 

0.1 18 

O.I7I 

0.016 

0.813 

183 

0.137 

0.183 

0.019 

0.799 

181 

0.137 

0.193 

0.033 

0.786 

180 

0.146 

o.ao3 

0.035 

0.772 

179 

0.155 

0.3I3 

0.038 

0.760 

178 

0.164 

0.332 

0.031 

0.747 

177 

0.173 

0.231 

0.034 

0.735 

176 

0.183 

0.240 

0.037 

0.723 

175 

0.I9I 

0.250 

0.040 

0.710 

174 

0.199 

0.357 

0.043 

0.700 

173 

0.307 

0.365 

0.047 

0.688 

172 

0.316 

0.373 

0.050 

0.677 

171 

0.334 

0.380 

0.054 

0.666 

170 

0.333 

0.387 

0.058 

0.655 

169 

0.343 

0.394 

0.063 

0.644 

168 

0.351 

0.301 

0.066 

0.633 

167 

0.359 

0.307 

0.071 

0.632 

166 

0.367 

0.314 

0.075 

0.61 1 

16S 

0.37s 

0.330 

0.080 

0.600 

164 

0.383 

0.336 

0.084 

0.590 

163 

0.393 

0.332 

0.089 

0.579 

163 

0.300 

0.338 

0.093 

0.569 

161 

0.308 

0.344 

0.098 

0.558 

z6o 

0.316 

0.349 

0.103 

0.548 

159 

0.334 

0.356 

0.107 

0.537 

158 

0.332 

0.363 

O.III 

0.527 

157 

0.340 

0.369 

O.I  15 

a5i6 

156 

0.348 

0.373 

O.I3I 

0.506 

155 

0.3S6 

0.378 

0.136 

0.496 

154 

0.36s 

0.383 

0.130 

0.487 

153 

0.373 

0.388 

0.135 

0.477 

152 

0.381 

0.392 

O.I4Z 

0.467 

151 

0.389 

0.397 

0.146 

0.457 

150 

0.397 

0.401 

O.I  S3 

0.446 

149 

0.404 

0.405 

0.157 

0.438 

148 

0.413 

0.408 

0.163 

0.429 

147 

0.419 

0.413 

0.164 

0.420 

146 

0.427 

0.415 

0.174 

0.41 1 

US 

0.43s 

0.415 

0.183 

0.403 

144 

0.443 

0.421 

0.186 

0.393 

143 

0.450 

0.423 

0.193 

0.385 

HYDROI«YStS  OP   STARCH   BY   ACIDS.  883 


C«]«.... 

^w 

^m- 

^«at* 

^.•.• 

14a 

0.458 

0.4*5 

0.199 

0.376 

HI 

0.465 

0.427 

0.2Q5 

0.368 

140 

0.473 

0.428 

0.212 

0.360 

139 

0.481 

0.431 

0.217 

0.352 

138 

0.488 

0.432 

0.22  J 

0.344 

137 

0.496 

0.434 

0.231 

0.335 

136 

■ 

0.503 

0.436 

0.237 

0.327 

135 

0.510 

0.437 

0.243 

0.310 

134 

0.517 

0.438 

0.249 

0.313 

133 

0.524 

0.439 

0.256 

0.305 

132 

o.53« 

0.439 

0.263 

0.298 

131 

0.538 

0.440 

0.270 

0.290 

»30 

0.546 

0.440 

0.277 

0.283 

129 

0.553 

0.441 . 

0.284 

0.275 

138 

0.560 

0.441 

0.291 

0.268 

127 

0.567 

0.440 

0.298 

0.262 

Z26 

0.574 

0.440 

0.305 

0.255 

"5 

0.580 

0.439 

0.313 

0.248 

124 

0.588 

0.438 

0.320 

0.242 

123 

0.595 

0.438 

0.327 

0.235 

122 

0.602 

0.437 

0.335 

0.228 

X2I 

0.608 

0.436 

0.343 

0.221 

120 

0.614 

0.435 

0.350 

0.215 

119 

0.621 

0.433 

0.358 

0.209 

XI8 

0.628 

0.431 

0.366 

0.203 

XX7 

0.655 

0.429 

0.374 

.).I97 

X16 

0.642 

0.428 

0.381 

0.191 

"5 

0.649 

0.425 

0.390 

0.185 

114 

0.656 

0.422 

0.398 

0.180 

113 

0.663 

0.420 

0.408 

0.174 

112 

0.669 

0.417 

0.414 

0.169 

III 

0.675 

0.414 

0.423 

0.164 

no 

0.681 

0.408 

0.432 

0.160 

109 

0.687 

0.407 

0.439 

0.154 

108 

0.694 

0.403 

0.448 

0.149 

107 

0.700 

0.400 

0.456 

0.144 

Z06 

0.707 

0.396 

0.465 

0.139 

105 

0.713 

0.392 

0.474 

0.134 

104 

0.719 

0.387 

0.483 

0.130 

103 

0.725 

0.383 

0.492 

0.125 

102 

0.732 

0.379 

0.500 

0.121 

ZOI 

0.738 

0.375 

0.508 

0.117 

100 

0.744 

0.370 

0.518 

0.112 

99 

0.750 

0.366 

0.527 

0.107 

884  Ono.   W.   ROLPB  AND  GBO.   DBPRBN. 


[«]?.... 

^w 

Jlf„^. 

^«M* 

^•M- 

98 

0.757 

0.361 

0.537 

0.102 

97 

0.763 

0.356 

0.546 

0.098 

96 

0.769 

0.350 

0.556 

0.094 

95 

0.775 

0.345 

0.565 

0.090 

94 

0.781 

0.341 

0.574 

0.085 

93 

0.787 

0*336 

0.583 

0.081 

92 

0.793 

0.331 

0.592 

0.077 

91 

0.799 

0.326 

0.601 

0.073 

90 

0.805 

0.320 

0.610 

0.070 

89 

0.810 

0.314 

0.620 

0.066 

88 

0.816 

0.308 

0.629 

0.063 

87 

0.822 

0.302 

0.638 

0.060 

86 

0.828 

0.295 

0.649 

ao56 

85 

0.834 

0.288 

0.658 

0.054 

84 

0.839 

0.282 

0.667 

0.051 

83 

0.844 

0.275 

0.677 

0.048 

82 

0.850 

0.267 

0.688 

0.045 

81 

0.856 

0.259 

0.698 

0.043 

80 

0.862 

0.251 

0.709 

0.040 

79 

0.867 

0.243 

0.719 

0.038 

78 

0.872 

0.234 

0.730 

0.036 

77 

0.878 

0.225 

0.741 

0.034 

76 

0.884 

0.217 

0.751 

0.032 

75 

0.889 

0.208 

0.762 

0.030 

74 

0.895 

0.200 

0.772 

0.028 

73 

0.901 

0.191 

0.783 

0.026 

72 

0.906 

0.182 

0.794 

0.024 

71 

0.91 1 

0.173 

0.805 

0.022 

70 

0.916 

0.163 

0.817 

0.020 

69 

0.921 

0.153 

0.828 

0.019 

68 

0.926 

0.143 

0.839 

0.018 

67 

0.932 

0.134 

0.850 

0.016 

66 

0.937 

0.125 

0.861 

0.014 

65 

0.943 

0.115 

0.872 

0.013 

64 

0.947 

0.105 

0.883 

0.012 

63 

0.952 

0.095 

0.895 

0.010 

62 

0.957 

0.085 

0.906 

0.009 

61 

0.962 

0.075 

0.917 

0.008 

60 

0.967 

0.065 

0.927 

0.008 

59 

0.972 

0.055 

0.938 

0.007 

58 

0.977 

0.045 

0.949 

0.006 

57 

0.982 

0.035 

0.960 

0.005 

56 

0.987 

0.025 

0.971 

0.004 

55 

0.992 

0.015 

0.982 

0.003 

54 

0.997 

0.005 

0.993 

0.002 

53-5 

1.000 

0.000 

1. 000 

0.000 

HYDROLYSIS  OF  STARCH   BY  ACIDS.  885 

It  would  seem  obvious  that  we  are  now  prepared  to  determine 
whether  a  sample  of  glucose  is  a  product  of  one  hydrolysis  ot  is 
a  mixture  of  two  separately  converted  products,  by  comparison 
of  the  actual  analytical  results  with  those  calculated  from  the 
rotatory  power. 

For  testing  this  method  we  have  made  a  few  analyses  of  com- 
mercial glucoses  obtained  in  open  market. 

In  the  manufacture  of  glucose  syrup  all  the  starch  is  not 
hydrolyzed  under  strictly  the  same  conditions,  as  the  factory 
practice  is  to  pump  the  starch  into  the  converter,  which  is  under 
steam  pressure  and  already  contains  the  hydrolyzing  acid.  As 
the  filling  of  a  converter  takes  about  one-third  of  the  total  time 
of  cooking,  it  is  clear  that  there  is  a  radical  difference  in  the 
time  of  hydrolysis  of  different  portions  of  starch.  Nevertheless, 
we  have  found  that  samples  known  to  have  been  made  under 
these  conditions  conform  to  the  laws  of  our  curve,  and  the  evi- 
dence seems  strong  that  those  which  depart  widely  from  these 
conditions  are  mechanical  mixtures. 

The  following  determinations  of  four  samples  of  commercial 
glucose  giving  the  cupric  reducing  power  as  found  and  as  calcu- 
lated for  the  corresponding  rotation  will  illustrate  the  method  : 

Sample.  ^»386.  -^386  (obUincd).  K^  (calculated). 

I.  C.  PopeCo.  (J) 131. f  0.566  0.537 

II.  C.  PopeCo.  (M) 125.4  0-578  0-578 

III.  Rockford  Co 141 -9  0.454  0.457 

IV.  Chicago  Co 137-2  0.505  0.495 

Evidently  II  and  III  are  normally  hydrolyzed.  IV  is  possi- 
bly a  mixture,  while  I  is  undoubtedly  so.  As  this  latter  is  a 
sample  of  jelly  goods  which  in  factory  practice  are  often  made 
by  mixing  two  lots,  our  conclusion  is  strengthened. 

From  the  results  as  a  whole  we  have  concluded  that  the  evidence 
is  strong,  ( i )  that  in  any  homogeneous  acid-converted  starch  prod- 
uct, irrespective  of  the  conditions  of  hydrolysis,  the  specific 
rotatory  power  always  represents  the  same  chemical  composition. 

(2)  That  but  three  simple  carbohydrates,*  possible  in  molec- 
ular aggregates,  exist  in  the  solution  of  a  starch  product  hydro- 
lyzed by  acids. 

^  Leaving:  out  of  consideration  the  possible  sm<(ll  amounts  of  products  formed  by 
ttrcrsion. 


886  GEO.   W.   ROI«PE  AND  GEO.   DEPREN. 

DETERMINATION   OP   THE   CONVERSION    OP   COMMERCIAL  GlrU- 

COSE. 

In  the  manufacture  of  gluco^  it  is  obviously  essential  to  have 
a  rapid  means  of  determining  the  degree  of  conversion  of  the 
starch  during  the  cooking  process.  The  usual  factory  practice 
is  to  control  the  conversion  by  means  of  iodine  color  tests. 
These  tests  are  usually  made  by  adding  a  definite  number  o£ 
drops  of  standard  }odine  solution  to  a  te^t-tube  of  the  cooled 
glucose  liquor.  The  tint  at  which  the  conversion  is  considered 
complete  varies  in  general  practice  from  that  corresponding  to 
[a]D  =  r28  to  [«]d  =  135,  the  variation  being  even  grater  in 
some  cases,  depending  on  the  ideas  of  the  manufacturer  and  the 
grade  of  goods  desired. 

By  daily  practice  workmen  become  quite  expert  in  making 
these  iodine  tints,  which  are  usually  carried  out  by  crude 
methods  and  read  off  without  comparison  with  any  standard. 
Nevertheless,  the  product,  when  examined  by  more  refined 
laboratory  processes,  shows  wide  variations  from  day  to  day, 
which  does  not  appear  surprising  when  we  examine  into  the 
errors  of  such  color  tests. 

Assuming  that  the  test  is  carried  out  under  uniform  condi- 
tions of  concentration  and  proportion  of  reagent  to  liquor  to  be 
tested,  which  is  by  no  means  always  the  case,  the  other  condi- 
tions affecting  the  color  are  (i)  temperature,  (2)  turbidity,  and 
(3)  illumination. 

Uniform  temperature  can  be  obtained  easily  by  some  simple 
cooling  device  as  a  stream  of  running  water. 

The  acid  converter  liquors  are  always  turbid  when  tested,  as 
filtration  in  this  rapid  testing  is  impracticable.  The  turbidity^ 
however,  is  fairly  constant.  It  is  the  third  condition,  that  of 
illumination,  which  is  constantly  variable  and  which  gives  rise 
to  the  greatest  error.  This  source  of  error  can  be  largely  elimi- 
nated by  the  use  of  a  comparison  standard,  prepared  of  the  same 
volume  as  that  used  in  the  color  test  and  hermetically  sealed  in 
a  glass  tube  of  the  standard  size  used  in  testing.  Mixtures  of 
solutions  of  iron  salts  with  finely  pulverized  glass  giving  the 
requisite  turbidity  when  shaken,  can  be  easily  made  to  exactly 
match  the  iodine  tint,  and  will  preserve  their  intensity  indefi- 


HYDROLYSIS  OP  STARCH   BY   ACIDS.  887 

nitely.  When  properly  adjusted  by  means  of  polariscopic  tests 
such  standards  have  served  well  to  fix  the  point  of  conversion 
within  narrow  limits  and  have  done  much  to  insure  a  uniform 
product. 

It  is  of  course  important  that  these  should  be  in  the  hands  of 
the  chemist  or  superintendent  of  the  works,  a  much  more  exact 
means  of  testing  the  degree  of  conversion.  This  is  most  natur- 
ally accomplished  by  determining  the  specific  rotator>'  power. 

We  have  arranged  a  table  for  quickly  calculating  specific 
rotatory  power,  and  found  it  so  useful  that  we  venture  to  publish 
it.  The  following  simple  calculation  will  sufficiently  explain 
the  principles  on  which  the  table  has  been  worked  out : 

TABLE  D. 

Table  for  Determining  Specific  Rotatory  Power  of  Solutions  of 
7.50^-10^  Brix  by  Reading  of  Ventzke  Saccharimetbr. 


Srix. 

Sp.  gr. 

/r=  grram 
per  100  cc. 

Log  '7.20 

Brix. 

Sp.   gT. 

/r=  gram 
per  xoo  cc. 

i^kCV^). 

7.50 

1.0298 

7-724 

0.3477 

S.80 

1.0352 

9.1 10 

0.2760 

7.55 

1.0300 

7-777 

0.3447 

8.85 

1.0354 

9.163 

0.2735 

7.60 

X.0302 

7.829 

0.341S 

S.90 

1.0356 

9.217 

0.2709 

7.65 

1.0304 

7-883 

0.33S8 

8.95 

X.035S 

9.270 

0.2684 

7.70 

1.0306 

7.936 

0.3359 

9.00 

1.0360 

9324 

0.2657 

■■75 

1.030S 

7.0S9 

0-3330 

9.05 

X.0362 

9.378 

0.2634 

7.''«o 

1.03x0 

S.042 

0.3301 

9.10 

io:M 

9.430 

0.2610 

7.»5 

1.0312 

S.096 

0.3272 

9.15 

1.0366 

9.4S4 

02585 

7.90 

1.0315 

S.149 

0.3244 

9  20 

1.036S 

9.538 

0.2560 

7-95 

1.P317 

S.202 
1 

0.3216 

9.25 

1.0370 

9-592 

0.2536 

S.00 

1.0319 

S.255 

0.3187 

930 

1-0372 

9.646 

0.25x0 

S.05 

1.0321 

S.30S 

0.3160 

9.^5 

1.0374 

9.690 

0.2488 

R.10 

1.0323 

S.361 

0.3x32 

9.40 

1.0376 

9.753 

0.2464 

8.15 

1.0325 

0.415 

0.3104 

945 

1.0378 

9.S07 

0.2440 

8.30 

1.0327 

S.468 

0.3077 

9.50 

1.03S1 

9.862 

0.24x5 

S.35 

1.0329 

8.522 

0.3050 

9.55 

1.03S3 

9.916 

0.239X 

S.50 

1.033: 

8.575 

0.3022 

9.60 

10585 

9.970 

0.2368 

».35 

10.V3 

S.629 

0.2995 

965 

1.03S7 

10.023 

o.?346 

R.40 

1.0335 

S.6S2 

0.2969 

9.70 

1.03S9 

X0.077 

0.2323 

8.45 

1.0337 

8.7.^5 

0.2943 

9-75 

10391 

X0.X30 

0.2300 

S.50 

1.0339 

S.7SS 

0.2916 

9.  So 

1.0393 

10.185 

0.2277 

8-55 

1.0341 

$.842 

0.2S89 

985 

1.0395 

10.239 

0.2252 

R.60 

10343 

S.895 

0.2S64 

9.90 

10397 

10.293 

0.223X 

8.65 

10345 

8.9*9 

0.283S 

9.95 

10399 

10.347 

0.2207 

8.70 

1.0347 

9.002 

0.2812 

10.00 

1.Q401 

10.401 

0.2x85 

8.75 

1.0350 

9.056 

0.2786 

Taking  the  usual  formula   for  the  specific  rotatory  power, 

«=  ~,  where  a  is  the  angle  of  rotation  of  the  solution  of  w 

Iw 


888  GEO.   W.    ROI«PB  AND  GBO.    DBPRBN. 

gram  of  the  active  substance  in  v  cc.  of  water  observed  through 
a  column  /  decimeters  long.  If  we  make  a^=^  a  it  is  plain  'w  is 
the  weight  of  substance  under  standard  conditions  which  will 
give  a  direct  reading  of  the  specific  rotatory  power  without  calcu- 
lation. In  an  instrument  reading  in  angular  degrees  under  the 
usual  conditions  of  v  =  loo  and  /=  2,  w  is  therefore  50^. 

If  a  is  the  reading  of  a  saccharimeter  with  the  Ventzke  scale, 
a^.:^  50  X  0.344  =17.20,  and  the  specific  rotatory  power  of  any 
solution  of  known  concentration  of  an  optically  active  substance 

will  be  -^ —  .     The  easiest  way  of  finding  the  concentration 
w 

of  glucose  solutions  with  sufficient  exactness  for  this  work  is  by 
the  Brix  (or  Balling)  hydrometer,  as  this  instrument  is  now 
made  of  great  accuracy. 

Brix  hydrometers  are  carried  in  regular  stock  of  the  larger 
houses  dealing  in  chemical  apparatus  for  brewers  and  sugar 
manufacturers,  with  scales  having  a  range  of  about  five  degrees 
and  easily  read  to  0.05  per  cent.  Thermometers  are  attached 
having  corrections  for  temperature  marked  on  the  scale.  Con- 
centrations of  about  ten  per  cent,  are  mo.st  convenient  for  polar- 
izing ;  hence  a  spindle  will  be  needed  reading  from  five  to  ten 
per  cent. 

The  method  of  determining  rotatory  powers  is  as  follows  :  The 
glucose  is  diluted  to  an  approximately  ten  per  cent,  solution. 
An  exact  Brix  (or  Balling)  reading  is  taken,  corrected  for 
standard  temperature  and  the  solution  polarized  in  a  200  mm. 
tube  in  any  saccharimeter  with  the  Ventzke  scale.     The  loga- 

rithm  of  the  factor  -^ —  corresponding  to  the  Brix  reading  is 

w 

then  found  in  the  table.  Therefore,  the  calculation  which  is, 
log   [«]d  "=■  log    \— — )  +  log  a,  simply  requires  finding  the 

logarithm  of  the  saccharimeter  reading  and  the  number  corres- 
ponding to  the  sum  of  this  and  the  logarithm  given  in  the  table. 
This  number  is  the  required  specific  rotatory  power.' 

1  Obviously  a  table  made  on  the  scheme  of  the  well-known  Schmits  table  for  cane- 
sugrar  aymps  would  do  away  with  all  calculation.  Such  a  table  is,  however,  rather 
bulky  for  insertion  here. 


HYDROLYSIS  OF  STARCH   BY  ACIDS.  889 

Thus  a  solution  of  7.85  Brix  having  a  reading  of  51.7*, 
Ventzke  has  the  rotatory  power  of  its  anhydrous  carbohydrates 
determined  as  follow^ : 

By  the  table,  the  corresponding  logarithmic  factor  is  0.3272. 

Log  51.7  =  1.7135 
Factor  0.3272 

2,0407  =  log  109.8 
which  is  the  required  rotatory  power. 

In  this  calculation  no  correction  is  made  for  ash,  which,  as  a 
rule,  does  not  affect  the  results  appreciably. 

The  errors  due  to  the  slight  variations  in  the  concentration  of 
the  solutions  used  and  changes  in  the  temperature  of  the  labora- 
tory are  too  small  to  be  taken  into  consideration  in  factory  work 
or  in  general  commercial  analysis.  The  method  in  practice  is 
^uite  as  rapid  as  the  '*  quotient  of  purity"  determination  of  cane- 
sugar  syrups.  We  suggest  that  this,  or  some  similar  scheme, 
be  uniformly  used  for  expressing  the  results  of  all  polarimetric 
investigations  of  honeys,  syrups,  and  similar  indeterminate  mix- 
tures of  carbohydrates  met  with  in  commercial  analysis,  instead 
of  merely  giving  the  polari2ations,  or  the  specific  rotatory  powers 
referred  to  the  weights  of  the  sample.  The  advantages  are 
obvious.  Such  analytical  results  would  be  close  approxima- 
tions to  the  exact  specific  rotatory  powers  of  the  mixed  anhydrous 
carbohydrates,  and  would  be  convenient  of  interpretation  by 
inspection  as  being  directly  comparable  on  what  is  for  all  prac- 
tical purposes  an  absolute  standard  and  the  one  used  in  all 
strictly  scientific  work  of  the  kind. 

THE  SPBBD  OF  THE  HYDROLYSIS  OF  STARCH  BY  ACIDS.* 

The  laws  of  the  speed  of  hydrolysis  of  the  carbohydrates  with 
the  exception  of  that  of  cane-sugar  have  been  but  little  studied. 
Solomon*  has  collected  some  data  on  the  action  of  various  acids 
at  boiling  temperature.  Welhelmy'  showed  in  the  case  of  the 
catalytic  action  of  hydrochloric  acid  on  cane-sugar  that  if  the 

1  Wc  are  greatly  indebted  to  Prof.  A.  A.  Noyes*  of  this  department,  for  valuable  aid 
in  calculating  the  results  of  this  work  on  speed  of  hydrolysis. 
^J.prakt,  Chem.,  (a).  a6. 
•  Ber.  d,  chem.  Gts,^  il,  »xi. 


890  GEO.   W.    ROI,FE   AND   GEO.    DEFREN. 

amount  of  acid  and  the  temperature  remained  constant  the  rate 
of  the  inversion  at  any  specified  moment  is  proportional  to  the 
amount  of  unchanged  sugar  present  at  that  moment. 

That  is,  if  A^  represent  the  amount  of  sugar  originally  pres- 
ent, X  the  amount  of  this  sugar  changed  over  in  any  period  of 

time,  /,  and  cthe  reaction-constant,  we  have  ~-  =z  c  (A^ — x). 

at 

The  relative  values  of  the  constant,  c,  of  the  various  acids  in 
their  action  on  cane-sugar  have  been  determined  by  several 
observers,  notably  Ostwald,*  who  has  compared,  by  means  of 
their  constants,  the  relative  effect  of  chemically  equivalent  quan- 
tities of  a  large  number  of  acids,  taking  the  constant  of  hydro- 
chloric acid  as  a  standard  with  the  arbitrary  value  of  100. 

Recent  work  shows  that  acids  act  on  salicin,*one  of  thegluco- 
sides,  in  a  manner  analogous  to  that  of  cane-sugar,  the  speed  of 
hydrolysis  of  this  body  by  the  different  acids  bearing  the  same 
relation  to  hydrochloric  acid. 

The  observations  noted  above  suggested  the  possibility  that 
in  the  hydrolysis  of  starch  the  acids  would  show  the  same  pro- 
portional speed  of  reaction.  This  is  an  especially  interesting 
problem  because  the  starch  molecule  is  exceedingly  complica- 
ted, the  molecular  weight  being  undoubtedly  very  high.  Starch 
hydrolysis,  however,  must  be  considered  as  somewhat  different 
from  that  of  cane-sugar  or  salicin.  While  these  are  easily  solu- 
ble in  cold  water,  starch  is  totally  insoluble  at  ordinary  room 
temperature.  On  the  other  hand,  amylodextrin,  the  product  of 
decomposition  of  starch  by  boiling  water,  is  somewhat  soluble  in 
cold  water,  its  solubility  increasing  with  rise  of  temperature. 

As  by  the  customary  procedure  in  determining  speed  of 
hydrolysis,  it  would  be  necessary  to  ascertain  the  exact  moment 
when  all  the  starch  has  been  converted  into  the  soluble  form,  a 
point  not  conveniently  determined,  we  have  adopted  a  method 
of  measurement,  based  on  the  following  principles  : 

The  conversion  products  of  starch,  with  the  possible  excep- 
tion of  those  of  very  high  rotatory  power,  are  easily  soluble  in 
water,  and  can  be  looked  upon  as  mixtures  of  maltose,  dextrose 
and  dextrin. 

\J,prakt.  Chem.,  18S4,  401. 

2  Noyes  and  Hall :  ZUchr.  phys.  Otem.^  /SgtS^  MO- 


HYDROLYSIS  OP  STARCH  BY  ACIDS.  89 1 

The  starch  first  changes  to  amylodextrin.  The  hydrolysis 
then  proceeds  by  successive  stages  through  the  so-called  malto- 
dextrin;  maltose,  and  dextrose.  **  Reversion,"  so-called,  may 
take  place  to  some  extent,  a  small  amount  of  the  dextrose  form- 
ing dextrin-like  bodies,  **  gallisin,"  **  isomaltose,"  etc.,  but  this 
point  is  not  considered  in  this  work.  The  dextrin  may  there- 
fore be  looked  upon  as  the  original  substance  hydrolyzed,  and 
maltose  and  dextrose  as  successive  products  of  the  reaction. 

Further,  we  have  shown  that  whatever  the  condition  of 
hydrolysis  by  acids,  the  specific  rotatory  power  of  any  conversion 
product  corresponds  to  a  definite  chemical  composition,  tables 
for  determining  which  we  have  constructed. 

Thus,  for  instance,  a  conversion  product  of  i6o^  has  been 
proved  to  contain  54.8  per  cent,  dextrin,  the  remainder  being 
maltose  and  dextrose. 

Hence,  the  time  of  taking  any  sample  after  the  contents  of 
the  autoclave  has  acquired  constant  temperature,  which  re- 
quires about  ten  minutes,  can  be  taken  as  the  initial  point  for 
determining  speed  of  hydrolysis,  and  all  subsequent  samples 
referred  to  this,  as  it  is  obvious  that  in  any  sample  we  can  ascer- 
tain the  dextrin  unacted  upon  at  that  stage  of  the  hydrolysis. 
The  same  holds  true  of  maltose. 

We  have  to  deal  with  two  reactions,  the  first  being  the  hydrol- 
ysis of  dextrin  to  maltose. 

If  Ao  is  the  amount  of  dextrin  at  the  initial  point  taken,  A^ — x, 
the  amount  remaining  at  any  time,  /,  and  c  the  constant  depend- 

dx 
ing  on  conditions  of  hydrolysis  we  get,  -~-  =c  (A^  — x), 

at 

A  I 

This,  on  integrating,    gives    log        °     =   ci,    or  —    log 

A—^x  t 

=f  f ,  which  is  the  general  equation  of  a  first-order  reac- 


^0 — X 

tion.  The  second  decomposition  is  that  in  which  maltose  is 
hydrolyzed  to  dextrose,  and  is  peculiar  in  so  far  as  it  pro- 
ceeds simultaneously  with  that  by  which  the  maltose  is  formed. 
As  a  result  of  the  hydrolysis  of  the  dextrin  the  maltose  increases 
rapidly  to'  a  maximum  of  44.1  per  cent,  at  a  rotation  of  129**. 


892  GEO.    W.    ROLFE   AND   GEO.    DEPREK. 

It  then  gradually  diminishes,  while  the  dextrose  percentage 
always  increases. 

Consequently,  the  equation  expressing  accurately  the  rate  of 
change  in  the  total  amount  of  maltose  present  is  quite  compli- 
cated, and  we  have  therefore  used  an  approximate  formula, 
which  is  sufficiently  exact  for  the  work  in  hand.  The  formula 
is  derived  from  the  exact  differential  equation 

which  states  that  the  amount  of  dextrose  formed  at  each  moment 
is  proportional  to  the  amount  of  maltose  present  by  replacing 
the  differential  quantities  by  finite  differences,  which  in  applica- 
tions of  the  formula  must  of  course  be  taken  small.  In  the 
place  of  M  the  average  amount  of  maltose  present  during  the 
interval  of  time  considered  is  also  substituted.  That  is,  if  M^ 
and  M^  are  the  amounts  of  maltose  present  at  the  time,  /.  and  /,, 
and  D^  and  D^  the  amounts  of  dextrose  present  at  these  same 
times,  and  c^  is  the  reaction  constant,  we  get  as  a  result  of  the 
above  mentioned  substitutions  : 


or, 

1^.  +  ^,  ' 


(^) 


I 


2 

The  results  are  contained  in  the  following  tables  : 

TABLE  E. 
Speed  of  Hydrolysis  of  Starch. 


Time  /. 
(minutes) 

[«]o... 

A  Q'~'X. 

-4. 

Ci. 

2 

c,. 

Hydrochloric  acid 

:  0.02  normal ;  aX2  A  T=  135° 

c. 

io  =  20\ 

c 

"]'i?«..= 

161 ;  Wo 

=  55.8. 

10 

137 

35-5 

0,2216 

0.02216 

0.3581 

0.0358 

20 

IIS 

20.3 

0.4391 

0.02196 

O.3II8 

0.0312 

30 

100 

II. 2 

0.6784 

0.02261 

0.3790 

0.0379 

40 

88 

6.3 

0.9684 

0.02421 

0.3274 

0.0327 

50 

76 

32 

I.24I5 

0.02483 

0.4638 

0.0464 

60 

69 

1.9 

1.4678 

0.02446 

0.4162 

0.0416 

70 

64 

1.2 
0.02344. 

1-6674 

0.02382 

0.4264 

C,=  0.0373. 

0.04.36 

HYDROLYSIS  OP  STARCH   BY  ACIDS. 


893 


40 

70 
100 

X40 

180 


Snlphnric  acid  :  0.0a  normal ;  ata  A  T^  135°  C. 
/o  »  ao ;  [a]  D  ,M  =  177° ;  ^o  —73.5. 


to 

X63 

57.9 

ao 

15a 

46.7 

3«> 

140 

36.0 

40 

139 

375 

60 

109 

15.4 

80 

90 

7.0 

100 

77 

3.4 

lao 

66 

M 

azQ36 
a3i48 
0.3100 
0.4270 
0.6788 
1. 021 3 
1.3348 
1.7103 


O.OX(^ 
O.OII34 

0.0x033 
0.01068 
0.0ZI3X 
0.0x377 
0.0133s 
0.01434 


01954 
O.X436 

O.X703 
0.X678 
a.36s6 
0.4700 
0.4809 
0.6915 
Cg  ss  o.oaxx. 


aoz95 

0.0x44 
0.0x70 
0.0168 
0.0x88 
ao335 
0.0340 
0.0346 


Ci  s  0.01x8. 

Oxalic  acid  :  0.04  normal ;  at  2  ^  T^t  135^  C. 
/oa2o;  [«]i?,„s=l8oO;  -4«=s77.2. 


30 

157 

51.6 

O.X75O 

0.00875 

0.3147 

0.0157 

40 

137 

33.5 

a  ^36 

0.00907 

0.3890 

0.0x45 

60 

X30 

31.5 

0.555a 

0.00935 

0.3738 

aoi37 

80 

106 

13.9 

0.7446 

0.0093X 

0.3763 

0.0x38 

xoo 

93 

8.x 

0.9791 

0.00979 

0.3334 

*  0.0x61 

130 

83 

4.5 

1.2344 

O.00039 

0.3426 

0.0x71 

140 

73 

3.6 

B  0.00957. 

1.4726 

0.01053 

0.4149 

Cn  ss  0.0x59. 

Sulphnrous  acid :  0.02  normal  ;At2A  T^  135^  C. 
/o  — 50;  [«]d„.=  i87^;  ^0=87. 


50 

179 

76.0 

0.0587 

0.00x17 

O.X354 

0.0035X 

xoo 

17a 

67.7 

O.X069 

0.00x09 

0.0907 

o.oox8x 

150 

165 

60.0 

o.x6x3 

0.00X0B 

O.IOX3 

O.OQ303 

300 

159 

53.7 

0.3095 

0.00105 

0.0799 

0.00x59 

ago 

151 

45.7 

0.3796 

O.0OXX3 

o.xq36 

300 

144 

39-3 

0.3451 

O.OOXX5 

0.0978 

0.00X96 

350 

137 

335 

0.4145 

0.00X19 

0.1053 

o.ooaxx 

400 

131 

39.0 

0.4773 

O.OOXX9 

0.0893 

0.00x79 

C^ 

3S  O.OOXX3. 

c^ 

s  0.00x98. 

Acetic  acid 

:  0.5  normal ; 

at  2  ^  r« 

« 135^  c. 

/o  — 50; 

[«]"„.  =  «7o°;  A 

=  65.5. 

50 

143 

38.5 

a3307 

O.OQ46X 

0.3775 

0.00755 

100 

I3X 

33.x 

0.4718 

aoQ473 

0.35x6 

0.00703 

ISO 

103 

X3.5 

0.7193 

aoo48o 

0.3643 

0.00739 

300 

86 

5.6 

1.0680 

O.00S34 

0^638 

0.00938 

350 

74 

3.8 

1.3690 

0.00548 

0.4969 

0.00994 

C^ 

K  0.00449. 

c. 

s  0.00633. 

Hydrochloric  acid  :  0.01  normal ;  at  i  ^  7"=  I2Z^  C. 

4«4o;  [«]",„  =183°;  ^0  =  81.3. 

0.00373 
0.00369 
0.00369 


350 


168 
158 
149 
137 

136 
130 
107 


63.3 

52-7 
43.8 
33.5 

25.5 
31.5 

14.4 


10 

!>»••' 

0.1087 

O.X883 

0.3686 
0.3851 
0.5036 
0.5777 
0.7517 


0.00375 
0.00380 
0.003S9 
0.0030X 


Cj  SE  0.00279. 


0.31X8 

O.X358 

0.1199 
0.1764 
0.1693 

O.X038 

0.2539 
Ci  s  0.00467. 


0.00539 
0.00453 
0.00400 
0.0044X 
0.00423 
0.005x4 
0.00506 


894  OBO.   W.   ROI«PE  AND   GEO.    DBPRBN. 

Hydrochloric  acid :  o.oi  normal ;  at  2  A  Ts=  135^ C. 
/os2o;  [«]"„.=  176°;  Wo«»7a-3- 


10 

163 

56-9 

0.1040 

0.0104 

O.X937 

0.0x94 

90 

148 

4a.9 

0.3366 

0.0x13 

0.1877 

0.0x88 

40 

138 

36.8 

0.4310 

0.0x06 

0.30x5 

O.OI5X 

60 

XIO 

16.0 

0.6550 

0.0x09 

0.3*59 

0.0x63 

80 

93 

8.1 

0.9506 

0.0XX9 

0.4x02 

0.0305 

100 

81 

4.3 

X.3356 

0.0x33 

0.3830 

0.0183 

xao 

70 

3.0 
0.0x15. 

1.5581 

0.0x30 

0.4479 
Cf  as  0.0x87. 

0.0394 

5 

X58              5a-7 

xo 

X40             36.0 

15 

X3S              14.8 

ao 

XIO              x6.o 

30 

88               6.3 

40 

74               a.8 

50 

65               1.3 

Ci  =  0.03x4. 

0.0347 

0.3766 

aQS53 

0.0389 

0.3538 

0.0506 

0.0300 

0.3351 

OXH70 

0.0331 

0.3756 

OJ055X 

0.0349 

0.5544 

0.0554 

0.0350 

0.5630 

0.0563 

0.0344 

0.6349 
C«  9:0.00548. 

0.0640 

Hydrochloric  acid  :  0.01  normal ;  at  3  y^  T^at  145^  C. 
/o«io;  [«]i?,M  — 174°;  ^0  =  70. 

O.X333 
a3888 

0.4506 

0.64x0 

1.0458 

1-3979 

1.73X3 

Hydrochloric  acid  :  o.oi  normal ;  at  4  yf  7^=:  153^  C. 
/o  =*  10 ;  [a]  *D  ,„  =  147° ;  Ao  =  42.0. 

5  XX7  X9.7  0.3387  0.0657  0.4900  0.0980 

xo  96  94  0.650X  0.0650  0.467X  0.0934 

15  79  3-8  1.0434  0.0696  0.5443  0.X088 

30         68         X.8         X.3679        0.0684         0.6060        0.13X3 

25  61       0.85      1.6938      0.0678       0.7x57      O.X43X 

30       57       0.5       X.9343      0.0641       0.7818      0.X564 

Cj  =  0.0668.  Ci  SSO.X303. 

Hydrochloric  acid  :  0.04  normal ;  At  ^  A  T=  145^  C. 


3 

XX5              X8.5 

0.383X 

O.X377 

0.5736 

0.X9X3 

5 

95                9.0 

0.696X 

0.1393 

0.4541 

0.3370 

7 

80               4.0 

X.048X 

O.X497 

0.4833 

0.34x6 

xo 

66                X4 

1.504a 

0.1504 

0.8083 

0.3694 

X3 

58                0.6 

1.8731 

0.1440 

1.0350 

0.3450 

«5 

66               0.4 

C|  s  0.14x3. 

3.0482 

0.X365 

c. 

0.6385 
s  0.3648. 

0.3143 

Hydrochloric  acid  :  0.02  normal ;  At  ^  A  Tss  145^  C. 
/o  =  10  ;  [«]  D „a«  148" ;  Ao  =?42.9. 


5 

xx6              19.  X 

0.3515 

0.0703 

0.5346 

0.X019 

xo 

96               9.4 

0.6594 

0.0659 

0.4478 

0.0896 

X5 

80               4.0 

1.0304 

0.0687 

0.5075 

0.X015 

30 

69               1.9 

1.3537 

0.0677 

0.5889 

0.XX78 

15 

6x                0.85 

X.7031 

0.0681 

0.7739 

0.X548 

30 

56               0.45 
Cj  =0.0678. 

1-9793 

0.0660 

X.0600 
C,  =  0.1304. 

0.3x60 

HYDROI^YSIS  OP   STARCH  BY  ACIDS.  S95 

Hydrochloric  acid :  o.ox  normal ;  at  3  yl  Tss  145°  C. 

^— s;  Wd,„  =  i74°;  a— 70. 


5 

158 

Sa.7 

0.XJ33 

0.0147 

0.3766 

0.0553 

10 

X40 

36.0 

o.a888 

0.OS89 

0.3538 

0.0506 

15 

185 

H.8 

0.4506 

0.0300 

0.335X 

0.0470 

ao 

txo 

x6.o 

0.64x0 

0.0331 

0.3756 

0.0551 

3» 

88 

6.3 

X.Q458 

o.<^J49 

0.5544 

0.0554 

40 

74 

3.8 

X.3974 

0.0350 

0.5^ 

0.0563 

50 

65 
Ci 

u  0.03x4. 

x.yaxa 

0.0344 

0.6349 
C,s  0.0548. 

0.0640 

Hydrochloric  acid : 

0.005  normal ;  at 

ZAT^  145° 

c. 

/o  =  ao ; 

[«]V...- 

172° ;  Ao 

=  67.7. 

ao 

Ua 

37.6 

o.a554 

o.oxaB 

0.4370 

o.oax4 

40 

1x3 

17.4 

0.590X 

0.0148 

0.0345 

60 

9X 

7-3 

0.9673 

o.oi6x 

0.5336 

0.036X 

80 

77 

3-4 

x.a99X 

0.0163 

0.5083 

0.0354 

xoo 

66 

M 

1.6845 

0.0x68 

0.7439 

0.0371 

190 

59 

0.7 
=E  0.0x55. 

X.9855 

0.0x65 

0.8X73 
C|  TT 0.0379. 

0.0409 

At  the  head  of  each  table  are  given  data  as  to  the  concentra- 
tion and  nature  of  the  acid,  the  temperature  corresponding  to 
the  steam  pressure  given  in  atmospheres  and  [^]d386  at  the  initial 
time  period  /«  with  the  corresponding  value  of  A^,  Time  values 
are  expressed  in  minutes,  and  the  constants  c^  for  the  hydrolysis 
of  dextrin,  c^  for  that  of  maltose,  are  calculated  according  to  the 
formulas  given  above. 

The  results  show  that  the  constants  in  general  are  satisfac- 
tory, and  that  therefore  the  reaction  like  the  sucrose  inversion 
follows  the  law  of  the  first  order.  It  will  also  be  seen  that  the 
values  c^  are  much  more  uniform  than  those  of  c^,  which  is  to  be 
expected  since  c^  is  absolute  and  c^  only  approximate.  Devia- 
tions of  c^  may  be  fairly  ascribed  to  variations  in  temperature 
which,  though  slight,  are  significant,  owing  to  the  high  temper- 
ature coe£Gicient  of  the  reaction. 

The  dextrin  values  in  Table  C  are  consequently  correct  within 
the  hmits  of  error  of  analysis.  It  will  be  seen  that  the  values 
of  c^  are  much  more  constant  in  those  determinations  in  which  t 
is  larger  and  the  values  of  [^]d  decrease  slowly.  This  was  to 
be  expected  from  the  conditions  of  the  approximate  formula 
given  above  for  the  decomposition  of  maltose,  these  requiring 
that  the  amount  of  substance  changed  in  a  period  of  time  must 
be  small.     The  question  of  reversion  may  possibly  have  some 


896  GEO.   W.   ROLFB  AND  GBO.    DBPRBN. 

influence  on  the  values  of  r,  but  as  yet  we  are  not  prepared  to 
express  ourselves  definitely  on  this  subject. 

The  relative  effects  are  shown  in  the  following  tabic :  Table 
I  shows  the  influence  on  the  speed  of  hydrolysis  of  various  acids 
at  the  same  temperature,  135®  C. 

Table  II  shows  the  influence  of  temperature  on  the  speed  of 
hydrolysis  when  the  same  amount  of  acid  is  used. 

Table  III  gives  the  influence  of  varying  amounts  of  acid. 

The  mean  value  of  constants  are  given  in  column  II.    Column 

III  gives  the  relative  value  of  the  constants  referred  to  that  of 
y^A^  hydrochloric  acid  at  135°  taken  as  100.  Column  IV 
gives  the  velocity  constants  determined  by  Ostwald'  for  cane- 
sugar  inversion  by  the  same  acids  at  half-normal  concentration. 

Tablb  I. 

Acid.  Concentration.  II.  III.  IT. 

Hydrochloric 0.02  N  0.02344  100  100 

Sulphuric 0.02  N  0.0118  50.35  53.6 

Oxalic 0.04  N  0.00957  40.83  .... 

(      *'      ) (0.02  N)  (0.00479)  (20-42)  18.6 

Sulphurous 0.02  N  0.00113  4.82  .... 

Acetic , 0.5    N  0.00499  21.29  *"■ 

(     **      ) (0.02  N)  o.oo(i20  0.8  0.4 

Tabi«b  II. 

Acid.                     Concentration.  Temp.                 I.  II. 

Hydrochloric o.oi  N            121  0.00279  11.91 

**           0.01  N            134  0.0115  49-07 

**            0.01  N            145  0.0314  13.40 

'*            0.01  N            153  0.0668  28.50 

Tablb  III. 

Acid.                                   Concentration.  II.                         III. 

Hydrochloric 0.04    N  0.1413  602.9 

*'            0.02    N  0.0678  289.3 

'*            O.OI    N  0.0314  134.0 

"            0.005  N  0.0155                 66.13 

It  is  seen  that  the  corresponding  numbers  of  columns  III  and 

IV  agree  fairly  well.  The  relative  influence  of  the  various  acids 
upon  the  hydrolysis  of  starch,  sucrose  and  salicin  are  therefore 
nearly  identical.  It  should  be  noted  however  that  the  chemical 
activity  of  hydrochloric  acid  on  starch,  as  in  the  case  of  salicin 

1  Loc,  ciU 


HYDROLYSIS  OF   STARCH    BY   ACIDS.  897 

and  cane-sugar,  increases  in  a  greater  ratio  than  the  concentra* 
tion,  while  the  electrical  conductivity  increases  more  slowly. 

The  influence  of  temperature  can  be  explained  graphically  by 
a  curve  approximating  a  parabola. 


89S  GEO.   W.   ROLPE  AND   GEO.   DEFREN. 

Plate  II  shows  the  influence  of  the  various  acids. 


HYDROLYSIS  OP  STARCH  BY  ACIDS.  899 

Plate  III  shows  the  tnflaeoce  ot  the  conceDtration,  or  amount 
of  acid  used. 


900  HYDROLYSIS   OF   STARCH    BY    ACIDS. 

Plate  IV  shows  the  relative  curves  due  to  temperature. 


NICKELO-NICKELIC  HYDRATE,  Ni,0,.an,0. 

By  William  L.  Dudlby. 
Received  Aufust  ttf,  M»gii. 

IN  Studying  the  action  of  fused  sodium  dioxide  on  metals,  I 
have  obtained  interesting  crystalline  compounds,  some  of 
which,  at  least,  have  never  been  described.  Only  one  of  them 
has  been  carefully  investigated  and  it  proves  to  be  nickelo- 
nickelic  hydrate,  having  the  formula  Ni,0^.2H,0. 

It  is  prepared  by  fusing  sodium  dioxide  in  a  nickel  crucible 
with  metallic  nickel  at  a  cherry-red  heat.  The  action  of  the 
oxide  upon  the  nickel  proceeds  with  moderate  rapidity,  and  in 
a  few  minutes  scaly  crystals  appear  floating  in  the  fused  mass. 
The  crystals  multiply  steadily  until,  in  the  course  of  an  hour, 
the  contents  of  the  crucible  is  thick  with  them,  and  comparatively 
little  liquid  remains.  After  cooling,  the  crucible  is  submerged 
in  a  beaker  of  distilled  water  and  the  undecomposed  sodium 
dioxide  together  with  the  sodium  oxide  dissolves  out,  leaving 
the  crystals  which  rapidly  settle  to  the  bottom  of  the  liquid. 
The  crystals  should  be  washed  several  times  with  boiling  water 
by  decantation,  and  finally  thrown  in  a  filter.  It  is  quite  diffi- 
cult to  wash  out  all  of  the  alkali,  which  adheres  with  unusual 
persistence.  Probably  the  best  plan  to  adopt  is  to  put  the  crys- 
tals in  a  Soxhlet  extraction  apparatus  and  wash  with  water  until 
no  coloration  is  obtained  with  phenolphthalein.  This  requires 
about  fifty  hours  of  continuous  washing.  The  crystals  should 
then  be  dried  at  iio^  C.  and  a  magnet  passed  carefully  through 
them  to  remove  ^ny  particles  of  metallic  nickel  which  may  have 
eroded  and  not  been  completely  acted  upon. 

The  crystals  are  lustrous  and  almost  black,  with  a  slight  brown- 
bronze  hue.  They  are  soft,  and  grind  in  a  mortar  much  like 
graphite.  The  crystals  seem  to  be  hexagonal  plates,  but  meas- 
urements of  the  angles  have  not  been  made.  They  dissolve 
slowly  in  acids,  forming  uickelous  salts.  Hydrochloric  acid 
evolves  chlorine  ;  sulphuric  and  nitric  acids,  oxygen.  They  are 
insoluble  in  water  and  in  solutions  of  the  alkalies.  The  com- 
pound is  not  magnetic.     The  specific  gravity  is  3.41 15  at  32**  C. 

At  130*  C.  the  compound  does  not  undergo  decomposition,  but 
at  about  140*  C.  it  begins  to  lose  weight ;  at  240''  C.  the  weight 


902  NICKBU>-NICKBUC  HYDRATE. 

remains  constant.  At  a  red  heat  further  loss  is  sustained  and 
the  residue  remaining  is  nickelous  oxide.  The  loss,  from  130**  C. 
to  240^  C.  is  due  to  water  driven  off,  and  at  a  red  heat  this  loss 
is  due  to  the  evolution  of  oxygen. 

The  compound  proved  to  be  Ni,0«.2H,0,  as  is  shown  by  the 
results  of  the  analysis : 

Loss  of  H,0  on  heating  from  130®  C.  to  240*  C. : 

Per  cent. 

First  determination 13.00 

Second  **  13.13 

Theory  for  Ni,04.3H,0 13.06 

The  residue  remaining  after  heating  to  240**  C.  is  Ni,0^. 
On  heating  this  residue  to  redness  the  loss  of  oxygen  was  found 
to  be: 

Per  cent. 

Loss  of  oxygen 6.63 

Theory 6.67 

The  total  loss  of  water  and  oxygen  obtained  on  heating  the 
compound  from  130^  C.  to  redness  was  : 

Per  cent. 

First  determination 18.91 

Second         •*  18.88 

Theory  for  NijO^.aHjO 18.86 

The  oxygen  given  off  on  heating  to  redness  was  determined 
by  calcining  the  compound  in  an  atmosphere  of  carbon  dioxide 
and  collecting  in  Schiff's  apparatus  over  potassium  hydroxide 
solution.    The  result  gave  : 

Per  cent 

Oxygen 5.93 

Theory  for  NijO*. 2HiO 5 .84 

The  nickel  was  determined  and  found  to  be  : 

Per  cent 

Nickel 63.67 

Theory 63.72 

In  all  of  the  calculations  the  atomic  weight  of  nickel  was  taken 
to  be  58.56  and  oxygen  16. 

The  compound  made  in  a  nickel  crucible  of  commerce  is  not 
perfectly  pure,  as  the  sample  obtained  was  found  to  contain  0.71 
per  cent,  of  cobalt,  the  presence  of  which,  however,  would  make 
no  appreciable  difference  in  the  results  of  the  analyses.  No 
method  has  been  found  for  freeing  the  compound  from  this  im- 


TABLE  OF   FACTORS. 


903 


purity,  and  it  appears  at  present  as  if  the  only  plan  would  be  to 
use  a  chemically  pure  nickel  crucible  in  making  it,  forno  crucible 
will  withstand  the  action  of  fused  sodium  dioxide.  Porcelain,, 
iron,  silver,  gold  and  platinum  crucibles  are  rapidly  attacked. 

The  presence  of  water  in  this  compound  seems  curious,  but  it 
may  be  due  to  the  presence  of  sodium  hydroxide  in  the  sodium 
dioxide.  Again  it  may  be  due  to  the  water  added  to  dissolve 
the  soluble  residue  from  the  crystals.  The  first  explanation 
seems  to  be  the  more  plausible  since  the  crystals  are  formed  in 
the  mass  while  it  is  fused,  and  they  are  not  produced  upon  the 
addition  of  the  water.  If  such  is  the  case  it  would  seem  that 
the  water  driven  off  between  130**  C.  and  240**  C.  is  from  the 
breaking  down  of  a  true  hydrate,  rather  than  the  expulsion  of 
water  of  crystallization. 

A  cobalto-cobaltic  hydrate,  Co,0,.2H,0,  has  been  described,' 
but  it  was  obtained  by  exposing  to  moist  air,  Co,0^,  prepared 
by  heating  cobalt  carbonate.  Ni,0^,  prepared  by  heating  nickelo- 
nickelic  h3'drate  to  240*^  C.  is  hygroscopic  and  absorbs  about 
seven  and  four- tenths  per  cent,  of  water  from  the  air  at  30®  C, 
which  is  completely  lost  at  iio^  C,  showing  that  no  hydrate  is 
formed  under  these  conditions. 

The  study  of  the  action  of  fused  sodium  dioxide  on  the  metals 
will  be  continued  here,  and  it  is'  hoped  that  some  more  data 
can  be  contributed  soon. 

Vandbrbilt  university. 


TABLE  OF   FACTORS. 

By  Edmund  H.  Miller  and  J.  A.  Mathews. 

Received  August  6,  xt96. 

ATOMIC  masses^  based  on  0=  i6,  taken  from  an  article  by 
F.  W.  Clarke,  this  Journal,  March,  1896. 


AIPO4 

Sb,04 

Sh,S, 

A8,S, 

Mg,As,OT 

AgiAsOf 

BaSO* 


Required. 
Al. 
A1,0,. 
Sb. 
Sb. 
As. 
As. 
As. 
BaO. 


PRCtor. 
0.221976 
0.418489 
0.790067 
0.714570 
0.609522 
0.483268 
0.162234 
0.657088 


IpOSTftrithm. 

I.346307I 
I.6216835 

1.8976643 

i.8540446 

i.7849890 
1.6841870 
I.2101418 

r.8i 76234 


I  Gentta  and  Gibbs :  Am.  J.  Set',,  S3,  257. 


904 


TABLE   OF   FACTORS. 


Required. 

Factor. 

Logarithm. 

SO,. 

0.342912 

".5351829 

s. 

0.137342 

i.1378121 

Bi,0, 

Bi. 

0.896600 

1.9525990 

CaCO, 

CaO. 

0.560296 

1. 7484173 

CaSO, 

CaO. 

O.411899 

1.6147904 

CaCO,. 

0.735145 

1.8663731 

CO, 

C. 

0.272893 

1-4359916 

Cr,0, 

Cr. 

0.684791 

f.8355581 

3K,S04.2CoS04 

Co. 

0.1415" 

1. 1507892 

CuO 

Cu. 

0.798995 

j. 9025440 

Cu,S 

Cu. 

0.798644 

1.9023531 

Fc,0, 

Fe. 

0.700076 

i.8451446 

Fe 

Fe,0,. 

1.42842 

0.1548554 

FeO. 

1. 28561 

O.IO91100 

Fe,04. 

1.38082 

O.I401359 

PbCrO* 

Pb. 

.640500 

I. 8065 193 

PbSO^ 

Pb. 

.682927 

1.8343742 

Mg,PA 

P. 

.278681 

1. 445 1076 

PA. 

.638038 

1.8048465 

MgO. 

.361962 

"•5586631 

MgCO,. 

.757343 

1.8792934 

MnjO^ 

Mn. 

.720490 

1.8576283 

MnjPjO^ 

Mn. 

.387226 

£.5879648 

CNH,),PtCl. 

Pt. 

.439205 

1.6426669 

N. 

.063281 

2.8012744 

NH,. 

.076911 

2.8859881 

NH,C1. 

.241235 

I  3824396 

Pt  from 

f     N. 

.144081 

1. 1 586075 

(NH,),PtCl, 

NHj. 

.175114 

1. 2433212 

I     NHjCl. 

.549253 

1.7397727 

KjPtCl, 

KCl. 

.306951 

1.4870695 

KjO. 

.193944 

i.  2876767 

KCl 

K,0. 

.631840 

1.8006072 

K,SO, 

K,0. 

.540593 

1.7328706 

SiO, 

Si. 

.470199 

1. 6722814 

AgBr 

Br. 

.425560 

i. 628961 1 

Agl 

I. 

.540313 

1.7326479 

AgCl 

CI. 

.247262 

i393i579 

Ag. 

.752738 

1.8766436 

NaCl 

Na,0. 

•530769 

1.7249057 

Na,SO, 

Na,0. 

.436801 

1.6402836 

SnO, 

Sn. 

.788150 

i. 8966087 

TiO, 

Ti. 

.600749 

1.7786928 

ZnO 

Zb. 

.803464 

1.9049663 

ZdjPjOt 

Zn. 

.429115 

1.6325737 

ZnNH^PO* 

Zn. 

.366438 

1. 564001 1 

RAPID  MEASURING  PIPETTE. 

Bv  Bdward  L.  Smith. 
RocelTad  August  4,  1896. 

THE  apparatus  described  below  is  a  device  for  rapidly  meas- 
uring and  discharging  a  definite  volume  of  liquid.  It 
may  be  well  to  state  at  this  point  that  the  principle  is  not  appli- 
cable in  all,  or  in  even  the  majority  of  cases,  where  it  is  desired 
to  measure  and  discharge  liquid  reagents  in  the  laboratory.  Where 
extreme  accuracy  is  essential,  the  ordinary  pipette  or  a  burette 
must  still  be  used.  Perhaps  the  best  way  to  explain  the  utility 
of  the  apparatus  will  be  to  state  the  exact  use  to  which  it  is  put 

in  our  laboratory.  In  the  course 
of  some  experiments  with  sand  fil- 
ters, samples  of  the  different  efflu- 
ents as  well  as  of  the  applied  sew- 
age were  taken  daily,  treated  with 
a  small  quantity  of  a  concentrated 
sterilizing  agent,  and  an  analysis 
made  each  week  of  the  combined 
daily  samples. 

It  was  to  measure  and  discharge 
this  sterilizing  solution  that  the 
appsCratus  was  devised .  The  quan- 
tity added  in  each  case  was  five  cc. 
Of  course  a  variation  from  that 
amount  of  one-  or  two- tenths  cc. 
would  not  materially  affect  the  re- 
sults and  the  great  advantage  in  con- 
venience and  rapidity  over  the  use  of 
the  common  pipette  for  the  same 
purpose  is  admitted  by  ail  who  have 
seen  the  apparatus  work.  A  large 
bottle  forms  the  reservoir.  The  stopper  of  this  bottle  carries 
two  tubes.  One  simply  serves  to  admit  air  and  contains  a  loose 
plug  of  cotton  to  exclude  dust,  etc.  The  other  tube  is  bent  to 
form  an  ordinary  siphon  and  the  end  of  the  longer  limb  is 
attached  to  a  short  glass  tube  by  means  of  a  rubber  connection. 


906  CHARLES  H.   HBRTY   AND  J.   G.   SMITH. 

provided  with  a  pinch-cock.  The  short  glass  tube  to  which  ref* 
erence  was  just  made  passes  through  a  stopper  inserted  into  the 
mouth  of  an  ordinary  test-tube.  Through  a  hole  blown  in  the 
side  of  this  tube  another  glass  tube,  bent  to  form  a  siphon,  is 
inserted  and  fastened  in  place  by  a  piece  of  rubber  tubing  of  the 
proper  size,  slipped  on  over  the  tube.  The  leg  of  the  siphon 
inside  the  test-tube  is  of  such  a  length  that  when  the  pinch-cock 
above  is  opened  and  the  liquid  allowed  to  enter  the  test-tube, 
five  cc.  will  be  automatically  discharged  when  the  level  of  the 
liquid  has  reached  a  mark  on  a  line  with  the  top  of  the  bend  in 
the  siphon  tube. 

The  apparatus  can  be  constructed  in  a  few  moments  in  any 
laboratory,  and  for  purposes  to  which  it  is  adapted,  it  will,  I  am 
sure,  be  found  satisfactory.  It  may  be  asked,  what  is  the 
advantage  of  the  form  suggested  over  the  ordinary  burette  with 
supply  tubes  ?  The  answer  is,  it  does  away  with  the  necessarily 
oft-repeated  filling  of  the  burette,  and  there  is  but  one  mark  to 
watch  in  making  the  measurement — ^that  previously  mentioned, 
on  the  test-tube.  The  tubing  used  is  of  such  size  that  a  rapid 
discharge  is  insured,  the  time  required  being  less  than  would  be 
the  case  were  a  burette  employed. 


MERCURIC  CHLOROTHIOCYANATE. 

By  Chaklbs  H.  Hbrty  and  J.  G.  Smith. 

Received  August  8.  it96. 

IT  has  been  shown  by  one  of  us'  that  the  so-called  compound 
lead  iodochloride,  PblCl,  is  not  a  true  chemical  compound, 
but  a  mixture  of  lead  iodide  and  lead  chloride. 

It  has  seemed  advisable,  therefore,  to  study  more  fully  the 
nature  of  the  compound  mercuric  chlorothiocyanate,  HgCl(CNS), 
described  by  McMurtry.*  To  this  end  a  series  of  solutions  was 
prepared,  in  one  of  which  was  used  the  exact  proportions  of 
mercuric  thiocyanate  and  mercuric  chloride  given  by  McMurtry 
for  the  preparation  of  mercuric  chlorothiocyanate  ;  in  the  other 
members  of  the  series,  arbitrarily  taken  quantities  of  the  one 
salt  were  replaced  by  equivalent  quantities  of  the  other.     The 

^Am.  Ckem.J.^  i8,  290. 
sy.  ckem.  Soc.,  1889,  50. 


MERCURIC   CHLOROTHIOCYANATB.  907 

mixed  salts  were  completely  dissolved  in  hot  water  and  the  solu- 
tions allowed  to  cool  and  crystallize.  The  quantities  actually 
used  were : 

Mercuric  Mercuric 

Name.  thiocyanate.  chloride.  Water. 

Grams.  Grams.  cc. 

A 9-5000  3*1439  20cx> 

B 6.7500  65.5004  750 

C 5-5000  6.5715  550 

Z>(McMQrtry) 5.0000  7.0000  450 

E 4-5000  7-4285  350 

F' 3-5000  8.2853  300 

G i.oooo  10.4296  200 

On  cooling,  crystals  separated  from  all  of  the  solutions  except 
G,  The  crystals  from  Ay  B,  C,  and  D  were  fern-shaped,  while 
those  from  E  and  /^were  prismatic. 

By  evaporating  the  solution  G  one-half,  quite  a  good  crop  of 
prismatic  crystals  was  obtained.  The  crystals  from  all  of  the 
solutions  were  separated  from  the  mother-liquor  by  filtration 
and  rapid  pressing  between  folds  of  drying  paper. 

From  the  mother-liquor  of  D  two  crops  of  prismatic  crystals 
were  obtained  by  evaporating  to  one-half  and  then  to  three- 
fourths  of  original  volume.  These  were  designated  D^  and 
D\ 

The  character  of  the  various  crops  of  crystals  was  determined 
by  estimating  the  mercury  present  in  each.  This  was  done  by 
reducing  the  compounds  with  sodium  peroxide,  as  recommended 
by  Schuyten,'  and  weighing  the  mercury.     Analysis  showed  : 

Mercury  calculated  for 
Mercury  Mercuric  Mercuric  Mercuric 

found.  thiocyanate,  chlorotbiocyanate,  chloride, 

per  cent.  per  cent.  per  cent.  per  cent. 

A 62.72  63.28  68.12  73-85 

B 62.76 

C 62.74 

D 63.91 

E 68.24 

F 68.67 

G 72.59 

I> 68.45  — 

Z>" ....  72.41 

These  results  show  that  the  various  crops  of  crystals  fall  into 
three  classes,  mercuric  thiocyanate,  mercuric  chlorotbiocyanate, 
and  mercuric  chloride.     This  was  confirmed  by  inspection  with 

1  Chem,  Ztg.t  so,  339. 


•  •  a...  •••■ 


*•  ••*■  ... 


.•  ....  .*• 


....  ....  .. 


•  . 


.«  ....  .... 


a  . 


... 


908  MBRCURIC  CHU>ROTHIOCYANAT9. 

the  microscope.  Further,  the  three  successive  crops  of  crystals 
from  solution  D  are  seen  to  be  the  first  mercuric  thiocyanate, 
slightly  contaminated  by  mercuric  thiocyanate,  as  proved  both 
by  the  high  analytical  result  and  by  microscopic  inspection,  the 
second  crop  is  mercuric  chlorothiocyanate,  and  the  third  mer- 
curic chloride. 

The  low  results  in  the  case  of  the  pure  salts  is  undoubtedly 
due  to  the  fact  that  the  filters  containing  the  reduced  mercury 
were  dried  at  the  ordinary  temperature  with  consequent  slight 
volatilization  of  mercury. 

The  effect  of  crystallization  upon  the  salt  mercuric  chloio- 
thiocyanate  was  next  tried.  A  portion  of  the  salt  was  dissolved 
in  hot  water  just  sufficient  for  complete  solution.  On  cooling 
crystals  separated,  which,  under  the  microscope,  were  seen  to  be 
only  mercuric  thiocyanate.  The  mother-liquor  from  these,  on 
evaporating  one-half,  yielded  only  mercuric  chlorothiocyanate. 
On  evaporating  the  mother-liquor  from  this  last  two-thirds  the 
crystals  formed  are  seen  to  be  a  mixture  of  crystals  of  mercuric 
chlorothiocyanate  and  mercuric  chloride.  Finally,  on  evapora- 
ting this  mother-liquor  to  dryness  spontaneously,  only  crystals 
of  mercuric  chloride  were  obtained.  The  substance  therefore 
undergoes  dissociation  when  dissolved  in  water. 

From  all  of  the  above  it  would  seem  that  mercuric  chlorothio- 
cyanate is  a  true  chemical  compound,  and  further,  that  the  only 
compound  which  can  be  prepared  from  solutions  of  mercuric 
chloride  and  mercuric  thiocyanate  is  that  represented  by  the 

formula    Hg<^5^g  or  HgCl,.Hg(CNS).. 

These  results  varying  so  widely  from  those  obtained  in  the 
case  of  lead  iodochloride  suggest  the  question:  is  the  difference 
due  to  the  fact  that  in  the  one  case  we  have  the  more  closely 
related  groups,  iodine  and  chlorine,  while  in  the  other  we  have 
the  more  different  groups,  thiocyanogen  and  chlorine,  or  is  the 
difference  due  to  the  fact  that  in  the  one  case  we  have  a  lead 
compound  while  in  the  other  a  mercury  salt  ?  To  test  this  point 
work  will  be  begun  at  once  on  mixtures  of  lead  chloride  and 
lead  thiocyanate. 

University  of  Gborgia. 


[Contributions  prom  the  Chbmical  Laboratory  op  Ca»b  School  op 

Applied  Science.] 

XXIV — COMPOSITION  OF  AHERICAN  KAOLINS. 

By  Charx,bs  p.  Mabb&y  andOtzs  T.  Ki.ooz.1 

Received  July  ae,  1^96. 

ALTHOUGH  great  advances  have  been  made  in  recent  years 
toward  a  better  knowledge  of  American  clays  and  suita- 
ble methods  for  the  manufacture  of  ware  from  them,  much  more 
extended  investigation  is  necessary,  both  concerning  the  com- 
position of  the  great  clay  deposits  and  in  the  details  of  manufac- 
ture. The  first  and  most  essential  information  is  a  correct 
knowledge  of  the  composition  of  all  clays  available  for  use.  Of 
scarcely  less  importance  is  masterly  skill  in  the  purification  of 
crude  materials,  shaping  the  ware  and  burning.  In  the  prepa- 
ration of  materia  Isit  is  questionable  whether  American  manufac- 
turers can  wait  patiently  several  months  for  the  slow  processes 
of  lixiviations  and  kneading  that  European  porcelain  makers 
have  found  indispensable  in  the  production  of  the  finest  porce- 
lain. The  great  porcelain  factories  in  Europe  are  founded  on 
the  application  of  scientific  skill  and  a  personality  in  shaping 
and  burning,  handed  down  by  lineal  descent  through  many 
generations.  Is  it  possible  to  procure  for  American  factories 
scions  for  those  ancient  families,  or  must  we  wait  for  its  perfec- 
tion by  our  own  ready  facility  and  ingenuity  ? 

As  already  mentioned,  the  porcelain  manufacturer  must  be 
perfectly  familiar  with  the  composition  of  all  materials  within  his 
reach.  In  making  suitable  mixtures  he  must  have  before  him 
as  one  of  the  most  essential  features  of  composition,  the  propor- 
tions of  free  and  combined  silica,  as  well  as  the  percentages  of 
lime,  iron,  alkalies  and  water. 

Having  at  hand  a  collection  of  clays,  including  representa- 
tives of  American  deposits,  as  well  as  several  specimens  from 
famous  factories  in  Germany,  it  seemed  of  interest  to  compare 
the  composition  of  clays  from  different  sources.  For  the  manu- 
facture of  the  finest  porcelain,  the  kaolin  used  in  the  Royal  Ber- 
lin factory,  at  Charlottenburg,  may  be  accepted  as  a  standard  of 
comparison.      As  every  one  knows  who  is  familiar  with  the 

1  This  work  was  offered  by  Mr.  Klooz  in  a  thesis  for  the  degrree  of  Bachelor  of 
Science. 


9IO  CHARLBS  P.   MABBRY  AND  OTIS  T.    KLOOZ. 

qualities  of  true  porcelain,  the  products  from  this  factory  are 
approached  by  no  other  in  the  world.  The  composition  of  the 
kaolin  used  in  the  manufacture  of  this  ware  is  shown  by  the  fol- 
lowing analysis  of  the  clay,  two  specimens  selected  at  different 
times,  from  great  quantities  within  the  Berlin  factory. 

I.  11. 

Combined  water 6.00  7.65 

Silica 72.16  65.70 

Alumina 20.05  24.49 

Iron o.io  1.03 

Lime 1.14  0.60 

Magnesia  0.02  0.26 

Sodium  oxide 0.12  6.23 

Potassium«ozide 0.41  0.03 

Free  silica 49-84  44*93 

The  different  percentages  in  these  analyses  indicate  that  some 
latitude  is  permissible,  although  a  high  percentage  of  silica  is 
evidently  essential.  These  analyses  show  nearly  the  same  com- 
position as  is  given  in  the  numerous  analyses  of  the  most  cele- 
brated clays  of  the  German  factories,  especially  in  the  low  per- 
centages of  lime,  iron  and  alkalies,  and  the  large  proportion  of 
silica.  An  analysis  of  biscuit  ware  from  the  same  factor)'  shows 
nearly  the  same  composition.  Apparently  the  clay  has  the 
required  proportion  of  silica  without  further  addition  : 

Silica 68.24 

Alumina 29.16 

Iron 0.10 

Lime 1. 18 

Magnesia o.  12 

Alkalies 0.17 

Free  silica 57«50 

Of  the  American  clays,  analyses  showed  that  some  contained 
a  considerable  excess  of  silica  above  the  amount  required  for  the 
oxygen  ratio  of  silica  to  that  of  the  alumina,  2  :  i,  or  the  formula 
Al,0,.3SiO„  which  is  accepted  in  the  manufacture  of  the  best 
German  ware ;  others  only  a  small  excess  of  silica.  Of  the  high 
silica  clays,  a  specimen  from  a  deposit  in  Maryland  gave  the 
following  results : 


COMPOSITION   OF   AMERICAN    KAOLINS.  9II 

Combitied  water 11.23 

Silica 47.60 

Alumina 37'38 

Iron 1.66 

Lime 1.50 

Sodium  oxide 0.22 

Potassium  oxide 0.34 

Free  silica 17.10 

Another  clay  of  this  class  is  a  Missouri  kaolin  which  was 
analyzed : 

Combined  water 4.15 

Silica 82.64 

Iron 12.41 

Lime • 0.05 

Magnesia 0.1 1 

Sodium  oxide 0.08 

Potassium  oxide 0.53 

Free  silica 69.45 

The  following  analysis  represents  another  high  silica  clay 
from  Black  Rock,  Arkansas: 

Combined  water 3.98 

Silica 84.24 

Alumina 11.50 

Iron 0.08 

Lime 0.52 

Magnesia 0.02 

Sodium  oxide trace 

Potassium  oxide 0.42 

Free  silica 69.93 

Another  high  silica  clay  is  from  Milton  Hollow,  Middlesex 
Co.,  N.  J.  : 

Combined  water 5.52 

Silica 75.06 

Alumina 18.32 

Iron 0.08 

Lime 0.80 

Magnesia 0.14 

Potassium  oxide 0.25 

Free  silica 59»7i 

A  specimen  of  clay  from  a  deposit  in  Washington,  Middlesex 
Co.,  N.  J.,  also  showed  a  high  percentage  of  silica : 


912  CHARLBS   F.    MABBRY   AND  OTIS  T.    KLOOZ. 

Combined  water 2.00 

Silica 89. 16 

Alumina 5.77 

Iron 0.07 

Lime 0.70 

Magnesia o.  12 

Potassium  oxide 1.29 

Sodium  oxide 1.31 

Free  silica 80.30 

A  clay  having  nearly  the  same  composition  as  the  specimen 
from  the  Berlin  factory,  is  from  a  deposit  at  Hockessen,  Dela- 
ware : 

Combined  water 6.55 

Silica 71.46 

Alumina 21.02 

Iron 0.08 

Lime 0.54 

Magnesia 0.14 

Potassium  oxide 0.33 

Sodium  oxide 0.36 

Free  silica 53.13 

It  should  not  be  inferred  from  the  foregoing  analyses  that  all 
American  clays  are  high  in  silica.  Some  of  the  largest  and  most 
important  deposits  contain  very  little  free  silica.  One  of  the 
purest  kaolins  is  found  in  large  quantities  in  Indiana,  and  the 
following  analysis  shows  its  composition  : 

Combined  water I5»09 

Silica 44.23 

Alumina - 40.56 

Iron 0.07 

Lime 0.13 

Magnesia o.io 

Potassium  oxide o.io 

Sodium  oxide 0.15 

Free  silica 2.41 

A  clay  of  somewhat  similar  quality  is  found  in  Northampton 
Co.,  Pa.  : 


COMPOSITION  OF  AMERICAN   KAOLINS.  913 

Combined  water 11.20 

Sil ica 48 . 1 6 

Alumina 37-24 

Lime 2.00 

Magnesia 0.29 

Iron 1. 16 

Potassium  oxide 0.25 

Sodium  oxide 0.08 

Free  silica 2.8s 

A  paper  clay  from  South  Amboy,  N.  J.,  Middlesex  Co.,  gave 
the  following  results  on  analysis : 

Combined  wflter I3'35 

Silica 43-30 

Alumina 42.45 

Iron 0.09 

Lime 0.34 

Magnesia o- 10 

Potassium  oxide 0.44 

Sodium  oxide • 0.08 

Free  silica 3.55 

A  washed  clay  used  in  the  manufacture  of  china,  from  New 
Castle,  Del.,  gave  the  following  composition  : 

Combined  water 12.95 

Silica 47.42 

Alumina 38.42 

Iron 0.08 

Lime 0.70 

Magnesia 0.12 

Potassium  oxide 0.30 

Sodium  oxide 0.12 

Free  silica - 4.79 

A  clay  in  Woodbridge,  Middlesex  Co.,  N.  J.,  also  used  in  the 
manufacture  of  ware,  is  nearly  pure  kaolin : 

Combined  water 14-34 

Silica ' 44.34 

Alumina 38.09 

Iron 0.15 

Lime 0.96 

Magnesia o.io 

Potassium  oxide i.oo 

Sodium  oxide 0.79 

Free  silica 1.33 


914  COMPOSITION   OF    AMERICAN    KAOLINS. 

It  is  interesting  to  compare  the  composition  of  American  kao- 
lins with  a  standard  kaolin  used  in  England : 

Combined  water 13.00 

Silica 46.00 

Alumina 40.00 

Iron » 0.33 

Lime 0.33 

Magnesia 0.33 

Several  of  the  clays  analyzed  are  used  in  the  manufacture  of 
ware.  From  some  of  these  deposits  specimens  have  been  ana- 
lyzed, and  the  results  given  in  the  "  Chemistry-' of  Pottery,"  by 
K.  Langenbeck,  are  not  essentially  different  from  those  given  in 
this  paper. 

It  is  evident  that  the  wide  differences  in  the  proportions  of  clay 
and  silica  in  American  kaolins  render  it  imperatively  necessar>' 
that  they  be  taken  into  account  in  the  selection  of  materials  for 
the  manufacture  of  ware.  It  is  also  evident  that  the  United 
States  is  not  wanting  in  an  abundance  of  material  for  the  manu- 
facture of  ware  equal  to  the  best  foreign  production . 

DISCUSSION. 

IVm .  McMurtrie  :  It  is  an  interesting  fact  not  brought  out  here, 
that  in  many  of  the  clays  of  New  Jersey,  and  I  think  particu- 
larly from  some  of  the  deposits  represented  in  the  tables.  Prof. 
Geo.  H.  Cook  reported  appreciable  quantities  of  titanic  oxide 
amounting  to  one-half  per  cent,  more  or  less.  The  same  con- 
stituent has  been  found  in  clays  from  other  localities  which  I  do 
not  now  exactly  remember,  but  I  have  been  led  to  believe  that 
the  existence  of  titanic  oxide  may  be  expected  in  a  good  many 
American  clays. 

W.  A,  Noyes  :  I  have  analyzed  a  number  of  Indiana  clays  and 
have  found  titanic  oxide  with  but  one  exception.  The  Indiana 
clay  given  corresponds  closely  with  one  I  analyzed  last  fall,  and 
that  particular  one  is  free  from  titanic  oxide,  or  practically  so. 
All  the  other  claj's,  and  I  feel  safe  to  say  that  all  these  clays 
must  contain  titanic  oxide. 

The  President :  Does  anyone  know  the  effect  of  titanium  on 
the  ware  ? 


COMPOSITION  OF   CERTAIN   MINBRAL  WATERS.  915 

A,  A,  Brenema7i :  My  impression  is  that  Seger  says  there 
seems  to  be  a  connection  between  the  peculiar  light  gray  of 
salt-glazed  stoneware,  a  color  which  is  unique,  and  the  presence 
of  titanium.  That  is  a  very  interesting  statement,  because  that 
peculiar  form  of  whitish  or  bluish  gray  stoneware  is  very  char- 
acteristic, and  I  see  nothing  in  the  presence  of  iron  alone  in  the 
clay  sufficientlj'  to  account  for  it.* 


[Contributions  from  thb  Chemicai,  Laboratory  of  Case  Schooi«  of 

Appukd  Science.  ] 

XXV.    COMPOSITION  OF  CERTAIN  MINERAL  WATERS   IN 
NORTHWESTERN  PENNSYLVANIA.* 

By  a.  K.  Robinson  and  Charles  P.  Mabery. 

Received  July  a6,  1896. 

THE  therapeutic  qualities  of  mineral  springs  throughout 
northwestern  Pennsylvania  have  long  been  recognized, 
and  recently  some  of  these  springs,  notably  those  at  Saegertown 
and  Cambridgeboro,  have  come  into  prominence  through  the 
enterprise  of  persons  interested  in  hotels  and  sanitariums.  The 
desirable  qualities  of  these  waters  are  doubtless  dependent  on 

INOTE  ov  Titanium  xn  Clays.— In  the  course  of  a  discussion  of  Prof.  Mabery's 
paper  on  American  clays  at  the  Buffalo  meeting  I  alluded  to  the  peculiar  color  of  salt- 
glazed  stoneware,  and  ascribed  to  Seger  the  suggestion  that  it  was  due  to  the  presence 
of  tiUnium.  On  referring  to  Seger's  article  (Wagner's  Jahresbericht,  1883.  p.  625).  I 
find  that  he  says  that  titanic  acid  (13.3  per  cent.)  heated  with  a  very  pure  kaolin  to  a 
temperature  between  the  melting  points  of  wrought  iron  and  platinum  fuses,  and  that 
titanic  acid  is,  under  similar  conditions,  more  of  a  flux  for  clay  than  silicic  acid  is.  In 
the  proportion  of  6.65  per  cent,  of  TiOt,  the  mass  became  only  semi-fused,  and  exhibited 
a  dark-blue  gray  color.  He  says  this  color  suggests  the  tint  given  by  many  clayp 
when  strongly  heated. 

Horgenroth  (Wag.  Jahr.,  1884, 638)  says,  however,  that  rutile  gives  to  clay  ware  a  gray 
color  under  the  glaze  when  impure  ferruginous  clays  are  used,  but  a  yellow,  ivory-like 
tint  with  pure  clays.  As  rutile  was  used  in  the  proportion  of  only  0.4  per  cent.,  the 
minute  proportion  of  iron  which  it  carries  (1.5  to  3.4  per  cent.  Pe,0|)  would  have  little 
effect. 

The  interpretation  of  these  facts  to  explain  the  peculiar  gray  color  of  salt-glazed 
stoneware,  was  probably  a  suggestion  of  my  own,  made  at  the  time  of  reading  these 
articles  a  dozen  years  ago.  It  was  ascribed  in  the  course  of  the  discussion  to  Seger,  as 
my  "  impression." 

Nevertheless,  in  view  of  the  peculiarity  of  this  color,  the  gray  of  salt-glazed  ware 
which  is  uniform  throughout  the  body  and  becomes  more  bluish  in  overbumed  pieces, 
and  in  new  also  of  the  presence  of  iron  in  the  rather  crude  clays  used  for  the  ware,  and 
the  (act  that  iron  alone  tends  to  escape  as  volatile  chloride  in  presence  of  the  salt  used 
for  glazing,  the  suggestion  is  worthy  of  note.  A.  A.  BRBXRMAif .. 

*  This  work,  with  a  study  of  the  methods  of  analysis,  was  offered  by  Mr.  Robinson 
in  a  thesis  for  the  degree  of  Bachelor  of  Science.    Read  at  the  Buffalo  Meeting.  August, 

X896. 


9l6  A.  £.  ROBINSON   AND   CHARI^BS  F.  MABERY. 

iron  and  certain  other  salts,  especially  on  the  bromides,  and  it  is 
a  popular  view  that  lithium  salts  sometimes  present  impart 
valuable  medicinal  qualities.  A  quantity  of  water  was  col- 
lected from  one  of  these  surface  springs  at  Conneautville  by  one 
of  us  (Robinson)  audits  composition  as  shown  by  analysis  may 
serve  as  a  representative  of  the  springs  in  this  region.  The 
total  solids  in  this  water  is  equivalent  to  6.586  grains  per 
imperial  gallon,  or  9.83  parts  per  100,000.  Evidently  the  com- 
bination of  bases  and  acids  is  to  a  certain  extent  arbitrary,  but 
this  distribution  accounts  for  the  total  quantities  of  the  various 
elements  given  by  analysis  : 

Grains 
per  gallon. 

Potassium  carbonate * 0.985 

Lithium  carbonate 0.002 

Sodium  chloride 0-925 

Calcium  bicarbonate  • 2.879 

Calcium  sulphate 1.291 

Magnesium  chloride 0.204 

Ferrpus  carbonate 0.743 

Silica 0.233 

Hydrogen  sulphide trace 

The  specific  gravity  of  this  water  was  foitnd  to  be  i  .0002  at 
20°.  Evidently  the  analysis  shows  the  composition  of  a  good 
potable  water.  Any  medicinal  qualities  it  possesses  must  be 
referred  to  the  iron  and  perhaps  to  a  less  extent  to  the  lithium. 

At  greater  depths  in  this  section  of  Pennsylvania  and  in  cer- 
tain  portions  of  Ohio,  water  may  be  found  that  partakes  in  a 
greater  degree  of  the  qualities  imparted  by  the  constituents  of 
bittern.  Wells  sunk  to  depths  of  1,000  to  3,000  feet  have  pene- 
trated strata  enclosing,  frequently  under  great  pressure,  large 
quantities  of  bittern  waters.  While  in  general  conforming  in 
composition  to  the  salts  contained  in  bittern,  occasionally  these 
wells  have  yielded  peculiar  results  on  analysis.  Such  an  aque- 
ous stratum  was  reached  several  years  ago  at  Conneautville, 
Crawford  County,  Pa.,  in  an  endeavor  to  obtain  oil  or  gas.  The 
drill  penetrated  the  formation  enclosing  water  at  a  depth  of  2,667 
feet  and  the  drilling  tools  were  forced  upwards  to  a  height  of 
1,800  feet  by  the  water  which  prevented  further  drilling.     This 


COMPOSITION  OF   CERTAIN   MINERAL  WATERS.  917 

level  was  maintained  notwithstanding  vigorous  attempts  to  clear 
the  well  by  pumping.  A  slight  examination  then  showed  that 
this  water  possessed  peculiar  qualities,  but  the  well  received  no 
further  attention  until  within  a  few  months  ago  when  it  was  cleared 
and  a  quantity  of  the  water  was  procured  for  a  more  thorough 
examination.  The  total  solids  is  equivalent  to  21,334.34  grains 
per  gallon  or  to  30,536  parts  per  100,000.  The  specific  gravity 
of  the  water  is  1.205  at  15®.  Its  composition  as  shown  by  the 
results  of  analyses  is  as  follows : 

Grains  per  Parts  per 

gallon,  100.000. 

Potassium  chloride 528.577  755-6 

Lithium            "         56.432  80.3 

Ammonium     *'         151*879  216.6 

Soilium             •*         9902.57S  14430.0 

Potassium  bromide 137.010  345*7 

*'         iodide 2.078  2.96 

Magnesium  chloride 2172.499  3096.0 

Calcium              '*         8335.537  1 1880.0 

sulphate 7.886  11. i 

Ferrous  carbonate 1 14.836  163.5 

Aluminum  chloride 21.816  31.  i 

Silica 3.220  4.6 

Hydrogen  sulphide 0.033  0.05 

There  are  certain  features  of  this  water  that  deserve  especial 
mention.  The  large  proportion  of  ammonium  chloride  is  quite 
unusual  in  waters  from  such  depths.  Lithium  chloride  is  fre- 
quently found  in  surface  springs,  and  in  brines  from  deep  wells, 
but  rarely,  if  ever,  in  such  quantities  as  this  water  contains. 
If  lithium  salts  impart  to  spring  water  the  therapeutic  qualities 
claimed  for  them,  it  is  not  difficult  to  account  for  the  beneficial 
effects  that  have  been  observed  in  the  use  of  this  water.  No 
doubt  the  large  proportion  of  potassium  bromide  has  much  to  do 
with  the  marked  sedative  effect.  The  large  percentage  of  potas- 
sium iodide  is  also  phenomenal,  and  it  must  intensify  the  min- 
eral characteristics  of  the  water.  Besides  the  characteristics  of 
a  bromo-lithia  water  the  large  percentage  of  iron  assures  the 
desirable  qualities  of  an  iron  water.  The  peculiar  composition 
of  this  water,  especially  in  the  large  quantities  of  the  rarer  ele- 
ments, offered  a  favorable  opportunity  to  ascertain  whether  these 


9l8  B.    B.    ROSS. 

bittern  deposits  contain  also  the  elements,  cesium  and  rubid- 
ium, which  are  rarely  found  in  springs.  Forty-five  liters  of 
the  water  were  evaporated  to  a  small  volume,  removing  the  great 
quantities  of  salt  as  they  separated.  When  the  volume  was 
reduced  to  less  than  fifty  cc.  this  solution  as  well  as  the  lixivia- 
ted salts  that  had  separated  during  evaporation  were  carefully 
examined  in  the  spectroscope.  But  not  a  trace  of  rubidium  nor 
cesium  could  be  detected.  It  is  therefore  safe  to  conclude  that 
the  bittern  deposits  from  the  ancient  sea  do  not  contain  these 
rarer  elements. 

It  may  not  be  out  of  place  to  remark  that  the  chemical  com- 
position of  this  water  explains  the  remarkable  therapeutic  quali- 
ties especially  for  rheumatism  and  nervous  diseases  that  it  has 
been  found  to  possess. 


SOriE  ANALYTICAL  HETHODS  INVOLVING  THE  USE  OF 

HYDROGEN  DIOXIDE.' 

Bv  B.  B.  Ross. 

Received  Au);ust  31.  i8g6. 

THE  use  of  hydrogen  peroxide  as  a  laboratory  reagent, 
although  originally  restricted  to  a  few  operations  of 
minor  importance,  has  within  recent  years  met  with  a  m\ich 
wider  extension,  and  its  numerous  applications  in  both  qualita- 
tive and  quantitative  analysis,  render  it  at  present  almost  indis- 
pensable in  every  well-equipped  analytical  laboratory. 

Among  the  more  interesting  applications  of  this  substance  in 
quantitative  estimations  are  those  which  are  based  on  the  reac- 
tion which  takes  place  when  an  excess  of  hydrogen  dioxide  is 
brought  in  contact  with  an  acid  solution  of  chromic  acid,  and 
Baumann'  several  years  since  described  quite  fully  a  number  of 
analytical  processes  growing  out  of  the  reaction  referred  to. 

In  the  process  for  the  estimation  of  chromic  acid  in  soluble 
chromates  as  outlined  by  Baumann,  the  substance  under  exami- 
nation is  first  brought  into  a  state  of  solution,  and  the  not  too 
concentrated  liquid  is  transferred  to  a  generating  flask  of  special 
construction. 

iRead  at  the  Buffalo  meeting^,  Augrust  23,  1S96. 
3  Ztichr.  anal.  CMem.,  31,  436. 


USE   OF    HYDROGEN   DIOXIDE.  919 

Ten  cc.  of  dilute  sulphuric  acid  are  next  added,  after  which 
from  five  to  ten  cc.  of  commercial  hydrogen  peroxide  are  run  in 
from  a  small  closed  vessel  connected  with  the  generating  flask, 
while  the  oxj'gen  which  is  evolved,  after  the  vigorous  shaking 
of  the  contents  of  the  flask,  is  collected  over  water  in  an  azotom- 
eter. 

The  following  equations  given  by  Baumann  illustrate  the 
chemical  changes  connected  with  the  above  described  reaction  : 

K.Cr,0,  +  H,0,  +  H,SO,  =  k,SO,  +  2H,0  +  Cr A  ; 
CrA  +  3H,SO,  +  4H  A  =  Cr,{SO  J,  +  j^fi  +  O.. 

From  these  equations  it  will  be  seen  that  for  two  molecules  of 
chromic  acid  or  one  molecule  of  potassium  dichromate,  there  are 
evolved  eight  atoms  of  oxygen,  giving  an  equivalent  of  445.3  cc. 
of  oxygen  (measured  at  0°  C.  and  760  mm.  pressure)  for  each 
gram  of.  chromic  acid  which  may  be  present. 

The  writer,  soon  after  the  appearance  of  the  original  article 
by  Baumann,  made  a  number  of  experimental  tests  of  this  method 
with  a  view  to  applying  it  to  some  other  analytical  processes, 
and  still  more  recently  has  conducted  a  series  of  tests  for  the 
purpose  of  determining  the  adaptability  of  Baumann's  method  to 
the  indirect  volumetric  estimation  of  iron. 

In  the  dichromate  method  for  the  volumetric  determination  of 
iron,  as  commonly  employed,  the  end  point  of  the  oxidation 
process  is  ascertained  b}*  the  reaction  with  potassium  ferricy- 
anide. 

As  the  end  of  this  reaction  is  almost  invariably  difficult  to 
determine  particularly  if  zinc  has  been  employed  as  a  reducing 
agent,  the  dichromate  process  has  met  with  but  limited  applica- 
tion. 

In  order  to  apply  the  principle  of  the  chromic  acid  method  of 
Baumann  to  the  estimation  of  iron,  an  excess  of  dichromate 
solution  was  employed  in  all  of  the  tests  and  experimental  deter- 
minations, the  amount  of  the  excess  of  chromic  acid  being  deter- 
mined by  the  volume  of  oxygen  evolved  upon  treatment  with 
hydrogen  dioxide. 

The  mode  of  procedure  adopted  was  as  follows  : 

A  dichromate  solution  was  prepared  by  dissolving  4.913  grams 


920  B.    B.    ROSS. 

of  C.  P.  crystallized  potassium  dichromate  in  water  and  dilu- 
ting to  a  bulk  of  one  liter. 

The  iron  solution  employed  in  standardizing  the  dichromate 
and  permanganate  solutions  was  obtained  by  dissolving  iron 
wire  in  dilute  sulphuric  acid,  the  solution  being  reduced  vrith 
metallic  zinc,  as  usual,  previpus  to  titration. 

The  dichromate  solution  was  also  titrated  against  a  freshly 
prepared  solution  of  ammonium  ferrous  sulphate,  the  strength  of 
which  had  been  determined '  by  titration  with  permanganate 
solution,  which  had  also  been  carefully  standardized  by  means 
of  iron  wire. 

In  order  to  ascertain  the  strength  of  the  dichromate  solution 
by  the  hydrogen  dioxide  method,  about  fifteen  cc.  of  the 
dichromate  solution  is  run  into  the  generating  flask  above 
referred  to,  and  there  is  also  added  an  amount  of  ferric  sul- 
phate solution  (free  from  ferrous  sulphate)  equivalent  to  about 
0.06  to  o.io  gram  of  iron.  The  object  of  employing  the  ferric 
sulphate  in  this  standardization  is  to  supply  approximately  the 
same  conditions  as  obtain  in  the  process  for  the  actual  deter- 
mination of  iron. 

The  amount  of  oxygen  given  off  from  chromic  acid  in  the 
presence  of  ferric  sulphate  is  slightly  less  than  thate\'olved  when 
ferric  sulphate  is  absent,  but  the  amount  of  ferric  iron  present 
may  vary  considerably  without  affecting  the  volume  of  oxygen 
liberated. 

To  the  contents  of  the  generating  vessel  about  ten  cc.  of  dilute 
sulphuric  acid  are  now  added,  and  the  flask  is  then  connected 
by  means  of  a  rubber  tube  with  a  Schulze's  azotometer,  which 
has  been  filled  with  water  to  the  zero  point. 

From  five  to  ten  cc.  of  hydrogen  dioxide  are  next  run  in  from 
a  small  closed  vessel  connected  with  the  generating  flask  and  the 
mixed  liquid  is  then  shaken,  at  first  gently,  and  afterwards  vig- 
orously. The  tube  leading  from  the  flask  to  the  azotometer 
should  be  provided  with  a  stop-cock,  which  should  be  closed 
before  and  opened  immediately  after  each  shaking. 

The  last  trace  of  the  oxygen  liberated  will  not  be  disengaged 
until  after  the  lapse  of  about  five  minutes,  but  it  is  notnecessj'r>' 
to  continue  the  shaking  during  the  whole  of  this  period.     After 


USE   OP   HYDROGEN    DIOXIDE.  921 

* 

eqaalizing  the  height  of  the  water  in  the  two  tubes  of  the 
azotometer,  the  volume  of  oxygen  is  noted  and  is  easil}'  cor- 
rected for  temperature  and  pressure  by  reference  to  proper  tables. 

In  order  to  test  the  strength  of  the  dichromate  solution  by 
means  of  iron  wire,  a  given  weight  of  the  wire  is  dissolved  in 
dilute  sulphuric  acid,  the  solution  reduced  with  zinc,  as  usual, 
and  rapidly  transferred  to  the  generating  flask  (filtering,  if 
necessar>0 . 

An  excess  of  dichromate  solution  is  now  run  in,  hydrogen 
dioxide  is  added,  and  the  oxygen  is  set  free  and  collected  as 
belore  described. 

If  a  large  excess  of  dichromate  has  been  used  in  the  prelimi- 
nary test,  duplicate  tests  should  be  made  with  employment  of  a 
small  excess,  say  from  two  to  three  cc,  of  the  dichromate. 

The  strength  of  the  solution  can  then  be  readily  calculated  by 
difference,  and,  if  necessary,  the  results  can  be  checked  by  still 
further  tests. 

In  the  determination  of  iron  in  ores  by  this  process,  the  solu- 
tions of  ferric  iron  are  reduced  by  zinc,  as  in  the  common  per- 
manganate method,  and  the  remainder  of  the  process  is  con- 
ducted just  as  described  for  the  standardization  of  the  dichro- 
mate b}'  means  of  iron  wire. 

In  addition  to  numerous  tests  of  solutions  of  pure  iron,  several 
estimations  of  iron  in  iron  ores  were  made  by  this  process,  the 
results  obtained  being  compared  with  those  secured  by  the  per- 
manganate method. 

The  following  are  the  results  of  the  tests  of  the  iron  ores 
referred  to : 

Permaiigauatc  method. 
Mean  of  several  deter mitiatiotis.    Dichromate  method. 

Iron  ore  No.  i 40.92  f ?'5? 

41-23 

55-35 
Iron  ore  No.  2 54.71  55.43 

55-50 
In  the  determination  of  iron  in  ores  b}'  this  process,  it  is  best, 
as  in  the  case  of  the  tests  with  iron  wire,  to  employ  only  a  small 
excess  of  the  dichromate  solution,  after  making  a  preliminary 
determination,  as  the  results  are  much  more  accurate  with  a 
small  than  with  a  large  excess  of  chromic  acid. 


922  USE   OF    HYDROGEN    DIOXIDE. 

While  a  sufficient  number  of  determinations  have  not  been 
made  to  ascertain  the  probable  value  of  this  method  as  an  inde- 
pendent process  for  the  estimation  of  iron,  nevertheless  some  of 
the  results  secured  would  seem  to  warrant  the  conclusion  that  it 
might  prove  of  utility  as  a  check  method,  it  being  easy  of  execu- 
tion and  not  at  all  time-consuming. 

The  following  equation  represents  the  changes  which  take 
place  when  the  dichromate  is  brought  in  contact  with  the  iron 
solution  after  reduction  : 

6FeS0,  +  K,Cr,0,  +  yH^SO,  =3Fe.(SOJ.  +  K.SO,+ 
Cr.(SO,).  +  jH^O. 

The  writer  has  also  attempted  to  apply  the  principle  of  the 
chromic  acid  method  above  described  to  the  estimation  of  invert 
sugar,  or  rather  to  the  determination  of  the  amount  of  cuprous 
oxide  thrown  down  from  Fehling's  solution  in  the  process  com- 
monly employed  for  estimating  reducing  sugars. 

The  following  equation  represents  the  changes  which  take 
place  when  cuprous  oxide  is  brought  in  contact  with  potassium 
dichromate  in  the  presence  of  dilute  sulphuric  acid  : 

3Cu,0  +  K,Cr,0,  +  ioH,SO,  =  6CuS0,  +  K.SO,  + 

Cr,(S0j3-|-  ioH,0. 

The  cuprous  oxide  thrown  down  from  the  sugar  solution 
under  examination  is  brought  upon  an  asbestos  filter  connected 
with  a  filter  pump  and  thoroughly  and  rapidly  washed  with  hot 
water.  The  filter  and  contents  are  next  transferred  to  the  gen- 
erating flask  of  the  apparatus  before  described,  and  after  the 
addition  of  dilute  sulphuric  acid,  an  excess  of  dichromate  is 
run  in. 

Very  thorough  and  long  continued  agitation  of  the  contents  of 
the  flask  is  necessary  in  order  to  effect  the  complete  oxidation 
and  solution  of  the  cuprous  oxide,  and  the  hydrogen  peroxide 
must  not  be  added  until  the  solution  is  complete. 

The  oxygen  liberated  on  the  addition  of  the  hj-drogen  dioxide 
is  collected  and  the  volume  noted  as  before  described.  The 
equivalent  amounts  of  chromic  acid,  cuprous  oxide  and  invert 
sugar  can  be  easily  calculated  from  the  data  thus  secured. 

This  method,  while  apparently  satisfactory  from  a  theoretical 


INTERNATIONAL  CONGRESS  OF  APPLIED  CHEMISTRY.       923 

standpoint,  has  so  far  failed  to  give  sufficiently  uniform  results, 
one  of  the  chief  objections  to  the  process  being  the  difficulty 
attendant  upon  the  solution  of  the  cuprous  oxide. 

With  improvements  in  the  details  of  manipulation  of  the  pro- 
cess, however,  it  is  quite  po.ssible  that  more  satisfactory  results 
could  be  obtained. 


SECOND  INTERNATIONAL  CONGRESS  OF  APPLIED 

CHEniSTRY. 

By  H.  W.  Wiley. 

Rece«\e«l  September  15.  1S96. 

At  the  first  congress  held  in  Brussels,  in  1S94.  it  was  decided 
to  hold  the  meetings  bi-annually  and  Paris  was  selected  as  the 
mo.st  desirable  place  for  the  reunion  this  y/ear.  As  has  already 
been  announced  to  the  readers  of  the  Journal,  the  present  con- 
gress is  organized  under  the  patronage  of  the  French  govern- 
ment and  under  the  immediate  direction  of  1' Association  des 
Chimistes  de  Sucrerie  et  de  Distillerie  de  France  et  des  Colonies. 
The  late  Professor  Pasteur  had  accepted  the  honorary  presidency 
of  the  congress,  and  all  delegates  from  foreign  countries  have 
felt  an  especial  regret  that  his  death  has  prevented  them  from 
listening  to  his  words  of  welcome  and  from  forming  his  personal 
acquaintance. 

To  promote  the  interests  of  the  congress,  committees  were 
organized  in  most  countries.  The  personnel  of  the  one  in  the 
United  States  has  already  been  published  in  this  Journal. 
Through  the  French  Foreign  Office  all  the  principal  govern- 
ments were  invited  to  send  delegates  to  the  congress.  Official 
representatives  were  present  from  Belgium,  Germany,  Italy, 
Russia,  Switzerland,  Austria,  Portugal,  Denmark,  and  the 
United  States.  So  far  as  I  can  learn,  and  the  fact  is  worthy  of 
remark,  there  is  no  representative  in  attendance  from  England, 
either  official  or  otherwise.  The  official  delegate  from  the 
United  States  is  Mr.  C.  A.  Doremus,  of  New  York,  while  the 
writer  has  a  commission  as  a  delegate  from  the  Department  of 
Agriculture,  and  one  from  the  American  Chemical  Society, 
sent  through  the  courtesy  of  the  president  and  council.  Bel- 
gium has  the  largest  representation  of  any  foreign  country,  and, 
since  these  gentlemen  are  all  French  in  their  language,  the  con- 
gress, as  is  natural,  is  essentially  French. 

The  congressrwas  formally  opened  July  27,  at  10  a.  m.,  in  the 
grand  amphitheater  of  the  Sorbonne.  Perhaps  there  is  no  other 
spot  in  the  whole  world  so  well  suited  by  its  history  and  tradi- 


924  H.  W.  WILEY.      SECOND   INTERNATIONAL 

tions  for  the  seat  of  a  scientific  congress,  especially  of  chemistr>'. 
In  or  near  the  Sorbonne  were  made  those  advances  in  chemical 
science  which  have  made  famous  the  names  of  Lavoisier,  Chev- 
reul,  Dumas,  Deville,  Wurtz,  Pasteur,  Berthelot,  and  maii3' 
others  scarcely  less  renowned.  The  address  of  welcome  was  fitly 
made  by  Mr.  Berthelot,  rendered,  by  the  death  of  Pasteur,  the 
head  and  front  of  French  science.  The  response  was  pronounced 
by  Mr.  Lindet,  provisional  president.  After  these  addresses,  the 
provisional  secretary  of  the  congress  presented  a  report  showing 
the  activity  of  the  French  and  other  committees  and  giving  the 
number  of  chemists  who  had  become  members  of  the  congress. 

The  congress  is  organized  with  ten  sections,  as  follows  : 

r*  Section. — Sucrerie. 

2*  Section. — Industries  de  la  fermentation  :  alcools,  vins, 
bieres,  cidres,  vinaigres. 

3*  Section. — Industries  agricoles  :  laiterie,  fromagerie,  f6cu- 
lerie,  amidonnerie,  glucoserie,  mati^res  alimentaires. 

4'  Section. — Chimie  agricole :  engrais,  terres,  eanx  residu- 
aires  ;  alimentation  du  b6tail. 

5*  Section. — Analyses  officielles  et  commerciales  des  matieres 
soumises  k  Timpot. — Appareils  de  precision. 

6*  Section. — Indu.stries  chimiques  :  produits  chimiques,  phar- 
maceutiques ;  corps  gras,  caoutchouc,  matieres  colorantes, 
papiers,  tannerie,  verrerie,  cferamique,  etc. 

7'  Section. — Photographie. 

8*  Section. — M^tallurgie,  mines,  explosifs,  etc. 

g""  Section. — Chimie  Appliqu6e  a  la  medecine,  k  la  toxicologic, 
k  la  pharmacie,  k  I'hygiene  et  k  Talimentation.  Matiferes  ali- 
mentaires :  alterations  et  falsifications. 

lo*  Section. — Electricity  :  61ectro-chimie. 

The  meetings  of  the  congress  are  held  in  the  Hotel  de  la  So- 
ci^tfe  d*  Encouragement  del' Industrie  Nationale,  44  rue  de  Rcn- 
nes,  opposite  the  church  of  St.  Germains  des  Pr6s  and  in  the 
H6teldesSoci6tesSavantes,  situated  in  rue  Serpente,  opposite  rue 
Danton.  Only  four  or  five  of  the  sections  are  in  session  at  any 
one  time,  thus  affording  an  opportunity  to  the  members  of  the 
congress  of  attaching  themselves  to  several  sections. 

In  the  afternoon  of  the  first  day  visits  were  made  to  the  Gobe- 
lin tapestries,  the  Museum  of  Natural  Histor}-,  botanical  gardens, 
the  National  Tobacco  Factory,  and  the  Eiffel  tower,  the  latter 
being  reached  by  boats  on  the  Seine.  At  the  end  of  these  visits 
a  banquet  was  served  on  the  first  floor  of  the  tower  and  from  the 
tables  a  pleasing  vision  of  Paris  by  night  was  obtained. 

On  the  second  day  of  the  congress,  an  interesting  paper  was 
read  by  Mr.  Moissan  on  the  electric  furnace.     A  large  number 


CONGRESS  OF  APPLIED   CHEMISTRY.  925 

of  samples  of  tlie  typical  compounds  obtained  at  the  intense  heat 
of  the  furnace  was  exhibited  and  a  description  of  their  physical 
and  chemical  properties  given.  The  possibilities  of  the  electric 
furnace  in  the  near  future  were  outlined.  Mr.  Moissan  described 
in  some  detail  the  construction  of  the  furnace.  It  is  best  made 
by  carving  a  block  of  quicklime  into  the  proper  shape.  The 
high  infusibility  of  the  quicklime  and  its  non-conducting  power 
are  points  in  its  favor.  The  electrodes  should  be  of  the  purest 
carbon  and  there  should  be  no  deflection  of  the  arc  into  the  cru- 
cible. The  control  of  the  current  is  of  the  greatest  importance. 
For  instance,  in  the  case  of  titanic  oxide  it  is  reduced  to  tiianous 
oxide  with  a  current  of  thirty  amperes ;  at  300  amperes  tita- 
nium nitride  is  produced  and  at  3,000  amperes  titanium  carbide. 
Many  metallic  carbides,  as,  for  instance,  calcium,  yield  a  gas 
when  moistened,  but  the  gases  are  not  identical.  In  addition  to 
acetylene,  hydrogen,  marsh  gas,  and  petroleum  have  been 
obtained,  the  latter  from  uranium  carbide.  This  fact  is  of  great 
interest  in  respect  of  the  origin  of  natural  gas  and  petroleum, 
which,  by  many,  are  supposed  to  be  of  organic  derivation.  In 
the  furnace,  molybdenum  and  manganese  are  capable  of  form- 
ing compounds  similar  to  cast  iron.  Fine  samples  of  chromium 
obtained  in  the  furnace  were  shown  and  many  specimens  of 
various  nitrides,  carbides,  and  borides.  Chromium  oxide  was 
reduced  to  metal  before  the  audience  and  silica  was  sublimed. 

In  addition  to  Mr.  Moissan's  paper,  a  general  discussion  of 
electrolytic  problems  was  held  including  electrolytic  methods  of 
preparing  chlorine,  chlorinated  soda,  and  calcium  carbide. 

Mr.  Moissan  has  accepted  an  invitation  to  attend  the  Prince- 
ton College  celebration  in  the  autumn  and  has  made  arrange- 
ments to  give  some  lectures  in  the  United  States.  Our  chem- 
ists, therefore,  will  have  an  opportunity  in  the  near  future  to 
bear  him  and  to  note  the  great  progress  which  the  elec- 
tric furnace  has  made  possible  in  the  line  of  discoveries  in  min- 
eral chemistry. 

Another  discussion  of  unusual  interest  was  devoted  to  the 
official  graduation  of  instruments  of  precision.  It  was  the 
general  consensus  of  Opinion  that  a  uniform  100  gram  weight  of 
platinum  should  be  adopted  by  all  countries,  and  that  all  instru- 
ments and  utensils  for  w^eight  and  volume  should  be  referred  to 
this  standard.  The  ofRcial  meter  was  regarded  by  all  to  be  the 
ultimate  standard  of  instruments  to  measure  length.  Some  of 
the  members  favored  a  standard  of  brass  coated  with  gold  or 
platinum,  in  order  to  have  an  ultimate  standard  of  greater  volume 
than  the  one  made  of  platinum.  The  difficulty  of  securing  brass 
of  uniform  and  definite  constitution  was  considered  as  an  insu- 
perable objection  to  this  proposition. 


926  H.  W.  WILEY.      SECOND   INTERNATIONAL 

Among  the  man}'  papers  of  special  interest  read  on  this  day 
only  a  few  can  be  mentioned  here  by  title,  viz,.  Application  of 
Electro-Chemistrj'  to  the  manufacture  of  Chemical  Products,  by 
M.  Joly  ;  The  Difficult  Digestibility  of  Sterilized  Milk,  by  M. 
Laurent ;  Determination  of  Soil  Elements  Assimilable  by  Plants, 
by  M.  Garola  ;  Plan  and  Installation  of  an  Agricultural  Experi- 
ment Station,  by  M.  Soillard. 

At  4  p.  M.  the  sections  were  adjourned  to  visit  the  new  city 
hall  (Hotel  de  Ville),  which  has  finally  been  completely  restored 
from  its  destruction  by  the  Commune.  There  the  members  were 
.received  by  the  mayor  of  the  city  (Prefet  de  la  Seine),  the  chief 
of  police  and  the  chief  of  the  fire  department.  After  enjoying  a 
delightful  collation,  such  as  the  city  of  Paris  knows  so  well  how 
to  prepare,  we  were  conducted  b}'  the  mayor  throughout  the 
building  and  had  described  to  us  the  mural  decorations  and  the 
various  groups  of  statuary-  In  the  opinion  of  experts,  the  new 
Hotel  de  Ville  is  quite  equal  in  its  artistic  decorations  to  the 
magnificent  structure  so  wantonly  destroyed  by  the  Communists 
in  187 1. 

On  the  third  daj-  of  the  congress  sessions  of  the  sections  were 
held  only  in  the  morning.  A  communication  was  presented  to 
the  second  section  by  Mr.  Chas.  J.  Murphy,  describing  a  new 
process  of  fermenting  maize  and  showing  the  way  to  a  more 
extended  use  of  this  product  in  the  European  distilleries. 
Before  the  third  section  was  read  several  papers  giving  the  latest 
European  processes  for  the  manufacture  of  starch.  Mr.  Grau- 
deau,  an  agronomist  well  known  in  the  United  States,  presented 
a  communication  to  the  fourth  section  on  the  assimilability  of 
phosphates.  Methods  of  analysis  of  phosphates,  especially  those 
applicable  to  phosphatic  slags  were  discussed  by  Mr.  Class,  of 
Halle,  and  by  many  others.  The  Wagner  method  of  solution 
in  ammonium  citrate,  of  a  definite  constitution,  was  advocated 
by  nearly  all  those  taking  part  in  the  discussion.  A  paper  on 
the  official  German  method  of  determining  iron  and  alumina  in 
phosphates,  was  presented  by  Dr.  von  Grueber.  The  method 
of  E.  Glaser,  as  modified  by  Jones,  is  the  one  which  the  German 
chemists  regard  as  the  most  reliable.  This  method  has  already 
been  described  in  the  Jownial  of  Ayialytical  and  Applied  Chemis- 
try, 6,  671.  It  was  pointed  out  that  analysts  had  received  an 
impression  that  E.  Glaser  had  acknowledged  that  this  method 
was  unsound.  This,  however,  is  not  the  case,  but  the  impres- 
sion arose  by  reason  of  a  critique  of  the  method  by  C.  Glaser,  of 
Baltimore.  Mr.  E.  Glaser  died  soon  after  publishing  his  method 
and  it  devolved  on  Dr.  Grueber  to  continue  his  work.  The 
modifications  of  the  original  method,  as  proposed  by  E.  Glaser, 


CONGRESS  OF   APPLIED   CHEMISTRY.  927 

which  have  been  accepted  by  the  German  chemists  are  princi- 
pally those  made  bj-  Jones  and  with  which  American  chemists 
are  quite  familiar.  The  process,  as  conducted  by  the  German 
official  chemists,  is  as  follows  : 

Ten  grams  of  the  sample  are  dissolved  in  twenty-five  cc. 
hydrochloric  acid,  sp.  gr.  1.20,  and  the  volume  completed  to  a 
half  liter.  Fifty  cc.  of  this  solution,  corresponding  to  one  gram 
of  the  substance,  are  evaporated  to  half  that  volume  in  a  beaker, 
ten  cc.  of  sulphuric  acid  (one  part  to  four  of  water)  added  and 
the  mixture  shaken.  150  cc.  of  absolute  alcohol  are  added, 
shaken,  and  the  beaker  placed  aside  for  three  hours.  The 
deposited  calcium  sulphate  is  separated  by  filtration  and  washed 
with  absolute  alcohol.  The  washing  is  finished  when  ten  drops 
of  the  filtrate,  diluted  with  the  same  volume  of  water,  does  not 
become  red  when  a  drop  of  a  solution  of  methyl  orange  is  added. 
The  alcohol  from  the  filtrate  and  washings  is  recovered  by  dis- 
tillation, and  the  residue  oxidized  by  bromine  and  hydrochloric 
acid,  a  slight  excess  of  ammonia  added  and  heated  until  the 
excess  is  expelled.  This  operation  is  very  important  to  prevent 
the  incorporation  of  magnesia  in  the  precipitate.  The  residual 
precipitate  is  separated  by  filtration,  any  remaining  on  the  walls 
of  the  beaker  being  washed  off  with  cold  water  and  a  rubber- 
tipped  tube.  The  whole  is  washed  on  the  filter  with  boiling 
water  until  all  traces  of  sulphuric  acid  have  disappeared.  The 
precipitate  is  dried,  ignited  and  weighed  and  consists  of  the 
phosphates  of  iron  and  alumina.  One-half  of  the  weight  of  the 
precipitate  consists  of  the  oxide  of  iron  and  alumina. 

The  quantity  of  iron  is  determined  by  reducing  the  iron  in  fifty 
cc.  of  the  first  solution  made,  by  means  of  zinc,  and  titrating  the 
amount  reduced  by  a  solution  of  potassium  permanganate  in  the 
usual  way.  The  quantity  of  iron  having  thus  been  determined, 
it  is  calculated  to  oxide  and  subtracted  from  half  the  weight  of 
the  iron  and  aluminum  phosphates.  The  difference  is  the  alumina. 

The  members  of  the  photographic  section  were  provided  with 
an  interesting  program,  but  the  writer  was  not  able  to  be  present, 
and  the  total  absence  of  any  reports  of  the  meetings  in  any  of 
the  daily  papers,  or  in  any  other  accessible  form,  makes  it  impos- 
sible to  give  even  a  summary  of  what  was  accomplished.  I  do 
not  think  it  advisable  to  encumber  the  pages  of  the  Journal  with 
a  complete  list  of  the  papery  presented,  inasmuch  as  the  present- 
ing of  the  titles  of  the  papers  would  fill  many  pages  and  give 
but  little  idea  of  the  proceedings.  Moreover  the  published  pro- 
ip-am,  although  extensive,  does  not  include  perhaps  more  than 
half  the  titles  of  the  papers  presented,  and  I  am  not  sufficiently 
acquainted  with  the  French  way  of  doing  things  to  be  able  to 


928  H.  W.  WII.EY.      SECOND   INTERNATIONAL 

complete  the  list.  Only  one  program  of  papers  and  proceedings 
has  been  printed,  and  that  evidently  is  to  serve  for  the  whole 
congress.  The  French  in  this  particular  might  well  imitate  the 
practice  of  the  American  Association  for  the  Advancement  of 
Science  in  providing  daily  programs. 

Interesting  communications  were  presented  to  the  ninth  sec- 
tion on  food  adulteration,  and  Mr.  Doremus  read  a  paper  on  the 
nature  of  the  gases  contained  in  canned  goods.  He  showed  that 
these  gases  were  chiefly  hydrogen  and  probably  the  hydrogen  is 
produced  by  galvano-electric  action  in  the  metals  of  the  can. 
In  all  cases  where  much  gas  was  found,  the  sides  of  the  can  were 
found  deeply  corroded.  There  was  no  evidence  in  these  cases 
of  the.  action  of  ferments  and  in  every  case  the  sterilization  of 
the  canned  goods  was  perfect.  Mr.  Thomas  Taylor  sent  to  the 
section  a  communication  on  the  crystals  of  butter  fat  embodying 
the  results  of  his  observations  while  chief  of  the  Division  of 
Microscopy  of  the  Department  of  Agriculture.  Mr.  F.  Jean  read 
a  communication  on  the  distinction  between  butter  and  marga- 
rine as  determined  by  his  instrument,  the  oleorefractometer. 
This  instrument  has  been  carefully  tested  in  the  Chemical  Divi- 
sion of  the  Department  of  Agriculture,  and  while  it  has  been 
found  to  give  valuable  indications  it  is  by  no  means  so  definitely 
diagnostic  as  its  inventor  claims. 

Before  the  eighth  section  were  presented  memoirs  on  the 
methods  of  determining  sulphur,  phosphorus,  nickel  and  carbon. 

The  afternoon  of  the  third  day  (Wednesday)  was  given  over 
to  a  visit  to  the  celebrated  agricultural  school  and  experiment 
station  at  Grignon.  The  members  of  the  congress  traveled  by 
railway  to  Versailles  where  carriages  were  provided  to  conduct 
us  to  Grignon.  Passing  the  palace  and  garden  of  Versailles, 
we  entered  the  forest  and  after  two  miles  reached  a  stretch  of 
fields  which  for  beauty  and  fertility  are  scarcely  equaled  in  the 
world.  The  wheat  and  oats  harvests  were  going  on  and  I  was 
impressed  with  the  primitive  methods  employed.  The  cradle  and 
the  sickle  are  almost  universally  used,  only  one  reaping  machine 
being  seen  in  a  drive  of  ten  miles.  At  Grignon  the  tourists  were 
received  by  Mr.  Philippar,  the  principal  of  the  school,  and  by  Mr. 
Deherain,  the  director  of  the  station,  whose  name  and  fame  are 
well  known  to  all  chemists,  especially  those  engaged  in  agricul- 
ture in  the  United  States. 

The  experimental  plots  of  thestatibnwej-e  explained  by  Mr.  De- 
herain, and  thereafter,  in  his  laboratory,  he  gave  a  brief  explanation 
of  the  charts  representing  the  results  of  the  experiments  for  many 
years.  After  leaving  the  experiment  station,  the  members  of 
the  congress  were  driven  over  the  farm   connected  with  the 


CONGRESS  OF   APPLIED   CHEMISTRY.  929 

school  and.  they  also  inspected  the  barns,  stables,  horses  and 
herds  of  sheep  and  cows.  I  noticed  that  much  of  the  agricul- 
tural machinery,  especially  the  reapers,  hay-rakes  and  plows,  were 
of  American  manufacture.  The  college  buildings  are  part  of  an 
old  chateau  which,  under  the  first  empire,  belonged  to  one  of  the 
marshals  of  France.  The  school  at  Grignon  is  the  largest  and 
most  important  of  the  three  national  colleges  of  agriculture. 
The  other  two  are  established  at  Montpellier  and  Rennes  respect- 
ively. Three  classes  of  pupils  are  admitted;  viz.,  internes,  who 
pay  $240  a  year,  demi-internes,  who  pay  $120,  and  externes, 
who  pay  $80.  Others  known  as  free  auditors  are  also  admitted 
to  all  the  lectures  and  pay  $40  a  year.  The  course  of  instruc- 
tion lasts  two  years  and  a  half  and  includes  zoology,  botany, 
mineralogy,  agricultural  geology,  physics,  meteorology,  general 
and  agricultural  chemistry,  agriculture,  horticulture,  arboricul- 
ture, viticulture,  sylviculture,  rural  economy,  entomology,  seri- 
culture, apiculture,  technology,  agricultural  legislation,  hygiene 
and  military  exercises.  The  number  of  pupils  admitted  to  each 
class  is  fixed  annually  by  ministerial  decree,  and  is  limited  also 
in  the  class  of  internes  by  the  number  of  beds.  The  total  num- 
ber of  pupils,  excluding  the  free  auditors,  is  about  250.  On  the 
completion  of  the  course  and  passing  a  satisfactory  examination, 
which  shall  merit  at  least  sixty-five  out  of  a  possible  100  points, 
the  pupil  receives  the  diploma  of  the  National  School  of  Agricul- 
ture, and  four-fifths  of  the  whole  number  thus  graduating,  com- 
prising those  who  have  received  the  highest  marks,  are  excused 
in  time  of  peace  from  all  military  service,  except  one  year. 

Examinations  for  admission  to  the  school  are  competitive  and 
include  arithmetic,  algebra,  geometry',  trigonometry,  elementary 
physics,  chemistry,  zoology,  botany  and  geology.  The  chem- 
ical instruction  is  given  by  Mr.  Deherain  and  his  assistants  and 
consists  of  lectures  and  demonstrations  in  general  and  agricul- 
tural chemistry,  including  the  chemical  study  of  plants,  soils 
and  fertilizers.  It  is  evident,  however,  that  in  the  short  time  at 
their  disposal  the  students  can  not  acquire  great  efficiency  in 
chemical  manipulations  and  in  fact  it  is  not  the  object  of  the 
school  to  train  agricultural  chemists,  but  rather  to  provide  young 
men  with  that  character  of  instruction  which  will  enable  them  to 
manage  with  intelligence  and  in  harmony  with  the  most 
advanced  teachings  of  science,  large  landed  estates. 

Those  members  of  the  congress  who  did  not  desire  to  visit 
Grignon  were  offered  an  alternative  excursion  to  the  nickel 
works  of  Messrs.  Christofle,  Bouilhet  and  Cie.,  at  Saint  Denis.. 
I  have  not  been  able  to  secure  any  reports  of  this  visit. 

The  fourth  day  of  the  congress,  July  30,  was  devoted  exclu- 


930  H.  W.  WILEY.      SECOND   INTERNATIONAL 

sivelj'  to  the  honor  of  the  late  M.  Pasteur.  At  9.30  in  the 
morning,  the  members  assembled  in  the  chapel  of  Notre 
Dame  and  placed  a  memorial  wreath  on  Pasteur's  coffin.  The 
body  of  the  illustrious  savant  lies  in  an  alcove  near  the  middle 
of  the  north  side  of  Notre  Dame,  the  coffin  scarcely  visible 
beneath  a  mountain  of  wreaths  and  crowns.  Not  onl}*  is  the 
alcove  in  which  the  coffin  rests  full  of  these  offerings,  but  they 
have  been  stored,  in  cart-loads,  in  all  the  adjoining  alcoves. 
They  come  from  individuals  and  learned  societies  from  all  parts 
of  the  world  and  from  nearly  every  municipality  in  France. 
The  coffin  rests  here  temporarily  until  the  tomb  and  monument, 
to  be  erected  by  popular  subscription  from  all  parts  of  the  world, 
are  ready.  The  final  resting  place  of  the  body  of  Pasteur  is  to 
be  in  the  court  of  the  Pasteur  Institute.  With  bowed  heads  the 
members  of  the  congress  marched  by  the  coffin  holding  onlj-  the 
motionless  brain  whose  activity  has  done  so  much  to  advance 
knowledge  and  benefit  mankind.  Thence  the  carriages  con- 
veyed us  to  the  Pasteur  Institute  where  the  laboratories  were 
inspected.  A  collection  of  many  compounds  of  historical  inter- 
est, prepared  by  Pasteur,  was  on  exhibition,  among  which  were 
all  the  tartaric  acids  and  tartrates  used  b}'  Pasteur  in  demon- 
strating molecular  asymmetry  as  displayed  by  the  ^ame  chemical 
substance  having  opposite  relations  to  polarized  light.  A  large 
collection  of  original  cultures  of  the  ferments  leading  to  the  dis- 
covery of  antidotes  for  rabies  was  also  on  exhibition.  A  large 
number  of  microscopes  showing  the  specific  microbes  of  phthisis, 
cancer  and  diphtheria  attracted  g^ieral  interest.  In  the  clinical 
rooms  we  were  permitted  to  see  one  of  the  daily  inoculations 
with  antirabic  serum.  About  thirty  patients  were  treated  in  less 
than  half  that  number  of  minutes.  About  two  or  three  cc.  of 
serum  are  administered  by  hypodermic  injection  to  each  patient. 
The  serum  is  inserted  in  the  skin  on  the  right  or  left  side  of  the 
abdomen,  the  most  convenient  place  on  account  of  the  infre- 
quency  of  nerves.  Each  patient  receives  from  ten  to  fifteen 
injections  on  successive  days.  About  150  patients  are  received 
monthly,  and  the  treatment  is  entirely  gratuitous.  Those  who 
are  able,  however,  usually  give  generously  to  the  funds  of  the 
institute.  A  large  collection  of  rabbits,  guinea  pigs  and  dogs, 
serving  for  experimental  purposes,  was  also  inspected.  We  were 
next  driven  to  St.  Cloud  and  through  its  beautiful  gardens  and 
forests  to  Garches,  where  a  delightful  breakfast  was  served  at 
one  o'clock.  After  breakfast  a  visit  was  made  to  the  stables 
containing  the  horses  used  to  furnish  the  anti-diphtheritic  serum. 
There  are  120  of  these  and  all  seemed  to  be  in  perfect  health. 
Each  one  of  these  horses  has  been  inoculated  with  the  diphthe- 


CONGRESS  OF  APPLIED   CHEMISTRY.  93 1 

ritic  poison  and  the  blood  thereafter  serves  as  the  source  of  the 
serum.  Two  horses  were  operated  on  as  an  illustration  of  the 
method  of  work.  A  large  vein  in  the  neck  of  the  animal  is 
opened,  a  tube  inserted  and  the  blood  collected  in  a  sterilized 
jar.  So  skillfully  is  this  accomplished  that  scarcely  a  drop  of 
blood  is  lost.  From  four  to  six  liters  of  blood  are  collected  from 
each  animal,  when  the  vein  is  closed  and  the  horse  returned  to 
his  stall.  In  three  or  four  weeks  he  is  ready  to  supply  another 
quantity  of  blood.  The  jars  containing  the  blood  are  placed  in  a 
cupboard  for  about  forty-eight  hours,  when,  if  the  horse  has  been 
properly  inoculated,  their  contents  will  be  found  sharply  separated 
into  clots  and  serum.  The  serum,  which  is  of  a  light  yellow 
color,  is  removed  by  decantation  and  by  an  ingenious  appara- 
tus, which  prevents  all  danger  of  infection,  is  bottled  in  vials 
containing  ten  cc.  each.  One  horse  was  shown  us  that  had  fur- 
nished in  the  past  few  years  several  hundred  liters  of  serum.  He 
appeared  to  be  good  for  many  hundred  more.  The  serum  thus 
prepared  is  used  directly  by  subcutaneous  injection  on  patients 
suffering  from  diphtheria.  Every  appointment  in  these  stables 
was  such  as  to  impress  the  visitors  with  a  new  and  a  noble  idea  of 
science,  ministering  thus  directly  to  saving  life  and  especially 
the  lives  of  children.  No  wonder  the  body  of  him  who  did 
so  much  to  establish  the  lines  of  investigations  which,  under  his 
immediate  direction,  if  not  by  his  own  hands,  have  led  to  such 
ameliorations  in  the  sufferings  of  men,  lies  to-day  in  honor  in 
one  of  the  most  magnificent  churches  in  the  world,  buried  under 
flowers  and  wreaths,  while  the  piemory  of  his  work  lives  immor- 
tal in  the  hearts  of  the  people  it  has  blessed. 

Next  was  inspected  the  national  porcelain  works  at  Sevres, 
reached  after  a  pleasant  drive  from  Garches.  The  officials  of 
the  factory  received  the  guests  at  the  entrance  and  dividing  the 
visitors  into  small  parties  each  was  personally  conducted  through 
the  works.  Beginning  with  the  crude  materials,  kaolin,  quartz, 
etc.,  the  methods  of  grinding  and  mixing  were  first  explained. 
The  character  of  the  mixing  is  of  course  suited  to  the  nature  of 
the  object  in  view,  the  massive  urns  and  vases  having  a  different 
proportion  of  the  several  ingredients  from  the  delicate  cups  and 
saucers.  The  molding  of  the  objects  was  shown  in  detail  in  its 
three  forms;  viz.,  by  carving  the  solid  moist  mass,  by  allowing  it 
in  a  pasty  state  to  flow  into  moulds,  and  by  turning  the  waxy 
mass  on  a  table  and  imparting  the  desired  form  by  the  hands  of 
the  operator.  The  urns  and  vases  are  made  by  the  first  and 
third  methods,  while  the  thinner  vessels,  such  as  cups,  etc.,  are 
made  by  the  second  method.  After  drying,  the  glaze  is  applied 
by  dipping  the  objects  in  a  creamy  bath  of  the  silicates  serving 


932  H.  W.  WILEY.      SECOND  INTERNATIONAL 

to  form  the  glaze.  After  the  glazing  is  fixed  by  firing,  the 
objects  are  passed  to  the  decorating  room  to  receive  their  final 
colorings.  After  each  color  is  applied,  it  is  fixed  by  firing. 
The  ingenious  hoods  used  to  secure  an  even  firing  of  the  objects 
were  exhibited  and  the  manner  of  using  them  shown.  The  con- 
struction of  the  large  furnaces  where  hundreds  of  vases  and  other 
objects  are  fired  at  once,  was  described,  and  the  furnaces  cold 
and  in  action  exhibited.  The  visit  concluded  with  an  inspec- 
tion of  the  museum  and  salesrooms  with  their  artistic  and  costly 
contents.  These  are  known  to  all  visitors,  but  the  process  of 
manufacture  which  was  so  minutely  shown  us  is  not  open  to  the 
public  in  general.  The  day  was  finished  by  a  drive  back  to 
Paris  through  the  parks  of  Meudon  and  Boulogne. 

Having  taken  the  whole  of  the  fourth  day  for  the  interesting 
and  instructive  excursions  which  have  just  been  briefly 
described,  the  fifth  day,  Friday,  July  31,  was  wholly  devoted  to 
the  scientific  work  of  the  congress.  Sections  i,  2,  4,  5,  9,  and 
10  held  morning  sessions,  and  2,  3,  5,  6,  8,  and  9  met  in  the 
afternoon.  The  time  of  Section  i  was  devoted  to  a  discussion 
of  the  crystallization  of  sugars  and  the  methods  of  suppressing 
the  molasses  in  the  manufacture  of  sugar  from  canes  and  beets. 
The  papers  presented  and  the  discussions  thereon  were  more 
technical  than  chemical. 

In  the  second  section,  the  difficulties  attending  the  detection 
and  estimation  of  the  higher  alcohols,  aldehydes  and  ethers  in 
brandies  and  whiskies  were  set  forth  and  Mr.  Tavildaroff,  of  St. 
Petersburg,  gave  a  rfesumfe  of  the  best  methods  of  procedure. 

In  the  third  section  the  methods  of  determining  phosphoric 
acid  in  soils  and  fertilizers  were  again  the  subject  of  discussion 
and  papers  on  this  subject  were  presented  by  Messrs.  Garola 
and  Sidersky.  Mr.  Lasne  presented  a  rfesumfe  of  his  work  on 
the  detection  and  estimation  of  iron  and  alumina  in  phosphates. 
In  section  5,  papers  on  the  analysis  of  fats,  estimation  of  acetic 
acid  in  pyroligneous  acid,  a  new  method  of  estimating  alcohol 
by  means  of  the  ebuUioscope,  and  a  rapid  method  of  analyzing 
denaturalized  alcohol  were  presented  by  Messrs.  Jean,  Kestner, 
Wiley,  and  Guillier,  respectively. 

In  the  ninth  section,  the  application  of  the  spectroscope  in 
medico-legal  cases  was  discussed.  In  connection  with  a  discus- 
sion of  the  influence  exerted  by  ptomaines  on  the  detection  of 
alkaloids  in  medico-legal  cases,  Mr.  Dorenius  presented  a  paper 
entitled  *' Recovery  of  Morphine  from  a  Cadaver  Embalmed 
with  Arsenical  Solution .'' 

The  subject  of  the  possible  detection  of  toxines  in  potable 


CONGRESS  OF   APPLIED   CHEMISTRY.  933 

waters  was  also  discussed  and  the  influence  exerted  on  them  by 
organic  matters  in  process  of  decomposition  pointed  out. 

In  the  afternoon,  in  section  2,  a  paper  was  presented  by  Mf. 
Kayser  on  the  properties  of  yeasts  of  different  origin.  A  sub- 
ject of  interest  to  the  wine  growers  of  our  southern  states  and 
California  was  a  paper  on  the  vinification  in  warm  climates,  by 
Mr.  Dugast.  The  pasteurization  of  wines  was  discussed  in  a 
paper  by  Mr.  Malvezin.  Other  papers  of  interest  to  wine  makers 
were  presented  and  discussed. 

To  chemists  and  bacteriologists  engaged  in  the  manufacture 
and  study  of  butter  and  cheese,  the  proceedings  in  section  3  were 
of  great  interest.  The  best  methods  of  disinfecting  stables  and 
creameries  by  chemical  means  were  presented  by  Mr.  Bordas.  A 
rteum6  of  our  knowledge  concerning  the  influence  of  food  on  the 
composition  and  character  of  milk  and  butter  was  presented  by 
Mr.  Martin.  A  general  discussion  of  the  best  means  of  provi- 
ding cities  with  pure  milk  was  led  by  Mr.*Saillard.  The  impor- 
tance of  selecting  ferments  in  the  manufacture  of  butter  and 
cheese  was  discussed  by  the  section,  but  the  work  done  by  Conn 
and  others  in  the  United  States  did  not  seem  to  be  appreciated. 

In  section  3  a  paper  on  the  effect  of  impurities  on  the  proper- 
ties of  metals  was  presented  by  Mr.  Le  Verrier,  and  the  methods 
of  micrographic  and  photomicrographic  examination  of  metals 
and  alloys  were  described  by  Mr.  Osmond. 

In  section  5,  Mr.  Jobin  presented  a  paper  giving  the  data  for 
comparing  the  different  saccharimetric  scales  in  use  in  the  deter- 
mination of  sugar  by  the  polariscope,  and  the  method  of  secur- 
ing a  uniform  scale  was  discussed  by  Mr.  Sidersky.  It  was 
voted  that  a  quartz  plate  of  exactly  one  millimeter  thickness  was 
the  most  scientific  standard  by  which  to  measure  or  fix  a  sac- 
charimetric scale.  The  most  probable  value  of  this  standard 
at  the  present  time  is  expressed  by  an  angular  rotation  of 
21**  40'. 

In  section  8,  papers  were  read  by  Mr.  Lasne  on  the  phosphate 
industry,  by  Mr.  Th.  Schloesing  on  the  condensation  of  vapors 
at  a  high  temperature,  on  the  ammonia  industry  by  Mr.  Truchot, 
and  several  other  papers  of  less  importance. 

In  section  9,  the  subject  of  the  analysis  of  urine  and  the 
determination  of  urea  was  discussed  by  Messrs.  Monfet,  Taffe, 
Hodencq,  Vicario,  Hugnei,  Barthe,  Girard,  and  Doremus,  the 
latter  describing  an  apparatus  for  the  purpose,  invented  some 
time  ago  by  himself,  and  also  the  use  of  bromine  dissolved  in 
sodium  bromide,  as  proposed  by  Rice. 

In  the  evening  a  lecture  was  given  to  the  congress  in  the 
amphitheatre  of  the  Sorbonne  on   color  photography  by  Mr. 


934  H.  W.  WII.EY.      SECOND  INTERNATIONAI, 

Lippmann,  who  has  achieved  an  international  reputation  by  his 
researches  into  this  important  process.  The  principles  of  color 
photography  were  described  and  illustrated  by  apt  experiments 
in  conjunction  with  a  projecting  lantern.  The  process  developed 
by  Lippmann  is  based  on  the  well  known  properties  of  thin  films, 
as,  for  instance,  a  soap  bubble  to  show  colored  bands  due  to  the 
relation  between  the  thickness  of  the  film  and  the  length  of  the 
waves  of  light.  Mr.  Lippmann  has  succeeded  in  depositing  on 
a  glass  plate  superimposed  films  of  silver  of  extreme  tenuousness 
and  each  of  these  films  differs  in  thickness  for  each  variation  of 
color  in  the  object  producing  the  photograph.  When  the  photo- 
graph is  thus  constructed  it  happens  that  when  it  is  viewed  by 
reflected  light,  every  color  of  the  object  photographed  is  exactly 
reproduced.  A  large  number  of  these  photographs,  representing 
paintings,  flowers,  landscapes  and  persons,  was  projected  by 
reflection  with  the  most  vivid  verisimilitude.  Perhaps  the  most 
interesting  of  these  was  the  spectrum  of  argon,  in  which  the  blue 
bands  were  shown  in  perfectly  natural  colors  and  clearly  defined. 
The  photographic  effect  is  secured  by  exposing  a  perfectly  trtins- 
parent  sensitive  plate,  backed  by  metallic  mercury,  in  contact 
with  the  film.  The  sensitive  surface  of  the  plate  is  turned 
away  from  the  object  to  be  photographed.  The  plate  holder 
for  this  operation  was  shown  and  is  remarkable  alike  for 
its  ingenuity  and  simplicity.  The  importance  of  color  photog- 
raphy, as  a  means  of  fixing  objects  for  study,  is  as  great  as  its 
usefulness  will  prove  to  be  in  preserving  with  all  the  tints  of 
vitality'  the  faces  of  friends  and  the  beguilements  of  beauty. 

Mr.  Lippmann  kindly  granted  to  Mr,  Doremus  and  myself  a 
private  interview  after  the  lecture,  where  we  had  a  better  oppor- 
tunity to  examine  the  negatives.  They  resemble  the  daguerreo- 
types of  forty  years  ago  and  a  distinct  view  of  the  image  is  only 
obtained  by  inclining  the  plate  in  the  proper  manner  to  secure 
the  reflection  of  the  light.  Unfortunately,  these  negatives  are 
not  capable  of  being  reproduced  as  positives  as  in  the  case  of 
ordinary  photography,  and  we  are  apparently  as  far  away  as 
ever  from  multiple  printing  color  photography. 

Sixth  day,  Saturday,  August  i.  Sessions  of  the  sections  were 
held  only  in  the  morning  and  those  meeting  were  i,  2,  4,  6,  7, 
8  and  10. 

In  the  fourth  section  Mr.  Kjeldahl  gave  a  brief  statement  of 
the  present  methods  of  conducting  his  process  for  the  determina- 
tion of  nitrogen  by  moist  combustion.  Papers  on  methods  of 
detecting  and  preventing  frauds  in  the  sale  of  commercial  fertil- 
izers were  presented  by  Mr.  Petermann.  A  paper  on  the 
importance  of  international  agreement  in  methods  of  agricultural 


CONGRESS  OF   APPLIED   CHEMISTRY.  935 

analysis  was  presented  by  the  writer.  A  general  discussion  of 
official  methods  of  analyzing  fertilizers  was  carried  on,  and  at 
the  end  it  was  voted  that  the  congress  collect  and  publish  in 
German  and  French  the  official  methods  of  France,  Germany 
and  the  United  States.  Mr.  Sidersky  was  selected  as  editor  of 
this  brochure, 

Messrs.  Roy  and  Jean  gave  a  paper  in  section  6  on  tannins, 
their  nature  and  analysis.  It  contained  little  that  is  new  to 
American  chemists  and  showed  a  lack  of  familiarity  with  the 
American  publications  on  that  subject. 

In  section  7  Mr.  Vogel  presented  a  paper  on  photograph}'  in 
colors,  and  one  on  the  same  subject  was  presented  by  Mr.  Vidal. 
These  papers  gave  in  detail  the  points  given  en  risumi  in  Mr. 
Lippmann's  lecture. 

In  section  9,  Mr.  Guichard  read  a  paper  on  alcohol  from  a 
hygienic  point  of  view. 

The  employment  of  aluminum  in  the  construction  of  cooking 
utensils  arid  its  influence  on  the  wholesomeness  of  food  prepared 
therein  was  the  subject  of  a  paper  by  Mr.  Boronia.  It  was  shown 
that  with  proper  precautions  aluminum  could  be  safely  used, 
but  that  it  presented  few  if  any  advantages  over  copper  or  other 
metals  in  common  use. 

So  widely  has  aluminum  come  into  use  for  cooking  utensils 
that  a  brief  abstract  of  our  present  knowledge  concerning  its 
merits  may  be  presented.  The  utility  of  an  aluminum  dish,  in 
respect  to  its  fitness  for  culinary  vessels,  depends  on  the  purity 
of  the  metal.  A  pure  aluminum  dish  is  almost  if  not  quite  as 
resistant  to  solvent  effects  of  ordinary  foods  as  any  common 
metal.  The  impurities  which  do  the  most  harm  are  sodium  and 
carbon.  When  the  aluminum  contains  carbon  an  electric  cur- 
rent is  at  once  set  up  when  a  suitable  liquid  is  applied.  In  such 
cases  after  water,  especially  if  it  be  saline,  has  stood  in  the  dish 
for  one  or  two  weeks,  the  surface  will  be  found  dotted  with  bril- 
liant rings,  and  on  scraping  off  the  aluminum  the  particle  of  car- 
bon will  be  disclosed.  If  a  strong  solution  of  salt  be  used,  the 
action  may  be  sufficient  to  cause  a  perforation  of  the  metal. 
The  aluminum  of  commerce,  unfortunately,  is  not  very  pure,  and 
it  is  for  this  reason  that  so  many  aluminum  dishes  have  shown 
a  rapid  deterioration.  The  French  troops  in  Madagascar  have 
been  supplied  with  15,000  sets  of  aluminum  dishes,  and,  when  a 
soldier  has  to  carry  his  kitchen  with  him,  the  importance  of 
lightness  is  not  to  be  despised.  But  even  granting  that  in  cook- 
ing in  aluminum  dishes  a  small  amount  of  alumina  is  introduced 
into  the  food,  it  has  not  been  shown  that  it  exercises  the  least 
harmful  action  on  the  digestion.     The  experience  of  two  men 


93^        H.  W.  WILEY.   SECOND  INTERNATIONAL 

may  be  cited  who  lived  for  a  year  on  food  prepared  exclusively 
in  aluminum  dishes  without  the  slightest  impairment  of  their 
health. 

In  the  afternoon  the  members  were  driven  in  carriages  to 
Gennevilliers,  where  they  inspected  the  irrigation  works,  lately- 
constructed  to  supplement  those  at  Asni^res  in  disposing  of  tbe 
sewage  of  Paris.  It  has  now  been  more  than  a  quarter  of  a 
century  since  the  city  of  Paris  has  been  using  its  sewage  for  irri- 
gation. The  fact  that  in  the  light  of  that  long  experiment  it  has 
recently  more  than  doubled  the  area  under  irrigation,  shows  that 
the  process  is  considered  a  practical  success.  The  sewage  of 
Paris  consists  mostly  of  the  water  used  for  washing  the  streets. 
Water-closets  are,  to  a  large  extent,  connected  with  vaults 
whose  contents  are  removed  by  means  of  wagons,  pumps  and 
closed  tanks  during  the  night.  The  sewage,  therefore,  is  not  so 
highly  polluted  nor  so  rich  in  fertilizing  materials  as  might  have 
been  supposed.  For  summers  like  the  present  one,  which  has 
been  excessively  dry,  the  disposal  of  the  sewage  by  irrigation  is 
easily  accomplished.  But  in  summers  of  excessive  rainfall  and 
in  the  winter,  the  problem  is  much  more  complex. 

We  first  were  shown  a  plan  on  a  large  chart  of  the  system  of 
sewers  and  the  distribution  of  the  waters.  Next  the  pumping^ 
house  was  visited  where  the  sewage  is  raised  to  a  sufficient 
height  to  carry  it  under  the  Seine  by  a  siphon  aqueduct  and  dis- 
tributed to  the  irrigated  fields.  The  fields  which  were  inspected 
are  only  a  part  of  the  vast  system  of  irrigation  now  in  operation. 
They  contain  799  hectares,  a  part  of  which  was  once  covered  by 
the  old  forest  of  St.  Germain.  The  city  of  Paris  spent  200,000,000 
francs  in  the  purchase  of  the  grounds,  the  building  of  the  aque- 
duct, erecting  the  pumping  machinery  and  building  the  irrigating 
canals.  The  work  on  the  aqueduct  of  Ach6res  was  commenced  in 
1893  and  the  whole  work  was  completed  in  1895.  The  aqueduct  is 
eleven  kilometers  long  and  is  three  meters  interior  diameter,  and 
it  crosses  the  Seine,  which  below  Paris  forms  a  loop,  twice.  For- 
tunately, the  soil,  forming  the  basin  of  the  Seine  in  this  locality, 
is  of  a  sandy  nature  and  permits  a  somewhat  rapid  filtration. 
A  clay  subsoil  would  render  the  whole  process  inapplicable. 
The  gardens,  though  only  two  years  old,  presented  a  scene  of 
almost  tropical  exuberance.  Many  dwarf  fruit  trees  were 
already  in  bearing  and  older  trees  showed  the  existence  of 
orchards  before  the  present  system  was  inaugurated. 

The  methods  of  irrigation  are  exactly  those  practiced  in 
the  arid  regions  of  the  United  States.  The  water  is  conducted 
in  furrows  on  the  surface  between  the  rows  of  growing  crops. 
Aside  from  a  slightly  unpleasant  odor  arising  from  the  sewage, 
there  is  nothing  in  the  scene  to  cause  the  observer  to  look  on 


CONGRESS  OF   APPLIED   CHEMISTRY.  937 

the  perfect  vegetables  and  flowers  with  suspicion.  In  harmony 
with  the  French  devotion  to  art,  the  borders  of  all  the  plots  are 
planted  in  roses  and  other  flowers  and  these,  at  the  time  of  our 
visit,  were  all  in  full  bloom,  recalling  in  their  floral  exuberance 
the  gardens  of  California.  Here,  as  a  result  of  the  applications 
of  science,  tj-phoid  fever  is  turned  into  turnips,  dysentery  dances 
in  the  dew  on  the  dahlias,  and  cholera  comes  chortling  as  cab- 
bage. The  one  unpleasant  reflection  is  found  in  the  fact  that 
this  extensive  harvest  is  sold  exclusively  in  the  Paris  markets 
and  one  can  hardly  avoid  thinking  in  the  restaurants  over  his 
cauliflower  and  artichoke  of  the  long  race  they  may  have  run  in 
the  aqueduct  of  Ach^res.  At  the  end  of  the  experimental  field, 
next  to  the  river,  the  sewage  which  has  passed  through  the  soil 
reappears  as  a  large  stream  of  pure  water,  absolutely  colorless 
and  bright.  Glasses  of  the  attractive  fluid  were  offered  the 
visitors,  many  of  whom,  unmindful  of  miasm  and  microbes, 
drank,  willing  martyrs  to  science  or  curiosity.  The  number 
of  micro-organisms,  which  is  many  millions  in  the  sewage,  is 
diminished  to  2,500  in  each  cubic  centimeter  of  the  filtered  water. 

Seventh  day,  Sunday,  August  2.  An  excursion  was  offered 
to  the  members  of  the  congress  on  Sunday  to  Comp^igne.  On 
reaching  the  station,  a  band  of  music  welcomed  the  excursion- 
ists. They  were  driven  through  the  gardens  and  forests  in  car- 
nages and  at  one  o'clock  a  breakfast  was  ser\'ed. 

Eighth  day,  Monday,  August  3.  In  section  i  papers  were 
presented  on  the  methods  of  determining  water  in  organic  vis- 
cous liquids,  by  Mr.  Pellet.  The  process  recommended  is  by 
absorption  with  pumice  stone  and  subsequent  drying,  first  at 
60°  to  80°  and  finally  at  100°.  Molasses  and  solids  should  first 
be  dissolved  in  water  to  promote  absorption  by  the  pumice.  A 
dr>'ing  dish  was  exhibited  with  a  circular  depression  in  the  cen- 
ter, into  which  the  body  is  weighed  and  mixed  with  enough 
water  to  make  it  flow  easily.  The  fragments  of  pumice  are 
placed  on  the  flat  bottom  of  the  dish,  exterior  to  the  depression, 
and  the  dissolved  mass  is  absorbed  by  the  pumice  on  inclining 
the  dish.  The  dish  and  cover  are  made  of  aluminum.  The 
diameter  of  the  dish  is  about  seven  and  its  depth  two  cm.  The 
composition  of  molasses  derived  from  the  sugar  cane  was  dis- 
cussed at  some  length.  Raffinose,  to  the  extent  of  three  per 
cent.,  has  been  detected  in  samples  of  cane  molasses  of  Egyptian 
origin.  The  reducing  sugar,  in  cane  molasses,  according  to  the 
statement  of  Pellet,  is  composed  solely  of  invert  sugar,  a  con- 
clusion which  he  has  reached  by  applying  the  method  of  esti- 
mating levulose  described  by  the  writer  in  this  Journal  a  few 
months  ago.^ 

1  Vol.  z8.  No.  X.  p.  81. 


938  H.  W.  WII^EY.      SECOND   INTERNATIONAL 

An  interesting  paper  by  Mr.  Herzfeld,  of  Berlin,  gave  a 
r6suni6  of  the  best  methods  of  separating  sugars  in  mixtures. 

In  section  3 ,  the  session  was  devoted  to  the  chemical  study  of  pro- 
cesses of  bread  making,  and  especially  to  the  methods  of  analysis 
of  moist  and  dry  gluten .  The  processes  presented  are  al most  iden- 
tical with  those  in  use  in  the  United  States.  Mr.  Lindet,  the 
president  of  the  congress,  read  a  communication  on  the  methods 
of  determining  starch  in  grains  and  flours,  in  which  the  separa- 
tion by  a  ferment  or  by  water  under  steam  pressure  was  recom- 
mended as  the  best.  These  are  the  processes  which  we  have 
preferred  for  several  years  in  the  agricultural  laboratory  at 
Washington. 

In  section  6,  papers  were  presented  on  gutta  percha,  paper, 
and  paint  used  to  prevent  corrosion  of  ship  bottoms. 

In  section  9  a  paper  on  the  analysis  of  wines  and  vinegar  was 
presented  by  Mr.  Leroy.  The  detection  of  glucose  in  beer  was 
discussed  by  Mr.  Pad6.  The  question  of  fermentation  and  the 
germicidal  methods  of  controlling  it  by  means  of  fluorides  was 
discussed  by  Mr.  Effront. 

An  interesting  exhibition  was  given  of  the  workings  of  the 
latest  form  of  bomb  calorimeter  for  the  determination  of  the  ther- 
mal equivalents  of  foods. 

Among  the  more  interesting  papers  presented  in  the  afternoon 
may  be  mentioned  one  by  Mr.  Fernback,  director  of  the  labora- 
tories of  the  Pasteur  Institute,  on  the  utilization  of  the  carbon 
dioxide  arising  from  fermentation,  in  section  2  ;  the  influence  of 
culture  on  the  chemical  and  physical  properties  of  the  soil,  by 
Mr.  Deherain,  in  section  4,  and  the  estimation  of  lactose  and 
sucrose  in  condensed  milks,  by  Mr.  F.  Dupont,  the  general  sec- 
retary of  the  congress,  in  section  5. 

In  the  evening  a  banquet  was  given  to  the  chairmen  of  com- 
mittees of  organization  and  to  the  delegates  of  foreign  govern- 
ments, in  theSalledesgrandesF^tesof  the  Grand  Hotel,  under  the 
presidency  of  Mr.  Cochery,  Minister  of  Finance,  at  which  nearh' 
500  sat  down.  An  orchestra  rendered  beautiful  music  during 
the  repast,  giving  among  other  things  the  national  airs  of  the 
various  governments  represented.  '*  Yankee  Doodle''  doubtless 
was  heard  with  equanimity,  but  one  can  imagine  the  feelings  of 
the  Frenchmen  present  when  '*Die  Wacht  am  Rhein"  was 
given.  Short  addresses  were  made  by  Mr.  Lindet,  the  presi- 
dent of  the  congress,  by  Mr.  Doremus,  on  the  part  of  the  foreign 
delegates,  and  a  rather  long  one  by  the  Minister,  who  greeted  the 
chemists  for  many  reasons,  and  especially,  he  said,  **  because  you 
are  the  precious  auxiliaries  of  my  department  in  promoting  the 
production  of  articles  that  can  be  taxed."  Mr.  Doremus  intro- 
duced his  address  by  quoting  one  of  the  inscriptions  on  the  statue 


CONGRESS  OP  APPLIED   CHEMISTRY.  939 

of  Danton  :  '*  Apres  le  pain  reducation  est  le  plus  grand  besoin 
du  peuple."  He  alluded  to  the  addresses  of  Berthelot,  Moissan, 
and  Lippmann,  as  illustrations  of  a  few  of  the  accomplishments 
of  applied  chemistry,  and  said  the  congress  had  shown  in  a 
striking  manner  the  necessity  of  a  close  alliance  between  applied 
and  research  science.  Pasteur  will  owe  his  immortality  to  the 
great  faculty  he  possessed  of  finding  a  practical  application  for 
his  discoveries.  He  concluded  as  follows :  **  Hon.  Minister  of 
Finance,  representing  the  French  Republic,  M.  Berthelot,  the* 
illnstrious  president  of  honor  of  this  congress,  M.  Wndet,  the 
president  of  the  congress,  M.  Dupont,  the  secretary,  I  wish  to 
thank  you  in  behalf  of  the  foreign  delegates,  for  the  hospitality, 
friendship,  and  good  fellowship  with  which  we  have  been 
received.  In  the  name  of  the  foreign  delegates,  I  propose  this 
toast,  the  French  Republic,  patron  not  only  of  this  congress,  but 
also  of  science,  art  and  industry,  the  mother  of  men  famous  in 
each  science,  but  especially,  in  chemistry.'* 

The  strangers  present  were  given  a  very  favorable  opportunity 
to  understand  the  heartiness  of  French  hospitality  and  the  excel- 
lence of  French  cooking.  We  might  learn  more  things  than 
good  cooking  from  a  French  banquet  and  among  others  the  art 
of  limiting  the  post  prandial  speeches.  At  ten  o'clock  the 
guests  left  the  table  and  assembled  in  the  grand  salon,  where 
coffee  and  liqueurs  were  served  and  an  hour  or  more  spent  in 
social  intercourse. 

Ninth  day,  Tuesday,  August  4.  I  have  already  used  so  much 
space  in  giving  even  a  few  of  the  details  of  the  congress  that  it 
is  not  advisable  to  mention  even  the  more  important  communi- 
cations presented  to-day.  Morning  sessions  only  were  held.  In 
the  afternoon  the  Conservatoire  des  Arts  et  Metiers  was 
visited,  where  the  congress  was  received  by  Mr.  Aim6  Girard, 
the  professor  of  applied  chemistry,  and  shown  through  the 
laboratories  and  museums.  In  the  latter  alone  are  enough 
objects  of  interest  to  employ  the  time  of  a  scientist  for  a  month 
for  a  careful  study.  We  can  only  mention /asiigta  rerum.  The 
pendulum  used  by  Foucault  in  his  classical  experiments  is  still 
swinging  and  showing  by  its  deflections  the  rotation  of  the  earth. 
All  the  important  apparatus  used  by  Lavoisier  is  collected  here. 
The  globes  employed  by  him  for  determining  the  composition  of 
water  are  remarkably  well  made  and  even  to-day  would  be 
regarded  as  entirely  convenient.  But  they  have  their  chief  value 
as  the  remains  of  those  era-making  investigations,  cut  short  by  the 
guillotine,  which  laid  the  foundation  of  modern  chemistry.  A 
wooden  wheel,  preserved  by  the  copper  sulphate  in  an  aban- 
doned copper  mine  since  the  fifth  century,  illustrates  in  a  most 


940  NOTE. 

Striking  way  one  of  the  best  methods  of  preventing  decay  in  rail- 
road ties.  The  standard  measures  of  all  nations  make  an  inter- 
esting collection,  but,  unfortunately,wewere  not  permitted  to  see 
the  original  meter,  which  is  preserved  from  view  in  the  vaults  of 
the  building.  In  the  courtyards  are  bronze  statues  of  Le  Blanc, 
who  made  the  fortunes  of  so  many  and  committed  suicide  by 
reason  of  his  own  poverty,  and  of  Boussingault,  the  contem- 
porary of  lyiebig  and  the  father  of  French  agricultural  chemis- 
try. A  photographic  view  of  the  congress  was  made  on  the 
steps  of  the  \yest  facade  of  the  building. 

Tenth  day,  Wednesday,  August  5.  In  the  morning  the  sec- 
tions held  their  final  sessions  for  hearing  papers  and  discussions. 
In  the  afternoon  the  closing  meeting  of  the  congress  was  held  in 
the  grand  amphitheatre  of  the  Sorbonne  under  the  presidency  of 
Mr.  Henri  Boucher,  Minister  of  Commerce  and  Industry. 
Addresses  were  made  by  Mr.  Lindet  and  the  Minister  and  a 
report  of  the  proceedings  of  the  cot^ress  presented  by  the  sec- 
retary, Mr.  Dupont.  Turin  and  Vienna  were  placed  in  nomina- 
tion as  the  places  of  meeting  of  the  congress  in  1898.  Vienna 
was  selected  by  a  large  majority.  An  invitation  was  extended 
by  Mr.  Lindet  to  hold  the  congress  of  1900  in  Paris  during  the 
World^s  Exhibition,  and  that  invitation  will  doubtless  be 
accepted  at  Vienna. 

After  the  adjournment  of  the  meeting,  the  new  laboratories  of 
organic  chemistry,  constructed  by  Friedel,  were  inspected  by- 
Mr.  Doremus  and  myself.  In  the  confusion  of  the  summer 
cleaning,  we  could  hardly  form  any  favorable  judgment  of  their 
points  of  excellence.  The  ultra  impressionist  painting  of  Para- 
dise Lost,  a  mural  ornamentation  back  of  the  professor's  lecture 
table,  was  the  most  original  and  inexplicable  feature  of  the  lab- 
oratory. 

Paris,  August  lo.  1896. 


NOTE. 

The  fourteenth  annual  report  of  the  Committee  on  Indexing 
Chemical  Literature  was  presented  to  the  American  Association 
for  the  Advancement  of  Science  at  the  Buffalo  meeting,  August 
24.  A  large  amount  of  work  has  been  done  in  this  field  during 
the  year.  The  committee  is  an  active  one  and  has  done  a  val- 
uable work  in  encouraging  and  recording  biographical  under- 
takings. Copies  of  the  report  may  be  obtained  of  the  chairman, 
Dr.  H.  Carrington  Bolton,  Cosmos  Club,  Washington,  D.  C. 


Vol.  XVIII.  [November,  1896.]  No.  11. 


THE  JOURNAL 


OF  THE 


AMERICAN  CHEMICAL  SOCIETY. 


A  NEW  FORM  OF  POTASH  BULB.* 

By  M.  Gomberg. 

Received  August  aS,  1896. 

THE  potash  bulb  most  frequently  used  at  present  in  elemen- 
tary organic  analysis  is  that  known  as  Geissler*s  bulb. 
"While  neat  and  compact,  it  still  has  the  same  drawback  as  pos- 
sessed by  other  forms  of  potash  bulbs;  namely,  that  even  with 
tlie  most  careful  handling  it  is  not  unfrequently  broken.  Some 
two  years  ago  I  drew  up  a  design  for  a  different  form  of  bulb, 
^wherein  all  the  connections  should  be  enclosed.  Several  attempts 
to  have  it  made  in  this  country  have  proven  unsuccessful.  The 
design  was  then  sent  to  Greiner  &  Friedrichs,  of  Thtiringen, 
and  I  have  recently  received  from  them  samples  of  such  bulbs. 
Meanwhile,  it  came  to  my  notice  that  a  bulb  based  on  similar 
principles  has  been  put  upon  the  market  by  Bender  &  Hobein, 
of  Miinchen.  A  comparison  of  the  two  bulbs  shows  them,  how- 
ever, to  be  sufficient!^  different  to  justify  me  in  presenting  a 
description  of  the  one  made  according  to  my  design,  without 
claiming  priority  as  to  the  principle  of  construction. 

The  arrangement  and  working  of  the  bulb  will  appear  clear 
from  the  subjoined  diagram,  which  presents  the  apparatus  re- 
duced to  one-half  its  actual  size. 

The  potash  bulb  is  divided  into  three  compartments, /f,  ^ 
and  C,  B  and  C  contain  the  potash  solution  for  the  ab- 
sorption of  the  carbon  dioxide,  while  A  serves  as  a  safety  res- 
ervoir in  case  of  backward  suction.  The  bulb  is  filled  by  dipping 
a  into  the  solution,  and  applying  suction  at  ^,  until  the  two  com- 

I  Communicated  by  A.  B.  Prescott.  Read  at  the  meeting  of  the  American  Chemical 
society,  Buffalo,  N.  Y.,  August  32,  1896. 


942 


CHARLBS  BASKERVILLB.      REDUCTION  OP 


partments  B  and  C  contaia  as 
much  of  the  solution  as  would 
completely  fill  A^  which  is 
about  thirty-five  to  forty  grams 
of  a  2  :  3  solution.  />,  which 
is  fastened  to  the  bulb  by 
means  of  a  ground-glass  joint, 
contains  solid  potassium  hy- 
droxide, or  soda-lime,  sup- 
ported by  a  plug  of  glass-wool. 
The  liquids  in^  and  Ccan  be 
easily  mixed  when  desired, 
by  applying  suction  at  a. 
The  bulb,  when  filled  and 
ready  for  use,  weighs  from 
sixty-five  to  seventy  grams, 
and  undoubtedly  can  be  made 
even  much  lighter. 

The  total  number  of  compartments  is  thus  reduced  from  five 
in  Geissler's  form  to  three  in  the  form  here  presented,  while  the 
absorbing  chambers  are  reduced  only  from  three  to  two.  The 
construction  of  the  bulb  is  such  that  C  can  never  get  overfilled 
by  the  solution  from  B. 

This  form  of  a  potash  bulb  possesses  the  advantages  first, 
that  it  can  be  easily  handled  and  wiped,  presenting  the  out- 
side surface  of  an  ordinary  small  flaslc,  and  second,  that  it 
can  be  set  without  any  support,  and  can  be  weighed  without 
suspending  it  if  so  desired. 

I  wish  to  express  my  thanks  to  the  firm  of  Greiner  &  Fried- 
richs,  of  Thiiringen,  who  have  kindly  made  the  bulb  for  me  in 
a  most  satisfactory  manner. 

CBBMICAI.  LABOKATORV,  UNrVBRSITY  OP  MICHIGAN. 


REDUCTION  OF  CONCENTRATED  SULPHURIC  ACID  BY 

COPPER. 


I 


Bt  Charlbs  Baskervillb. 

Received  August  97,  1890. 


KCCVITCU    August    S7(    I«9P. 

N  a  previous  communication*  the  writer  noted  that  copper  was 
acted  upon  by  concentrated  sulphuric  acid  (i.84sp.  gr.)  not 

1  This  Journal.  17,  90. 


r- 


CONCENTRATBD  SULPHURIC  ACID  BY  COPPER.  943 

only  at  the  ordinary  temperatures  of  the  air »  20^-30'*  C . ,  but  at  zero 
as  well.  Andrews'  states  that  the  assertion  is  ihcorrect  and  that 
it  does  not  occur  until  the  temperature  86^  C.  has  been  reached, 
or  a  point  above  the  dissociation  temperature  of  the  concentra- 
ted sulphuric  acid  I  67*  C,  according  to  him.  Andrews  further 
says  that  the  author's  statements  were  based  "not  upon  any 
demonstrations  of  the  formation  of  sulphurous  acid,  but  solely 
on  the  formation  of  copper  sulphate,"  which,  he  says,  occurs 
only  '*  in  consequence  of  the  presence  of  the  air.**  It  is  to  be 
regretted  that  Dr.  Andrews  did  not  note  carefully  the  statements 
of  the  author  in  his  previous  communication,  as  no  reason  what- 
ever exists  for  any  such  conclusions,  because  it  was  distinctly 
stated  that  not  only  the  copper  as  sulphate,  but  as  sulphide  was 
determined,  as  well  as  sulphurous  acid,  and  moreover,  that  the 
experiments  were  carried  out  when  the  air  had  been  replaced  by 
a  neutral  gas,  either  hydrogen  or  carbon  dioxide. 

The  author,  although  confident  of  the  correctness  of  his  for- 
mer statement,  carried  out  further  experiments  to  correct  the 
error,  if  committed  or  to  establish,  beyond  question,  the  fact  that 
concentrated  sulphuric  acid  of  1.84  sp.  gr.  is  reduced  b}'  copper 
below  86®  C,  the  limit  postftve/y  .set  by  Dr.  Andrews. 

The  fact  that  these  experiments  but  confirmed  the  former 
statement  of  the  author  allows  the  incorporation  of  the  results  in 
this  paper. 

As  far  back  as  1834  the  fact  that  copper  is  acted  upon  by 
concentrated  sulphuric  acid  at  ordinary  temperatures,  if  suffi- 
cient time  be  given,  was  made  known  by  Barruel.*  Calvert  and 
Johnson,'  however,  failed  to  obtain  any  action  below  130®  C.  and 
considered  that  none  took  place.  Pickering/  however,  stated 
that  "  sulphuric .  acid  attacks  copper  at  all  temperatures  from 
19**  C,  (and  probably  even  still  lower)  upwards.*' 

Ftrs/  Experiment. — Copper  ribbon  in  strips,  i  x  3-4  cm.,  was 
submerged  in  concentrated  sulphuric  acid  in  a  clean  glass  stop- 
pered flask  for  a  month.  At  the  end  of  that  time  not  only  were 
there  white  crystals  of  anhydrous  copper  sulphate  clinging  to 

1  This  Journal,  i8»  353. 
sy.  depkarm.^  90,  13, 18J4. 
•y.  Chem.  See.,  19,  438,  iS66. 
«y.  Oum.  Soc.,  Trans.,  1878,  115. 


944  CHARLES  BASKERVILLE.      REDUCTION  OP 

the  sides  of  the  containing  yessel,  but  there  was  a  very  appre- 
ciable amount  of  brownish  black  cuprous  sulphide  and  sulphur 
dioxide  was  easily  detected  by  its  strong  odor  when  the  vessel 
was  opened. 

Andrews'  states  *'  that  in  the  presence  of  air  sulphuric  acid  is 
attacked  by  copper  at  ordinary  temperatures,  but  without  reduc- 
tion of  the  acid.  The  reaction  must  take  place  according  to  the 
equation, 

2CU  +  O,  +  2H,S0,  =  2CuSO,+  2H.O.'' 

Formerly  the  author*  stated  that  the  presence  of  the  oxygen 
of  the  air  when  it  comes  into  contact  with  the  copper  in  the  acid 
has  great  influence  on  the  reaction.  Fifty  years  ago,  Maumen^' 
proved  that  when  a  current  of  oxygen  gas  was  passed  through 
the  boiling  acid,  the  amount  of  insoluble  residue,  e,  g,^  cuprous 
sulphide,  was  diminished,  that  is,  less  than  there  would  be 
formed  if  the  experiment  were  carried  out  with  a  current  of  car- 
bon dioxide.  The  copper  must  be  directly  exposed  to  the  oxy- 
gen by  only  partial  submersion  or  the  bubbling  of  the  air  against 
or  around  the  submerged  copper ;  but  the  air  in  a  confined 
space,  not  at  all  in  contact  with  the  copper,  but  separated  by  a 
thick  layer  of  concentrated  sulphuric  acid,  has  little  or.no  effect. 

Yet  grant  that  the  oxygen  of  the  air  (volume  of  air  about  200 
cc.)  confined  in  the  flask,  had  been  utilized  in  the  formation  of 
the  copper  sulphate  produced.  According  to  the  formula  given 
above,  the  oxygen  would  be  absorbed  and  no  corresponding 
amount  of  any  other  gas  would  be  eliminated;  consequently 
there  should  be  a  greater  external  pressure  at  the  close  than  at 
the  beginning  of  the  experiment.  When  the  smoothly  fitting 
glass  stopper  was  removed,  not  only  no  extra  internal  pressure 
was  noticed,  but  in  fact  a  pressure  from  within.  This  was  evi- 
dently produced  by  the  sulphur  dioxide  generated.  The  sulphur 
dioxide  was  swept  out  by  a  current  of  air  through  a  dilute  solu- 
tion of  potassium  permanganate,  which  was  quickly  bleached. 
The  presence  of  sulphur  dioxide  was  further  proven  by  the  addi- 
tion of  barium  chloride  to  the  bleached  potassium  permanganate 

^  This  Journal,  i8,  252. 

a  Ibid,  17.  912. 

^  Attn,  chim,  phys.,  1B46,  [3I,  il,  311. 


CONCENTRATED   SULPHURIC   ACID   BY   COPPER.  945 

solution.     Nor  does  the  formula  given  above  account-  for  the 
cuprous  sulphide  which  is  always  produced. 

Second  Experiment. — Realizing  the  possibility  of  some  organic 
matter  or  dust  remaining  in  the  flask,  although  it  had  been  care- 
fully cleansed,  the  first  experiment  was  repeated  with  the  great- 
est precaution  to  ensure  the  absence  of  dust.  The  flask  was  scoured 
with  boiling  concentrated  pure  sulphuric  acid  containing  potas- 
sium bichromate  and  carefully  cleansed  with  distilled  water. 
The  last  traces  of  water  were  removed  by  four  subsequent  wash- 
ings with  the  same  kind  of  concentrated  acid  used  throughout 
the  experiments.  The  experiment  was  carried  out  in  the  same 
manner  as  the  first,  the  same  results  l)eing  obtained. 

A  blank  experiment  was  carried  out  at  the  same  time.  The 
flask  was  rendered  dust  free  in  the  manner  just  mentioned 
and  fifty  cc.  of  the  same  acid  allowed  to  remain  in  the  flask  for 
six  months.  At  the  end  of  that  period  not  a  trace  of  sulphur 
dioxide  could  be  detected  in  the  blank,  therefore  the  sulphur 
dioxide  produced  when  the  copper  was  inserted  could  not  be  due 
to  the  reduction  of  the  sulphuric  acid  by  an  extraneous  sub- 
stance, but  solely  by  the  copper.  The  conclusion  is  that  sul- 
phuric acid  is  reduced  by  copper  when  air  is  present  at  the  ordi- 
nary temperatures,  2o**-30°  C. 

Third  Experiment,— Kxi  ordinary  Kjeldahl  digestion  flask  was 
made  dust  free  by  the  treatment  noted  above.  100  cc.  sul- 
phuric acid,  1.84  sp.  gr.,  were  placed  therein  and  clean  dry 
strips  of  copper  ribbon  were  completely  submerged  in  the  acid. 
Now  air-free  carbon  dioxide  was  passed  through  the  flask  for 
three  hours.  The  inlet  tube  was  just  dipped  into  the  acid. 
The  flask  was  then  attached  to  a  suction  pump,  with  a  sulphuric 
acid  dr>'ing  flask  intervening  to  prevent  a  possible  return  flow  of 
gas  or  air  which  might  carry  moisture  or  dust  into  the  flask. 
The  flask  was  exhausted  of  the  carbon  dioxide  present  for  three 
hours  at  a  pressure  of  150  mm.  It  was  then  sealed  with  the 
blast  lamp  and  placed  aside  in  a  darkened  cupboard.  Obser\'a- 
tions  were  made  every  few  days  to  note  any  reaction  taking 
place.  Within  two  days  it  could  be  easilj'  seen  that  copper  sul- 
phate had  been  formed  and  the  liquid  was  somewhat  clouded  by 
very  finely  divided   suspended   cuprous   sulphide.     Continued 


94^         REDUCTION  OP  CONCBNTRATBD  SULPHURIC  ACID. 

observations  extending  over  a  period  of  seven  weeks  showed 
only  an  increase  in  the  amounts  of  both  of  these  substances. 
The  temperature  of  the  cupboard  had  at  no  time  risen  above  20* 
C,  and  was  for  most  of  the  time  much  lower.  The  flask  was 
then  opened  as  any  other  sealed  tube,  and  instead  of  an  external 
pressure  inward,  which  had  been  sufficient  to  heavily  dent  the 
tube  in  sealing,  there  was  a  strong  internal  pressure  outward. 
The  gas  evolved  was  sulphur  dioxide,  easily  detected  by  its 
strong  odor  and  bleaching  effect  upon  a  dilute  solution  of  potas- 
sium permanganate.  The  sulphuric  acid  produced  by  the  oxi- 
dation of  the  sulphur  dioxide  by  the  permanganate  was  precipi- 
tated by  barium  chloride.  All  solutions  and  apparatus  were 
proven  to  be  free  from  traces  of  sulphur  dioxide  and  sulphuric 
acid  by  a  blank  experiment. 

Coficlusian, — Concentrated  sulphuric  acid,  1.84  sp.  gr.,  is 
reduced  by  copper  when  air  is  absent  and  at  temperatures  far 
below  86^  C,  in  fact  at  the  ordinary  atmospheric  temperatures 
with  the  formation  of  copper  sulphate  and  cuprous  sulphide  and 
the  production  of  sulphur  dioxide. 

Finally. — Apparatus  similar  to  that  made  use  of  by  Andrews* 
was  employed  with  the  modification  of  having  three  drying 
flasks  containing  concentrated  sulphuric  acid  instead  of  one,  and 
a  Meyer  absorption  tube  was  substituted  for  a  single  small  flask. 
These  served  merely  as  extra  precautions  against  dust  and 
insured  an  intimate  mixing  of  the  outgoing  gases  with  the  per- 
manganate. Within  twelve  hours  the  permanganate  was 
bleached.  Andrews'  experiment  lasted  only  fifteen  minutes. 
The  presence  of  the  sulphur  dioxide  produced  was  easily 
detected  by  the  odor  when  the  apparatus  was  opened,  and  in  the 
bleached  permanganate  solution  by  barium  chloride.  Copper 
sulphate  and  cuprous  sulphide  were  also  formed. 

Concentrated  Sulphuric  Acid  is  Acted  upon  by  Copper  at  Zero, — 
Quantitative  experiments  were  carried  out  by  the  author  when 
the  concentrated  sulphuric  acid  in  which  the  copper  was  sub- 
merged was  practically  at  zero.'  In  stating  the  results,  how- 
ever, the  author  gave  the  temperature  as  *  *  o'-io'  C. "     The  flask 

1  This  Journal,  xl,  251. 
a  Ibid.  17,  908. 


SEPARATION   OP   THORIUM.  $47 

containing  the  acid  was  buried  in  an  ice-bath  and  the  tempera- 
ture of  the  liquid  noted  by  a  thermometer  inserted  through  a 
rubber  stopper.  The  apparatus  was  air-tight.  A  stream  of 
hydrogen  gas  was  continued  through  the  apparatus  in  one 
experiment  for  six  Weeks  and  in  another  two  months.  On  two 
occasions  when  the  ice  in  the  bath  had  melted  in  going  over 
Sunday,  the  temperature  rose  to  lo**  C.  The  temperature  could 
not  possibly  have  remained  that  high  for  over  twelve  hours,  which 
would  have  had  small  influence  when  the  experiments  lasted 
through  a  number  of  days.  The  temperature  was  reported 
o'^-io^  C,  however.  Not  only  copper  sulphate,  but  cuprous  sul- 
phide and  sulphur  dioxide  had  also  formed.  Copper,  therefore, 
decomposes  concentrated  sulphuric  acid  (sp.  gr.  1.84)  practi- 
cally at  zero. 

From  my  own  experiments  and  from  experiments  performed 
with  apparatus  similar  to  that  used  by  Andrews  and  under  the 
same  conditions,  except  with  regard  to  the  important  element, 
time,  which  consideration  is  necessary  for  all  chemical  reac- 
tions, the  author  must  adhere  to  his  former  statement. 

UifrvBRSiTY  OF  North  Carolina. 


THE  SEPARATION  OF  THORIUfl  FROil  THE  OTHER  RARE 
EARTHS  BY  MEANS  OF  POTASSIUfl  TRINITRIDE. 

By  h.  M.  Dennis. 

Received  Scptenber  4.  1896. 

SOME  time  ago  the  author  and  F.  L.  Kortright*  briefly 
described  the  action  of  a  solution  of  potassium  trinitride 
upon  a  neutral  solution  of  the  rare  earths.  It  was  found  at  that 
time  that  the  flocculent  precipitate  which  is  produced  was  most 
probably  thorium  hydroxide,  but  our  supply  of  potassium  trini- 
tride having  been  exhausted  it  was  impossible  to  further  inves- 
tigate the  reaction  or  ascertain  the  completeness  of  the  separa- 
tion. The  immediate  continuation  of  the  work  was  prevented 
by  unexpected  difficulties  which  were  encountered  in  the  prepa- 
ration of  pure  hydronitric  acid  on  a  large  scale.  These  difficul- 
ties have  since  been  removed,  and  it  has  been  possible  to  prepare 
an  amount  of  the  reagent  sufficient  for  the  investigation 
described  below. 

>  Xtschr.  anoirg.  Ckem.,  6,  35 ;  Am.  Chem.J.,  x6,  79. 


94^  L.    M.    DENNIS. 

The  solution  of  potassium  trinitride  which  was  used  was  pre- 
pared by  careful!}'  neutralizing  a  dilute  solution  of  hydronitric 
acid  with  a  dilute  solution  of  pure  caustic  potash  and  then  add- 
ing hydronitric  acid  sufficient  to  give  to  the  solution  a  distinctly 
acid  reaction.  The  solution  first  employed  contained  about 
three  and  two-tenths  grams  of  potassium  trinitride  to  the  liter. 

Before  studying  the  separation  of  thorium  from  the  other  rare 
earths,  the  reaction  between  potassium  trinitride  and  pure  tho- 
rium chloride  was  first  investigated.  The  thorium  employed 
was  from  a  sample  of  thorium  oxalate,  which  had  been  very 
kindly  presented  to  me  by  Dr.  Theodor  Schuchardt,  of  Goerlitz. 
It  was  found  to  be  of  a  very  high  grade  of  purit}',  but  to  guard 
against  the  possible  presence  of  other  rare  earths,  the  oxalate 
was  converted  to  the  oxide  b}'  ignition,  treated  with  concentra- 
ted sulphuric  acid,  the  anhydrous  sulphate  dissolved  in  distilled 
water  at  a  temperature  of  o°,  and  this  solution  was  precipi- 
tated with  pure  oxalic  acid.  The  precipitated  thorium  oxalate 
was  thoroughly  washed  with  hot  water  containing  one  per  cent. 
hydrochloric  acid,  and  was  then  dropped  into  a  hot,  concen- 
trated solution  of  ammonium  oxalate.  It  dissolved  completely 
and  no  precipitate  formed  when  the  solution  was  diluted  and 
cooled.  From  this  solution  the  thorium  was  again  precipitated 
as  oxalate  by  means  of  strong  hydrochloric  acid  and  was  then 
brought  into  solution  as  thorium  sulphate  in  the  manner 
described  above.  It  was  then  precipitated  by  ammonium 
hj-droxide  and  the  precipitate  thoroughly  washed  with  water. 
The  thorium  hydroxide  was  then  dissolved  in  hydrochloric  acid, 
ammonium  hydroxide  was  added  until  a  faint  but  permanent 
precipitate  remained,  and  this  was  then  removed  by  filtration. 
There  was  thus  obtained  a  neutral  solution  of  thorium  chloride 
containing  a  very  small  amount  of  ammonium  chloride. 

The  strength  of  this  solution  of  thorium  chloride  was  ascer- 
tained  by  precipitating  portions  of  ten  cc.  each  with  ammonium 
hydroxide,  filtering,  washing,  igniting,  and  weighing  as  ThO,. 
Two  determinations  gave,  for  thorium  oxide  in  ten  cc.,  0.0591 
gram  and  0.0595  gram.  The  mean  of  these  results  is  equivalent 
to  0.00521  gram  thorium  in  one  cc. 

Upon  adding  to  this  thorium  solution  a  few  cc.  of  the  solution 


SEPARATION   OP  THORIUM.  949 

of  potassium  trinitride,  the  precipitate  which,  in  the  previous 
work  with  Dr.  Kortright,  had  formed  at  once,  failed  to  appear  ; 
upon  heating  the  solution  to  boiling,  however,  there  was  quickly 
formed  a  white,  flocculent  precipitate,  closely  resembling  in 
appearance  aluminum  hydroxide,  but  settling  rapidly  when  the 
flame  was  removed.  In  the  first  determinations  the  solution 
was  boiled  for  five  minutes,  but  it  was  later  found  that  boiling 
for  one  minute  is  sufficient.  During  the  boiling  the  odor  of 
hydronitric  acid  was  distinctly  noticeable.  The  precipitate  was 
washed  by  decantation  with  hot  water,  transferred  to  the  filter, 
ignited,  and  weighed  as  ThO,.  Twenty  cc.  of  thorium  chlo- 
ride, containing,  according  to  the  determination  with  ammo- 
nium hydroxide,  0.1186  gram  thorium  dioxide  gave,  by  precipi- 
tation with  potassium  trinitride,  0.1183  gram  thorium  dioxide, 
equivalent  to  0.00520  gram  thorium  in  one  cc.  instead  of  0.00521 
as  obtained  with  ammonia. 

It  is  apparent,  therefore,  that  thorium  can  be  quantitatively 
precipitated  by  potassium  trinitride. 

The  pre\^ious  work  of  Dr.  Kortright  showed  that  the  thorium 
is  probably  precipitated  as  the  hydroxide,  but  the  tendency  of 
the  precipitate  to  absorb  carbon  dioxide  rendered  the  analyses 
unsatisfactory.  If,  however,  the  thorium  is  precipitated  as  the 
hydroxide,  then  all  of  the  hydronitric  acid  of  the  potassium  salt 
first  added  must  reappear  in  the  filtrate  from  the  thorium 
hydroxide  and  in  the  gas  evolved  during  the  boiling.  To  ascer- 
tain whether  this  took  place  the  precipitation  was  made  in  a 
round  bottomed  flask.  In  the  neck  of  the  flask  there  was 
inserted  a  two  hole  rubber  stopper,  through  one  opening  of 
which  a  current  of  purified  air  was  admitted,  the  other  opening 
carrying  an  upright  condenser.  The  condenser  was  connected 
at  the  upper  end  with  two  absorption  vessels  containing  neutral 
silver  nitrate  solution.  As  the  hydronitric  acid  was  to  be  deter- 
mined by  precipitation  with  silver  nitrate,  a  neutral  thorium 
nitrate  solution,  containing  0.0075  gram  thorium  in  onecc,  was 
substituted  for  the  thorium  chloride.  The  thorium  nitrate  solu- 
tion was  placed  in  the  flask,  potassium  trinitride  was  added,  and 
after  starting  a  current  of  air  through  the  apparatus,  the  con- 
tents of  the  flask  was  heated  to  boiling  and  kept  boiling  for  two 


950  h.    M.    DENNIS. 

minutes.  Soon  after  the  heating  began  a  white  precipitate  of 
silver  hydronitride  formed  in  the  first  absorption  flask  contain- 
ing the  silver  nitrate ;  by  the  time  the  reaction  was  complete 
this  precipitate  had  become  quite  voluminous.  The  absorption 
of  the  gas  by  silver  nitrate  seems  to  be  both  rapid  and  complete^ 
for  nothing  more  than  a  slight  opalescence  ever  appeared  in  the 
second  absorption  flask.  After  the  apparatus  had  become  cool 
the  thorium  hydroxide  was  filtered  off  and  the  filtrate  was  pre- 
cipitated by  silver  nitrate.  The  silver  trinitride  thus  obtained, 
together  with  that  in  the  absorption  flasks,  was  washed  by 
decantation  with  cold  water,  the  washings  being  passed  through 
a  hardened  filter.  When  the  wash  water  gave  no  further  reac- 
tion for  silver  the  funnel  with  the  filter  was  placed  in  the  neck 
of  the  flask  containing  the  main  part  of  the  precipitate,  and 
quite  dilute,  hot  nitric  acid  was  poured  upon  the  paper.  The 
silver  trinitride  on  the  paper  dissolves  almost  immediately. 
After  washing  the  paper  with  water,  the  funnel  was  removed 
and  the  contents  of  the  flask  was  boiled  until  all  of  the  silver  tri- 
nitride had  dissolved.  The  silver  was  then  precipitated  by 
hydrochloric  acid  and  weighed  as  silver  chloride.  Ten  cc.  of 
thorium  nitrate  and  ten  cc.  of  potassium  trinitride  were  used. 
The  silver  chloride  resulting  weighed  o.  1447  gram,  equivalent 
to  0.0434  gram  hydrpnitric  dcid.  The  strength  of  the  potas- 
sium hydronitride,  which  was  a  different  solution  from  the  one 
first  employed,  was  then  determined  in  the  same  manner. 

5  cc.  gave  0.0744  AgCl  =  0.02232  HN,, 
5  cc.  gave  0.0745  AgCl  =  0.02235  HN,. 

Using  the  mean  of  these  results,  it  appears  that  0.0446  gram 
of  h3'dronitric  acid  was  used  in  the  precipitation  of  the  thorium 
nitrate,  of  which  0.0434  gram  was  recovered  from  the  filtrate 
and  distillate.  That  this  latter  result  is  somewhat  low  is  doubt- 
less due  to  the  loss  of  hydronitric  acid  by  volatilization  during 
the  filtration  of  the  liquid  in  the  flask.  These  results,  together 
with  those  given  in  the  preceding  article  •  already  referred  to, 
enable  us  to  represent  the  reaction  by  the  equation 

Th(NO,),  +  4KN.  +  4H.O  =  Th(OH),  +  4KNO,  +  4HN.. 

This  reaction  is  interesting  not  only  because  of  the  quantita- 


SEPARATION  OP  THORIUM.  95 1 

tive  precipitation  of  thorium  by  this  means,  but  also  because  6f 
the  peculiar  behavior  of  the  potassium  hydronitride.  As  Ost- 
wald  has  stated,  hydronitric  acid  is  but  slightly  stronger  than 
glacial  acetic  acid,  and  the  above  equation  reminds  one  of  the 
behavior  of  acetates  towards  ferric  iron,  the  solution  of  ferric 
acetate  being  fairly  stable  in  the  cold,  but  breaking  down  upon 
heating,  into  acetic  acid  and  ferric  hydroxide. 

The  experiments  detailed  below  were  then  made  to  ascertain 
whether  thorium  could  be  quantitatively  separated  from  the 
other  rare  earths  by  means  of  the  above  reaction.  A  neutral 
solution  of  pure  lanthanum  chloride  was  first  prepared  and  its 
strength  determined  by  precipitating  with  ammonium  hydroxide 
and  weighing  the  lanthanum  as  La^O,.  The  solution  contained 
0.00431  gram  lanthanum  in  one  cc.  This  solution  gave  no  pre- 
cipitate when  boiled  for  some  minutes  with  potassium  trinitride. 
Fifteen  cc.  of  this  solution  and  fifteen  cc.  of  the  thorium  chloride 
solution  were  placed  in  an  Erlenmeyer  flask,  twenty-five  cc.  of 
potassium  trinitride  (three  and  two-tenths  grams  to  the  liter) 
was  added  and  the  solution  was  boiled  'or  one  minute.  The 
precipitate  was  filtered  ofiF  and  washed  with  hot  water,  ignited, 
and  weighed.  To  the  filtrate  five  cc.  more  of  potassium  trini- 
tride was  added  and  the  solution  boiled  for  two  minutes. 
No  further  precipitation  resulted.  The  solution  was  then  pre- 
cipitated with  ammonia  and  the  lanthanum  weighed  as  the 
oxide.     The  results  were  : 

Taken.  Pound. 

Thorinm 0.0781  0.0777 

Lanthanum 0.0646  0.0642 

A  mixture  of  the  rare  earths  in  Brazilian  monazite  was  then 
freed  from  thorium  by  repeatedly  digesting  the  mixed  oxalates 
with  a  hot,  concentrated  solution  of  ammonium  oxalate.  The 
residual  oxalates  were  then  transformed  into  chlorides  and  dis- 
solved in  water.  The  solution  showed  the  pink  color  and 
absorption  bands  of  didymium  and  gave  a  strong  reaction  for 
cerium  when  treated  with  hydrogen  peroxide  and  ammonia. 
When  boiled  with  potassium  trinitride  it  gave  a  very  faint  pre- 
cipitate which  was  filtered  off.  By  precipitation  with  ammonia 
this  solution  of  cerium,  lanthanum,  didymium,  etc.,  free  froni 


952  SEPARATION   OF  THORIUM. 

thorium,  was  found  to  contain  0.0166  gram  of  the  mixed  oxides 

in  one  cc.  The  precipitation  was  made  as  in  the  separation 
from  lanthanum  and  an  excess  of  potassium  trinitride  was  used 
in  each  case. 

Taken.  Found. 

I.  Thorium... 0.1300  0.1294 

Ce,  t^a,  Di  oxides 0.0332 

II.  Thorium 0.0785  0.0783 

Ce,  La,  Di  oxides 0.0830  .... 

III.  Thorium 0.0535  0.0526 

Ce,  La,  Di  oxides 0.2490  .... 

IV.  Thorium 0.0535  0I0531 

Ce,  La,  Di  oxides 0.2490  .... 

V.  Thorium.,  r 0.0550  0.0541 

Ce,  La,  Di  oxides 0.4980  . .  • . 

VI.  Thorium 0.0555  •  0.0550 

Ce,  La»  Di  oxides 0.5810  .... 

VII.  Thorium 0.0570  0.0558 

Ce,  La,  Di  oxides 0.8300 


•   a 


The  recovery  of  the  thorium  is  in  all  cases  fairly  exact  and  the 
variation  in  the  relative  amounts  of  thorium  and  the  other  earths 
does  not  influence  the  sharpness  of  the  separation.  That  tho- 
rium alone  is  precipitated  by  potassium  trinitride  is  to  be 
explained  by  its  weak  basicity.  It  is  the  weakest  base  in  the 
whole  group  of  the  rare  earths  with  the  possible  exception  of 
cerium  in  the  eerie  condition,  and  this  higher  form  of  cerium  is 
probably  incapable  of  existence  in  the  presence  of  hydronitric 
acid. 

We  have,  then,  in  potassium  trinitride  a  reagent  which  can  be 
used  both  for  the  qualitative  detection  of  thorium  and  for  its 
quantitative  determination  either  alone  or  in  the  presence  of 
other  rare  earths.  So  far  as  the  author  is  aware,  this  is  the 
only  method  as  yet  devised  by  which  one  of  these  earths  can  be 
quickly  and  accurately  separated  from  the  others,  and  that  in  a 
single  simple  operation. 

Cornell  University,  Augrust,  1896. 


NOTES  ON  REIN§CH'S  TEST  PORARSENIC  AND  ANTIMONY. 

By  Jas.  I«bwx6  Howe  and  Paul  S.  Mertin s. 

Received  September  la.  1896. 

THAT  Reinsch's  test  for  arsenic  possesses,  in  point  of  con- 
venience, marked  advantages  over  that  of  Marsh,  is  gen- 
erally acknowledged,  but  it  has  been  questioned  both  as  to  deli- 
cacy and  as  to  accuracy  in  distinguishing  between  arsenic  and 
antimony.  As  to  the  former  point,  Reinsch,  in  his  second  arti- 
cle on  the  test,*  states  that  arsenic  may  be  detected  in  a  solution 
of  one  part  per  million.  In  his  original  description*  of  the  test 
he  placed  the  accuracy  about  one-third  of  this.  Our  own 
experiments  show  that  this  accuracy  is  not  overstated.  The 
fact  that  arsenious  oxide  and  antimonous  oxide  (Sb.O.)  are 
isomorphous  in  their  crystallization  has  led  to  the  conjecture  that 
antimonous  oxide  subliming  from  the  copper  in  the  closed  tube 
might  appear  in  the  brilliant  octahedra,  characteristic  of  arsenic 
in  the  test. 
Experiments  bearing  on  this  point  were  made  as  follows : 
Reinsch's  test  was  applied  to  the  different  compounds  of  arse- 
nic in  this  laboratory  and  in  each  case  several  sublimation  tubes 
were  used.  The  test  was  carried  out  by  boiling  the  substance 
with  sixteen  per  cent,  hydrochloric  acid,  in  which  several  strips 
(2.5  X  0.5  cm.)  of  thin,  pure  copper  were  placed.  After  fifteen 
minutes  (except  in  cases  to  be  mentioned  later)  the  strips  of 
copper  were  removed,  washed  and  dried,  and  after  rolling  or 
folding  to  small  compass,  placed  in  open  tubes  five  cm.  long  and 
not  over  five-tenths  cm.  diameter.  These  tubes  were  held  in  an 
inclined  position  in  the  lowest  possible  flame  of  a  Bunsen  burner 
until  the  arsenic  sublimed  ;  a  second  or  two  usually  sufiices. 

The  test  was  similarly  carried  out  with  compounds  of  anti- 
mony and  also  with  various  organs  of  two  cats,  one  killed  by 
six  grains  of  tartar  emetic,  dying  six  hours  after  administration, 
and  the  other  dying  in  three  days  after  the  administration  of  the 
first  of  six  small  doses  given  every  twelve  hours.  Each  dose  was 
two  grains,  but  much  of  this  was  probably  not  taken  into  the 
system.     A  perceptibly  higher  degree  of  heat  was  necessary  to 

1  H.  Reinsch  :  De  1*  Bssai  de  1'  Arsenic  par  le  Cuivre  :  /.  pharm.  Chim.,  2,  361,  {1842). 
s  H.  Reinsch  :   Ueber  das  Vetlialten  des  metallischen  Kupfers  xu  cinigcn  M etall- 
MhiiBgeB:  /,prakt.  Chem.,  24,  244,  (/^). 


954       reinsch's  test  for  arsenic  and  antimony. 

sublime  the  antimony  than  was  the  ca^e  with  arsenic; 
altogether  185  tests  were  made,  most  of  them  furnishing  good 
sublimation  tubes.  Each  tube  was  numbered  as  made,  and 
later  the  whole  number  were  mixed  and  sorted  for  arsenic  and 
antimony  by  examination  with  a  microscope  of  low  power. 
Reference  to  the  note  book  showed  that  in  no  case  had  a  mis- 
take been  made,  in  fact  in  every  case  the  arsenic  sublimation 
could  easily  be  distinguished  from  that  of  antimony  by  the  naked 
eye.  In  no  case  did  the  sublimate  of  antimonous  oxide  show  a 
trace  of  crystallization  under  the  microscope  used,  nor  did  the 
arsenious  oxide  fail  in  any  case  to  show  the  characteristic  bril- 
liant octahedral  crystals. 

The  evidence  that  the  antimonous  oxide  cannot  appear  in 
crystals  which  might  be  mistaken  for  arsenic  is  of  course  nega- 
tive, but  owing  to  the  variety  of  forms  used  it  must  be  con- 
sidered to  have  the  weight  of  positive  evidence. 

As  regards  the  substances  tested,  the  following  may  be  re- 
corded : 

All  arsenious  compounds  soluble  in  hydrochloric  acid  gave 
the  deposit  on  copper  immediately  on  heating. 

Commercial  **  metallic  '*  arsenic  gave  the  deposit  readily. 

Freshly  sublimed  **  metallic'*  arsenic  (bright  crystals)  gave 
no  deposit. 

Arsenates  gave  a  deposit  only  after  several  minutes  boiling. 

In  the  presence  of  nitric  acid  or  chlorates  no  test  is  obtained 
owing  to  the  solution  of  the  copper. 

Whenever  aqua  regia  or  potassium  chlorate  is  necessary  for 
solution  of  an  arsenic  compound,  the  solution  should  be  evapo- 
rated to  dryness  with  hydrochloric  acid.  The  test  can  then  be 
carried  out  as  with  arsenates. 

The  presence  of  organic  matter  in  the  arsenic  solution  does 
not  affect  the  test,  hence  it  can  be  applied  directly  to  any  organs 
without  any  previous  destruction  of  tissue.  If  much  arsenic  is 
present  it  is  best  to  use  but  a  small  portion  of  the  substance,  since 
if  much  arsenic  is  deposited  on  the  copper,  it  will  not  adhere 
with  firmness. 

Antimony  is  not  precipitated  on  the  copper  as  rapidly  as  arse- 


PHOSPHORUS  IN   STEBL  AND   CAST   IRON. 


955 


nic,  and  the  deposit  has  a  decidedly  violet  tint,  very  distinct 
from  the  iron  gray  deposit  of  arsenic. 

The  following  distribution  of  antimony  in  the  two  cats  may 
be  added : 


Acnte  poisonins:  (6  hours}. 

Stomach, — Heavy  deposit  and  sub- 
limate. Good  test  with  ^\^ 
of  stomach. 

Liver, — Not  so  heavy  deposit  as 
stomach.    Good  sublimate. 

Heart. — Good  deposit  after  sev- 
eral hours  boiling.  Good  sub- 
limate. 

I^ncreas, — Faintdeposit.  No  dis- 
tinct sublimate. 

'Spleen, — Paint  deposit.  No  dis- 
tinct sublimate. 

JCidney. — Faint  deposit.  No  dis- 
tinct sublimate. 

Intestine, — Good  deposit  and  sub- 
limate. 

Muscle.— VeXvX  deposit  on  two 
days  boiling.     No  sublimate. 

£rain, — No  deposit. 
Spinal  Chord, — No  deposit. 

V/ASH1NCTON   AND  LBE  UNIVERSITY. 

Lbxinoton,  VA. 


Slow  poi!M>utug  (73  hours), 
Good  tests. 


Heavy  deposit  and    good    subli- 
mate. 

Good    deposit  on  ninety  minutes 
boiling.    Good  sublimate. 

Good  deposit  and  sublimate. 


Good  deposit  and  sublimate. 

Good  deposit  and  sublimate. 

Slight  violet  tinge  to  copper.    No 
sublimate. 

Marked  violet  tint  to  copper.  No 
sublimate. 


NOTES   ON    THE   DETERMINATION  OF  PHOSPHORUS  IN 

STEEL  AND  CAST  IRON. 


By  George  Auchv. 

Received  August  97,  1196. 


OF  the  many  improvements  made  in  recent  years  in  the 
method  of  determining  phosphorus  in  steel,  that  of  Jones 
— ^the  use  of  the  **reductor'* — is  not  the  least.  There  has  been, 
however,  some  difference  of  opinion  as  to  the  completeness  of 
the  reduction  accomplished  by  its  use.  Quoting  from  three 
most  recent  pu'blications  on  the  subject:  Doolittle  and  Eavenson 
consider  the  reduction  of  the  molybdic  acid  to  be  to  a  point  cor- 


95^  GEORGE   AUCHY. 

responding  to  the  ratio  of  89.16  iron  to  molybdic  acid  ;  Noyes 
and  Royse,  by  special  precautions,  obtain  a  reduction  completely 
to  Mo,0,  (factor  85.71)  ;  and  Blair  and  Whitfield  find  the  ratio 
88.  i6,  a  reduction  to  Mo,^0„  only,  even  with  the  precautions  of 
Noyes  and  Royse  observed.  Doolittle  and  Eavenson  heat  the 
solution  before  passing  it  through  the  reductor.  Noyes  and 
Royse  do  not.  The  first  named  chemists  do  not  use  the  pre- 
cautions of  Noyes  and  Royse.  It  appears  from  a  result  by 
Prof.  Noyes  given  in  this  Journal,  10,  759,  that  he  does  not 
invariably  get  a  reduction  to  Mo,0,  by  his  method,  his  result 
there  given  corroborating  Blair  and  Whitfield's  hypothesis  of  a 
reduction  to  Mo„0„  only. 

It  was  thought  by  the  writer  that  perhaps  the  reduction  to  Mo,0, 
could  invariably  be  accomplished  by  combining  the  precautions 
of  Noyes  and  Royse  with  the  practice  of  Doolittle  and  Eavenson 
of  passing  the  solution  through  the  reductor  hot.  The  following 
results  were  obtained «  using  yellow,  phosphomolybdate  precipi- 
tate dried  six  hours  at  150**  C. 


Phosphomolybdate 
taken. 
Gram. 

Phosphorus 

present. 

Per  cent. 

Phosphorus  found. 

Noyes'  factor. 

fPcr  cent. 

Phosphorus  foui 
Blair's  factor. 
Per  cent. 

O.OIOO 

1.63 

1.63 

1.68 

O.OIOO 

1.63 

1.63 

1.68 

0.0200 

1.63 

1.59 

1-63 

0.0300 

1.63 

1.63 

1.68 

0.0300 

1.63 

1. 61 

1.65 

0.0200 

1.63 

'•55 

1-59 

0.0400 

1-63 

1.59 

1.63 

0.0500 

1.63 

1.62 

1.67 

0.0500 

1.63 

1.63 

1.68 

0.0600 

1.63 

1.63 

1.68 

0.0700 

1.63 

1.61 

1.65 

0.0700 

1.63 

r.6o 

1.64 

O.IOOO 

1.63 

1.63 

1.68 

0.0400 

1.63 

1.59 

1.63 

0.0400 

1.63 

1-59 

1.63 

0.0200 

1.63 

1.58 

1.62 

0.0400 

1.63 

1.56 

1.60 

0.0300 

1.63 

1.59 

1.63 

0.0900 

1.63 

'•57 

1. 61 

0.0400 

1.63 

1.60 

i6i 

0.0500 

1.63 

1.61 

1.65 

PHOSPHORUS   IN   STEEL  AND   CAST   IRON.  957 

Passing  the  solution  through  the  reductor  hot  does  not  seem 
to  insure  an  invariable  reduction  to  Mo,0,,  and  perhaps  adds 
nothing  to  the  effectiveness  of  the  process.  The  following  tests 
were  made  in  the  cold  : 


Phosphomolybdate 
taken. 
Gram. 

Phosphorus 
present. 
Per  cent. 

Phosphorus  found. 

Noyes*  factor. 

Per  cent. 

Phosphorus  found. 
Blair's  factor. 
Per  cent. 

0.0500 

1.63 

1.58 

1.62 

0.0400 

1.63 

1.59 

1.63 

0.0300 

1.63 

1.59 

1.63 

0.0400 

1.63 

1-57 

I.61 

0.0500 

1.63 

1.63 

1.68 

But  in  these  last  tests,  and  also  in  the  first  series  of  tests  in 
nearly  all  cases  where  the  result  calculated  by  Noyes'  factor 
came  low,  the  point  of  the  reductor  had  been  washed  off,  and  the 
sides  of  the  flask  ivashed  down  by  the  jet.  Noyes  warns  against 
any  dilution  of  the  reduced  solution  before  titration,  but  it  was 
thought  that  such  a  slight  dilution  would  do  no  harm.  For  a 
test  of  this  the  following  determinations  were  made  (cold)  and 
without  washing  down  : 


»homolybdai 

e       Phosphomolybdate 

Phosphorus  found. 

Phosphorus  found. 

taken. 

present. 
Per  cent. 

Noyes'  factor, 
rer  cent. 

Blair's  factor. 

Gram. 

Per  cent. 

0.0500 

1.63 

I.61 

1.65 

0.0900 

1.63 

1.60 

1.64 

0.0300 

I. .63 

I.61 

1.65 

0.0300 

1.63 

1.63 

1.68 

0.0400 

1.63 

1.60 

1.64 

0.0400 

1.63 

1.63 

1.68 

0.0300 

1.63 

1.63 

1.68 

0.0400 

1.63 

1.60 

1.64 

0.0400 

1.63 

1.63 

1.68 

0.0500 

1.63 

1.62 

1.67 

Comparing  these  results  with  those  of  the  preceding  series  it 
is  seen  that  a  complete  avoidance  of  any  dilution,  however  slight, 
after  reduction,  will  bring  higher  results  than  if  this  precaution 
be  neglected.  But  it  is  further  seen  that  the  observance  of  this 
precaution  does  not  invariably  assure  a  result  agreeing  with  a 
reduction  to  Mo,0„  although  it  generally  does  so.  Of  the 
eleven  results  in  the^  first  series  of  experiments  (solution  passed 
through  the  reductor  hot),  obtained  by  an  observance  of  this 
precaution,  seven,  calculated  by  Noyes'  factor,  are  over  1.61  ; 


95^  GEORGB  AUCHY. 

and  of  the  ten  results  of  the  last  series  (reduced  cold) ,  seven 
are  1.61  per  cent,  or  over.  On  the  other  hand»  of  the  thirteen 
results  obtained  by  washing  down  the  sides  of  the  flask  after 
leduction,  ten  fall  short  of  the  theoretical  f.63  per  cent,  by  more 
than  0.02  per  cent.,  calculated  by  Noyes*  formula,  and  do  bring 
1.63  per  cent,  calculated  by  Blair*s  factor.  The  inability  of 
Messrs.  Blair  and  Whitfield  to  accomplish  a  reduction  to  Mo,Og 
by  an  observance  of  the  precautions  given  by  Messrs.  Noyes 
and  Prohman,  and  also  the  still  higher  factor  found  by  Messrs. 
Doolittle  and  Eavenson,  may  perhaps  be  due  to  the  fact  that 
the  zinc  in  each  case  used  differed  in  reductive  power  from  that 
of  the  others.  The  writer  had  on  one  occasion  zinc  which  when 
used  in  the  reductor  with  all  care  and  precautions,  never  gave  a 
/eduction  of  more  than  one-half ;  and  in  his  opinion  it  is  safer 
and  more  accurate  to  use  the  old  Emmerton  method  of  reduc- 
tion and  filtration,  but  with  the  modifications  and  precautions 
described  later  in  this  article. 

The  phosphomolybdate  employed  in  the  above  tests,  was,  for 
part  of  them,  made  by  precipitating  from  sodium  phosphate 
solution ;  for  another  part  of  the  tests,  made  by  precipitation 
from  pig  iron  solution,  exactly  as  is  done  in  the  determination 
of  phosphorus  in  pig  iron.  Messrs.  Blair  and  Whitfield  have 
shown  the  constancy  of  the  composition  of  phosphomolybdate 
made  under  varying  circumstances. 

The  volume  of  the  solution  passed  through  the  reductor  in 
each  of  the  above  experiments  was  100  cc,  as  recommended  by 
Blair  and  Whitfield.  Noyes  and  Frohman  use  200  cc,  but  this 
seems  an  unnecessary  bulk.  Fifteen  cc.  of  sulphuric  acid  (2:1) 
was  used  for  acidifying. 

For  washing  100  cc.  of  hot  water  was  used  containing  ten  cc. 
of  sulphuric  acid,  ( 2 :  i ),  followed  by  100  cc.  cold  water,  and 
again  by  fifty  to  seventy-five  cc.  of  cold  water. 

The  reductor  was  of  the  form  described  b)'  Blair  and  Whit- 
field,* except  that  it  was  considerably  wider  at  the  top  than  the  bot- 
tom— in  shape  like  a  common  tinhorn.  This  shape  holds  more 
zinc  for  the  given  height  (ten  inches)  of  the  column,  and  so 
makes  the  necessity  of  filling  less  frequent. 

1  This  Journal.  17.  74- 


PHOSPHORUS   IN   STEEI,  AND   CAST  IRON.  959 

The  redactor  may  be  used  without  refilling  till  the  column  of 
zinc  falls  to  five  or  six  inches  without  any  diminution  of  effective- 
ness. AH  of  the  results  of  the  two  preceding  series,  and  some  of 
the  last  results  in  the  first  series  were  obtained  by  the  use  of  five 
to  seven  inches  of  zinc  in  the  reductor. 

It  adds  somewhat  to  the  facility  of  the  working  of  the  apparatus 
to  have  the  beaker  containing  the  phosphorus  solution  above 
the  level  of  the  zinc  in  the  reductor  so  that  the  connecting  tube 
may  work  as  a  siphon.  And  the  last  washing  may  then  con- 
veniently be  made  by  diminishing  the  force  of  the  suction  of  the 
pump,  loosening  the  stopper  of  the  reductor,  and  allowing  the 
water  to  be  siphoned  over  and  fill  up  the  vacant  space  in  the 
reductor  above  the  zinc  column. 

The  passage  of  the  solution  through  the  reductor  was  not  pre- 
ceded by  the  passage  of  dilute  sulphuric  acid,  and  in  many  of 
the  tests  some  little  air  was  accidentally  drawn  over  into  the 
reductor  at  the  time  of  washing,  although  care  was  uniformly 
taken  to  allow  no  air  to  enter  at  the  first  washing. 

Messrs.  Noyes  and  Royse  direct  that  the  reductor  should  be 
rinsed  with  dilute  sulphuric  acid  before  using,  even  if  it  has 
stood  but  a  few  minutes.  This  is  some  little  trouble,  and  to  test 
the  necessity  of  it,  the  following  tests  were  made  : 


Ptaotptaomolybdate 

Phosphorus 
taken. 

Phosphorus 

Uken. 

found. 

Reductor 

Grmm. 

Per  cent. 

Per  cent. 

stood. 

0.03CX> 

1.63 

I.61 

one  hour 

0.0700 

1.63 

1.60 

all  night 

O.IOOO 

1.63 

1.63 

six  hours 

0.0300 

1.63 

1.63 

all  night 

0.0300 

1.63 

1.63 

three  hours 

0.0300 

1.63 

1.63 

two  days 

0.0300 

1.63 

1.63 

two  days 

0.Q500 

1.63 

1.63 

two  days 

0.0500 

1.63 

1.62 

two  days 

These  results  seem  an  indication  that  this  precaution  is  not 
absolutely  necessary.  But  if  the  reductor  stand  nearly  a  week 
or  more,  the  sulphuric  acid  will  take  up  considerably  more  of 
the  impurity  of  the  zinc  than  ordinarily.  Zinc,  for  instance, 
which  ordinarily  will  require  a  deduction  of  two-tenths  cc.  from 
the  amount  of  permanganate  used  in  the  titration,  will  require 


960  GBORGB  AUCHY. 

a  deduction  of  four-tenths  if  the  reductor  has  stood  that  length 
of  time  unused. 

It  is  necessary  to  remove  the  zinc  from  the  reductor  at 
intervals  for  cleaning,  best  done  by  stirring  up  in  a  capacious 
dish  with  hot  water,  adding  a  little  sulphuric  acid  to 
clear  the  liquid,  pouring  ott,  washing  by  decantation  and. drying 
in  the  dish  on  the  hot  plate.  But  after  such  a  treatment  the 
zinCi  after  being  replaced  in  the  reductor,  should  be  rinsed  with 
dilute  sulphuric  acid  before  being  used  in  analysis,  as  much 
more  than  the  ordinary  impurity  of  the  zinc  will  be  taken  up  by 
the  sulphuric  acid  the  first  time  it  is  used. 

Perhaps  a  more  convenient  way  of  cleaning  the  zinc  is  to  soak 
it  (in  the  reductor)  in  water  for  a  day  (conveniently  over  Sun- 
day) ,  plugging  up  the  ends  of  the  reductor  to  retain  the  water. 
After  such  a  treatment  the  reductor  will  go  a  long  time  without 
becoming  clogged  up  with  zinc  oxide. 

Instead  of  using  the  reductor,  it  is  a  trifle  quicker  and  more 
convenient,, especially  when  the  phosphorus  present  is  consider- 
able as  in  pig  iron,  to  use  the  following  slight  modification  of 
the  old  Bmmerton  method  of  reduction  and  filtration. 

The  yellow  precipitate  in  a  seven  cm.  filter  paper  is  dissolved 
in  as  little  ammonia  as  possible,  allowing  to  run  into  the  eight- 
ounce  Erlenmeyer  flask  in  which  the  precipitation  occurred; 
washed  five  minutes  with  hot  water ;  the  solution  acidified  with 
twenty-five  cc.  of  sulphuric  acid  (two  parts  water  to  one  part 
acid) ;  a  mustard  spoonful  of  granulated  zinc  added  (five  grams), 
and  the  flask  heated  gently  on  the  hot  plate  for  five  minutes,  or 
until  the  zinc  is  nearly  dissolved  (ten  minutes  is  required  for 
some  zinc).  The  flask  is  removed  from  the  plate,  a  little  dry 
sodium  carbonate  added,  and  when  effervescence  has  nearly 
ceased  the  flask  is  corked  tightly  and  cooled  in  cold  water  with- 
out agitating  the  contents  any  more  than  can  be  helped.  The 
solution  is  then  filtered  from  the  undissolved  zinc  through  a 
little  cotton  wool  in  a  Hirsch  funnel,  smallest  size,  using  the 
pump,  and  the  flask  rinsed  out  with  cold  water  three  times  and 
the  rinsings  drawn  through  the  cotton  wool.  The  sides  of  the 
sixteen-ounce  gas  flask  which  receives  the  liquid  are  washed 
down  with  the  jet,  and  the  solution  titrated  in  the  flask  without 
further  dilution. 


PHOSPHORUS   IN   STBBL  AND   CAST  IRON. 


961 


If  the  zinc  is  of  the  sort  not  dissolving  very  readily,  thirty-five 
cc.  of  sulphuric  acid  should  be  used  for  acidifying  the  phos- 
phorus solution  instead'  of  twenty-five  cc. 

The  reduction  is  to  Mo„0,,.  Factor  of  iron  to  molybdic  acid 
90.76.  More  correctly  speaking,  the  reduction  is  to  Mo,0,, 
which  filtering  and  dilution  oxidizes  to  M0j,0„. 

It  will  be  found  upon  trial  that  this  way  of  reduction  and  fil- 
tration is  somewhat  easier  and  more  rapid  than  the  usual  reduc- 
tor  method,  as  the  filtration  through  cotton  wool  in  a  Hirsch 
funnel  and  with  aid  of  the  pump  is  performed  as  easily  and 
quickly  as  merely  pouring  and  rinsing  from  one  vessel  into 
another.  While  the  zinc  is  dissolving  in  one  determination^ 
the  yellow  precipitate  of  the  next  determination  may  be  filtered 
off. 

The  following  results  show  that  the  reduction  and  filtration 
through  cotton  wool,  as  described,  brings  the  molybdenum  oxide 
to  the  form  Mo„0„. 

Considerable  phosphomolybdate  (four-tenths  to  eight-tenths 
gram)  taken  for  each  test. 


Phosphorus 
present. 

Phosphorus 
founcf.    Fac- 
tor 90.76. 

Phosphorus 
present. 

Phosphorus 
founa.  Fac- 
tor 90.76. 

Per  cent. 

Per  cent. 

Per  cent. 

Per  cent. 

1.63 

1.63 

1.63 

1.63 

1.63 

1.63 

1.63 

1.62 

1.63 

1.62 

1.63 

1.63 

1.63 

1.63 

1.63 

1.63 

1.63 

1.63 

1.63 

1.62 

Small  amounts  of  phosphomolybdate  taken. 

Phosphomolyb- 
date taken. 

Phosphorus 
present. 

Phosphorus 

found. 
Factor  90.76. 

Phosphorus 
present,  reck- 
oned as  if 
from  1.8333 
grams  steel. 

Phosphorus 

found  if 

from  i.8a33 

grams 

steel. 

Gram. 

Per  cent. 

Per  cent. 

Per  cent,  in 
the  steel. 

Per  cent,  in 
the  steel. 

0.2000 

1.63 

1.63 

0.179 

0.179 

0.2000 

1.63 

1.63 

0.179 

0.179 

0.2000 

1.63 

1.65 

0.179 

O.181 

0.2000 

1.63 

1.63 

O.I  79 

0.179 

0.1500 

1.63 

1.64 

0.134 

0.135 

0.0890 

1.63 

1.62 

0.071 

0.071 

0.0700 

1.63 

1.64 

0.062 

0.063 

0.0700 

1.62 

1.63 

0.062 

0.063 

962 


GEORGE   AUCHY. 


Phoftphotnolyb- 
date  taken. 

Phosphorus 
Phosphorus             found, 
present.           Factor -90. 77. 

Phosphorus 
present,  reck- 
oned as  if 
from  x.823^ 
firrams  steel. 

phosphorus. 

found  if 
fr»m  X.8933 

fframs 

steel. 

Gram. 

Per  cent             P« 

ir  cent. 

Per  cent,  in 
the  steel. 

Per  cent.  in. 
the  steel. 

0.0600 

1.63 

r.64 

0.054 

0.054 

0.0600 

1.63 

r.63 

0.054 

0.054 

0.0500 

1.63                    ] 

[.61 

0.045 

0.044 

0.0400 

1.63                 : 

t.64 

0.036 

0.036 

0.0300 

1.63                    1 

[.64 

0.027 

0.027 

0.0300 

1.63                    ] 

[.64 

0.027 

0.027 

0.0300 

1.63                     ] 

[.61 

0.027 

0.027 

0.0300 

1.63                    1 

t.6o 

0.027 

0.026 

0.0400 

1.63                 : 

[.62 

0.036 

0.036 

0.0400 

1.63                 ] 

[.63 

0.036 

0.036 

0.0400 

1.63                 : 

t.63 

0.036 

0.036 

0.0350 

1.63                 ] 

[.62 

0.031 

0.031 

0.0380 

1.63                 ] 

^63 

0.034 

0.034 

0.0250 

1.63                 J 

[.6z 

0.022 

0.022 

0.0230 

1.63                 ] 

r.6i 

0.020 

0.020 

0.0200 

1.63                 ] 

C.55 

0.018 

0.017 

0.0200 

1.63                 : 

t.51 

0.018 

0.017 

0.0200 

1.63                 ] 

1.64 

0.018 

0.018 

0.0200 

1.63                 ] 

r.50 

0.018 

0.017 

0.0200 

1.63                 1 

.60 

0.018 

0.018 

0.0200 

1.62                ] 

^55 

0.018 

0.017 

0.0200 

1.63                 ] 

t.57 

0.018 

0.017 

0.0180 

1.63                 1 

1.55 

0.016 

0.015 

0.0150 

1.63                 1 

.55 

0.013 

0.013 

0.0150 

1.63                 1 

.55 

0.013 

0.013 

0.0130 

1.63                 ] 

^51 

0.012 

0.01 1 

0.0120 

1.63                 ] 

r.63 

O.OIO 

0.010 

O.OIOO 

1.63                 1 

^51 

0.0089 

0.008 

O.OIOO 

1.63                 ] 

t.55 

0.0089 

0.0085 

O.OIOO 

1.63                 ] 

[.46 

0.0089 

0.008 

O.OIOO 

1.63                 1 

[.64 

0.0089 

0.0089 

O.OIOO 

1.63                 ] 

[.64 

0.0089 

0.0089 

The  figures  in  the  last  two  columns  were  obtained  by  reckon- 
ing as  though  1.8233  grams  of  steel  had  in  each  case  been  taken 
for  analyses.  In  other  words,  these  percentages  in  the  last  two 
columns  are  what  they  would  have  been  had  the  phosphomolyb- 
date  taken  been,  in  each  case,  obtained  from  1.8233  grams  of 
steel,  in  the  regular  course  of  analysis,  for  phosphorus. 

It  will  be  noticed  that  when  the  amount  of  phosphomolybdate 


PHOSPHORUS  IN  STEEL  AND   CAST  IRON.  963 

taken  is  very  small  (equivalent  to  0.008  to  0.017  per  cent,  in 
steel)  there  is  frequently  some  oxidation,  the  percentage  of 
phosphorus  in  the  yellow  precipitate  thus  falling  short  of  1.63 
by  as  much  as  0.17  per  cent,  in  one  case.  But,  as  will  be  seen 
by  reference  to  the  last  two  columns  of  results,  this  affects  the 
result  in  steel  but  slightly. 

This  proneness  to  oxidation  when  very  little  phosphorus  is 
present  in  the  solution  indicates  that  the  stability  of  the  Mo^.O,, 
solution  is  greater  when  concentrated  than  when  dilute.  And 
the  solution  should  therefore  be  in  as  small  bulk  as  possible. 
Other  necessary  precautions  are :  to  have  a  large  excess  of  sul- 
phuric acid  present,  to  avoid  a  boiling  temperature  when  dis- 
solving the  ^inc,  to  cool  the  liquid  before  filtering  from  the  un- 
dissolved zinc,  to  exclude  air  while  cooling,  and  to  filter  rapidly 
through  cotton  wool  in  a  Hirsch  funnel,  with  aid  of  the  pump. 
But  where  considerable  phosphorus  is  present,  as  in  pig  irons, 
these  precautions  may  be  neglected,  except  the  cooling  before 
filtering.  That  is,  the  liquid  may  be  cooled,  after  the  reduction 
with  zinc,  without  the  addition  of  sodium  carbonate,  and  with 
free  access  of  air,  and  the  filtration  may  be  made  through  a 
seven  cm.  coarse  paper  (instead  of  cotton  wool)  by  aid  of  the 
pump.  The  results  given  under  the  head  * '  considerable  phos- 
phomolybdate  taken  for  each  test'*  were  obtained  in  this  way, 
air  not  excluded,  and  filtered  through  paper  instead  of  cotton 
wool. 

The  advantage  of  making  the  reduction  and  filtration  in  this 
way  in  the  case  of  pig  iron  is  very  marked  when,  as  frequently 
happens,  the  yellow  precipitate  separates  out  when  its  solution 
in  ammonia  is  acidified  with  sulphuric  acid.  For  if  the  reduc- 
tion be  made  as  described,  this  separation  may  be  ignored  as  in 
contact  with  the  zinc  and  sulphuric  acid  the  yellow  precipitate 
becomes  reduced  and  goes  into  solution.  This  is  shown  by  the 
following  tests,  in  which  no  ammonia  was  used  at  all.  That  is, 
the  yellow  phosphomolybdate  precipitate  was  weighed  directly 
into  the  reducing  flasks,  and  thirty-five  cc.  of  sulphuric  acid 
(2:1)  poured  over,  a  mustard  spoonful  of  zinc  added,  heated 
gently,  etc. 


964  GBORGB  AUCHY. 

Phosphomolybdate  Phosphorus  Plosphorus 

taken.  present.  found. 

Gram.  Per  cent.  Per  cent. 
About  0.4000                                         1.63  1.63 

**      0.4000  1.63  1.62 

**      0.4000  1.63  1.63 

"      0.4000  1.63  1.62 

•*      0.4000  1.63  1.63 

In  experimenting  with  this  process  some  interesting  results 
were  had.  The  port  wine  Mo„0,,  solution  is  apparently  not  so 
stable,  especially  in  dilute  solution  or  with  small  amounts  of 
phosphorus  present,  as  Emmerton  supposed,  and  certain  precau- 
tions are  necessary. 

In  the  first  place  considerable  amounts  of  phosphomolybdate 
were  taken,  dissolved  and  reduced  as  described,  and  filtered 
through  seven  cm.  filter  papers  by  aid  of  the  pump.  The  re- 
sults showed  1.63  per  cent,  phosphorus,  the  theoretical  amount. 

Several  tests  were  then  made  in  the  same  way  and  with  the 
same  weights  of  yellow  precipitates,  but  not  waiting  for  the  solu- 
tions to  cool  before  filtering  from  the  undissolved  zinc.  Instead 
of  the  theoretical  1.63  per  cent.,  1.57  per  cent.,  and  1.58  per 
cent,  were  obtained,  showing  the  necessity  of  filtering  cold. 

Next  the  stability  of  the  reduced  solution  was  tested. 

Before  filtering  from  the  Phosphorus  Phosphorus 

undissolved  zinc.  present.  found. 

Per  cent  Per  cent. 

Stood  two  hours 1.63  1.59 

"         **        **       and  poured  back  and 

forth  four  times i  .63  1.55 

Stood  one  hour 1.63  1.59 

**      one-half  hour 1.63  1.61 

The  flasks  were  not  corked  while  standing. 

Smaller  weights  of  phosphomolybdate  precipitate  were  then 
taken.  The  results  obtained  fell  very  much  short  of  the  theo- 
retical, 1.63  per  cent.,  and  varied  considerably.  It  was  at  first 
thought  that  the  filtration  by  aid  of  the  pump  oxidized  the 
solutions  more  than  by  the  original  Emmerton  way  of  filtering 
through  a  large  ribbed  filter.  But,  upon  making  four  tests  and 
filtering  in  that  way  (Emmerton's)  the  results  gave  1.51  per 
cent.,  1.52  per  cent.,  1.46  per  cent.,  and  1.48  per  cent.,  respect- 
ively, instead  of  the  theoretical,  1.63  per  cent.,  although  about 


PHOSPHORUS  IN  STBEL  AND   CAST  IRON.  965 

four-tenths  gram  yellow  precipitate,  was  in  each  case  taken ; 
an  amount  of  yellow  precipitate  which,  when  taken  for  the 
foregoing  tests  made  by  filtering  through  a  seven  cm.  filter  paper 
by  aid  of  the  pump,  never  failed  of  bringing  a  result  equal  to 
the  theoretical.  In  filtering  through  a  seven  cm.  filter  by  the 
pump  the  oxidation  of  the  solution  is  therefore  considerably  less 
than  the  oxidation  by  filtering  through  a  large  ribbed  filter. 

An  article  by  Blair  and  Whitfield'  contains  a  description  of  an 
experiment  made  by  reducing  the  phosphorus  solution  by  boiling 
with  zinc,  keeping  an  atmosphere  of  hydrogen  continually  in  the 
flask,  and  boiling  till  the  zinc  was  completely  dissolved ;  then 
cooling  (maintaining  the  atmosphere  of  hydrogen  in  the  flask) 
and  titrating,  the  result  falling  considerably  below  the  theo- 
retical. In  the  case  of  the  writer's  low  results  just  spoken  of, 
obtained  by  filtering  through  a  seven  cm.  filter  paper  by  suction, 
the  reduction  had  also  been  effected  by  boiling  with  the  zinc, 
though  not  in  an  atmosphere  of  hydrogen,  and  not  to  complete 
solution  of  the  zinc.  Remembering  the  experiment  of  Blair  and 
Whitfield,  above  quoted,  it  was  thought  that  the  reason  for 
the^  low  results  in  both  cases  lay,  perhaps,  in  the  boiling  of 
the  phosphorus  solutions  while  being  reduced,  the  sulphuric 
acid  having  an  oxidizing  effect  perhaps  in  that  case.  No 
otheri  reason  could  be  offered  at  least  for  the  low  result  in 
Blair  and  Whitfield's  experiment,  since,  in  that  experiment,  air 
had  been  so  carefully  excluded  from  the  flask  during  the  solu- 
tion of  the  zinc  and  the  cooling  of  the  liquid.  To  test  the  mat- 
ter, Other  determinations  were  made,  exactly  as  before,  except 
that  the  zinc  was  dissolved  at  a  gentle  heat  instead  of  by  boiling. 
Results  were  much  better,  as  will  be  seen  in  the  following  table. 
Hence  the  necessity  for  the  precaution  of  avoiding  a  boiling  tem- 
perature while  dissolving  the  zinc. 

Phosphorus  present — 1.63  per  cent. 

1  This  Journal,  17,  757. 


966 


GBORGB   AUCHY. 

Phosphosnolybdate 
Uken. 

Gram. 

Zinc  dUtoWed 
by  boiling. 

Pbosphoma  fonnd. 

Zinc  diMoIved  at 
a  gentle  heat. 

Phoaphorui  found 

O.OIOO 

1.37 

1.55 

0.0200 

1-37 

1-50 

0.0200 

1.37 

1. 41 

0.0300 

1. 31 

1.46 

0.0300 

1.37 

■  •  ■  • 

0.0400 

1.46 

1.55 

0.0500 
0.0600 

1-39 
1.28 

1.57 
1.60 

0.0700 

1.52 

1.57 

0.0700 
0.0800 

1.53 
1.16 

a  a  ■  • 
1.57 

0.2000 

1.58 

1.63 

O.IOOO 

1-59 

1.64 

In  the  second  column  of  results,  the  third  and  fourth  results 
are  considerably  lower  than  the  rest  of  them.  But  it  had  been 
noticed  that  in  these  two  determinations  the  green  color  of  the 
reduced  phosphorus  solution  had  faded  to  the  port  wine  shade 
during  the  cooling  of  the  liquid  and  before  the  filtration  from 
the  undissolved  zinc,  while  in  all  the  other  determinations  the 
green  color  had  persisted  till  the  moment  of  filtration.  This 
pointed  to  the  necessity  of  excluding  air  during  the  cooling  of 
the  liquid,  preparatory  to  filtration,  from  the  undissolved  zinc, 
and  the  precaution  was  accordingly  adopted  of  corking  the  flask 
while  cooling,first  adding  a  little  sodium  carbonate  to  fill  the 
flask  with  carbon  dioxide.     Results  by  this  procedure  follow. 

As  the  flask  is  already  filled  with  hydrogen  gas  from  the  solu- 
tion of  the  zinc,  and  vapor  from  the  heating  of  the  liquid,  it  is 
perhaps  unnecessary  to  add  the  sodium  carbonate  at  the  end  of 
the  reduction.  In  that  case  the  flask  should  be  corked  with  a 
one-hole  cork  with  drawn-out  glass  jet»  during  the  solution  of 
the  zinc ;  and  the  jet  closed  when  the  reduction  is  completed. 

It  was  thought  that  results  agreeing  more  closely  and  uni- 
formly with  the  theoretical  might  be  obtained  by  filtering 
through  cotton  wool  instead  of  paper,  as  the  filtration  can  be 
considerably  more  quickly  accomplished  in  that  way,  even  when 
much  suction  is  used  in  the  latter  way.  Results  showed 
this  to  be  the  case,  and  are  also  given  below  in  comparison  with 
results  by  filtering  through  paper. 


PHOSPHORUS  IN   STBBI«  AND   CAST  IRON. 


967 


Phosphomolyb- 
date  taken. 

Gram. 
O.OIOO 
O.OICX) 
O.OIOO 

0.0100 

O.OIOO 

o.oaoo 
0.0200 
0.0200 
0.0200 
0.0200 
0.0200 
0.0300 
0.0300 
0.0300 
0.0300 
0.0400 
0.0400 
0.0500 
0.0600 
0.0600 
0.0700 
0.0800 


Piltration 
through  paper. 

Through  cot- 
ton wool. 

Phoaphorua 
present. 

Phosphorus 
found. 

Phosphorus 
found. 

Per  cent. 

Per  cent. 

Per  cent. 

1.63 

1. 41 

X.55 

i.63 

•  •  a  . 

1.64 

X.63 

•  .  •  ■ 

1.46 

1.63 

•  •  •  ■ 

1.64 

1.63 

•  .  ■  • 

1.46 

1.63 

1.46 

1.50 

1.63 

1.48 

1.60 

1.63 

1.48 

1.64 

1.63 

•  •  a  • 

1.63 

1.63 

•  •  a  a 

1.55 

1.63 

m  »  »  • 

1.57 

1.63 

1.58 

1.64 

1.63 

1.57 

1.63 

1.63 

•      •      •     • 

1.64 

1.63 

■     •     •      • 

I.61 

1.63 

1.55 

1.64 

1.63 

•     •     •     • 

1.62 

1.63 

1-57 

I.61 

1.63 

1-57 

1.64 

1.63 

1.60 

1.63 

1.63 

1.60 

1.63 

1.63 

1.62 

a  •  •  a 

Prom  these  results  it  is  seen  that  cotton  wool  is  much  better 
for  use  in  filtering  from  the  undissolved  zinc  than  paper.  Very 
little  pressure  is  required  with  the  former  and  very  little  cotton 
wool  is  required.  A  small  Hirsch  funnel  is  necessary.  But  the 
cotton  wool  should  not  be  pressed  down  with  the  finger  after  it 
is  wet,  but  sucked  down  by  the  pumpa  In  the  above  experi- 
ments the  filtrations  through  paper  were  also  accomplished  by 
a  Hirsch  funnel,  smallest  size.     (Paper  size,  seven  cm.) 

Using  cotton  wool,  no  oxidation  of  the  port  wine,  Mo,,0,„ 
solution  need  be  feared  where  the  amount  of  phosphorus  pres- 
ent is  that  which  in  a  sample  of  steel  (one  and  eight-tenths 
grams)  would  be  equivalent  to  0.020  per  cent,  or  over ;  while 
with  percentages  under  0.020  the  oxidation  is  never  greater  than 
will  make  a  difference  of  o.ooi  per  cent,  in  the  result. 

All  the  foregoing  experiments  were  made  with  the  use  of  zinc, 
requiring  about  ten  minutes  for  solution.     This  supply  becom- 


968  GEORGE  AUCHY. 

ing  exhausted,  new  zinc  was  procured  which  happened  to  dis- 
solve much  more  freely  in  acid,  and  experiments  were  therefore 
made  as  before  but  using  only  fifteen  cc.  of  sulphuric  acid  for 
solution  of  the  zinc  instead  of  thirty-five  cc.  as  before  with  the 
first  lot  of  zinc.  Results  were  noticeably  lower,  pointing  to  the 
inference  that  a  large  excess  of  sulphuric  acid  present  is  neces- 
sary as  favoring  the  stability  of  the  Mo,,Oj,  port  wine  solution. 
Other  determinations  were  then  made,  using  twenty-five  cc.  of 
acid. 

Phosphorus  present,  1.63  per  cent. 

Phosphomolybdate 
taken. 

Gram. 
0.0400 
0.0380 
0.0350 
0.0320 
0.0300 
0.0280 
0.0250 
0.0230 
0.0230 
0.0200 
0.0200 
0.0180 
0.0180 
0.0150 
0.0150 
0.0160 
0.0140 
0.0130 
O.OIOO 

This  shows  the  necessity  for  the  precaution  of  using  plenty 
of  sulphuric  acid  for  solution  of  the  zinc. 

As  before  pointed  out,  results  by  the  foregoing  procedure, 
using  all  precautions,  never  fail  of  the  theoretical,  1.63  per  cent., 
or  a  reasonable  approximation  thereto,  except  when  the  amount 
of  phosphomolybdate  taken  is  only  0.0200  gram  (equivalent  to 
0.018  per  cent,  in  all  steel  determinations)  or  less,  and  the 
error  in  that  case  in  a  steel  never  amounts  to  more  than  o.ooi 
per  cent,  with  about  two  grams  of  steel  taken  for  analysis  ;  and 


Fifteen  cc.  sul- 
phuric acid. 

Phosphorus  found. 

Per  cent. 

1.63 
1.63 
1.62 

Twenty-five  cc.  sul 
phunc  acid. 

Phosphorus  found 

Per  cent. 

«  •  ■  * 
«  ■   •  • 
•  •  •  • 

I-5I 

•  •  •  • 

1-55 

1.60 

1-53 
1.50 

•  •  •  • 

1. 61 

1.50 

^.55 
1.48 

■  •  •  • 

I.61 
1. 51 

1.55 

1.55 

1.47 
1.48 
1.46 

1.55 

•  •  •  • 

1.55 

•  ■  *  • 

1.55 

•  •  •  • 

•  •  •  • 

1.54 
1.56 

1.40 

I.5I 

1.40 

I-5I 

PHOSPHORUS  IN   STBBL  AND  CAST  IRON.  969 

the  writer  therefore,  on  the  sc«re  of  accuracy,  prefers  this 
method  to  the  reductor  method. 

A  convenience  in  phosphorus  determinations  is  a  Mohr 
burette  for  the  sulphuric  acid,  attached  to  the  sulphuric  acid 
bottle  by  tubing  reaching  just  to  the  zero  mark  of  the  burette 
according  to  the  well  known  plan.  The  bottle  should  stand 
high,  and  the  tubing  be  wide  so  that  too  much  lung  power  will 
not  be  required  to  fill  the  burette.  The  delivery  tube  of  the  burette 
should  also  be  of  a  good  width,  so  that  the  acid  may  run  quickly 
into  the  phosphorus  solutions.  The  apparatus  is  also,  conve- 
nient for  Blliott  sulphur  determinations,  using  sulphuric  acid  for 
acidifying  the  caustic  soda  sulphur  solution  instead  of  hydro- 
chloric acid. 

There  is  some  difference  of  opinion  among  chemists  as  to  the 
advisability- of  using  sugar  for  reducing  the  manganese  precipi- 
tate formed  by  the  addition  of  permanganate  to  the  boiling  nitric 
acid  solution  of  the  steel.  Sugar  was  originally  recommended 
by  Dr.  Drown,  but  Mr.  Clemens  Jones,  obtaining  varying 
results  which  he  attributed  to  its  use,  substituted  ferrous  sul- 
phate with  very  satisfactory  results.  Dr.  Dudley  also  states  that  in 
using  sugar  a  different  result  is  obtained  than  when  ferrous  sul- 
phate is  used.  On  the  other  hand.  Handy  and  others  have 
claimed  that  sugar  has  no  harmful  effect.  The  following  tests 
were  made  by  the  writer : 

Usinf  ferrous  sulphate.  Using  sugar. 

No.                                              Phosphorus.  Phosphorus. 

Per  cent.  Per  cent. 

Steel  618 0.017  0.018 

*'    690 0.018  0.018 

Gray  pig  iron 0'7i9  0.720 

Test  bar 0.016  0.016 

Steel  684 0.049  0.049 

Phosphate  solution o.  123  0.121 

Thes^  results  were  considered  sufficient  evidence  that  sugar 
does  not  interfere  with  the  precipitation  of  the  phosphorus.  Its 
use  is  more  advantageous  in  several  respects  :  it  is  cheaper  than 
ferrous  sulphate  ;  less  of  it  is  required  ;  it  may  be  added  to  the 
boiling  solution  without  fear  of  the  solution  boiling  over ;  and  it 
never  contains  phosphorus. 


970  I«.  M.  DENNIS,    MARTHA  DOAN,    AND   A.  C.  GILI«. 

The  merest  pinch  of  sugar  will  suffice  to  reduce  a  very  abun- 
dant precipitate  of  manganese  peroxide  if  the  boiling  be  con- 
tinued for  some  time  after  its  addition  to  the  liquid. 

For  the  filtration  of  the  yellow  phosphomolybdate  precipitate 
with  the  aid  of  the  pump,  it  is  the  writer's  experience  that  noth- 
ing succeeds  so  well  as  two  seven  cm.  Schleicher  &  Schtill 
No.  579  filter  papers,  folded  and  placed  in  the  funnel  together. 
The  filtration  may  be  made  very  rapidly,  yet  without  any  of  the 
precipitate  going  through  the  paper. 

After  the  solution  of  the  yellow  precipitate  on  the  filter  paper 
in  ammonia  and  washing,  the  same  filter  may  be  used  (without 
removal  from  the  funnel)  for  another  phosphomolybdate  filtra- 
tion, and  so  on  for  a  number  of  consecutive  determinations. 

No.  579  is  a  very  loose  and  porous  paper.  No.  589  black  rib- 
bon also  serves. 


50nE  NEW  COMPOUNDS  OF  THALLIUM. 

By  L.  M.  DBNms  and  Maktha  Doam,  with  Crystallogrnpbic  NoCeft,  by  A.  C.  Gill. 

ReoeiTcd  September  4.  iflQ^ 

THAI.i:X>US    TRINITRIDE,  TIN,. 

WHEN  a  concentrated  solution  of  potassium  trinitride  con- 
taining a  little  free  hydronitric  acid  is  added  to  a  solu- 
tion of  thallous  sulphate,  a  white,  finely  crystalline  precipitate 
is  formed.  This  compound  is  soluble  in  hot  water,  and  when 
recrystallized  from  a  hot  aqueous  solution,  it  separates  in  ortho- 
rhombic  needles  of  a  light  straw  color. 

The  thallium  in  this  salt  was  determined  volumetrically  by 
means  of  a  standard  solution  of  potassium  permanganate,  accord- 
ing to  the  method  of  Willm.' 

In  the  case  of  the  hydronitric  acid,  a  volumetric  method  also 
was  first  attempted.  A  weighed  portion  of  the  salt  wasdissolved 
in  water  and  placed  in  a  Hempel  distilling  bulb,  which  was  con- 
nected by  fused  joints  to  a  condenser.  A  separatory  futoel  was 
inserted  in  the  neck  of  th^  distilling  bulb.  The  hydronitric  acid 
was  set  free  by  the  addition  of  an  excess  of  dilute  sulphuric  acid 
and  was  distilled  into  an  Erlenmeyer  flask  containing  a  known 
amount  of  ammonia,  the  excess  of  ammonia  being  then  deter- 

1  Ann.  chtm. phys.^  (4),  5,  79. 


SOMB  NBW  COMPOUNDS  OP  THALUUM.  97 1 

mined  by  titration.  It  was  at  first  difficult  to  drive  over  all  of  the 
hydronitric  acid,  the  results  being  uniformly  low  with  one  excep- 
tion, and  in  that  case  the  distillate  gave  a  reaction  for  sulphuric 
acid.  The  results  continued  poor  in  spite  of  various  modifica- 
tions which  were  tried,  so  that  finally  recour^  was  had  to  the 
gravimetric  method,  this  not  having  been  used  before  because  of 
the  explosive  character  of  the  silver  trinitride.  A  weighed  por- 
tion of  the  salt  was  dissolved  in  water  and  precipitated  with  a 
neutral  silver  nitrate  solution.  The  silver  trinitride  was  thor- 
oughly washed  by  decantation  with  cold  water,  the  washings 
being  passed  through  a  Schleicher  and  Schtill  hardened  filter 
No.  575.  The  precipitate  was  then  transferred  to  the  paper,  the 
point  of  the  filter  carefully  perforated  and  the  precipitate 
washed  through  into  a  weighed  porcelain  crucible.  Hydro- 
chloric acid  was  then  added  to  the  contents  of  the  crucible  and 
the  whole  evaporated  to  dryness.  By  this  treatment  the 
silver  trinitride  is  decomposed  and  *  the  hydronitric  acid 
expelled,  together  with  the  excess  of  hydrochloric  acid.  The 
silver  chloride  remaining  in  the  crucible  was  then  weighed,  and 
from  its  weight  the  amount  of  nitrogen  in  the  salt  was  computed. 
The  results  were : 

Calculated  for 

T1N|.  Pound. 

Thallium 204.18  82.9  82.87 

Nitrogen 42.09  17.1  17.2 

246.27  loo.o  100.07 

The  prism  angle  could  be  measured  on  the  goniometer,  but 
the  end  faces  were  too  small  to  give  good  reflections.  The  trace 
of  the  macrodome  on  the  prism  face  was  measured  repeatedly  on 
the  microscope  stage,  giving  an  angle  of  51®  30' with  the  vertical 
edge.  The  prism  angle,  no  :  110=^79"  50'.  Hence  the  axial 
ratio : 

a  :  b  :  c  =  0.8366  :  i  :  1.2407. 

The  crystals  were  composed  of  many  fine  needles,  sometimes 
twinned  on  the  prism  face  (no),  but  more  frequently  in  parallel 
growth.  The  double  refraction  was  strong,  and  the  plane  of 
the  optical  axes  is  at  right  angles  to  the  long  direction  of  the 
needles,  i.  ^.,=  o.ooi. 


972  I<.  M.  DENNIS,    MARTHA   DOAN,   AND  A.  C.  GILL. 

Thallous  trinitride  is  somewhat  soluble  in  cold  water  and  is 
easily  soluble  in  hot  water.  It  is  not  explosive,  resembling  in 
this  particular  the  trinitrides  of  potassium  and  sodium.  It  melts 
without  decomposition  when  heated  in  an  atmosphere  of  carbon 
dioxide.  Its  melting  point  was  determined  by  placing  some  of 
the  crystals  in  a  small  glass  tube  in  the  top  of  which  was  inserted 
a  cork  with  two  holes.  Carbon  dioxide  was  passed  into  the 
tube  through  one  of  these  openings,  and  a  small  exit  tube  was 
inserted  in  the  other.  The  tube  was  heated  by  immersing  it  in 
a  bath  containing  an  easily  fusible  alloy,  and  the  temperature 
was  measured  with  a  carbon  dioxide  filled  thermometer  corrected 
by  the  Physikalisch-Technische  Reichsanstalt  of  Charlotten- 
burg.  The  corrected  temperature  at  which  the  crystals 
melted  was  334"*. 

When  exposed  to  the  sunlight,  the  crystals  of  thallous  trini- 
tride assume  a  dark  brown  appearance,  which  is  probably  due 
to  the  formation  of  thallous  oxide.  This  change  must  be  very 
superficial,  however,  as  no  change  in  weight  could  be  detected 
in  a  sample  which  had  been  in  a  southern  exposure  for  two 
months. 

When  heated  in  a  current  of  dry  nitrogen,  thallous  trinitride 
was  easily  reduced.  The  hydrogen  on  leaving  the  combustion 
tube,  in  which  the  boat  containing  the  thallous  trinitride  was 
placed,  was  passed  through  two  bulbs  containing  water.  The 
aqueous  solution  thus  obtained  had  a  very  distinct  odor  of  ammo- 
nia, turned  turmeric  paper  brown,  and  when  neutralized  with 
hydrochloric  acid  and  allowed  to  spontaneously  evaporate  over 
sulphuric  acid  and  caustic  potash,  it  yielded  crystals  which 
under  the  microscope  were  identical  with  those  of  ammonium 
chloride.  The  ammonia  found  in  two  of  the  reductions  in  hydro- 
gen was  titrated  with  standard  acetic  acid,  this  acid  being  used 
in  order  that  only  the  free  ammonia  might  be  neutralized  and 
any  ammonia  which  might  be  present  combined  with  hydronitric 
acid  would  remain  as  such.' 

In  one  case  29.83  per  cent,  of  the  nitrojg^en  in  the  trinitride 
acid  was  converted  into  ammonia  ;  in  the  other  27.37  per  cent. 
of  the  nitrogen  was  thus  changed. 

1  HN3  is  somewhat  stronger  than  glacial  acetic  acid.   /.  prakt.  Chem.^  (2),  4s,  207. 


SOME   NEW   COMPOUNDS  OF   THALLIUM.  973 

Hydroiiitric  acid  was  tested  for  in  the  aqueous  solution  by 
addition  of  silver  nitrate  to  the  solution  in  which  the  ammonia 
had  been  neutralized,  and  in  each  case  only  a  trace  was  found. 
It  was  thought  that  perhaps  the  formation  of  the  acid  might  be 
due  to  the  presence  of  a  small  amount  of  moisture  in  the  hydro- 
gen, so  a  reduction  was  made  with  hydrogen  which  had  been 
passed  through  a  piece  of  moist  cotton.  In  this  case  21.55  per 
cent,  of  the  nitrogen  was  converted  into  ammonia,  and  as  before 
only  a  small  amount  of  hydronitric  acid  was  formed. 

The  highest  results  for  the  nitrogen  converted  into  ammonia 
approximate  one-third  of  the  total  nitrogen  present,  and  inas- 
much as  only  a  trace  of  the  nitrogen  is  found  to  exist  in  the  form 
of  hydronitric  acid,  it  is  possible  that  the  molecule  of  the  acid 
breaks  down  thus  :* 

II  )N— H  +  5  =  N,+  N— H. 
N^  ^  \H 

THALLOUS  THALLIC  TRINITRIDK,  T1N,.T1N,. 

It  was  thought  .that  thallic  trinitride  might  be  obtained  by  the 
solution  of  freshly  precipitated  thallic  hydroxide  in  hydronitric 
acid.  The  hydroxide  when  .treated  with  hydronitric  acid  and 
warmed,  dissolved  to  a  clear  straw-colored  solution,  but  when 
the  solution  was  allowed  to  stand  at  ordinary  temperature, 
hydronitric  acid  escaped  and  thallic  hydroxide  was  precipitated. 
Concentration  of  the  solution  was  tried  by  placing  it  in  a  freezing 
mixture  and  removing  the  water  as  ice.  Prom  the  liquid  thus 
concentrated,  bright  yellow  crystals  separated,  yet  so  much  of 
the  salt  solution  was  occluded  in  the  ice  that  this  method  proved 
wasteful.  The  best  yield  of  crystals  was  obtained  by  dissolving 
the  thallic  hydroxide  in  a  one  and  six-tenths  per  cent,  solution 
of  hydronitric  acid  and  allowing  the  solution  to  stand  at  a  tem- 
perature of  about  zero  in  a  Hempel  desiccator  which  was 
exhausted  by  means  of  a  common  suction  pump.  Glistening, 
yellow,  needle-shaped  crystals  appeared.  They  were  removed 
in  five  fractions,  which  under  the  microscope  seemed  to  be  alike 
and  homogeneous. 

1  The  further  investigation  of  thin  reaction  is  now  being  carried  on  in  this  labora- 
tory.   D. 


974  ^-  ^'  DBNNIS,    MARTHA   DOAN,    AND   A.  C.  GII,I<. 

These  sharply  outlined  crystals  verged  toward  a  brown  color 
in  the  larger  specimens.  On  the  stage  of  a  microscope  they 
showed  either  parallel  extinction,  or  an  extinction  of  42^.  That 
is,  the  long  direction  of  the  crystals  varied  in  different  individuals. 
The  crystals  were  probably  triclinic,  though  there  is  a  possibility 
that  they  furnished  a  case  of  flattening,  parallel  to  the  face  of 
the  orthorhombic  pyramid.  An  optical  axis  emerged  obliquely 
from  the  tabular  face,  showing  that  it  was  not  really,  as  would 
otherwise  appear,  the  pinacoid  of  an  orthorhombic  crj-stal.  The 
plain  angles  were  132**,  132**  and  96**.  The  double  refraction 
was  not  very  strong. 

The  thallium  was  determined  by  dissolving  some  of  the  crj's- 
tals  in  dilute  hydrochloric  acid,  reducing  the  thallium  to  the 
thallous  condition  by  sulphurous  acid,  driving  off  the  excess  of 
the  latter  acid  by  heating  the  solution  and  then  titrating  with 
potassium  permanganate.  The  nitrogen  could  not  be  determined 
by  the  method  used  for  thallous  trinitride,  because  the  salt 
could  not  be  dissolved  either  in  water  or  dilute  acids  without 
evolution  of  hydronitric  acid.  For  this  reason  the  absolute 
method  was  used.  We  had  already  found  that  the  salt  was 
highly  explosive,  but  the  behavior  of  the  thallous  trinitride,  when 
heated  in  an  atmosphere  of  carbon  dioxide,  led  us  to  attempt  the 
decomposition  of  a  small  portion  of  this  substance  in  a  similar 
manner.  A  few  milligrams  were,  accordingly,  spread  over  the 
bottom  of  a  long  porcelain  boat,  which  was  placed  in  a  combus- 
tion tube  containing  granular  copper  oxide.  The  tube  was  con- 
nected at  one  end  to  a  carbon  dioxide  generator,  and  at  the  other 
to  a  Schiff  nitrometer.  The  exit  end  of  the  tube  was  heated  to 
redness  and  the  heat  was  then  run  back  very  carefully  toward 
the  boat.  Gradual  decomposition  of  the  compound,  however, 
was  not  attained,  for  when  the  temperature  in  the  neighborhood 
of  the  boat  had  risen  but  slightly,  the  salt  exploded  violently, 
shattering  the  boat  and  tube.  Another  portion  of  the  hydroni- 
tride  was  then  mixed  with  granular  copper  oxide  and  heated  as 
before.  The  decomposition  in  this  case  was  quiet  and  gradual. 
The  nitrogen  in  the  nitrometer  amounted  to  27.32  per  cent,  of 
the  salt  taken.  It  seemed  possible,  however,  that  in  mixing  the 
hydronitride  with  the  coarse  copper  oxide,  some  of  the  salt 


SOME   NEW   COMPOUNDS  OP  THALLIUM.  975 

might  have  been  decomposed  by  the  friction,  and  that  conse- 
quently the  above  per  cent,  of  nitrogen  might  be  too  low.  To 
ascertain  if  this  were  true,  a  fresh  portion  of  copper  oxide  was 
ground  very  fine  and  was  then  carefully  mixed  with  a  small  por- 
tion of  the  salt.  In  this  way  higher  results  were  obtained. 
The  analysis  gave : 


Thallium 304.18 

Nitrogen 84.18 


Calculated  for 
TIN,. 

Pound. 

70.81 

70.70 

29.19 

^.3 

288.36  100.00  100.00 

If  this  were  a  simple  compound,  the  thallium  would  seem  to 
be  in  the  bivalent  condition,  but  as  this  is  at  variance  with  the 
nsual  behavior  of  the  element,  it  seemed  more  probable  that  the 
compound  is  a  double  salt  containing  thallium  in  both  the  thal- 
lous  and  thallic  condition.  This  supposition  was  confirmed  by 
the  behavior  of  the  crystals  when  treated  with  hot  water.  Brown 
thallic  hydroxide  separated,  and  upon  filtering  this  off  and  add- 
ing potassium  iodide  to  the  filtrate,  a  precipitate  of  thallous 
iodide  resulted.  Instead,  however,  of  finding  only  fifty  per 
cent,  of  thallium  in  the  thallous  condition,  as  would  be  required 
by  the  formula  TIN,. TIN,,  there  was  obtained  63.7  per  cent. 
This  excess  of  thallous  thallium  is  doubtless  due  to  the  reduc- 
tion of  some  of  the  thallic  hydroxide  by  the  hydronitric  acid  set 
free  when  the  salt  is  treated  with  hot  water. 

Thallous-thallic  trinitride  is  highly  explosive,  the  decomposi- 
tion being  accomplished  by  a  sharp  report  and  a  vivid  flash  of 
green  light.  The  explosion  can  be  brought  about  by  heat,  per- 
cussion or  even  gentle  friction. 

THALLOUS  TELLURATE,  Tl,TeO,. 

In  1878  p.  W.  Clarke  prepared  what  he  supposed  to  be  thal- 
lous tellurate  by  precipitating  a  thallous  nitrate  solution  with 
ammonium  tellurate.*  The  amount  obtained  was  so  small  that 
no  analysis  was  made. 

To  avoid  the  presence  of  other  salts  in  the  solution,  we  used  a 
solution  of  thallous  hydroxide  and  precipitated  that  by  adding  a 

1  Ber.d.  dUm.  Ges.^  xi,  1507. 


976  L.  M.  DENNIS,    MARTHA   DOAN,    AND   A.  C.  GILL. 

solution  of  pure  telluric  acid.  The  white,  flocculent  precipitate 
which  formed  was  washed  with  cold  water,  transferred  to  a  fil- 
ter and  dried  over  calcium  chloride. 

In  the  analysis  of  this  substance,  the  thallium  was  determined 
by  the  method  above  described.  Considerable  difficulty  was 
encountered  in  the  determination  of  the  tellurium,  the  presence 
of  thallium  making  it  impossible  to  use  either  the  potassium  per- 
manganate titration  or  the  method  recently  described  by  Gooch.* 
The  thallous  tellurate  was  soluble  in  water,  but  the  amount  of 
water  required  for  its  solution  was  so  great  that  the  telluric  acid 
could  be  precipitated  by  neither  lead  nor  barium  solutions.  For 
these  reasons  the  method  of  Kastner*  was  used,  the  tellurium 
being  precipitated  in  alkaline  solution  by  means  of  grape  sugar. 
As  some  thallium  separated  with  the  tellurium,  the  precipitate 
was  treated  with  nitric  acid  and  the  acid  then  driven  off  by 
evaporation.  The  thallous  nitrate  was  removed  by  washing  the 
residue  with  water  and  the  tellurous  oxide  was  filtered  in  a  Gooch 
crucible,  dried  and  weighed.     The  results  were  : 

Calculated  for 

TljTc04.  Found. 

Thallium 408.39  68.13  68.17 

Tellurium 127.00  21.19  21.19 

Oxygen 64.00  10.68  (diff . )  10.64 

599-36  100.00  100.00 

Thallous  tellurate  is  slightly  soluble  in  water,  and  it  was 
hoped  that  there  might  be  obtained  from  the  aqueous  solution 
crystals  sufficiently  well  defined  to  admit  of  a  comparison  of 
them  with  those  of  thallous  sulphate  and  thallous  selenate. 
Unfortunately,  however,  it  was  found  impossible,  in  spite  of 
many  and  varied  attempts,  to  obtain  anything  but  a  white  amor- 
phous powder.  Even  when  a  solution  saturated  at  40**  was 
allowed  to  slowly  cool  to  15"  through  a  period  of  eight  days,  no 
cr>'stals  resulted. 

THALLOUS  CYANPLATINITE,  Tl,Pt(CN)^. 

Carstanjen  prepared  what  he  reported  to  be  thallous  cyanplat- 

1  Ztichr,  anorg.  Chem.y  7,  132. 
a  Ztschr.  anal.  Chem.,  14,  142. 


SOME   NEW   COMPOUNDS  OF  THALLIUM.  977 

inite  by  neutralizing  cyanplatinousacid  with thallous carbonate.' 
The  compound  was  given  the  formula  TlCN.PtCN,  although  no 
analytical  results  were  given. 

The  cyanplatinous  acid  used  by  us  in  the  preparation  of  the 
thallium  cyanplatinite  was  obtained  according  to  the  method  of 
Schafarik.'  It  was  neutralized  by  thallous  hydroxide,  which 
was  prepared  by  precipitating  a  thallous  sulphate  solution  with 
the  calculated  amount  of  baryta  water.  The  crystals  separated 
out  in  the  form  of  thin  plates. 

A  determination  of  the  thallium  and  cyanogen  gave  the  follow- 
ing results : 

Calculated  for 
Tl>Pt(CN)4.       Observed. 

Thallium 408.36  57.73  57.7 

Platinum I9S'(^  37>S6 

Cyanogen 104.12  14.71  14.5 


707.48  100.00 

The  crystals  are  nearly  colorless  plates,  usually  very  thin  and 
occurring  irregularly  grown  together  on  the  flat  sides.  The 
crystal  system  was  not  positively  determinable  from  the  material 
at  hand,  but  is  probably  triclinic,  possibly  monoclinic  with 
crossed  dispersion.  In  converged  polarized  light,  a  bisectrix  is 
seen  nearly  or  quite  normal  to  the  large  face  of  the  plates,  and 
the  dispersion  of  the  planes  of  the  optic  axes  is  remarkably 
strong,  so  that  the  crystals  simply  change  color  without  becom- 
ing dark  on  rotation  between  crossed  Nicols.  The  double 
refraction  is  high.  The  plates  are  bounded  by  crystal  faces, 
giving  them  a  six-sided  outline,  but  on  the  material  used  no 
goniometric  measurements  could  be  made. 

Cornell  Univbrsity, 
AUGUST  1896. 

1  J.prakt.  Ch€m.^  xoa,  144. 
s/^uf..  66,401. 


NOTES  ON  THE  ESTIHATION  OF  CAFFEIN. 

By  w.  a.  Puckker. 

Received  September  a.  1896. 

SOME  time  ago  Gomberg  published  a  method  for  the  estima- 
tion of  caffein,  by  means  of  Wagner's  reagent/  wherein 
appear  certain  statements  from  which  is  to  be  inferred  the 
superiority  of  this  method  over  such  where  the  caffein  is  shaken 
out  of  an  aqueous  solution  by  means  of  chloroform,  and  which, 
if  true,  would  show  that  most  methods  now  in  use,  give  low 
results  since  but  an  imperfect  separation  of  caffein  is  attained. 
Thus  Spencer*  is  said  to  have  demonstrated  the  diflSculty  with 
which  the  alkaloid  is  abstracted  from  watery  solutions,  he 
,  directing  that  at  least  seven  portions  of  chloroform  be  used  for 
this  purpose,  but  offering  no  proof  of  the  necessity  for  this 
departure  from  the  usual  direction  of  shaking  out  the  liquid 
with  three  or  four  portions  of  the  solvent.  Spencer  is  at  vari- 
ance with  Allen/  who  investigated  this  matter  and  found  that 
from  a  solution,  slightly  acidulated  with  sulphuric  acid,  one 
treatment  with  chloroform  removed  seventy  to  eighty-five  per 
cent,  of  the  amount  present,  while  four  usually  effected  com- 
plete extraction,  especialh'  if  toward  the  end  the  solution  is 
rendered  faintly  alkaline. 

This  agrees  well  with  the  results  of  my  own  experiments, 
where  anhydrous  caffein,  in  quantities  from  one-tenth  to  four- 
tenths  gram,  dissolved  in  fifty  cc.  one  per  cent,  sulphuric  acid, 
was  shaken  successively  with  twenty-five,  ten  and  ten  cc.  chloro- 
form, the  united  chloroform  solution  evaporated  at  a  gentle 
heat  and  the  residue  dried  over  sulphuric  acid  to  constant 
weight.  In  each  case  the  solution  was  shaken  with  a  further 
quantity  of  ten  cc.  chloroform  and  the  weight  of  the  caffein  so 
extracted  ascertained  as  before. 


Caffeiti 

Residue  from  first,  secon 

d      Residue  from 

taken. 

and  third  extraction. 

fourth  extraction. 

Total  percent 

Gram. 

Gram. 

Gram. 

recovered. 

0. 1 285 

0.1277 

0.0004 

99.69 

0.1852 

0.1820 

0.0026 

99.67 

0.1988 

0.1980 

0.0002 

99.69 

0.201 I 

0.1977 

0.0025 

99-55 

0.2559 

0.2552 

0.0005 

99,92 

0.4416 

0.4355 

0.0043 

9958 

1  This  Journal, 

18, 

331. 

- 

iy.  AnaL  Chem 

.4i 

300. 

8  Com.  Org. 

Anal. 

.3,  Part  11.  485- 

ESTIMATION  OP  CAPPBIN.  979 

This  shows  that  the  extractioil  of  ca£Fein  from  an  aqueous 
solution  presents  no  difficulties  since,  even  when  the  solution  is 
quite  acid,  practically  the  entire  amount  is  obtained  when  four 
portions  of  chloroform  are  used  ;  while,  even  if  the  fourth  be 
omitted  the  results  will  be  sufficiently  correct  for  most  purposes. 

In  the  article  referred  to  we  are  also  told,  although  it  is 
usually  stated  caffein  may  be  shaken  out  of  an  acid  solution, 
since  its  salts  are  broken  up  by  water,  that  this  is  but  relatively 
true;  as  a  proof  thereof  the  following  is  offered  : 

•*  1.0085  grams  of  caffein  were  dissolved  in  sixty  cc.  of  sul- 
phuric acid  (i.io),  and  this  solution  was  repeatedly  shaken  with 
chloroform,  twenty-five  cc.  at  a  time  : 

m 

Ten  consecutive  portions  of  chloroform  gave  a  total  of  0.3514  gram  caffein. 

Three  additional        "        •*  '*  made      *'      "0.4859     " 

Three  more  "  *'        "  **  "  '*      "0.5034     '*        "      •» 

Since  the  degree  of  dissociation  of  caffein  salts  is  inversely 
proportional  to  the  acid  strength  of  the  solution,  it  is  to  be 
expected  that  it  will  be  extremely  difficult  to  shake  out  the  alka- 
loid from  a  solution  containing  so  great  a  quantity  of  free  acid  ; 
but  while  at  times  it  may  be  advantageous  to  extract  caffein 
from  a  solution  having  an  acid  reaction,  in  no  instance  would 
there  seem  need  of  a  sufficient  amount  to  render  the  method 
inapplicable ;  further,  according  to  Knox  and  Prescott'  Gom- 
berg*s  method  becomes  uncertain  under  similar  conditions. 

Ill  the  experiments  just  quoted  ten  extractions  with  chloro- 
form yielded  but  34.85  per  cent,  of  the  total  caffein,  or  on  an 
average  each  treatment  removed  only  3.485  per  cent.,  while  the 
three  subsequent  treatments  removed  an  additional  13.33  per 
cent,  of  the  whole,  or  4.44  per  cent,  for  each  extraction,  i,  ^., 
although  the  total  substance  in  solution  had  been  decreased  by 
more  than  one  third  the  average  amount  given  up  to  chloroform 
increased  in  the  nth,  12th  and  13th  treatment;  while  in  the 
14th,  15th  and  i6th  but  1.735  per  cent.,  or  on  an  average  of 
0.578  per  cent,  for  each  shaking  was  obtained. 

Although  the  writer  had  never  attempted  a  caffein  determina- 
tion under  the  conditions  mentioned,  he  was,  from  theoretical 
considerations,  inclined  to  question  the  figures  given,  and 
accordingly  made  the  following  experiments. 

1  Proceedings  Am.  Pharm.  Ass.,  1896. 


980  ESTIMATION   OF   CAPPEIN. 

1. 0137  gi'ain  caffein,  rendered  anhydrous  by  keeping  in  a 
desiccator  over  sulphuric  acid  until  its  weight  remained  con- 
stant, was  dissolved  in  sixty  cc.  ten  per  cent,  sulphuric  acid  and 
shaken  successively  with  nine  portions  of  chloroform,  twenty- 
five  cc.  each  ;  the  chloroform  solutions  evaporated  sit  a  gentle 
heat  and  the  residue  dried  6ver  sulphuric  acid  to  constant 
weight. 

1st  portion  of  twenty-five  cc.  yielded  a  residue  of  0.5525  gram. 


and 
3rd 
4tli 
5th 
6th 
7th 
8th 
9th 


(i  (<  li 

tt  if  If 

tt  t<  it 

ft  f<  <( 

«i  ii  (( 

<(  (•  II 

II  <i  <i 


0.2514  " 

It        11        i<      «  0.1155  " 

II      I*  0.0535  ** 

II        II        II      II 


11      II  0.0114      •* 

It  II  II  li     M   <^nfiQ  (I 


0.0237 

o.oi  14 

0.0058 


n  II  II         II   0.0029         " 

II       .1       II     II  0.0015     " 


I.Ol82»       ** 

In  the  second  experiment  i.oooi  gram  anhydrous  ca£Fein  in 
sixty  cc.  ten  per  cent,  sulphuric  acid,  extracted  as  before,  with 
chloroform  in  proportions  of  twenty-five  cc.  each  : 


I  St,  2d  and  3rd  portions  gave  a  total  residue  of  0.9086  gram. 
4th,  5th  and  6th       **  ••     **     **  "         **    0.0854 

7th,  8th  and  9th       "  '*     **     "  '*         "  0.0134 


II 
II 


1.0074'     •* 

The  sulphuric  acid  used  in  Gomberg*s  experiments  was 
designated  as  '*  ( i:  10)  ''  by  which  it  is  presumed  an  acid  contain- 
ing ten  per  cent,  by  weight  of  sulphuric  acid  was  meant ;  since, 
however,  it  was  possible  that  sulphuric  acid  1:10  ^  volume 
was  the  strength  of  the  acid  used,  a  determination  was  made 
with  an  acid  with  such  concentration,  t.  e.,  ten  cc.  concentrated 
sulphuric  acid  mixed  with  water  enough  to  make  when  cold, 
100  cc.  In  sixty  cc.  of  this  were  dissolved  0.9790  gram  caffein 
and  extracted  with  chloroform  in  portions  of  ti^'enty-five  cc.  each 
as  before. 

1st,  2nd  and  3rd  portion  yielded  a  total  residue  of  0.6484  gram. 
4th,  5th  and  6th        '*  "        '•     **  "         "0.2222 

7th,  8th  and  9th        "  **        "     "  "         "0.0756 

loth,  nth,  1 2th,  13th,  14th,  15th  and  i6th  *'        **  0.0379 


It 
II 
It 


II 


0.9841* 

1  No  explanation  is  offered  to  account  for  the  plus  error  in  the  at>ovc.     Contamina- 
tion with  sulphuric  acid  was  suspected,  but  disproved. 


RUTHENOCVANIDE.  98 1 

As  was  to  be  expected,  this  confirms  in  a  general  way,  the 
statement  relative  the  difficulty  with  which  caffein  is  shaken  out 
of  solutions  containing  a  large  proportion  of  sulphuric  acid  ;  in 
no  way,  however,  does  it  agree  with  the  data  given  by 
Gomberg,  who  by  ten  successive  treatments  with  chloroform 
removed  only  34.85  per  cent.,  while  my  figures  show  that  when 
a  ten  per  cent,  sulphuric  acid  was  used,  with  but  three  extrac- 
tions, fully  ninety  per  cent,  was  recovered,  and  even  with  a  still 
stronger  acid  (1  +  9  by  volume),  three  portions  of  chloroform 
removed  about  sixty-five  per  cent. 

UiriVB&SITY  OF  ILLIVOI8.  SCHOOL  OP  PHiRMACY. 


CONTRIBUTION  TO  THE  KNOWLEDGE  OF  THE 

RUTH  ENOC  Y  AN  I DES. 

By  J  as.  I<bwib  Howb. 
Received  August  97.  1996. 

POTASSIUM  ruthenocyanide  was  described  by  Claus,  in 
1854,  in  his  *•  Beitrage  zur  Chemie  der  Platinmetalle." 
The  salt  was  formed  by  fusing  ammonium  rutheninitrosochlo- 
ride'  (tetrachloride  of  Claus)  with  potassium  cyanide.  The 
attempt  was  also  made  to  form  it  by  fusing  potassium  ferrocya- 
nide  with  ruthenium,  but  it  was  found  impossible  to  separate 
the  ferrocyanide  and  ruthenocyanide.  It  is  probable  that  some 
of  Claus'  experiments  were  carried  out  with  a  ruthenocyanide 
cx>ntaminated  with  ferrocyanide,  from  the  fact  that  he  describes 
copper  ruthenocyanide  as  brown,  whereas,  when  free  from  the 
ferrocyanide,  it  is  pale  green.  Potassium  ruthenocyanide  in 
reactions  and  crystallization  resembles  ver>'  closely  the  ferro- 
cyanide, except  that  when  pure  it  is  white.  Its  crystallography 
as  well  as  that  of  the  isomorphous  ferrocyanide  and  osmocya- 
nide  are  described  by  A.  Dufet.* 

Preparation  of  potassium  ruthenocyanide  for  the  purpose  of 
carrying  out  experiments  upon  it  not  yet  completed,  gave  occa- 
sion to  the  work  recorded  in  this  paper. 

In  the  Claus  method  of  preparation,  a  large  proportion  ol  the 
ammonium  rutheninitrosochloride  is  decomposed  with  separa- 
tion of  metallic  ruthenium,  and  while  a  part  of  the  ruthenocya- 

1  Joly  :  Compi.  rend.,  108,  854, 1889  ;   Howe  :  J.  Am.  Ckem.  Soc.,  16.  38S.  1S94. 
*  Compt.  rend.,  (189$)*  *ao.  377. 


982  JAS.    I.EWIS  HOWE. 

nide  formed  crystallizes  out  from  a  solution  of  the  melt,  in  large 
square  pseudorhombic  plates,  much  is  left  in  the  solution  and 
cannot  be  directly  separated  from  the  potassium  cyanide  and 
other  salts  present.  Attempts  were  therefore  made  to  use  other 
methods  of  formation  with  the  following  results  : 

1.  Potassium  rutheninitrosochloride,  K,RuCl^NO,  fused  with 
potassium  cyanide,  gave  rather  better  results  in  ruthenocyanide, 
there  being  rather  less  decomposition  than  was  the  case  with  the 
ammonium  salt. 

2.  Ruthenium  trichloride,  RuCl.,  fused  with  potassium  cya- 
nide gave  a  fair  product  of  ruthenocyanide. 

3.  Metallic  ruthenium,  fused  with  potassium  cyanide,  was 
slightly  acted  upon,  giving  a  trace  of  ruthenocyanide. 

4.  Metallic  ruthenium,  fused  with  potassium  cyanide  and  a 
little  potassium  hydroxide,  gave  rather  stronger  reaction  than 
case  3,  but  the  amount  of  ruthenocyanide  formed  was  very 
small. 

5.  The  melt  formed  by  fusion  of  ruthenium  in  potassium 
hydroxide  and  nitrate,  containing  potassium  ruthenate,  K.RuO^, 
was  dissolved  in  water  and  boiled  with  potassium  cyanide.  The 
deep  orange-red  solution  was  quickly  decolorized  and  the  ruthe- 
nium was  converted  into  ruthenocyanide  with  little  loss.  A 
considerable  proportion  could  be  obtained  in  the  usual  square 
crystals.  This  process  could,  by  modiQcation,  probably  be 
made  the  most  satisfactory  method  of  forming  the  ruthenocya- 
nide, presenting  one  decided  advantage  that  metallic  ruthenium, 
or  oxides,  can  be  used,  thus  avoiding  the  necessity  of  preparing 
the  nitrosochloride  or  chloride. 

6.  Ruthenium  trichloride  was  boiled  with  a  strong  solution  of 
potassium  cyanide.  The  ruthenocyanide,  crystallizing  in  the 
usual  square  form,  was  obtained,  but  very  much  contaminated 
with  a  greenish  by-product  not  yet  investigated,  probably 
analogous  to   Prussian  blue. 

7.  Potassium  rutheninitrosochloride  was  boiled  with  a  strong 
solution  of  potassium  cyanide.  The  solution  was  slowly  decol- 
orized, considerable  of  the  greenish  by-product  being  formed. 
From  this  solution  there  crystallized  thick  straw-coh>red  hexago- 


RUTHENOCYANIDE.  983 

nal  plates,  which  will  be  considered  further  on.     The  quantity 
of  the  product  is  not  satisfactor>\ 

8.  The  Weselsky  method*  of  forming  double  cyanides  was 
tried.  Hydrocyanic  acid  was  led  into  a  solution  of  the  nitroso- 
chloride,  in  which  barium  carbonate  was  suspended,  until  effer- 
vescence ceased.  The  solution  gave  no  reaction  for  ruthenocya- 
nide.  Its  color  had  changed  to  the  brown-yellow  of  the  trichlo- 
ride, but  gave  no  reaction  for  this  with  potassium  thiocyanate, 
or  with  ammonia  and  sodium  thiosulphate.  On  warming,  the 
solution  gelatinized  to  a  firm  hydrogel,  insoluble  in  hot  aqua 
regia,  but  soluble  in  boiling  potassium  hydroxide.  This  last 
solution  was  unchanged  on  acidification  with  hydrochloric  acid, 
and  gave  the  potassium  ferrocyanide  reaction  for  nitrosochlo- 
ride,  but  no  reaction  for  trichloride.  The  dried  jelly  was  easily 
explosive  on  heating.  It  presents  an  interesting  analogy  to 
Jackson's*  hydrogel  of  cobaltocyanide  and  is  being  further 
studied. 

9.  The  Weselskj'  method  was  also  applied  to  ruthenium  tri- 
chloride. The  merest  trace  of  ruthenocyanide  was  formed,  and 
the  solution,  little  changed  in  color,  no  longer  gave  reactions  for 
the  trichloride. 

10.  The  nitrosohydroxide  of  Joly,  formed  by  the  precipitation 
of  the  chloride  by  potassium  carbonate,  is  easily  soluble  in  potas- 
sium cyanide  and  converted  into  ruthenocyanide  by  prolonged 
boiling. 

The  following  reactions  of  ruthenocyanide  may  be  noted : 

No  precipitates  are  formed  with  the  caustic  alkaline  earths, 
their  ruthenocyanides  being  soluble  in  water. 

Lead  acetate  gives  a  fine  white  precipitate,  soluble  in  nitric 
acid. 

Silver  nitrate  gives  a  white  curdy  precipitate,  insoluble  in 
both  ammonia  and  in  nitric  acid. 

Ferric  chloride  gives  a  rich  purple  precipitate,  closely  resem- 
bling Prussian  blue  in  its  chemical  properties.  In  pure  water  it 
is  soluble,  but  is  precipitated  from  this  solution  by  salts  or  alco- 
hol.    It  forms  a  very  beautiful  and  intense  dye,  adhering  with 

1  WeseUky,  Sitzber.  Akad.  Wien.,  60,  ii.  (/870),  36t :  Ber.  d.  chem.  Ges.,  2, 588,  1869, 
s  Jackson  :  Ber.  d.  chem.  Gef.,  39,  1020,  iB^. 


984  JAS.    LEWIS   HOWE. 

great  persistence  to  cotton  fiber,  on  which  it  has  been  precipi- 
tated. It  is  decomposed  very  readily  by  alkalies  with  precipita- 
tion of  ferric  hydroxide,  re-forming,  however,  on  the  addition  of 
acids,  unaffected  by  dilute  acids,  but  permanently  decomposed 
by  strong  acids.  It  is  a  most  delicate  reaction  for  the  detection 
of  ruthenocyanide. 

Ferrous  sulphate  gives  a  pale  blue  precipitate,  which  gradually 
changes  to  the  purple  above  mentioned,  and  instantly  if  bromine 
water  is  added. 

Copper  sulphate  gives  a  very  pale  green  flocculent  precipitate 
(not  brown  as  given  by  Claus). 

With  salts  of  the  following  metals  precipitates  are  formed 
insoluble  in  hydrochloric  acid  :  Cadmium,  white  (soluble  in  hot 
acid);  zinc,  white;  tin  (both  stannous  and  stannic) ,  white ; 
mercury,  white ;  bismuth,  white  (insoluble  in  nitric  acid) ; 
nickel,  dirty  green  (changing  to  blue  with  hydrochloric  acid) ; 
cobalt,  pale  red  ;  platinum,  yellow-green ;  manganese  gives  a 
white  precipitate  soluble  in  hydrochloric  acid.  With  gold  there 
is  no  immediate  precipitate,  but  a  gradual  darkening  and  sepa- 
ration of  a  dark  precipitate,  the  solution  becoming  green. 

Bromine  water  changes  the  solution  to  a  dark  red,  which  does 
not  give  the  trivalent  ruthenium  reaction.  Iodine  also  seems  to 
alter  the  solution. 

No  reaction  with  hydrogen  sulphide,  ammonium  sulphide,  or 
thioacetic  acid. 

Nitric  acid  has  no  effect  in  the  cold,  but  when  heated  slightly 
reddens  the  solutions.  It  then  shows  no  signs  of  a  reaction 
analogous  to  that  of  the  nitroprussides. 

It  is  acted  on  by  potassium  nitrite  with  sulphuric  acid,  and 
when  neutralized  gives  a  fugitive  rose  red  with  ammonium  sul- 
phide. 

It  gives  no  apparent  reaction  with  ruthenium  trichloride  or 
nitrosochloride. 

Two  methods  of  purification,  applicable  to  such  portions  of  the 
ruthenocyanides  as  cannot  be  separated  by  crystallization,  may 
be  used.  The  most  satisfactory  is  the  precipitation  in  dilute 
solution  by  lead  acetate  and  thorough  washing  with  hot  water 
to  remove  any  lead  chloride  present.     Suspension  of  the  lead 


RUTHENOCYANIDE.  985 

ruthenocyanide  (carbonate,  cyanide,  etc.)  in  much  water  and 
decomposition  with  dilute  sulphuric  acid.  Filtration  and  addi- 
tion of  baryta  water  till  nearly  neutral  and  then  of  barium  car- 
bonate in  excess  ;  warming,  filtration,  and'  evaporation  to  crys- 
talUzation  of  the  barium  ruthenocyanide  from  which  other  ruthe- 
nocyanides  may  be  formed  by  double  decomposition. 

The  other  method  of  purification  which  is  applicable  especially 
to  all  residues,  is  precipitation  with  ferric  chloride  in  slightly 
acid  solution,  washing  with  acidified  water,  as  far  as  possible 
(the  purple  begins  to  dissolve  as  the  salts  are  washed  out)  and 
decomposing  with  baryta  water.  This  method,  while  very  use- 
ful for  recovery  of  residue,  does  not  give  so  pure  a  product  as 
the  first  method.  n 

The  hexagonal  crystals  described  above,  in  process  7,  pre- 
sented points  of  interest,  in  that  it  seemed  not  impossible  that 
they  contained  the  nitroso  group  of  the  nitrosochloride  from 
which  they  were  formed.  When  dissolved  in  water  they  showed 
every  reaction  of  the  ordinary  square  crystals  of  the  ruthenocya- 
nide, but  they  could  not  be  converted  into  the  square  form  by 
recrystallization  nor  could  their  yellowish  tint  be  removed.  The 
crystals  are  anhydrous  while  the  white  crystals  contain  three 
molecules  of  water  of  crystallization.  On  heating  they  explode 
with  considerable  violence  while  the  square  crystals  decompose 
very  gently.  On  recrystallization  they  show  prismatic  forms,  with 
many  twins  resembling  staurolite  crosses,  and  others  resembling 
aragonite  twins.  Though  perfectly  hexagonal  in  form,  they  do 
not  seem  to  belong  to  the  hexagonal  system.  After  conversion 
into  the  lead,  hydrogen,  barium,  and  back  into  the  potassium 
salt  by  the  first  method  of  purification  described,  and  further 
precipitation  of  this  potassium  salt  by  alcohol  and  recrystalliza- 
tion from  water,  crystals  were  obtained  which  were  square, 
white,  and  in  every  respect,  crystallographically  as  well  as 
chemically »  resembled  the  ordinary  potassium  ruthenocyanide. 
This  was  verified  by  analysis  of  the  barium  salt  and  partial  analy- 
sis of  the  potassium  salt. 

It  is  evident  that  the  hexagonal  crystals  are  not  a  nitrosocya- 
nide,  and  it  seems  possible  that  the  form  may  be  conditioned  by 


986  JAS.   LEWIS  HOWE. 

some  trace  of  impurity.     They  are  being  further  studied  at  pres- 
ent. 

ANALYSIS  OF  POTASSIUM  AND   BARIUM  RUTHENOCYANIDES. 

Potassium  ruthenocyanide,  K^Ru(CN),,3H,0,  formed  by  boil- 
ing a  solution  of  potassium  rutheninitrosochloride  with  potassium 
cyanide  ;  purified  by  conversion  through  lead,  hydrogen,  and 
barium  salts. 

Per  cent. 

I.  Loss  of  water  in  four  days  standing  over  sulphuric  acid.  10.84 

II.  "     •*     "      at  120^ 10.90 

III.  "     "     **      in  30  hours  standing  over  sulphuric  acid .  11.25 

Theory  for  K4Ru(CN)e,3H,0.  3H,0   =    11.53 

The  crystals,  especially  when  small,  are  so  efflorescent  that  it 
is  difficult  to  obtain  uneffloresced  salt  for  analysis,  and  the  fol- 
lowing are  calculated  for  the  dehydrated  salt. 

Theory  for 
K4Ru(CN),. 

Potassium 37-76 

Ruthenium 24.53 

This  corresponds  to  the  potassium  ruthenocyanide  described 
by  Claus. 

Barium  ruthenocyanide,  Ba,Ru(CN)„6H,0,  (new)  formed 
from  the  ordinary  form  of  the  potassium  salt. 

Pale  straw-colored,  diamond-shaped  (up  to  one-half  cm.  long) 
monoclinic  crystals,  or  larger  crystal  rosettes,  slightly  soluble 
in  cold,  more  easily  in  hot  water,  slowly  lose  water  of  crystalli- 
zation over  sulphuric  acid,  lose  five  and  a  half  molecules  of 
water  at  100°  but  retain  one-half  molecule  to  nearly  200*",  thus 
resembling  barium  ferrocyanide.  The  barium  ruthenocyanide 
from  the  hexagonal  form  of  the  potassium  salt  was  similar,  but 
was  not  obtained  in  well  enough  defined  crystals  to  identify  pos- 
itively with  the  preceding,  but  analysis  shows  the  constitution 
to  be  the  same. 

The  method  of  analysis  was  the  following :  The  salt  was 
heated  in  a  platinum  boat  (in  two  cases  porcelain  was  used  and 
attacked,  so  that  the  ruthenium  was  contaminated  by  silica — 
Analyses  I   and  V)  in   an    oxygen    current,  and    the    carbon 


I. 

37.22 

Found. 
XL. 

38.32 

in. 
37.28 

23.90 

34.22 

24.44 

RUTHENOCYANIDE. 


987 


dioxide  evolved  collected  in  an  absorption  apparatus.  The  pro- 
portion given  off  was  variable,  but  usually  a  little  more  than  five 
atoms.  The  boat  was  then  heated  in  a  hydrogen  current,  to 
reduce  the  oxide  of  ruthenium  formed.  The  boat  was  then 
placed  in  a  carbon  dioxide  apparatus  and  treated  with  hydro- 
chloric acid  and  the  remainder  of  the  carbon  dioxide  collected. 
The  barium  chloride  was  then  filtered  off  from  the  ruthenium  and 
determined  as  sulphate ;  the  ruthenium,  after  burning  the  filter 
paper  and  heating  in  a  hydrogen  current  in  a  porcelain  boat, 
was  estimated  as  the  metal.  It  was  not  found  possible  to  arrive 
at  any  agreement  in  different  analyses  as  to  the  loss  on  heating 
the  barium  salt  in  air,  or  oxygen,  or  subsequently  in  hydrogen. 
While  most  of  the  carbon  of  the  cyanogen  is  burned  to  carbon 
dioxide,  a  part  remains  as  barium  carbonate.  The  remainder 
of  the  barium  seems  to  fluctuate  between  oxide  and  peroxide, 
while  a  variable  portion  of  the  ruthenium  is  oxidized.  The 
analyses  show  conclusively  that  six  atoms  of  carbon  are  present 
in  the  salt  derived  from  the  nitrosochloride,  hence  one  cyanogen 
group  cannot  be  replaced  by  the  nitroso  group. 
The  results  of  several  analyses  are  as  follows  : 


Theory  From 

for  RuCli 

Ba«Ru  and 

(CN),.  KCN 

6M,0.  sol. 


Found. 


From 
RuClaNO 

fusion. 


From    • 
RuCIsNO 

by 
solution. 


I.  11. 

Barium 42.90      

Ruthenium 15.86    16.41     15.67 

(with  SiO,) 

5JH,0  (ioqO) 15.67  15.67    15.63 

6H,0  (200°) 16.83  16.68    16.85 

5C 9.37      

6C 11.23      

C  from  combustion      

^  in  residue  •••••••     •■••       •■••       ■• 

xotsl  csroon ••>••■     •••■      •••«      •• 

WaSHINGTOIT  and  LEB  UKIVBRStTV. 

I«BXiifCTON,  Va.,  June.  1896. 


III.  IV.  v.  VI. 

—    42.46   43.32    42.27 
....     15.72    17.92    15.80 

(with  SiO,) 

•  • • •       X  ^ • Oo         ••■•  •••• 

...     16.69    16.71     16.56 


VII. 

42.54 
1560 

15.53 
16.68 


9.91      9.16 

2.65     2.03 

12.56    1 1. 19 


9.96 


9.98 

2.30 
12.28 


DIPYRIDINE    METHYLENE    IODIDE   AND  THE  NON-FOR- 
MATION OF  THE  CORRESPONDING  MONOPYRIDINE 

PRODUCTS.' 

By  S.  H.  Babr  and  A.  B.  Prbscott. 
Received  Sepceniber  %  i4p6. 

THE  addition  compound  of  pyridine  and  methylene  iodide 
was  formed  in  different  ways,  varying  the  conditions  of 
mass,  temperature,  pressure,  and  time,  as  follows.  The  method 
of  preparation  recommended  is  that  of  No.  V. 

Preparation  I, — Pyridine  and  methylene  iodode  in  equimolec- 
ular  proportions,  reacting  at  laboratory  temperature,  for  two 
days,  form  a  dark-red  crystalline  mass.  This  was  washed  in 
cold  alcohol,  which  does  not  dissolve  it. 

Preparations  11  and  III, — The  same  proportions  (those  of  a 
monopyridine  product)  were  taken  in  reaction  at  120°  C.  The 
methyl  iodide  for  I  and  II  was  colored  with  free  iodine,  that  for 
III  was  obtained  colorless  by  distillation  in  vacuum.  In  each 
case  the  crystals,  washed  with  cold  alcohol,  were  dark-red.  This 
color  was  not  afifected  by  treating  the  crystals  with  thiosulphate 
solution,  and  therefore  not  due  to  free  iodine  or  to  periodides. 

Preparation  IV, — By  reaction  of  colorless  methylene  iodide, 
in  the  same  proportions,  with  the  pyridine,  but  without  heat,  an 
orange  precipitate  settles  slowly .     This  was  washed  as  the  others. 

Preparation  V, — Pyridine  of  boiling  point  118*  C,  and 
methylene  iodide  either  colorless  or  tinged  with  iodine,  in  abont 
equal  molecular  quantities,  are  placed  in  a  flask,  alcohol  in 
volume  equal  to  the  two  reacting  materials  is  added,  a  return- 
condenser  adjusted,  and  the  heat  of  a  water-bath  applied  for  an 
hour.  On  cooling,  long  yellow  needles  separate  out.  To  purify 
further,  dissolve  in  hot  fifty  per  cent,  alcohol,  cool,  and  add  a 
little  ether,  when  fine  crystals  are  formed. 

So  obtained,  the  product  is  in  fine  needles,  of  yellow  color, 
decomposing,  not  melting,  at  220"*  C. ;  soluble  in  water,  from 
which  it  crystallizes  at  0°  C. ;  insoluble  in  cold  alcohol,  spar- 
ingly soluble  in  hot  alcohol ;  insoluble  in  ether,  or  chloroform, 
or  benzene,  or  amyl  alcohol ;  sparingly  soluble  in  methyl  alcohol. 

Analysis  gave  us  percentages  as  follows : 

Calculated  for  Found. 

(C»H|N),CH«I,.       I.  II.  III.  IV.  V. 

I    59.61  58.95         58.91  57.98         58.4  59.6 

N».« ^'57  ••••  ....  ....  ..•  0.73 

1  Rend  at  the  Buffalo  meeting  of  the  American  Asaociation  for  the  Advancement  of 
Science. 


DIPYRIDINE   METHYLENE   IODIDE.  989 

The  product,  therefore,  not  quite  pure  in  the  first  four  experi- 
mental preparations,  is  substantially  the  same  under  the  differ- 
ent conditions  employed,  and  with  whatever  excess  of  the  diiodo- 
methane,  is  always  the  dipyridine  addition  compound.  And  its 
formula,  agreeing  with  those  of  its  bromine  homologues,*  maybe 
confidently  written,  to  express  the  relations  of  the  methylene 
group  and  the  halogen  atoms  :^ 

^CH— CH.         .CH,.        ^CH=CH. 

^CH=CH^      ^I     V       ^CH— CH^ 

Kleine  found*  that  trimethylamine,  in  combination  with  dihalo- 
gen  substituted  hydrocarbons,  forms  both  the  monammonium 
and  the  diammonium  products,  the  former  prevailing,  especially 
when  there  are  not  more  than  two  atoms  of  carbon  in  thehalide. 

It  seemed  now  desirable  to  subject  pyridine  to  various  condi- 
tions of  additive  reaction  with  various  dihalides,  in  order  to 
know  whether  it  can  in  au}*^  case  form  such  monamine  com- 
pounds as  the  fatty  amines  sometimes  form.'  Pyridine  and 
ethylene  bromide,  in  equal  molecular  proportions,  were  digested 
together  in  a  sealed  tube  for  two  weeks,  when  the  entire  content, 
a  crystalline  mass,  was  dissolved  in  hot  alcohol  of  ninety-five  per 
cent.,  and.  fractionally  crystallized  in  successive  crops,  washing 
each  with  cold  absolute  alcohol.  These  crops  of  crystals  gave, 
of  bromine,  respectively,  46.15,  46.15,  46.18,  and  46.03  percent., 
the  calculated  per  cent,  in  (C^HfcN),C,H^Br,  being  46.21. 

Next  pyridine  with  excess  of  ethylene  bromide  was  digested 
in  a  pressure  flask,  in  water-bath,  with  agitation.  T^e  crystal- 
lized product  gave  46.26  per  cent,  of  bromine.  Finally  dipyri- 
dine ethylene  bromide  was  heated  with  excess  of  ethylene  bro- 
mide in  a  sealed  tube  to  170*"  C.  .  There  was  some  charring  in 
the  mixture.  By  recrystallizing  from  it  a  product  was  obtained 
which  gave  45.84  per  cent,  of  bromine. 

Dipyridine  ethylene  bromide  crystallizes  in  colorless  plates, 
insoluble  in  ether,  and  melting  with  decomposition  at  295°  C. 

UitiVERSiTV  OP  Michigan. 

1  The  ethylene  bromide,  (Hofmann)  Davidson,  1861 :  Proc,  Roy.  Soc.^  pa?e  a6i  ;  The 
trimethylene  bromide,  Plintennann  and  Prescott,  1895  :J,  Am.  Chem,  Soc.,  i8.  28. 

*G.  Kleine,  1894:  Chem.  Centrdi.,  page  161. 

s  This  in  continuation  of  the  inquiry  of  Plintermann  and  Pretcotti  1895:  J.  Am.  Chem. 
Sof.,  18,  33. 


[Contribution  prom  thb  John  Harrison  Laboratory  of  Chemistry. 

No.  13.] 

DETERMINATION  OF  THE  ATOMIC  HASSBS  OF  SILVER. 
HERCURY  AND  CADMIUM  BY  THE  ELECTRO- 
LYTIC METHOD.' 

By  Wzllett  Lbplby  Hahdin. 
Received  September  a6, 1896. 

INTRODUCTION. 

A  glance  at  the  literature  on  the  determinations  of  the 
atomic  masses  of  silver,  cadmium  and  mercury  will  show 
that,  with  the  exception  of  cadmium,  the  electrolytic  method  has 
not  been  tried.  A^ide  from  the  fact  that  certain  errors  involved 
in  the  washing  and  drying  of  the  precipitates  are  eliminated  by 
this  method,  its  simplicity  at  once  gives  it  preference  over  the 
usual  methods  of  gravimetric  determinations.  Inasmuch  as 
these  three  metals  are  completely  precipitated  from  certain  of 
their  solutions  by  the  electric  current,  and  as  it  is  desirable  to 
determine  the  atomic  mass  of  any  element  by  different  meth- 
ods, it  was  thought  advisable  to  apply  this  method  in  a  redeter- 
mination of  the  atomic  masses  of  these  elements. 

GENERAL  CONSIDERATIONS. 

Before  taking  up  the  different  metals  separately,  the  following 
general  considerations  may  be  mentioned  : 

1.  A  careful  preliminary  study  was  made  in  the  selection  of 
compounds.  Some  compounds,  which  from  a  theoretical  stand- 
point seemed  to  offer  certain  advantages,  were  found  by  experi- 
ment not  to  meet  the  requirements  of  exact  determinations. 
Salts  which  can  be  sublimed  were  used  whenever  possible  ;  and 
in  all  cases  only  those  salts  were  used  which  form  well  defined 
crystals. 

2.  All  reagents  used  were  either  prepared  or  purified  by  my- 
self and  carefully  tested  for  impurities. 

3.  The  metals  were  deposited  in  platinum  dishes  of  about  200 
cc.  capacity  and  about  sixty-five  grams  in  weight.  When  the 
precipitation  was  complete,  before  interrupting  the  current,  the 

1  Prom  the  author's  thesis  presented  to  the  Faculty  of  the  Uaiversity  of  Pennsylvt- 
nia  for  the  degree  of  Ph.D.,  1896. 


ATOMIC  MASSES  OP  SILVERi  MERCURY  AND  CADMIUM.       991 

solution  was  siphoned  from  the  platinum  dish,  pure  water  being 
added  at  the  same  time  ;  this  was  continued  until  the  solvent 
used  was  completely  removed  from  the  dish.  The  current  was 
then  interrupted  and  the  deposit  washed  several  times  with  boil- 
ing water,  with  the  hope  of  removing  any  occluded  hydrogen. 
After  drying,  the  dishes  were  placed  in  a  vacuum  desiccator 
over  anhydrous  calcium  chloride  and  allowed  to  remain  in  the 
balance  room  until  their  temperature  was  the  same  as  that  of  the 
room.  Atmospheric  dust  was  excluded  from  the  platinum 
dishes  during  the  process  of  deposition  by  means  of  two  glass 
plates  which  formed  a  complete  cover ;  the  moisture  which  col- 
lected on  this  cover  was  washed  back  into  the  dish  from  time  to 
time.  The  dishes  were  handled  with  nickel  tongs  tipped  with 
rubber. 

4.  The  balance  used  was  made  expressly  for  this  work  by 
Henry  Troemner,  of  Philadelphia.  The  beam  and  pans  were 
made  of  aluminum,  the  beam  being  about  twenty  centimeters 
long.  The  framework  was  plated  with  gold  to  prevent  corro- 
sion. The  sensibility  for  different  loads  and  the  ratio  of  the 
length  of  the  two  arms  were  carefully  determined.  The  balance 
is  sensitive  to  the  fortieth  of  a  milligram,  and  the  sensibility  is 
almost  independent  of  the  load  up  to  seventy-five  grams.  The 
difference  in  the  length  of  the  two  arms  is  so  slight  that  no  cor- 
rection need  be  applied.  The  balance  was  kept  in  a  large  quiet 
room  of  nearly  constant  temperature. 

The  larger  weights  used  were  made  of  brass  and  the  fractions 
01  a  gram  made  of  platinum.  The  weights  were  all  previously 
compared  against  each  other  and  standardized  with  reference  to 
the  largest  weight.  The  small  corrections  found  in  comparing 
them  were  tabulated  and  applied  to  all  results.  The  weighings 
were  made  by  the  method  of  oscillations.  The  temperature  and 
barometic  pressure  were  noted  at  the  time  of  each  weighing, 
and  all  weighings  were  reduced  to  a  vacuum  standard.  As  the 
density  of  the  atmosphere  at  the  time  of  weighing  the  empty 
platinum  dish  was  different  from  that  at  the  time  of  weighing 
the  dish  and  deposit  together,  the  following  formula  was  applied 
to  obtain  the  weight  of  the  deposit  in  vacuo  : 


992  WII.LETT  LEPtEY   HARDIN. 

weight  of  dish ( i  +  -^ 7~) 

Weight  of  (dish  +  deposit) ^l ii — 

I      A-  A 

'  +  — / 

I  +  -^ j\  =  weight  of  deposit  in  vacuo. 

Where  X  =  density  of  air  at  the  time  of  weighing  the  empty 

dish. 
A.'  =  density  of  air  at  the  time  of  weighing  the  dish  + 

deposit. 
J  =  density  of  platinum  dish. 
J'  =  density  of  metallic  deposit. 
/  =  density  of  weights. 
As  the  weights  were  all  standardized  with  reference  to^  the 
hundred-gram  brass  weight,  it  is  evident  that  they  must  all  be 
calculated  as  having  the  same  density,  equal  to  that  of  brass. 

5.  The  atomic  masses  of  the  different  elements  involved  in 
the  calculation  of  results  were  taken  from  Clarke's  latest  report.^ 

PART  I. 

DETERMINATION   OF  THE   ATOMIC   MASS  OF   SILVER. 

The  mean  of  all  the  earlier  determinations,  as  calculated  by 
Clarke,  gives  107.923  for  the  atomic  mass  of  silver;  a  result 
almost  identical  with  the  mean  (107.93  ;  O  =  16)  of  the  deter- 
minations of  Stas. 

PREPARATION  OF  PURE  METALLIC  SILVER. 

The  silver  used  in  this  work  was  purified  by  the  Stas  method. 
Two  hundred  grams  of  silver,  about  ninety-nine  per  cent,  pure, 
were  dissolved  in  dilute  hot  nitric  acid.  The  solution  was 
evaporated  to  dryness,  the  nitrate  heated  to  fusion  and  main- 
tained in  a  fused  condition  until  the  oxides  of  nitrogen  were  no 
longer  evolved.  The  residue,  after  cooling,  was  dissolved  in  as 
little  cold  water  as  possible,  and  after  standing  forty-eight  hours 
the  solution  was  filtered  through  a  double  filter  to  remove  any 
suspended  matter.  The  clear  solution  was  then  diluted  with 
thirty  times  its  volume  of  distilled  water,  and  to  it  was  added  an 

ly.  Am.  Chem.  Soc.,  x8,  197. 


ATOMIC  MASSES  OF  SILVER,  MERCURY  AND  CADMIUM.      993 

excess  of  pure  hydrochloric  acid.  The  silver  chloride  which 
separated  was  allowed  to  subside  and  was  then  thoroughly 
washed  by  decantation,  at  first  with  water  containing  a  little 
hydrochloric  acid,  and  finally  with  pure  water.  The  precipitate 
was  then  collected  on  a  cheese  cloth  filter,  pressed  strongly  and 
allowed  to  dry.  When  perfectly  dry,  the  silver  chloride  was 
powdered  finely  and  digested  for  three  days  with  aqua  regia  ;  it 
was  then  thoroughly  washed  by  decantation  with  distilled  water. 
After  obtaining  the  pure  chloride  of  silver,  it  was  necessary  to 
purify  the  caustic  potash  and  milk  sugar  used  in  reducing  the 
chloride  to  the  metallic  state .  The  caustic  potash  was  heated  to  the 
boiling  point  and  to  it  was  added  a  concentrated  solution  of  po- 
tassium sulphide  to  precipitate  any  heavy  metals  which  might  be 
present.  The  solution  was  filtered  and  the  filtrate  digested  for 
some  time  with  freshly  precipitated  silver  oxide  and  again  fil- 
tered to  remove  the  excess  of  potassium  sulphide.  The  milk 
sugar  was  purified  in  a  similar  manner.  The  silver  chloride  was 
then  placed  in  large  porcelain  dishes  and  covered  with  a  solution 
of  caustic  potash  and  milk  sugar.  The  dishes  were  placed  on  a 
water-bath  and  heated  to  a  temperature  of  yo'^-So®  until  the 
reduction  to  finely  divided  metallic  silver  was  complete.  The 
alkaline  solution  was  then  poured  off,  and  the  gray  metallic  sil- 
ver was  washed  with  distilled  water  until  the  alkaline  reaction 
disappeared.  The  metal  was  then  digested  with  pure  dilute 
sulphuric  acid,  and  finally  washed  with  dilute  ammonia  water. 
The  silver  thus  obtained  was  mixed,  when  dry,  with  five  per 
cent,  of  its  weight  of  fused  borax  containing  ten  per  cent,  of  pure 
sodium  nitrate.  The  mixture  was  fused  in  a  clay  crucible  and 
the  silver  poured  into  a  mold.  The  metal  obtained  in  this  way 
was  almost  snow  white  in  appearance,  and  dissolved  completely 
in  nitric  acid  to  a  colorless  solution. 

PREPARATION  OF  PURE  NITRIC  ACID. 

To  obtain  pure  nitric  acid,  one-half  liter  of  the  commercial 
C.  P.  acid  was  mixed  with  an  equal  volume  of  concentrated  C. 
P.  sulphuric  acid  and  distilled  from  a  retort  provided  with  a 
knee  tube  and  condenser.  The  first  portion  of  the  dis- 
tillate Was  rejected.     The  process  was  stopped  when  half  of  the 


994  WII^LETT  LEPLEY  HARDIN. 

nitric  acid  present  had  been  distilled  over.  The  distillate  was 
mixed  with  an  equal  volume  of  pure  sulphuric  acid  and  redis- 
tilled. The  second  distillate  was  collected  in  a  flask,  the  mouth 
of  which  was  closed  with  glass  wool.  When  the  process  was 
complete,  the  flask  was  closed  with  a  doubly  perforated  cork 
and  placed  in  a  water-bath  at  a  temperature  of  40**.  A  current 
of  pure  dry  air  was  then  conducted  through  the  acid  to  remove 
any  oxides  of  nitrogen.     The  acid  was  kept  in  a  dark  place. 

EXPERIMENTS  ON  SII^VER  OXIDE. 

If  pure,  dry  silver  oxide  could  be  prepared,  the  atomic  mass  of 
silver  could  be  compared  directly  with  that  of  oxygen.  A  large 
number  of  experiments  were  made  on  this  compound  with  the 
hope  of  determining  the  ratio  of  the  atomic  masses  of  these  two 
elements. 

PREPARATION  OF  SILVER   OXIDE. 

A  portion  of  the  pure  metallic  silver  was  dissolved  in  pure 
dilute  nitric  acid  and  the  solution  evaporated  to  crystallization. 
The  crystals  of  silver  nitrate  were  dissolved  in  pure  water  and  to 
the  solution  was  added  a  solution  of  pure  sodium  hydroxide, 
prepared  by  throwing  pieces  of  metallic  sodium  on  distilled 
water  in  a  platinum  dish.  The  twenty-five  grams  of  silver 
oxide  prepared  in  this  way  were  washed  by  decantation  with 
twenty  liters  of  water.  The  material  was  then  dried  at  the 
ordinary  temperature,  after  whicn  it  was  finely  powdered  and 
dried  for  twenty-four  hours  in  an  air-bath  at  100°.  The  oxide 
was  kept  in  a  weighing  tube  in  a  dark  place. 

Several  analyses  were  made  by  dissolving  a  weighed  portion 
of  the  material  in  pure  potassium  cyanide,  electrolyzing  the  solu- 
tion and  weighing  the  resulting  metallic  silver.  The  observa- 
tions invariably  gave  less  than  ninety-five  for  the  atomic  mass  of 
silver.  The  oxide  was  redried  at  a  temperature  of  125*  and 
analyzed  as  before,  but  the  quantity  of  silver  obtained  was  far 
below  that  calculated  for  the  compound  Ag,0.  Observations 
were  also  made  on  material  dried  at  140®  and  150**.  The  results 
showed  that  it  was  impossible  to  prepare  the  silver  oxide  in  a 
pure,  dry  condition. 


ATOMIC  MASSES  OF  SILVER,  MERCURY  AND  CADMIUM.     995 

After  making  these  observations,  my  attention  was  called  to 
an  article  by  M.  Carey  Lea,*  in  which  were  given  the  results  of 
a  series  of  analyses  of  silver  oxide  dried  at  different  tempera- 
tures varying  from  lOO**  to  170°.  These  observations  prove  con- 
clusively that  oxygen  is  given  off  at  a  much  lower  temperature 
than  that  required  to  remove  the  last  traces  of  moisture.  From 
these  observations  and  the  results  obtained  by  myself,  it  was 
evident  that  any  further  attempt  to  determine  the  atomic  mass 
of  silver  from  the  oxide  would  be  useless. 

Although  no  careful  study  was  made  as  to  the  nature  of  this 
compound,  it  might  be  added  that,  from  my  own  observations, 
it  seems  very  probable  that  the  oxide  contains  some  hydrogen  in 
the  form  of  hydroxyl. 


FIRST  SERIES. 


EXPERIMENTS  ON  SILVER  NITRATE. 

The  nitrate  of  silver  seems  to  fulfil  the  conditions  necessary 
for  accurate  analyses,  inasmuch  as  it  is  stable  and  crystallizes 
in  well  defined  crystals  which  can  be  fused  without  decomposi- 
tion. 

PREPARATION  OF  SILVER  NITRATE. 

The  material  used  in  these  experiments  was  prepared  by  dis- 
solving pure  silver  in  pure  aqueous  nitric  acid  in  a  porcelain 
dish.  An  excess  of  silver  was  used,  and  after  complete  satura- 
tion the  solution  was  poured  off  from  the  metal  into  a  second 
dish  and  evaporated  to  crystallization.  The  perfectly  transpar- 
ent, rhombic  plates  of  silver  nitrate  which  separated  were  dis- 
solved in  pure  water  and  recrystallized.  The  crystals  were  then 
carefully  dried,  placed  in  a  platinum  crucible  which  rested  in  a 
larger  platinum  dish  and  gradually  heated  to  fusion.  After 
cooling,  the  perfectly  white  opaque  mass  was  broken  up  and 
placed  in  a  ground-glass  stoppered  weighing  tube  and  kept  in  a 
desiccator  in  a  dark  place. 

MODE  OF  PROCEDURE. 

The  platinum  dish  in  which  the  deposit  was  made  was  care- 

^  Am.J.  Set'.,  44.  240. 


99^  WILLBTT  I^EPtEY  HARDIN. 

fully  cleaned  with  nitric  acid  and  dried  to  constant  weight.  It 
was  then  placed  in  a  desiccator  over  anhydrous  calcium  chloride, 
and  this,  together  with  the  desiccator  containing  the  tube  of 
silver  nitrate,  was  placed  in  the  balance  room,  where  they  were 
allowed  to  remain  until  their  temperatures  were  the  same  as  that 
of  the  room.  After  weighing  the  platinum  dish,  the  tube  of  sil- 
ver nitrate  was  weighed  and  part  of  the  salt  removed  to  the  dish, 
after  which  the  tube  was  reweighed.  The  difference  in  the  two 
weighings,  of  course,  represented  the  weight  of  silver  nitrate 
used  in  the  experiment.  Enough  water  to  dissolve  the  nitrate 
was  added  to  the  dish,  and  then  a  solution  of  potassium  cya- 
nide, made  by  dissolving  seventy-five  grams  of  pure  potassium 
cyanide  in  one  liter  of  water,  was  added  until  the  silver  cyanide 
first  formed  was  completely  dissolved.  The  dish  was  then  filled 
to  within  a  quarter  of  an  inch  of  the  top  with  pure  water  and  the 
solution  electrolyzed  with  a  gradually  increasing  strength  of 
current.  The  following  table  will  show  the  strength  of  current 
and  the  time  through  which  it  acted  : 

Time  of  action.  Strength  of  current. 

2  hours N.Djgo  =  0.015  amperes. 

4      "      N.D,oo  =  o.030 

6      "      N.Dioo  =  o.o75 

4      *•      N.D,oo  =  o.i5o 

4      "      N.Dioo  =  o.4oo 

By  gradually  incre^ing  the  strength  of  current  in  this  way 
the  silver  came  down  in  a  dense,  white  deposit.  When  the  depo- 
sition was  complete,  before  interrupting  the  current,  the  liquid  was 
siphoned  from  the  dish,  pure  water  being  added  at  the  same  time. 
This  was  continued  until  the  cyanide  was  completely  removed. 
The  dish  with  the  deposit  was  washed  several  times  with  boiling 
water  and  carefully  dried.  It  was  then  placed  in  a  desiccator 
and  aHowed  to  remain  in  the  balance  room  until  its  temperature 
was  the  same  as  that  of  the  room,  when  it  was  reweighed. 

Weight  of  platinum  dish  =  71.27302  grams. 

Weight  of  silver  nitrate  =  0.31 198  grams. 

Temperature,  22*^. 

Barometric  pressure,  770  mm. 

Weight  of  platinum  dish  +  silver  deposit  =  71.47104  grams. 


(I 
<< 


ATOMIC  MASSES  OF  SILVSR,  MERCURY  AND  CADMIUM.     997 

Temperature,  22". 
Barometric  pressure,  760  mm. 
Density  of  silver  nitrate  =  4.328. 

**  brass  weights  =  8.5. 

"  platinum  dish  =  21.4. 

**  metallic  silver  =  10.5. 

*'  atmosphere  at  the  time  of  weighing  the  empty  dish 
and  silver  nitrate  =  0.0012 12. 

*'  atmosphere  at  the  time  of  weighing  the  platinum 
dish  +  silver  deposit  =  0.001196. 
Computing  on  this  basis  we  have  the  following  : 

^  /     ,    0.001212       0.001212  \  .  ,        < 

o.3ii98^iH -— g — — j   =  0.31202  =  weight   of 

AgNO,  in  vacuo. 


cc 
( I 

CI 

i( 

C  ( 


71.27302 


_       ,     O.OOI2I2  O.OOI2I2_ 

n + n 

21.4 8.5 

,    O.OOII96         O.OOII96 
L  I  +- 


=  71.27291  =  weight  of 


21.4  8.5 

platinum  dish  at  22^  and  760  mm. 

71.47104  —  71.27291  =  0.19813  =  weight  of  deposit  at  22**  and 

760  mm. 

^      /     ,  0.001196       o.ooii96\  -  •  u^    /  J 

0.198131  iH r — ^— )  =0.19812  =  weight  of  de- 

^     *^\  10.5  8.5       /  ** 

posit  tn  vacuo. 

Taking  0=  16  and  N  =  14.04,  the  atomic  mass  of  silver  = 
0.19812  X  62.04    =  ,  . 

(31202  —  I98I2) 

Ten  observations  on  silver  nitrate  computed  in  the  foregoing 
manner  are  as  follows  : 

Atomic  mass 
WeiflTbt  of  AgNO*.  Weight  of  Ag.  of  silver. 

Gram.  Gram. 

1  0.31202  O.I9812  107.914 

2  0.47832  0.30370  107.900 

3  0.56742  0.36030  107.923 

4  0.57728  0.36655  107.914 

5  0.69409  0.44075  107.935 

6  0.86367  0.54843  107.932 

7  0.8681 1  0.55130  107.960 


99^  WItLETT  LEPI^EY   HARDIN. 


Weight  of  AgNOt. 

Weight  of  Ag. 

Atomic  mass 

Gram. 

Gram. 

of  silver. 

8 

0.93716 

0.59508 

107.924 

9 

1.06170 

0.67412 

107.907 

TO 

1. 19849 

0.76104 

107.932 

Mean 

=  107.924 

Maximum 

—  107.960 

Minimum 

=  107.900 

Difference         =     0.060 
Probable  error  =  ±0.005 

Computing  the  atomic  mass  of  silver  from  the  total  quantity 
of  material  used  and  metal  obtained,  we  have  107.926. 


SECOND  SERIES. 


EXPERIMENTS  ON  SILVER  ACETATE. 

The  fact  that  silver  forms  well  crystallized  salts  with  a  num- 
ber of  organic  acids  makes  the  comparison  of  the  atomic  mass  of 
silver  with  the  combined  atomic  masses  of  carbon,  hydrogen, 
and  oxygen,  a  matter  of  no  great  difficulty.  From  certain  pre- 
liminary experiments,  the  acetate  of  silver  seemed  to  fulfill  the 
conditions  necessary  for  accurate  determinations. 

PREPARATION  OP  SIIyVER  ACETATE. 

The  purest  commercial  sodium  acetate  was  dissolved  in  water, 
the  solution  filtered  and  recrystallized.  After  three  crystalliza- 
tions the  material  was  dissolved  in  pure  water,  and  to  the  rather 
concentrated  solution  was  added  a  solution  of  silver  nitrate,  pre- 
pared in  the  manner  already  indicated.  The  white  curdy  pre- 
cipitate which  separated,  after  washing  with  cold  water,  was 
dissolved  in  hot  water,  the  solution  filtered  and  evaporated  to 
crystallization.  The  silver  acetate  separated  in  brilliant  sword- 
shaped  crystals.  After  pouring  off  the  solution  the  crystals 
were  quickly  rinsed  with  cold  water  and  placed  between  filters 
to  remove  the  adhering  moisture.  The  material  was  allowed  to 
remain  in  contact  with  the  filters  only  for  a  short  time.  It  was 
then  placed  in  a  platinum  dish,  and  when  apparently  drj'  the 
crystals  were  broken  up  into  a  finely  divided  condition  and  dried 
forty-eight  hours  in  a  vacuum  desiccator.  This  work  was  car- 
ried on  in  a  darkened  room,  and  the  silver  acetate  obtained  was 


ATOMIC  MASSES  OF  SILVER,  MERCURY  AND  CADMIUM.     999 


placed  in  a  weighing  tube,  and  kept  in  a  desiccator  in  a  dark 
place. 

MODE  OP   PROCEDURE. 

The  method  of  operation  was  similar  to  that  described  under 
silver  nitrate.  After  weighing  the  silver  acetate,  its  aqueous  or 
cyanide  solution  was  electrolyzed  and  the  weight  of  the  result- 
ing metallic  silver  determined.  The  results  obtained  from  the 
aqueous  solution  were  sometimes  vitiated  by  the  separation  of 
silver  peroxide  at  the  anode.  To  prevent  this,  potassium  cya- 
nide was  sometimes  added.  The  results,  however,  from  the  two 
solutions  were  practically  the  same  when  no  peroxide  separated. 
Prom  the  aqueous  solution  the  silver  was  deposited  in  a  crystal- 
line form.  The  strength  of  current  and  time  of  action  were  the 
same  as  for  silver  nitrate. 

Ten  observations  on  silver  acetate  reduced  to  a  vacuum  stand- 
ard on  the  basis  of 

3.241  =  density  of  silver  acetate, 
10.5      =         **  metallic  silver, 

24.4      =         '*  platinum  dish, 

8.5      =         *'  weights, 

and  computed  for  the  formula  AgC,H,0„  assuming  the  atomic 
masses  of  carbon,  hydrogen  and  oxygen  to  be  12.01,  1.008  and 
16,  respectively,  are  as  follows  : 


Weight  of  ARCfHsO,. 

Weif  bt  of  Ag. 
Gram. 

Atomic  mass  of 

Grams. 

silver. 

I 

0.32470 

0.20987 

107.904 

2 

0.40566 

0.26223 

107.949 

3 

0.52736 

0.34086 

107.913 

4 

0.60300 

0.38976 

107.921 

5 

0.67235 

0.43455 

107.896 

6 

0.72452 

0.46830 

107.916 

I 

0.78232 

0.50563 

107.898 

0.79804 

0.51590 

107.963 

9 

0.92  lOI 

0.59532 

107.925 

10 

1.02495 

0.66250 

107.923 

Mean 

=  107.922 

Maximum 

=  107.963 

Minimum 

=  107.896 

Difference  =     0.067 
Probable  error  =  ±      0.005 

Computing  from  the  total  quantity  of  material  used  and  metal 

obtained  we  have  107.918  for  the  atomic  mass  of  silver. 


lOOO  WILLETT  LEPLEY   HARDIN. 

EXPERIMENTS  ON  SILVER  SUCCINATE. 

Silver  succinate  was  prepared  in  a  manner  similar  to  that  of 
silver  acetate.  The  commercial  C.  P.  succinic  acid  was  recr>'s- 
tallized  three  times  ;  the  ammonium  salt  was  then  prepared  and 
its  aqueous  solution  precipitated  with  a  solution  of  pure  silver 
nitrates.  The  precipitate  of  silver  succinate  was  thoroughly 
washed  bj*  decantation  with  pure  water  and  carefully  dried. 
After  drying  for  several  hours  in  an  air-bath  at  50**,  the  material 
was  ground  in  an  agate  mortar  to  a  finely  divided  powder,  and 
was  then  redried  for  twenty-four  hours  in  an  air-bath  at  a  tem- 
perature of  60**.  The  white  powder  obtaine;d  in  this  way  was 
placed  in  a  weighing  tube  and  kept  in  a  desiccator. 

The  method  of  analysis  was  similar  to  that  of  silver  acetate. 
A  weighed  portion  of  the  material  was  dissolved  in  a  little  potas- 
sium cyanide  in  a  platinum  dish.  After  diluting  wath  pure 
water,  the  solution  was  electrolyzed  and  the  resulting  deposit 
weighed.  The  strength  of  current  and  time  of  action  were  the 
same  as  for  silver  nitrate.  The  results  computed  for  the  formula 
C^H^O^Ag,  were  not  constant,  and  were  invariably  from  one  to 
two  units  lower  than  those  obtained  from  silver  nitrate  and  sil- 
ver acetate.  The  material  was  then  dried  at  a  temperature  of 
75®,  but  the  results  obtained  were  not  satisfactor>^ 

The  two  most  probable  causes  for  these  low  results  are  : 

First,  the  difficult j"^  of  removing  the  last  traces  of  impurities 
from  a  precipitate  like  that  of  silver  succinate.  The  experience 
throughout  this  work  has  been,  that,  to  remove  all  the  impuri- 
ties from  a  finely  divided  precipitate  by  washing  is  almost  impos- 
sible. 

Second,  the  difficulty  met  in  drying  material  of  this  kind. 
This  same  difficulty  was  met  in  the  experiments  on  silver  oxide 
which,  as  shown  by  Lea,  retained  moisture  up  to  165**. 


THIRD   SERIES. 


EXPERIMENTS  ON  SILVER  BENZOATE. 

The  preceding  work  on  silver  acetate  and  silver  succinate 
shows  the  necessity  of  selecting  compounds  which  form  well 
defined  crystals.      Perhaps  no  organic  salt  of  silver  fulfils  the 


ATOMIC  MASSES  OP  SII^VER,  MERCURY  AND  CADMIUM.     lOOI 

conditions  necessary  for  accurate  analyses  better  than  silver 
benzoate. 

PREPARATION  OP  SILVER  BENZOATE. 

The  purest  commercial  benzoic  acid  was  resublimed  thfee 
times  from  a  porcelain  dish  into  a  glass  beaker.  The  product 
thus  obtained  was  dissolved  in  pure  aqueous  ammonia  and  the 
solution  evaporated  to  crystallization.  The  ammonium  salt  was 
then  dissolved  in  distilled  water  and  to  the  solution  was  added  a 
solution  of  pure  silver  nitrate.  The  white  precipitate  of  silver 
benzoate  which  separated  was  washed  with  cold  water ;  it  was 
then  dissolved  in  hot  water,  the  solution  filtered,  and  evaporated 
to  crystallization.  The  salt  separated  in  fine  needles,  which 
clung  together  in  arborescent  masses.  After  removing  the 
liquid  from  the  beaker,  the  crystals  were  quickly  rinsed  with 
cold  water  and  placed  between  filters  to  remove  the  adhering 
moisture.  When  apparently  dry  they  were  broken  up  into  small 
fragments  and  dried  forty-eight  hours  in  a  vacuum  desiccator. 
The  material  was  then  placed  in  a  glass  stoppered  weighing  tube 
and  kept  in  a  dark  place. 

MODE  OP  PROCEDURE, 

The  details  of  the  method  of  operation  are  the  same  as  those 
given  under  silver  nitrate.  A  weighed  portion  of  the  material 
was  dissolved  in  a  dilute  solution  of  potassium  cyanide  in  a  plat- 
inum dish.  The  solution  was  then  electrolyzed  and  the  result- 
ing metal  weighed.  The  strength  of  current  and  time  of  action 
were  the  same  as  for  silver  nitrate. 

Before  the  results  could  be  reduced  to  a  vacuum  standard  it 
was  necessary  to  determine  the  specific  gravity  of  silver  ben- 
zoate. This  was  done  by  means  of  «  specific  gravity  bottle, 
the  liquid  used  being  chloroform.  The  mean  of  two  determina- 
tions gave  2.082  for  the  specific  gravity  of  silver  benzoate. 

Ten  results  on  this  compound,  reduced  to  a  vacuum  standard 
on  the  basis  of 

2.082  =  density  of  silver  benzoate, 
10.5=       **        *'  metallic  silver, 
21.4=       **        **  platinum  dish, 
8.5  =       •*        '*  weights, 


I002  WILLETT   LEPLEY   HARDIN. 

and  computed  for  the  formula  C^H^AgO,,  assuming  12.01,  1.008, 
and  16  to  be  the  atomic  masses  of  carbon,  hydrogen  and  oxy- 
gen, respectively,  are  as  follows  : 

Atomic  masi 
of  silver. 

107-947 
107.976 

107.918 

107.918 

107.964 

107.935 
107.936 

ro7.9i4 
107.908 
107.962 


Weight  of  CyHsAkO,. 

Weight  of  Ag. 
Gram. 

Grams. 

I 

0.40858 

0.19255 

2 

0.46674 

0.21999 

3 

0.48419 

0.22815 

4 

0.62432 

0.29418 

5 

0.66496 

0.31340 

6 

0.75853 

0.35745 

7 

0.76918 

0.36247 

8 

0.81254 

0.38286 

9 

0.95673 

0.45079 

10 

1.00840 

0.47526 

Mean 

=  107.938 

Maximum 

=  107.976 

Minimum 

=  107.908 

Difference         =     0.068 
Probable  error  =  ±0.005 

Computing  from  the  total  quantity  of  material  used  and  metal 
obtained  we  have  107.936  for  the  atomic  mass  of  silver. 

SUMMARY. 

In  discussing  the  work  on  the  atomic  mass  of  silver,  two  pos- 
sible sources  of  error  suggest  themselves. 

First,  the  hydrogen  which  is  continually  being  set  free  in  the 
process  of  electrolysis  may,  in  part,  be  occluded  by  the  metallic 
silver.  As  already  pointed  out,  the  metallic  deposits  were 
washed  several  times  with  boiling  water,  with  the  hope  of  remov- 
ing any  occluded  gases  ;  but  whether  this  effected  a  complete 
removal  of  all  the  occluded  gases  was  not  determined. 

Second,  the  condensation  of  moisture  on  the  platinum  dish 
might  be  urged  as  a  possible  source  of  error.  But  it  must  be 
remembered  that  the  dish  was  dried  in  the  same  manner  each 
time  and  kept  for  several  hours  in  a  desiccator,  and  that  the 
atmosphere  inside  the  balance  was  kept  dry  by  means  of  several 
beakers  of  anhydrous  calcium  chloride,  and  that  the  temperature 
of  the  balance  room  throughout  the  work  was  almost  constant. 
Under  these  conditions  there  is  but  little  chance  of  error  irom 


r 


ATOMIC  MASSES  OP  SILVER,  MERCURY  AND  CADMIUM.    IOO3 

different  amounts  of  moisture  condensed.  Moreover,  the  varia- 
tion in  the  different  weighings  of  the  same  dish  was  very 
slight. 

The  advantages  of  the  method  are  evident. 

First,  the  great  advantage  of  the  method  is  its  extreme  sim- 
plicity. 

Second,  the  nature  of  the  compounds  used  and  of  metallic  sil- 
ver renders  them  well  adapted  to  weighing. 

Third,  the  method  was  such  as  to  eliminate  the  errors  inci- 
dent to  the  ordinary  gravimetric  methods  of  analysis. 

Of  the  three  series,  the  first  is  probably  entitled  to  the  greatest 
weight.  That  the  silver  nitrate  was  pure  and  free  from  moisture 
seems  beyond  question.  However,  the  close  agreement  of  the 
last  two  series  with  the  first  indicates  that  the  acetate  and  ben- 
zoate  of  silver  were  also  free  from  moisture. 

Giving  equal  weight  to  each  of  the  three  series,  we  have  the 
following  as  the  general  mean  computed  from  the  separate  obser- 
vation : 

Atomic  mass  of  silver. 

First  series 107.924 

Second    '*  107.922 

Third       *'  107.938 

General  mean  ^  107.928 

Computing  the  general  mean  from  the  total  quantities  of 
material  used  and  metal  obtained  we  have  : 

Atomic  mass  of  silver. 
First  series 107.926 

Second    **  107.918 

Third       '*  107.936 

General  mean  =  107.927 

Combining  this  with  the  first  general  mean  we  have  107.9275 
as  the  final  result  for  the  atomic  mass  of  silver. 


PART  II. 


DETERMINATION  OF  THE  ATOMIC  MASS  6F  MERCURY. 

From  all  the  earlier  determinations  Clarke  gives  200  as  the 


I004  WILLETT  I^EPLEY   HAKDIN. 

most  probable  value  for  the  atomic  mass  of  mercury,  assuming 
oxygen  equal  to  i6. 

EXPERIMENTS  ON  MERCURIC  OXIDE. 

A  large  number  of  experiments  were  made  with  a  view  of 
determining  the  ratio  of  mercury  to  oxygen  in  mercuric  oxide. 
The  method  proved  to  be  unsatisfactory,  although,  apparently 
very  good  results  were  obtained  in  some  preliminary  experi- 
ments. The  cause  of  this  close  agreement  of  results  will  be 
explained  in  the  details  of  the  work. 

PREPARATION  OF  PURE  MERCURIC  OXIDE. 

The  purest  commercial  mercuric  chloride  was  carefully  sub- 
limed from  a  porcelain  dish  into  a  glass  funnel.  The  sublimed 
portion  was  dissolved  in  water,  the  solution  filtered,  and  evap- 
orated to  crystallization.  The  crystals  were  then  thoroughly 
dried  and  carefully  resublimed.  The  product  obtained  in  this 
way  consisted  of  white  crystalline  leaflets  which  dissolved  com- 
pletely in  water.  Pure  sodium  hydroxide  was  then  prepared  by 
throwing  pieces  of  metallic  sodium  on  pure  water  contained  in  a 
platinum  dish.  To  the  pure  sodium  hydroxide  was  added  a 
solution  of  mercuric  chloride,  the  former  always  being  in  excess. 
The  yellow  mercuric  oxide  which  separated  was  washed  for 
several  days  by  decantation  with  hot  water.  The  material  was 
then  dried  twenty-four  hours  in  an  air-bath  at  105**. 

MODE  OP  PROCEDURE. 

In  a  series  of  'preliminary  experiments  made  in  the  spring  of 
1895,  a  weighed  portion  of  mercuric  oxide  prepared  in  the  above 
manner  was  dissolved  in  a  dilute  solution  of  potassium  cyanide 
in  a  platinum  dish.  The  solution  was  then  electrolyzed  and  the 
weight  of  the  resulting  metallic  mercury  determined.  Inasmuch 
as  the  results  obtained  in  these  preliminary  experiments  were 
not  reduced  to  a  vacuum  standard,  it  was  thought  advisable  to 
weigh  the  empty  platinum  dish  after  removing  the  metallic 
deposit  in  order  that  the  two  weighings  might  be  made  under 
approximately  the  same  conditions.  The  results  for  the  most 
part  agreed  very  closely  and  differed  very  little  from  the  results 
obtained  by  other  methods.  Six  observations  computed  for  the 
formula  HgO,  assuming  the  atomic  mass  of  oxygen  to  be  16,  are 
as  follows : 


ATOMIC  MASSES  OP  SII^VER,  MERCURY  AND  CADMIUM.    IOO5 


Atomic  mass 

Weight  of  HfirO. 

Weight  of  Hg. 

of  mercury. 

Gram. 

Gram. 

I 

0.26223 

0.24281 

200.05 

2 

0.23830 

0.22065 

200.02 

3 

0.23200 

0.21482 

200.06 

4 

0.14 148 

O.13100 

200.00 

5 

0.29799 

0.27592 

200.03 

6 

O.I9631 

O.18177 

200.02 

Mean  ss  200.03. 

These  results  were  selected  from  a  larger  series.  After  ma- 
king the  above  observations  it  was  noticed  that  the  platinum  dish 
had  gradually  decreased  in  weight  throughout  the  work.  This 
decrease  in  weight  indicated  that  the  mercury  deposit  had 
formed  an  amalgam  with  the  platinum  dish,  which  was  soluble 
in  hot  nitric  acid.  To  ascertain  whether  such  was  the  case  or 
not  the  platinum  dish,  after  weighing*  was  filled  with  a  solution 
of  the  double  cyanide  of  mercury  and  potassium  and  the  solu- 
tion electrolyzed.  On  dissolving  the  mercury  deposit  in  cold 
nitric  acid  a  dark  colored  film  remained  on  the  sides  of  the  dish. 
The  dish  was  then  carefully  washed,  dried  and  reweighed,  and 
found  to  be  heavier  than  at  the  beginning  of  the  operation,  show- 
ing that  the  mercury  had  not  been  completely  removed.  The 
dark  film  was  then  dissolved  in  hot  nitric  acid  and  the  dish 
again  weighed.  This  last  weight  being  less  than  that  at  the 
beginning  showed  that  some  of  the  platinum  had  been  dissolved 
from  the  dish.  The  nitric  acid  solution  of  the  dark  film  was 
evaporated  to  dryness  and  ignited  to  remove  the  mercury.  The 
residue  was  dissolved  in  aqua  regia,  the  solution  evaporated  to 
dryness,  and  enough  water  added  to  dissolve  the  small  residue. 
A  little  concentrated  ammonium  chloride  was  then  added  to  the 
solution,  and  the  double  chloride  of  ammonium  and  platinum 
separated  as  a  yellow  crystalline  powder.  This  proved  conclu- 
sively that  the  mercury  deposit  had  united  with  the  platinum 
dish  to  form  an  amalgam  which  was  soluble  in  hot  nitric  acid. 
Hence  the  results  given  for  mercuric  oxide  are  of  no  value  in 
determining  the  atomic  mass  of  mercury. 

A  series  of  careful  experiments  was  then  made  on  the  oxide 
dried  at  different  temperatures.      To  avoid  any  error  from  the 


ICX)6  WILLETT   LEPLEY   HARDIN. 

amalgam  which  formed  with  each  deposit,  the  platinum  dish  was 
weighed  at  the  beginning  of  each  observation,  the  temperature 
and  barometric  pressure  being  noted  at  the  same  time.  The 
results  obtained  from  the  oxide  dried  at  a  temperature  of  105** 
gave  from  180  to  185  for  the  atomic  mass  of  mercury.  The  mate- 
rial was  then  dried  at  a  temperature  of  125'',  but  the  increase  in 
the  amount  of  mercury  obtained  was  very  slight.  Finallj-  with 
material  dried  at  150**,  the  results  obtained  for  the  atomic  mass 
of  mercur>' were  all  below  195**. 

The  most  probable  causes  for  these  low  results  are  : 

First,  the  difficult^'  of  removing  the  last  traces  of  alkalies  from 
the  mercuric  oxide. 

Second,  the  difficulty  met  in  the  complete  removal  of  the 
moisture  from  an  amorphous  precipitate.  This  difficulty  as  well 
as  the  first  was  referred  to  in  the  experiments  on  silver  oxide. 

Third,  mercuric  oxide  does  not  form  a  clear  solution  with 
potassium  cyanide.  There  seems  to  be  a  slight  reduction  of  the 
oxide  to  the  metallic  state.  It  is  difficult  to  determine  whether 
this  reduced  portion  unites  completely  with  the  metallic  deposit 
or  is  partially  removed  in  the  process  of  washing.  The  latter  is 
probably  true,  and  it  may  be  that  a  dififerent  method  of  analysis 
would  give  more  accurate  results  for  this  compound. 


FIRST  SERIES. 


EXPERIMENTS  ON  MERCURIC  CHLORIDE. 

The  material  used  in  this  series  of  experiments  was  prepared 
from  the  commercial  C.  P.  mercuric  chloride.  The  productwas 
first  dissolved  in  water,  the  solution  filtered  and  evaporated  to 
crystallization.  The  crystals  were  dried  and  carefully  sublimed 
from  a  porcelain  dish  into  a  glass  funnel.  The  sublimed  portion 
was  dissolved  in  water,  the  solution  filtered  and  evaporated  to 
crystallization.  These  crystals  were  dried  as  before  and  care- 
fully resublimed.  The  material  was  then  placed  in  a  weighing 
tube  and  kept  in  a  desiccator. 

'MODE  OF  PROCEDURE. 

The  method  of  operation  was  similar  to  that  already  described 


ATOMIC  MASSES  OF  SILVER,  MERCURY  AND  CADMIUM.     lOOy 

under  the  different  compounds  of  silver.  A  weighed  portion  of 
the  mercuric  chloride  was  dissolved  in  a  little  potassium  cyanide 
and  the  solution  electrolyzed.  The  deposit  was  washed  and 
dried  and  handled  in  every  way  like  the  deposits  of  silver.  The 
strength  of  the  current  and  time  of  action  were  as  follows  : 

Time  of  action.  Strength  of  current. 

4  hours N.Djoo  =  0.02  amperes. 

6      "     N.Dioo  =  o.05 

6      "     N.D,oo  =  o.io 

6      **     N.Dioo=o.30 

A  current  of  gradually  increasing  strength  deposits  the  mercury 
in  extremely  small  globules,  which  can  be  washed  and  handled 
more  easily  than  the  larger  globules  obtained  by  using  a  strong 
current  at  first.  In  cases  where  more  than  one-half  gram  of 
metal  was  deposited  the  strong  current  was  allowed  to  act  two 
hours  longer. 

Ten  results  on  mercuric  chloride  reduced  to  a  vacuum  stand- 
ard on  the  basis  of 

5.41  =  density  of  mercuric  chloride, 
13.59=:       *'        •*  metallic  mercury, 
21.4    =       *'        **  platinum  dish, 
8.5    =       **        •'  weights, 

and  computed  from  the  formula  HgCl,,  assuming  35.45  to  be  the 
atomic  mass  of  chlorine,  are  as  follows : 

Weight  of  Hgd,.  Weight  of  Hg.  Atomic  mass  of 

Grams.  Grams.  mercury. 

1  0.45932  0.33912  200.030 

2  0.54735  0.40415  200.099 

3  0.56002  0.41348  200.053 

4  0.63586  0.46941  199-947 

5  0.64365  0.47521  200.026 

6  0.73281         0.54101         199-988 

7  0.86467  0.63840  200.838 

8  1.06776  0.78825  199-946 

9  1.^7945  0.79685  199-917 
10                         1. 5 1402                       1. 1 1 780                       200.028 

Mean  =  20o.cx)6 
Maximum  =  200.099 
Minimum  3=5  199.917 

Difference  ss     0.182 
Probable  error  s=  ±0.011 


I008  WILLETT  LEPLEY  HARDIN. 

Computing  from  the  total  quantity  of  material  used  and  metal 
obtained  we  have  199.996  for  the  atomic  mass  of  mercury. 


SECOND  SERIES. 


EXPERIMENTS  ON  MERCURIC  BROMIDE. 

The  bromine  used  in  these  experiments  was  prepared  by  dis- 
tilling the  commercial  C.  P.  bromine  twice  over  manganese 
dioxide.  Any  trace  of  chlorine  which  might  be  present  would 
be  removed  by  this  method. 

PREPARATION  OP  MERCURIC  BROMIDE. 

Fifty  grams  of  metallic  mercury  were  placed  in  a  beaker  and 
covered  with  water.  Pure  bromine  was  then  added  until  the 
mercury  was  completely  saturated.  The  contents  of  the  beaker 
were  then  digested  with  hot  water  until  the  mercuric  bromide 
dissolved  ;  the  solution  was  filtered  and  evaporated  to  crystalli- 
zation. The  white  cr^-^stals  of  mercuric  bromide  which  sepa- 
rated were  thoroughly  dried  and  carefully  sublimed  from  a  por- 
celain dish  into  a  glass  funnel.  Only  the  middle  portion  of  the 
sublimate  was  used  in  the  experiments.  The  product  obtained 
in  this  way  consisted  of  brilliant  crystalline  leaflets  which  dis- 
solved completely  in  water.  The  material  was  kept  in  a  weigh- 
ing tube  in  a  desiccator. 

MODE  OP  PROCEDURE. 

The  method  of  analysis  was  exactly  like  that  described  under 
mercuric  chloride.  A  weighed  portion  of  the  mercuric  bromide 
was  dissolved  in  dilute  potassium  cyanide  in  a  platinum  dish. 
The  solution  was  then  electrolyzed  and  the  resulting  metal 
weighed.  The  strength  of  current  and  time  of  action  were  the 
same  as  for  mercuric  chloride. 

Ten  results  on  mercuric  bromide  reduced  to  a  vacuum  stand- 
ard on  the  basis  of 

5.92  =  density  of  mercuric  bromide, 
13.59  =  **  '*  metallic  mercury, 
21.4    =        **       *'  platinum  dish, 

8.5    =        **       **  weights. 


ATOMIC  MASSES  OP  SILVER,  MERCURY  AND  CADMIUM.    IOO9 

and  computed  for  the  formula  HgBr,,  assuming  79.95  to  be  the 
atomic  mass  of  bromine,  are  as  follows  : 

Atomic  mass  of 
mercury. 

199.898 

199.876 

199938 
199-832 
199.814 
199.91 1 
199.869 
199.840 
199.899 

199952 


Weight  of  HgBr,.           Weight  of  H 
Grams.                           Grams. 

I 

0.70002                         0.58892 

2 

0.56430                         0.31350 

3 

0.57142                         0.31750 

4 

0.77285                         0.42932 

5 

0.80930                         0.44955 

6 

0.85342                         0.47416 

7 

I.I  1076                         0.61708 

8 

1. 17270                         0.65145 

9 

1. 36186                         0.70107 

10 

1. 40142                         0.77870 

Mean  =  199.883 

Maximum  =  199.952 

Minimum  =  199.814 

Difference  ^     0.138 

Probable  error  =  ±0.010 

Computing  from  the  total  quantity  of  material  used  and  metal 
obtained,  the  atomic  mass  of  mercury  is  199.885. 


THIRD   SERIES. 


EXPERIMENTS  ON  MERCURIC  CYANIDE. 

A  series  of  observations  was  made  on  several  organic  salts  of 
mercury  with  a  view  of  selecting  a  compound  suitable  for  atomic 
mass  determinations.  Mercuric  acetate  and  other  similar  salts 
were  found  to  be  unstable  in  the  air  and  unsuited  for  accurate 
analyses.  Mercuric  cyanide,  on  the  other  hand,  was  found  to 
be  perfectly  stable  and  to  form  well  defined  crystals. 

PREPARATION  OF  HYDROCYANIC  ACID. 

Five  hundred  grams  of  potassium  ferrocyanide  were  placed  in 
a  two  liter  retort  connected  with  a  condenser.  A  cooled  mixture 
of  300  grams  of  pure  sulphuric  acid  and  700  cc.  of  distilled  water 
was  then  poured  into  the  retort  and  the  mixture  carefully  heated 
until  the  hydrocyanic  acid  was  distilled  over  into  the  receiver. 
The  product  obtained  was  redistilled  and  used  immediately  in 
the  preparation  of  mercuric  cyanide. 


lOIO  WILLETT  LEPLEY   HARDIN. 

PREPARATION  OF  MERCURIC  CYANIDE. 

Fifty  grams  of  mercuric  oxide,  prepared  as  already  described 
in  the  experiments  on  mercuric  oxide,  were  dissolved  in  pure, 
warm  hydrocyanic  acid.  The  solution  was  then  filtered  and 
evaporated  to  crystallization.  The  transparent  crystals  of  mer- 
curic cyanide  which  separated  were  dissolved  in  pure  water  and 
recrystallized.  The  product  obtained  by  the  second  crystalliza- 
tion was  quickly  rinsed  with  cold  water  and  dried  for  six  hours 
in  an  air  bath  at  a  temperature  of  50**.  The  crystals  were  then 
ground  to  a  finely  divided  powder  in  an  agate  mortar  and  redried 
for  twenty-four  hours  in  an  air  bath  at  a  temperature  of  55**.  The 
dry,  white  powder  was  then  placed  in  a  weighing  tube  and  kept 
in  a  desiccator. 

MODE  OF  PROCEDURE. 

The  mode  of  procedure  with  mercuric  cyanide  was  somewhat 
different  from  that  of  the  preceding  experiments,  in  that  no  po- 
tassium cyanide  was  used  in  preparing  the  solution  for  electroly- 
sis. A  weighed  portion  of  the  material  was  dissolved  in  pure 
water  in  a  platinum  dish.  When  the  crystals  had  completely 
dissolved,  the  dish  was  filled  to  within  a  quarter  of  an  inch  of  the 
top  with  water,  after  which  one  drop  of  pure  sulphuric  acid  was 
added.  The  solution  was  then  electrolyzed  and  the  resulting 
metal  weighed.  The  strength  of  the  current  and  the  time  of 
action  were  the  same  as  for  mercuric  chloride.  In  the  last  four 
experiments,  where  rather  large  quantities  of  mercury  were 
deposited,  the  strong  current  was  allowed  to  act  from  two  to  six 
hours  longer. 

The  results  of  ten  experiments  on  mercuric  cyanide,  reduced 
to  a  vacuum  standard  on  the  basis  of 

4.0  =  density  of  mercuric  cyanide, 
13.59  z=  *'  **  metallic  mercury, 
21.4    =       "        "  platinum  dish, 

8.5    =       *'         **  weights, 

and  computed  for  the  formula  Hg(CN)j,  assuming  12.01  and 
14.04  to  be  the  atomic  masses  of  carbon  and  nitrogen,  respect- 
ively, are  as  follows  : 


ATOMIC  MASSES  OP  SILVER,  MERCURY  AND  CADMIUM.    Id  I 


Weight  of  Hg(CN),. 
Grams. 

Weight  of  Hg. 
Grams. 

Atomic  mass 
of  mercury. 

I 

0.55776 

0.44252 

200.063 

2 

3 
4 

0.63290 
0.70652 
0.86241 

0.50215 
0.56053 
0.63663 

200.092 
200.038 
200.075 

5 
6 

7 
8 

9 

0.65706 
0.81678 
1.07628 
1. 22615 
1.66225 

0.52130 
0.64805 
0.85392 
0.97282 
I.31880 

200.057 
200.103 
200.077 
200.071 
200.057 

lO 

2. 1 1 170 

1. 67541 

200.077 

Mean 

=  200.071 

Maximum 

=  200.103 

Minimum 

=  200.038 

Difference  =    0.065 

Probable  error  =     0.005 

From  the  total  quantity  of  material  used  and  metal  obtained, 
the  atomic  mass  of  mercury  is  200.070. 


FOURTH  SERIES. 


According  to  Faraday's  law  the  quantities  of  different  metals 
deposited  from  their  solutions  by  the  same  current  are  propor- 
tional to  their  equivalent  weights.  In  this  series  of  experi- 
ments an  attempt  was  made  to  determine  the  ratio  of  the  atomic 
mass  of  mercury  to  that  of  silver  by  passing  the  same  current 
through  the  solutions  of  the  two  metals  and  weighing  the  two 
resulting  deposits.  If  the  proper  conditions  could  be  obtained, 
this  would  certainly  be  the  simplest  and  most  direct  method  for 
comparing  the  equivalent  weights  of  different  metals.  But  so 
many  difficulties  were  met  that  the  method  on  the  whole  was 
not  satisfactory. 

In  the  **  Revision  of  the  Atomic  Weight  of  Gold,*'*  Mallet 
made  use  of  this  method,  and  in  a  series  of  careful  preliminary 
experiments  determined  the  conditions  most  favorable  to  its 
application.  From  a  number  of  experiments  made  by  passing 
the  same  current  through  two  different  solutions  of  copper  sul- 

^  Am.  Chem.J.,  la,  182. 


IOI2  WILLETT  LEPLEY  HARDIN. 

phate,  using  pure  electrot)rpe  copper  for  both  anode  and  cathode 
in  each  solution,  Mallet  found  : 

First. — Other  conditions  being  the  same,  the  difference  in  the 
quantities  of  metal  deposited  from  solutions  of  unequal  concen- 
trations was  very  slight  and  somewhat  variable,  but  the  ten- 
dency was  toward  a  slightly  larger  quantity  from  the  more  con- 
centrated solution. 

Second. — With  equal  quantities  of  metal  in  the  two  solutions 
and  unequal  quantities  of  free  acid,  the  difference  in  the  results 
obtained  were  almost  insignificant  and  somewhat  variable  in 
direction,  the  tendency  being  toward  a  slightly  larger  quantity 
from  the  less  acid  solution. 

Third. — Other  conditions  being  the  same,  a  difference  in  the 
temperature  of  the  two  solutions  invariably  caused  a  slightly 
larger  deposit  from  the  cooler  solution. 

Fourth. — Other  conditions  being  the  same,  a  difference  in  the 
size  of  the  copperplates,  and  hence  a  difference  in  the  ** current 
density,"  caused  a  slightly  greater  deposit  on  the  smaller  plate. 

Fifth. — A  difference  in  the  distance  between  the  two  plates  did 
not  produce  a  constant  difference  of  result,  but  the  tendency  was 
toward  a  slightly  larger  deposit  on  the  cathode  plate  farther 
separated  from  its  anode. 

From  the  foregoing  experiments  it  is  evident  that  the  condi- 
tions most  favorable  to  this  method  are,  that  the  two  solutions 
should  be  equally  concentrated,  of  the  same  temperature,  and 
should  contain  equal  amounts  of  free  acid,  or  when  the  double 
cyanides  are  used,  equal  quantities  of  free  potassium  cyanide. 
And,  moreover,  that  the  two  cathodes  and  also  the  two  anodes 
should  be  of  the  same  size,  and  that  the  distance  between  the 
anode  and  cathode  should  be  the  same  in  both  solutions.  These 
conditiohs  were  closely  observed  throughout  this  work. 

ARRANGEMENT  OP  APPARATUS. 

The  deposits  in  this  series  of  experiments  were  made  in  two 
platinum  dishes  of  equal  capacity  and  equal  internal  area.  The 
anode  in  each  case  consisted  of  a  coil  of  rather  large  platinum 
wire,  the  two  coils  being  of  the  same  shape  and  size.  The  dishes 
were  insulated  from  each  other  by  means  of  two  glass  stands. 


ATOMIC  MASSES  OF  SILVER,  MERCURY  AND  CADMIUM.    IOI3 

The  platinum  coils  were  completely  immersed  in  the  solutions  and 
the  portion  of  the  wire  near  the  surface  of  the  liquid  was  covered 
with  paraffin  to  prevent  surface  contact.  The  current,  after 
passing  through  the  two  solutions,  was  allowed  to  pass  through 
a  hydrogen  voltameter  in  order  that  its  strength  might  be 
observed  at  any  time. 

In  the  second  arrangement  of  apparatus  the  platinum  dishes 
were  made  the  anodes,  and  two  pieces  of  platinum  foil  of  the 
same  shape  and  size  were  used  for  the  cathodes.  The  results, 
however,  from  this  second  arrangement  were  not  as  satisfactory 
as  from  the  first. 

MODE  OF  PROCEDURE. 

A  solution  of  the  double  cyanide  of  silver  and  potassium  was 
placed  in  one  of  the  platinum  dishes  and  a  solution  of  the  double 
cyanide  of  mercury  and  potassium  in  the  other.  The  quantities 
of  silver  and  mercury  present  in  their  solutions  were  approxi- 
mately proportional  to  their  equivalent  weights.  Each  solution 
contained  a  slight  excess  of  potassium  cyanide.  The  dishes 
were  placed  in  their  positions  and  the  anodes  immersed  some- 
time before  the  current  was  allowed  to  act.  When  the  tempera- 
ture of  the  two  solutions  was  the  same  as  that  of  the  room,  the 
connection  was  made  and  the  same  current  allowed  to  pass 
through  the  two  solutions.  The  quantity  of  metal  deposited 
was  never  allowed  to  exceed  one-half  of  the  metal  present  in  the 
solution  at  first.  Before  interrupting  the  current,  the  solutions 
were  siphoned  from  the  two  platinum  dishes  at  the  same  time 
with  two  siphons  of  the  same  bore.  The  deposits  were  then 
washed  several  times  with  boiling  water,  carefully  dried  and 
their  weights  determined.  Experiments  were  made  with  cur- 
rents of  different  strength  and  with  solutions  of  various 
degrees  of  concentration.  The  results  obtained  were  far  from 
being  satisfactory.  The  strength  of  current  which  seemed  best 
adapted  to  the  work  was  that  which  deposited  about  one-tenth 
of  a  gram  of  silver  per  hour. 

From  a  large  number  of  experiments,  only  seven  results  were 
obtained  which  seem  of  any  value  in  determining  the  atomic 
mass  of  mercury.     And  it  must  be  added  that  many  others  were 


IOI4  WILLETT  LEPLEY  HARDIN. 

rejected,  not  because  they  were  known  to  be  vitiated  in  any  way, 
but  because  the  results  obtained  for  the  atomic  mass  of  mercury 
differed  from  those  obtained  by  other  methods.  It  is  possible 
that,  in  a  large  number  of  experiments,  the  condition  would  be 
more  favorable  in  some  than  in  others,  but  whether  the  close 
agreement  of  the  results  selected  was  due  to  this  or  to  the  bal- 
ancing of  errors,  could  not  be  determined. 

Seven  results  computed  on  the  basis  of  107.92  for  the  atomic 
mass  of  silver  are  as  follows  : 

Atomic  mass 
Weight  of  Ug.  Weight  of  Ag.  of  mercury. 

Gram.  Gram. 

1  0.(16126  0.06610  200.036 

2  0.06190  0.06680  200.007 

3  0.07814  0.08432  200.021 

4  0.1036 1  O.I  1 181  200.011 

5  O.1520I  0.16402  200.061 

6  0.26806  0.28940  199-924 

7  0.82808  0.89388  199929 


Difference  =     0.137 

Computing  from  the  total  quantities  of  mercury  and  silver 
obtained,  we  have  199.971  for  the  atomic  mass  of  mercury. 

Although  the  cause  of  the  large  variation  in  the  rejected 
observations  could  not  be  definitely  determined,  several  sources 
of  error  suggest  themselves. 

First,  small  quantities  of  hydrogen  were  undoubtedly  set  free 
in  the  process  of  electrolysis,  and  unless  these  quantities  were 
always  equal  in  the  two  solutions,  which  is  not  probable,  an 
error  would  be  introduced. 

Second,  in  some  solutions  an  error  might  easily  be  introduced 
by  a  change  in  the  atomicity  of  mercury,  but  in  a  solution  of 
the  double  cyanide  of  mercury  and  potassium  this  change  is 
hardly  probable. 

Third,  the  occlusion  of  hydrogen  by  the  two  metallic  deposits 
would  also  be  a  possible  source  of  error ;  but  only  small  errors 
could  be  introduced  in  this  way. 

To  account  for  the  difference  of  several  units  in  the  results,. 


ATOMIC  MASSES  OF  SILVER,  MERCURY  AND  CADMIUM.    IOI5 

the  source  of  error  first  mentioned  seems  by  far  the  most  proba- 
ble. 

SUMMARY. 

In  the  discussion  of  the  results  obtained  in  the  different  series 
of  observations  on  the  compounds  of  silver,  the  probable  sources 
of  error  and  likewise  the  advantages  of  the  method  were  pointed 
out.  The  same  discussion  applies  equally  well  to  the  observa- 
tions on  mercury. 

It  is  evident  that  the  first  tliree  series  of  observations  on  mer- 
cur>-  are  entitled  to  more  weight  than  the  last  series.  Just  why 
the  results  on  mercuric  bromide  should  be  lower  than  those  on 
mercuric  chloride  is  not  clear.  Both  compounds  are  certainly 
well  adapted  to  atomic  mass  determinations,  inasmuch  as  they 
can  be  purified  by  both  crystallization  and  sublimation. 
The  most  probable  impurity  in  mercuric  bromine  would  be  mer- 
curic chloride,  but  that  would  tend  to  increase  rather  than  lower 
the  results.  The  series  of  observations  on  mercuric  cyanide 
have,  perhaps,  one  advantage  over  the  others,  in  that  no  potas- 
sium cyanide  was  used.  The  results  obtained  in  this  series  are 
still  higher  than  those  obtained  from  mercuric  chloride  and 
almost  two-tenths  of  a  unit  higher  than  those  obtained  from 
mercuric  bromide.  However,  as  the  same  care  was  exercised 
in  the  purification  of  the  material  for  each  of  the  three  series, 
and  as  there  was  no  apparent  error  in  either  case,  equal  weight 
must  be  given  to  each  of  the  three  series  in  determining  the 
most  probable  value  of  the  atomic  mass  of  mercury.  And,  as 
the  mean  of  the  last  series  is  almost  identical  with  the  mean  of 
the  first  three,  equal  weight  can  be  given  to  this  series  without 
introducing  any  error. 

Computing  the  general  mean  from  the  separate  observations, 
we  have : 

Atomic  mass  of  mercury. 

First  series 200.006 

Second   •'    199.883 

Third     "    200.071 

Fourth   **    199*996 

General  mean  ^  199-989 


IOl6  WILLETT   I^BPLEY  HARDIN. 

Prom  the  total  quantities  of  material  used  and  metal  obtained, 
the  general  mean  is  : 

Atomic  mass  of  mercury. 

First  series -   199.996 

Second   **    199.885 

Third     "    200.070 

Fourth**    Z99>97i 

General  mean  =  199.981 
Combining  this  with  the  first  general  mean  we  have  : 

Atomic  mast  of  mercuty. 
First  general  mean  ^  199.989 
Second    **  **     =  199.981 

Most  probable  mean  of  all  the  results  =s  199.985 
or  200  for  the  atomic  mass  of  mercury. 


PART  III. 


DETERMINATION  OP  THE  ATOMIC  MASS  OF  CADMIUM. 

Nine  experimenters  have  determined  the  atomic  mass  of  cad- 
mium by  many  different  methods,  but  the  large  variations  in  the 
results  given  by  different  chemists  leave  the  true  value  of  this 
constant  still  uncertain. 

Stromeyer'  gave  no  details  of  his  method  of  operation,  but 
found  that  100  parts  of  cadmium  combined  with  14,352  parts  of 
oxygen.  On  the  basis  of  O  =  16,  this  ratio  gives  11 1.483  for 
the  atomic  mass  of  cadmium.  This  result  is  mucji  lower  than 
those  obtained  by  other  experimenters  and  is  perhaps  only  of 
historical  interest. 

In  a  series  of  nine  experiments,  Von  Hauer'  determined  the 
ratio  of  cadmium  sulphate  to  cadmium  sulphide.  The  sulphate 
used  was  purified  by  repeated  recrystallizations  and  was  finally 
dried  at  a  temperature  of  20Q**.  After  weighing  the  sulphate 
was  always  dried  a  second  time  and  reweighed.  The  two 
weighings  never  differed  as  much  as  one  milligram.  The  sul- 
phide obtained  was  in  each  case  tested  for  sulphate.  The  reduc- 
tion of  the  sulphate  to  sulphide  was  accomplished  by  heating 

1  Berzelius'  Lehrbuch,  5th  Ed.,  3, 13x9. 
^J.praki.  Ck€m.,  7a,  350. 


ATOMIC  MASSES  OF  SILVER,  MERCURY  AND  CADMIUM.    IOI7 

the  sulphate  in  a  current  of  dry  hydrogen  sulphide  under  pres- 
sure. The  mean  of  nine  observations  computed  on  the  basis) 
0=  16  and  S  =  32.06  gives  11 1.93  for  the  atomic  mass  of  cad- 
mium. Considering  the  large  quantity  of  material  used  each 
time  and  the  precautions  taken  to  insure  accuracy,  there  seems 
to  be  little  objection  to  the  method. 

Dumas'  determined  the  ratio  of  cadmium  chloride  to  metallic 
silver  by  titrating  a  solution  containing  a  weighed  quantity  of 
cadmium  chloride  with  a  silver  nitrate  solution  of  known 
strength.  The  cadmium  chloride  was  prepared  by  dissolving 
metallic  cadmium  in  boiling  hydrochloric  acid.  The  solution 
was  evaporated  to  dryness  and  the  chloride  fused  for  six  hours 
in  a  current  of  hydrochloric  acid  gas.  The  mean  of  six  deter- 
minations gives  1 1 2. 24  for  the  atomic  mass  of  cadmium  (0=  16). 

Maximum  result,  Cd  =  112.759 
Minimum       '*       Cd  =  1 11.756 

DifFerence  =      1.003 

This  large  variation  in  the  results  obtained  indicates  the  pres- 
ence of  impurities  in  the  material  used.  In  the  first  three  ex- 
periments the  cadmium  was  not  purified ;  the  mean  of  these 
three  is  Cd  =  112.476.  The  metal  used  in  the  last  three  exper- 
iments was  considered  by  Dumas  to  be  absolutely  pure ;  the 
mean  of  the  last  three  results  is  Cd  =  112.007.  From  the 
degree  of  purity  of  the  cadmium  chloride  used  in  the  different 
experiments,  Dumas  was  inclined  to  reject  the  higher  results 
and  concluded  that  the  true  atomic  mass  of  cadmium  was  about 
112. 

Lensen*  prepared  pure  cadmium  oxalate  by  precipitating  a 
solution  of  cadmium  chloride,  purified  by  repeated  crystallization, 
with  pure  oxalic  acid.  The  precipitate  was  washed  and  care- 
fully dried  at  a  temperature  of  150**.  The  mean  of  three  results 
obtained  by  converting  a  weighed  portion  of  the  oxalate  to  oxide 
gives  112.06  for  the  atomic  mass  of  cadmium  (O  =  16).  The 
small  quantity  of  material  used  in  the  different  experiments  is 
somewhat  objectionable. 

1  Ahh.  cAim.phys.t  [3],  55,  158. 
ij.^akt.  Ckem.,  79,  281. 


IOl8  WILLETT  LEPLKY   HARDIN. 

Huntington/  under  the  direction  of  Cooke,  determined  the 
ratio  of  cadmium  bromide  to  silver  bromide  and  also  the  ratio  of 
cadmium  bromide  to  metallic  silver.  The  bromide  used  was 
prepared  by  dissolving  cadmium  carbonate^  which  had  been 
carefully  purified,  in  pure  hydrobromic  acid.  The  product 
obtained  was  dried  at  a  temperature  of  200^  and  finally  sublimed 
in  a  porcelain  tube  in  a  current  of  dry  carbon  dioxide.  In  the 
first  series  of  experiments  the  silver  bromide  corresponding  to 
the  cadmium  bromide  used  was  weighed.  The  mean  of  eight 
determinations  computed  from  the  total  quantity  of  material 
used  and  silver  bromide  obtained,  on  the  basis  of  Ag  =  107.93 
and  Br  =  79.95  is  Cd  =  112.24.  In  the  second  series  of  experi- 
ments the  quantity  of  metallic  silver  required  to  precipitate  a 
known  quantity  of  cadmium  bromide  was  determined.  The 
mean  of  eight  determinations  computed  as  in  the  first  series 
gives  112.245  for  the  atomic  mass  of  cadmium.  The  separate 
determinations  in  both  series  agree  very  closely. 

Partridge*  made  three  series  of  determinations.  The  first 
depended  upon  ttie  conversion  of  cadmium  oxalate  into  oxide, 
the  second,  on  the  reduction  of  the  sulphate  to  sulphide,  and  the 
third,  on  the  conversion  of  the  oxalate  into  sulphide.  The  cad- 
mium used  in  these  experiments  was  purified  by  distilling  twice 
in  vacuo.  Ten  observations  on  the  conversion  of  the  oxalate 
into  oxide,  computed  on  the  basis  of  0=  16  and  C  =  12,  give 
1 1 1. 801  as  a  mean  for  the  atomic  mass  of  cadmium.  Recalcu- 
lated by  Clarke,'  on  the  basis  of  0=  16  and  C=  12.005,  the 
atomic  mass  of  cadmium  becomes  11 1.8 18.  The  mean  of  ten 
results  obtained  by  reducing  the  sulphate  to  sulphide,  computed 
on  the  basis  of  O  =  16  and  S  =  32,  gives  11 1.797  for  the  atomic 
mass  of  cadmium.  Recalculated  by  Clarke  on  the  basis  of  0  = 
16  and  S  =  32.074,  the  atomic  mass  of  cadmium  is  111.711.  In 
the  third  series  the  oxalate  of  cadmium  was  converted  into  sul- 
phide by  heating  in  a  current  of  dry  hydrogen  sulphide.  The 
mean  of  ten  determinations,  computed  on  the  basis  of  O  =  16 
and  S=32,  gives  in. 805  for  the  atomic   mass  of  cadmium. 

1  Proc.  Amer.  Acad.,  17,  28. 
^  Am.  J.  Set.,  r3].40,  377- 
^  Am.  Chem.J.,  13,  34.. 


ATOMIC  MASSES  OF  SILVEP.,  MERCURY  AND  CADMIUM.    IOI9 

Recalculated  by  Clarke  on  the  basis  of  O  =  16  and  S  =  32.074, 
the  mean  becomes  1 1 1 .589.  Partridge  gives  11 1 .8  for  the  atomic 
mass  of  cadmium,  as  a  mean  of  the  three  series.  If  the  higher 
values  for  carbon  and  sulphur  be  introduced  this  value  becomes 
somewhat  lower. 

Jones*  determined  the  atomic  mass  of  cadmium  by  two  differ- 
ent methods.  The  first  was  based  on  the  conversion  of  the 
metal  into  oxide,  and  the  second  on  the  conversion  of  the  oxa- 
late into  oxide.     The  cadmium  used  was  distilled  six  times  in 

* 

vacuo.  The  last  distillate  was  tested  spectroscopically  and  found 
to  be  free  from  impurities.  In  the  first  series  of  experiments  a 
weighed  portion  of  the  pure  metal  was  dissolved  in  pure  nitric 
acid  in  a  porcelain  crucible.  The  solution  was  evaporated  to 
dr3mess  and  the  resulting  cadmium  nitrate ,  ignited  to  oxide. 
The  final  decomposition  was  accomplished  by  means  of  a  blast 
lamp.  Reducing  gases  were  carefully  excluded  from  the  crucible 
during  the  process  of  ignition.  The  weighings  were  all  made 
against  a  tared  crucible.  The  mean  of  ten  observations,  com- 
puted on  a  basis  of  O  =  16  gives  112.07  for  the  atomic  mass  of 
cadmium.  The  different  determinations  agree  very  closely.  In 
the  second  series  of  experiments  cadmium  oxalate,  prepared  by 
precipitating  pure  cadmium  nitrate  with  pure  oxalic  acid,  was 
converted  into  oxide.  The  material  was  carefully  ignited  until 
the  oxalate  was  decomposed  ;  it  was  then  treated  with  nitric  acid 
and  again  ignited  in  a  manner  similar  to  that  described  in  the 
first  series.  The  mean  of  five  determinations  computed  on  the 
basis  of  O  =  16  and  C  =  12.003  is  Cd  =  1 11.032.  From  all  the 
observations,  Jones  concludes  that  112.07  represents  very  closely 
the  atomic  mass  of  cadmium  (O  =  16). 

Lorimer  and  Smith'  determined  the  ratio  of  the  atomic  mass 
of  cadmium  to  that  of  oxygen  by  dissolving  pure  cadmium  oxide 
in  potassium  cyanide  and  electrolyzing  the  solution.  To  obtain 
pure  material,  the  commercial  cadmium  was  dissolved  in  nitric 
acid  and  the  solution  evaporated  to  crystallization.  The  crys- 
tals of  cadmium  nitrate  were  removed  from  the  liquid,  dissolved 
in  pure  water  and  recrystallized.     The  product  obtained  by  the 

I  Am,  Chem.J,,  14,  a6i. 

9  Ztsckr.  anorg.  Chem.,  x,  364. 


I020  WII^LETT   LEPLEY   HARDIN. 

second  recrystallization  was  dissolved  in  a  little  water  and  treated 
with  a  slight  excess  of  potassium  cyanide  in  a  platinum  dish. 
From  this  solution  the  metallic  cadmium  was  thrown  out  by 
means  of  the  electric  current.  The  nitrate  obtained  by  dissolv- 
ing the  electrolytic  cadmium  in  pure  nitric  acid  was  tested  spec- 
troscopically  and  found  to  be  free  from  impurities.  The  pure 
cadmium  nitrate  was  digested  with  ammonium  hydroxide  and 
ammonium  carbonate  and  the  resulting  cadmium  carbonate 
ignited  to  oxide  in  a  platinum  crucible.  The  method  of  opera- 
tion was  very  simple,  a  weighed  portion  of  the  oxide  was  dis- 
solved in  pure  potassium  cyanide,  the  solution  electrolyzed  and 
the  resulting  metallic  cadmium  weighed.  The  mean  of  nine 
observations  computed  on  the  basis  of  O  =  i6  gives  112.055  for 
the  atomic  mass  of  cadmium. 

Bucher*  made  six  series  of  experiments.  The  cadmium  used 
was  purified  by  nine  distillations  in  vacuo.  The  weighings  were 
all  reduced  to  a  vacuum  standard  and  computed  on  the  basis  of 
0=16.  8  =  32.059,  C=  12.003,   CI  =  35.45,   Br  =79.95,  and 

Ag  =  107.53- 

In  the  first  series  cadmium  oxalate,  dried  for  fifty  hours  at 
150*",  was  ignited  to  oxide.  The  mean  of  eight  observations 
gives  1 1 1.89  for  the  atomic  mass  of  cadmium. 

In  the  second  series,  cadmium  oxalate  was  converted  into 
sulphide  by  heating  in  a  current  of  dry  hydrogen  sulphide.  The 
mean  of  four  determinations  is  Cd  =  112. 15. 

In  the  third  series  a  weighed  quantity  of  cadmium  chlo- 
ride, dried  at  a  temperature  of  300*  in  hydrochloric  acid  gas, 
was  precipitated  with  silver  nitrate  and  the  resulting  silver 
chloride  weighed.  The  mean  of  twenty-one  determinations  is 
Cd=  112.39.  The  separate  observations  in  this  series  agree 
very  closely. 

The  fourth  series  was  similar  to  the  third,  except  that  cad- 
mium bromide  was  used  instead  of  the  chloride.  The  mean  of 
five  determinations  is  Cd=  112.38,  a  resuft  almost  identical 
with  that  obtained  from  the  chloride. 

In  the  fifth  series  a  weighed  portion  of  metallic  cadmium  was 
converted  into  sulphate,  which  was  dried  at  400'  and  weighed. 

1  Thesis,  Johnn  Hopkins  University,  iSg4. 


ATOMIC  MASSES  OP  SII«VBR,  MBRCURY  AND  CADMIUM.    I02I 

The  excess  of  sulphuric  acid  which  remained  with  the  sulphate 
was  estimated  and  its  weight  deducted.  The  only  result  given 
isCd  =  115.35. 

In  the  last  series  metallic  cadmium  was  converted  into  oxide 
by  dissolving  in  nitric  acid  and  igniting  the  resulting  cadmium 
nitrate.  The  mean  of  two  determinations  made  by  igniting  the 
material  ip  a  porcelain  crucible  gives  112.08  for  the  atomic  mass 
of  cadmium.  Three  similar  determinations  made  with  a  plati- 
num crucible  gave. as  a  meanCd=  11 1.87.  Prom  a  series  of 
experiments  on  cadmium  oxide,  Bucher  concluded  that  a  cor- 
rection should  be  applied  to  the  last  and  also  the  first  series. 
By  making  this  correction,  the  results  in  these  two  series  would 
be  very  close  to  those  obtained  from  the  chloride  and  bromide. 

Prom  all  the  preceding  determinations  Clarke  gives  11 1.93  as 
the  most  probable  value  for  the  atomic  mass  of  cadmium.  The 
large  variation  in  the  results  of  different  experimenters  has  not 
been  fully  explained.  Some  chemists  think  that  the  larger 
values  are  due  to  a  higher  degree  of  purity  in  the  metallic  cad- 
mium used,  and  hence  regard  these  values  as  being  more  nearly 
correct.  But  it  must  be  remembered  that  the  reverse  is  true  in 
the  experiments  of  Dumas.  From  material  which  had  not  been 
purified,  Dumas  obtained  results  ranging  from  112.32  to  112.76 
for  the  atomic  mass  of  cadmium,  while  from  material  which  he 
considered  absolutely  pure,  the  results  were  from  11 1.76  to 
112. 13. 

PREPARATION  OF  PURE  CADMIUM. 

The  metallic  cadmium  used  in  these  experiments  was  purified 
by  distillation  in  a  current  of  hydrogen  which  had  been  passed 
through  solutions  of  caustic  potash^  lead  nitrate,  potassium  per- 
manganate, and  sulphuric  acid.  A  hard  glass  combustion  tube 
was  heated  to  redness  and  the  walls  of  the  tube  indented  at  two 
points  with  a  three-cornered  file.  This  divided  the  tube  into 
three  parts.  Commercial  cadmium  was  placed  in  one  end  Of  the 
tube  and  connection  made  with  the  hydrogen  generator.  After 
complete  removal  of  the  air,  the  tube  was  carefully  heated  in  a 
combustion  furnace  until  one-lialf  of  the  metal  had  distilled  over 
into  the  middle  portion  of  the  tube.     The  metal  was  cooled  in  a 


I022  WILLETT   LEPLEY   HARDIN. 

current  of  hydrogen.  The  tube  was  then  broken  and  the  metal 
removed.  The  portions  in  the  first  and  last  sections  of  the  tube 
were  rejected.  The  middle  portion  was  placed  in  a  second  com- 
bustion tube,  similar  to  the  first,  and  the  distillation  repeated. 
After  three  distillations  the  metal  was  examined  spectroscop- 
icallj'  and  found  to  be  free  from  impurities. 


FIRST  SERIES. 


EXPERIMENTS  ON  CADMIUM  CHLORIDE, 

Dumas  and  Bucher  have  both  determined  the  ratio  of  cadmium 
to  chlorine  in  cadmium  chloride.  The  results  given  for  the 
atomic  mass  of  cadmium  by  the  latter  experimenter  are  almost 
four- tenths  of  a  unit  higher  than  those  given  by  the  former. 

PREPARATION  OF  CADMIUM  CHLORIDE. 

Hydrochloric  acid  was  purified  by  first  passing  chlorine 
through  the  commercial  C.  P.  acid  to  remove  any  sulphur  diox- 
ide ;  the  excess  of  chlorine  was  removed  by  a  current  of  carbon 
dioxide.  The  acid  was  then  distilled  from  calcium  chloride  and 
the  hydrochloric  acid  gas  collected  in  pure  water.  Pure  metal- 
lic cadmium  was  then  dissolved  in  the  acid  and  the  solution 
evaporated  to  crystallization.  The  crj^stalsof  cadmium  chloride 
were  removed  from  the  liquid  and  thoroughly  dried.  The 
material  was  then  placed  in  a  hard  glass  combustion  tube,  simi- 
lar to  that  used  in  the  distillation  of  metallic  cadmium,  and  care- 
fully sublimed  in  a  current  of  dry  carbon  dioxide.  The  first 
and  last  portions  of  the  sublimate  were  rejected.  The  middle 
portion,  which  consisted  of  pearly  leaflets,  was  placed  in  a 
weighing  tube  and  kept  in  a  desiccator.  As  only  a  small  quan- 
tity of  the  material  could  be  sublimed  at  a  time,  the  different 
analyses  were  made  from  different  sublimations. 

MODE  OF  PROCEDURE. 

A  weighed  portion  of  the  cadmium  chloride  was  dissolved  in 
a  little  water  in  a  platinum  dish.  A  slight  excess  of  potassium 
cyanide  was  added  and,  after  diluting  to  200  cc.  with  pure  water, 
the  solution  was  electrolyzed.  Before  interrupting  the  current, 
the  liquid  was  siphoned  from  a  dish  in  a  manner  already  outlined 


ATOMIC  MASSES  OF  SILVER,  MERCURY  AND  CADMIUM.    IO23 

in  the  experiments  on  silver.  The  metallic  deposit  was  washed 
several  times  with  boiling  water  and  carefully  dried.  The 
strength  of  the  current  and  time  of  action  were  as  follows : 

Time  of  action.  Strensrth  of  current. 

labours N.D,oo  =  o.i   amperes. 

4       •*     N.Dioo=o.i5 

4       "     N.Dioo  =  o.30 

The  cadmium  was  thrown  down  as  a  dense  white  deposit. 
Ten  results  on  cadmium  chloride  reduced  to  a  vacuum  stand- 
ard on  the  basis  of : 

3.3    =  density  of  cadmium  chloride, 
8.55  =       **        **  metallic  cadmium, 
21.4    =       *'        **  platinum  dish, 
8.5    =       •*        ''  weights, 
and  computed  for  the  formula  CdCl,,  assuming  35.45  to  be  the 

atomic  mass  of  chlorine,  are  as  follows : 

• 

Weight  of  CdClf.  Weight  of  Cd.  Atomic  mass  of 

Grams.  Gram.  cadmium. 

1  0.43140  0.26422  112.054 

2  0.49165  O.30112  112.052 

3  0.7175a  0.43942  112.028 

4  0.72188  0.44208  1 12.021 

5  0.77264  0.47319  112.036 

6  0.81224  0.49742  112.023 

7  0.90022  0.55135  112.041 

8  1.02072  0.62505  112.002 

9  1.26322  0.77365  112.041 
10  1.52344  0.93314  112.078 

Mean  =  112.038 
Maximum  =  112.078 
Minimum  =  112.002 

Difference  =     0.076 
Probable  error  =  ±0.005 

From  the  total  quantity  of   material  used  and  metal  obtained, 
we  have  112.040  for  the  atomic  mass  of  cadmium. 


SECOND  SERIES. 


PREPARATION  OF  CADMIUM  BROMIDE. 

The  bromine  used  in  this  series  was  purified  as  outlined  in  the 
experiments  on  mercuric  bromide.  The  cadmium  bromide  was 
prepared  by  allowing  bromine  water  to  act  on  metallic  cadmium 
for  several  days  at  the  ordinary  temperature.     When  the  action 


I024  WII^LETT   LEPI^EY   HARDIN. 

was  complete,  the  solution  was  filtered  and  evaporated  to  crys- 
tallization. The  crystals  of  cadmium  bromide  were  removed 
from  the. liquid  and  thoroughly  dried.  The  material  was  then 
placed  in  a  hard  glass  combustion  tube  and  carefully  sublimed 
in  a  current  of  dry  carbon  dioxide.  The  first  and  last  portions 
of  the  sublimate  were  rejected.  The  middle  portion  was  removed 
from  the  tube,  placed  in  a  weighing  bottle  and  kept  in  a  desic- 
cator. The  product  obtained  in  this  way  consisted  of  a  crystal- 
line, pearly  leaflet  which  dissolved  immediately  in  water  with- 
out leaving' a  residue. 

MODB  OP  PROCEDURE. 

The  method  of  operation  was  the  same  as  for  cadmium  chlo- 
ride. A  weighed  portion  of  the  material  was  dissolved  in  a  little 
water  in  a  platinum  dish.  A  slight  excess  of  potassium  cyanide 
was  then  added  and  after  diluting  to  200  cc.  the  solution  was 
electrolyzed  and  the  resulting  metal  weighed.  The  strength  of 
current  and  time  of  action  were  the  same  as  for  cadmium  chlo- 
ride. 

Ten  observations  on  cadmium  bromide  reduced  to  a  vacuum 
standard  on  a  basis  of : 

4.8    =  density  of  cadmium  bromide, 
8.55=       **       •*  metallic  cadmium, 
31.4    =       **       '*  platinum  dish, 
8.5    =       **       **  weights, 
and  computed  for  the  formula  CdBr,,  assuming  79.95  to  be  the 
atomic  mass  of  bromine,  are  as  follows  : 


I 

2 

3 

4 

5 
6 

7 
8 

9 
10 


Weight  ol  CdBr,. 

Weight  of  Cd. 
Gram. 

Atomic  masfl  of 

Grams. 

cadmium. 

0.57745 

0.23790 

1 1 2.03 1 

0.76412 

0.31484 

112.052 

0.91835 

0.37842 

112.067 

1. 01460 

0.41808 

112.068 

1.15074 

0.47414 

112.053 

1. 2475 1 

0.51392 

II2.OI9 

I.25051 
i.5i«>5 

O.SI905 

112.087 

0.62556 

1x2.076 

1.63543 

0.67378 

112.034 

2.15342 

0.88722 

1 12.041 

Mean 

=  112.053 

Maximum 

=  112.087 

Minimum 

=  II2.OI9 

Difference 

3=      0.068 

Probable  error 

=  ±0.005 

ATOMIC  MASSES  OF  SILVER,  MERCURY  AND  CADMIUM.    IO25 

From  the  total  quantity  of  material  used  and  the  metal  obtained, 
Cd  =  112.053. 


THIRD  SERIES. 


In  these  experiments  an  attempt  was  made  to  determine  the 
ratio  of  the  atomic  mass  of  cadmium  to  that  of  silver  by  allowing 
the  same  electric  current  to  pass  successively  through  solutions  of 
the  two  metals  and  weighing  the  resulting  deposits.  The  arrange- 
ment of  apparatus  and  the  details  of  the  method  were  described 
under  the  mercur^*^  silver  series.  The  results  were  not  as  satisfac- 
tory as  the  corresponding  results  obtained  for  mercury-.  A  large 
number  of  determinations  were  made  with  currents  of  different 
strength  and  solutions  of  different  concentration,  but  the  results 
were,  in  most  cases,  far  below  those  obtained  in  the  first  two 
series.  A  current  which  deposited  about  twelve  hundredths  of  a 
gram  of  silver  per  hour  seemed  to  give  the  best  results.  From 
all  the  observations,  five  results  were  selected  which  differed 
only  about  one-tenth  of  a  unit  from  those  of  the  first  two  series. 
Results  selected  in  this  way  are  entitled  to  but  little  weight,  and 
perhaps  should  not  be  used  in  determining  the  general  mean  of 
all  the  observations. 

Computed  on  the  basis  of  107.92  for  the  atomic  mass  of  silver, 
the  only  admissible  results  are  as  follows : 

Atomic  mass 
of  cadmium. 

1  0.24335  0.12624  II  1.928 

2  0.21262  O.IIO32  III.99I 

3  0.24515  0.12720  1 1 1.952 

4  0.24331  O.12616  1 1 1. 916 

5  0.42520  0.22058  1 11.971 


Weight  of  Ag. 

Weiflrht  of  Cd. 

Gram 

Gram. 

0.24335 

0.12624 

0.21262 

0.  II 032 

0.24515 

0.12720 

0.24331 

O.I 2616 

0.42520 

0.22058 

Mean 

=s  III.952 

Maximum 

=  111.991 

Minimum 

=s III.916 

Difference 

—    0.07s 

This  method  was  discussed  under  mercury.  The  probable 
sources  of  error  pointed  out  there  apply  equally  well  in  the  case 
of  cadmium.     Until  the  large  variations  can  be  accounted  for 


I026    ATOMIC  MASSES  OF  SILVER,  MERCURY  AND  CADMIUM. 

and  the  difficulties  overcome,  the  method  must  be  regarded  as 
unsatisfactory. 

SUMMARY. 

Inasmuch  as  but  one  method  of  analysis  has  been  used 
throughout  this  work,  it  is  useless  to  discuss  it  here.  The 
advantages  and  objections  pointed  out  under  silver  apply  also  to 
cadmium. 

In  summing  up  the  work  on  cadmium,  equal  weight  must  be 
given  to  the  first  two  series.  The  last  series  must  be  considered 
alone  and  all  that  need  be  said  of  it  is,  that  the  results  obtained 
for  the  atomic  mass  of  cadmium  never  exceeded  112.  In  the 
corresponding  series  on  mercury,  the  variations  were  in  both 
directions  from  200. 

The  general  mean  of  the  first  two  series  calculated  from  the 
separate  observations  is  : 

Atomic  mass  of  Cd. 
First  series       =  112.038 
Second  series   =  112.053 

General  mean  =  11 2.1)455 

From  the  total  quantity  of  material  used  and  metal  obtained 
we  have: 

Atomic  mass  of  Cd. 
First  series       =s  112.040 
Second  series   =  112.053 

General  mean  =112.0465 
Combining  this  with  the  first  general  mean  we  have  112.046  as 
the  most  probable  result  of  all  the  work,  for  the  atomic  mass  of 
cadmium.  This  result  is  lower  than  those  obtained  by  Hunt- 
ington and  Bucher,  but  agrees  very  closely  with  the  results 
obtained  by  von  Hauer,  Dumas,  I^ensen,  Jones,  and  Lorimer 
and  Smith. 

I  wish  here  to  express  my  sense  of  gratitude  to  Professor 
Edgar  F.  Smith,  at  whose  suggestion  this  work  was  undertaken 
and  under  whose  personal  supervision  it  was  carried  out. 


BOOKS  RECEIVED. 

Bulletin  No.  128.  Pests  of  Grain  Crops.  Issued  by  the  North  Caro- 
lina Agricultural  Experiment  Station,  Raleigh,  North  Carolina.  July  i, 
1896.       II  pp. 

Bulletin  No.  45.  Varieties  of  Apples.  University  of  Illinois  Agricul- 
tural Experiment  Station,  Urbana,  Illinois.    July,  1896.    52  pp. 

The  Solvay  Process  Alkali ;  Its  Various  Forms  and  Uses,  with  Notes  on 
Alkalimetry  and  Chemical  and  Commercial  Tables,  conveniently  arranged 
for  the  Use  of  the  Consumer.  The  Solvay  Process  Company,  Syracuse, 
New  York.     39  pp.     7  plates. 

Special  Bulletin,  Second  Edition.  Commercial  Fertilizers.  Purdue 
University,  Lafayette,  Indiana.     August,  1896.    8  pp. 

The  University  Scientific  Monthly.  Published  by  the  Engineering 
Society  of  the  University  of  Tennessee,  Knoxville,  Tennessee.  August, 
1896.     43  pp. 

Bulletin  No.  127.  Parasites  of  Domestic  Animals.  Issued  by  the 
North  Carolina  Agricultural  Experiment  Station,  Raleigh,  North  Caro- 
lina.    May  15,  1896.     42  pp. 

Chemistry  in  Daily  Life.  Popular  Lectures  by  Dr.  Lassar-Cohn.  Trans- 
lated by  M.  M.  Pattison  Muir.  Philadelphia:  J.  B.  Lippincott  Co. 
With  21  wood  cuts  in  text,     x,  324  pp.     Price,  $1.75. 

Bulletin  No.  40.  I.  Analyses  of  Manurial  Substances  Sent  on  for 
Examination.  II.  Analyses  of  Licensed  Fertilizers  Collected  by  the  Agent 
of  the  Station  during  1896.  Hatch  Experiment  Station  of  the  Massa- 
chusetts Agricultural  College,  Amherst,  Mass.    July,  1896.    19  pp. 

Bulletin  No.  41.  The  Application  of  Tuberculin  in  the  Suppression  of 
Bovine  Tuberculosis.  Hatch  Experiment  Station  of  the  Massachusetts 
Agricultural  College,  Amherst,  Mass.     August,  1896.     27  pp. 

Humphry  Davy,  Poet  and  Philosopher.  By  T.  E.  Thorpe,  LL.D., 
F.R.S.     New  York:  Macniillan  &  Co.,  Ltd.     1896.     240pp.     Price,  ^1.25. 

A  Manual' of  Quantitative  Chemical  Analysis  for  the  Use  of  Students. 
By  Frederick  A.  Cairns,  A.M.  Third  Edition.  Revised  and  enlarged 
by  Elwyn  Waller,  Ph.D.  New  York  :  Henry  Holt  &  Co.  1896.  xii,  417 
pp.     Price,  $2.00. 

The  Chemical  Analysis  of  Iron.  A  Complete  Account  of  all  the  Best 
Known  Methods  for  the  Analysis  of  Iron,  Steel,  Pig  Iron,  Iron  Ore, 
Limestone,  Slag,  Clay,  Sand,  Coal,  Coke,  and  Furnace  and  Producer 
Gases.  By  Andrew  Alexander  Blair.  Third  Edition.  Philadelphia : 
J.  B.  Lippincott  Co.     1896.     322  pp.     Price  ;^t.oo. 

Bulletin  No.  38.  Canaigre,  the  new  Tanning  Plant.  By  H.  H.  Har- 
rington and  Duncan  Adriance.     Agricultural  and  Mechanical  College  of 


I028  BOOKS   RECEIVED. 

Texas,  College  Station,  Brazos  County*  Texas.     March,  1S96.     11  pp.    7 
plates. 

Bulletin  No.  24.  North  Dakota  Soils.  Government  Agricultural  Ex- 
periment Station  for  North  Dakota,  Fargo,  N.  D.    June,  1896.    17  pp. 

Roentgen  Rays  and  Phenomena  of  the  Anode  and  Cathode.  Princi- 
ples, applications  and  theories.  By  Edward  P.  Thompson,  M.B.,  E.E. ; 
concluding  chapter  by  Prof.  William  A.  Anthony.  60  diagrams.  45  half- 
tones,   xiv,  190  pp.     N.  Y. :  D.  Van  Nostrand  Co.    Price  I1.50. 

Foods :  Their  Composition  and  Analysis.  By  Alexander  Wynter 
Blyth,  M.R.C.S.,  F.I.C.,  F.C.S.,  etc.  With  numerous  tables  and  illus- 
trations. Fourth  edition,  revised  and  enlarged.  1896.  xxxii,  755  pp. 
New  York  :  D.  Van  Nostrand  Co.     Price  $7.50. 

Bulletin  No.  30.  Wheat-cutting  at  different  dates.  Agricultural  Ex- 
periment Station  of  Nevada  State  University,  Reno,  Nevada.    December, 

1895-     7  pp. 

Bulletin  No.  32.  Commercial  Fertilizers  and  Chemicals.  Inspected, 
analyzed  and  admitted  for  sale  in  the  State  of  Georgia  up  to  September  i, 
1896,  and  other  information  concerning  fertilizers.  Under  the  supervi- 
sion of  Hon.  R.  T.  Nesbitt,  Commissioner  of  Agriculture  of  the  State  of 
Georgia,  Atlanta,  Ga.     132  pp. 

Tables  for  Iron  Analysis.  By  John  A.  Allen,  First  Edition,  vii  +85 
pp.     New  York  :  John  Wiley  &  Sons. 

Geological  Survey  of  Canada.  Reports  and  Maps  of  Investigation  and 
Surveys.  Annual  Report.  Vol.  VII.  1894.  xiv  +  124A  +  427B  -f-  40C+ 
54F  -|-  157J  +  149M  -|-  68R  -|-  187S  4-  xvii  pp.,  with  maps.  Ottawa  :  Gov- 
ernment Printing  Bureau. 


Vol.  XVIII.  [December,  1896.]  No.  12. 


THE  JOURNAL 


OF  THE 


AMERICAN  CHEMICAL  SOCIETY. 


[Contribution  prom  the  John  Harrison  Laboratory  of  Chemistry. 

No.  14.] 

METAL   SEPARATIONS   BY  MEANS   OP   HYDROCHLORIC 

ACID  QAS/ 

Bt  J.  Bird  Mover. 

RecciTcd  September  s6,  1896. 

INTRODUCTION. 

THE  action  of  gaseous  haloid  acids  upon  metallic  oxides  and 
their  salts,  is  a  field  of  investigation,  which,  though  not 
of  recent  origin,  has  been  but  lately  developed.  It  was  Debray* 
who  first  called  attention  to  the  volatility  of  molybdic  acid  in  a 
stream  of  hydrochloric  acid  gas,  with  the  formation  of 
MoO(OH),Cl,. 

E.  Pochard'  applied  this  and  showed  that  molybdic  acid  was 
completely  eliminated  and  separated  from  tungstic  acid,  by  its 
volatility  in  a  current  of  hydrochloric  acid.  Since  that  time 
nothing  further  has  been  done  with  single  haloid  acids,  in  gas 
form,  until  quite  recently.  Compounds  have  been  decomposed, 
salts  volatilized,  and  separations  made,  by  means  of  other  gases 
and  mixtures,  which  may  be  as  effective  as  hydrochloric  acid, 
but  are  not  devoid  of  trouble  nor  nearly  so  neat. 

Smith  and  Oberholtzer*  repeated  and   confirmed   Pochard's 

I  Prom  author's  thesis  presented  to  the  Paculty  of  the  University  of  Pennsylvania 
for  the  degree  of  Doctor  of  Philosophy.  1S96.  ^ 

S  Compt.  rend,,  46  1098,  tiud  Ann.  Chem.  (Liebig).  108.  250. 
•  Compt,  rend,,  114.  173. 
4  /.  Am.  Chem.  Sac.,  15.  z. 


I030  J.    BIRD    MOVER.       METAI.  SEPARATIONS 

work  in  regard  to  the  separation  of  molj'bdic  acid  from  tungstic 
acid,  and  in  addition  showed  that  gaseous  hydrobromic,  hydri- 
odic,  and  hydrofluoric  acids  acted  similarl}'.  Later,  Smith  and 
Maas'  made  use  of  the  volatilization  of  molybdic  acid  for  a  close 
atomic  mass  determination  of  molybdenum. 

Smith  and  Hibbs*  showed  that  vanadium  behaved  like  molyb- 
denum. Hydrochloric  acid  gas  completely  eliminates  vanadic 
acid  from  sodium  vanadate.  A  little  later  they  investigated  the 
action  of  hydrochloric  acid  upon  the  members  of  Group  V  of  the 
periodic  system.' 

The  sodium  salts  of  nitric,  pyrophosphoric,  pyroarsenic  and 
pyroantimonic  acids  were  used.  They  found  nitrogen,  arsenic, 
and  antimony  to  be  volatile  in  gaseous  hydrochloric  acid,  and 
made  it  the  basis  of  a  separation  of  phosphoric  acid  from  nitric 
acid.  Lead  arsenate  changed  completely  to  chloride,  the  arse 
nic  being  volatilized,  thus  affording  a  good  quick  separation. 
Smith  and  Meyer*  tried  the  action  of  all  the  haloid  acids  upon 
the  elements  of  Group  Vof  the  periodic  system.  They  worked 
with  sodium  salts  and  observed  :  I.  That  nitrogen  was  expelled 
completely  by  all  the  haloid  acids.  II.  That  phosphoric  acid 
was  not  acted  upon.  III.  That  arsenic  acid  was  fully  expelled 
by  hydrochloric,  hydrobromic,  and  hydriodic  acids,  but  only 
partially  by  hydrofluoric  acid.  IV.  That  antimony"  was  com- 
pletely volatilized  by  hydrochloric  acid.  There  was  no  work 
done  on  bismuth.  V.  Vanadium  went  over  completely  in 
hydrochloric  acid,  but  only  partially  in  hydrobromic  and  hydro- 
fluoric acids.  VI.  Columbium  forms  volatile  products  with 
hydrochloric  and  hydrobromic  acids.  No  knowledge  of  didym- 
ium  was  obtained.  VII.  Tantalum  is  onlj'^  slightly  volatile 
in  hydrochloric  acid. 

P.  Jannasch  and  F.  Schmidt''  repeated  some  of  the  work  of 
Smith  and  Hibbs,  in  which  they  confirmed  the  separation  of 
arsenic  from  lead.  They  anticipated  a  slight  portion  of  my 
work,  and  in  addition  separated  arsenic  acid  from  iron,  tin  from 

1  Ztsch*-.  anoi-g.  Chem.,  5.  2S0. 

ay.  Am.  Chem.  Sot.,  16/578. 

•  Ibid.,  17,  682. 

4  Ibid.,  X7,  735. 

6  Ztschr.  anorg.  Chem.,  9,  274. 


BY   MEANS  OF   HYDROCHLORIC   ACID   GAS.  IO3I 

lead,  tin  from  copper,  and  tin  from  iron,  in  a  stream  of  hydro- 
chloric acid  gas. 

The  position  of  bismuth  in  the  periodic  system  makes  it 
natural  to  suppose  that  it  too  will  be  volatile  in  hydrochloric 
acid  gas.  This  I  have  shown  to  be  true,  and  was  thus  enabled 
to  separate  it  from  lead  and  copper.  The  action  of  hydrobromic 
acid  on  bismuth  trioxide  was  also  tried  ;  it  formed  the  bromide 
and  then  volatilized.  It  requires  a  higher  temperature  and 
longer  action  than  with  hydrochloric  acid.  Because  of  lack  of 
time,  I  have  been  compelled  to  abandon  the  experiments  in- 
stituted with  a  view  of  affecting  separations,  in  atmospheres 
of  hydrobromic  acid  and  hydriodic  acid  gas  and  have  confined 
my  labors  to  hydrochloric  acid  gas. 

METHOD  OF   WORK. 

The  hydrochloric  acid  gas  was  generated  by  dropping  con- 
centrated sulphuric  acid  from  a  separatory  funnel,  upon  concen- 
trated hydrochloric  acid  contained  in  a  three  liter  flask.  The 
gas  evolved  at  the  ordinary  temperature  was  dried  by  passing  it 
through  two  sulphuric  acid  drying  bottles  and  then  through  a 
calcium  chloride  tower,  when  it  was  considered  sufficiently  dry 
for  the  purpose.  The  substance  to  be  acted  upon  was  weighed 
out  in  a  porcelain  boat  and  the  latter  was  placed  in  a  combus- 
tion tube  of  hard  glass. 

The  tube  had  previously  been  rinsed  with  alcohol  and  then 
with  ether,  to  remove  all  moisture.  The  ether  was  removed  by 
drawing  a  current  of  drj'-  air  through  the  tube.  This  tube  was 
connected  to  a  two-necked  bulb  receiver  containing  about  300 
cc.  of  distilled  water.  When  working  with  arsenic  ten  cc.  of 
nitric  acid  were  added.  The  connecting  tube  from  the  combus- 
tion tube  to  the  bulb  receiver  was  made  to  enter  the  receiver  and 
dip  below  the  surface  of  the  water,  thus  catching  all  volatile 
products,  as  well  as  taking  up  the  hydrochloric  acid  gas.  To 
insure  safety  from  the  loss  of  volatile  products,  a  small  flask 
containing  water  was  attached  to  the  bulb  receiver.  The  appa- 
ratus was  controlled  at  both  ends  by  stop-cocks.  This  is  neces- 
sary to  prevent  backward  suction  on  disconnecting  the  appara- 
tus.    After  the  reaction  was  completed  the  boat  was  removed  to 


I032  J.    BIRD   MOYBR.      MBTAI<  SEPARATIONS 

a  sulphuric  acid  desiccator  from  which  the  air  could  be 
exhausted.  In  general,  the  procedure  was  similar  to  that 
employed  by  Hibbs.* 

I. — BBHAVIOR  OP  ANTIMONY  TRIOXIDE. 

Antimony  oxide,  labelled  chemically  pure,  was  dissolved  in 
hydrochloric  acid  and  precipitated  with  a  large  amount  of 
water.  After  washing  by  decantation  it  was  redissolved  and 
reprecipitated.  This  procedure  was  repeated  several  times, 
when  it  was  precipitated  by  ammonium  carbonate,  washed,  and 
ignited.  The  pure  oxide  obtained  in  this  manner  was  sub- 
jected to  the  action  of  hydrochloric  acid  gas  and  it  was  "found  to 
volatilize  completely.  In  each  trial  a  one-tenth  gram  of  the  oxide 
was  acted  upon.  The  temperature  varied  between  150**  and  190**  C. 
It  was  determined  in  the  following  way  :  The  combustion  tube 
was  slipped  through  two  holes  made  in  the  sides  of  a  copper 
drying  oven. 

A  very  slow  current  of  gas  was  used  as  the  antimony  seemed 
to  volatilize  more  readily  and  completelj',  if  the  current  was 
slow  and  the  heat  gentle.  This  I  attribute,  on  reflection,  to  the 
fact  that  I  ignited  the  oxide  too  strongly,  (to  a  red  heat)  in  its 
preparation.  It  dissolved  with  difficulty  in  concentrated  hydro- 
chloric acid.  Lack  of  time  prevented  the  repetition  of  this 
experiment  and  the  separation  of  antimony  from  lead  and  copper. 
in  which  this  substance  was  used.  About  eight  hours  was  the 
time  required  for  the  volatilization;  very  probably  a  shorter  time 
w^ould  be  required  if  the  oxide  had  been  obtained  by  gentle  igni- 
tion. 

II. — BEHAVIOR  OP   LEAD  OXIDE. 

Pure  lead  oxide  was  obtained  from  recrystallized  nitrate,  by 
careful  ignition.  This  oxide  changed  completely  into  chloride 
at  the  ordinary  temperature  and  it  was  only  necessary  to  apply 
a  gentle  heat  to  complete  the  change  and  entirely  remove  the 
water  formed.  No  volatilization  was  noticed  until  a  tempera- 
ture of  225**  was  reached  ;  at  this  point  the  lead  chloride  slightly 
volatilized. 

I  think  it  possible  to  estimate  lead  as  chloride,  if  the  tempera- 
ture is  kept  under  200^.     A  weighed  amount  of  lead  oxide  was 

1  Thesis,  1896. 


it 


BY  MBANS   OF   HYDROCHLORIC   ACID   GAS.  IO33 

acted  upon  by  hydrochloric  acid  gas  in  the  told,  for  two  hours, 
and  then  heated  sufficiently  to  remove  all  the  water  formed. 

The  boat  was  cooled  in  the  gas,  and  then  placed  in  a  sulphuric 
acid  desiccator  and  allowed  to  stand  one-half  hour.  It  was  then 
weighed. 

Experiments. 

Lead  oxide     Lead  chlo-      Lead  chlo- 
taken.      ride  obtained,  ride  required.   Differeuce. 

Gram.  Gram.  Gram.  Gram. 

Experiment  I 0.1017  0.1267  0.1267  0.0000 

II 0.1015  0.1258  0.1265  —0.0007 

III 0.1 169  0.1454  0.1447  -f-0.0007 

The  lead  chloride  dissolved  in  hot  water  without  residue. 

III. — THE  SEPARATION  OF  ANTIMONY  EROM  LEAD. 

The  oxides  were  carefully  weighed  and  thoroughly  mixed  in 
a  porcelain  boat.  Hydrochloric  acid  gas  was  passed  over  them 
in  the  cold,  until  the  lead  oxide  had  been  entirely  changed  to 
,the  chloride.  It  was  then  heated  with  the  smallest  flame 
obtainable  from  a  fish-tail  burner,  placed  about  two  inches  below 
the  tube. 

Antimony  tri-    Lead  chlo-        Lead  chlo-       Lead  chlo- 
chloride  taken,  ride  taken,    ride  obtained,  ride  required. 

Gram.  Gram.  Gram.  Gram. 

Ezperiment  I  •  •  •  •  0.1015  0.1189  0.1470  0.1482 

"           II>'->  0.1090  0.1021  0.1266  0.1272 

**           III...  0.1350  0.0852  0.1057  0.1062 

"           IV...  0.1250  0.1671  0.2069  0.2083 

The  time  required  was  seven  hours..  The  lead  chloride  was 
immediately  weighed.  It  dissolved  completely  in  hot  water  and 
this  solution  was  tested  by  means  of  Marsh's  apparatus  for  anti- 
mony, without  finding  the  latter  present.  Experiment  II  was 
slightly  varied  by  first  moistening  the  oxides  with  a  drop  of 
hydrochloric  acid.  ^ 

IV. — BEHAVIOR  OF  BISMUTH  OXIDE. 

Bismuth  nitrate,  as  pure  as  could  be  obtained,  was  dissolved 
in  nitric  acid  and  then  thrown  down  with  a  large  quantity  of 
water.  The  precipitate  was  carefully  washed  by  decantation. 
This  operation  was  repeated  several  times. 

It  was  then  dissolved  in  acidulated  water  and  precipitated 


I034  J.    BIRD   MOVER.      METAL  SEPARATIONS 

with  ammonium  hydroxide  and  ammonium  carbonate.  This, 
on  ignition,  gave  pure  oxide,  which,  heated  in  a  stream  of 
hydrochloric  acid  gas,  completely  volatilized  as  chloride.  Here 
the  same  treatment  is  necessary  as  obtained  for  antimony.  A 
slow  current  of  gas  and  a  low  heat  were  best  adapted  for  the 
volatilization  (a  temperature  of  130°,  or  roughly,  the  heat 
afforded  by  a  fish-tail  burner  placed  two  inches  below  the  com- 
bustion tube,  with  a  flame  an  eighth  of  an  inch  high).  The  bis- 
muth chloride  sublimed  nicely,  forming  awhite  crystalline  mass 
beyond  the  boat,  which  could  be  readily  driven  along  by  a  gen- 
tle heat. 

v. — THE  SEPARATION  OF  BISMUTH  FROM  LEAD. 

The  same  material  was  used  as  in  the  preceding  experiments. 
The  weighed  oxides  were  thoroughly  mixed  in  a  porcelain  boat. 
Usually  the  gas  was  allowed  to  act  in  the  cold  for  an  hour, 
which  changed  the  oxides  completely  to  chlorides. 

The  same  conditions  prevailed  as  under  bismuth  oxide  alone. 
If  an  attempt  was  made  to  hasten  the  reaction  by  heating  higher 
than  iSo"*,  a  little  lead  would  volatilize.  This  sublimate,  slightly 
yellow  in  color,  w^ould  appear  directly  over  the  boat  and  could 
not  be  driven  along  the  tube  like  bismuth,  hence  it  was  readily 
detected. 

The  separation  of  bismuth  from  lead  requires  much  care,  as 
it  is  not  as  sharp  as  could  be  desired.  It  is  also  difiicult  to  tell 
exactly  when  the  last  traces  of  bismuth  have  been  driven  out  of 
the  boat,  as  there  was  no  color  change  to  indicate  it,  both  metals 
forming  white  chlorides.  The  separation  is  complete  in  from 
six  to  seven  hours.  At  the  end  of  the  separation  the  position  of 
the  boat  was  changed  and  the  action  continued  ;  if  no  further 
sublimation  occurred  it  was  cooled  and  removed  to  a  desiccator. 
The  weight  was  taken  after  standing  one-half  hour  over  sul- 
phuric acid.  With  care  bismuth  can  be  separated  from  lead  in 
this  manner. 


Bismuth 

Lead 

Lead 

Lead  oxide 

trioxide 

chloride 

chloride 

taken. 

taken. 

obtained. 

required. 

Difference. 

Gram. 

Gram. 

Gram. 

Gram. 

Gram. 

Experiment!""   0.1014 

0.2020 

O.I 261 

0.1264 

—0.0003 

"            II....   0.1006 

0.0642 

0.1252 

0.1254 

— 0.0002 

III...  0.1038 

0.1003 

0.1294 

0.1302 

—0.0008 

IV  ...  0.1412 

0.1260 

0.1759 

01 759 

4-0.0000 

BY    MEANS  OF    HYDROCHLORIC   ACID   GAS. 


1035 


The  chloride  of  lead  dissolved  completely  in  hot  water.  It 
showed  no  bismuth.     The  sublimate  contained  no  lead. 

VI. — BEHAVIOR  OF  CUPRIC  OXIDE. 

Pure  copper  nitrate  was  made  by  recrystallization.  It  was 
then  ignited  in  a  porcelain  crucible  at  a  dull  red  heat,  until  it 
became  constant  in  weight.  The  pure  black  oxide  was  then 
subjected  to  the  action  of  hydrochloric  acid  gas.  In  Experi- 
ment I,  the  boat  containing  the  oxide  was  heated  at  the  outset 
to  175**.  It  was  taken  out  after  two  hours,  placed  over  sulphuric 
acid  for  half  an  hour,  and  weighed.  The  weight  showed  that 
the  copper  oxide  had  hardly  been  acted  upon.  It  had  only  been 
superficially  changed  to  chloride.  It  was  then  moistened  with 
two  or  three  drops  of  hydrochloric  acid,  dried  in  a  rapid  current 
of  the  gas,  and  heated  two  hours  longer.  This  resulted  in  the 
complete  transformation  into  chloride.  The  anhydrous  chloride 
thus  obtained,  liver  brown  in  color,  was  placed  in  a  desiccator 
from  which  the  air  was  exhausted.  This  was  done  to  remove 
all  the  gas  that  might  be  retained  and  prevented  a  too  rapid 
absorption  of  moisture. 

Copper  chloride  absorbs  moisture  but  not  so  rapidly  as  to  pre- 
vent weighing  in  this  form,: 


« 

Copper 
oxide 
taken. 

Copper 

chloride 

obtained. 

Copper 

chloride 

required. 

Difference. 

Gram. 

Gram. 

Gram. 

Gram. 

Experiment  I  . . . . 

0  ion 

0.1708 

O.1713 

—0.0005 

II.... 

0.1025 

0.1726 

0.1736 

— O.OOIO 

III... 

0.1034 

0.1756 

0.1752 

4-0.0004 

In  Experiment  II,  the  change  was  completed  in  the  cold  by 
prolonged  action  through  four  hours.  It  was  then  heated  about 
ten  minutes  at  the  end  to  drive  out  the  moisture  that  had 
formed.  In  all  the  experiments  cited,  the  copper  chloride,  after 
weighing,  was  found  to  dissolve  completely  in  cold  water. 

VII. — THE  SEPARATION  OF  ANTIMONY  FROM  COPPER. 

The  same  material  was  used  as  in  the  preceding  experiments. 

The  weighed  oxides  were  thoroughly  mixed.     The  antimony 

was  completely  volatilized,   leaving  copper  chloride  which  was 


1036 


J.    BIRD   MOVER.      METAL  SEPARATIONS 


weighed  as  such.  The  volatile  antimony  chloride  was  caught 
in  the  bulb  receiver  at  the  end  of  the  tube.  The  bulb  and  tube 
were  washed  out  with  acidulated  water  into  a  beaker  and  the 
antimony  thrown  down  with  hydrogen  sulphide.  The  antimony 
sulphide  was  filtered,  thoroughly  washed,  and  while  moist  dis- 
solved in  strong  hydrochloric  acid.  The  hydrogen  sulphide 
evolved  was  conducted  into  bromine  water  and  oxidized  to  sul- 
phuric acid,  which  was  estimated  as  usual  and  the  antimony 
calculated. 

The  length  of  time  required  was  eight  hours.  On  several 
occasions  the  experiment  was  interrupted  at  the  end  of  four 
hours,  but  invariably  the  separation  was  incomplete  and  on  dis- 
solving out  the  copper  chloride  formed,  black  copper  oxide  and 
white  antimony  oxide  were  plainly  evident.  In  some  cases  the 
mixture  of  oxides  was  moistened  with  a  couple  of  drops  of 
hydrochloric  acid  and  then  evaporated  down  in  a  stream  of  acid 
gas,  by  heating  the  tube  over  a  water- bath.  This  treatment 
seemed  to  facilitate  matters  but  it  is  not  altogether  advisable, 
because  the  copper  chloride  has  a  tendency  to  creep  over  the 
sides  of  the  boat.  It  is  quicker  in  the  end  to  separate  them  in 
the  dry  condition,  allowing  plenty  of  time  for  the  reaction.  The 
copper  chloride  obtained  was  perfectly  soluble  in  cold  water  and 
contained  no  antimony.  It  could  readily  be  changed  to  oxide 
and  weighed  if  thought  necessary. 


Antimony 

trioxide 

taken. 

Copper 
oxide 
taken. 

Copper 
chloride 
obtained. 

Copper 
chloride 
required 

Gram. 

Gram. 

Gram. 

Gram. 

Experiment  I  ..  0.1068 

0.1040 

0.1750 

0.1745 

II..  0.1062 

0.1053 

0.1774 

0.1784 

•*               III.    0.1022 

0. 1020 

0.1726 

0.1728 

IV  .  0.1198 

0.1020 

0.1722 

0.1728 

Antimonv  tri-  Antimony  tri- 
oxide taken,     oxide  found. 


Experiment  I 


Gram. 
0.1068 


Gram. 
0.1039 


Difference. 
Gram. 

+0.0005 
KOOIO 
).0002 
).O006 


Difference. 
Gram. 

+0.0009 


VIII. — THE  SEPARATION  OF  BISMUTH  PROM  COPPER. 

The  pure  oxides  were  mixed  and  treated  as  directed  under 
bisiiiuih  and  lead. 


BY   MEANS  OF   HYDROCHLORIC   ACID   GAS. 


1037 


oxiae 
taken. 

Gram. 

Experiment  I  ••  0.1030 

'*  II..  0.1004 

III.  0.1026 

**  IV  .  0.1019 


Bismuth 

trichloride 

taken. 

Grams. 
0.1069 
0.1077 
0.1060 
0.1058 


Copper 
chloride 
obtained. 

Grams. 
0.1738 

O.I  701 

0.1741 
O.1718 


Copper 

chloride 

required. 

Grams. 

0.1745 
O.I713 

0.1738 

0.1726 


Difference. 
Grams. 

.0007 

L00I2 

+00003 

i.oooS 


Bismuth 

trioxide 

obtained. 

Gram. 


Bismuth 
trioxide 
required. 

Gram. 


Difference 

Gram. 


Experiment  1 0.1076  0.1069  +0.0007 

The  time  required  in  each  of  these  trials  was  seven  hours. 
It  seemed  to  be  advantageous  to  raise  the  temperature  and  heat 
sharply  for  about  ten  minutes  at  the  end,  to  insure  the  complete 
removal  of  the  bismuth. 

Moistening  with  acid  helped  the  reaction  but  subjected  it  to 
the  same  danger  of  creeping  as  noted  under  antimony  and  cop- 
per. 

The  bismuth  was  estimated  as  follows  :  It  was  washed  out  of 
the  tube  and  bulb  with  acidulated  water  and  then  precipitated 
as  sulphide.  The  bismuth  sulphide  was  filtered,  washed,  and 
dissolved  in  nitric  acid.  It  was  thrown  out  of  the  solution  with 
ammonium  hydroxide  and  ammonium  carbonate,  as  hydrated 
oxide,  and  then  filtered,  dried,  and  ignited.  It  was  weighed  as 
oxide.  The  residue  of  copper  chloride  in  the  boat  dissolved  in 
cold  water  and  showed  no  bismuth. 

IX. — ACTION  OF  GASEOUS  HYDROCHLORIC  ACID  ON  SODIUM 

PYROARSENATE. 

Hibbs'  showed  that  arsenic  was  completely  volatilized  from 
sodium  pyroapsenate,  leaving  weighable  sodium  chloride.  In 
fact,  so  clean  was  the  elimination  of  arsenic  that  he  made  it  the 
basis  of  an  arsenic  atomic  mass  determination,  with  admirable 
success. 

In  working  up  the  separation  of  arsenic  from  other  metals  it 
was  necessary  to  start  with  the  pure  sodium  salt.  After  purifi- 
cation I  decided  to  test  it,  by  weighing  the  salt  produced  by  the 
action  of  the  acid  gas  upon  it.  Several  determinations  gave 
close  results,  proving  the  salt  pure. 

^  See  next  paper,  page  1044. 


1038  J.    BIRD   MOVER.      METAL  SEPARATIONS 

Chemically  pure  arsenate  was  procured.  It  was  recrystallized 
and  then  ignited  (not  too  strongly)  for  an  hour.  Thepyroarse- 
nate  obtained  was  used  in  precipitating  the  various  arsenates 
investigated. 

Sodium  pyroar-      Sodium  chlo-         Sodium  ctalo- 
senate  taken.       ride  obtained.       ride  required. 

Gram.  Gram.  Gram. 

Experiment  1 0.2021  0.1330  0.1335 

"  II 0.1039  0.0691  0.0686 

X. — THE  SEPARATION  OF  ARSENIC  FROM  COPPER. 

Pure  sodium  pyroarsenate  was  used  to  precipitate  the  copper 
salt. 

Copper  sulphate  was  recrystallized  five  times,  a  few  good 
crystals  were  dissolved  and  the  two  solutions  mixed.  A  green 
copper  arsenate  was  precipitated.  It  was  washed  and  dried  at 
100°.  Salkowski*  observes  that  copper  arsenate  still  contains 
water  above  130°.  My  salt  had  the  composition  Cu,As,0,  + 
2H,0. 

Hydrochloric  acid  gas  completely  changes  it  in  the  cold  to 
chloride.  A  slight  heat  drives  out  the  arsenic  and  water  and 
leaves  a  brown  anhydrous  copper  chloride,  which  can  be  weighed 
as  such.  Care  was  taken  to  remove  all  the  acid  gas  before 
weighing. 

The  arsenic  was  washed  out  of  the  bulb  into  a  beaker,  this 
was  warmed  with  nitric  acid  to  insure  oxidation,  and  then  it 
was  precipitated  from  an  ammoniacal  solution  with  **  a  magne- 
sia mixture.'*     It  was  weighed  as  Mg,As,0,. 

Copper  arse-    Copper  chlo-  Copper  chlo- 


nate  takeu. 

ride  obtained. 

ride  required. 

Differem 

Gram. 

Gram. 

Gram. 

Gram. 

Experiment 

;  I  ....   0.1067 

0.0850 

0.0851 

— O.OOOI 

1 1 

II....   0.1240 

0.0998 

0.0991 

-ho.  0007 

( ( 

III.. .   0.1072 

0.0860 

0.0856 

-|-o.ooa4 

i  t 

IV...   O.I  155 

0.0924 

0.0923 

+0.0001 

( 1 

V 0.1042 

0.0832 

0.0833 

— O.OOOI 

Experiment  I.  ASjOj  obtained,  0.0498  gram  ;  As^Oj  required,  0.0487  gram. 

The  residue  of  copper  chloride  completely  dissolved  in  water. 
It  showed  no  arsenic  when  tested  in  a  Marsh  apparatus. 

'^J.prakt.  Chem.,  104,  129. 


BY   MEANS  OF   HYDROCHLORIC   ACID   GAS.  IO39 

XI. — THE  SEPARATION   OP  ARSENIC  FROM  SILVER. 

Silver  arsenate  was  made  by  precipitating  silver  nitrate  with 
sodium  arsenate.  Care  was  taken  to  have  the  nitrate  in  excess. 
The  reddish- brown  arsenate  of  silver  was  washed  with  boiling 
water,  until  the  washings  no  longer  showed  silver,  when  tested 
with  hydrochloric  acid.     It  was  dried  at  no**. 

As  was  expected,  the  acid  gas  attacked  it  even  in  the  cold. 
In  fact  the  action  was  so  vigorous  that  a  couple  of  analyses  were 
spoiled  by  spattering.  The  trouble  arose  from  the  fact  that  the 
arsenate  was  not  finely  powdered.  Heat  was  generated  in  the 
reaction  sufficiently  to  send  over  a  portion  of  the  water  formed. 
Experiment  I  was  run  in  the  cold  for  one  hour  and  then  heated 
sharply,  for  a  few  minutes,  to  expel  the  arsenic  and  water.  The 
result  was  only  0.46  per  cent,  too  high,  but  indicated  that  the 
salt  should  be  heated  longer,  and  not  necessarily  as  high  to 
remove  all  the  arsenic. 

The  succeeding  experiments  were  heated  from  one  to  two 
hours  at  150**  with  better  results  : 


Silver 

arsenate 

taken. 

Silver 
chloride 
obtained. 

Silver 
chloride 
required. 

Difference. 

Gram. 

Gram. 

Gram. 

Gram. 

Experiment  I 

0.2542 

0.2381 

0.2363 

-I-0.0018 

II.... 

0.2325 

0.2163 

O.2161 

-f-0.0002 

III... 

0.2084 

0.1952 

0.1938 

-fo.0014 

IV... 

0.2070 

0.1927 

0.1924 

-I-O.OOO3 

Experiment  i.  Ag  obtained  =  70.45  per  cent ;  Ag  required  =69.99  P^r 
cent. 

The  residues  in  Experiments  II,  III,  and  IV  were  dissolved 
and  tested  for  arsenic.     None  was  found. 

XII.— THE  SEPARATION  OF  ARSENIC  FROM  CADMIUM. 

Chemically  pure  cadmium  sulphate  was  precipitated  by  a 
solution  of  sodium  pyroarsenate.  Stirring  brought  out  a  gelat- 
inous arsenate,  which  changed  by  additional  stirring  to  a  granu- 
lar salt.  This  was  thoroughly  washed  and  dried  at  no®.  It 
had  the  composition  Cd,As,0^  +  2H,0.  Salkowski'  observes 
that  a  red  heat  is  necessary  to  fully  dehydrate  this  salt. 

The  moisture  and  arsenic  were  completely  expelled  at  150**, 

1  Loc.  cit. 


I040 


J.    BIRD  MOVER.      METAI,  SEPARATIONS 


leaving  a  uniform  mass  of  cadmium  chloride.  It  was  weighed 
as  such  after  standing  over  sulphuric  acid  for  one-half  hour. 
The  arsenic  was  determined  as  usual. 


cd,A»,o,  -h 

3H,0 
taken. 

Cadmium 
chloride 
obtained. 

Cadmium 
chloride 
required. 

Difference. 

Gram. 

Gram. 

Gram. 

Gram. 

Experiment  I . . .  •  0.2359 

0.1965 

0.1977 

—0.0012 

<i 

II  ...  O.I 166 

0.0968 

0.0968 

—0.0000 

<< 

III...  0.1030 

0.0857 

0.0855 

+0.0002 

<( 

IV  ...  0.1138 

0.0947 

0.0946 

+0.0001 

i( 

V  .-..  0.1043 

0.0870 

0.0867 

+0.0003 

Cd,Aa,0,  + 
3H«0  taken. 

AS|09 
obtained. 

Aa^O, 
required. 

Difference 

Gram. 

Gram. 

Gram.. 

Gram. 

Ezperitneni 

t  I  ....   0.2359 

0.0813 

0.0822 

0.0009 

The  cadmium  chloride  dissolved  perfectly  in  water  and  showed 
no  arsenic,  when  tested  in  a  Marsh  apparatus. 

XIII. — THE    ACTION    OF    HYDROCHLORIC    ACID   GAS    ON   FERRIC 

OXIDE. 

Pure  oxide  of  iron  was  heated  in  a  stream  of  acid  gas.  The 
behavior  of  iron  is  rather  peculiar,  as  it  very  readily  changes 
into  chloride,  and  then  only  partially  volatilizes.  On  heat- 
ing to  200**  the  greater  part  is  driven  over  as  flaky  crystals 
of  ferric  chloride.  The  remainder  consists  of  a  white  mass, 
which  refuses  to  go  over  on  prolonged  action  and  also  on  raising 
the  temperature. 

This  residue  was  soluble  in  water  and  did  not  react  with 
potassium  thiocyanate,  but  immediately  gave  a  blue  precipi- 
tate with  ferricyanide.  Reduction  was  therefore  evident ;  this 
is  also  noted  by  Jannasch  and  Schmidt.*  The  temperature  at 
which  ferric  chloride  usually  goes  into  the  ferrous  condition  is 
above  1000**. 

Care  was  taken  to  prepare  perfectly  pure  hydrochloric  acid 
gas.  Chemically  pure  acids  were  used  to  this  end.  The  action 
however  was  the  same  in  all  cases. 

XIV. — ^THE   SEPARATION   OF  ARSENIC   FROM  IRON. 

Chemically  pure  ferrous  ammonium  sulphate  was  carefully 

1  Loc.  cit. 


BY   MEANS  OF    HYDROCHLORIC   ACID   GAS.  IO41 

oxidized  with  nitric  acid,  it  was  taken  up  in  water,  filtered  and 
then  crystallized  several  times.  The  best  crystals  were  selected 
and  a  solution  made  to  precipitate  the  arsenate.  A  white  pre- 
cipitate tinged  with  yellow  was  formed.  It  was  washed  by 
decantation  and  then  filtered  and  washed  until  the  washings  no 
longer  gave  Prussian  blue  with  ferrocyanide.  It  was  then  dried 
and  gently  ignited. 

The  acid  gas  acts  on  it  quickly  in  the  cold  and  it  becomes  a 
light  green  liquid.  In  evaporating  off  the  moisture  the  chloride 
of  iron  was  carried  over  with  the  arsenic. 

In  a  second  trial,  .with  the  temperature  lower  and  occasionally 
removing  the  source  of  the  heat  altogether,  when  ebullition 
threatened  to  cause  spattering,  ferric  chloride  was  obtained  with- 
out loss.  This  was  gradually  heated  a  little  higher  to  remove 
all  the  arsenic. 

The  chloride  of  iron  was  dissolved,  oxidized,  precipitated  with 
ammonium  hydroxide  and  estimated  as  usual.  The  result  was 
fair  and  the  product  tested  showed  the  absence  of  arsenic,  but 
all  succeeding  experiments  failed.  Either  the  substance  spat- 
tered or  the  iron  went  along  with  the  arsenic. 

Jannasch  and  Schmidt*  separated  arsenic  from  iron  by  placing 
their  material  in  a  large  hard  glass  bulb  and  evaporating  down 
to  dryness  with  nitric  acid,  in  an  air  current.  This  is  not  appli- 
cable when  a  porcelain  boat  is  employed.  They  then  volatilized 
the  arsenic  in  hydrochloric  acid  gas  at  120**. 

XV. — SEPARATION  OF  ARSENIC  PROM  ZINC. 

In  some  preliminary  work  zinc  oxide  was  treated  with  acid 
gas  at  200**.  It  completely  changed  to  chloride  and  was  not 
volatile.  Pure  zinc  sulphate  was  used  to  precipitate  the  arse- 
nate; it  was  washed,  dried  and  ignited  to  150**.  The  same  diffi- 
culty appeared  as  was  encountered  under  iron.  Zinc  arsenate 
melts  down  to  a  liquid  mass  as  soon  as  the  acid  gas  strikes 
it,  which  is  extremely  hard  to  evaporate  without  spattering.  A 
small  glass  cover  was  placed  over  the  boat,  which  tended  to  les- 
sen the  spattering,  but  did  not  entirely  prevent  it. 

The  zinc  was  estimated  by  taking  the  chloride  up  in  a  little 

^Loccit. 


I042  J.    BIRD   MOVER.      MBTAI.  SEPARATIONS 

hydrochloric  acid  and  running  it  down  with  pure  mercuric 
oxide.  It  was  then  ignited  and  weighed  as  zinc  oxide.  One 
good  result  was  obtained,  but  generally  the  residues  of  zinc  con- 
tained arsenic  and  the  results  were  far  from  being  concordant. 

XVI. — THE    SEPARATION    OF    ARSENIC    PROM    COBALT  AND 

NICKEL. 

Cobalt  and  nickel  were  precipitated  as  arsenates  in  the  usual 
manner,  with  a  solution  of  pyroarsenate. 

Cobalt  nitrate,  a  Merck  preparation,  was  carefully  purified; 
considerable  manganese  w^s  found  and  eliminated. 

This  gave  the  pink  salt  Co,As,0^  +  8H,0,  which  was  ignited 
to  the  blue  anhydrous  compound. 

Cobalt  arsenate  is  ver>'  readily  attacked  by  the  acid  gas  in  the 
cold,  yielding  a  pink  chloride.  A  slight  heat,  not  much  above 
120°,  changed  it  to  the  blue  chloride  and  drove  out  the  arsenic. 
At  first  it  was  quickly  weighed  as  chloride,  then  it  was  taken  up 
in  a  little  hydrochloric  acid  and  evaporated  down  with  mercuric 
oxide.     On  ignition,  black   Co,0^  was  obtained  and  weighed. 

The  arsenic  was  estimated  as  usual. 

Experiment  I.         Experiment  II. 
Gtum.  Gram. 

Co„As,0(,  taken '••       0.1509  0.2029 

CoCl,  obtained 0.1309  

CoCl,  required o.  1293                     .... 

CojOi  obtained 0.0738                   0.0969 

CojO,  required 0.0731                   0.0983 

Difference +0.0007               —0.0014 

AsjOj  obtained 0.0770 

AS7O5  required 0.0764 

Difference +0.0006 


.  a  •  . 


... 


.... 


On  testing  the  cobalt  residue  by  the  Marsh  test,  no  trace  of 
arsenic  was  found.  No  cobalt  was  found  in  the  sublimate. 
Some  of  the  first  experiments  gave  cobalt  too  low;  it  was  thought 
that  they  had  been  heated  too  high,  but  testing  showed  no  vola- 
tilized cobalt. 

A  temperature  of  125**  is  sufficient  to  drive  out  all  of  the  arse- 
nic, and  at  this  temperature  there  is  no  danger  of  volatilizing 
the  cobalt. 

In  working  with  nickel,  the  green  arsenate  was  simply  dried 


BY   MEANS  OF    HYDROCHLORIC   ACID   GAS.  IO43 

in    the   first  experiment.     It  'therefore    had    the   composition 
Ni,As,0,  +  8H,0. 

Hydrochloric  acid  gas  attacked  it  in  the  cold.  A  slight  heat 
drives  out  the  arsenic  and  moisture  and  leaves  a  salmon-colored 
chloride.  The  nickel  chloride  was  changed  to  oxide  by  evapo- 
rating it  with  nitric  acid  and  igniting. 

Experiment  I. 
Gram.     ^ 

Ni,As,Og  +  8H,0  taken 0.1502 

NiO  obtained 0.0554 

NiO  required 0.0561 

Difference — 0.0007 

In  Experiments  II  and  III  the  salt  was  made  anhydrous  by 
ignition. 

Experiment  II.  Experiment  III. 

Gram.  Gram. 

NisAs^O^  taken o.i  166  0.1040 

NiO  obtained 0*0577  0.0523 

NiO  required 0.0575  0.0513 

Difference -I-0.0002  -|-o.ooio 

ASfOj  obtained 0.0515 

AsjOj  required .  • . .  0.0526 

Difference ....  — o.ooii 

The  Marsh  test  showed  no  arsenic  with  the  nickel. 

XVII. — BEHAVIOR  OF  MINERALS   IN   HYDROCHLORIC   ACID  GAS. 

Niccolite.  One-half  gram  of  the  mineral  was  finely  powdered 
and  subjected  to  the  action  of  acid  gas  for  a  day,  at  a  tempera- 
ture of  200''  C.     It  was  only  ver>'  slightly  affected. 

A  second  portion  was  dissolved  in  nitric  acid  and  evaporated 
doivn  in  a  porcelain  dish.  It  was  then  transferred  to  a  boat  and 
evaporated  to  dryness.  To  remove  all  the  acid,  it  was  heated  in 
an  oven  to  no''  for  one-half  hour.  The  dry  substance  was 
acted  upon  bj'  the  acid  gas  in  the  cold  for  five  hours.  It 
changed  completely  to  chloride.  A  temperature  of  150"  for  an 
hour  removed  all  the  moisture  and  arsenic. 

The  nickel  chloride  was  evaporated  dowm  with  nitric  acid, 
ignited,  and  weighed  as  NiO.  The  arsenic  was  estimated  as 
usual. 


I044  JOSEPH   GILLINGHAM   HIBBS.      ATOMIC 

Per  cent 

Nickel  found 43*79 

Nickel  calculated 43-6o 

Difference o.  19 

Arsenic  found 56.66 

Arsenic  calculated 56.40 

Difference 0.26 

Undoubtedly  there  is  still  a  wide  field  open  in  regard  to  the 
behavior  of  hydrochloric  acid  gas  upon  mineral  species.  Smith 
and  Hibbs*  showed  that  mimetite  lost  its  arsenic  quantitatively, 
when  heated  in  a  stream  of  acid  gas.  In  this  laboratory  others 
are  being  investigated  with  favorable  indications.  The  direct 
employment  of  hydrochloric  acid  gas  upon  a  powdered  mineral 
would  simplify  many  a  tedious  gravimetric  process,  leaving  the 
separated  elements  in  a  desirable  condition  for  further  treat- 
ment. 

In  the  case  of  a  mineral  such  as  niccolite,  where  it  must  first 
be  decomposed  with  nitric  acid  and  then  transferred  to  a  boat, 
the  advantage  is  not  so  great.  This,  however,  can  be  modified, 
so  that  the  time  factor  is  reduced  and  the  advantage  of  the 
method  still  retained.  Instead  of  using  a  boat,  which  has  no 
advantage  unless  the  non-volatile  chlorides  are  to  be  weighed 
directly,  a  hard  glass  bulb  can  be  substituted.  The  mineral  is 
placed  in  the  bulb,  dissolved  in  nitric  acid,  and  evaporated 
down  by  the  aid  of  a  current  of  air  drawn  through  the  bulb. 

The  residual  oxides  are  then  separated  in  a  stream  of  hydro- 
chloric acid  gas  as  usual. 


[Contribution  from  thk  John  Harrison  Laboratory  of  Chbbcistry, 

No.  15.] 

THE  ATOMIC  WEIGHTS  OF  NITROGEN  AND  ARSENIC 

By  Joseph  Dillingham  Hibbs. 

Received  September  a6.  i8g6. 

THE  atomic  weight  of  the  metal  molybdenum  had  been 
determined  by  expelling  molybdic  acid  from  sodium 
molybdate  with  hydrochloric  acid  gas,  then  weighing  the  resid- 
ual sodium  chloride. 

1  Loc.  cit. 

2  From  author's  thesis  presented  to  the  Faculty  of  the  UniverBity  of  PeoDsylvania 
for  the  degree  of  Doctor  of  Philosophy,  1896. 


WEIGHTS  OF   NITROGEN  AND  ARSENIC.  IO45 

Having  found  that  nitric  acid  and  arsenic  acid  were  driven 
from  their  alkali  salts  with  ease,  leaving  a  chloride  that  was 
absolutely  pure,  and  believing  that  the  atomic  masses  of  nitro- 
gen and  arsenic  determined  in  this  manner  would  a£Ford  a  valu- 
able contribution  to  the  literature  relating  to  these  constants,  a 
carefully  conducted  series  of  experiments  was  made  with  two 
nitrates  and  one  arsenate.  The  results  are  given  in  detail  in  the 
following  lines : 

THE  ATOMIC  WEIGHT  OP  NITROGEN. 

In  the  past,  determinations  of  the  atomic  weight  of  nitrogen 
have  been  made  from  the  density  of  the  gas  itself,  from  the  ratio 
between  ammonium  chloride  and  silver,  and  from  the  decompo- 
sition of  certain  nitrates.  The  first  method  in  particular  has 
been  frequently  applied.  Thomson,  Dulong,  Berzelius,  and 
Lavoisier  brought  to  light  many  new  facts  relating  to  the  atomic 
weight  of  nitrogen;  unfortunately,  however,  considerable  that 
they  have  presented  has  been  affected  by  complications  that 
have  introduced  inaccuracies. 

Dumas  and  Boussingault*  found  the  mean  density  of  nitrogen 
to  be  0.972  ;  for  hydrogen  they  found  a  mean  density  of  0.0693, 
which  would  give  nitrogen  an  atomic  weight  of  14.026.  Reg- 
nault  obtained  a  more  concordant  series  of  results,  the  mean 
being  0.97137,  and  a  density  for  hydrogen  of  0.0692,  which 
makes  the  atomic  weight  of  nitrogen  equal  to  14.0244. 

Clarke  gives  in  detail  his  computation  of  the  means  of  the 
results  obtained  by  Penny,  Stas,  and  Marignac.  Their  work  on 
the  determination  of  the  atomic  weight  of  this  particular  ele- 
ment was  mainly  on  the  ratio  of  ammonium  chloride  and  silver, 
and  the  decomposition  of  certain  nitrates.  A  great  degree  of 
accuracy  was  maintained  throughout  the  entire  investigation  ; 
but  the  amount  of  work  required  to  obtain  a  single  result  neces- 
sarily lays  the  method  open  to  a  serious  error  of  manipulation. 

In  this  connection  a  paragraph  from  Clarke's  '*  A  Recalcula- 
tion of  the  Atomic  Weights"  may  be  cited:  **The  general 
method  of  working  upon  these  ratios  is  due  to  Penny.  Applied 
to  the  ratio  between  the  chloride  and  nitrate  of  potassium,  it  is 

1  Compi.  rend.^  1841—12.    1005. 


1046  JOSEPH  GILLINGHAM   HIBBS.      ATOMIC 

as  follows :  A  weighed  quantity  of  the  chloride  is  introduced 
into  a  flask  which  is  placed  upon  its  side  and  connected  with  a 
receiver.  An  excess  of  pure  nitric  acid  is  added,  and  the  trans- 
formation is  gradually  brought  about  by  the  aid  of  heat,  the 
nitrate  being  brought  into  a  weighable  form.  The  liquid  in  the 
receiver  is  also  evaporated,  and  the  trace  of  solid  matter  which 
has  been  mechanically  carried  over,  is  recovered  and  also  taken 
into  account.*' 

The  method  indicated  in  this  study,  and  actually  applied  with 
the  results  appended,  is  decidedly  less  objectionable.  In  this 
method  there  is  no  distillation,  no  precipitate,  in  fact,  nothing 
that  could  involve  serious  error. 

Clarke  summarizes  the  results  of  Penny,  Stas,  and  Marignac 
as  follows : 

1.  From  specific  gravity  of  N N  =  14.0244 

2.  ' *  ammonium  chloride N  ^  14.0336 

3.  **  ratio  number  four N  =s  14.0330 

4.  *'  silver  nitrate N  =  13.9840 

5.  *'  potassium  nitrate N  =  13.9774 

6.  '*  sodium  nitrate N  =  13.9906 

Mean  of  results  for  N N  =  14.0210 

If  oxygen  is  16,  this  becomes  14.0291.  Stas  found  the  atomic 
weight  of  nitrogen  to  be  14.044.  Dumas  found  14  by  experi- 
ments on  the  combustion  of  ammonia  and  cyanogen  (0=  16). 
Pelouze  found  14.014  by  bringing  a  known  weight  of  silver 
nitrate  in  contact  with  a  known  and  slightly  excessive  weight  of 
ammonium  chloride,  which  excess  was  titrated.  Anderson 
found  13.95  by  the  decomposition  of  the  nitrate  of  lead,  with 
just  enough  heat  for  decomposition  (the  same  method  that  was 
used  by  Berzelius).  Marignac  found  14.02  by  dissolving  a 
known  weight  of  silver  in  nitric  acid  and  then  melting  and 
weighing  the  nitrate  found. 

A. — ATOMIC   WEIGHT   OF   NITROGEN   BY   ACTION  OF   HYDROGEN 

CHLORIDE  UPON  POTASSIUM  NITRATE. 

The  purest  salt  obtainable  was  dissolved  in  water,  filtered, 
and  recrystallized  six  times,  a  solution  of  which  was  tested  for 
chlorides,  sulphates,  etc.,  but  no  impurity  was  found.  One 
more   crystallization   was   made   and    the    best   crystals    were 


WEIGHTS  OF    NITROGEN  AND   ARSENIC.  IO47 

selected.  These  were  washed  with  distilled  water  and  dried  at 
210®  C.  for  three  hours,  powdered,  and  again  dried,  and  finally 
placed  in  a  weighing  bottle.  This  compound  was  dried  before 
each  experiment.  It  was  also  allowed  to  stand  in  a  balance 
case  one  hour  before  weighing.  The  same  degree  of  care  was 
exercised  in  the  preparation  of  the  boat  for  weighing. 

The  weighing  bottle  was  placed  on  the  scale  pan  and  allowed  to 
stand  several  minutes  in  order  to  regain  its  normal  temperature. 
After  weighing  it  was  quickly  opened  and  a  portion  of  the  salt 
removed  to*  the  boat  and  again  closed  and  allowed  to  stand  in 
the  balance  case  for  several  hours  before reweighing.  The  boat 
was  then  introduced  into  the  combustion  tube  and  the  gas 
passed  over  it.  The  characteristic  action  took  place.  The  only 
difference  in  the  method  of  procedure  adopted  here  and  that 
described  in  the  first  section  of  this  paper,  was  a  longer  time 
being  given  to  complete  the  action,  using  a  lower  temperature, 
in  order  to  do  away  with  all  possibility  of  fusion  of  the  salt.  It 
was  then  carefully  removed  to  a  vacuum  desiccator  and  allowed 
to  stand  over  night  before  weighing.  It  may  be  said  also  that 
experiments  were  only  conducted  on  clear  days  to  insure  the 
non-entrance  of  moisture. 

With  potassium  nitrate,  no  great  variation  of  amount  was 
taken. 

Five  determinations  were  made  in  this  case  : 


«.-  -  ***  •*•  sua  «*•  Oi-z 


S«           SoS         £Si         ^lo         tt.=        t-^lc      .S^Sgl  5'=| 

0        l«            ol-          o|«          51-=          "S'^l         S'S2"S       o'S|i:.=  2 --3 

Z       Zt          (SoS         u&i:         5s.o         u^B       u^Sr      S^a  =  5  <'o5 

Gram.        Gram.        Gram.        Gram.        Gram.        Gram. 

1.  O.I  1084    0.08173    0.00006    o.cxxxu    O.I  1090    0.08177    O.IOII2I  14.01 1 

2.  0.14864    0.00960    0.00007    0.00005    0.14871    o.ro965    o. ion 20  14.010 

3.  0.21056    0.15525    o.oooii    0.00008    0.21067    0.15533    0.101123  14.013 

4.  0.23248    0.17214    0.00012    0.00009    0.23360    0.17223    0.101121  14.011 

5.  0.24271    0.17894    0.00013    0.00009    0.24284    0.17903    o. ion 24  14.014 

Atomic  weight  of  nitrogen  =  14.01 18  ±  0.000472. 

The  atomic  values  used  in  these  calculations  were  taken  from 

'*  Table  of  Atomic  Masses,"  revised  by  F.  W.  Clarke,  in  Octo- 
ber, 1891. 


1048  JOSEPH  GILLINGHAM   HIBBS.      ATOMIC 

The  figures  deduced  from  these  values  are,  of  course,  subject 

to  any  change  made  by  later  revision  of  atomic  weights.    It  is 

not  so  much  the  exact  figure  to  which  attention  is  called,  as  to 

the  constancy  of  result  brought  forward  by  this  method.    The 

values  used  were  : 

Oxygen 16.00 

Potassium 39-ii 

Chlorine 3S.45 

Specific  gravity  potassium  nitrate 2.1 

Specific  gravity  potassium  chloride 1.99 

B. — ATOMIC  WEIGHT    OP   NITROGEN   BY  ACTION  OF  'HYDROGEN 

CHLORIDE  UPON  SODIUM  NITRATE. 

The  same  degree  of  care  and  method  of  procedure  were  here 
observed  as  in  Division  A.     The  results  are  as  follows : 

fig  «•§  s-  s^  s-  o'^  ^     ^  tZ 

iJS  I  .2a  .2  .2=  o  «^-3         ^S 

•a5  E-g  ^6  oS  «a  wS  sss        if2 

5*^  -SJi  ta  ts «'  ts  us; «      "s--^        §a 

Z       ^i:  -xc  uS  uSc  uz  uSc      S^S        <o 

Gram.  Gram.  Gram.         Gram.          Gram.  Gram. 

1.  0.01550  0.01064         ....              ....  0.01550  0.01066  85.061  I4.OII 

2.  0.20967  0.14419  0.00009  0.00007  0.20976  0.14426  85.061  14.OII 

3.  6.26217  0.18029  0.00012  o.oooc>9  0.26229  0.18038  85.064  14.014 

4.  0.66610  0.46805  0.00035  0.00024  0.66645  0.45829  85.064  14.014 

5.  0.93676  0.64422  0.00042  0.00034  0.93718  0.64456  85.058  14.008 

Atomic  weight  of  nitrogen  =3  14.01 16  ±  0.000741. 

Atomic  values  used  were 

Oxygen 16.00 

Sodium 23.05 

Chlorine 35.45 

Specific  gravity  sodium  chlari(^e 2.16 

Specific  gravity  sodium  nitrate 2.26 

When  these  results  are  compared  with  those  obtained  by 
Penny  and  Stas  by  treatment  of  potassium  chloride  with  nitric 
acid,  and  the  treatment  of  potassium  gitrate  with  hydrochloric 
acid  (likewise  for  sodium),  a  close  comparison  can  be  made. 

Penny.                                                 Hydrogen  chloride  method. 
For  potassium  nitrate  >  *  •  >  13.9774       For  potassium  nitrate ....  14.0118 
"    sodium  nitrate 13.9906  "    sodium   nitrate 14.0116 

Sfiowing  a  difference  of 

0.0344  for  potassium  salt, 
0.0210  for  sodium  salt. 


WEIGHTS  OF  NITROGEN   AND   ARSENIC.  IO49 

When  a  mean  of  the  above  results  is  taken,  the  atomic  weight 

of  nitrogen  equals 

13.9996  for  potassium  salt, 
14.0011  for  sodium  salt. 

Taking  now  a  mean  of  these  values,  the  atomic  weight  of 
nitrogen  would  be  14.0003. 

C. — THE  ATOMIC  WEIGHT  OF  ARSENIC. 

The  atomic  weight  of  arsenic  has  been  obtained  from  the  chlo- 
ride (AsCl,),  the  bromide  (AsBr,),  and  the  trioxide  (As,0,). 

Pelouze,  in  1845,^  and  Dumas,  in  1859,  determined  it  by  the 
titration  with  known  quantities  of  pure  silver  in  the  analysis  of 
arsenic  trichloride.  The  mean  of  their  results,  as  computed  by 
Clarke,  giveS  the  atomic  weight  of  arsenic,  74.829.  Wallace* 
makes  the  same  titration  with  silver  in  the  analysis  of  arsenic 
tribromide.  His  value  is  74.046.  Kessler  made  a  set  of  deter- 
minations  by  estimating  the  amount  of  potassium  bichromate 
required  to  oxidize  100  parts  of  arsenic  trioxide  to  arsenic  pent- 
oxide.     He  obtained  a  mean  value  of  75.002. 

A  mean  of  these  results  gives  the  following  : 

From  AsClj 74  829 

**     AsBF) 74-046 

**      ASjO, 75002 

General  mean  # 74-9i8 

If  oxygen  =  16,  then  the  atomic  weight  of  arsenic  will  equal 

75090. 

Berzelius,  in  1826,  heated  sulphur  and  arsenic  trioxide 
together  in  such  a  way  that  sulphur  dioxide  alone  escaped  ;  this 
method  gave  74.840  as  the  atomic  weight  of  arsenic.  But  one 
experiment  was  made,  so  that  it  does  not  possess  much  value. 
In  the  above  method  there  seems  to  be  a  wide  variation  in  the 
results  obtained,  the  difference  between  the  extreme  values  is 
but  little  less  than  one  unit. 

By  the  hydrogen  chloride  method,  we  have  but  the  weighing 
of  the  material  used  in  the  determination — which  must  neces- 
sarily enter  every  estimation  or  analysis — and  a  single  weighing 
after  the  action  of  the  acid  gas.  As  in  the  case  of  nitrogen,  the 
method  seems  to  be  as  short  and  concise  as  possible. 

1  Compt.  rend.^  xo,  X047. 
sphiLMag.(4),  tt,  279. 


I050        ATOMIC   WEIGHTS  OP   NITROGEN   AND   ARSENIC. 

The  methods  and  modus  operandi  were  exactly  the  same  as 
those  used  in  the  determination  of  the  atomic  weight  of  nitrogen. 

The  sodium  chloride  obtained  was  perfectly  white  in  color. 
In  no  instance  was  it  fused.  After  weighing  the  salt  residue  it 
showed  no  traces  of  arsenic,  and  was  readily  soluble  in  cold 
water  without  residue.  The  same  conditions  of  atmosphere 
were  obser\'ed. 

As  the  specific  gravity  of  sodium  pyroarsenate  could  not  be 
obtained,  it  was  determined  by  means  of  the  specific  gravity  bot- 
tle, against  chloroform,  and  was  found  to  be  2.205,  while  the 
specific  gravity  of  sodium  chloride  was  taken  as  2.16.  The 
atomic  values  used  were  : 

Oxygen i6.oo 

Sodium 23.Q5 

Chlorine 35.45 

The  results  here  obtained,  besides  being  to  a  great  degree 
constant,  compare  favorably  with  those  obtained  by  Pelouze 
(74.829)  and  Kessler  (75.002). 

A  coincidence  may  here  be  shown  by  the  fact  that  the  mean 
of  these  values  gives  74.9155,  while  the  hydrogen  chloride 
method  gives  74.9158. 

In  order  to  give  the  method  a  thorough  trial,  the  amounts 
taken  cover  a  wide  range.  The  smallest  amount  used  was 
0.02176  gram  of  sodium  pyroarsenate,  and  the  largest  3.22485 
grams.  It  will  also  be  noticed  that  the  variation  in  result  is  but 
0.027  ^or  ten  determinations. 

25         i-c        U        Ii         og         ^i  i        \ 

0-5  ^.5  =0.  c*'  30.  aW  w^g,  vj 

-i  ^2  -2    ii         .2  S     V         .2  «oo:  ^-g 

=  5  =•£  weed  "5  S  «S"S  oS  Z'^^v         JH^ 

Gram.         Gram.         Gram.         Gram.         Gram.         Gram. 

1.  0.02176  0.01439  o.ooooi  0.00000  0.02177  0.01439  354-oo8  74.904 

2.  0.04711  0.03 1 14  0.00002  0.00001  0.04713  0.031 15  354.042  74.921 

3.  0.05792  0.03828  0.00003  0.00002  0.05795  0.03830  354.054  74.927 

4.  0.40780  0.26970  0.00021  o.oooii  0.40801  0.26981  354.002  74.901 

5.  0.50440  0.33028  0.00026  0.00017  0.50466  0.33045  354.033  74.916 

6.  0.77497  0.51222  0.00041  0.00027  077538  0.51249  354.034  74.917 

7.  0.82853  0.547^  0.00044  0.00029  0.82897  0.54791  354.034  74.917 

8.  1. 19068  0.78690  0.00056  0.00041  1.19K24  0.78731  354.053  74.926 

9.  1.67464  1.10681  0.00081  0.00051  1.67545  1.10732  354.057  74.928 
10.  3.22485  2.13168  0.00152  0.00099  3-"637  2.13267  354.002  74,901 

Atomic  weight  of  arsepic  =  74.9158  ±  0.00222. 


[Contribution  from  the  John  Harrison  Laboratory  of  Chemistry, 

No  i6.] 

THE  SEPARATION  OF  VANADIUM  FROM  ARSENIC. 

Bv  Charles  Field,  3rd,  and  Edgar  F.  Smith. 

Received  October  a.  1896. 

AS  vanadium  and  arsenic  occur  associated  in  minerals  and 
likewise  in  artificial  products,  their  separation  becomes  a 
matter  of  consequence. 

The  course  usually  pursued  in  carrying  out  this  separation  is 
that  long  since  recommended  for  the  removal  of  vanadic  acid 
from  its  solutions ;  namely,  its  precipitation  as  ammonium  meta- 
vanadate.  Other  methods  have  recently  appeared  in  the  litera- 
ture bearing  on  analysis.  Reference  is  here  made  especially  to 
the  publication  of  Fischer.* 

Experiments  made  in  this  laboratory  on  the  behavior  of  vana- 
dates' and  arsenates'  heated  in  an  atmosphere  of  hydrochloric 
acid  gas,  in  which  both  acids  were  volatilized,  suggested  the 
thought  that  if  the  sulphides  of  vanadium  and  arsenic  were 
exposed  to  the  same  vapors  perhaps  they  would  show  a  variation 
in  deportment.  And  so  it  has  proved.  Perfectly  dry  arsenic 
trisulphide,  previously  washed  with  alcohol,  carbon  disulphide, 
and  ether,  then  dried  at  ioo°  C,  when  exposed  in  a  porcelain 
boat,  placed  in  a  combustion  tube,  was  almost  completely 
expelled  from  the  retaining  vessel  at  the  ordinary  temperature. 
The  last  traces  were  driven  out  at  a  temperature  little  above 
150**  C.  Brown  vanadium  sulphide,  in  a  perfectly  dry  condition, 
treated  in  the  same  manner,  was  not  altered.  It  only  remained 
then  to  prepare  mixtures  of  known  amounts  of  the  two  sulphides 
and  subject  them  to  the  action  of  the  acid  vapor.  To  this  "end 
the  following  experiments  were  made  : 

I.  0.1303  gram  of  vanadium  sulphide, 
0.1302  gram  of  arsenic  sulphide. 

The  arsenic  sulphide  was  volatilized  without  difficulty  and  left 
0.1297  gi*^™  of  vanadium  sulphide. 

1  Bestimraung  von  Vanadins&ure :  Dissertation,  Rostock,  1894. 
3/.  Am.  Chem.  Soc.,  16,  578. 
t/Mf..  17.6S2. 


1 052  SEPARATION   OF   VANADIUM    FROM   ARSENIC. 

II.  0.1290  gram  of  vanadium  sulphide, 
0.2242  gram  of  arsenic  sulphide, 

gave  after  exposure  of  one  hour  to  hydrochloric  acid  vapor  a 
residue  of  vanadium  sulphide,  weighing  0.1297  gram. 

III.  0.0828  gram  of  vanadium  sulphide, 
0.0582  gram  of  arsenic  sulphide, 

left  0.0827  gi'sni  of  vanadium  sulphide. 

IV.  0.1306  gram  of  vanadium  sulphide, 
0.2028  gram  of  arsenic  sulphide, 

gave  a  residue  of  0.1308  gram  of  vanadium  sulphide. 

V.  0.1403  gram  of  vanadium  sulphide, 
0.2409  gram  of  arsenic  sulphide, 

left  0.1404  gram  of  vanadium  sulphide. 

The  temperature  in  these  experiments  was  not  allowed  to 
exceed  250*^  C,  as  beyond  that  point  there  is  danger  of  affecting 
the  vanadium  and  causing  its  partial  volatilization. 

The  method  worked  so  well  and  with  such  evidently  favorable 
results  that  the  following  course  was  adopted  in  the  analysis  of  a 
specimen  of  the  mineral  vanadinite.  0.2500  gram  of  air-dried 
and  finely  divided  material  was  placed  in  a  porcelain  boat :  the 
latter  was  then  introduced  into  a  combustion  tube  and  gently 
heated  in  a  current  of  dry  hydrochloric  acid  gas.  By  this  treat- 
ment vanadic  and  arsenic  oxides  were  expelled,  leaving  lead 
phosphate  and  chloride.  The  receiver  containing  the  vanadium 
and  arsenic  was  made  alkaline  and  digested  with  ammonium 
sulphide.  From  the  solution  of  the  sulpho-salts  the  vanadium 
and  arsenic  sulphides  were  set  free  by  a  dilute  acid.  After 
washing  and  careful  drying  these  sulphides  were  separated  as 
indicated  in  the  preceding  lines,  then  changed  to  oxides  and 
determined  in  the  usual  manner.  The  sum  of  the  total  con- 
stituents determined  as  lead  oxide,  phosphoric  oxide,  vanadic 
and  arsenic  oxides,  with  some  lead  chloride,  amounted  to  0.2501 
gram. 

The  method  in  addition  to  being  satisfactory  in  the  analytical 
way,  certainly  forms  a  very  excellent  means  of  purifying  and 
freeing  vanadium  from  arsenic. 


[Contribution  prom  the  John  Harrison  Laboratory  of  Chem- 
istry, No.  17.] 

THE    SEPARATION    OF    HANQANESE    FROM    TUNGSTIC 

ACID. 

By  Walter  T.  Taggart  and  Edgar  F.  Smith. 

Received  Oc toiler  3,  1896. 

THE  necessit}'  of  obtaining  pure  tungstic  acid  from  time  to 
time,  using  wolframite  as  the  starting  out  material,  has 
frequently  suggested  the  inquiry  as  to  what  course  would  proba- 
bly prove  the  best  in  the  quantitative  separation  of  this  acid 
from  oxides,  such  as  those  of  iron  and  manganese. 

In  the  experiments  recorded  in  this  communication  only  the 
results  obtained  from  a  study  of  mixtures  of  a  manganous  salt  and 
a  soluble  alkali  tungstate  will  be  given.  The  directions  taken 
in  the  experimentation  were,  ist,  to  effect  the  separation  by  the 
use  of  3'ellow  ammonium  sulphide  in  the  presence  of  ammonium 
chloride  ;  2nd,  to  eliminate  the  acid  oxide  by  the  use  of  an 
alkaline  carbonate. 

Following  thefirst  course, mixturesof  definite  amounts  of  ammo- 
nium tungstate  and  manganous  chloride  were  made.  To  these 
was  added  water  and  a  considerable  excess  of  yellow  ammonium 
sulphide,  together  with  ammonium  chloride.  The  mixtures 
were  digested  on  a  water-bath  at  70°  C,  for  several  hours,  and 
the  vessels  containing  them  were  then  closed  and  allowed  to 
stand  during  the  night.  The  manganese  sulphide  was  filtered 
out,  and,  after  solution,  was  changed  into  sulphate  and  weighed 
as  such,  or  it  was  finally  obtained  as  protosesquioxide  in  the 
customarj'  way. 

Results, 

Manganous  oxide 
found. 

Gram. 
0.2I2I 
0.2255 

O.I 70S 

0.1720 
0.1760 

III  every  trial  tungstic  acid  adhered  to  the  metallic  oxide. 
In  trying  the  second  suggestion  the  soluble  tungstate  and  the 


Manganous  oxide 
present. 

Gram. 

0.1950 

0. 1949 

0.1290 

0.1287 

O.1291 

I054      SEPARATION  OF  MANGANESE  FROM  TUNGSTIC  ACID. 

soluble  manganous  salt  were  digested  for  some  hours  in  a  plati- 
num dish,  upon  a  water-bath,  with  an  excess  of  a  ten  pe:  cent, 
potassium  carbonate  solution,  after  which  the  whole  was  evapo- 
rated to  dryness,  the  residue  boiled  up  with  water,  the  mangan- 
ous  carbonate  filtered  out,  washed,  and  finally  converted  into 
protosesquioxide. 

Results. 

Manganous  oxide  Manganous  oxide 

present.  found. 

Gram.  Gram. 

0.194)9  O.1516 

0.1949  0.1534 

Several  trial  were  made  using  a  fifty  per  cent,  solution  of 
potassium  carbonate. 

Results. 

Manganous  oxide  Manganous  oxide 

present.  found. 

Gram.  Gram. 

O.195I  0.1745 

0.1950  0.1528 

The  experimental  evidence  given  in  the  preceding  paragraphs 
leaves  no  doubt  as  to  the  insufficiency  of  the  two  methods, 
which  were  tried,  in  effecting  the  desired  separation.  It  is 
probable  that  fusion  with  an  alkaline  carbonate  will  alone 
answer  for  this  purpose.  How  complete  that  course  would  be 
can  only  be  ascertained  by  careful  experimentation. 


In  the  course  of  analysis  molybdenum  is  quite  often  obtained 
as  sulphide.  Its  conversion  into  a  weighable  form  is  attended 
with  more  or  less  difficulty.  Trials  made  in  connection  with  its 
estimation  show  that  if  the  sulphide,  as  generally  obtained,  be 
dried,  then  intimately  mixed  with  anhydrous  oxalic  acid,  its 
careful  ignition  to  trioxide  can  be  made  quite  rapidly. 

Results. 

Molybdenum  trioxide  Molybdenum  trioxide 

taken.  /ound. 

Gram.  Gram. 

0.3000  0.3009 

0.3000  0.2990 

0.1007  O.IOII 


[Contribution  prom  thb  John  Harrison  Laboratory  op  Chemistry, 

No.  i8.] 

THE  SEPARATION  OP  BISMUTH  PROM  LEAD. 

Bt  Arthur  I^  Bbnxbrt  akd  Bdoar  P.  Smith. 

Received  October  t.  i8q6. 

MANY  methods  have  been  suggested  to  effect  this  separa- 
tion. 'In  a  recent  issue  of  the  Zcitschrift  fitr  angewandte 
Chemie  (1895,  p.  530),  Olav  Steen  reviews  thirteen  of  these 
methods  and  concludes  that  an  early  proposal  of  Rose/  in  which 
the  lead  is  thrown  out  as  chloride  and  weighed  as  sulphate^ 
another  by  Lowe,'  in  which  the  bismuth  is  removed  as  basic 
nitrate,  and  a  late  suggestion  madef  by  Jannasch,'  viz,^  the 
expulsion  of  the  bismuth  as.  bromide  from  a  mixture  of  lead  and 
bismuth  sulphides  by  an  air  current  carrying  bromine  are  the 
most  satisfactory.  At  least  these  methods  gave  Steen  the  best 
results.  The  separation  of  bismuth  from  lead  frequently  con- 
fronts the  analyst,  and  any  novelty  in  this  direction  cannot  be 
absolutely  devoid  of  interest,  hence  the  present  communication, 
which  brings  data  that  may  perhaps  prove  of  service  in  the 
hands  of  others  who  are  interested  in  the  solution  of  this  analyt- 
ical problem. 

It  will  be  recalled  that  Herzog^  proposed  to  separate  bismuth 
from  lead  by  precipitating  the  former  as  basic  acetate.  The 
method  required  considerable  time  for  execution,  and  in  other 
hands  than  those  of  its  author  apparently  has  not  yielded  entirely 
satisfactory  results. 

An  idea  closely  related  to  that  of  Herzog  wouM  be  the  sub- 
stitution of  a  formate  solution  for  that  of  the  acetate.  This  was 
done  with  results  that  are  very  interesting. 

Solutions  of  lead  nitrate  and  bismuth  nitrate  in  nitric  acid 
were  made  up  of  such  strength  that  twenty  cc.  of  the  first  con- 
tained 0.2076  gram  of  lead  oxide,  and  twenty  cc.  of  the  second 
0.1800  gram  of  bismuth  trioxide.  The  lead  and  bismuth  were 
accurately  determined  after  dilution  to  a  liter.  Twenty  cc.  of 
these  two  nitrate  solutions  were  then  introduced  into  a  beaker 

1  Ann.  ekem, phys.  Pi^g.*  no,  425. 
sy.  prakt.  Chtm.^  74, 348. 
*  Ber.  d.  ckem,  Ges.^  as,  134. 
4  Ztsckr.  anal.  Chem.,  §7,  650. 


1056  SEPARATION  OF   BISMUTH   PROM   LEAD. 

glass,  carefully  diluted  aud  almost  neutralized  with  sodiisn  car- 
bonate, or  until  the  incipient  precipitate  dissolved  slowly,  when 
considerable  sodium  formate  solution  of  sp.  gr.  1.084  an^  a  few 
drops  of  aqueous  formic  acid  were  added.  The  total  dilution  of 
the  liquid  was  250  cc.  It  was  gradually  heated  to  boifing  and 
held  at  that  point  for  five  minutes.  The  precipitate  was  then 
allowed  to  subside,  but  was  filtered  while  yet  hot.  The  basic 
formate  separates  rapidly  and  is  easilj'  washed  if  not  boiled  too 
long.  It  was  washed  with  hot  water,  then  dissolved  in  dilute 
nitric  acid  and  precipitated  with  ammonium  carbonate.  The 
'ignited  bismuth  trioxide  weighed  too  much  ;  it  contained  lead. 
However,  the  impure  oxide  was  dissolved  in  nitric  acid,  diluted 
to  250  cc,  and  after  the  addition  of  sodium  carbonate  to  almost 
complete  neutralization,  sodium  formate  and  free  formic  acid 
were  added  as  before,  and  the  precipitation  of  basic  formate 
repeated.  This  precipitate  after  solution  and  the  bismuth  thrown 
out  by  ammonium  carbonate  gave  0.1804  gi'am  of  bismuth 
oxide  instead  of  0.1800  gram  as  required  by  theory.  Seven 
additional  separations,  in  which  the  quantities  of  bismuth  and 
lead  were  the  same  as  indicated  above,  gave  : 

0.1806  gram  of   Bi^Os. 

0.1806  *•      " 

0.1803  *' 

0.1S04  '' 

0.1804  " 

0.1805  *' 

0.1796  *' 

The  conditions  in  these  determinations  were  similar  to  those 
previously  outlined. 

With  a  solution  containing  0.3600  gram  of  bismuth  oxide  and 
0.2076  gram  of  lead  oxide,  operating  in  an  analogous  manner, 
two  results  were  obtained : 

0-3595  gram  of  Bi,Oj. 
0.3605      "      •*      " 

instead  of  the  required  0.3600  gram. 

The  residual  bismuth  trioxide  was  examined  for  lead,  but 
none  was  found. 


i(  • 


c< 

(I 
<( 
(« 


[Contributions  prom  the  Chbmicai,  Laboratory  op  ths  Univbrsity 

OF  Cincinnati.] 

XLVIII.— ON  SOME  NEW  FORMS  OF  QAS  GENERATORS.* 

By  Thomas  H.  Norton. 

Received  August  tj,  SS96. 

IMPROVEMENTS  in  the  construction  of  the  automatic  gener- 
ators, for  the  gases  most  frequently  used  in  our  laboratories, 
are  always  welcome.  The  following  three  types,  which  I  devised 
some  time  since,  have  been  subjected  to  prolonged  trial  in  the  lab- 
oratory of  the  University,  and  have  given  such  satisfactory  results , 
that  a  detailed  description  would  seem  worthy  of  publication. 
In  Fig.  I  is  represented  a  gas  generator  for  hydrogen,  hydro- 
gen sulphide,  etc.,  which  differs  in  several 
details  from  well  known  types  of  the  same 
general  outline.  It  is  constructed  of  glazed 
P  earthenware,  and  is  easily  made  in  our  ordi- 
nary potteries.  A,  the  outside  container,  is 
provided  with  handles  on  the  outside,  and  is 
ordinarily  sixty  cm.  in  height.  Its  chief 
peculiarity  is  the  presence  on  opposite  sides  of 
the  inner  wall,  of  the  shoulders  DDy  each 
about  four  cm.  wide  and  slightly  concave  on 
the  lower  surface.  By  the  gas  reservoir,  is  of 
the  ordinary  bell-jar  construction,  with  orifice 
at  the  top  for  the  introduction  of  a  perforated 
stopper  and  outlet  tube.  It  is  provided  with 
projecting  shoulders  three  cm.  wide,  corresponding  to  DD,  and 
at  such  a  height  that  they  barely  slip  beneath  the  latter.  At  the 
bottom  are  frequent  circular  perforations,  one  centimeter  in 
diameter,  to  allow  of  the  easy  passage  of  the  acid  charge.  The 
recipient  C,  designed  to  hold  zinc  or  any  solid  charge,  is  pro- 
vided with  a  loose  disk  perforated  with  many  fine  openings  and 
resting  upon  the  shoulder  of  the  constriction.  Beneath  the  con- 
striction are  perforations  corresponding  to  those  in  B.  A  strong 
copper  wire  or  rod,  passing  through  the  perforations  of  both 
parts  of  the  apparatus,  holds  B  and  C  in  their  mutual  position  to 
each  other. 

iRead  before  the  American  Association  for  the  Advancement  of  Science  at  the 
Buffalo  Meeting. 


Fio.  I. 


I058 


THOMAS  H.    NORTON.      SOME   NEW 


The  working  of  the  generator  is  exceedingly  simple.  C 
receives  its  charge  of  zinc,  marble  or  ferrous  sulphide,  ^is 
placed  over  it.  The  copper  rod  is  passed  through  the  perfora- 
tions at  the  bottom.  B  with  Cis  then  introduced  into  A,  and 
turned  until  the  shoulders  of  B  are  beneath  DD.  A  is  then 
filled  with  the  acid  charge.  The  buoyancy  of  B  is  partly 
overcome  by  the  rigid  attachment  of  C,  and  entirely  prevented 
by  DD,  Gas  can  be  drawn  off  as  desired,  by  opening  the  tap 
at  the  outlet  tube.  When,  as  naturally  occurs,  the  acid  in  the 
lower  portion  of  the  generator  becomes  weak  and  the  evolution 
of  gas  sluggish,  the  exit  tap  is  closed,  B  is  turned  slightly  so  as 
to  be  free  from  DD,  and  is  then  lifted,  by  grasping  the  neck, 
along  with  the  holder  C,  until  entirely  above  the  surface  of  the 
acid.  Both  are  then  plunged  to  the  bottom  of  A,  and  a  few 
repetitions  of  this  churning  movement  renders  the  acid  charge 
of  uniform  strength. 

This  style  of  generator  has  rendered  excellent  ser\nce.  For 
example,  one  sixty  centimeters  in  height  easily  supplies  all  the 
hydrogen  sulphide  required  by  a  class  of  thirty  in  qualitative 
analysis.  The  special  advantages  of  this  generator  are  to  be 
found  in  the  ease  and  simplicity  with  which  the  buoyancy  of  the  gas 
reservoir  is  overcome  and  the  acid  charge  is  maintained  at  a 
uniform  strength  until  practically  exhausted. 

In  Fig.  2  we  have  a  less  compact  and  less  transportable  form, 

yf  but  one  which  maintains  the  uni- 
j][)L  form  strength  of  the  acid  charge 
until  it  is  exhausted,  without  the 
need  of  special  manipulation,  as 
described  above.  It  is  particularly 
designed  for  use  where  small  amounts 
of  hydrogen  sulphide  are  in  constant 
requisition,  as  in  the  laborator)-  for 
qualitative  analysis,  and  it  has  the 
advantage  of  being  capable  of  easy 
construction  from  the  glassware 
found  in  an}'  well  equipped  labora- 
tor}'. /4  is  a  capacious  tubulated 
bell-jar  inverted  and  resting  upon 


Pig.  a. 


FORMS  OP  GAS  GENERATORS.  IO59 

either  a  tripod  or  the  ring  of  an  ordinary  support.  The  perfo- 
rated stopper  in  the  neck  is  traversed  by  a  J  tube.  One  terminal 
of  this  tube  is  connected  with  a  simple  Bunsen  valve,  B,  t\  e,^ 
a  piece  of  rubber  tubing,  closed  at  one  end  and  provided  with  a 
clean  cut  slit  in  the  rubber  some  two  cm.  in  length.  The  other 
terminal  of  the  J  tube  is  connected  with  C  in  the  upper  portion 
of  A,  The  attachment  C  is  similar  to  that  frequently  introduced 
between  suction  pumps  and  filtering  flasks.  It  is  the  reverse  of 
B  in  its  construction,  allowing  a  current  of  liquid  to  enter  from 
the  outside  through  the  rubber  valve.  A  serves  as  a  reservoir 
for  the  acid  charge.  The  third  external  terminal  of  the  \  tube 
is  connected  with  the  tubulus  of  the  lower  portion  of  an  ordinary' 
lime  drj'ing  tower,  /?,  preferably  of  the  largest  size  constructed. 
Z>  serves  as  the  recipient  for  the  ferrous  sulphide,  etc.,  which 
may  be  used,  and  is  provided  with  a  perforated  disk  at  E  and 
the  outlet  tube  F,  the  latter  on  a  level  with  the  top  of  A .  The 
working  of  the  generator  is  exceedingly  simple.  A  is  charged 
with  acid  and  D  with,  say,  ferrous  sulphide.  When  /^is  opened 
the  acid  flows  through  C  into  D,  When  F  is  closed  the  pressure 
of  the  gas  evolved  forces  the  acid  back  into  A  through  B,  The 
result  is  that  the  supply  of  acid  furnished  D  is  alwa3's  from  the 
top  of  the  reservoir -^,  and  hence  stronger  than  that  found  in 
the  lower  strata,  which  are  successively  of  greater  specific 
gravity,  weaker  in  acid  and  richer  in  saline  matter,  as  the  bot- 
tom is  approached.  The  arrangement  permits  of  a  very  com- 
plete utilization  of  the  acid.  When  the  current  of  gas  is  in  con- 
tinuous demand,  and  evolution  becomes  sluggish,  it  is  necessary' 
to  close  the  tap  at  F  for  a  short  time  until  the  liquid  in  D  is 
driven  back  into  A . 

Care  must  be  exercised  in  constructing  the  valve  at  C  so  that 
it  will  yield  to  a  very  slight  pressure.  To  effect  this  the  slit  in 
the  rubber  should  be  at  least  two  cm.  in  length.  When 
the  apparatus  is  used  exclusively  for  the  evolution  of  hydrogen 
sulphide  to  be  employed"  in  qualitative  analysis,  it  is  desirable  to 
have  beyond  F  some  device  which  regulates  uniformly  the 
strength  of'the  current  of  gas  and  keeps  it  within  the  limits  of 
easy  absorption.  In  practice  this  has  been  accomplished  most 
simply  by  introducing  into  the  rubber  tube  attached  to /'a  short 


io6o 


SOME   NEW  FORMS  OP  GAS  GENERATORS. 


piece  of  glass  tubing,  one  end  of  which  is  drawn  out  so  as  to 
form  a  very  narrow  opening. 

Essentially  the  same  principle  for  the  control  of  the  strength 
of  the  acid  charge  is  to  be  found  in  the  generator  devised 
recently  by  Professor  Harris.  In  consequence  of  the  costly 
character  of  the  latter,  due  largely  to  the  use  of  valves  of  elabo- 
rate construction,  the  form  of  generator  just  described  may  be 
welcome  to  many  on  account  of  its  simplicity  and  inexpensive- 
ness. 

An  automatic  chlorine  generator  based  upon  the  use  of  manga- 
nese dioxide,  has  long  been  desired.  In  Fig.  3  is  shown  such 
^"'C^^  a  generator  which  for  six  years  has 

rendered  satisfactory  service,  both 
on  the  lecture  table  and  in  the 
laboratory.  The  essential  parts 
only  are  outlined  without  the 
accompanying  supports.  ^  is  a 
copper  funnel,  provided  with  a 
hollow  projection  C,  on  one  side, 
perfectly  similar  in  make  to  the 
funnels  used  for  hot  water  filtra- 
tion. It  can  be  advantageously  re- 
placed by  the  more  graceful  and 
modem  type  of  aluminum  funnel, 
resting  in  a  ring  burner.  The  res- 
ervoir B  is  of  glass,  and  is  an  article  of  current  manufacture, 
obtainable  from  all  dealers  in  chemical  glassware.  The  long, 
tapering  neck  is  tightly  fastened  in  the  neck  of  the  funnel  by 
means  of  a  section  of  rubber  tubing.  A  large  opening  at  the 
top,  closed  by  a  rubber  stopper,  serves  for  the  admission  of 
the  charge.  In  a  smaller  tubulure  on  the  side  is  a  perforated 
rubber  stopper  with  outlet  tube  and  tap.  The  funnel  with 
its  reservoir  is  held  firmly  in  a  support,  so  that  the  end  of  C  is 
about  two  cm.  above  the  top  of  an  ordinary  burner.  A  perfo- 
rated plate  is  introduced  into  j9  so  as  to  prevent  solid  matter  from 
falling  into  the  nalrow  neck.  The  latter  is  connected  at  Z?  with 
a  large  tubulated  bottle  E^  which  serves  as  a  reser\'oir  for  hydro- 
chloric acid,  and  is  attached  to  a  support  so  that  it  can  be  raised 


Fio.  3. 


MINERAI.  CONSTITUBNTS  OF  THE  WATERMELON.      IO61 

or  lowered  at  will.  When  in  use  B  is  filled  to  two-thirds  of  its 
capacity  with  manganese  dioxide,  large  lumps  alone  being  used, 
as  powdered  mineral  may  easily  cause  a  stoppage  of  the  connec- 
tions. E  is  filled  with  hydrochloric  acid  and  raised  to  a  level 
slightly  above  the  top  of  B,  Water  is  poured  into  the  funnel  A 
until  it  is  nearly  full,  and  a  lamp  is  placed  under  C  As  soon  as 
the  temperature  has  reached  about  80°,  a  very  small  flame  suf- 
fices to  maintain  the  activity  of  the  generator.  When  th&  exit 
from  ^  is  open,  the  acid  enters  and  the  evolution  of  chlorine 
continues  until  checked  by  closing  the  tap,  when  the  acid  is 
driven  back  into  E.  A  slight  agitation  of  the  latter  before 
opening  the  tap  serves  to  prevent  the  accumulation  of  a  stratum 
of  weak  acid  at  the  bottom.  It  is  advisable  to  lower  the  reser- 
voir E  when  a  current  is  not  required,  so  as  to  avoid  pressure 
and  any  possible  escape  through  minute  leaks.  In  practice  it 
is  also  found  desirable  to  connect  the  opening  of  ^  by  a 
flexible  tube  with  a  bottle  of  caustic  soda  solution,  the  tube  ter- 
minating at  the  surface  of  the  solution.  This  prevents  any 
escape  into  the  surrounding  air  of  chlorine,  with  which  the  con- 
tents of  E  are  soon  saturated.  When  thus  arranged  a  current 
of  the  gas  can  be  taken  at  will  from  the  generator,  the  sole  con- 
dition being  the  maintenance  of  a  small  flame  beneath  C  The 
manifold  advantages  of  such  a  device,  especially  for  the  lecture 
table,  will  be  appreciated  by  all  who  attempt  an  extended  series 
of  experiments  with  chlorine.  As  described  above  the  genera- 
tor can  be  readily  constructed  from  pieces  of  apparatus  ordinarily 
found  in  a  well  equipped  laboratory.  I  have  found  a  generator 
in  which  the  reservoir  B  contains  1500  cc,  a  very  convenient 
size  for  use  in  the  lecture  room. 


MINERAL  CONSTITUENTS  OF  THE  WATERHELON, 

By  Gbokob  F.  Paymb. 

Received  September  a8.  iap6. 

THE  watermelon  is  not  a  crop  that  is  widely  grown  even  in 
this  country  with  great  success.  It  is  this  very  reason 
which  makes  it  a  desirable  crop  to  handle  in  Georgia,  as  the 
watermelons  in  this  state  attain  finer  flavor,  crispness,  juiciness 
and  sweetness  than  anywhere  else  in  the  world. 


I062        MINER  At  CONSTITUENTS  OP  THE  WATERMELON. 

Upon  analysis  of  two  medium-sized  watermelons  cut  up  and 
mixed  together,  we  found  them  to  contain  just  one-third  per 
cent,  of  pure  ash,  calculated  as  free  from  carbonic  acid.  The 
exact  figures  were  0.3338,  which  in  our  calculations  we  will 
round  off  into  an  even  one-third,  which  it  practically  is. 

The  composition  of  watermelon  ash  is  as  follows  : 

Per  cent. 

Sulphur  trioxide 4.41 

Calcium  oxide 5.54 

M&gnesium  oxide 6.74 

Potassium  oxide 61. 18 

Sodium  oxide 4.31 

Silicon  dioxide 2.15 

Phosphorus  pentoxide 10.35 

Chlorine 4.94 

Iron  sesquioxide 0.48 

Total 100.00 

A  good  average  crop  of  watermelons  is  considered  to  be  about 
one-half  carload  to  the  acre,  though  much  larger  crops  than  this 
are  sometimes  made.  Large  watermelons  are  also  considered 
desirable,  hence  in  considering  what  is  carried  off  from  the  land 
by  the  removal  of  the  crop,  it  is  well  to  consider  how  much 
would  be  taken  off  by  a  large  crop,  as  it  is  the  large  crops  which 
we  desire  to  produce.  We  have  before  us  a  report  of  a  crop  of 
watermelons  upon  an  acre  of  land  which  is  an  unusually  large 
one,  but  which  was  weighed  up  in  the  presence  of  disinterested 
witnesses  and  sworn  to  by  them  as  being  honestly  grown  upon  an 
acre  and  correctly  weighed.  This  crop  weighed  39,766  pounds. 
One-third  percent,  of  such  a  crop  would  be  pure  ash,  and  conse- 
quently the  mineral  plant  food  taken  out  of  an  acre  of  land  b^ 
such  a  crop  would  be  as  follows : 

*  Pounds. 

Sulphur  trioxide 5.85 

Calcium  oxide 7.34 

Magnesium  oxide 8.93 

Potassium  oxide 81.09 

Sodium  oxide 5.71 

Silicon  dioxide 2.85 

Phosphorus  pentoxide 13-59 

Chlorine 6.55 

Iron  sesquioxide 0.64 

ToUl 132.55 


A    MODIFIED   FORM  OF  THB  BBULLIOSCOPB.  IO63 

In  the  crop  mentioned  above  to  replace  the  phosphoric  acid 
and  potash  carried  off  from  one  acre  by  the  melons  alone,  not 
taking  into  account  the  vines  and  roots,  would  require : 

Pounds. 

Acid  phosphate  (thirteen  per  cent.  PsO^) 100 

Muriate  of  potash  (fifty  per  cent.  K,0) 160 

A  fair  crop  of  melons  upon  good  land,  however,  is  usually 
considered  to  be  about  one-third  of  the  above  large  crop  or  about 
one-half  carload.  II  we  estimate  then  the  amounts  of  phos- 
phoric acid  and  potash  required  for  an  average  crop  of  fair  char- 
acter, such  a  crop  will  take  from  the  soil  materials  to  replace 
which  will  require  about : 

Pounds. 

Acid  phosphate 33  J 

Muriate  of  potash 53} 

This  will  give  about  four  and  one-half  pounds  of  available 
phosphoric  acid  to  an  acre,  and  about  twenty-seven  pounds  of 
pure  potash  to  an  acre.  The  usual  goods  on  the  market  guar- 
antee about  ten  per  cent,  of  available  phosphoric  acid  and  about 
one  per  cent,  of  potash.  The  use  of  300  pounds  of  such  goods 
upon  each  acre  of  watermelons,  furnishes  thirty  pounds  of  avail- 
able phosphoric  acid,  or  about  six  and  one-half  times  as  much 
as  is  needed  to  replace  what  is  carried  off  by  the  watermelons. 
It  also  furnishes  about  three  pounds  of  potash,  which  is  only 
one-ninth  of  what  is  carried  off  by  the  crop  removed.  This  being 
the  case  it  shows  with  what  advantage  and  economy  the  water- 
melon grower  can  replace  a  large  proportion  of  his  phosphoric 
acid  with  potash. 

[Contribution  prom  ths  Chemical  Laboratory  op  the  U.  S.  Depart- 
ment OF  Agriculture,  No.  22.] 

A  MODIFIED  FORM  OF  THE  EBULLIOSCOPE. 

BY  H.  W.  Wiley. 

Received  September  a6,  1896. 

THE  determination  of  the  alcohol  in  wines  and  beers,  from 
the  temperature  of  the  vapors  given  off  on  boiling  at 
atmospheric  pressures,  has  long  been  practiced.  The  instru- 
ment by  means  of  which  this  determination  is  made  is  known  as 
the  ebuUioscope  or  ebulliometer.     The  use  of  this  instrument 


I064  H.    W.   WILEY. 

was  proposed  many   years   ago  by  Tabari6,  and  it  has  been 
improved  by  Malligand,  Salleron  and  others. 

It  is  evident  that  if  so  simple  an  apparatus  could  be  made  to 
give  accurate  data,  it  would  come  into  general  use  for  ordinary 
purposes.  The  difficulties  which  have  attended  the  use  of  the 
ebullioscope,  however,  have  been  of  such  a  nature  as  to  render 
the  data  given  by  it  somewhat  unreliable.  Among  these  difiS- 
culties  may  be  mentioned  the  fact  that  a  wine  or  beer  contains 
a  considerable  quantity  of  dissolved  matters,  which  serve  to 
render  the  temperature  of  the  boiling  liquid  higher  than  the 
temperature  of  a  mixture  of  a  similar  percentage  of  alcohol  with 
water.  While  the  temperature  of  the  vapors  emitted  are,  theo- 
retically, not  influenced  in  a  marked  degree  by  the  initial  tem- 
perature at  which  they  are  formed,  nevertheless,  in  practice  it 
has  been  shown  that  the  tendency  of  the  higher  initial  boiling 
point  is  to  give  a  higher  reading  to  the  thermometer  whose  bulb 
is  sutrounded  by  the  emitted  vapors. 

Another  difficulty  attending  the  use  of  the  ebuUioscope  is 
found  in  the  fact  that  the  percentage  of  alcohol  in  the  vapors 
emitted  is  much  greater  than  in  the  residual  liquid.  As  a  result, 
it  is  difficult  to  establish  a  balance  between  the  condensed  vapors 
and  the  liquid  remaining  in  the  flask,  in  such  a  manner  as  to 
secure  a  continuous  evolution  of  a  vapor  containing  a  definite 
proportion  of  alcohol. 

In  the  third  place,  it  has  been  customary  to  return  the  con- 
densed vapors  through  the  apparatus  in  such  a  way  that  they 
come  in  contact  with  the  uncondensed  vapors  surrounding  the 
thermometer.  By  this  means  the  vapors  surrounding  the  bulb 
of  the  thermometer  are  subjected  to  changes  of  temperature 
which  render  it  difficult  to  get  a  mean  reading  of  the  height  of 
the  mercurial  column  in  the  instrument.  The  variations  which 
the  mercurial  column  may  undergo  amount,  in  some  instances, 
to  two  or  three-tenths  of  a  degree  and  as  each  tenth  of  a  degree 
represents  approximately  a  tenth  of  a  per  cent,  of  alcohol,  it  is 
not  difficult  to  see  that  these  variations  would  tend  to  lead  to 
erroneous  results. 

In  the  fourth  place,  barometric  changes,  which  are  constantly 
taking  place  in  the  atmosphere,  change  the  boiling  point  of  the 


presented,  an  effort  has  been  made 


A  HODIFIBD  FORM  OF  THS  BBULLIOSCOPE.  lo6s 

vapor  of  water  so  that  it  is  frequently  necessary  to  clieck  the 
instrument  with  pure  water,  in  order  to  have  an  initial  tempera- 
ture for  the  calculations. 

In  the  apparatus  which  is 
to  remedy  the  difficulties 
which  hive  been  mentioned 
above.  The  apparatus  con> 
siats  of  the  flask  F.  which 
is  closed  by  a  rubber  stop- 
per carrying  the  large  ther- 
mometer B  and  a  tube  lead- 
ing to  the  condenser  D. 
The  vapors  which  are  given 
oS  during  ebullition  are 
condensed  in  D  and  return 
to  the  flask  through  the 
tube,  as  indicated  in  the 
figure,  entering  the  flask 
below  the  surface  of  the 
liquid  therein. 

The  flask  is  heated  by  a 
gas    lamp    and   is    placed 
upon  a  circular  disk  of  as- 
bestos in  such  a  way  as  to 
entirely  cover  the  hole  in 
the  center  of  the  asbestos 
disk ,  which  is  a  little  small- 
er than  the  bottom  of  the 
flask.     The  whole  appara- 
tus is  protected  from  exter-  \ 
nal  influences  of  tempera-   <  s^ 
ture  by  the  glass  cylinder  E,  which  rests  upon  the  asbestos  disk 
below  and  is  covered  with  a  detachable,  stiff  rubber  cloth  disk 
above. 

The  thermometer  C  indicates  the  temperature  of  the  ambient 
air  between  FatiA  E.  The  reading  of  the  thermometer  ^  should 
always  be  made  at  a  given  temperature  of  the  ambient  air,  as 
indicated  by  C.     The  tube  leading  from  the  top  of  the  conden- 


I066  H.   W.    WILEY. 

ser  D  to  the  left,  is  made  long  and  is  left  open  at  its  lower 
extremity,  in  order  to  secure  atmospheric  pressure  in  /%  and  at 
the  same  time  prevent  the  diffusion  of  the  alcohol  vapors 
through  D, 

The  flame  of  the  lamp  is  so  regulated  as  to  bring  the  tem- 
perature of  the  thermometer  C  to  about  ^  in  ten  minutes 
for  substances  not  containing  over  five  per  cent,  of  alcohol. 
After  boiling  for  a  few  minutes,  the  temperature,  as  indicated  in 
the  thermometer  B,  is  constant,  and  the  readings  of  the  ther- 
mometer should  be  made  at  intervals  of  about  half  a  minute  for 
two  minutes.  Some  pieces  of  scrap  platinum  placed  in  the  flask 
will  prevent  bumping  and  secure  a  more  uniform  evolution  of 
vapor. 

Slight  variations,  due  to  the  changes  in  temperature  of  the 
vapor,  are  thus  reduced  to  a  minimum  effect  upon  the  final 
results. 

The  apparatus  is  easily  operated,  is  quickly  charged  and  dis- 
charged and  with  it  at  least  three  determinations  of  alcohol  can 
be  made  in  an  hour. 

The  thermometer  used  is  the  same  as  is  employed  for  the 
determination  of  freezing  and  boiling  points  in  the  ascertain- 
ment of  molecular  weights.  The  reading  of  the  thermometer  is 
arbitrary,  but  the  degrees  indicated  are  centigrade.  The  ther- 
mometer is  set  in  the  first  place  by  putting  the  bulb  in  water 
containing  sixteen  grams  of  common  salt  to  loo  cc.  When  the 
water  is  fully  boiling,  the  excess  of  mercury  is  removed  from 
the  column  in  the  receptacle  at  the  top  and  then,  on  placing  in 
ordinary  boiling  water,  the  column  of  mercury  will  be  found  a 
little  above  the  5°  mark.  This  will  allow  a  variation  in  all  of  5** 
in  the  temperature,  and  a  thermometer  thus  set  can  be  used  for 
the  estimation  of  percentages  of  *alcohol  from  one  to  five  and  a 
half,  by  volume.  When  the  liquor  contains  a  larger  percentage 
of  alcohol  than  this,  it  is  advisable  to  dilute  it  until  it  reaches 
the  standard  mentioned. 

In  order  to  avoid  frequent  checking  of  the  thermometer,  ren- 
dered necessary  by  changes  in  barometric  pressure,  I  use  a  sec- 
ond apparatus  made  exactly  as  the  one  described,  in  which 


A  MODIFIED  FORM  OF  THE   BBULLIOSCOPE.  IO67 

water  is  kept  constantly  boiling.  It  is  only  necessary  in  this 
case  to  read  the  two  thermometers  at  the  same  instant  in  order 
to  make  any  necessary  correction  required  by  changes  in  baro- 
metric pressure. 

It  is  not  my  purpose  here  to  submit  a  table  showing  the  per- 
centages of  alcohol  corresponding  to  any  given  depression  in  the 
temperature  of  the  boiling  vapor.  It  is  only  necessary  to  call 
attention  to  the  fact  that  for  the  percentages  named,  the  platted 
line  showing  the  variation  in  depression  from  o**  to  five  per 
cent,  by  volume  of  alcohol  is  practically  straight  and  that  for 
each  0.8°  change  in  the  boiling  point  of  the  vapor,  there  is  a 
change  of  about  one  per  cent  by  volume  of  alcohol.  This  rule 
can  be  safely  applied  for  practical  purposes  to  all  liquors  con- 
taining not  more  than  five  and  five- tenths  per  cent,  of  alcohol. 
For  instance,  if,  in  a  given  case,  the  temperature  of  the  vapor  of 
boiling  water,  as  marked  by  the  thermometer,  is  5.155'',  and  the 
temperature  of  the  vapor  of  a  sample  of  beer  is  ^2.345°,  the  depres- 
sion is  equivalent  to  2.810°,  and  the  percentage  of  alcohol  by  vol- 
ume is  therefore  2.81  divided  by  0.80=  3.51. 

The  thermometer  used  is  graduated  to  hundredths  of  a  degree 
and  is  read  by  means  of  a  cathetometer,  which  will  easily  give 
readings  to  five  thousandths  of  a  degree. 

The  reading  of  the  thermometer  is  facilitated  by  covering  the 
bulb  with  a  test-tube  containing  water.  The  high  specific  heat 
of  the  -water  distributes  evenly  any  little  variations  of  tempera- 
ture which  otherwise  would  cause  the  mercurial  column  in  ther- 
mometer B  to  oscillate.  The  water  jacket  also  serves  as  a  pro- 
tection against  the  projection  of  any  particles  of  the  boiling 
liquor  directly  against  the  bulb  of  the  thermometer. 

It  is  believed  that  this  apparatus  is  the  best  form  of  ebullio- 
scope  which  has  yet  been  offered  for  practical  use  to  analysts. 


VOLUHETRIC  DETERHINATION  OF  ACETONE.' 

By  Edward  R.  Squibb. 

Received  November  9.  it90. 

IN  the  Afoniteur  Scientijique  oi  1893,  41,  4  Serie,  Vol.  7,  i^'.p. 
272-274,  MM.  J.  Robineau  and  G.  Rollin  publish  a  paper 
entitled,  **  Dosage  Volumetrique  de  L'Acetone,"  and  the  fol- 
lowing is,  first,  a  free  translation  and  abridgement  of  this  paper; 
and  second,  the  detail  of  an  improvement  of  the  process  whereby 
it  is  rendered  easier,  more  simple,  quicker,  and  better  adapted 
to  technical  uses,  whilst  still  sufficiently  accurate  for  most  pur- 
poses. 

FIRST,  FREE  TRANSLATION. 

The  common  way  of  determining  the  proportion  of  acetone  in 
a  liquid  containing  it  is  to  convert  the  acetone  into  iodoform  by 
means  of  iodine  in  the  presence  of  soda  after  eliminating  from 
the  liquid  everything  that  would  interfere  with  the  proper  reac- 
tion. 

For  this  process  binormal  solutions  of  iodine  and  of  sodium 
hydroxide  are  used,  and  the  precipitated  iodoform  is  washed, 
dried,  and  weighed,  or  is  dissolved  in  ether,  and  the  whole  or 
a  fraction  of  the  ethereal  solution  is  dried  over  sulphuric  acid 
and  weighed. 

The  appreciable  volatility  of  iodoform  at  ordinary  tempera- 
tures introduces  a  source  of  error  that  is  objectionable,  especially 
when  dealing  with  small  quantities. 

But,  aside  from  this,  the  time  required  for  this  process  is  rela- 
tively so  long  that  we  have  sought  to  change  it  to  a  volumetric 
process  that  is  mor^  rapid. 

Our  proceeding  consists  in  mixing  the  acetone  with  a  solution 
of  potassium  iodide  and  sodium  hydroxide,  and  then  transform- 
ing it  into  iodoform  with  a  titrated  solution  of  a  hypochlorite. 
The  end  reaction  is  indicated  by  the  appearance  of  a  blue  color, 
when  a  drop  of  the  liquid  is  touched  with  a  drop  of  bicarbonated 
starch  solution. 

From  the  quantity  of  hypochlorite  used  the  quantity  of  ace- 
tone is  deduced. 

1  Read  before  the  New  York  Section  of  the  American  Chemical  Society,  November 
6th,  1896. 


VOLUMETRIC   DETERMINATION   OF   ACETONE.  IO69 

For  it  happens  that  the  presence  of  even  the  smallest  trace  of 
an  alkaline  hypoiodite,  in  a  solution  of  soda,  gives  a  blue  color 
with  a  starch  solution  which  contains  an  excess  of  sodium  bi- 
carbonate. 

Again,  a  liquid  containing  acetone,  an  iodide  and  caustic  soda 
in  excess,  and  into  which  a  solution  of  hypochlorite  is  passed, 
gives  no  reaction  with  bicarbonated  starch  until  the  whole  of 
the  acetone  is  converted  into  iodoform. 

This  proceeding,  however,  only  gives  constant  results  when 
certain  precautions  are  taken.  Unless  the  liquid  containing  the 
acetone  be  sufficiently  alkaline,  an  excess  of  hypochlorite  will 
be  required  to  decompose  aH  the  acetone. 

The  potassium  iodide  must  be  in  excess. 

The  dilution  must  be  fairly  uniform,  and  the  concentration  of 
the  hypochlorite  about  the  same  for  the  different  titrations. 

The  process  should  not  be  used  in  too  strong  a  light. 

It  is  very  important  that  the  liquid  should  be  constantly 
stirred  during  the  additions  of  the  hypochlorite. 

The  strength  of  the  hypochlorite  solution  is  ascertained  by 
trial  against  a  pure  acetone  made  by  the  bisulphite  process. 

PREPARATION  OF  THE  HYPOCHLORITE. 

For  the  titration  of  liquids  containing  considerable  proportions 
of  acetone,  the  hypochlorite  solution  is  prepared  as  follows  : 

To  500  cc.  of  the  concentrated  solution  of  sodium  hypochlorite 
of  commerce,  which  tests  from  forty-five  to  fifty-five  volumes  of 
chlorine,  an  equal  measure  of  water,  and  ten  cc.  of  solution  o* 
pure  soda  of  36°  B.,  are  added,  and  the  solution  is  kept  in  an 
amber  colored  bottle,  well  corked. 

TITRATION  OF  THE  HYPOCHLORITE  SOLUTION. 

About  two  grams  of  pure  acetone  from  bisulphite  is  weighed 
off  and  diluted  to  500  cc. 

Then  ten  grams  of  pure  potassium  iodide  is  put  into  a  conical 
Bohemian  beaker  and  100  cc.  of  the  diluted  acetone  and  twenty 
cc.  of  solution  of  caustic  soda  of  28°  B.  are  successively  added, 
and  the  whole  is  stirred  until  the  iodide  is  dissolved  and  the 
liquid  is  homogeneous. 


lOyO  EDWARD  R.    SQUIBB. 

Into  this  the  hypochlorite  solution  is  passed  drop  by  drop 
from  a  burette,  with  constant  stirring,  precipitating  the  iodoform 
in  large  flakes  which  easily  settle  out.  When  farther  additions 
of  the  hypochlorite  give  but  a  light  cloudiness  a  drop  of  the 
liquid  is  transferred  to  a  white  porcelain  plate  by  means  of  a 
glass  rod  and  is  there  brought  in  contact  with  a  drop  of  the 
bicarbonated  starch  solution.  As  soon  as  the  hypochlorite  is  in 
excess  the  blue  color  appears  very  distinctly.  The  volume  of 
hypochlorite  used  is  then  read  off  from  the  burette,  and  then 
for  security  of  result  the  titration  is  repeated. 

Example, — 2.081  grams  pure  acetone  is  weighed  off  and  diluted 
to  500  cc ;  100  cc.  of  this  solution  requires  22.5  cc.  of  the  hypo- 
chlorite. This  gives  for  each  cc.  of  hypochlorite  0.01874  gram 
of  pure  acetone.  These  results  are  liable  to  vary  a  little  if  the 
conditions  of  the  experiment  vary  much.  The  stirring  is  sup- 
posed to  be  constant,  and  the  hypochlorite  solution  to  be  regu- 
larly added. 

If  to  100  cc.  of  the  diluted  acetone  100  cc.  of  water  be  added 
and  the  same  quantities  of  iodide  and  soda  as  above,  22.05  <^- 
of  hypochlorite  is  required  instead  of  22.5  cc.  In  using  forty  cc. 
of  the  soda  solution  instead  of  twenty  cc,  twenty-two  cc.  of  the 
hypochlorite  is  required.  In  using  sixty  cc.  of  soda  solution 
instead  of  twenty  cc. ,  21.6  cc.  of  the  hypochlorite  is  required. 
In  using  ten  cc.  of  soda  instead  of  twenty  cc,  twenty-three 
cc  of  the  hypochlorite  is  required. 

These  results  show  that  dilution  of  the  acetone  and  a  small 
excess  of  soda  have  but  little  influence,  but  that  a  deficiency  in 
alkalinity  has  a  very  considerable  effect  on  the  quantity  of  hypo- 
chlorite required.  And  farther,  that  the  alkalinity  indicated  by 
twenty  cc  of  soda  solution  of  28**  B.  appears  to  be  normal. 

Under  the  given  conditions  of  alkalinity  and  dilution  the  rela- 
tions between  acetone  and  the  available  chlorine  of  the  hypo- 
chlorite is  obviously  one  molecule  of  acetone  to  six .  atoms  of 
chlorine. 

The  solution  of  hypochlorite  used  by  us  was  the  liquor  of 
Penot,  testing  21.56  volumes. 

The  titration  of  the  hypochlorite  with  pure  acetone  may  be 


,  39'9S  per  cent,  ace- 
tone. 


VOLUMETRIC  DETERMINATION  OP  ACETONE.  IO7X 

omitted,  simply  determining  the  available  chlorine  of  the  liquor 
of  Penot  instead,  but  we  prefer  the  titration  with  pure  acetone. 

DETERMINATION  OF  ACETONE  IN  A  COMPLEX  LIQUID. 

We  give  as  an  example  of  this  the  titration  of  a  complex 
liquid  made  with  precision,  which  liquid  has  served  us  to  con- 
trol the  accuracy  of  the  process. 

The  complex  liquid  contained  : 

I  •  5 10  grams  water, 

1 .677     *  *      ethyl  alcohol, 

1.550     **     methyl  alcohol,  pure 

3.149     **      acetone,  pure  from  bisulphite  ^ 

Of  this  mixture  3.2445  grams  was  weighed  o£f  and  diluted  to 
500  cc.  Proceeding  as  before  100  cc.  of  this  dilution,  ten  grams 
of  potassium  iodide,  and  twenty  cc.  of  soda  solution  of  28**  B. 
required  13.85  cc.  of  the  hypochlorite.  This  by  calculation 
gives  39.99  per  cent,  of  acetone,  and  thus  verifies  the  composi- 
tion of  the  complex  liquid ;  and  it  is  seen  that  the  presence  of 
ethyl  alcohol  is  without  influence  on  the  result. 

The  effect  of  the  presence  of  paraldehyde  in  the  same  complex 
liquid  was  tried  by  a  similar  titration. 

To  100  cc.  of  the  complex  solution  corresponding  to  0.4162 
gram  of  pure  acetone,  ten  cc.  of  an  aqueous  solution  containing 
five  per  cent,  of  pure  paraldehyde  was  added  (say  one-half 
gram)  or  a  little  more  paraldehyde  than  acetone.  This  mixture 
took  22.4  cc.  of  the  hypochlorite  instead  of  22.2  cc.  as  required. 

This  variation  is  slight  for  the  relatively  large  proportion  of 
paraldehyde,  and  is  greater  for  larger  proportions,  but  instances 
are  rare  in  which  paraldehyde  is  present  in  such  proportions. 

In  all  such  instances  where  the  presence  of  the  aldehyde  has 
been  established  by  the  process  of  Bardy,  the  acetone  should  be 
purified  by  this  process  before  titration. 

For  the  determination  of  acetone  in  very  dilute  solutions  a 
solution  of  hypochlorite  of  one-fifth  of  the  above  strength  is  pre- 
ferred. That  is,  a  solution  containing  four  or  five  volumes  of 
available  chlorine,  and  the  degree  of  alkalinity  should  be  pro- 
portionately reduced. 

With  a  little  practice  it  is  easy  to  judge  as  to  how  much  ace- 
tone is  present  in  a  liquid  to  be  titrated,  and  from  this  to  judge 


I072  EDWARD   R.   SQUIBB. 

of  the  corresponding  quantity  of  hypochlorite  required,  and  in 
this  way  keep  the  conditions  of  the  method  nearly  uniform,  and 
the  more  uniform  the  conditions  the  more  constant  the  results. 

This  process  has  the  great  advantage  of  being  rapid,  and  thus 
of  permitting  a  number  of  titrations  being  made  in  a  short  time 
with  results  sufficiently  accurate, 

REMARKS. 

The  reaction  used  in  this  titration  is  very  delicate,  and  where 
traces  of  acetone  are  concerned  it  is  better  seen  when  there  is 
excess  of  iodide  and  of  soda  and  but  little  hypochlorite.  An 
aqueous  solution  of  0.004  gram  of  acetone  in  the  liter  gives 
a  heavy  cloudiness  immediately.  The  reaction  with  0,0012  gram 
of  acetone  in  the  liter  is  seen  in  a  few  moments.  With  0.0008 
gram  to  the  liter  the  reaction  is  difficult  to  see.  This  reaction 
should  not  be  made  in  a  bright  light.  In  sunshine  or  in  a  very 
bright  light  the  traces  of  iodoform  produced  disappear  vexy 
rapidly,  the  liquid  becoming  clear,  but  in  a  dim  light  the  pre- 
cipitate does  not  disappear. 

The  titrated  solution  of  hypochlorite  should  be  kept  in  amber- 
colored  glass  in  a  cool  place  and  sheltered  from  bright  light. 
The  titration  should  be  frequently  repeated,  because  it  varies 
rather  rapidly,  especially  when  diluted.  We  have  made  a  series 
of  experiments  on  this  point,  which  strikingly  show  these  varia- 
tions under  different  influences.  A  solution  of  hypochlorite 
prepared  for  titrations  gave  22.16  volumes  of  available  chlorine; 
kept  in  a  cool  place,  in  obscurity  for  six  days,  it  gave  21.96;  kept 
in  colorless  glass,  corked,  in  a  bright  light,  most  of  the  time  in 
sunlight,  for  seven  days,  it  gave  12.32.  In  a  water-bath  at 
100^  C.  for  a  quarter  of  an  hour  it  gave,  when  cooled,  19.48. 

END   OF    TRANSLATION. 

• 

The  rapidly  increasing  uses  of  acetone  in  the  three  years  that 
have  passed  since  the  publication  of  this  important  paper  of 
Robineau  and  Rollin  have  given  to  it  so  much  additional  impor- 
tance that  it  seemed  well  to  the  present  writer — who  ejirly 
adopted  this  volumetric  method — to  attempt  to  modify  the 
method  in  the  direction  of  greater  simplicity  and  rapidity,  even 


VOLUMETRIC   DETERMINASTION   OF   ACETONE.  IO73 

if  this  should  be  at  the  cost  of  a  little  of  its  accuracy.  As  ace- 
tone comes  more  and  more  to  take  the  place  of  both  ethyl  and 
methyl  alcohol  as  much  the  better  solvent  for  most  purposes, 
and  as  its  manufacture  is  cheapened,  it  becomes  more  and  more 
desirable  to  have  a  rapid  and  easy  way  of  estimating  its  propor- 
tions in  mixtures  or  under  conditions  to  which  specific  gravity 
is  not  applicable. 

Therefore,  taking  the  above  quoted  paper  as  a  basis,  and  giv- 
ing full  credit  to  the  authors  of  it  for  every  important  principle 
and  step  of  the  method ,  the  following  slight  modifications  are  offered 
as  the  result  of  about  three  months'  experience  with  the  original 
process  and  over  a  year's  experience  with  the  modifications. 

STANDARD   SOLUTION  OP    ACETONE. 

A  flask  of  TOO  cc.  capacity  containing  about  fifty  cc.  of  dis- 
tilled water  is  carefully  weighed.  To  this  is  added  about  thir- 
teen cc.  of  pure  acetone,  made  by  the  bisulphite  process.  The 
weight  is  then  again  taken,  when  it  will  be  found  that  the  ace- 
tone added  is  a  fraction  more  or  less  than  ten  grams.  The  dilu- 
tion is  then  transferred  to  a  measuring  flask,  the  weighing  flask 
being  rinsed  in  and  is  farther  diluted  with  distilled  water  until 
each  ten  cc.  of  the  dilution  contains  one-tenth  gram  of  acetone. 
This  is  kept  in  a  well  stoppered  bottle  of  dark  glass,  for, 
although  the  writer  has  no  evidence  of  any  change  taking  place 
in  acetone,  and  believes  it  to  be  quite  as  permanent  as  ethyl 
alcohol,  still  it  may  be  well  to  keep  a  dilute  standard  solution 
protected  against  bright  light. 

Of  this  solution  or  dilution  ten  cc.  equal  to  one-tenth  gram  of 
acetone,  is  accurately  measured  off  for  each  titration  of  the  solu- 
tion of  hypochlorite. 

SOLUTION  OP  POTASSIUM  IODIDE. 

Of  this  salt  250  grams  are  dissolved  in  distilled  water,  and 
the  solution  is  made  up  to  one  liter,  when  each  ten  cc.  will  con- 
tain two  and  a  half  grams  of  the  iodide. 

SOLUTION   OP  SODIUM   HYDROXIDE. 

Of  commercial  caustic  soda,  purified  by  alcohol,  257  grams  is 
dissolved  in  distilled  water,  the  solution  made  up  to  one  liter, 


I074  BDWARD  R.   SQUIBB. 

and  set  aside  until  it  settles  quite  clear.  Then  850  cc.  of  clear 
solution  is  poured  off  and  added  to  the  solution  of  potassium 
iodide,  making  1,850  cc.  of  total  solution. 

Of  this  solution  twenty  cc.  is  taken  for  each  titration. 

The  remainder  of  the  soda  solution  is  again  allowed  to  settle 
clear  for  farther  use  in  the  hypochlorite  solution. 

SOLUTION  OF  SODIUM  HYPOCHLORITE. 

The  officinal  solution  of  chlorinated  soda  of  the  U.  S.  Pharma- 
copoeia (**Liquor  Sodae  Chloratae,"  U.  S.  P.)  answers  very  well 
for  this  process,  the  officinal  strength  of  two  and  six-tenths  per 
cent,  of  available  chlorine  being  quite  convenient. 

To  a  liter  of  this  solution  in  a  bottle  of  dark  glass,  twenty-five 
cc.  of  the  above  described  clear  soda  solution  is  added  and  the 
mixture  well  shaken. 

If  in  buying  the  **  Solution  of  Chlorinated  Soda"  of  the  U.  S. 
P.  for  this  process  it  should  be  found,  as  is  not  unfrequently  the 
case,  weaker  than  is  required  by  the  U.  S.  P.,  or,  if  by  keep- 
ing it  becomes  weaker,  this  will  be  at  once  discovered  on 
balancing  it  against  the  standard  acetone  solution,  and  so  long 
as  the  one-tenth  gram  of  acetone  does  not  require  more  than  say 
twenty  cc.  of  the  more  dilute  hypochlorite,  the  formula  need 
not  be  modified. 

If  there  be  much  of  this  titration  to  do  it  is  very  convenientto 
fit  this  bottle  with  an  automatic  zero  burette,'  as  shown  in  the 
following  illustration,  this  form  being,  so  far  as  is  known,  origi- 
nal with  the  writer  and  very  convenient  for  general  rapid  work- 
ing with  a  burette.  The  advantage  is,  beside  that  of  rapid  and 
easy  working,  that  it  does  not  require  a  special  burette  and  is 
easily  fitted  up  from  the  resources  of  any  laboratory. 

BICARBONATED   STARCH   SOLUTION. 

Starch,  0.125  gram,  is  mixed  with  five  cc.  of  cold  water,  and 
then  added  to  twenty  cc.  of  boiling  water  and  boiled.  When 
cold  two  grams  of  sodium  acid  carbonate  is  added  and  stirred 
until  dissolved.  Kept  in  a  colorless  bottle  this  solution  does  not 
sensibly  diminish  in  delicacy  or  reaction  in  three  months.  But 
for  how  much  longer  it  would  remain  good  for  this  reaction  was 
not  tried. 

1  This  Journal,  i6, 145. 


1076  EDWARD   R.    SQUIBB. 

THE   TITRATION. 

The  burette  being  filled  with  the  solution  of  sodium  hypo- 
chlorite, ten  cc.  of  the  standard  solution  of  acetone  (equal  to  one- 
tenth  gram  of  acetone)  is  measured  into  a  beaker  of  about  fifty 
cc.  capacity,  and  twenty  cc.  of  the  mixed  solution  of  iodide  and 
soda  is  added  and  stirred  well.  Into  this  the  hypochlorite  solu- 
tion is  passed  in  rapid  dropping,  with  constant  stirring,  until 
eight  or  ten  cc.  has  been  run  in.  Then  the  precipitated  iodo- 
form is  allowed  to  settle  out,  and  a  drop  or  two  more  hypo- 
chlorite is  added.  Should  this  produce  a  dense  cloudiness  one- 
half  cc.  more  hypochlorite  is  added,  and  well  stirred  and  again 
allowed  to  settle.  Then  a  drop  or  two  more  of  hypochlorite  is 
added.  If  there  should  still  be  a  cloudiness,  another  one-half 
cc.  of  the  hypochlorite  is  added  and  well  stirred,  and  so  on  until 
the  cloudiness  is  very  slight.     Then  the  starch  testing  begins. 

A  small  drop  of  the  liquid  is  transferred  by  a  rod  to  a  white 
porcelain  tile  or  plate,  and  a  similar  small  drop  of  the  starch 
solution  is  placed  very  near  it.  Then  with  the  first  rod  the 
drops  are  made  to  connect  by  a  fine  line,  so  that  the  whole  has 
a  dumb-bell  form.  If  there  be  no  blue  color,  one  or  two-tenths 
cc.  more  of  the  hypochlorite  is  added  and  well  stirred,  and  the 
testing  is  repeated,  until  finally  a  blue  line  will  be  seen  at  the 
moment  of  contact  of  one  drop  with  the  other.  If  the  last  nega- 
tive testing  has  taken  10.4  cc.  from  the  burette,  and  this  posi- 
tive testing,  which  has  given  the  blue  line,  required  10.6  cc., 
then  the  accepted  reading  would  be  10.5  cc,  and  this  would  be 
the  hypochlorite  equivalent  of  one-tenth  gram  ^f  acetone.  If 
the  blue  line  be  very  faint,  it  will  be  momentary  only,  and  will 
indicate  that  the  excess  of  hypochlorite  is  very  small,  and  that 
10.6  cc.  is  a  closer  reading  than  10.5,  but  the  process  is  not 
sufficiently  accurate  to  take  much  account  of  such  differences, 
since  even  with  much  experience  and  great  care  it  is  hardly 
practicable  to  get  any  two  titrations  to  agree  within  one-tenth 
cc.  of  hypochlorite. 

Having  then  10.5  cc.  as  the  hypochlorite  equivalent  of  one- 
tenth  gram  of  acetone  at  this  time,  it  is  easy  to  estimate  any 
smaller  or  larger  quantity  of  acetone  that  requires  a  smaller  or 


VOI,UM£TRIC   DETERMINATION  OI^  ACETONE.  IO77 

larger  quantity  of  the  hypochlorite  by  the  equation  10.5  :  o.i  :  : 
a  :  X. 

But  this  hypochlorite  solution  is  liable  to  diminish  in  strength 
by  keeping,  and  therefore  must  be  standardized  by  this  standard 
acetone  solution  as  often  as  the  accuracy  .of  the  determinations 
may  require.  At  times  the  change  in  strength  is  scarcely  per- 
ceptible from  day  to  day  in  several  successive  day's  work,  but 
in  standing  for  a  week  or  two  there  will  always  be  a  falling  off 
in  strength  to  the  extent  of  one-tenth  to  five-tenths  cc.  in  the 
h3rpochlorite.  The  addition  of  the  soda  solution  appears  to  ren- 
der the  h3rpochlorite  more  permanent,  just  as  the  sodium  bicar- 
bonate renders  the  starch  solution  more  permanent.  But  in  the 
case  of  the  starch  the  blue  reaction  does  not  occur  if  the  bicar- 
bonate be  not  present. 

The  titration  of  the  acetone  present  in  unknown  dilutions  re- 
quires first  that  the  strength  should  be  estimated  by  known  con- 
ditions or  by  sensible  properties,  in  order  to  keep  the  proportions 
of  the  reagents  and  the  dilutions  approximately  the  same,  or  at 
least  not  differing  very  widely  when  close  determinations  are  re- 
quired. If  then  the  taste  and  smell  should  indicate  that  the  ace- 
tone to  be  tested  is  below  twenty-five  per  cent.,  four- tenths  cc. 
may  be  taken  for  the  testing.  If  over  twenty-five  per  cent,  and 
under  fifty  per  cent.,  two-tenths  cc.  may  be  taken.  If  over  fifty 
per  cent.,  one-tenth  cc.  is  sufficient. 

For  the  adjustment  of  these  small  quantities  with  a  sufficient 
degree  of  accuracy  for  rapid  technical  working,  it  is  convenient 
to  have  a  five-tenths  cc.  pipette  divided  in  o.oi  cc.  fitted  with  a 
rubber  bulb,  as  shown  in  the  illustration.  By  screwing  the  neck 
of  this  bulb  up  or  down  upon  the  glass,  with  the  point  in  the 
liquid,  close  measurements  may  be  quickly  made. 

A  beaker  of  fifty  cc.  capacity  containing  ten  cc.  of  water  is 
weighed  and  the  weight  noted.  The  four- tenths,  two-tenths,  or 
one-tenth  cc.  of  the  sample  to  be  titrated  is  delivered  in  the  water 
and  the  weight  again  taken  to  give  the  qyantity  of  the  sample 
taken  for  the  titration.  Then  the  twenty  cc.  of  the  iodide  and 
soda  solution  is  added,  the  whole  well  stirred,  and  the  hypo- 
chlorite dropped  in,  and  the  end  reaction  managed  precisely  as 
described  in  standardizing  the  h3rpochlorite,  and  the  quantity  of 


1078  VOI^UMBTRIC   DETERMINATION  OP  ACETONE. 

h3rpochIorite  used  is  noted.  Then  as  10.5  cc.  of  the  hypochlorite 
is  to  one-tenth  gram  of  acetone,  so  is  the  quantity  of  hypochlorite 
now  used  to  the  quantity  of  acetone  present  in  the  portion  of  the 
sample  taken  for  titration.  Then  as  the  weight  of  this  portion 
taken  for  titration  is  to  the  quantity  of  acetone  found  in  it,  so  is 
100  to  the  percentage  of  acetone  in  the  sample. 

For  example,  a  sample  supposed  to  be  not  far  from  absolute  is 
to  be  titrated.  A  fifty  cc.  beaker  with  ten  cc.  of  water  weighs 
25-283  grams;  with  one-tenth  cc.  of  the  sample  added  the 
weight  becomes  25.360  grams,  giving  0.077  gram  as  the  weight 
taken  for  the  titration.  To  this  is  added  the  twenty  cc.  of  iodide 
and  soda  solution,  and  the  mixture  being  well  stirred,  the  hypo- 
chlorite is  dropped  into  saturation  when  seven  and  nine-tenths 
cc.  is  found  to  have  been  used.  Then  as  10.5  is  to  one-tenth,  so 
is  seven  and  nine-tenths  to  0.0752  gram  of  acetone  in  the  0.077 
gram  of  the  sample  taken.  Then  as  0.077  gram  of  the  sample 
taken  is  to  the  0.0752  of  acetone  indicated,  so  is  100  to  97.66  per 
cent,  of  acetone  in  the  sample. 

This  is  the  rationale  of  the  operation,  but  the  calculation  is 
shortened  by  simply  dividing  the  standard  hypochlorite  (10.5  cc. ) 
into  the  hypochlorite  required  (seven  and  nine-tenths  cc.)toget 
the  corresponding  acetone  (0.0752  gram),  and  then  dividing  the 
weight  of  the  sample  taken  (0.077  gram)  into  the  weight  of  ace- 
tone obtained  from  it  (0.0752  gram)  to  get  the  percentage  pro- 
portion of  the  acetone.  (97.66  per  cent.). 

Of  course  the  method  of  definite  dilution,  and  the  titration  of 
an  aliquot  part,  as  described  in  the  original  paper  of  Robineau 
and  RoUin  (see  translation)  is  available  and  more  accurate  than 
that  here  recommended,  and  takes  but  little  more  time. 

Where  acetone  is  made,  or  is  much  used,  and  especially  in 
processes  where  it  is  recovered  by  distillation  to  be  used  over 
again,  there  is  often  much  need  of  testing  the  strength  of  very 
weak  dilutions,  and  of  knowing  when  acetone  is  absent.  In 
many  such  uses  accuracy  is  not  required  and  rough  estimates 
are  sufficient.  For  work  of  this  kind,  •  especially  when  the 
strength  is  below  ten  per  cent.,  the  weighing  of  the  sample  to  be 
tested  may  be  omitted,  because  the  specific  gravity  is  so  nearly 


DFTBRMINATION  OF  SULPHUR  IN  CAST  IRON.  IO79 

that  of  water  that  the  measura  may  be  accepted  as  cubic  centi- 
meter for  gram. 

DETERMINATION  OP  ACETONE  IN  THE  PRESENCE  OP  ETHYL 

ALCOHOL. 

The  standard  dilution  of  acetone  containing  ten  grams  in  the 
liter  was  used,  and  ten  cc.  of  this  required  14.3  cc.  of  the  h3rpo- 
chlorite  solution.     On  repetition  14.4  cc.  was.  required. 

A  dilution  of  ethyl  alcohol  was  made  containing  ten  grams  in 
the  liter,  and  ten  cc.  of  this  requires  one-tenth  cc.  of  the  hypo- 
chlorite.    On  repetition  0.125  cc.  was  required. 

To  ten  cc.  of  the  acetone  dilution  two-tenths  cc.  of  the  alcohol 
dilution  was  added,  and  this  mixture  required  14.4  cc.  of  the 
hypochlorite  solution.  On  repetition  14.4  cc.  again  was  required. 

To  ten  cc.  of  the  alcohol  dilution  two-tenths  cc.  of  the  acetone 
dilution  was  added,  and  this  mixture  required  0.35  cc.  of  the 
hjrpochlorite.     On  repetition  four-tenths  cc.  was  required. 

In  each  case  ten  cc.  of  the  iodine  and  soda  solution  was  used 
and  all  other  conditions  were  kept  fairly  uniform. 

In  the  case  wherein  the  hypochlorite  was  added  to  alcohol 
alone  no  precipitate  nor  cloudiness  was  visible,  although  o.i 
to  0.125  cc.  was  required  to  obtain  the  starch  reaction.  When 
acetone  had  been  added  to  the  alcohol  one-half  this  quantity  of 
the  hypochlorite  was  suflScient  to  give  decided  cloudiness. 

These  results  appear  to  confirm  the  conclusions  of  Robineau 
and  RoUin  to  the  effect  that  the  presence  of  ethyl  alcohol  has  no 
effect  upon  the  titration  of  acetone  by  this  method,  although 
ethyl  alcohol  is  an  iodoform-yielding  substance.  The  small 
quantity  of  hypochlorite  required  to  obtain  the  starch  reaction 
when  alcohol  alone  was  titrated  was  probably  in  consequence  of 
traces  of  impurity  in  the  alcohol. 


THE  DETERMINATION  OF  SULPHUR  IN  CAST  IRON. 

By  Francis  C.  Phillips. 
Received  November  10,  18B6. 

IN  a  paper  read  before  the  American  Chemical  Society  in 
August,  1895,'  I  have  detailed  some  experiments  made  in 
the  determination  of  sulphur  in  white  cast  iron  by  the  evolution 
method,  and  have  attempted  to  show  that  the  loss  of  sulphur  in  its 

iThU  Journal,  17,  891. 


I080  FRANCIS  C.    PHILLIPS. 

determination  in  such  iron  may  be  due  to  the  formation  of 
organic  sulphur  compounds  not  oxidizable  to  sulphuric  acid  by 
the  usual  means. 

By  passing  the  gases  evolved  during  the  solution  of  the  iron 
in  hydrochloric  acid  through  a  heated  porcelain  tube  it  was  found 
that  the  volatile  organic  sulphur  compounds  may  be  decom- 
posed and  nearly  all  the  sulphur  recovered  by  conversion  into 
hydrogen  sulphide,  oxidation  and  precipitation  as  barium  sul- 
phate. 

In  judging  of  the  correctness  of  an  analytical  method  it  has 
been  necessary  in  the  case  of  the  majority  of  the  constituents  of 
iron  to  depend  upon  a  single  criterion ;  that  method  is  regarded 
as  most  accurate  which,  being  correct  in  its  details,  yields  the 
highest  percentage  of  the  constituent  sought  to  be  determined. 
For  it  is  hardly  possible  to  add  to  pure  iron  a  known  percentage 
of  sulphur,  phosphorus  or  carbon,  and  test  the  method  by  a 
determination  of  the  added  constituent.  For  the  determination 
rf  sulphur  in  iron  it  has  been  common  to  regard  the  method  of 
oxidation  and  solution  of  the  iron  by  nitric  acid,  followed  by 
precipitation  of  the  sulphur  in  form  of  barium  sulphate  as  the 
most  accurate,  inasmuch  that  it  yields  results  somewhat  higher 
than  those  obtained  by  other  modes  of  procedure. 

It  does  not  seem  probable  that  an  appreciable  error  could 
occur  in  the  use  of  this  method  unless,  in  the  simultaneous  oxi- 
dation of  the  carbon  and  sulphur  of  the  iron,  an  organic  sulphur 
compound  should  be  formed. 

It  has  seemed  to  be  of  interest,  however,  to  apply  a  method 
for  the  determination  of  sulphur  by  which  all  the  constituents  of 
the  metal  could  be  completely  oxidized  in  a  dry  state  and  at  a 
high  temperature,  in  order  to  avoid  as  eflfectually  as  possible 
the  chances  of  loss  due  to  the  conversion  of  sulphur  into  a  vola- 
tile compound  not  oxidizable  by  ordinary  means  to  sulphuric 
acid. 

In  searching  for  a  method  which  should  answer  these  require- 
ments, it  seemed  possible  that  by  heating  the  iron  in  the  form  of 
fine  powder  in  presence  of  a  mixture  of  alkaline  carbonate  and 
nitrate  the  sulphur  might  be  oxidized  directly  and  completely 
to  the  condition  of  a  sulphate  without  affording  an  opportunity 


DETERMINATION  OF  SULPHUR  IN   CAST  IRON.  IO81 

for  the  escape  of  a  trace  of  sulphur  in  some  intermediate  volatile 
or  soluble  compound.  Accordingly  an  experiment  was  tried  in 
the  following  way : 

An  iron  containing  its  carbon  in  the  combined  form  was 
melted  in  a  crucible  and  poured  while  fused  into  water.  The 
granulated  metal  was  crushed  in  a  steel  mortar  to  an  extremely 
fine  powder.  The  powder  so  obtained  was  sifted  through  bolt- 
ing sheeting. 

Two  and  one-half  g^ams  of  the  sifted  iron  were  mixed  with 
ten  grams  of  a  mixture  of  equal  parts  of  sodium  nitrate  and  car- 
bonate in  a  platinum  crucible.  The  crucible  was  covered  and 
heated  over  a  Bunsen  burner.  At  a  red  heat  a  sudden  and 
rather  violent  reaction  occurred,  and  having  been  begun,  was 
easily  maintained  with  very  little  aid  from  the  burner  flame. 
The  reaction  appeared  to  be  complete  in  a  few  minutes.  After 
heating  for  a  half  hour  the  crucible  was  cooled  and  its  contents 
softened  in  water.  A  residue  of  a  reddish  brown  powder,  con- 
sisting of  ferric  oxide  with  a  little  ferrous  oxide,  w*as  obtained. 
This  residue  was  found  to  contain  no  sulphuric  acid,  and  on 
digesting  with  hydrochloric  acid  dissolved  without  effervescence, 
showing  that  none  of  the  particles  of  the  original  iron  had 
remained  unoxidized.  Prom  the  results  of  this  experiment  and 
others  which  need  not  be  detailed  here,  it  seemed  to  be  possible 
to  oxidize  finely  divided  iron  so  completely  by  heating  with 
sodium  carbonate  and  nitrate,  that  its  sulphur  might  be  con- 
verted quantitatively  into  sulphuric  acid. 

The  mixture  of  sodium  carbonate  and  nitrate  although  tend- 
ing to  oxidize  finely  divided  iron,  seems  to  exert  a  less  powerful 
acUon  upon  the  carbon  contained  in  the  iron,  and  this  carbon 
may  appear  as  a  black  residue  after  the  fused  mass  has  been 
softened  and  extracted  by  water  and  the  ferric  oxide  dissolved 
in  hydrochloric  acid.  \ 

It  seems  to  be  important  for  the  success  of  the  method  that  in 
the  oxidation  of  the  iron  the  carbon  should  also  be  nearly  or 
completely  oxidized,  for  if  the  carbon  remained  unburned  a  por- 
tion of  the  sulphur  might  escape  oxidation.  In  general  it  may 
be  said  that  the  order  of  oxidation  of  these  three  elements  by  the 
method  used  is  as  follows:  i,  iron;  2,  carbon;  3,  sulphur  ;  the 


I082  FRANCIS  C.   PHILLIPS. 

iron  being  the  most  easily  oxidized,  and  the  sulphur  the  most 
difficult  to  oxidize.  This  order  is  not  exactly  what  w^  should 
anticipate,  but  it  is  to  be  remembered  that  unless  the  iron  grains 
are  fine  enough  to  be  penetrated  by  oxygen,  and  changed  com- 
pletely into  a  soft  powder  of  ferric  oxide,  the  sulphur  and  car- 
bon have  no  opportunity  to  oxidize  at  all.  If  the  iron  could  be 
used  as  an  impalpable  powder  the  order  of  oxidation  would  prob- 
ably be  different.  The  marked  resistance  of  the  carbon  to  oxi- 
dation has  been  frequently  observed,  even  when  using  more 
sodium  nitrate  in  the  fusion  than  is  theoretically  enough  to  com- 
pletely oxidize  both  iron  and  carbon,  supposing  that  the  sodium 
nitrate  is  reduced  only  to  nitrite  in  the  process. 

Experiments  of  a  similar  kind  were  tried  with  ferromanganese. 
A  metal  containing  about  eighty  per  cent,  of  manganese  was  used. 
By  crushing  in  a  steel  mortar  this  iron  was  very  easily  reduced 
to  a  powder  fine  enough  to  pass  through  bolting  sheeting.  On 
heating  the  powder  with  the  mixture  of  sodium  nitrate  and  car- 
bonate a  most  violent  reaction  occurred,  the  metal  burning  with 
a  long  flame,  extending  several  inches  above  the  crucible.  In 
order  to  control  the  reaction  it  was  found  necessary  to  melt  one- 
half  of  the  fusion  mixture  to  be  used  in  the  crucible  and  then 
add  slowly  the  other  half,  previously  mixed  with  the  powdered 
metal,  while  stirring  constantly.  In  this  way  the  reaction  could 
be  easily  controlled.  On  softening  the  fused  mass  in  water  it 
was  found  that  the  iron  had  been  peroxidized  and  the  manga- 
nese changed  to  binoxide.  No  trace  of  sodium  manganate  was 
ever  formed,  the  solution  in  water  being  after  filtration  invaria- 
bly colorless.  No  carbon  was  found  in  the  residue.  The  oxi- 
dation of  the  carbon  is  much  more  easily  effected  in  the  case  of 
iron  containing  a  high  percentage  of  manganese.  In  all  the 
trials  made  the  silicon  of  the  iron  was  oxidized,  but  it  was  found 
that  when  the  fused  mass  is  softened  in  water  very  little  silica 
enters  into  solution  as  an  alkaline  silicate,  the  greater  portion 
remaining  insoluble  and  in  a  flocculent  form. 

Experiments  were  then  tried  with  a  gray  iron.  This  form  of 
iron  could  not  be  crushed  to  a  fine  powder,  and  an  experiment 
was  made  in  reducing  it  from  small  drillings  by  means  of  a 


DBTBRMINATION  OP  SULPHUR  IN   CAST  IRON.  IO83 

chilled  iron  rubber  and  plate,  such  as  is  ordinarily  used  for 
grinding  ores.  Several  gray  irons  were  tried  in  this  way. 
Some  could  not  be  powdered  by  the  method  just  mentioned,  the 
grains  tending  to  flatten  instead  of  being  crushed.  Others  were 
readily  reduced,  but  the  powder  was  not  in  any  case  fine  enough 
for  sifting  through  bolting  sheeting.  It  was  found  in  the  case 
of  a  gray  iron  reduced  to  powder  by  the  method  of  grinding, 
that  on  fusion  with  the  mixture  of  sodium  nitrate  and  carbonate, 
used  in  the  preceding  experiments,  the  graphitic  carbon  of  this 
iron  was  more  readily  burnt  than  the  combined  carbon  of  white 
iron. 

As  it  had  proved  to  be  a  somewhat  difficult  matter  to  oxidize 
completely  the  carbon  of  the  iron  in  the  various  experiments 
made  with  the  fusion  method,  notably  in  the  case  of  white  iron, 
some  trials  were  made  in  the  use  of  sodium  peroxide.  This 
proved  to  be  a  more  efficient  oxidizing  agent  for  iron  and  its 
contained  carbon  than  sodium  nitrate.  For  these  trials  a  mix- 
ture was  used  consisting  of  forty-five  parts  each  of  sodium  per- 
oxide and  sodium  nitrate,  together  with  ten  parts  of  sodium  car- 
bonate. 

White  iron  was  oxidized  and  its  carbon  burnt  during  a  fusion 
la.sting  less  than  ten  minutes. 

On  heating  ferromanganese  with  this  mixture  the  iron  was 
found  to  be  completely  oxidized.  The  carbon  was  burnt  and 
the  manganese  was  oxidized  and  converted  into  sodium  manga- 
nate,  yielding  a  deep  green  solution  when  the  fused  mass  was 
digested  in  water. 

An  admixture  of  sodium  carbonate  to  sodium  peroxide  tends 
in  all  cases  to  diminish  its  action  upon  finely  divided  iron  at  a 
high  temperature  and  renders  the  process  more  easily  controlled. 
It  seemed  to  be  possible  to  base  a  method  for  the  quantitative 
determination  of  sulphur  in  certain  kinds  of  cast  iron  upon  the 
reactions  described  above. 

An  indispensable  condition  of  success  in  the  use  of  the  method 
is  found  in  the  extreme  fineness  of  the  iron.  In  the  case  of 
white  irons  the  fineness  of  the  powder  has  been  secured  by 


I084  FRANCIS  C.    PHILLIPS. 

crushing  in  a  steel  mortar  until  tfae  powder  passed  through  a 
sieve  of  bolting  sheeting  or  bolting  cloth. ^ 

Some  gray  irons  cannot  be  crushed  or  ground.  To  these  the 
method  is  not  applicable.  For  gray  irons,  however,  the  evolu- 
tion method  answers  all  requirements. 

The  following  details  are  given  of  the  method  finally  employed  : 

1 .  White  iron .  — About  one  and  one*half  grams  of  the  finely  pow- 
dered and  sifted  metal  was  intimately  mixed  with  eight  gramsof  the 
sodium  peroxide  mixture  above  mentioned,  or  with  four  grams 
each  of  sodium  carbonate  and  nitrate.  The  somewhat  violent 
reaction  set  up  on  the  application  of  strong  heat  to  the  platinum 
crucible  was  completed  in  a  few  minutes.  The  crucible  was 
heated  for  about  twenty  minutes  in  all.  After  cooling  the  con- 
tents were  softened  in  water,  the  solution  decanted  and  the  resi- 
due ground,  while  wet,  in  a  mortar.  The  solution  and  residue 
were  then  digested  in  a  beaker  on  the  water-bath  for  one  hour 
after  addition  of  two  cc.  of  strong  bromine  water.  The  liquid 
was  then  filtered,  acidulated  with  hydrochloric  acid,  evaporated 
to  dryness  to  separate  the  small  portion  of  silica  which  had 
entered  in  solution  and  filtered.  The  sulphuric  acid  was  deter- 
mined in  the  filtrate  in  the  usual  manner.  The  barium  sulphate 
obtained  was  always  white.  If  the  fusion  mixture  contains 
sodium  carbonate  and  nitrate,  but  no  sodium  peroxide,  the  cru- 
cible must  be  heated  for  a  longer  time,  but  a  portion  of  the  car- 
bon of  the  iron  may  still  remain  unoxidized. 

2.  Ferromanganese. — In  this  case  it  is  better  to  use  a  mixture 
of  equal  parts  of  sodium  nitrate  and  carbonate,  omitting  the 
sodium  peroxide. 

Ten  grams  of  the  mixture  were  divided  into  two  portions,  one 
of  which  was  fused  in  a  crucible.  The  other  portion  mixed 
with  two  or  two  and  one-half  grams  of  the  finely  powdered  iron 

1  Two  different  materials  are  sold  which  are  suitable  for  the  siftin^r*  One  is  caUed 
t>olting  cloth,  the  other  bolting  sheeting.  The  bolting  cloth  used  in  these  experimenta 
contained  about  eighty-five  meshes  to  the  linear  inch,  while  in  the  bolting  sheeting 
about  one  hundred  and  thirty-five  were  counted.  The  muterial  having  the  smaller 
number  of  meshes  is  made  of  coarser  threads,  however,  and  yields,  on  account  of  the 
smaller  openings,  a  finer  powder.  Bolting  cloth  is,  on  this  account,  better  suited  to  the 
preparation  of  a  sample  of  white  iron  for  a  determination  of  sulphur  by  the  method 
described. 


DETERMINATION   OF   SULPHUR   IN   CAST  IRON.  IO85 

was  then  slowly  added.  Although  too  violent  combustion  of 
the  iron  is  to  be  avoided,  it  seems  to  be  important,  for  the  suc- 
cess of  the  method,  that  a  reaction  of  decided  intensity  should 
occur  during  the  fusion. 

Sodium  nitrate  possesses  an  advantage  over  sodium  peroxide 
in  its  greater  purity,  the  formet  compound  being  readily  obtain- 
able with  practically  insignificant  traces  of  sulphur. 

Natural  gas  was  the  fuel  used  for  the  Bunsen  burner  in  heat- 
ing the  charges.  This  gas  was  found  by  repeated  experiments, 
not  to  contain  a  sufficient  quantity  of  sulphur  to  affect  the  purity 
of  the  sodium  carbonate  when  heated  in  a  platinum  crucible  in 
the  same  manner  as  in  the  case  of  the  determinations  described. 

The  usual  occurrence  of  sulphur  compounds  in  coal  gas  would 
preclude  its  use  in  the  application  of  the  method. 

From  the  experiments,  the  results  of  which  are  stated  in  the 
accompanying  table,  there  seems  to  be  some  reason  to  suppose 
that  not  quite  all  the  sulphur  of  the  iron  is  converted  into  barium 
sulphate  when  the  metal  is  oxidized  and  dissolved  by  nitric  acid. 
That  it  has  been  completely  recovered  by  the  process  of  fusion 
cannot  be  positively  asserted.* 

The  method  I  have  described  is  not  proposed  as  a  substitute 
for  any  existing  method.  The  purpose  of  the  present  work  was 
merely  to  ascertain  as  far  as  possible  whether  by  a  process  of  direct 
oxidation  of  the  iron  in  a  dry  state  a  larger  proportion  of  the  sul- 
phur could  be  recovered  in  weighable  form  than  by  the  usual 
method  of  oxidation  and  solution  in  nitric  acid. 

My  thanks  are  especially  due  to  Mr.  F.  B.  Smith  for  great 
care  and  attention  to  detail  in  conducting  the  experiments!  have 
detailed. 

1  The  meUiod  of  preparation  of  a  sample  for  analysis  in  the  case  of  the  more  brittle 
forms  of  iron,  by  crushing  in  a  steel  mortar  and  sifting,  is  suggested  in  Regnault's  Ele- 
ments of  Chemistry,  translated  from  the  French  by  Betton,  1867,  a,  iia. 


io86 


DETERMINATION  OF  SULPHUR   IN   CAST  IRON. 


Character  of  irou  used. 

White  iron  A  crushed 
in  mortar  and  sifted 
through  bolting 
sheeting. 


Fusion  mixture  employed. 
Contained  equal  parts 

of  sodium  carbonate 

and  nitrate. 


5  w  o 

A«  •B.o 
O.II2 
O.I  12 
O.III 
0.107 

O.I  14 
O.I  14 

0.106 
0.108 
0.107 
0.103 


Means 


White  iron  B  crushed 
and  sifted. 


Contained 
45  parts  NaNOs 
45  parts  Na^Oj 
10  parts  Na,COs 


Means 


Ferromanganese 
crushed  and  sifted. 


Contained  equal  parts 
of  sodium  nitrate 
and  carbonate. 


0.155 
0.150 

0.130 

0.139 
0.166 

0.156 

0.156 

0.161 

0.151 

0.151 

0.022 
0.027 
0.018 
0.018 
0.018 
0.019 
0.016 


Means 0.020 


Gray   iron    drillings 
powdered  by  rubber 
and  plate. 
Not  sifted. 


Contained  equal  parts 
of  sodium  nitrate 
and  carbonate. 


0.034 
0.030 
0.036 
0.034 
0.033 
0.034 


pi4  C.O0  a 

O.IOI 

0.098 
0.096 

0.099 

O.IOO 

0.102 
0.102 

O.IQ4 


0.109  O.IOO 


0.143 
0.149 
0.143 
0.147 


0.145 

0.012 
0.013 
0.012 

O.OIO 


0.012 

0.027 
0.030 
0.026 
0.028 
0.028 
0.022 


Means 


0.033        0.027 


CARBON   DETBRniNATIONS  IN   PIQ   IRON. 

By  BERTIUND  S.  8UMSCBS.S. 
Received  October  3,  i8|p6. 

THOSE  chemists  who  have  had  occasion  to  do  many  carbon 
determinations  in  pig  iron,  to  which  was  allotted  but  little 
time,  have  probably  felt  the  need  of  improvements  in  some  of 
our  standard  methods. 

The  old  oxygen  combustion  method,  although  accurate,  re- 
quires more  time  than  can  usually  be  spared  if  use  is  made  of  a 
porcelain  or  glass  tube.  However,  it  has  the  greatest  of  all  ad- 
vantages, that  of  accuracy.  The  writer  has  used  for  some  time 
a  regular  Bunsen  furnace  with  a  glass  tube,  and  while  the 
results  were  all  that  could  be  desired,  the  time  required  for  a 
refractory  residue  was  almost  three  hours. 

A  series  of  experiments  was  conducted  with  the  ordinary 
chromic  acid  process,  but  the  results  were  quite  unsatisfactory. 
Every  precaution  was  taken  to  insure  accuracy,  but  with  high 
carbon  residue  low  results  were  obtained  in  nearly  every  case 
when  checked  by  the  oxygen  combustion  method.  This  was 
particularly  noticeable  when  a  considerable  content  of  graphite 
was  present.  The  results  checked  quite  well  with  each  other 
and  gave  satisfactory  results  when  working  on  steel. 

As  this  state  of  affairs  greatly  embarrassed  matters  in  the  lab- 
oratory, an  effort  was  made  to  devise  some  means  by  which  the 
carbon  could  be  determined  with  reasonable  speed  and  accuracy. 

Recognizing  the  advantages  of  the  combustion  method,  it  was 
decided  to  make  use  of  a  platinum  tube.  To  avoid  delay  and 
expense  the  tube  was  manufactured  in  the  factory.  It  was  made 
of  0.200  stock  twelve  inches  long  and  eleven-sixteenths  inch  in 
diameter.  A  perfectly  tight  tube  was  constructed  by  using 
ordinary  gold  solder,  which  may  be  obtained  from  any  jeweler. 
Around  each  end  of  the  tube  copper  coolers  were  brazed,  in 
order  to  cool  the  tube  in  the  neighborhood  of  the  rubber  stop- 
pers. The  inlet  of  the  coolers  served  the  double  purpose  of 
supports  and  water  supplies.  In  spite  of  this  precaution  it  was 
found  that  the  air  circulating  through  the  heated  portion  of  the 
tube  was  hot  enough,  on  reaching  the  stoppers,  to  seriously 
affect  them.  In  order  to  prevent  this,  the  scheme  shown  in  Fig.  3 


io88 


BBRTRAND  S.   SUMMERS. 


was  devised.  The  funnel  shape  protuberance  here  seen  was 
filled  with  ignited  asbestos,  and  the  whole  was  removed  with  the 
stoppers.      This  appliance  proved  an  effectual  preventive  for 

further  heating  of  the 


r 


I 


ro5. 


Fic.3. 


I 


i' 


Tife^f 


F/G.2 


Stoppers,  as  a  red  heat 
could  be  maintained 
two  inches  from  them 
and  they  remain  per- 
fectly cool. 

With  this  arrange- 
ment it  was  found  that 
a  high  carbon  residue 
could  be  burned  com- 
pletely in  twenty  min- 
utes. It  became  evi- 
dent from  this  that  if 
the  aspirating  space 
were  decreased,  good 
results  could  be  ob- 
tained in  a  compara- 
tively short  time. 
With  this  idea  in  view , 
the  train  depicted  in 
the  accompanying  pic- 
ture was  designed  and 
made  by  our  own 
glass-blowers.  The 
train  has  the  further 
advantage  that  rub- 
ber  connections    are 


Pig.  \—AAy  Platinum  tube  ;  BB^  Support  and  water 
outlet ;  CC,  Coolers  ;  DD^  Water  supplies. 

Fig-  ^—bb^  Sockets  for  BB  \  dd.  Connection  for  water 
supply,  DD :  E,  Main  water  supply  ;  F^  waste  pipe ; 
(7,  Gas  connection. 

Pig.  3'-<^s,  Stoppers  ;  yf  j,  Glass  cup  for  asbestoses ; 
y4^  Outlet. 

Pigs.  4  and  5— Showing  connections  for  mercury  joint. 

avoided,  the  only  rubber  tubing  in  use  being  at  the  ends  of  the 
combustion  tube. 

The  purifying  train  consists  of  a  large  ({-tube  of  one  and  one- 
half  inch  stock  and  twelve  inches  long.  The  first  limb  is  filled 
with  broken  caustic  potash,  and  the  second  with  fused  calcium 
chloride.  The  first  limb  connects  with  a  Drechsel  bottle  par- 
tially filled  with  strong  sulphuric  acid,  and  the  second  with  the 
combustion  tube. 


CARBON  DETERMINATIONS  IN  PIG  IRON. 


I090  CARBON   DETERMINATIONS  IN  PIG  IRON. 

The  purifying  train  on  the  absorption  end  is  made  in  one 
piece.  It  consists  of  a  five  inch  ||-tube  of  thick  walled  glass 
five-eighths  inch  in  diameter,  into  the  sides  of  the  limbs  of  which 
are  fused  arms.  These  arms  are  made  of  one-inch  stock  and 
about  seven  inches  long.  The  first  arm  is  filled  with  anhydrous 
cuprous  chloride  and  anhydrous  cupric  sulphate.  The  ||-tube 
serves  as  the  receptacle  for  the  sulphuric  acid,  and  the  second 
arm  is  filled  with  calcium  chloride  previously  treated  with  an 
excess  of  carbon  dioxide. 

The  connection  with  the  Geissler  bulbs  is  established  by  means 
of  mercury  joints.  These  serve  to  facilitate  removal  of  the  bulbs 
and  make  a  joint  which  is  perfectly  secure.  The  joint  can 
readily  be  made  by  any  glass  blower,  an  illustration  of  which  is 
seen  in  Fig.  4.  The  end  of  the  Geissler  bulb  is  so  reamed  as  to 
fit  loosely  over  the  tube  inside  the  cup  (Fig.  5).  A  small  piece 
of  rubber  tubing  (b,  Fig.  4)  is  slipped  over  the  tube  and  makes 
a  moderately  tight  joint  with  the  end  of  the  Geissler  bulb. 
When  the  cup  is  filled  with  mercury  a  perfect  connection  is 
obtained.  The  method  of  connecting  the  Geissler  bulb  with 
rubber  tubing  was  both  awkward  and  liable  to  leakage.  These 
junctions  have  been  in  use  for  some  time  in  our  laboratory  and 
have  given  thorough  satisfaction. 

With  this  apparatus  as  described  the  most  refractory  residues 
are  burned  in  an  hour  and  a  half.  With  residues  of  less  refrac- 
tory nature  and  lower  carbon  content  an  estimation  may  be  com- 
pleted in  less  time.  The  blank  on  the  apparatus  never  exceeds 
three-tenths  of  a  milligram,  and  is  usually  one-tenth  or  nil. 

Some  results  are  here  appended,  thinking  they  may  be  of 
interest.  Those  obtained  by  the  chromic  acid  process  were  quite 
scattering  unless  great  care  was  exercised  and  sufficient  time  was 
allowed.  The  results  from  this  method,  given  below,  are  those 
where  much  time  was  given  and  great  pains  taken  to  insure 
complete  oxidation.  V^ues  from  the  Bunsen  furnace  are  given 
to  serve  for  comparison. 

Chromic  Acid  Method.  Bunsen  Furnace. 

Total  Carbon.  Total  Carbon. 

323  3-31 

3.27  3-33 

3.23 
%,2S 


SOLUBILITY  OP  BISMUTH  SULPHIDE.  IO91 

Results  from  the  above  described  process,  when  compared 
with  the  Bunsen  furnace,  were  very  good. 

Platinum  Furnace.  Bunsen  Furnace. 

Total  Carbon.  Total  Carbon. 

303  303 

3.03  3.05 

3.05 

The  convenience  of  this  apparatus  in  expediting  work  in  the 
laboratory  has  led  me  to  write  this  description,  in  the  hope  that 
it  might  be  of  service  to  other  chemists. 

Cbbmicax«  I«aboratort»  Wbstbrn  Slbctrxc  Company, 

Chicago. 


NOTE  ON  THE  SOLUBILITY  OF  BISMUTH  SULPHIDE  IN 

ALKALINE  SULPHIDES. 

By  Gborob  C.  Stonb. 

Received  November  9,  1896. 

IN  the  August  number  of  this  Journal  there  is  a  note  by 
Prof.  Stillman  on  this  subject ;  he  shows  that  if  a  solution 
containing  bismuth  is  made  alkaline  by  sodium  hydroxide  and 
then  heated  with  an  excess  of  an  alkaline  sulphide  a  considera- 
ble amount  of  bismuth  is  held  in  solution.  On  repeating  his 
experiments  qualitatively  I  obtained  the  same  result,  but  when 
the  bismuth  was  first  precipitated  as  sulphide  from  an  acid  solu- 
tion and  then  treated  with  an  alkaline  sulphide  but  little  if  any 
was  dissolved. 

To  test  the  solubility  quantitatively  I  made  a  solution  of  about 
one  and  two-tenths  grams  of  bismuth  hydroxide  in  500  cc.  of 
very  dilute  hydrochloric  acid ;  .in  two  portions,  of  fifty  cc.  each. 
I  determined  the  bismuth  by  precipitation  by  ammonium  car- 
bonate, finding  0.0966  and  0.0965  gram. 

I  next  precipitated  the  bismuth  in  two  more  lots  of  the  same 
solution  by  hydrogen  sulphide,  filtered  and  heated  the  precipi- 
tate for  half  an  hour  with  a  large  excess  of  potassium  sulphide, 
filtered,  dissolved  and  reprecipitated  by  ammonium  carbonate, 
the  bismuth  weighed  0.0981  and  0.0970  g^am. 

Two  more  lots  treated  in  the  same  manser,  except  that  they 
^wexe  heated  with  ammonium  sulphide,  gave  0.0970  and  0.0976 
gram  of  bismuth. 

From  the  above  it  seem?  fair  to  conclude  that  bismuth 
sulphide  precipitated  from  an  acid  solution  is  not  dissolved  by 
subsequent  treatment  with  an  alkaline  sulphide. 


[Contribution  from  the  Laboratory  op  Agricui^tural  Chemistry 

OP  THB  Ohio  State  UNivBRSiry.] 

ON  THE  BEHAVIOR  OF  COAL-TAR  COLORS  TOWARD  THE 

PROCESS  OP  DIQESTION. 

By  H.  a.  Webbr. 
Received  October  to,  x9g6. 

IT  is  very  well  known  that  the  coal-tar  colors  have  come  into 
general  use  for  coloring  confectionery  and  other  articles  of 
food  and  drink.  In  fact  they  have  almost  completely  superseded 
the  vegetable  colors,  which  have  been  used  from  time  immemo- 
rial for  a  similar  purpose.  The  indiscriminate  use  of  these 
colors,  some  of  which  are  derived  from  bodies  of  a  decidedly 
poisonous  nature,  has  often  been  regarded  with  suspicion  by 
persons  who  are  interested  in  public  health.  On  account  of  the 
uncertainty  existing  in  regard  to  these  colors  from  a  sanitary 
point  of  view,  Austria  has  prohibited  their  use  tn  toto  in  all  arti- 
cles of  food  and  drink.  Other  countries  prohibit  certain  of  the 
colors,  which  have  been  shown  to  be  injurious,  and  allow  all 
others  to  be  used. 

The  experiments  made  upon  the  lower  animals  have,  in  the 
main,  revealed  negative  results.  Thus  the  writer  about  eight 
years  ago  fed  some  of  the  colors  most  commonly  employed  by 
confectioners  to  rabbits  in  order  to  test  this  question.  One-half 
gram  of  the  colors,  among  which  magenta  and  corallin  were 
included,  was  fed  to  as  many  rabbits  per  day  for  ten  days  in  suc- 
cession without  any  apparent  ill  effects.  The  exhaustive  treatise 
of  Dr.  Weil,  translated  by  Leffmann,  ascribes  toxic  effects  to 
only  a  small  number  of  the  many  colors  employed  by  him  in  his 
experiments  upon  domestic  animals. 

The  effect  which  these  colors  might  exert  upon  digestive  fer- 
ments, however,  was  a  subject  which  had  as  yet  received  no 
attention,  and  the  following  experiments  were  undertaken  in 
order  to  throw  some  light  upon  this  question.  The  ferments 
employed  were  Armour's  pepsin  and  pancyeatine,  liberal  sam- 
ples of  which  were  kindly  furnished  by  Armour  &  Co.,  of  Chi- 
cago. 

For  the  purpose  of  showing  the  digestive  action,  blood  fibrin 
preserved  in  alcohol  was  employed.     The  fiibrin  was  soaked  and 


BEHAVIOR  OP  COAL-TAR  COIX>RS.  IO93 

thoroughly  washed  with  water  to  remove  the  alcohol,  then 
pressed  between  filter  paper,  and  the  amount  required  for  each 
experiment  weighed  off. 

In  each  set  of  experiments  a  control  experiment  was  carried 
on  without  the  addition  of  color.  The  mixture  was  made  as  fol- 
lows : 

Hydrochloric  acid  solntion  (two-tenths  per  cent.)  xoo  cc. 

Pepsin*  - 20  milligrams. 

Fibrin i  gram. 

This  mixture  placed  in  a  large  test-tube  was  digested  in  a 
water-bath  at  a  temperature  of  38^  to  40**  C.  until  the  fibrin  was 
dissolved. 

At  the  same  time  similar  mixtures  as  above  containing  in  addi- 
tion I,  0.5,  Q.250,  0.125,  And  0.062  gram  of  the  color  to  be  tested 
respectively,  were  digested  in  the  same  water-bath  for  the  time 
required  to  dissolve  the  fibrin  in  the  control  experiment.  Any 
fibrin  remaining  undissolved  in  the  latter  tests,  was  removed, 
thoroughly  washed,  pressed  between  filter  paper  as  before  and 
weighed. 

I .   PEPSIN  AND  OROUNB  YELLOW. 

This  color  was  one  of  a  series  employed  in  the  coloring  of  con- 
fectionery, and  was  found  to  be  what  is  known  in  the  trade  as 
Acid  Yellow  or  Fast  Yellow,  and  is  a  mixture  of  sodium  amido- 
azobenzenedisulphonate  with  sodium  amidoazobenzenemono- 
sulphonate. 

Amouut 
Amounfof        Amount  of     Amount  of     Duration  of     offibnn 
color.  fibrin.  pepsin.       experiment,    dlasolved. 

Gram  Gram.  Gram.  Hours.  Gram. 

I 0.0  I  0.020  3  I.O 

a 1.0  I  0.020  3  O.I 

3 0.5  I  0.020  3  0.12 

4 0.25  I  0.020  3  0.22 

5 0.125  I  0.020  3  0.35 

6 0.062  I  0.020  3  0.73 

From  this  it  will  be  seen  that  even  in  test  No.  6,  where  the 
color  employed  amounted  to  only  one  part  in  1600  parts  of  the 
solution,  the  presence  of  the  color  had  still  a  depressing  effect. 
For  fear  that,  owing  to  the  nature  of  this  color,  the  hydrochloric 
acid  might  have  been  neutralized  in  part,  the  experiment  was 


I094  H.   A.   WKBER. 

repeated  with  a  six-tenths  per  cent,  solution  of  hydrochloric  acid 
with  similar  results. 

Of  course  the  determination  of  the  fibrin  dissolved  is  only 
approximate,  as  can  readily  be  inferred  from  the  way  it  was 
done. 

In  tests  Nos.  2,  3  and  4  no  change  in  the  amount  of  fibrin  was 
apparent  to  the  eye.  That  a  smaU  part  of  the  fibrin  had  gone 
into  solution  was  confirmed  by  the  fact  that  a  slight  precipitate 
of  albuminoids  was  obtained  on  the  addition  of  a  solution  of  tan- 
nin. On  the  whole  it  must  be  conceded  that  this  color  has  a 
marked  and  injurious  effect  upon  peptic  digestion. 

2.    PEPSIN  AND  SAFFOLINE. 

This  is  also  a  candy  color  and  was  found  to  be  acridine  red. 


Amount  of 
color. 
Gram. 

Amount  of 
fibrin. 
Gram. 

Amount  of 
pepsin. 
Gram. 

Duration  of 

experiment. 

Hours. 

Amount 
of  fibrin 
dissolved 
Gram. 

0.0 

0.020 

3i 

I.O 

0.020 

5 

0.5 

0.020 

S 

0.25 

0.020 

S 

0.125 

0.020 

3i 

0.062 

0.020 

3i 

As  will  be  seen  from  the  table  above,  this  color  only  slightly 
retards  the  digestion  of  the  fibrin  in  the  three  stronger  solutions, 
while  in  the  last  two  tests  there  waS  no  interference  with  the 
process.  On  the  whole  it  may  be  said  that  the  effect  of  this  color 
on  peptic  digestion  is  practically  nil. 

3.    PEPSIN  AND  MAGENTA. 

It  is  needless  to  tabulate  the  results  of  this  experiment.  Suf- 
fice it  to  say  that  the  solution  of  the  fibrin  in  the  five  tests  con- 
taining the  same  proportions  of  the  color  as  employed  above  kept 
pace  throughout  the  whole  duration  of  the  experiment  with  the 
control  test,  the  fibrin  in  all  cases  dissolving  at  the  expiration  of 
three  and  one-half  hours. 

This  color,  therefore,  seems  not  to  interfere  with  peptic  diges- 
tion. 

These  four  colors  were  also  employed  with  pancreatin,  and 
the  method  was  as  follows : 


BEHAVIOR  OF  COAVTAR  COLORS.  IO95 

For  the  control  experiment  the  following  mixture  was  made  : 

Water 100  cc. 

Sodium  bicarbonate 1.5  grams. 

Pancreatin 0.3  g^am. 

Fibrin i.o  gram. 

This  mixture  contained  in  a  large  test-tube,  was  digested  in  a 
water-bath  until  the  fibrin  was  peptonized.  To  test  the  effect  of 
the  colors,  there  was  added  to  similar  mixtures  as  above  i,  0.5, 
0.25,  0.125  and  0.062  gram  of  each  color  respectively. 

5.    PANCREATIN  AND  OROLINE  YELLOW. 

To  the  great  surprise  of  the  writer,  this  color,  which  had 
proved  so  effective  in  stopping  and  retarding  peptic  digestion, 
was  found  to  exert  no  action  whatever  on  the  pancreatic  fer- 
ment ;  the  fibrin  in  all  five  of  the  tests  with  this  color,  dissolved 
as  freely  as  that  of  the  control  test.  The  solution  of  the  fibrin 
in  all  cases  was  completed  at  the  expiration  of  six  hours. 

PANCREATIN  AND  SAFFOUNE. 

The  action  of  this  color  was  quite  different  from  that  of  oroline 
yellow,  as  the  subjoined  table  will  show  : 


Amount 

Amount  of        Amount  of    Amount  of    Duration  of     of  fibrin 
color.  fibrin.       Pancreatin.  experiment,    dissolved. 

Gram.  Gram.  Gram.  Hours.  Gram. 


I 0.0  I  0.3  6|  1.0 

2 1 .0  I  0.3  6|  0.0 

3 0.5  I  0.3  6i  0.0 

4 0.25  I  0.3  6J  0.55 

5 0.125  I  0.3  6\  0.65 

6 0.062  I  0.3  6J  0.75 

These  results  show,  that  in  the  two  stronger  solutions  the 
action  of  the  pancreatic  ferment  was  entirely  stopped,  and  that 
even  in  test  No.  6,  which  contained  only  one  part  of  color  to 
1600  of  the  solution  the  action  of  the  ferment  was  retarded  to  a 
marked  extent. 

Tannin  precipitates  the  coloring  matter. 

7.    PANCREATINE  AND  MAGENTA. 

This  color  was  as  marked  in  retarding  and  stopping  the  action 
of  pancreatine  as  saffoline.  The  results  are  given  in  the  table 
below  : 


I 0.0 

2 I.O 

3 0.5 

4 0.25 

5 0.125 

6 0.062 


1096  JEROMB  KKLLEY,  JR.  AND  EDGAR  F.  SMITH. 

Amount 
Amount  of       Amount  of    Amount  of    Duration  of      of  fiber 
color.  fiber,      pancreatine,  experiment,    dissolved. 

Gram.  Qram.  Gram.  Hour.  Gram. 

0.3  6i  1.0 

0.3  6i  0.0 

0.3  6J  0.0 

0.3  6J  0.40 

0.3  6^  0.60 

0.3  6i  0.73 

The  solutions  of  tests  2  and  3  gave  no  precipitate  with  tannin. 
In  all  other  tests  the  precipitate  was  either  marked  or  heavy. 

8.  PANCREATINE  AND  METHYL  GRANGE. 

This  color  in  all  of  the  tests  behaved  like  the  last  three  colors 
described,  completely  stopping  the  action  of  the  pancreatine  in 
the  two  strongest  solutions  and  retarding  it  to  a  marked  extent 
in  the  weakest.  The  tabular  statement  would  be  similar  to  the 
last. 

It  seems  then,  so  far  as  these  four  colors  are  concerned,  that 
none  interfere  with  both  peptic  and  pancreatic  digestion,  but  that 
each  color  interferes  seriously  with  either  the  one  or  the  other. 
What  the  action  of  other  coal  tar  colors  may  be,  can,  of  course, 
not  be  iiiferred  from  this  limited  number  of  experiments,  but  it 
may  safely  be  said  that  bodies  which  have  such  a  decided  action 
in  retarding  the  most  important  functions  of  the  animal  economy, 
cannot  properly  have  a  place  in  our  daily  food  and  drink. 


[CONTRIBtrriUN  PROM  THE  JOHN  HARRISON  LABORATORY  OF  CHBMISTRT» 

No.  19.] 
THE  ACTION  OF  ACID  VAPORS  ON  ilETALLIC  SULPHIDES. 

*  By  Jekomb  Kbllby,  Jr.  and  Exx»ar  P.  Smith. 

RecclTcd  October  a,  tSgS. 

EXPERIMENTS  made  in  this  laboratory  on  the  action  of 
the  vapors  of  hydrochloric  acid  upon  the  sulphide  of 
arsenic  proved  that  the  latter  is  wholly  volatilized.  The  purpose 
of  the  present  communication  is  to  record  further  observations 
along  analogous  lines.  Thus,  when  washed  and  dried  arsenic 
trisulphide  is  exposed  to  the  action  of  hydrobromic  acid  gas,  it 
volatilizes  completely.  Indeed  the  action  commences  in  the  cold 
with  the  formation  of  a  liquid  that  passes  out  of  the  containing 


ACTION  OF  ACID  VAPORS  ON  METALLIC  SULPHIDES.     IO97 

vessel  upon  the  application  of  a  very  gentle  heat.     In  evidence 
of  this,  two  quantitative  experiments  may  be  given  : 

Arsenic  sulphide  taken.  Arsenic  sulphide  expelled. 

Gram.  Gram. 

0.2945  0.2941 

0.4632  0.4628 

Antimony  trisulphide,  like  that  of  arsenic,  is  volatilized  by 
hydrochloric  acid  gas.  It  was  quite  probable  that  a  like  deport- 
ment would  be  observed  if  hydrobromic  acid  gas  should  be  sub- 
stituted. This  was  found  to  be  the  case.  When  the  gas  came 
in  contact  with  the  sulphide  it  became  liquid  and  volatilized  as 
soon  as  a  gentle  heat  was  played  upon  the  boat  in  which  the  sul- 
phide was  contained. 

Antimony  sulphide  taken.  Antimony  sulphide  expelled. 

Oram.  Gram. 

0.1473  0.1469 

0.0938  0.0935 

Upon  substituting  stannic  sulphide  for  antimony  sulphide,  an 
experience  similar  to  that  observed  with  antimony  and  arsenic 
sulphides  followed.  There  was  a  complete  volatilization  with 
but  a  trifling  residue,  which  proved  to  be  carbon  from  filter 
paper  that  had  adhered  to  flie  metallic  sulphide. 

stannic  sulphide  taken.  Stannic  sulphide  expelled. 

Gram.  Gram. 

0.1880  0.1880 

0.5527  0.5521 

0.4174  0.4169 

The  oxides  of  arsenic,  antimony  and  tin  (at  least  in  the  stan- 
nic form)  can  be  volatilized  in  a  current  of  hydrochloric  acid 
gas.  This  is  also  true  of  the  sulphides  of  arsenic  and  antimony, 
but  how  the  two  sulphides  of  tin  would  act  under  like  conditions 
was  not  known. 

Experiments  recently  made  demonstrate  the  perfect  volatility 
of  stannic  sulphide  in  this  way.  With  stannous  sulphide  it  was 
found  that  by  the  continued  action  of  the  gas  in  the  cold  there 
followed  a  complete  conversion  into  chloride  without  any  vola- 
tilization. That  the  residue  was  the  chloride  was  evident  from 
its  action  upon  a  mercuric  salt  solution.  The  figures  obtained 
in  the  several  trials  were  : 


1098        HERBERT  A.   SCHAFPER   AND   EDGAR   F.    SMITH. 

Stannous  chloride  found.  Stannous  chloride  tbcory. 

Gram.  Gram. 

0.3544  0.3523 

0.4893  0.4903 

Several  attempts  were  made  to  separate  stannous  and  stannic 
sulphides  by  this  procedure.  The  results  were  unsatisfactory. 
In  order  to  drive  out  the  stannic  salt  completely  it  is  necessary 
to  heat  the  mixture,  and  this  caused  a  partial  volatilization  of 
the  stannous  chloride,  so  that  quantitative  results  could  not  be 
obtained. 

Comparatively  few  metallic  sulphides  have  been  studied  in 
the  direction  indicated  in  the  preceding  lines,  so  that  it  is  prob- 
able a  wider  application  of  the  method  will  disclose  interesting 
behaviors,  and  that  probably  new  separations  can  be  brought 
about  in  this  way.  The  action  of  the  vapors  of  haloid  acids  has 
also  been  tried  on  natural  sulphides  with  a  fair  degree  of  suc- 
cess. 


[Contribution  from  the  John  Harrison  Laboratory  of  Chemistry, 

No.  20.] 

TUNGSTEN  HEXABROMIDE. 

By  Herbert  A.  Schapfbr  an%Edoar  P.  Smith. 

Received  October  to,  1896. 

THE  most  recent  work  upon  tungsten  bromides  is  that  of 
Roscoe,'  who  endeavored  to  prepare  a  hexabromide,  but  ob- 
tained instead  a  penta  derivative  from  which  the  dibromide  was 
subsequently  made.  By  reference  to  the  literature  bearing  upon 
this  subject  it  will  be  noticed  that  bromine,  diluted  with  carbon 
dioxide,  was  made  to  act  upon  tungsten  metal  exposed  to  a 
red  heat.  Experimental  evidence  is  at  hand  that  tungsten  at 
high  temperatures  deoxidizes  carbon  dioxide,  thus  allowing 
ample  opportunity  for  the  production  of  oxybroraides,  which,  in 
spite  of  the  greatest  care,  are  sure  to  appear  in  larger  or  smaller 
amount.  The  thought  also  suggested  itself  that  possibly  the 
**red  heat"  at  which  the  action  was  allowed  to  occur  might 
have  been  detrimental  and  have  indeed  prevented  the  formation 
of  the  hexabromide. 

Hence,  we  determined  to  operate  in  an  atmosphere  of  nitro- 

i  Ann.  Chem,  (Liebig),  xte,  363. 


TUNGSTEN  HBXABROMIDB.  IO99 

gen  and  to  apply  a  very  gentle  heat  to  the  vessel  containing  the 
tungsten.  In  this  connection  it  may  be  mentioned  that  the 
nitrogen  was  conducted  through  a  series  of  vessels  charged  with 
chromous  acetate,  sulphuric  acid,  caustic  potash,  and  phos- 
phorus pentozide,  respectivel3\  It  then  entered  an  empty  ves- 
sel into  which  dry  bromine  was  dropped  from  a  tap-funnel,  and 
after  passing  through  a  tall  tower,  filled  with  calcium  chloride, 
entered  a  combustion  tube  resting  in  a  Bunsen  furnace.  The 
anterior  portion  of  the  combustion  tube  was  contracted  at  inter- 
vals, forming  a  series  of  bulbs,  and  at  its  extremity  was  con- 
nected with  an  empty  Woulff  bottle,  followed  by  a  calcium  chlo- 
ride tower,  and  finally  a  receiver  filled  with  soda  lime  and 
broken  glass.  A  steady  current  of  nitrogen  was  conducted 
through  this  system  for  a  period  of  three  days.  On  the  fourth 
day  bromine  was  introduced.  The  tungsten  contained  in  the 
combustion  tube  was  heated  very  gently.  In  a  short  time 
brown  vapors  appeared.  These  condensed  to  a  liquid  beyond 
the  boat  and  eventually  passed  into  blue-black  crystalline  masses 
that  separated  from  the  walls  of  the  tube,  when  perfectly  cold, 
with  a  crackling  sound.  Very  little  heat  was  required  to  melt 
them  and  they  could  with  care  be  resublimed  in  distinct,  blue- 
black  needles.  The  latter  was  collected  in  one  of  the  bulbs  (No. 
2)  previously  mentioned.  Other  products  were  observed  and 
isolated.  All  were  analyzed.  Bulb  No.  i— that  nearest  the 
tungsten  metal — contained  a  black,  velvety  compound,  which 
upon  analysis  showed  the  presence  of  tungsten  dibromide.  Bulb 
No.  2  contained  0.2103  gram  of  the  blue-black  crystals,  which 
yielded  0.0577  gram  of  tungsten,  or  27.43  per  cent.,  and  0.1543 
gram  of  bromine,  or  73.53  per  cent.  The  theoretical  require- 
ments of  tungsten  hexabromide are  27.72  percent,  tungsten  and 
72.28  per  cent,  bromine.  The  bromine  percentage  found  is 
high.  This  may  be  due  to  traces  of  bromine  that  had  not  been 
driven  out  from  the  crystalline  deposit,  or  to  adherent  silver 
tungstate,  as  some  tungstic  acid  remained  in  the  solution  from 
which  the  silver  bromide  was  precipitated. 

A  fresh  portion  of  the  blue-black  crystals  was  prepared  as 
before  and  analyzed.  The  bromine  determination  was  unfortu- 
nately lost.     The  determination  of  the  tungsten  resulted  as  fol- 


IICX)  EDMUND   H.   MILLER. 

lows  :  0.4351  gram  oi  material  gave  0.1222  gram  of  tungsten  or 
28.08  per  cent. 

A  third  preparation  was  made.     On  subjecting  0.1775  gram 
of  it  to  analysis  these  results  were  obtained  : 

0.0496  gram  tungsten  or  27.94  per  cent. 
0.1266  gram  bromine  or  71.32  per  cent. 
Tabulating  the  series,  we  have  : 


pound. 

Mean 

Required  for 
hexabromide 

Per  cent. 

Per  cent. 

Per  cent. 

Per  cent 

Per  cent 

Tungsten  •  •  • 

...   27.4} 

28.08 

27.94 

27.81 

27.72 

Bromine*  ••• 

•••   73-53 

71-62 

•  •  •  • 

72.33 

72.28 

These  figures  give  evidence  that  the  body  analyzed  is  tung- 
sten hexabromide. 

In  analyzing  the  third  portion  of  the  blue-black  needles  the 
bromine  was  determined  by  placing  the  material  in  a  small 
Erlenmeyer  bulb,  covering  it  with  nitric  acid  and  then  distill- 
ing. The  liberated  bromine  was  passed  into  a  silver  nitrate 
solution. 

The  tungsten  hexabromide  prepared  by  us  consists,  as  already 
observed,  of  blue-black  needles.  Moderately  elevated  tempera- 
tures decompose  the  compound.  It  gives  off  fumes  when 
brought  in  contact  with  the  air.  Water  decomposes  it  with  the 
formation  of  a  royal-blue  colored  oxide.  Ammonia  water  dis- 
solves it,  the  solution  remaining  colorless.  A  vapor  density 
determination  resulted  negatively,,  as  decomposition  was  appar- 
ent early  in  the  experiment. 


NOTES   ON   THE   FERROCYANIDES  OF   ZINC  AND  MAN- 
GANESE. 

By  Edmund  H.  Millbr. 

Received  October  lo,  1896. 

THE  composition  of  the  ferrocyanides  of  zinc  and  manganese, 
formed  when  salts  of  these  metals  are  precipitated  by 
potassium  ferrocyanide,  is  given  by  Prescott  dnd  Johnson'  as 
Zn,Fe(CN),  and  Mn,Fe(CN)„  while  the  books  on  volumetric 
analysis,  such  as  Sutton's  and  Beringer*s,  ignore  the  composi- 
tion of  this  precipitate. 

1  QualitAtive  Analysis,  pa^cs  67  and  57. 


FERROCYANIi:>ES  OF   ZINC   AND   MANGANESE.  HOI 

The  prevailing  idea  is  that  in  the  titration  of  zinc  by  potas- 
sium ferrocyanide,  a  normal  zinc  ferrocyanide  is  formed.  This 
I  believe  to  be  incorrect,  for  if  the  reaction  is 

K,Fe(CN).  H-  2ZnCl,  =  Zn,Fe(CN).+  4KCI, 

a  solution  of  potassium  ferrocyanide,  one  cc.  of  which  is  equiva- 
lent to  ten  milligrams  of  zinc,  would  contain  32.32  grams  of 
K^Fe(CN),.3H,0  to  the  liter,  not  43.2'  to  45*  grams,  as  has  been 
found  by  experiment.  Using  forty-four  grams  per  liter  as  a 
basis  for  calculation,  the  reaction  becomes 

2K,Fe(CN),  +  3ZnCl,  =  Zn.K,(Fe(CN).),  +  6KC1. 

This  reaction  is  not  merely  one  that  may  possibly  be  true,  but 
according  to  Wyrouboff,'  the  precipitate  formed  by  the  action  of 
potassium  ferrocyanide  on  a  zinc  salt,  whichever  is  in  excess,  is 
3Zn,Fe(CN)..K,F,(CN)..i2H,0,  white,  while  the  normal  salt, 
Zn,Fe(CN),.4H,0,  is  formed  only  by  the  action  of  hydroferro- 
cyanic  acid  on  a  zinc  salt. 

This  statement  agrees  both  with  the  preceding  reaction  and 
with  the  results  obtained  in  standardizing  potassium  ferrocya- 
nide solution. 

The  manganese  precipitate  with  potassium  ferrocyanide,  as 
obtained  in  titration,  is  given  by  Stone*  as  Mn,Fe,(CN),,.  This 
is  a  ferri-,  not  a  ferrocyanide,  thus  making  necessary  a  change 
of  quanti valence.  Mr.  Stone  also  states  that  an  amount  of 
potassium  ferrocyanide  which  will  precipitate  four  atoms  of  zinc 
will  only  precipitate  three  of  manganese,  thus  basing  his  calcu- 
lation on  the  formation  of  a  normal  zinc  ferrocyanide. 

Wyrouboff*  gives  the  precipitate  obtained  from  potassium  fer- 
rocyanide and  a  manganese  salt,  whichever  is  in  excess,  as 

5Mn,Fe(CN),.4K,Fe(CN),.4H,0,  rose  white  ; 

while  the  normal  salt  Mn,Fe(CN)..7H,0,  cream,  is  formed  as 

in  the  case  of  zinc  by  hydroferrocyanic  acid. 

The  solution  used  by  Mr.  Stone  had  the  following  strength  : 

1  Sutton  t  Volumetric  Analysis,  p.  3>9  ;  Beringer :  Assaying,  p.  319. 
*  Purman  :  Assaying,  p.  205. 
9  Amh. cAim. pAys.,  fsj.  8-  485. 
^J.Am.  Ckem.  Soc.,  17,  473. 


II02  JOHN   FIELDS. 

One  cc.  =  o.CK)6c>6  gram  zinc. 

One  cc.  '-=  0.00384  gram  manganese. 

If  the  ratio  were  exactly  four  zinc  to  three  manganese,  using 
the  most  recent  atomic  weights,  the  strength  of  this  solution 
against  manganese  would  be  one  cc.  =  0.00382  gram,  while, 
according  to  Wyrouboff,  loMn  =  9K^Fe(CN),  and  6Zn  = 
4K4Fe(CN),,  or  loMn  =  13.5  Zn,  or  iMn=  1.35  Zn,  and  the 
strength  against  manganese  would  be  i  cc.  =  0.003774  gram. 

These  figures  show  but  little  difference  between  the  two  ratios 
and  while  Mr.  Stone's  experimental  results  are  undoubtedly 
accurate,  this  theory  based  on  the  formation  of  Zn,Fe(CN),and 
Mn,Fe,(CN)„  is  not  satisfactorily  proved. 

This  article  is  only  a  preliminary  note  regarding  the  composi- 
tion of  the  ferrocyanides  as  they  are  being  investigated  in  this 
laboratory. 

In  connection  with  the  ferrocyanide  of  zinc  I  have  found  a 
very  strong  solution  of  hydrochloroplatinic  acid,  H,PtCl„  acidi- 
fied with  hydrochloric  acid,  a  most  satisfactory  indicator  for  the 
titration  of  zinc  by  potassium  ferrocyanide,  when  performed  in  a 
hot  solution.  This  indicator  is  used  in  the  same  way  as  uranium 
acetate  and  is  less  affected  by  a  varying  amount  of  hydrochloric 
acid.  The  end  reaction  is  a  bright  emerald  green,  which  takes 
a  few  seconds  to  develop.     It  will  not  work  with  a  cold  solution. 

ASSAY  Laboratory.  Columbia  Univsrsxty. 


A  MODIFICATION  OF  THE  GUNNINQ  HETHOD  FOR 

NITRATES, 

By  John  Fields. 

Receired  October  ao,  xt9^ 

THE  full  text  of  the  official  Gunning^  method  is  as  follows : 
''In  a  digestion  flask,  holding  from  250  to  500  cc,  place 
from  seven- tenths  to  three  and  five- tenths  grams  of  the  substance 
to  be  analyzed,  according  to  the  amount  of  nitrogen  pres- 
ent. Add  thirty  to  thirty-five  cc.  of  salicylic  acid  mixture; 
namely,  thirty  cc.  sulphuric  acid  to  one  gram  of  salicylic  acid, 
shake  until  thoroughly  mixed  and  allow  to  stand  five  to  ten 

1  Ann.  chim, phys.^  [5],  8,  474. 

s  Bulletin  46.  U.  S.  Dept.  of  A.%r.,  p.  z8. 


MODIFICATION  OP  GUNNING  METHOD  FOR  NITRATES.     I IO3 

minutes,  with  frequent  shaking  ;  then  add  five  grams  sodium 
thiosulphate  and  ten  grams  of  potassium  sulphate.  Heat 
very  gently  until  frothing  ceases,  then  heat  strongly  until  nearly 
colorless.  Dilute,  neutralize,  and  distil  as  in  the  Gunning 
method." 

This  method  has  its  advantages  in  that  no  heavy  metals  are 
added,  such  as  zinc  and  mercury,  which  sometimes  interfere 
with  the  distillation.  It  has,  however,  a  few  disadvantages 
which  the  following  modification  partially  overcomes.  When 
working  with  some  materials,  there  is  considerable  trouble  due 
to  persistent  frothing,  and  in  some  cases,  it  has  taken  six  hours 
constant  attention  to  get  the  digestion  safely  over  this  point. 
Moreover,  unless  the  contents  of  the  flask  are  diluted  while  still 
warm,  there  is  a  tendency  for  the  sulphates  to  become  hard  and 
difficult  of  solution. 

In  the  modification  proposed,  the  following  reagents  are 
necessary  : 

1.  Chemically  pure  sulphuric  acid. 

2.  Salicylic  acid. 

3.  Potassium  sulphide. 

The  material  containing  the  nitrates  is  weighed  out  into  a 
digestion  flask  and  thirty  cc.  sulphuric  acid  containing  one 
gram  salicylic  acid  are  added,  and  gently  heated  to  facilitate  the 
solution  of  nitrates  and  prevent  frothing  later.  While  warm, 
six  to  seven  grams  of  potassium  sulphide  are  added  in  small 
portions,  the  flask  being  thoroughly  shaken  after  each  addition. 
It  is  then  placed  over  a  low  flame  and  the  heat  rapidly  increased 
until  the  acid  mixture  boils.  No  further  attention  is  required 
and  the  digestion  is  usually  complete  at  the  end  of  an  hour. 
When  cold',  the  liquid  is  diluted  and  distilled  in  the  usual  man- 
ner. 

The  average  difference  between  the  results  on  sixty  samples 
of  fertilizers  containing  nitrates  by  the  official  method  and  the 
proposed  modification  was  0.02  per  cent.,  those  by  the  latter 
being  higher. 

The  points  of  difference  between  the  modification  and  the  offi- 
cial modified  Gunning  are  the  following  : 


II04 


SEPARATION   OP  AI^KALOIDAL  BXTRACTS. 


1.  The  number  of  reagents  used  in  the  digestion  is  reduced 
from  four  to  three. 

2.  Frothing  is  obviated  and  the  operation  requires  no  atten- 
tion except  turning  up  the  lamps  until  full  heat  is  secured. 

3.  The  time  of  digestion  is  shortened. 

4.  Potassium  sulphide  is  made  to  do  double  work  by  acting  as 
a  reducing  agent  instead  of  sodium  thiosulphate  and  then 
being  converted  into  potassium  hydrogen  sulphate  serving  the 
end  secured  by  adding  potassium  sulphate  in  the  original  method. 


T 


THE  SEPARATION  OF  ALKALOIDAL  EXTRACTS. 

By  Charles  Platt. 

Received  October  ao,  i8p6. 

HE  writer  has  found  the  accompanying  simple  device  of 
great  value  in  the  separation  of  the  annoying  emulsions 

so  often  met  with  in  alkaloidal  anal- 
ysis, as,  for  instance,  in  the  petroleum 
ether  and  benzene  extractions  of  Dra- 
gendorff's  method.  The  filtering  tube 
is  nineteen  cm.  long,  the  upper  12.5  cm. 
having  an  inside  diameter  of  fourteen  mm., 
the  lower  contracted  portion,  an  inside  diam- 
eter pf  three  mm.  A  stout  platinum  wire 
bent  at  the  upper  end  is  so  placed  as  to 
pass  through  the  constricted  portion  of  the 
tube  to  the  bottom  of  the  eight-ounce 
Erlenmeyer  flask.  Washed  cotton  is  firmly 
packed  in  the  tube  to  a  depth  of  about 
four  cm.  and  the  apparatus,  connected 
with  a  filter  pump,  is  ready  for  use.  The 
filtered  liquids  may  finally  be  carefully 
poured  into  an  ordinary  separating  funnel 
and  manipulated  as  usual.  By  this  method 
the  most  persistent  emulsions  are  separated 
into  their  constituent  liquids  in  as  many 
minutes  as  ordinarily  are  required  hours  or 
days. 

Chemical  Laboratory.  Hahnemann  Medical  College, 

Philadelphia,  Pa. 


THE  PREPARATION  OP  DIETHYL  HALONIC  ESTER. 

By  W.  a.  Noybs. 
ReceiTed  October  99,  1896. 

HAVING  occasion  recently  to  prepare  considerable  quanti- 
ties of  malonic  ester,  it  has  been  found  that  the  process 
can  be  very  much  shortened  by  the  use  of  sulphuric -in  place  of 
hydrochloric  acid  and  of  acid  sodium  carbonate  in  place  of  potas- 
sium carbonate.  As  the  body  is  the  starting  point  for  a  great 
variety  of  syntheses  the  method  used  may  be  of  interest  to 
others. 

One  hundred  grams  of  chloracetic  acid  are  placed  in  a  porce- 
lain dish,  21  cm.  in  diameter,  and  200  cc.  of  wateradded.  Thesolu- 
tion  is  warmed  and  ninety  grams  of  acid  sodium  carbonate  added 
in  small  portions,  and  the  warming  continued  until  a  temperature 
of  55''-6o**  is  reached  and  effervescence  nearly  ceases.  Eighty 
grams  of  coarsely  powdered  potassium  cyanide  is  then  added, 
and  the  whole  stirred  without  further  warming,  till  the  some- 
what vigorous  reaction  is  complete.  The  solution  is  then  evap- 
orated rapidly  on  a  thin  sheet  of  asbestos  paper  till  the  ther- 
mometer with  which  it  is  vigorously  and  constantly  stirred  shows 
a  temperature  of  i3o''-;35**.  The  hand  should  be  protected  by  a 
glove  or  otherwise,  and  the  glass  of  the  hood  in  which  the 
evaporation  is  conducted  should  be  between  the  dish  and  the 
lace  during  this  part  of  the  process.  The  mass  should  be  stirred 
occasionally  while  cooling,  and  as  soon  as  it  solidifies  it  should 
be  broken  up  coarsely  and  transferred  to  a  liter  flask.  Add 
forty  cc.  of  alcohol  and  connect  with  an  upright  condenser. 
Through  the  latter  add,  in  small  portions  and  with  frequent 
shaking,  a  cooled  mixture  of  160  cc.  of  alcohol  with  i6occ.  of 
concentrated  sulphuric  acid.  The  whole  may  be  added  within 
five  or  ten  minutes  (instead  of  the  day  and  a  half  required  to 
saturate  with  hydrochloric  acid  by  the  old  method).  Toward 
the  close  there  is  a  considerable  evolution  of  hydrochloric  acid. 
Heat  on  a  water-bath  for  an  hour.  Cool  quickly  under  the  tap, 
with  shaking  to  prevent  the  formation  of  a  solid  mass  of  crys- 
tals. Add  200  cc.  of  water,  filter,  wash  the  undissolved  salts 
with  about  fifty  cc.  of  ether,  shake  up  with  the  filtrate  and  sepa- 
rate.    Add  a  solution  of  sodium  carbonate  and  shake  carefully 


II06  NOTB. 

with  the  ethereal  solution  till  alkaline.  Separate  again,  distil 
off  the  ether  and  dry  by  heating  for  fifteen  minutes  on  a  water- 
bath  under  diminished  pressure,  using  a  capillary  tube  as  for 
vacuum  distillations.  The  residue  gives,  after  one  distillation, 
an  almost  pure  malonic  ester. 

The  sodium  carbonate  solution  appears  to  contain  some  of 
the  acid  ester.  If  this  solution  is  added  to  the  first  acid  solu- 
tion, the  ester  separates  with  some  ether.  The  ethereal  solu- 
tion may  be  separated,  the  ether  evaporated  at  a  gentle  heat, 
and  the  residue  added  to  the  contents  of  the  flask  in  which  a 
second  saponification  of  the  cyanacetate  is  to  be  effected.  If 
this  is  done,  a  yield  of  malonic  ester  equal  to  the  weight  of  chlor- 
acetic  acid  used  can  be  obtained.  This  is  ten  to  fifteen  percent, 
better  than  by  the  old  method. 

Rose  Polttschnic  iNSTtruTB,  / 

Tbr&b  Haxttb,  Ind.,  Oct.  37,  X896. 


NOTE. 

Untaxed  Alcohol  for  Use  in  Manufacturing  and  in  the  Arts, — 
The  Joint  Select  Committee,  createdatthe  last  session  of  Congress, 
to  investigate  and  report  upon  the  question  of  the  use  of  alco- 
hol free  of  tax  in  the  manufactures  and  arts,  have  prepared 
a  series  of  interrogatories,  which  will  be  distributed  through- 
out the  country  to  such  parties  as  are  thought  to  be  interested 
in  the  question. 

The  report  of  Mr.  Henry  Dalley,  Jr.,  who  was  commissioned 
to  investigate  the  workings  of  foreign  laws  governing  the  use  of 
untaxed  alcohol  in  the  manufactures  and  arts  has  been  submit- 
ted, and  contains  very  full  and  extremely  valuable  data  covering 
Great  Britain,  Germany,  France,  Belgium  and  Switzerland. 

It  is  the  earnest  desire  of  the  committee  to  secure  all  possible 
information  bearing  upon  the  subject,  and  it  is  hoped  that  par- 
ties interested  will  submit  their  views  to  the  committee  promptly. 
Sets  of  the  circular  letter  and  blank  for  replies  will  be  supplied 
to  any  applicant  by  addressing  the  chairman.  Room  21,  Senate 
Annex,  Washington,  D.  C. 

The  committee,  which  is  composed  of  three  members  of  each 
House,  will  probably  assemble  in  Washington  soon  after  the 


OBITUARY  NOTICE.  IIO7 

middle  of  November  for  the  purpose  of  formulating  a  report  to 
Congress  accompanied  by  the  draft  of  a  law  which  will  place 
domestic  industries  on  as  favorable  a  basis  as  similar  industries 
in  foreign  countries..  During  their  sessions  in  Washington 
hearings  will  probably  be  given  in  order  to  supplement  the 
information  obtained  through  the  interrogatories  above  set 
forth.  Due  notice  of  the  time  of  such  hearings  will  be  given 
to  tibe  public. 


OBITUARY  NOTICE. 

Propbssor  ^ugust  Kbkiti«£'s  part  in  the  advancement  of 
chemistry  has  been  so  important  that  his  death  on  the  13th  of 
last  July  has  brought  a  feeling  of  sorrow  to  the  hearts  of  chemists 
throughout  the  world. 

Kekul6  was  bom  at  Darmstadt,  the  birthplace  of  Liebig,  on 
the  7th  of  September,  1829.  It  was  the  intention  of  his  parents 
that  he  should  become  an  architect,  and  he  entered  the  Univer- 
sity at  Giessen  as  a  student  of  architecture.  He  devoted  him- 
self with  application  to  the  studies  bearing  on  his  future  calling, 
but  like  many  another  student  who  came  within  the  range  of 
I/iebig's  influence,  he  was  filled  with  an  enthusiasm  for  chemis- 
try, which  changed  all  his  plans  for  the  future,  and  led  him  to 
devote  himself  to  this  science.  It  is  quite  possible  that  his  pre- 
liminary architectural  studies  had  much  to  do  with  turning  his 
mind  toward  the  ideas  of  molecular  structure  or  molecular 
architecture,  which  he  subsequently  developed.  Kekul^  also 
studied  in  Paris  under  Dumas,  and  in  London  under  William- 
son. In  1856  he  became  privatdocent  at  the  University  of  Hei- 
delberg. He  was  appointed  professor  of  chemistry  at  the  Uni- 
versity of  Ghent  (Belgium)  in  1858  ;  and  in  1865  was  called  to 
the  University  of  Bonn,  where  he  remained  until  his  death. 

Kekul6'sfirstpublished  work  appeared  in  hiehig^s  Annalaiior 
1850.  Four  years  later  he  published  his  second  paper,  in  which 
he  described  thiacetic  acid  and  discussed  the  action  of  phos- 
phorus pentasulphide  on  oxygen  acids. 

The  period  from  1854  to  1874  was  one  of  the  greatest  activity 
with  Kekul6.     Since  1874  he  has  made  comparatively  few  con- 


II08  OfilTUARY   NOTICE. 

tributions  to  chemistry,  although  occasional  papers  have 
appeared.  In  spite  of  the  great  number  of  investigations  he 
has  made,  chemistry  is  most  indebted  to  Kekul6  for  his  great 
generalizations  and  theoretical  suggestions. 

He  extended  Gerhardt's  t3rpe  theory  by  adding  the  marsh  gas 
tjrpe  and  introducing  the  idea  of  mixed  t3rpes.  These  types 
made  clear 'to  him  the  difference  in  the  power  of  the  elements  to 
hold  other  atoms  in  combination,  and  he  developed  the  idea  of 
valence,  first  put  forward  by  Frankland,  so  that  this  new  prop- 
erty of  the  elements  was  at  once  recognized  by  chemists,  the 
conception  of  atom-linking  followed  at  once,  and  this  made  pos- 
sible the  transition  from  the  type  theory  to  our  present  concep- 
tions in  regard  to  the  structure  of  compounds. 

In  this  paper  published  in  1858  Kekul6  says:  "  It  is  the  sub- 
stitution and  relation  of  the  atoms  and  not  radicals,  that  we  must 
look  to  in  order  to  get  a  clearer  idea  of  the  nature  of  these  com- 
pounds.** 

He  closes  this  remarkable  paper  with  the  following  words : 
'*  In  conclusion  I  believe  that  I  should  emphasize  that  I  do  not 
set  much  value  upon  this  kind  of  speculation.  But  since  chem- 
istry, in  its  entire  lack  of  exact  scientific  principles,  must  content 
itself  for  the  time  with  the  most  probable  and  useful  theories ;  it 
appears  proper  to  present  these  views,  for  they,  as  it  seems  to 
me,  give  a  simple  and  entirely  general  expression  for  the  latest 
discoveries,  and  because  moreover  their  application  may  be 
the  means  of  discovering  new  facts.'* 

It  is  not  too  much  to  say  that  the  ideas  thus  modestly  put  for- 
ward, supported  by  his  subsequent  work,  were  the  prime  cause 
which  led  to  the  abandoning  of  Gerhardt's  types  for  our  present 
structural  formulas. 

These  ideas  had  made  considerable  progress,  when  in  1865 
Kekul6  published  his  now  well  known  h3rpothesis  in  regard  to 
the  constitution  of  benzene.  Seldom  has  a  theory  in  chemistry 
been  so  suggestive  or  given  rise  to  so  much  investigation  as  this 
benzene  theory.  The  rich  and  manifest  results  accruing  from 
these  investigations  testify  sufficiently  to  the  utility  of  the  theory. 

Many  students  of  chemistry  were  attracted  to  Bonn ;  these 
Kekul6  inspired  with  a  love  of  investigation  that  has  been 


NEW  BOOKS.  1 109 

exceedingly  fruitful  for  the  science.  Besides  his  work  as  a 
lecturer  and  investigator,  he  began  in  i860  and  finished  in  1861 
the  first  volume  of  his  Lehrbuch  der  organischen  Chemie,  a 
book  that  was  epoch-making  with  its  new  ideas  and  new  methods 
of  presenting  this  complex  subject.  The  book  was  received 
with  enthusiasm  among  chemists,  and  has  served  as  a  model  for 
subsequent  works  in  the  same  field.  Three  volumes  of  this  work 
were  finally  published,  but  the  work  was  never  completed.  He 
was  also  for  many  years  one  of  the  editors  of  Liebig's  Annalen, 
During  his  last  years  he  suffered  much  from  ill  health,  having 
followed  too  literally  I^iebig's  advice:  **  If  you  would  become 
a  chemist,  you  must  ruin  your  health.  He  who  does  not  ruin 
his  health  by  hard  study  in  these  days  comes  to  naught  in  chem- 
istry." 

In  1890  the  German  Chemical  Society  celebrated  the  twenty- 
fifth  anniversary  of  Kekul^'s  benzene  theory.  The  meeting  was 
largely  attended  by  chemists  from  all  parts  of  the  world. 
Addresses  were  given  by  A.  W.  Hofmann,  the  President  of  the 
Society,  Adolph-von  Baeyer,  Kekulfe's  oldest  pupil,  and  by 
Kekul6  himself.  A  full  account  of  the  meeting  has  been  pub- 
lished.* G.  M.  Richardson. 


Oct.  17, 1896. 


NEW  BOOK5. 

Manual  op  DBTBRMiNATrvrs  Minbralocy  with  an  Introduction  on 
Blowpipb  Analysis.  By  Geor^  J.  Brush.  Revised  and  Enlarged  by 
Samuel  L.  Penfield.  14th  Edition,  pp.  iz  -f~  108.  John  Wiley  & 
Sons.     Price,  I3.50. 

This  revision,  with  the  exception  of  the  tables,  is  practically  a 
new  book.  The  author  states  that  **A  complete  revision  of  the 
tables  for  the  determination  of  minerals  will  be  made  as  soon  as 
possible,  and  a  short  chapter  on  crystallography  and  the  phys- 
ical properties  of  minerals  will  be  prepared,  but  until  this  work 
can  be  accomplished,  use  will  be  made  of  the  tables  and  of  the 
short  introduction  to  them  from  tb^  last  edition  of  Professor 
Brush." 

This  proposed  revision  of  von  Kobell's  table  is  greatly  needed. 
When  it  is  finished  the  book  bids  fair  to  be  as  nearly  perfect  as 
text-books  can  well  be.  The  introductory  chapter  has  been 
rewritten  with  evident  care  and  by  a  practiced  hand,  and  as  it 

1  Ber,  d,  chem.  Get.^  S3, 1265. 


I  no  NBW  BOOKS. 

now  stands  this  edition  is  a  great  improvement  over  preceding 
ones. 

' '  In  preparing  the  introductory  chapters,  great  pains  have  been 
taken  in  the  selection  of  the  tests  for  the  elements.  Many  of 
them  are  performed  by  means  of  the  blowpipe,  but  chemical  tests 
in  the  wet  way  are  recommended  when  it  is  believed  that  they 
are  more  decisive."  To  this  evidence  of  good  common  sense  it 
may  be  added  that  in  several  places. the  author  shows  a  desire 
and  ability  to  make  his  knowledge  of  practical  value.  This  is 
shown,  for  example,  under  gold,  where  careful  directions  are 
given  for  the  detection  of  gold  in  poor  gold  ores  and  the  like, 
first  by  the  use  of  mercury  and  thien  without  mercury.       B.  h. 

Thb  Elbmbnts  op  Chbmistry.  By  Paxil  C.  Prbbr,  Ph.D.  z+284  ppi 
Boston:  Allyn  &  Bacon.    XS95.    Introductory  price,  fi.oo. 

One  feature  in  particular  makes  this  book  especially  worth 
noticing,  and  that  is  its  outright  recognition  of  the  great  import- 
ance of  quantitative  work  in  an  elementary  course  in  chemistry. 
The  recognition  has  been  a  long  time  on  the  way,  and  its  absence 
has  been  a  great  detriment  to  the  chemical  instruction  in  second- 
ary schools. 

It  is  also  pleasant  to  find  Professor  Freer  recognizing  that  cer- 
tain so-called  physical  matters  are  best  reviewed  at  the  outset  of 
such  a  course.  Indeed  it  would  seem  as  if  some  such  matters 
which  are  taken  up  in  the  present  work,  rather  late  in  the  course, 
would  better  be  considered  earlier  (the  laws  of  Mariotte  and 
Charles  for  instance). 

The  book  cannot  be  used  to  advantage  by  an  inadequately 
trained  teacher,  but  will  certainly  be  found  valuable  to  the 
student  teacher  on  account  of  its  excellent  collection  of  experi- 
ments which  are  carefully  planned  and  digested. 

Joseph  Torrby,  Jr. 

Tabi«bs  and  DiRBcnoNS  for  Quautativb  Chbmicai,  Analysis.  B7 
M.  M.  Pattison  Muir. 

This  little  work  is  evidently  intended  to  increase  the  possi- 
bilities of  lecture  table  instruction  in  qualitative  analysis.  It 
consists  of  such  brief  statements  of  processes  and  methods  as  will 
enable  the  student  to  attend  to  what  is  going  on  on  the  lectore 
table  without  running  the  risk  of  losing  material  which  ought  to 
get  into  his  note  book.  The  analytical  methods  described  are, 
for  the  most  part,  such  as  have  stood  the  test  of  time  and  expe- 
rience .  Joseph  Torrey  ,  Jr  . 

The  Liquefaction  of  Gases.  Papers  by  Michabl  Faraday,  F.R.S. 
(1833-1845).  Alembic  Club  Reprints  No.  12.  79  pp.  Bdinbnrgh:  Wm. 
F.  Clay.    Price,  two  shillings. 

In  this  little  book  of  seventy-nine  pages  there  is  much  matter 


CORRBSPONDBNCH.  1 1 1 1 

that  will  be  of  practical  service  to  every  one  who  teaches  ele- 
mentary chemistry.  Its  value  to  investigators  and  advanced 
students  is  sufl&ciently  obvious.  Students  ought  to  be  intro- 
duced to  the  classics  of  chemistry  at  a  comparatively  early  stage 
of  their  development.  They  are  not-  as  a  rule,  at  present, 
because  the  original  papers  are  seldom  accessible  to  the  teacher. 
The  publication  of  Ostwald's  **  Klassiker  "  was  the  first  step  in 
the  right  direction,  but  the  fact  that  they  are  in  German  makes 
them  inaccessible  to  many  who  most  need  them. 

Joseph  Torrey,  Jr. 


CORRESPONDENCE. 


POLARIZATION  BY  DOUBLE  DILUTION. 


United  States  Department  op  Agriculture, 

Division  of  Chemistry, 

Washington,  D.  C,  Nov.  27,  1896. 

Editor  Journal  of  the  American  Chemical  Society  ^  Easton,  Pa, : 

Dear  Sir  :  By  accident  a  portion  of  the  rule  for  calculating 
polarizations  by  double  dilution  in  our  paper  published  in  this 
Journal,  1896,  Vol.  18,  pages  428  to  433,  was  omitted. 

Page  430,  beginning  at  the  end  of  line  9,  the  rule  for  the  ap- 
proximate calculation  of  results  obtained  by  Scheibler's  method 
of  double  dilution  should  have  this  addition  after  the  words 
'*  small  flask,"  **  multiply  the  difference  by  two  and  subtract  the 
product  from  the  reading  in  the  small  flask.*'  This  is  equiva- 
lent to  multiplying  the  reading  obtained  from  the  solution  in  the 
large  flask  by  four  and  subtracting  the  reading  obtained  from 
the  solution  in  the  small  flask  from  the  product.  The  re- 
sult is  the  corrected  reading  and,  when  a  solution  of  double  the 
normal  strength  is  polarized  in  a  tube  of  double  the  normal 
length,  must  be  divided  by  four  to  obtain  the  percentage.  In 
this  case  a  simpler  and  equivalent  rule  for  calculation  is  the  fol- 
lowing :  Subtract  one-fourth  the  reading  of  the  solution  in  the 
small  flask  from  the  reading  in  the  large  flask  and  the  result  will 
be  the  corrected  percentage. 

Page  430,  end  of  line  17,  the  word  sucrose  should  be  lactose. 

Page  432,  the  figures  in  the  table  in  the  column  headed  **Vol- 


1 1 12  BOOKS  RECEIVED. 

ume  of  precipitate/'  were  calculated  before  the  exact  formula  on 
P^S^^  43^  ^^s  evolved,  and  are  somewhat  at  variance  with  the  re- 
sults objjtained  by  use  of  the  formula.  The  formula  gives  the  fol- 
lowing numbers  :  5.26,  10.71,  4.88,  9.86,  5.05,  5.41,  4.53,  4.12, 
3.87,  4.99,  3.33,  4.22,  16.23.  The  numbers  in  the  column 
headed  "  True  volume  in  100  cc.  flask"  must  be  changed  ac- 
cordingly. 

Respectfully, 

H.  W.  Wiley, 

E.  E.  EWELL. 


BOOKS  RECEIVED. 

Bulletin  No.  33.  Commercial  Fertilizers  and  Chemicals,  and  Other  In- 
formation in  Regard  to  Fertilizers.  Under  the  supervision  of  Hon.  R.  T. 
Nesbitt,  Commissioner  of  Agriculture  of  the  State  of  Georgia.  Dr.  George 
P.  Payne,  State  Chemist.  Atlanta,  Ga. :  George  W.  Harrison,  Sttte 
Printer. 

Manual  of  Determinative  Mineralogy,  with  an  Introduction  on  Blow- 
pipe Analysis.  By  George  J.  Brush.  Revised  and  enlarged  by  Samuel 
L.  Penfield.  Fourteenth  Edition.  x4-  108  pp.  New  York :  John  Wiley 
&  Sons.    Price  $3.50. 

Jahrbuch  der  organischen  Chemie.  Heransgegaben  von  Gaetano  Min- 
unni.  Palermo.  Zweiter  Jahrgang.  993  pp.  1894.  Leipzig :  Johann  Am- 
brosius  Barth.     (Arthur  Meiner).     1896. 

A  Brief  Introduction  to  Qualitative  Analysis ;  for  Use  in  Instruction  in 
Chemical  Laboratories.  By  Ludwig  Medicus.  Translated  from  the 
Fourth  and  Fifth  German  Editions  by  John  Marshall.  Fourth  Edition. 
Philadelphia :  Printed  by  J.  B.  Lippincott  Co.  1896.  203  pp.  Price 
I1.50. 

Bulletin  No.  43.  Second  Series.  Bovine  Tuberculosis  in  North  Lou- 
isiana. Bulletin  of  the  Louisiana  State  Experiment  Station,  Baton 
Rouge,  La.     1896.     20  pp. 


ERRATUM. 

On  page  994  (November  number),  seventh  line  from  bottOLi,  instead  of 
'*  extra  internal  pressure'*  read  '*  extra  external  pressure.'* 


Index  to  Vol.  XVIII,  1896. 


References  to  the  pafea  of  the  Proceeding  are  given  in  Perentbetia. 

.CCURACY,  limits  of  in  metallurgical  analyaia.  35;  of  chemical  analyai8....8oS,  (88) 
▲csenaphthene,  heat  of  aolntion  in  methyl  alcohol,  iss  ;  in  ethyl  alcohol,  153 ;  in 

propyl  alcohol,  154;  in  chloroform,  155;  intoluene • 155 

▲cetamlde,  heat  of  solution  in  water,  151 ;  in  ethyl  alcohol 153 

▲cctanilid,  qualitative  examination  of,  143 ;  heat  of  solution  in  methyl  alcohol,  15a ; 

in  ethyl  alcohol,  153 ;  in  chloroform .«••    155 

Acetylene,  use  in  polariscopic  work  as  an  llluminant 179,  (37) 

Acetone,  manufacture  from  acetic  acid,  331,  (30) ;  volumetric  determination  of 106B 

Acid  vapors,  action  of,  on  metallic  sulphides » • 1096 

Address,  changes  of  post  office .*. 

va).  (*6),  (36),  (57)»  (65).  (71).  (M).  (97),  (99),  ("8).  (iX9).(ia4) 

Addresses  wanted,  post  office (2),  (36),  (57).  (70.  (1^4) 

Albuminoid  nitrogen,  source  of  error  in  determina tion  of 464 

Alcohol,  determination  by  means  of  the  ebullioscope,  1063;  the  tax  on 1106 

Alknloldal  extracts,  separation  of iiof 

Alloys  of  aluminum,  analysis  of 77a 

Almonds,  proteidsof 610 

Alumina,  determination  of,  in  phosphate  rock  by  the  ammonium  acetate  method, 

717:  estimation  of ,  in  phosphate  rock.  Tax ;  analysisof 779 

Aluminum,  anal]rsis  of,  766 ;  determination  of,  in  metallic  aluminum,  77a :  solders, 
analysis  of,  777 ;  use  in  cooking  vessels,  935 ;  in  pig  iron,  (3a):  See  also  erratum  (56) 

Amandin 611 

Amines,  occurrence  of,  in  the  juice  of  the  sugar  cane 743 

Ammonia,  the  separation  of  trimethylamine  from 670 

Ammonium  phosphomolybdate,  gravimetric  method  of  estimating  phosphoric  acid 

^••S3;  precipitation  of,  in  steel  analysis 170 

Amphoteric  reaction  of  milk (33) 

AifDnswa,  I^AUifCBLOT.  On  the  Reduction  of  Sulphuric  Acid  by  Copper  as  a  Func- 
tion of  the  Temperature  ^i 

Ahdkbwb,  W.  W.    Some  Sxtensions  of  the  Plaster  of  Paris  Method  in  Blowpipe 

Analysis 849 

Anhalonium  hydrochlorste,  64a ;  see  also (83) 

Antidiphtheritic  serum,  method  of  collecting 930 

Anti-friction  alloys,  determination  of  bismuth  in 683 

Antimony  cinnabar,  formation  of g 34a 

Antimony,  Reinsch's  test  for,  953 ;  trioxide,  behavior  with  hydrochloric  acid,  103a ; 
separation  from  lead  by  means  of  hydrochloric  acid,  X033  ;    separation  from 

copper  by  means  of  hydrochloric  acid 1035 

Argon ,  atomic  weight  of • ax  i 

Arsenic,  Reinsch's  test  for,  953 ;  separation  from  copper  by  means  of  hydrochloric 
acid,  XQ38 ;  separation  from  silver  by  means  of  hydrochloric  acid,  X039 ;  separa- 
tion from  cadmium  by  means  of  hydrochloric  acid,  1039 ;  separation  from  iron 
by  means  of  hydrochloric  acid,  X040 ;  separation  from  sine  by  means  of  hydro- 
chloric acid,  XQ41 :  separation  from  cobalt  and  nickel  by  means  of  hydrochloric 

acid,  104a  ;  atomic  weight  of,  X044 :  separation  of  vana<ttum  from lasx 

Asplialtum,'technical  analysis  of 97$ 

AsMidates  elected (i),  (a),  (a6),  (35).  (36),  (^),  (71),  (93),  (xx9).  (xa4) 

Atomic  masses  of  silver,  mercury  and  cadmium,  determination  of,  by  the  electro- 
lytic method 990 

Atomic  weights,  report  of  committee  on,  X97 ;  of  nitrogen  and  arsenic 1044 


1 1 14  INDBX. 

AVCBT,  Obokob.  The  Prcelpitetioii  of  Phosphomoiybdftte  in  Steel  Anftlysis,  170 ; 
Drown's  Method  of  Detemininff  Snlphiir  in  Tig-iron,  40S:  Sonrces  of  Error  in 
Volhard't  and  Similar  Methods  of  Determininir  Manganese  in  Steel,  49^ ;  Notes 
on  the  Determination  of  Phosphoms  in  Steel  and  CasUron • 9SS 

AUSTBN,  Pbtbr  TowMSBiiD.  A  New  Specimen  Bottle  (ttstc) ••••• 4M> 

AVBBT,  S.   See  Nicholson,  B.  H. 

IBACTBItlA,  study  of  sas-prodncittff,  tS7  :  inndtksogar • •..dS7,  (73) 

Babb,  S.  H.  AiTD  A.  B.  Frescott  Dipjrridinc  Methylene  Iodide  and  the  Non-Foraan* 
tion  of  the  Corresponding  Monopyridine  Frodncts,  9B8 ;  see  also  Presoott,  A.  B. 

Balance,  for  first  years  work  in  general  chemistry  (note) ••••••••.••. rty 

Barium  sulphate,  the  effect  of  an  excess  of  reagent  in  the  precipitation  ci  •»•»••*»••    799 
Basxbbviixb,  Cbaklbs.    Reduction  of  Concentrated  Sulphuric  Acid  by  Copper  •  • .    94s 

Bauxite,  analysis  of • tAs 

Bbadlb  C.   See  Cross,  C.  P. 

Beef  Pat,  microscopic  detection  in  lard •• •••••••.•    189 

Bbbson,  J.  L.  Occurrence  of  the  Amines  in  the  Juice  of  the  Sugar  Cane,  743 ;  A  Sina- 

ple  and  Convenient  Bxtraction  Apparatus  for  Pood-stuff  Analysis •••• 744 

Bbitkbrt,  Abthub  It.  and  Smith,  BdgarP.    The  Separation  of  Bismuth  from  Vemd  iaS9 
BsjfNBTT,  A.  A.  and  S.  B.  Pammel.    A  Study  of  Some  Gus-Produdng  Bacteria.  •  ••  •  •    157 
Bbniibtt,  a.  a.  AMD  h.  A.  Placeway.    The  QuantltatiTe  Determination  of  the  Three 
Halogens,  Chlorine,  Bromine  and  Iodine,  in  Mixtures  of  Their  Binary  Goos- 

pounds.** • • 6M 

Bensamidc,  heat  of  solution  in  ethyl  alcohol • • is# 

Bensoin,  action  of  acid  amides  upon,  mi ;  action  of  urea  upon,  118 ;   action  of 

thiourea  upon • • 119 

Berthelot's  contributions  to  the  history  of  chemistry  (review) 466 

Bbyah,  B.  J.  See  Cross  C.  P. 

BiGBLow,  W.  D.    Index  to  the  Literature  on  the  Detection  and  Bstimatlon  of  Pusel 

OilinSpiriU 3Sff 

Birch  Wood  Gum siS 

Bismuth,  estimation  of,  in  anti-friction  allojrs,  6B3 ;  oxide,  behaTior  of,  with  hydro- 
chloric add,  X033  ;  separation  from  lead  by  means  of  hydrochloric  acid,  1034 ; 
separation  from  copper  by  means  of  hydrochloric  acid,  1036 ;  separation  front 
lead,  IOS5 ;  sulphide,  solubility  of,  in  sodium  sulphide,  683 ;  sulphide,  solubility 

of,  in  alkaline  sulphides • •• • •  1091 

Blaix,  A.  A.    Method  for  the  Detemlination  of  Carbon  in  Steel n$ 

Blowpipe  analysis,  some  extensions  of  the  plaster  of  Paris  method  in..... 849 

Boiler  scale,  the  presence  of  oil  in • 741 

BOLTOH,  H.  Cabbimoton.    Beitbelot's  Contributions  to  the  History  of  Chemistry 

(review) .486.  (3^ 

Book  Reviews.  The  Scientific  Foundations  pf  Analytical  Chemistry  Treated  In  an 
Blementary  Msnner  (Ostwald),  98;  Practlcsl  Proofs  of  Chemical  I«aws  (Comlah), 
99 ;  Organic  Chemistry.  The  Fatty  Compounds  (Whiteley),  99;  Analjrtical  Chem^ 
irtry  (N.  Menschtttkin),  190 ;  On  the  Densities  of  Oxygen  and  Hydrogen  and 
on  the  Ratio  of  their  Atomic  Weights  (B.  W.  Morley),  19a;  Water  Supply,  Con- 
sidered Principally  from  a  Sanitary  Stand-point  (W.  P.  Mason),  gSa ;  Hints  on  the 
Teaching  of  Blementary  Chemistry  in  Schools  and  Science  Classes  (W.  A.  TO- 
den),  s^;  Chemistry  lor  Bngineersand  Manufacturers  (Betram  Blount),  745; 
Laboratory  Bxperiments  in  General  Chemistry  (C.  R.  Sanger),  747;  Bxpcrimeata 
in  General  Chemistry  and  QuaUlatlve  Analyiis  (C  R.  Sanger),  747 ;  Manual  off 
Determinative  Mineralogy  with  an  Introduction  on  Blowpipe  Analysis  (Bmah- 
Penfield),  1109;  The  Blements  of  Chemistry  (P.  C  Freer),  ixio;  Tables  and  I>lree> 
tions  for  Qualitative  Analysis  (M.  M.  Pattison  Mulr),  ixxo ;  The  Liquefactioii  of 

Gases  (Michael  Faraday) txio 

Books  received 100,196^3x0,4x4,474,964,660,747.848,1007.  iiis 


INDBX.  ZII5 

plMsphate,  distiiBclion  from  rock  pho^bate..... 491 

BoTAcic  acid,  acidity  of  milk  increased  by • • 847 

Bottle  for  tpecimena 4xj 

Bimall-attt,  protdda  of  the • 6ai 

Bromiae,  iodlsie  and  chlorine,  indirect  determination  of,  1815;  qnantitatiTc  determina- 
tion of,  in  miztnrea  of  the  binary  compoonda  of  the  halogena  • ••...  668 

BmYAHT,  A.  P.  A  Method  for  BeparaUnir  the  Iniolnble  Phoaphoric  Acid  in  Mixed 
FertiUaers  Derived  from  Bone  and  Other  Organic  Matter  from  that  Derived  from 

Sock  Phoaphatc 491 

Bvas,  CBAmLBS  S.    Petiolettm  Prodncta (68) 

Bynedeatin • 54a 

Bjnin .^.. 55a 

Bwtter,  vac  of  calorimeter  in  detecting  adulteratloa  of 174 

OACTACBAB,  the  chemistry  of  the 694,  (83) 

Cadminm,  atomic  weight  of,  aqs;  preparation  of  pure,  loai;  separation  from  arsenic 
by  means  of  hydrochloric  add,  1039 ;  bromide,  preparation  of  pnre,  loaj ;  chlo> 

ride,  preparation  of  pure • ion 

Caffeine,  new  method  for  estimation  of,  331 ;  perhalides  of,  347  ;  estimation  of 978 

Caldnm  cartiide,  manufacture  of ,  3x1 :  eatimatloa  of  aulphidea  in 740 

Caldnm  pboaphide,  preparation  of .^ (s8) 

Calorimeter,  uae  ol,  in  detecting  adulteration  of  butter  and  lard 174 

CaMPBSix,  B.  D.    A  Proposed  Schedule  of  Allowable  Difference  and  of  Probable 

limita  of  Accuracy  In  Quantitative  Analyses  of  Metallurgical  Materials 35 

CAMraxu.,  GBonOB  P.   See  Oabome,  Thomaa  B. 

Canned  gooda,  gaseain ^ 936 

Carbon,  atomic  weight  of ,  sis ;  determination  of,  in  steel,  233  ;  determination  of.  in 
aluminum,  771 ;  dctcrminationa  in  pig  iron,  10B7 :  dloodde,  determination  by  ab- 
aorption,  x ;  <ttDKide,  produced  by  growth  of  bacteria,  157 ;  monoidde,  cooatittt- 

tioaof 386 

Caxbonatea,  determination  of  cartMm  dioxide  in • i 

Caatorbean,  protdda  of  the • 6ai 

Cerium,  atomic  wdghtof sio 

Cedum  fluoride • • 57 

Chemical  Club  in  New  York (71) 

CHBmnrr,-V.  K.    Pdson  ivy  as  a  Skin  Irritant (8z) 

Chicago  Section.    See  Meetlnga  of  the  Sodcty. 

Chloral  hydrate,  heat  of  solution  In  water,  150;  in  ethyl  alcohol,  153 ;  in  doroform, 

135;  In  toluene... • 156 

Chlorine,  bromine  and  iodine.  Indirect  determination  of,  iSiK  quantitative  determi- 
nation of « In  mixtures  of  the  binary  compounds  of  the  halogens 688 

Chloroform  from  acetone  made  fromacetic  add S31 

Chromium,  volumetric  detcrminatloa  by  means  of  hydrogen  dioxide 9x8 

dndnnati  Sectioii,    See  Meetings  of  the  Sodety. 

dimabar,  electrolytic  determination  of  mercury  in 9& 

dtrate  aolutlon  uaed  in  analysis  of  fertilisers,  method  of  determining  neutrality  of   437 

CukMMM,  P.  W.    Third  Annual 'Report  of  Committee  on  Atomic  Weights.. ...•• 197 

Clasxb,  Thomai  W.   Ace  Venable,  P.  P. 

Coal-tar  «afors,  behavior  toward  the  process  of  digestion X099 

Ooaras,  Chablbs  B.,  mud  W.  &.  Dodson.    Nitrogen  Asdmllatlonln  the  Cotton  Plant   4as 

Cobalt,  atomic  weight  of sos 

OobaU  and  nickel,  separation  from  arsenic  by  meaiu  of  hydrochloric  add •  xop 

Ooooanut,  protddsof  the 6az 

OoLUBK,  Pbtbr.    Obitnaiy  note • 748 

Odor,  measurement  of,  In  imtural  waters,  364 ;  of  potable  water,  valuation  of,  484; 
photography,  Lippmaim's  work  on 9S9«93S 


tll6  INDBX. 


COlorlniT  BUitter  of  tMtunl  water* < «.•..«••...«• ,4*4A,{'jo) 

Columlyiiiiii,  reactionsofoaddes  with  carbon  tetrachloride • «•>••   S3> 

Colninbiiiiii  and  tantalom,  derivatiTeaof •••-•     af 

Compreaaed  sates,  noteaon •••..« •••«....«.••..  (31) 

COKB,  BvwtN  P.   The  Batimation  of  Pyrrhotltc  in  Pyrites  Orea « m 

Conirlotin • «• 609 

Congrreaa  of  applied  chemiatr7,aeoondinteniatioaaU...«« «•••« ,307,660.99 

Copper,  assay  by  iodide  method,  458 ;  accuracy  of  determinatioa  of,  814 ;  aa  redvcer 
of  stronir  sttlphnric  acid,  94a;  aeparation  from  antimony  by  meana  of  hydrochloric 
acid,  103s ;  aeparation  from  biamnth  by  meana  of  hydrochloric  acid,  1096 ;   aepa> 

ration  from  arsenic  by  meana  of  hydrochloric  acid—*.... ••.....«..•••.••••  n^ 

Correspondence ••>  ....... 660, 84B,ixn 

Corrosion  of  electric  mains « •...  (9) 

Cofylin • > • icf 

Council,  minntes  of  the..... (35)*  (S7).  (^).  (70.  (97).  ("9).  («^) 

Caoea,  C.  P.,  B.  J.  Bevan  and  C.  Beadle.    The  Katural  0<ycellu1oses. ......  ^ 8 

Cttpricoacide,  behavior  of  with  hydrochloric  acid «.«.....«...... »q} 


Is  CHAIrMOT,  G.    See  Morehead,  J.  T. 

DsFRior,  GBoaOB.  The  Determination  of  Reducing  Sugars  in  Terma  of  Cnprlc  Ox- 
ide,  749 ;  aee  alao  Bolfe,  Geo.  W. 

DBLAFOifTAiKB,  M.  and  C.  B.  Linebarger.  On  the  Reaction  between  Carbon  Tet- 
rachloride and  the  Ozidea  of  Niobium  and  Tantalum.......... 531 

Dbnnxs,  I..  M.  The  Separation  of  Thorium  from  the  Other  Rare  Bartha  by  Means 
of  Potassium  Trinitride 9(7 

Dbnwts,  L.  M.'andA.  B.  Spencer.    Zirconium  Tetraiodide 673 

DsNKia,  L.  M.  and  Martha  Doan,  with  cryatailographic  notea  by  A.  C  Gill.  Some 
Compounds  of  Thallium.... 979 

Db  ScHWBiifiTZ,  B.  A.  and  Marion  Doraett.  Purther  Notea  upon  the  Pats  con- 
tained in  the  Tuberculosis  Bacilli.... • 449 

Db  Schwbxmits,  B.  A.  and  James  A.  Bmery.  The  Use  of  the  Calorimeter  in  Detect- 
ing Adulterations  of  Butter  and  Lard 171 

Deuteroprotcose *• #..... ..••.,.. 557 

Dbwbt,  Pbbdbbic  p.  The  History  of  Blectric  Heating  Applied  to  Metallurgy  (Re- 
view), 367,  (37) ;  The  Sulphuric  Acid  Process  of  Refining  Liadvlaticm  Sulphides, 
^«  (76)  ;.The  Actual  Accuracy  of  Chemical  Analysis «.Ai8,  (81) 

Diastase,  the  chemical  nature  of ••••.. « s3^ 

Diethyl  malonic  ester,  preparation  of ins 

Digestion,  behaviorof  coal-tar  colors  toward.... 1091 

Dipyridine  methylene  iodide • 9B8 

Dipyridine  trimethylene  dibromide..... • jB 

Directora,  minutes  of  the.Board  of • (i),  (15) 

Doan,  Martha.    See  Dennis,  T..  M. 

Dox>aoN,  W.  R.    See  Coates,  Charles  B. 

DoRBMUS,  Chablbs  A.    Notc  on  the  Presence  of  Oil  In  Boiler  Scalc >...••  • 741 

D0R8BTT,  Marion.    See' DeSchweinits  B- A. 

Drying  oven,  steam  jacketed •....• , (58) 

Dubois,  H.  W.    See  Mixer,  C.  T. 

DuDLBT,  William  L.    Nickelo-Nickelic  Hydrate...... .......* 901 


IBULLIOSCOPB,  a  modified  form  of le^ 

Editor's  report « (18) 

Blectric  furnace,  Moissan's • «.     934 

Electric  heating  applied  to  metallurgy..... 387,  (37) 

Electroljrtic  determination  of  iron,  nickel  and  sine... fs4 

Electrolytic  stand,a  cheap  adjustable » 33$ 

£lx.m8,J.  W.    See  Richards,  Ellen  H. 


INDBX.  1 1 17 

\  jAitBS  A.    See  De  Schweinlts,  B.  A. 

Vrrato  .•••• • • 414,  ma 

HtidorplM » (66) 

B^TBix,  B&viif  B.    Tbe  Chemistry  of  the  Cactaceae •• 634,  (83) 

BwxLi.,  B.  B.  and  H.  W.  Wiley.    The  Bffect  of  Acidity  on  the  Development  of  the 

Mitzifyinf  Organisms.  47s ;  See  also  Wiley  H.  W. 
Hxcnrslons.  to  the  works  of  the  Orasselll  Chemical  Co.,  (24);  To  works  of  Otis  Steel 
Co.,  COntinenUl  Chemical  Co.,  Cleveland  Nitrous  Oxide  Co.,  (46);  To  works  of 
Cleveland  Varnish  Co.,  Cleveland  Rubber  Co.,  Glidden  Varnish  Co.,  and  War- 
ner and  Swasey,(47) ;  To  petroleum  works  of  Scofield.  Shurmers  and  Teasel, 
(51)  ;  to  Crescent  Sheet  and  Tin  Plate  Co's  works,  Bmma  Blast  Pumace,  Rolling 
Mills  0f  Union  Rolling  Mill  Co.,  Cleveland  Rolling  Mill  Co.,  (53);  To  Case  School 
of  Applied  Science  and  Adelbert  College,(54);  to  Lorain  Steel  Works,  (54);  R  ;cep- 

tion  by  Cleveland  Chamber  of  Commerce (55) 

B3Ktracti<m  apparatus  for  food-stuff  analysis 744 

^•ACTORS,  table  of 903 

pA&mnfOTON,  B.  H.    Acidity  of  Milk  Increased  by  Boracic  Acid 847 

Pat,  determination  of  solid  fat  in  artificial  mixtures  of  vegetable  and  animal  fata  and 

oils,  9S9*  (5)  ;  contained  in  bacillus  tul>erculo sis 449 

Pata,  see  oils. 

Perric  oxide,  action  of  hydrochloric  add  gas  on 1040 

Perrocyanides  of  sine  and  manganese iioo 

Pcrtilisers.  simple  method  for  determining  the  neutrality  of  ammonium  citrate  solu- 

tion,457;  distinction  between  bone  and  rock  phosphate > 491 

PIber  determinations  in  bagasse,  source  of  ertorin 469 

PiXLD,  CBAKLBa  AND  Bdoar  P.  SMITH.  The  Separation  of  Vanadium  from  Arsenic  1051 

PiBi^Dft,  JOBir.    A  Modification  of  the  Gunning  Method  for  Nitrates 1x03 

Pllbert,  proteids  of  the 618 

PmsMAir,  P.  A.    New  Mode  of  Pormadon  of  Tertiary  and  Quaternary  Phosphines.  •  (89) 

Pi^RCK,  Hmit ANN.    The  Seps  ration  of  Trimethylamine  from  Ammonia 670 

Plxntx&mann,  R.  p.  and  A.  B.  Prescott  Dipyridine  TrimethyUne  Dibromide.and  a 

Study  of  Certain  Additive  Reactions  of  Organic  Bases a8 

Pood-stuff  analysis,  extraction  apparatus  for..... : 744 

Poods,  experlmcntaon  calorimetric  value  of '(63^ 

POULK,  C  W.    The  Bffect  of  an  Bxcess  of  Reagent  in  the  Precipitation  of  Barium 

Sulphate 793 

Pnsel  oil  in  spirita,  index  to  literature  on  the  detection  and  estimation  of 397 


-AS  pipette  for  absorption  of  illuminanta.  67 ;  regulator,  simple  form  of.  511 :  gen- 
erators, some  new  forms  of .-.. 1057 

Gases  produced  by  growth  of  bacteria....* • 157 

Gill,  Auoustub  H.    ^u  Improved  Gas  Pipette  for  the  Absorption  of  Illuminanta, 

67 ;  see  also  Dennis,  L.  M. 
Gill,  Auoustus  H.  and  H.  A.  Richardson.      Notes  upon  the  Determination  of  Ni- 
trites in  Potable  Water » • at 

Gladding,  Tbom as  S.    A  Gravimetric  Method  of  Bstimating  Phosphoric  Acid  as 
Ammonium  Phosphomolybdate,  23 ;    Note  on  the  Microscopic  Detection  of  Beef 
Pat  in  I«ard,  189 ;   On  the  Estimation  of  Sulphur  in  Pyrites,  446 ;    Determination 
of  Iron  Oxide  and  Alumina  in  Phosphate  Rock  by  the  Ammonium  Acetate 
Method,  717;  A  New  Method  lor  the  Bstimation  of  Iron  Oxide  and  Alumina  in 

Phosphate  Rock • • 7ai 

Glasbr,  CHARLB8.    Estimation  of  Thorla.    Chemical  Analysis  of  Monasite  Sand. .    78a 

Glncinum ,  notes  on  pre  pa  ration  of ( 104) 

Glucose  and  other  invert  sugars,  indirect  determination  by  means  of  hydrogen  di- 
oxide   * 9« 

Glue  solntionsi  specific  gravity  of. .,.•.  (61) 


IIl8  INDBX. 

Gold,  assay  by  cyanide  process,  309:  sccarscy  of  determiiiatioii  of,  in  copper  mate- 
rials, 816  :  Cassel-Hinman  bromine  process  for  extraction  of 051,  C^) 

GoicBBftO,  M.  On  the  Action  of  Wagner's  Reagent  upon  Caffeine  and  a  New  Method 
for  the  Bstimation  of  Caffeine,  331 ;  Perhalides  of  Caffeine.  331 ;  New  POrm  of 
Potash  Bnlb 941 

Gunning  method  for  nitrates,  modification  of  • irot 


[AI^IDBS  of  platinum  and  potassium 13s 

Halogens,  quantitative  determination  of,  in  mixtures  of  their  binary  compounds. .  -  -  6B8 

Hakdt,  jAMssOns.   Aluminum  Analysis  •■ 766 

Hardin,  Willbtt  Lbplbt.  Determination  of  the  Atomic  Masses  of  Silver,  Mer- 
cury, and  Cadmium  by  the  Electrolytic  Method 990 

H  AftT,  Bdward.  The  Valency  of  Oxygen  and  the  Structure  of  Compounds  Contain- 
ing It  (review),  283;  Notes  on  the  Preparation  of  Olucinum (104) 

Hasel-nut,  proteids  of  the * 61S 

Hacbn,  Allbm .    The  Measurement  of  the  Colors  of  Natu ral  Waters a6« 

Hbath,  G.  h.    A  Cheap  Adjustable  Blectrolytic  Stand $59 

Heats  of  solution  of  carbon  compounds 14S 

Heat  of  bromination  of  oils (38) 

HuDBHHAiit,  H.    On  the  DeterminAtion  of  Carbon  Dioxide  by  Absorption..... i 

Helium,  atomic  weight  of..... sn 

Hepthyl  thiocyanaie • (7S) 

Hbrtt,  Chablbs  H.    Mixed  Double  Halides  of  Platinum  and  Potassium 130 

Hbbtt,  Cha&lbs  H.  and  J.  6.  Smith.   Mercuric  Chlorothiocyanate 906 

Heteroproteose S57 

H1BB8,  JOBBP^  GiixiNOHAJC    The  Atomic  Weights  of  Nitrogen  and  Arsenic... 1044 

HiBDS,  J.  I.  D.    Photometric  Method  for  the  Quantitative  Determination  of  Lime 

and  Sulphuric  Acid • tfi 

Honey,  estimation  of  levulose  in 81,  189 

Hopkins,  Ctbil  G.    A  New  Safety  Distillation  Tube  for  Rapid  Work  in  Nitrogen 

Determinations vf 

HowB,  JAS.  Lbwib.  Contributions  to  the  Knowledgeof  Ruthenocyanides $Bi 

HowB,  JA8.  LBW18  and  Paul  S.  Mertins.    Notes  on  Reinsch's  Test  for  Arsenic  and 

Antimony i 9S3 

Hydrochloric  add  gas,  metal  separations  by  means  of,  1099 :  action  on  sodium  py- 

roarsenate,  1037 ;  on  ferric  oxide,  1040 :  behavior  of  minerals  in 1Q13 

Hydrocyanic  acid,  preparation  of • • ..••• 1009 

Hydrofluoric  acid 415 

Hydrogen,  produced  by  growth  of  bacteria,  157 ;   density  of.  198 ;  dioxide,  structure 

of ,  a83 ;   dioxide,  some  analytical  methods  involving  the  use  of 918 

Hydrolysis  of  starch  by  acids,  an  analjrtical  investigation  of 869 

XLI/UMINANTS,  gas  pipette  for  absorption  of 67 

Indexing  chemical  literature,  report  of  committee  on 940 

Indirect  analysis tSt 

Inversion  of  sugar  by  salts 110 

Iodine,  bromine,  and  chlorine,  indirect  determination  of,  185  :  quantitative  determi- 
nation of,  in  mixtures  of  the  binary  compounds  of  the  halogens 688 

Iron,  electroljrtic  determination  of,  ^ ;  separation  from  arsenic  by  means  of  hydro- 
chloric add,  1040;  oxide,  determination  of,  in  phosphate  rotk  by  the  ammonium 
acetate  method,  717 ;  oxide,  estimation  of,  in  phosphate  rock,  721 ;   volumetric 

determination  by  means  of  hydrogen  dioxide 9rt 

Ironand  steel  analysis,  present  status  of.. (iB) 

JOHNSON,  S.W.    Composition  of  Wood  Gum U4 


\t  composition  of  American 9^9 

KbkulA,  August.    Obituary  notice luif 


INDBX:  1 1 19 

KxixsT.jBftoitx,  J&.AiidSiiiitli,BdgarP.   The  Action  of  Add  Vapors  on  McUlUc 

SulphM^t X096 

lKiJOO%,  Otis  T.    8«e  Mabeiy,  Charles  P. 

2C« Across  In  milk,  determination  by  doable  dilation  and  polarisatiim .......  438,  x iix 

Xjunns,  Bdwakd  K.    Indirect  Analysis... «• * x8s 

X»SLirB,N.  J.    Determination  of  Snlpharic  Acid... 683 

Lard,  nse  of  calorimeter  in  detectin^f  adulteration  of,  174 ;  microscopic  detection  of 

beef  fat  in , 189 

X«ead,  Tolnmetrfe  estimation  of,  737 ;  separation  from  antimony  by  means  of  hydro- 
chloric acid,  103^;  separation  from  bismuth  by  means  of  hydrochloric  acid,  1034 ; 

separation  of  bismuth  fiom,  xoss;  oxide,  t>ehaTior  of,  with  hydrochloric  acid...  1031 

IliCnther,  the  valueof  refuse 565 

I,»«Da.  AIpSbrt  R.    Standard  Prisms  in  Water  Analysis  and  the  Valuation  of  Color 

in  PoUble  Waters,  4A4;  Bacteria  in  Milk  Sugar 687 

Leirvmin  of  the  pea  and  vetch • 58I3 

Lehigh  Valley  Section.    See  Meetings  of  the  Society. 
tpBirBBR,  VicTon.    See  Rising,  W.  B. 

Lcucoain 547 

X«CTulose.  estimation  in  honey,  etc • ••*.. 8t,  189 

librarian's  report • (15) 

Ume,  quantitatiTe  determination  of ,  by  photometric  method 661 

LiXDSXT,  J.  B.    The  Value  of  Leather  Refuse • 565 

L1VXBAKOB&,  C.  B.    A  Rapid  Method  of  Determining  the  Molecular  Masses  of 

Liquids  by  Means  of  their  Surface  Tensions,  5x4 ;  see  also  Delafontaine,  M. 

Lnraoir,  Laura  a.    Technical  Analysis  of  Asphaltum.    No.  2 37S 

Liquids^  some  thoughts  about 734 

Litmus  paper,  best  method  of  using - (6o| 

LoifO,  J.  R.  On  the  Inversion  of  Sugar  by  Salts,  xao ;  On  the  Inversion  of  Sugar  by 

Salts,  No.  a,   693 ;  On  the  Ponnation  of  Antimony  Cinnabar • 34s 

Lonx>,  N.  W.    A  Simple  Method  for  Determining  the  Neutrality  of  the  Ammonium 

Citrate  Solution  Used  in  the  Analysis  of  Ptrtilisers • • 457 

Low,  AiABKT,  H.    The  Copper  Assay  by  the  Iodide  Method 458 

Ll7iiOB,0.    On  the  Bstimation  of  Sulphur  in  Pyrites 68s 

Lupin,  pioteids of • 6ea 

ACabSRY,  CHARLES  P.   Lecture  on  Petroleum,  (ao) :  See  also  Robinson.  A.  B. 

MABsnT,  Charlbs  p.  and  Otis  T.  Kloos.  Composition  of  American  Kaolins 909 

Magenta%  effect  of,  on  peptic  digestion,  X094;  on  pancreatic  digestion « 1095 

Malt,  the  globulin  of,  543 ;  the  proteids  of,  541 ;  the  albumin  of,  547 ;  the  pioteoses  of     549 
Manganese,  volumetric  determination  of,  338,  (30);  SftmstrOms  method  for  determin- 
ing manganese  in  iron  ores.  385  ;  separation  from  tungstic  scid.  1053 ;  ferrto- 
cyanide;  xxoo;  in  steel,  sources  of  error  in  Volhard's  method  for  determination 

of,  498;  in  steel,  accuracy  of  determination «• 808 

Mannan • * •• 219 

Manidte,  heat  of  solution  in  water • 151 

Mabon,  W.  p.    Chemical  vs.  Bacteriological  Examination  of  Water 166 

MARTixf,  WixxiAM  J.,  Jr.  The  Cyanide  Method  of  Extracting  Gold  from  its  Ores. 
Application  to  the  Assays  of  Ores  ^x>r  in  Gold  and  Silver.  Preliminary  Notice 

(note) 309 

Mathrws.  J.  A.  Phthalimid,  679;  See  Miller,  Edmund  H. 

MclLHiUBY,  Parkbr  C    The  Cassel-Hinman  Gold  and  Bromine  Process .451.  (61) 

Meetings  of  the  Society,  (7).  (loi):  of  the  Rhode  Island  Section,  (33),  (43).  (64},  (68). 
(99).  (i«».  ("5)  :  of  the  Cincinnati  Section, (5). (33).  (40),  (60).  (68),  (7a),  (93).  (xs6)  ; 
of  the  New  York  Section,  (a).  (a8),  (37),  (4a),  (61),  (69).  (7a),  (95).  (lao),  (las);  of 
the  Washington  8ectlon.(a7).  (36),  (38),  (57).  (66),  (75);  of  the  Lehigh  Valley  Sec- 


II20  INDEX. 

tioiif  (39) f  (45) ;  of  the  Nebruka  Section  (33),  (41).  (68).  (94)  ;  of  the  CldcagD  Sec- 
tion, (43);    of  the  North  Carolina  Section (60),  (97} 

Members  elected (x).  (25).  (a6),  (35).  (6s),  (7x),  f93).("9h  ("4) 

Mercuric,  chlorothiocsranate,  906 ;  oxide,  pre]>a ration  of  pure,  1004;  bromide,  prepa- 
ration of  pure,  1006 ;  cyanide,  preparation  of • iwo 

Mercury,  electrolytic  determination  in  cinnabar,  96  ;  electrolytic  determination  of, 

169;  determination  of  atomic  maMof 990 

MBarnts,  Paul  S.    See  Howe,  Jai.  I^wii. 

Meta-cerium • sit 

Metal  separationi  by  mean*  of  hjrdrochloric  acid^t 1009 

Metallic  sulphides,  the  action  of  acid  vapors  on 1096 

Metallurgy,  electric  heating  applied  to. 387 

Metaphosphinic  acids (90} 

Methyl  orange,  effect  on  pancreatic  digestion 1096 

Milk,  the  acidity  of,  increased  by  boracic  acid,  847 ;  amphoteric  reaction  of,  (53);  in- 
vestigation of (73) 

Milk  sugar,  bacteria  in « 6S7 

MiLLBR,  Bdmttnd  H.    Notes  on  the  Perrocyanldes  of  Zinc  and  Manganese xxoo 

MiLLsa,  Bdmund  H.  and  J.  a.  Mathews.    Table  of  Factors •••  goi 

Mineral  waters,  composition  of  certain  northwestemPennsylvania 9^5 

MiZBR,  C.  T.  and  H.  W.  DuBois.     S&mstr&m's  Method  of  Determining  Manganese 

in  Iron  Ores 3^5 

Molasses,  composition  of 937 

Molecular  masses  of  liquids,  determination  by  means  of  surface  tension SM 

Molybdate  solution,  a  modified 445 

Molybdenum,  atomic  weight  of 907 

Monasite  sand,  chemical  analjrsis  of 781 

MoKBHBAi>,  J.  T.  and  G.  de  Chalmot.    The  Manufacture  of  Calcium  Carbide 3" 

MOYBR,  J.  BniD.    Metal  Separations  by  Means  of  Hydrochloric  Acid  Gas xm9 

Muif BOB,  Chablbs  B.    On  the  Development  of  Smokeless  Powder  (review),  819;  (66) 

SlfAPHTHALHNBt  heat  of  solution  in  methyl  alcohol,  xsa;  in  ethyl  alcohol,  153; 
in  propyl  alcohol,  154 ;  in  chloroform,  155  :  in  toluene,  156 ;  synthetic  prepara- 
tion of (3) 

Nebraska  Section.    See  Meetings  of  the  Society. 
New  Books,  98, 190,  569.  745, 1x09.    See  Book  Reviews. 
New  York  Section     See  Meetings  of  the  Society. 

Nickel,  electrolytic  determination  of ^ 

Nickel  and  Cobalt  separation  from  arsenic  by  means  of  hydrochloric  acid i04> 

Nickelo-nickelic  hydrate 9QX 

NxcHOLSON,  H.  H.  and  S.  Avery.    Notes  on  the  Blectroljrtic  Determination  of  Iron. 

Nickel  and  Zinc -. ^ 

Niobium,  see  columbium. 

Nitrates,  modification  of  Gunning  method  for xna 

Nitric  acid,  preparation  of  pure 993 

Nitrifying  organisms,  effect  of  acidity  on  the  development  of « 475 

Nitrites  in  natural  water,  determination  of n 

Nitrogen  assimilation  in  the  cotton  plant,  425 ;  determination  of,  in  aluminum,  771 ; 

atomic  weight  of , 1044 

Nitrogen  determination,  safety  distillation  tube  for  rapid t*J 

North  Carolina  Section.    See  Meetings  of  the  Society. 

NOBTON,  Thomas  H.    On  Some  New  Forms  of  Gas  Generators ifl57 

Notes 189. 307.4x2. 558, 940."^ 

N0YB8,  W.  A.    The  Preparation  of  Dimethyl  Malonic  Bster ims 

»AT-KBRNBL,  proteids  of  the 6u 


INDEX.  1 121 

Obituary  notice  of  Peter  Collier 748 

Oils,  determinatioo  of  heat  of  brominatioo 378 

Oroline  yellow,  effect  of,  on  peptic  digestion,  Z093 :  on  pancreatic  digeation 1095 

Osborne,  T.B.  VegeUble  Proteids (4) 

OSBORMB.  Thomas  B.  and  George  P.  Campbell.    The  Chemical  Nature  of  DiaaUse, 
536;  The  Proteids  of  Malt.  54a :  The  Proteids  of  the  Potato.  575 :  Legumin  and 

Other  Proteids  of  the  Pea  and  the  Vetch,  583 ;  Conglutin  and  Vitcllin 609 

Oxycelluloses,  the  natural 8 

Oxygen,  density  and  atomic  weight,  198 ;  valency  of  and  structure  of  compounds 

containing  it  (review) 383 

Ozone,  constitution  of,  a86 ;  msnufacture  of,  use  as  a  disinfectant,  for  bleaching,  etc  (46) 

Z^AMMBL,  B.  E.    See  Bennett,  A.  A. 

Paratoluidine,  heat  of  solution  in  ethyl  alcohol,  154 ;  in  chloroform,  155 ;  in  toluene. .    156 

Pasteur  memorial  service 930 

Pasteur  monument  committee « (105) 

Patkb,  Gborob  p.    Mineral  Constituents  of  the  Watermelon 1061 

Pea.  the  proteids  of < 583 

Peach  kernel,  proteids  of  the 613 

pRNNTNOTO!f ,  Mary  Bnolb.    Derivatives  of  Columbium  and  tantalum 38 

Permanganate,  probable  production  by  combustion  of  metallic  manganese  ....230,  (30) 
Petroleum,  lecture  by  C.  P.  Mabery,  (ao);   Distillation,  (51);  origin  of.   (64);  petro- 
leum products • (68) 

Phenanthrene,  heat  of  solution  in  ethyl  alcohol,  153 ;  in  toluene 136 

Phit.i,ips.  Francis  C.    The  Determination  of  Salphur  in  Cast  Iron.  1079 

Phosphates,  neutrality  of  citrate  solution  used  in  analysing,  457 ;  methods  of  analy- 
sis of « 926 

Phosphate  rock,  distinction  from  bone  phosphate 491 

Phosphf nes.  tertiary  and  quaternary,  mode  of  formation  of (89) 

Phosphoric  acid,  gravimetric  estimation  as  ammonium  phosphomolybdate,  33  ;  In 
fertilisers,  various  modifications  of  Pemberton's  volumetric  method  for  deter- 
mining. 3^;  accuracy  of  determination  of  ..., 812 

Ptaosphorns.  precipitation  of  phosphomolybdate  in  steel  analysis.  170 :  modified 
molybdate  solution,  445 :  in  pig  iron,  accuracy  of  determination.  Sri ;  in  steel  and 

cast  iron,  determination  of » 9S5 

Photometric  method  for  the  quantitative  determination  of  lime  and  sulphuric  acid.    661 

Phthalimid 679 

Pipette,  rapid  measuring 905 

PtACRWAY.  L.  A.    See  Bennett,  A.  A. 

Plaster  of  Paris  method  in  blowpipe  analysis,  some  extensions  of  the  « . . . .    849 

Platinum  potassium  halides 130 

PI.ATT.  Charlrs.    The  Qualitative  Examination  of  Acctanilid,  T43 ;  The  Separation 

of  Alkaloidal  Extracts 1104 

Poison-ivy  as  a  skin  irritant.^ (81) 

Polariscopic  work,  use  of  acetylene  as  an  illuminant  in 179 

Polarizing  at  high  and  low  temperatures,  apparaturf'for 83,  85 

Popr,  Prrd.  J.  Volumetric  Estimation  of  Lead,  737 :  Estimation  of  Sulphides  in  Cal- 
cium Carbide 740 

Porcelain  factory  at  Sevres.  Prance,  description  of  931 

Potable  water,  determination  of  nitrites  in ai 

Potash ,  accuracy  of  determination 817 

Potash  bulb,  a  new  form  of > 941 

Potassium  chloride,  purificstion  of,  for  atomic  weight  determinations aoi 

Potassium  platinum  halides ijo 

Potansinm  and  sodium,  indirect  analysis  in  a  mixture  of  chlorides  of 184 

Potassium  trinitride,  use  of,  In  separating  thorium  from  the  other  rare  earths 947 

Potato,  proteids  of 575 


1 122  INDEX. 

Potter.  Witliam  R.    Fallacies  in  Urine  AulysU  (99) 

Pkbscott,  Albbrt  B.    Notes  on  a  Pew  Pyridine  Alkyl  Iodides,  91 ;  see  also  Flinter- 

mann,  R.  P.;  see  also  Baer,  S.  H. 

Prbscott.  A.  B.  and  S.  H.  Baer.   Pyridine  Alkyl  Hjrdrozides - M7 

Proteids  of  malt,  54a :   of  the  potato,  575 ;   of  the  pea  and  ▼etch,^583 ;  vef^eUble....    (4) 

Proteoses  of  malt s$9 

Protoproteose 556 

PucKif  BR,  W.  A.    Notesonthe  Estimation  of  CafTein 97^ 

Pyridine  alkyl  hjrdroxides,  247 ;  pjrridine  propyl  hydroxide.  348 ;  pyridine  isoprapyl 

hjrdroxide,  349 ;  pyridine  alkyl  iodides,  notes  on.  91 ;  pyridine  ethyl  iodide,  91 ; 

P3rridine  isopropyl  iodide,  93 ;  pyridine  methyl  iodide,  91 ;   pyridine   propyl 

iodide 9> 

Pyrites»  estimation  of  sulphur  in 446*  tiiS 

Pyrrhotite.  estimation  in  pyrites  ores 401 


.BPINING  liziviation  sulphides,  the  sulphuric  acid  process 643 

Reinsch*s  teat  for  arsenic  and  antimony • 953 

Resorcinol,  heat  of  solution  in  water,  153 ;  in  ethyl  alcohol tsi 

Rhode  Island  Section.    See  Meetinips  of  the  Society. 

Richards,  Bllbn  H.  and  J.  W.  BUms.  The  Coloring-Matter  of  Natural  Waters,  its 

Source,  Composition  and  Quantitative  Measurement O 

R1CHARD80K,  G.  M.   Obituary  notice  of  August  Kekul^ 1107 

RiCHARDBON,  H.  A.    See  Gill,  Augustus  H. 

Risiifo,  W.  B.  and  Victor  Lenher.    An  Electrolytic  Method  for  the  Determination  of 

Mercury  in  Cinnabar •  9^ 

Robinson,  A.  E.  and  Mabery.  Charles  P.  Composition  of  Certain  Mineral  Waters  of 

Northwestern  Pennsylvania 9'5 

RoLPB.  Gko.  W.  and  George  Defren.   An  Analytical  Investigation  of  the  Hjrdrolysis 

of  Starch  by  Acids 861 

Ross,  B.  B.    Some  Analytical  Methods  involving  the  Use  of  Hydrogen  Dioxide  ..•<  9'^ 

Rubber  manufacture (49) 

Rubidium  fluoride S7 

Ruthenocyanides,  contribution  to  the  knowledge  of 9B1 

fliAARBACH.  LUDWIG.    A  Simple  Form  of  Gas  Regulator 5" 

Saccharimeter.  comparison  of  scales 933 

Saffoline,  effect  of,  on  peptic  digestion,  1094  ;  pancrsatic  digestion  1095 

Sa'undbrb.  W.  M.    Amphoteric  Reaction  of  Milk (33) 

ScHAPPBR.  Hrrbrrt  A.  Bud  Edgar  P.  Smith.  Tungsten  Hexabromide 1098 

School  of  agriculture  at  Grignouy  Prance 9^ 

Sbal,  Alprbd  Nbwlin.     Action  of  Acid  Amides  upon  Bensoin  ...• loi 

Sewage,  disposal  of ,  Paris 93^ 

Srorby,  Edmund  C.   On  Two  Sources  of  Error  in  Sugar  House  Analyses 4^ 

Silicon,  determination  of,  in  aluminum 7^ 

Silver,  oxidation  of.  254 ;  accuracy  of  determination  in  cop^r  materials  816  ;  deter- 
mination of  atomic  mass  of,  990;  preparation  of  pure  metal,  99a;  separation 
from  arsenic  by  means  of  hydrochloric  acid,  1039 ;  acetate,  preparatiottof  pare, 
998:  benzoate.  preparation  of  pure,  looi :  nitrate,  preparation  of  pure.  995 ;  oxide, 
preparation  of  pure,  994;  sulphide,  refining  of 643 

Skin  irritanU (81) 

Slag,  barium,  in  blast  furnace (3>) 

Smith,  Edoar  P.    See  Benkert,  Arthur  L.;  Kelley,  Jerome,  Jr.;  Schaffer,  Hert>ert 

A.;  Taggart,  Walter  T.  and  Pield,  Charles. 
Smith,  Edgar  P.  and  Daniel  L.  Wallace.     The  Electrolytic  Estimation  of  Mercury   1^9 

Smitr,  Edward  L.    Rapid  Measuring  Pipette 9<9 

Smith,  J.  G.    See  Herty,  Charles  H. 

Smith,  Walter  E.    The  Origin  of  Petroleum {Hi 

Smithsonian  Institution,  account  of (>7) 


INDBX.  1 1 23 

Smokeless  powder,  (66) ;  the  development  of • 819 

fiodinm,  determination  of,  in  aluminum,  770 ;  pyroarsenate,  action  of  bjdrochlorlc 

acid  STMon i<Q7 

Sodium  and  potassium  chlorides.    Indirect  analysis  of  a  mixture  of 184 

Solution  of  carbon  compounds,  heats  of 146 

Specimen  bottle 4^2 

Sfshcrr.  a.  B.    See  Dennis,  I«.  M. 

Spktkrs,  C.  L.    Heats  of  Solution  of  Some  Carbon  Compounds,  146;  Some  Thouirbts 

about  Liquids • 7*4 

Squibb,  Hdwabd  R.    The  Manufacture  of  Acetone  and  of  Acetone  Chloroform  from 

Acetic  Acid.  331;  Volumetric  Determination  of  Acetone 106S 

Stabi.,  Karl  P.    Hydrofinoric  Acid -•   41S 

Starch,  an  analjrtical  investiflration  of  the  hjrdrolysis  of,  by  acids 869 

Stas  memorial « aoi 

Steel,  carbon  determination  in,  »3 ;  present  status  of  iron  and  steel  analysis (^) 

Stilucah,  Tboicas  B.  Note  on  tht  Solubility  of  Bismuth  Sulphide  in  Sodium  Sul- 
phide with  Special  Reference  to  the  Bstimation  of  Small  Amounts  of  Bismuth  in 

Anti-Priction  Alloys ^3 

Stoddabd,  John  Tafpan .  A  New  Balance  for  Pirst  Year's  Work  in  General  Chem- 
istry (Note) x89 

8TOKB8.H.  N.    MeUphosphinic  Acid •-.••  (go)* 

Stonb,  Gbobgr  C.  Remarks  on  Mr.  Auchy's  Paper  on  the  Volumetric  Detennina- 
tion  of  Manganese,  338,  (30) ;  Probable  Production  of  Permanganate  by  Direct 
Combustion  of  Metallic  Manganese,  230,  (30)  ;  Note  on  the  Solubility  of  Bismuth 

Sulphide  in  Alkaline  Sulphides 109Z 

STuifTC.  Prof.  C.  R.    Resolutions  adopted  by  Cincinnati  Section  on  death  of (34) 

Style  in  writing  chemical  papers,  discussion  on • •  (77) 

Succinimide.  heat  of  solution  in  water,  xsx  ;  in  ethyl  alcohol X54 

Sucrose,  heat  of  solution  in  water xsx 

Sugar,  inversion  by  salts,  130.  693 ;  determiimtlon  of  reducing,  in  terms  of  cuprie 

oxide 749 

Sugar  cane,  occurrence  of  the  amines  in  the  juice  of 743 

Sulphides,  estimation  of ,  in  calcium  carbide. .« 74o 

Sulphur  in  pig-iron,  Drown's  method  of  determining,  406 ;  estimation  in  pyrites, 

446 ;  estimation  of.  in  pyrites.  685  ;  determination  of,  in  cast  iron  1079 

Sulphuric  acid,  reduction  of.  by  copper,  351 ;  quantitative  determination  of,  by  pho- 
tometric method,  661 ;  determination  of,  683 ;  determination  as  barium  sulphate, 

793;  reduction  of.  by  copper 942 

SUMMRRS,  BSRTR AKD  S.    Carbou  Determinations  in  Pig  Iron X087 

Sunflower,  proteids  of  the * 6s3 

^F  AGO  ART,  Waltbr  T.  and  Smith,  Edgar  P.  The  Separation  of  Manganese  fxom 

Tungstic  Acid 1053 

Tantalum,  reactions  of  oxides  with  carbon  tetrachloride SSa 

Tantalum  and  columbium,  derivatives  of 38 

Tellurium,  atomic  weight  of • so6 

Tetraphenylasine,  study  of < ixa 

Thallium,  some  compounds  of ^ 970 

Thallons  trinitride  970 ;  thallous  thallic  trinitride  973  ;  thallous  tellurate  973 ;  thal- 

lou^i  cyanplatinite > 97^ 

Thoria,  estimation  of 79t 

Thorium,  separation  from  other  rare  earths 947 

Titanium,  occurrence  of • 409 

Titanium  cesium  fluoride  ...• 60 

Titanium  rubidium  fluoride  • S8 

Tolnidine.    Sec  paratoluidine. 

Treasurer's  report • (14) 


1 1 24  INDEX. 

Trlmethylamtne.  the  separatioo  of  ammonium  from 670 

TungT'ten  hexabromtde   T098 

TungTfitic  acid,  separation  of  raaniranese  from • '0S3 

XTRKA,  heat  of  nolution  in  water.  150;  in  ethyl  alcohol iS3 

Urethane,  heat  of  solution  in  water.  150 ;  in  methyl  alcohol.  152 :  in  ethyl  alcohol,  iss: 

in  propyl  alcohol.  154 ;  in  chloroform,  155 ;  in  toluene ISS 

Urine  analysis,  fallacies  in (99) 

ID^ANADIUM,  separation  from  arsenic IQSX 

Va  rnish-making (47) 

Veitch,  p.  p.  On  the  Various  Modifications  of  the  Pemberton  Volumetric  Method  for 

Determinins:  Phosphoric  Acid  in  Commercial  Fertilizers 389 

Vbnablb.  p.  p.  and  Thomas  W.  Clarke.  A  Study  of  the  Zirconates 4M 

Vetch,  the  proteidsof sSs 

Vitellin 609 

Vivisection,  objectionable  letislation  in  regard  to (87).  (94).  (95).  (97) 

Volhard's  method  for  mangranese.  sources  of  error  in 498 

Volumetric  analysis,  estimation  of  lead,  737 ;  determination  of  acetone 1068 


'ALLACH,  DANIEL  L.  See  Smith.  Bdffar  P. 
^Washington  Section.    See  Meetings  of  the  Society. 

Water,  constitution  of,  286 ;   determination  in  viscous  orfranic  liquids 937 

Water  analysis,  chemical  versus  bacteriological.  166;  measurement  of  the  colors  of 
natural  waters,  264  ;  the  valuation  of  color  in,  484 ;  recent  advances  in  milk  tn- 

vestigrations •• (7:^) 

Watermelon,  mineral  constituents  of  the io6t 

Waikwright,  J.  H.    The  Determination  of  the  Solid  Pat  in  Artificial  Mixtures  of 

Vegretable  and  Animal  Pats  and  Oils 9S9 

Wait.  Charles  K.    The  Oxidation  of  Silver.  254 ;  The  Occurrence  of  Titanium 402 

Weber.  H.  A.   On  the  Behavior  of  Coal-Tar  Colors  toward  the  Process  of  Diares- 

tion lOQ} 

Walnut,  proteidfl  of  the 616 

Wiley  H.  W.  On  the  Estimation  of  Levulose  in  Honeys  and  oth<»r  Substances.  81. 
189 ;  Note  on  the  Use  of  Acetylene  Gas  as  an  Illuminant  in  Polariscopic  'Work, 
179  ;  Determination  of  the  Heat  of  Bromination  in  Oils,  378.  (^);  Second  Interna- 
tional Congress  of  Applied  Chemistry,  923  ;  A  New  Pormof  Bbullioscope.  1063; 
see  also  Bwell.  E.  E- 
WiLEr,  H.  W.  and  E.  E.  Bwell.  Determination  of  Lactose  in  Milks  by  Double  Dilu- 
tion and  Polarization « iiH 

WiLTON,A.L.    A  Modified  Molybdate  Solution 445 

Wood  Gum,  composition  of .  2x4;  from  birch  wood 218 

:ylAN * 215 

'TTRTUM.  atomic  weight  of 209 

SBINC.  atomic  weight  of 203 

Zinc,  electrolytic  determination  of 6m 

Zinc,  separation  from  arsenic  by  means  of  hydfochloric  acid 1041 

Zinc  ferrocyanide  • 1100 

Zinc  oxide,  curious  forms  of (31) 

Zirconates.  a  study  of  the 434 

Zirconium  tetraiodide 673 


IftSYied  witb  January  Number.  1896. 

Proceedings. 


BOARD  OF  DIRECTORS. 

The  Board  of  Directors  have  passed  the  following  resolutions : 

**  Resolved^  That  the  Board  of  Directors  hereby  approve  and 
ratify  the  action  of  the  majority  of  said  Board,  as  obtained  by 
their  signatures,  in  granting  a  charter  for  a  Local  Section  of  the 
American  Chemical  Societ}*^  in  North  Carolina,  and  that  the 
charter  date  from  the  time  said  action  was  taken,  November  8, 

1895." 
*'  Resolved,    That  the  Finance  Committee  of  the  American 

Chemical  Society  is  hereby  authorized  to  approve,  and  the  treas- 
urer to  pay  to  the  General  Secretary  each  month  during  the  year 
1896  a  bill  or  bills  for  clerical  help,  provided,  however,  that  the 
total  sum  called  for  by  said  bills  does  not  amount  to  two  hun- 
dred and  fifty  dollars  ($250.00).*' 


NBW  MBMBBRS  BLKCTBD  XOVBMBBR  31,  1895. 

Bailey,  Ralph  Waldo,  Elizabeth,  N.  J. 

Bischoff,  Dr.  .Ernst,  87-89  Park  Place,  New  York  City. 

Broadhurst,  W.  Homer,  Polytechnic  Inst.,  Brooklyn,  N.  Y. 

.Doerflinger,  Wm.  F.,  Polytechnic  Inst.,  Brooklyn,  N.  Y. 

Uolbrook,  Frederick  A.,  75  Joralemon  St.,  Brooklyn,  N.  Y. 

Jameson,  A.  H.,  Cleveland  Linseed  Oil  Co.,  S.  Chicago,  111. 

Lc  Boutillier,  Clement,  High  Bridge,  N.  J. 

Morgan,  J.  Livingston  Rutgers,  New  Brunswick,  N.  J. 

Perry,  Frank  J.,  B.S.,  Polytechnic  Inst.,  Brooklyn,  N.  Y. 

Potter,  Charles  A.,  174  Weybosset  St.,  Providence^  R»  I. 

Shaw,  Wm.  T.,  Chem.  Lab.  Agn  Exp.  Sta.,  Bozeman,  Mont. 

Tucker,  S.  A.,  135  Madison  Ave.,  N.  Y.  City. 

Tyrer,  Thomas,  Stirling  Chem.  Works,  Stratford,  E.   England. 

ASSOCIATB  BI<BCTBD  NOVBMBBR  21)  1895. 

Tuckerman,  Alfred,  342  West  57th  St.,  N.  Y.  City. 

NBW  MBMBBRS  BLBCTBD  DBCBMBBR  I3,   1865. 

Bellam,  Henry  Lynch,  B.S.,  Anaconda,  Mont. 
Cameron,  Prof.   Frank  Kenneth,  Catholic  Univ.  of  America, 
Washington,  D,  C. 


(2) 

Cushman,  AUerton  S.,  Washington  Univ.,  St.  Louis,  Mo. 
Cutts,  Henry  E.,  care  Stillwell  &  Gladding,  55  Fulton  St.,  N.T 
Elliott,  E.  C,  care  Univ.  of  Nebraska,  Lincoln,  Nebr. 
Hobbs,  Perry  L.,  Western  Reserve  Medical  College,  Clevelanc , 

Ohio. 
Meisel,  C.  F.  A.,  402  Washington  St.,  New  York  City. 
Moore,  Chas.  C,  Jr.,  Dept.  Agr.  Div.  Chemistry,  Washingto. 

D.  C. 
Schmidt,  H.  B.,  215  E.  4th  St.,  Cincinnati,  Ohio. 
Stoddard,  Dr.  H.  T.,  57  Crescent  St.,  Northampton,  Mass. 
Thomas,  W.  S.,  Belt,  Cascade  Co.,  Mont. 

ASSOCIATB  EI^CTED  DBCKMBBR  13,  1895. 

Waldman,  Louis  I.,  P.  O.  Box  162,  Albany,  N.  Y. 

CHANGES  OF  ADDRESS. 

Appleton,  Prof.  J.  H.,  209  AngellSt.,  Providence,  R.  I. 

Benton,  Geo.  W.,  27  E.  St.  Joe  St.,  Indianapolis,  Ind. 

Dalton,  Parmly,  Swampscott,  Mass. 

Dunham,  E.  K.,  338  E.  26th  St.,  New  York  City. 

Ehrenfeld,  A  Clemens,  Steele  High  School,  Dayton,  O. 

Feid,  George  F.,  519  Findlay  St.,  Cincinnati,  Ohio. 

GriflSth,  Dr.  S.  H.,  U.  S.  Naval  Museum  of  Hygiene,  Wash- 
ington, D.  C. 

Guiterman,  Franklin,  care  Omaha  and  Grant  Sm.  Co.,  Durango, 
Colo. 

Hewitt,  Edward  R.,  119  E.  i8th  St.,  New  York  City. 

Lammers,  Theodore  L.,  Helena,  Mont. 

Textor,  Oscar,  158  Superior  St.,  Cleveland,  Ohio. 

Trubek,  M.,  Raceland,  La. 

Volckening,  Gustave  J.,  88  Clinton  Ave.,  Brooklyn,  N.  Y. 

Wood,  Edward,  Harvard  Medical  School,  Boston,  Mass. 

ADDRESS  WANTED. 

Grosvenor,  Wm.,  Jr.  Last  address.  Box  166,  Johns  Hopkins 
Univ.,  Baltimore,  Md. 


MEETINGS  OF  THE  SECTIONS. 

NEW  YORK   SECTION. 

The  regular  meeting  was  called  to  order  December  6th,  1895. 
at  8.25,  Prof.  P.  T.  Austen  in  the  chair.  There  were  about 
sixty  members  present. 

The  chairman  opened  the  meeting  with  the  statement  that  Dr. 
Webb,  the  President  of  the  College  of  the  City  of  New  York, 


(3) 

and  Prof.  R.  Ogden  Doremus,  had  put  the  chemical  lecture  room 
of  the  college  at  the  disposal  of  the  society ;  and  in  his  opinion 
it  was  the  most  satisfactory  and  most  favorably  situated  of  any 
that  had  yet  been  considered.  He  then  introduced  Prof.  Dore- 
mus,  who  said  that  he  was  as  much  surprised  as  any  one  at  the 
success  of  the  society's  request,  as  he  had  been  under  the  im- 
pression that  there  was  something  in  the  charter  of  the  college 
which  prevented  such  use  of  the  room.  He  remarked  that  the 
laboratory  was  now,  since  the  destruction  of  the  University 
building,  the  oldest  educational  chemical  laboratory  in  the  city. 

He  hoped  the  society  would  find  it  suitable  to  their  purpose 
and  assured  it  of  the  heartiest  welcome. 

The  minutes  were  then  read,  and  the  remarks  of  Prof.  McMur- 
trie  in  regard  to  Illinois  waters,  as  reported,  were  corrected,  and 
the  minutes  adopted.  Prof.  McMurtrie  moved  that  the  thanks 
of  the  Section  be  sent  to  President  Webb  and  Dr.  Doremus  for 
their  courtesy  in  giving  the  Section  the  use  of  the  lecture  room. 
The  motion  was  seconded  and  carried. 

A  letter  addressed  to  the  chairman  from  the  Secretary  of  the 
English  Society  of  Chemical  Industry  was  then  read,  thanking 
the  New  York  Section  and  the  Lehigh  Valley  Section  of  the 
American  Chemical  Society,  and  the  New  York  Section  of  the 
Society  of  Chemical  Industry  for  the  honor  done  to  the  President 
of  their  Society,  Mr.  Thos.  Tyrer,  and  the  Hon.  Foreign  Secre- 
tary, Mr.  Ludwig  Mond,  on  the  occasion  of  their  recent  visit  to 
New  York. 

The  letter  was  ordered  on  file. 

Prof.  Moale  read  a  paper  entitled  **A  Brief  History  of  Naph- 
thalene,*' in  which  the  work  of  the  earliest  investigators  of  this 
interesting  substance  as  well  as  those  in  recent  years,  was  re- 
viewed. 

Mr.  Neiman  was  called  upon  by  the  chairman  and  gave  his 
experiences  in  attempting  to  make  naphthalene  synthetically, 
for  the  purpose  of  deciding  its  theoretical  constitution. 

He  stated  that  the  decomposition  of  certain  amido-naphthol- 
sulpho-acids  having  a  tendency  to  show  that  the  position  of  the 
double  bonds  in  the  napthalene  ring  are  not  symmetrical, 
attempts  were  made  to  disprove  this  by  the  synthetic  production 


(4) 

of  naphthalene  from  ortho-xylene  tetrabromide  and  ethane.  By 
passing  ethane  over  a  heated  mixture  of  granulated  pumice 
stone  and  ortho-xylene  tetrabromide,  a  portion  of  naphthalene 
was  formed,  but  circumstances  prevented  the  further  investig-a- 
tion  in  this  line.  This  formation  would  seem  to  show  that  the 
central  bond  is  a  double  one,  and  the  formula  a  symmetrical  one 
as  far  as  the  bonds  are  concerned. 

The. second  paper  of  the  evening,  on  "Vegetable  Proteids," 
was  read  by  the  author,  Dr.  T.  B.  Osborne. 

Mr.  Hewitt  asked  if  a  ten  per  cent,  solution  of  sodium  hy- 
droxide would  extract  all  the  proteids  or  only  one,  or  only  a  few. 

Dr.  Osborne  replied  that  all  the  proteids  would  dissolve. 

Mr.  Hewitt  had  extracted  the  white  bean  in  large  quantities, 
agitating  the  bean  flour  in  dilute  alkali  by  machinery,  and  had 
obtained  a  clear  solution  which  filtered  readily. 

Mr.  Hewett  asked  if  there  was  any  difiference  in  the  product 
on  repeated  precipitations.  Dr.  Osborne  replied,  **  No,  not  if 
they  are  pure. '  * 

Prof.  Speyers  said  that  the  most  it\teresting  point  to  him  was 
the  solubility  of  the  glutinoids  in  a  mixture  of  alcohol  and 
water,  when  it  appeared  that  they  were  insoluble  in  either  water 
or  alcohol  alone. 

Dr.  Osborne  thinks  there  may  be  a  hydrate  formed  by  takings 
up  water  from  the  dilute  alcohol,  and  this  hydrate  then  dis- 
solves. 

Dr.  Smith  said  that  no  one  who  had  not  worked  in  this  diffi- 
cult subject  could  appreciate  the  value  of  Dr.  Osborne's  work, 
and  especially  the  classification  which  had  been  made  of  the 
compounds. 

Mr.  Hewitt  suggested  a  method  of  separating  the  proteids  by 
availing  of  the  different  behavior  of  solutions  of  different  osmotic 
pressures,  and  described  experiments  in  which  he  had  used  mem- 
branes prepared  with  gelatine  treated  with  formaline,  which  he 
found  more  satisfactory  than  potassium  bichromate  or  tannin  for 
making  the  gelatine  insoluble.  He  had  also  found  that  the  re- 
sults differed  when  bichromate  or  tannin  were  used. 

Prof  Austen  asked  if  the  vegetable  proteids  are  entirely  dif- 


(5) 

ferent  from  the  animal,  and  if  there  is  any  classification  of  the 
latter. 

Dr.  Osborne  said  that  superficially  they  were,  quite  similar, 
but  closer  study  revealed  marked  difEerences.  Nearly  all 
authors  of  physiological  chemistries  give  classifications  for  these 
proteids,  but  that  most  of  these  authorities  differ  to  a  greater  or 
less  extent  from  one  another.  The  most  comprehensive  classi- 
fication of  the  animal  proteids  that  he  had  seen  was  that  given  by 
Prof.  Chittenden  in  his  Cartwright  lectures  for  1894. 

Mr.  J.  H.  Wainwright  read  a  paper  on  the  **  Determination  of 
Solid  Pats  in  Artificial  Mixtures  of  Vegetable  and  Animal 
Fats.''  He  said  that  the  problem  was  to  make  analyses  of  mix- 
tures of  solid  fats  and  vegetable  oils,  as  cottonseed-oil,  lard,  and 
oleostearin,  which  might  be  classed  as  compound  lards,  of 
which  "  cottolene  "  was  an  example ;  the  chief  object  being  to 
ascertain  the  percentage  of  oleostearin. 

In  a  simple  mixture  as  cottonseed-oil  and  stearin,  the  analy- 
sis can  be  readily  made  by  determining  the  constants  of  the  fat, 
iodine  number,  etc.  But  in  a  compound  lard  containing  lard 
itself,  the  determination  of  constants  gives  very  little  satisfaction, 
owing  to  the  confusing  effect  of  the  lard.  Experiments  were 
made  on  special  mixtures  with  the  result  of  proving  that  under 
pressure  at  ordinary  temperatures  both  cottonseed-oil  and  lard 
are  removed,  leaving  the  stearin. 

At  temperatures  much  above  75**  F.  or  much  below  70**  F.  the 
error  was  considerable,  but  within  these  limits  he  had  obtained 
results  differing  not  more  than  a  half  per  cent,  from  the  correct 
figure.  Until  the  method  was  further  perfected,  he  allowed  a 
plus  or  minus  error  of  one  and  a  half  per  cent. 

CINCINNATI  SECTION. 

The  regular  meeting  of  this  section  was  held  on  Tuesday 
evening,  December  17,  1895,  Dr.  Alfred  Springer  presiding. 

Prof.  T.  H.  Norton  spoke  of  the  loss  sustained  by  the  section 
in  the  death  of  Chauncey  R.  Stuntz,  professor  of  physics  and 
chemistry  at  Woodward  High  School,  and  moved  that  Messrs. 
F.  Homburg,  E.  Twitchell,  S.  P.  Kramer,  and  H.  B.Foote,  all 
former  students  of  the  deceased,  be  appointed  a  committee  to 
draft  resolutions  of  respect.    Prof.  Stuntz  was  one  of  the  best 


(6) 

known  educators  in  the  Ohio  Valley  and  one  of  the  org^anizers 
of  the  chemical  society  of  Cincinnati  and  vicinity,  which  after- 
ward became  the  Cincinnati  Section  of  the  American  Chemical 
Society.  He  was  elected  chairman  for  1893,  and  his  earnest 
work  in  behalf  of  the  Section  was  highly  appreciated  by  all  the 
members. 

Prof.  J.  U.  Lloyd  read  a  paper  on  ** Percolation"  and  gave  a 
practical  demonstration  oi  packing  the  percolator. 

The  committee  appointed  to  nominate  o£Bicers  for  the  Section 
for  1896  reported  the  following  ticket : 

President,  E.  Twitchell. 

Vice  Presidents,  Prof.  O.  W.  Martin  and  Chas.  G.  Merrill. 

Treasurer,  H.  B.  Foote. 

Secretary,  E.  C.  Wallace. 

Directors,  Dr.  S.  P.  Kramer,  Dr.  S.  Waldbott,  Dr.  John  Mc- 
Crae, 

On  motion  the  secretary  was  instructed  to  cast  the  ballot  of 
the  Section  in  favor  of  the  above  ticket. 


IfltYicd  with  Pcbmary  Nvmber,  1896. 


Proceedings. 


TWELFTH  GENERAL  MEETING  OP  THE  AMER. 

ICAN  CHEMICAL  SOCIETY. 

The  twelfth  general  meeting  of  the  American  Chemical 
Society  was  held  in  Cleveland,  Ohio,  December  30th  and  31st, 

X895. 
The  first  session  was  called  to  order  by  the  president,  Dr.  £. 

P.  Smith,  at  9.15  a.  m.,  Monday,  December  30th,  in  the  Chemi- 
cal Lecture  Room  of  the  Western  Reserve  Medical  College. 

Mr.  M.  S.  Greenough,  President  of  the  Cleveland  Gas  Light 
and  Coke  Company,  was  introduced  and  gave  the  following 
words  of  welcome : 

GentUmen  of  the  American  Chemical  Society: 

It  is  with  great  pleasure  that  I  have  accepted  the  invitation  of 
Prof.  Mabery  to  act  as  spokesman  for  the  local  interests  to  which 
you  are  allied,  and  welcome  you  to  the  hospitalities  of  our  city. 
Cleveland  is  a  very  popular  city  for  conventions,  and  I  am  in- 
formed that  no  less  than  100  have  met  here  during  the  year  now 
closing.  I  venture,  however,  to  assert  that  no  body  of  men 
gathered  together  in  this  vicinity  or  elsewhere  could  represent  a 
profession  more  useful  or  honorable  than  your  own.  What  civ- 
ilization owes  to  chemistry  is  hardly  appreciated  by  the  ordinary 
citizen.  He  is  so  accustomed  to  enjoy  the  health,  comfort  and 
prosperity  which  comes  from  it,  that  he  looks  upon  his  blessings 
as  part  of  the  natural  order  of  things,  and  never  stops  to  consider 
to  what  he  is  indebted  for  them.  When  man  was  in  his  primi- 
tive state  and  dressed  in  skins,  lived  in  a  tent  on  what  he  could 
kill  and  changed  his  residence  daily,  he  naturally  had  not  much 
use  for  chemistry,  but  nowadays  without  the  expert  analyst  we 
should  be  simply  helpless.  We  depend  upon  him  to  know  that 
our  drinking  water  is  safe  to  use,  whether  our  children's  milk  is 
genuine  or  diluted,  whether  our  groceries  are  pure  or  adulterat- 
ed. We  invoke  his  assistance  in  every  department  of  manufac- 
ture. In  the  steel  business  for  instance,  which  is  the  right  hand 
of  this  city,  the  buyer  purchases  on  a  guaranteed  percentage  of 
ingredients,  and  every  blow  is  tested,  and  where  iron  ores  are 
saleable  or  unsaleable  according  to  their  chemical  composition, 
there  the  chemist  is  absolutely  indispensable.     It  might  be  truly 


(8) 

said  that  without  the  constant,  persistent,  almost  unnoticed  work 
of  the  chemist,  this  city  of  Cleveland,  with  its  diversified  inter- 
ests, including  every  sort  of  steel  work,  with  its  paint  works,  its 
refineries,  its  chemical  works  and  its  great  ship  yards,  would  be 
a  small  town,  one-tenth  its  present  size,  with  gardens  and  or- 
chards coming  down  to  the  Public  Square.  Cleveland  is  built  oa 
the  work  of  the  expert  chemist,  and  yet  not  one  man  in  a  hun- 
dred ever  stops  to  realize  that  fact. 

My  own  business  is  to  furnish  illuminating  gas  to  this  commu- 
nity and,  of  course,  no  man  here  knows  so  well  as  I  how  much 
that  industry  is  indebted  to  chemistry.  When  I  left  Harvard 
College,  twenty-seven  years  ago,  and  entered  the  service  of  the 
Boston  Gas  Co.,  gas  sold  at  $2.50  per  thousand;  now  it  is  sold 
there  at  a  dollar,  and  here  this  company  only  nets  seventy-five 
cents  for  its  product,  after  paying  to  the  city  our  franchise  tax  of 
six  and  one-half  per  cent.  This  is  not  all  due  directly  to  chem- 
istry, but  a  great  part  of  it  is.  An  old-fashioned  gas  manager  is 
reported  to  have  said  **I  don*t  care  a  d — n  about  your  hydro- 
gens and  your  oxygens ;  you  give  the  coal  and  I'll  cook  the  gas 
out  of  it."  Such  a  man  was  once  very  useful,  and  may  be  still, 
but  if  everybody  had  held  his  opinions  gas  would  not  now  be 
sold  so  cheap  or  pure.  There  is  not  a  large  company  in  this 
country  to-day  but  what  either  employs  a  permanent  chemist  or 
has  one  near  by  on  whom  to  call.  Before  the  chemist  lent  his 
aid  gas  companies  either  ran  their  tar  and  ammonia  into  the 
nearest  river,  or  else  burned  their  tar  and  left  their  ammonia  in 
the  gas.  Without  his  experiments  and  analyses  we  should  never 
have  applied  the  principles  of  regenerative  gas  firing  to  our  re- 
tort benches  or  our  gas  burners.  We  should  have  failed  in  the 
successful  enrichment  of  .decomposed  steam  by  petroleum  prod- 
ucts, which  has  been  a  development  of  the  last  twenty  years,  and 
which  furnishes  the  method  by  which  gas  can  be  most  cheaply 
made  in  many  localities  in  this  country. 

Auer  Von  Welsbach  was  an  Austrian  chemist  who  discovered 
the  luminous  qualities  of  some  rare  earths,  by  heating  which  a 
foot  of  gas  is  enabled  to  give  three  times  as  much  light  as  is  or- 
dinarily obtained  by  burning  it,  and  by  which  gas  has  been  fur- 
nished with  its  strongest  weapon  in  the  fight  for  business.  Last 
of  all  is  this  new  discovery  of  calcium  carbide  with  its  product  of 
acetylene,  which  affords  the  most  beautiful  artificial  light  yet 
produced ;  and  though  I  am  by  no  means  prepared  to  endorse 
the  claims  of  its  enthusiastic  advocates,  yet  it  must  have  its  effect 
upon  the  lighting  interests  of  the  whole  country  of  every 
description.  *' Every  man  to  his  last''  says  the  shoemaker,  and 
I  speak  of  these  things  because  they  have  come  under  my  own 


(9) 

eyes ;  but  I  have  no  doubt  that  every  manufacturer  in  this  city 
might  be  heard  from  in  a  similar  strain  as  to  the  cheapening  and 
improving  of  his  product  by  the  skill  of  your  profession.  I  know 
that  I  speak  for  them  all  when  I  welcome  you  to  our  city,  and 
dwell  upon  the  respect  in  which  we  hold  you. 

You  will  find  upon  the  programs  a  large  number  of  enter- 
prises which  you  are  invited  to  visit.  The  gas  works  are  not 
upon  the  list,  but  if  any  of  your  body  are  interested  in  that  di- 
rection you  will  find  our  works  on  Willson  Avenue  a  good  exam- 
ple of  modem  gas  engineering. 

I  sincerely  trust  that  you  will  find  your  stay  here  both  agree- 
able and  interesting,  and  that  we  may  have  the  pleasure  of  seeing 
you  again,  either  individually  or  collectively. 

If  there  are  any  eastern  members  of  your  body  who  are  con- 
sidering the  advisability  of  moving  westward,  you  will,  I  am 
sure,  take  home  with  you  for  reflection  the  great  futiu-e  which  is 
awaiting  this  city  and  its  environs,  and  will  also  realize  the  oppor- 
tunities which  are  awaiting  the  man  who  settles  among  us 
thoroughly  equipped  with  the  education  of  the  industrial  chem- 
ist. 

In  response.  Dr.  E.  F.  Smith,  President  of  the  American 
Chemical  Society,  said : 

Believe  me,  sir,  that  the  American  Chemical  Society  fully 
appreciates  the  cordial  and  •  hearty  reception  that  has  been 
extended  to  its  members,  and  through  me  returns  to  you  and 
those  whom  you  represent,  its  sincere  thanks. 

We  are  glad  to  be  here,  and  we  are  eager  to  avail  ourselves  of 
the  many  opportunities  which  we  shall  have  while  in  your 
midst,  of  inspecting  the  many  industrial  plants  within  the  bor- 
ders of  this  city,  and  within  its  immediate  neighborhood. 

We  feel  particularly  grateful  to  our  Council  for  having  called 
us  here,  where  there  is  such  a  centralization  of  enterprises 
founded  on  scientific  principles.  I  can  assure  you,  we  will  take 
advantage  of  the  privileges  which  you  have  offered  to  us. 

We  feel  happy,  too,  in  the  thought  that  in  coming  here  we 
have  a  chance  to  meet  with  your  local  chemists,  who  are  a  host 
within  themselves.  They  have  wrought  well,  and  they  have 
contributed  very  largely  to  placing  the  name  of  your  city  upon  a 
high  pedestal  among  the  cities  of  this  nation  which  encourage 
industries  founded  upon  chemical  principles  and  processes.  Two 
of  them  particularly  are  we  proud  of :  I  need  not  mention  the 
names  of  Dr.  Morley  and  Dr.  Mabery,  the  first  of  whom  has 
won  for  himself  a  reputation  by  his  investigations  on  certain 
constants  of  nature ;  and  the  second  has  achieved  equal  glory  by 


(lO) 

the  researches  which  he  has  made,  and  the  light  which  he  has 
thrown  upon  thedif&cultiessurroundingthe  petroleum  problem. 

For  these  reasons,  and  for  the  opportunities  which  we  hope  to 
have  while  here  of  inspecting  these  great  industries,  and  for  the 
kindly  reception  given  us  and  the  many  hospitalities  which  will 
be  ours  while  we  are  here,  we  thank  you. 

I  trust  that  you,  and  all  Cleveland  for  that  matter,  if  conveni- 
ent, will  attend  our  session  and  join  in  the  discussions  of  the 
papers  which  are  to  be  presented.     (Applause.) 

The  President  then  called  for  the  report  of  the  General  Secre- 
tary, which  was  read,  and  by  vote  of  the  Society  was  ordered 
placed  on  file.     The  Secretary's  report  is  as  follows : 

To  the  Members  of  the  American  Chemical  Society: 

Gentlemen  : — ^The  record  of  the  American  Chemical  Society 
during  the  past  year  has  been  one  of  enlarged  activities,  and  of 
higher  attainments  and  more  extended  usefulness  than  ever 
before.  The  membership  of  the  Society  has  steadily  increased ; 
three  new  local  sections  have  been  established  and  many  of  the 
older  sections  have  made  commendable  progress  in  numbers  and 
in  the  character  and  influence  of  their  work  ;  the  journal  has 
been  much  improved,  and  the  Society  to-day  exerts  a  more 
potent  influence  among  chemists,  both  in  the  old  world  and  the 
new,  than  it  did  one  year  ago. 

The  roll  of  membership  December  26,  1894,  was  as  follows:— 
Members,  720 ;  associates,  55  ;  honorary  members,  8 ;  total,  783. 
On  December  26,  1895,  there  were  884  members,  58  associates, 
and  8  honorary  members ;  total  950.  If  to  this  number  we 
add  the  names  of  54  persons,  who  have  been  elected,  but  have 
not  yet  qualified,  (a  great  majority  of  them  having  been  elected 
since  the  ist  of  November,  and  according  to  the  constitution, 
not  being  required  to  qualify  before  January  ist),  and  31  whose 
applications  for  membership  are  now  under  consideration,  we 
have  a  grand  total  of  1035,  which  n:\ay  be  considered  the 
present  numerical  strength  of  the  Society.  The  increase  in 
membership  during  1895  has  been  greater  than  in  any  previous 
year,  except  1894,  and  there  is  reason  to  believe  that  the  momen- 
tum which  the  Society  has  acquired  in  this  direction  during  the 
past  few  years  will  continue  for  a  long  time  to  come. 


(II) 

The  following  named  members  have  died  since  the  presenta- 
tion of  the  last  annual  report  of  the  General  Secretary  :  A.  A. 
Pesquet,  H.  B.  Nason,  J.  C.  Dittrich,  W.  H.  Whalen,  Mark 
Powers,  Lewis  W.  HofiFmann,  G.  E.  Moore,  W.  G.  Wallace  and 
Win.  C.  Wilson.  This  list  was  reported  to  the  Society  at  the 
Springfield  meeting  last  August,  and  sketches  of  Prof.  Nason, 
one  of  the  ex-Presidents  of  the  Society,  and  of  Dr.  Moore,  one 
of  the  most  highly  respected  members  of  the  New  York  Section, 
have  appeared  in  the  Journaf  of  the  Society. 

The  three  local  sections  established  during  the  past  year  are 
located  respectively  in  Chicago,  Nebraska  and  North  CaiDlina. 

There  are  now  nine  local  sections  of  the  Society,  viz  : 

Rhode  Island  Section  :  Presiding  Officer,  Charles  A.  Catlin, 
133  Hope  St.,  Providence,  R.  I.;  Secretary,  Walter  M.  Saun- 
ders, Olneyville,  R.  I. 

CincinnaH  Section :  Presiding  Officer,  Karl  Langenbeck,  27 
Orchard  St.,  Zanesville,  Ohio;  Secretary,  E.  C.  Wallace, 
Room  71,  Blymeyer  Bttilding,  Cincinnati,  Ohio. 

New  York  Section :  Presiding  Officer,  Peter  T.  Austen,  Poly- 
technic Institute,  Brooklyn,  New  York ;  Secretary,  Durand 
Woodman,  127  Pearl  St.,  New  York  City. 

Washington  Sectiofi:  Presiding  Officer,  Charles  K.  Munroe, 
Columbian  University,  Washington,  D.  C. ;  Secretary,  A.  C. 
Peale,  605  12th  St.,  N.  W.,  Washington,  D.  C. 

Lehigh  Valley  Section  :  Presiding  Officer,  Edward  Hart,  Lafay- 
ette College,  Easton,  Pa, ;  Secretary,  Albert  H.  Welles,  Lafay- 
ette College,  Easton,  Pa. 

New  Orleans  Section  :  Presiding  Officer,  A.  L.  Metz,  Tulane 
Medicfil  College,  New  Orleans,  La. ;  Secretary,  Hubert  Edson, 
Bartels,  La. 

Chicago  Section:  Presiding  Officer,  Prank  Julian,  South  Chicago, 
111. ;  Secretary,  F.  B.  Dains,  2421  Dearborn  St.,  Chicago,  111. 
«  Nebraska  Section :  Presiding  Officer,  H.  H.  Nicholson,  Uni- 
versity of  Nebraska,  Lincoln,  Nebraska ;  Secretary,  John  White, 
Box  675,  Lincoln,  Nebraska. 

North  Carolina  Section  .•  Officers  not  yet  reported. 

The  financial  outlook  of  the  Society  is  very  encouraging;  the 
report  of  the  Treasurer  shows  a  good  balance  after  paying  all 


(12) 

indebtedness.  A  little  well  directed  efifort  on  the  part  of  the 
members  to  secure  advertisiements  for  the  Journal,  and  increase 
the  regular  and  associate  membership,  would  add  very  mate- 
rially to  the  income  of  the  Society  and  would  enable  the  Com- 
mittee on  Papers  and  Publications  to  enlarge  the  scope  of  the 
Journal,  and  to  make  it  in  many  ways  even  superior  to  what  it 
is  at  present. 

The  membership  dues  have  been  collected  by  the  General  Sec- 
retary during  1895  ^^  1^  ^^^  previous  years.  This  work  has 
been  looked  after  very  closely,  and  the  results  have  been  of  two- 
fold advantage  to  the  Society — a  considerable  sum  has  been 
secured  that  otherwise  would  have  been  lost  in  unpaid  arrears, 
and  those  who  have  paid  their  dues,  after  repeated  reminders 
from  the  Secretary,  have  been  saved  to  the  membership  of  the 
Society,  and  have  been  more  prompt  the  next  time  in  their 
remittances. 

During  the  year  1895  it  has  been  necessary  to  drop  the  names 
of  only  twelve  persons  from  the  roll  of  the  Society,  as  required  by 
the  constitution,  for  non-payment  of  arrears. 

During  the  year  Prof.  F.  W.  Clarke  resigned  as  Chairman  of 
the  Committee,  appointed  by  the  Council,  for  considering  the 
question  of  revising  the  constitution,  and  Dr.  H.  W.  Wiley  was 
appointed  to  fill  the  vacant,  position.  This  committee  has  not 
yet  completed  its  work.  Under  the  authority  and  direction  of 
the  Society,  the  General  Secretary  secured  the  passage  of  a  bill 
by  the  New  York  Legislature,  enabling  the  Society  to  choose 
its  directors  without  regard  to  their  being  residents  of  the  State 
of  New  York,  or  any  other  State  or  locality,  and  also  legalizing 
whatever  action  the  Society  might  take  at  any  of  its  meetings 
held  outside  of  New  York  State. 

Prof.  Clarke  presents  his  annual  report  on  atomic  weights 
as  a  paper  to  be  read  at  this  meeting.  The  Society  is  to  be  con- 
gratulated in  having  among  its  members  a  person  so  able  and 
at  the  same  time  so  willing  to  present  fully  a  regular  annual 
report  upon  this  subject. 

During  the  year  the  President,  upon  the  authority  of  the 
Council,  appointed  Messrs.  Hale,  Austen  and  Breneman  as  a 
committee  to  consider  the  question  of  a  permanent  badge  for  the 


(13) 

Society.  The  committee  have  met  and  considered  the  subject 
and  requests  for  suggestions  and  designs  have  been  sent  to 
every  member  of  the  Society,  but  no  report  has  yet  been  pre- 
pared. 

The  Society  held  its  eleventh  general  meeting  in  Springfield, 
Mass.,  August  27th  and  28th,  1895,  just  previous  to  the  meeting 
of  the  American  Association  for  the  Advancement  of  Science  in 
the  same  city.  There  was  a  large  attendance  and  a  full  pro- 
gram of  papers.  The  meeting  was  one  of  unusual  interest  and 
inspiration  to  all  who  were  fortunate  enough  to  attend.  A  full 
account  of  the  proceedings  was  published  in  the  October  number 
of  the  Journal. 

Early  in  the  year  formal  invitations  were  received  from  the 
officials  of  the  city  of  Cleveland,  the  Chamber  of  Commerce  of 
Cleveland,  the  Western  Reserve  University,  the  Case  School  of 
Applied  Science  and  the  Cleveland  Chemical  Society,  for  the 
American  Chemical  Society  to  hold  their  annual  meeting  this 
year  in  Cleveland.  These  invitations  were  so  hearty,  and  Cleve- 
land is  so  desirable  a  city  in  which  to  hold  a  meeting  of  tiie 
Society,  that  the  Council  gladly  accepted  the  invitations,  with 
the  result  that  we  are  now  .the  favored  guests  of  these  bodies  and 
enjoying  their  cordial  and  unstinted  hospitalities. 

It  is  much  to  be  regretted  that  every  member  of  the  Society 
could  not  be  present  to  partake  of  the  rich  feast  we  find  pre- 
pared for  us  in  this  beautiful  and  enterprising  industrial  and  ed- 
ucational centre. 

In  looking  back  upon  the  past,  carefully  surveying  the  present 
condition  and  attainments,  and  anticipating  the  future,  the  mem- 
bers of  the  American  Chemical  Society  have  every  reason  for 
encouragement  and  gratification. 

If  the  Society  could  receive  from  all  its  members  the  loyaltj' 
and  active  support  which  has  always  been  given  by  those  who 
have  been  most  devoted  to  its  interests,  the  rapid  progress  of  the 
past  few  years  would  be  regarded  as  little  when  compared  with 
what  the  next  decade  would  witness.  May  we  not  hope  as  we 
begin  this  new  year  in  our  history  that  this  active  support  and 
loyalty  will  be  accorded,  and  that  every  member  will  to  the 
utmost  of  his  ability  exert  himself  to  increase  the  membership. 


(14) 

the  strength  and  the  influence  of  the  Society,  both  at  home  and 
abroad. 

We  sometimes  feel  that  we  need  to  establish  a  league  of  loyal 
Americans  in  the  realm  of  chemical  science  ;  whatever  Ameri- 
cans  accomplish  should  go  to  the  credit  of  America*.  Bat  this  is 
not  all ;  we  believe  it  is  a  mistake  for  any  chemist  under  the 
existing  conditions  to  think  that  his  best  path  to  recognition  by 
the  scientific  wolrld  lies  through  the  medium  of  foreign  periodi- 
cals. Articles  published  in  our  Journal  are  so  fully  abstracted 
and  so  often  copied  entire  by  foreign  scientific  periodicals  that 
it  seems  that  the  best  means  for  securing  general  publication  and 
wide  spread  recognition  for  any  deserving  paper  is  through  the 
columns  of  our  Journal.  Thus  not  only  loyalty  to  the  Ameri- 
can Chemical  Society,  but  also  self  interest  demands  that  the 
columns  of  our  Journal  be  kept  filled  with  the  records  of  the 
best  work  done  in  our  own  country. 

Respectfully  submitted, 

Albert  C.  Halb, 

Brooklyn,  N.  Y.,  Dec.  26, 1895.  General  Secretary, 

Financial  Report,  1895. 

Received  for  daes  from  Dec.  i,  1894  to  Dec.  14,  1895 %  4365.75 

Retained  as  Commission 436.50 

Balance  for  the  A.  C.  S.  Treasurer 39^-35 

Paid  A.  C.  S.  Treasurer  (as  per  vouchers) 3025.00 

Balance  not  yet  forwarded '^'S 

Interest ii^S 

Balance  on  Deposit 1 15.73 

Albbrt  C.  Hale, 
Dec.  14,  1895.  Gen.  Sec.  A.  C.  S. 

In  the  absence  of  the  Treasurer  his  report  was  read  by  the 
General  Secretary,  as  follows  : 

New  York,  December  26,  1895. 

TRBASURBR'S  REPORT  FOR  THE  YEAR  1895. 

Receipts, 

Balance  on  hand  Dec.  aist,  1894 %   505.95 

Net  dues  and  interest  received  from  the  Oen'l  Secretary 3f940-73 

Cash  received  for  subscriptions  to  Journal 624.93 

"          "          "    back  numbers 122.83 

''          "         "   advertisements  in  Journal 5'7'Z^ 

Interest  from  Farmers'  Loan  and  Trust  Co 9.S0 

l5.7«-97 


€t 
«< 
«« 


(15) 

Disbursements . 

expenses  of  Treasurer's  office $     1 1>35 

••  "         "  Gen'lSec'y  office 494-03 

"  ••  Editor'soffice 56.40 

"  Librarian's  office 78.18 

■  •  **  *  Springfield  meeting 42.17 

general  expenses • 67.30 

salary  of  editor 350.00 

publication  of  Journal 3,344.15 

insurance  •  •  •; 30.00 

"     rebates  to  Local  Sections,  as  follows : 

New  York  Local  Section 1 195.98 

Washington"  **        103.33 

Lehip;h  Valley  •*        30.00 

Cincinnati     **  *'         60.00 

Chicago  "  "        38.33 

Nebraska       "  *•        21.67 

449x3 

Balance -on  hand  Dec.  36th,  1895: 

In  Farmers  Loan  and  Trust  Co $  443.63 

In  Bank  of  Metropolis 356.90 

Checks  on  hand 184.15 

Cash  on  hand 12.00 

Postage  stamps 2.40 

99908 

I5.721.97 

The  following  report  of  the  Librarian  was  read  by  the  General 
Secretary : 

December  26,  1895. 

The  library  has  been  in  storage  during  the  year  and  therefore 
there  is  little  to  report.  It  is  hoped,  however,  that  a  suitable 
place  where  the  books  may  be  useful  to  the  members  will  soon 
be  found.  Several  places  have  been  suggested  but  none  as  yet 
that  meets  the  requirements  of  the  case. 

There  is  a  growing  call  for  back  numbers  of  the  Journal  and  I 
would  suggest  that  the  money  obtained  from  their  sale  be  used 
to  find  and  care  for  the  library. 

The  Librarian  has  received  the  following  exchanges : 

UNITED  STATES. 

American  Chemical  Journal. 

American  Joufnal  of  Pharmacy. 

American  Manufacturer  and  Iron  World. 

American  Naturalist. 

Annals  of  the  New  York  Academy  of  Arts  and  Sciences. 

Anthony's  Photographic  Bulletin. 


(i6) 

Bulletin  of  the  American  Museum  of  Natural  History. 

Deutsch-Amerikanische  Apotheker-Zeitung. 

Engineering  and  Mining  Journal. 

Kphemeris  (Squibb). 

Engineering  Magazine. 

Journal  of  the  Franklin  Institute. 

Journal  of  the  United  States  Artillery. 

New  York  Medical  Journal. 

Oil,  Paint»  and  Drug  Reporter. 

Popular  Science  Monthly. 

Proceedings  of  the  Academy  of  Natural  Sciences  (Philadel- 
phia). 

Proceedings  of  the  American  Academ}'  of  Arts  and  Sciences 
(Boston). 

Proceedings  of  the  American  Philosophical  Society  (Philadel- 
phia). 

School  of  Mines  Quarterly. 

Scientific  American. 

Technology  Quarterly. 

Textile  Colorist. 

Textile  Manufacturers*  Review  and  Industrial  Record. 

Transactions  of  the  American  Institute  of  Electrical  Engineers. 

Transactions  of  the  American  Institute  of  Mining  Engineers. 

Transactions  of  the  New  York  Academy  of  Sciences. 

CANADA. 

Journal  and  Proceedings  of  the  Hamilton  Association. 
Proceedings  of  the  Canadian  Institute. 

Proceedings  and  Transactions  of  the  Nova  Scotia  Institute  of 
Sciences. 

HOLLAND. 

Revue  Internationale  des  Falsifications. 

ITALY. 

Gazzetta  Chimica  Italiana. 

ENGLAND. 

Analyst. 

Chemical  News. 

Engineering. 

Journal  of  the  Chemical  Society. 

Journal  of  the  Society  of  Arts. 

Journal  of  the  Society  of  Chemical  Industry. 

Jil  and  Colorman*s  Journal. 

Pharmaceutical  Journal  and  Transactions. 

Sugar  Cane. 

Transactions  of  the  Institute  of  Brewing. 


(17) 

FRANCE. 

Annales  des  Mines. 

Bulletin  de  la  Soci^td  Chimique  de  Paris. 

Bulletin  de  la  Soci£t6  Industrielle  de  Rouen. 

Bulletin  de  la  Soci£t6  Industrielle  de  Amiens. 

Moniteur  de  la  Teniture. 

Moniteur  Scientifique  de  Quesneville. 

Repertoire  de  Pharmacie. 

GERMANY. 

Archiv  der  Pharmacie. 
Bierbrauer. 

Bulletin  de  la  Soci6t£  Industrielle  de  Mulhouse. 
Sitzungsberichte  der  K.  B.  Akademie  der  Wissenschaften  zu 
Munchen. 

AUSTRIA. 

AllgemeineOesterreichische  Chemiker  und  Techniker  Zeitung. 
Oesterreiches  Zeitschrift  ftir  Berg  und  Htittenwesen. 
(Proceedings)  Kaiserliche  Akademie  der  Wissenschaften  in 
Wien. 

RUSSIA. 

Bulletin  de  1*  Academic  Imperiale  des  Sciences  de  St.  Peters- 
burg. 

Memoirs  de  la  Soci^£  des  Naturalistes  de  Kiew. 

AUSTRALIA. 

Journal  and  Proceedings  of  the  Royal  Society  of  New  South 
Wales. 

ROUMANIA. 

Buletinul  Societatii  de  Sciinte  Fizice. 

The  Librarian  wishes  to  acknowledge  the  receipt  of  the  fol- 
lowing volumes  : 

Chemical  Bulletins  U.  S.  Department  of  Agriculture. 

Reports  and  Bulletins  of  the  Massachusetts  Experiment  Sta- 
tion. 

Reports  and  Bulletins  of  the  Connecticut  Agricultural  Exper- 
iment Station. 

One  hundred  years  of  business  life,  Wm.  J.  Schie£Felin. 

An  Introduction  to  the  Study  of  Rocks.  Presented  by  the 
Trustees  of  the  British  Museum  of  Natural  History. 

Respectfully  submitted, 

F.  E.  Dodge, 

Librarian. 


(i8) 

Dr.  Hart  was  then  called  upon  and  made  a  report  for  the  Com- 
mittee on  Papers  and  Publications. 

He  stated  that  last  year  915  pages  of  the  Journal  were  pub- 
lished, this  year  we  published  1092  pages,  and  we  have  enottg^h 
papers  left  over  to  fill  the  January  number  and  part  of  the  Feb- 
ruary number.  The  committee  have  been  hampered  in  their 
plans  by  the  financial  condition  of  the  society.  But  the  treasu- 
rer's report  is  an  encouraging  one,  and  we  hope  next  year,  if  inre 
are  to  go  on  with  the  work,  to  show  still  better  results. 

I  may  say  that  there  are  frequent  complai  nts  of  non-delivery  of  tbe 
Journal  from  members  of  the  society.  The  difficulty  in  most  cases 
is  not  in  the  sending  out,  but  with  the  post  office  authorities*  who 
are  so  overwhelmed  with  second-class  matter  that  they  become 
careless.  The  Journal  is  mailed  as  carefully  as  it  is  possible  to 
do  it,  the  address  printed  and  kept  standing,  and  there  are  very- 
few  mistakes  made  in  the  office  of  distribution.  I  hope  that 
members  who  do  not  receive  the  Journal  regularly  will  write  to 
us,  and  we  will  make  every  effort  to  get  the  Journal  to  them. 
Very  often  when  a  complaint  has  been  made  of  non-receipt  of 
the  Journal,  another  Journal  has  been  sent,  and  the  second  one 
has  not  been  received.  The  difficulty  seems  to  be  with  Uncle 
Sam's  method  of  conducting  business. 

Several  plans  have  been  suggested  and  considered  for  increas- 
ing the  efficiency  of  the  Journal,  but  nothing  that  has  been  sug- 
gested is  yet  ready  for  report. 

The  question  of  a  good  journal  is  largely  a  financial  question. 
If  we  have  money  to  print  and  circulate  a  journal,  we  can  have 
a  good  journal.  There  is  no  difficulty  about  papers.  We  have 
more  good  papers  now  than  we  can  manage. 

It  would  perhaps  be  interesting  to  the  members  to  know  some- 
thing about  the  actual  circulation  of  the  Journal,  which  is  con- 
siderably in  excess  of  the  membership.  We  sent  out  for  Decem- 
ber 11 25  Journals.  Of  this  number  less  than  fifty  are  exchan- 
ges, so  our  actual  paid  subscription  list  is  very  nearly  iioo.  The 
returns  for  the  next  year  are  beginning  to  come  in,  and  I  am 
able  to  report  large  accessions  to  the  number  of  subscribers,  es* 
pecially  foreign  subscribers. 


(19) 

Prof.  Sabin  reported  for  the  Finance  Committee. 

Formal  reports  were  then  made  by  members  of  special  com- 
mittees as  follows : 

Committee  on  Duty- Free  Importation,  C.  £.  Munroe  ;  Com- 
mittee on  Nomenclature  and  Spelling  of  the  Journal,  Edward 
Hart;  Committee  on  Triennial  Congre^  of  Chemists,  F.  W. 
Clarke. 

The  Secretary  then  read  a  letter  from  Dr.  T.  H.  Norton,  of 
Cincinnati,  expressing  his  regret  that  he  was  unable  to  attend 
the  meeting  and  sending  his  best  wishes  for  an  enjoyable  and 
profitable  occasion. 

The  following  communication  was  then  read  by  the  Secretary  : 

Department  of  the  Interior, 
United  States  Geoixxjicai,  Survey. 

Washington,  D.  C.  November  30,  1895. 

To  the  President  of  the  American  Chemical  Society  : 

Sir:  In  compliance  with  a  request  emanating  from  the 
Chemical  Division  of  this  Survey,  I  address  you  as  the  head  of 
the  most  representative  body  of  American  chemists  with  a  view 
to  securing  action  on  the  part  of  the  American  Chemical  Society 
looking  toward  the  general  adoption,  in  this  country  at  least,  of 
a  method  for  the  proximate  analysis  of  coal. 

The  prevailing  method  of  proximate  anal3rsis,  though  unscien- 
tific and  far  from  satisfactory,  is  still  capable  of  affording  infor- 
mation which  is  valuable,  as  chemists  and  geologists  know, 
both  as  a  preliminary  to  more  extended  scientific  examination 
and  as  to  the  value  of  coal  for  one  or  the  other  of  the  uses  to 
which  it  may  be  put  as  a  fuel.  But  in  practice  such  wide  diver- 
sity exists  in  the  details  of  this  method  that  the  analyses  of  dif- 
ferent series  of  coals,  made  by  different  chemists,  are  seldom  of 
much  value  for  purposes  of  comparison,  since  concordant  results 
are  only  to  be  attained  by  a  rigid  adherence  to  a  certain  order 
of  procedure. 

This  matter  is  of  great  importance  to  geologists  and  chemists 
as  well  as  to  those  who  contemplate  investing  in  coal 
properties  and  to  many  large  consumers  of  coal.  A  uniform 
method  of  analysis,  which  should  also  cover  the  determination 
of  sulphur  in  coals  seems  therefore  very  desirable,  and  the  adop- 
tion of  such  a  method  can  most  readily  be  brought  about  by  the 
authoritative  sanction  of  the  American  Chemical  Society. 

I  would  make  the  suggestion  that  a  committee  of  chemists 
experienced  in  coal  analysis  be  appointed  with  instruction  to 


(20) 

gather  from  all  sides  the  views  of  those  whose  opinions  are  likely 
to  be  of  value  in  connection  with  their  own,  and  from  the  data 
thus  collected  to  formulate  in  minute  detail  a  method  which  may 
come  to  be  accepted  as  the  one  by  which  all  analyses  of  coal  and 
coke  infthis  country  shall  be  made. 

It  is  not  necessary  that  a  nbvel  method  be  devised,  but  only 
that  the  diversity  in  detail  now  practiced  be  reduced  to  uniform- 
ity by  the  selection  of  those  features  which  in  the  judgment  of 
the  committee  will  most  nearly  meet  the  exigencies  of  the  case. 

Yours  with  respect, 

Chas.  D.  Walcott, 

Director. 

On  motion  of  Prof.  Edward  Hart  it  was  resolved  that  the 
President  appoint  a  committee  of  three  to  take  into  consideration 
Prof.  Walcott's  communication  and  present  a  report  upon  the 
same  at  the  Summer  meeting. 

After  some  announcements  by  the  General  and  Local  Secre- 
taries, A.  A.  Bennett  read  a  paper  on  *'  The  Quantitative  Deter- 
mination of  the  Halogens  in  the  presence  of  each  other;"  and 
Wm.  McPherson  presented  a  paper  on  ''Constitution of  Oxyazo- 
benzene."  The  latter  was  discussed  by  Drs.  Prescott,  Hart  and 
Mabery. 

In  the  absence  of  the  author  a  paper  by  Willis  K.  Everetteon 
the  ''Method  of  Analysis  of  Nickel  and  Cobalt  in  Ores,"  was 
read  by  the  General  Secretary,  and  was  afterwards  discussed  by 
Drs.  Mabery  and  C.  B.  Dudley. 

A.  B.  Prescott  then  presented  a  paper  prepared  by  himself  and 
S.  H.  Baer  on  the  "Melting  Points  of  Certain  Homologous  Pyr- 
idine Derivatives,*'  and  this  was  followed  by  another  paper  enti- 
tled * '  Pjrridine  Alkyl  Hydroxides,  *  *  by  the  same  authors.  These 
papers  were  discussed  by  Drs.  Fireman,  Smith  and  Hart.  After 
some  announcements  the  session  adjourned. 

In  the  afternoon  visits  were  made  to  various  works  in  Cleve- 
land, and  in  the  evening  the  laboratories  and  lecture  rooms  of 
Adelbert  College  and  the  Case  School  of  Applied  Science  were 
inspected,  after  which  the  Society  held  an  evening  session. 

The  evening  session  of  the  Society  was  called  to  order  by 
President  Smith,  at  8.15  p.  m.  in  the  Chemical  Lecture  Room  of 
the  Case  School  of  Applied  Science.  After  some  announcements 
by  the  General  Secretary,  Dr.  Chas.  F.  Mabery  was  introdnced 


(21) 

and  delivered  a  very  valuable  and  interesting  address  upon  pe- 
troleum. Prof.  Mabery  gave  an  account  of  the  experimental 
methods,  products,  and  results  connected  with  work  now  in 
progress  on  American  petroleums.  The  different  forms  of  stills 
employed  in  fractional  distillation  both  under  atmospheric  pres- 
sure and  in  vacuum  were  shown,  together  with  the  apparatus 
for  distillation  under  diminished  pressure  when  many  operations 
are  in  progress.  The  determination  of  sulphur  in  gases,  liquids, 
and  solids  was  described  and  illustrated  by  the  apparatus. 

Representative  crude  oils  from  the  Oil  Springs  and  Petrolia 
fields  in  Canada,  from  the  Lima  and  Pindlay  fields  in  Ohio,  and 
the  Berea  grit  sandstone  in  Ohio  were  exhibited  and  their  com- 
position given  as  well  as  the  composition  of  representative  oil 
rocks, — the  Corniferous  limestone,  the  Trenton  limestone,  and 
the  Berea  Grit  sandstone. 

A  distillation  now  in  operation  for  the  separation  of  the  bu- 
tanes and  pentanes  from  a  very  light  gasoline  (92^)  in  which  a 
distillate  was  collecting  below  — 10®  was  shown  in  operation, 
together  with  other  distillates  with  low  boiling  points,  and  their 
halogen  derivatives  which  are  now  under  examination  for  the 
purpose  of  establishing  the  identity  of  the  butanes.  The  puri- 
fied octanes  and  some  of  their  halogen  derivatives  were  also 
described. 

Prof.  Mabery  read  a  letter  from  Professor  Markownikow  of 
Moscow,  which  stated  that  Professor  Markownikow  had  given 
no  attention  to  Pennsylvania  petroleum.  In  one  of  his  papers, 
the  suggestion  had  been  made  that  the  Pennsylvania  oil  might 
prove  to  contain  the  naphtenes.  This  assertion  from  Professor 
Markownikow  was  obtained  to  correct  the  erroneous  statements 
in  German  and  American  works  on  petroleum  that  Markowni- 
kow had  examined  Pennsylvania  petroleum. 

Professor  Mabery  exhibited  many  specimens  of  hydrocarbons 
which  had  been  separated  from  Berea  Grit,  Ohio,  Canada,  and 
Pennsylvania  petroleums  for  the  purpose  of  ascertaining  the 
composition  of  these  oils  above  150°. 

A  number  of  specimens  of  sulphur  compounds,  including  sul- 
phides and  unsaturated  hydrocarbons  were  shown  that  had  been 
separated  from  Canadian  petroleum. 


(22) 

After  the  address  several  questions  were  asked  of  Dr.  Mabery 
and  various  points  were  discussed  by  Drs.  Dudley  andPiescott; 
Profs.  Moulton  and  Breneman  and  Mr.  Prasch.  Upon  motion 
of  Dr.  Hale,  the  Society  passed  a  unanimous  vote  of  thanks  to 
Dr.  Mabery.     The  evening  session  then  adjourned. 

The  morning  session  of  Tuesday,  December  31st,  was  called 
to  order  by  President  Smith  at  9. 10  A.  m.  After  some  announce- 
ments by  the  General  Secretary,  the  President  named  the  mem- 
bers of  the  Committee  on  Coal  Analysis,  in  accordance  with  the 
request  received  by  communication  from  Prof.  Walcott.  The 
committee  named  were:  Drs.  W.  P»  Hillebrand,  C.  B.  Dudley, 
and  W.  A.  Noyes. 

Mr.  James  Otis  Handy  then  read  a  paper  on  **  Improved 
Methods  for  the  Analysis  of  Aluminum,  Alumina  and  Bauxite;" 
this  was  followed  by  a  paper  on  "The  Cyanide  Method  of  Ex- 
tracting Gold  from  its  Ores,*'  by  Wm.  J.  Martin,  Jr.,  read  by 
the  General  Secretary  in  the  absence  of  the  author. 

A  paper  on  *  *  The  Use  of  the  Calorimeter  in  Detecting  Adul- 
terations of  Butter  and  Lard,"  by  E.  A.  de  Schweinitz  and  James 
A.  Emery,  was  read  by  Prof.  Sabin,  the  authors  of  the  paper 
being  absent.  Prof.  Sabin  also  discussed  some  of  the  points 
contained  in  this  paper. 

A  paper  by  H.  W.  Wiley  on  **  Determination  of  the  Heat  of 
Bromination  in  Oils,"  was  read  by  Dr.  C.  B.  Dudley,  Dr.  Wiley 
being  absent.  This  paper  was  discussed  by  Dr.  Dudley  and 
Prof.  McPherson. 

After  some  announcements  by  Dr.  Mabery  regarding  the  after- 
noon excursion,  a  paper  on  *' Technical  Analysis  of  Asphaltum" 
by  Miss  Laura  A.  Lynton  was  read  by  Dr.  Prescott,  and  was 
discussed  by  Prof.  Sabin,  Drs.  Mabery  and  Prescott. 

A  paper  on  *'  The  Microscopic  Detection  of  Beef  Fat  in  Lard" 
by  T.  S.  Gladding,  was  read  by  Dr.  Hart,  after  which  Prof.  P. 
W.  Clarke's  Annual  Report  on  the  Atomic  Weights  of  the  Ele- 
ments, was  read  by  Prof.  Breneman  and  discussed  by  Drs.  E.P. 
Smith,  Edward  Hart,  and  C.  B.  Dudley. 

The  report  of  the  canvassers  for  the  election  of  officers  for  the 
year  1896  was  presented  by  the  Secretary  and  the  following 


(23) 

named  persons  were  declared  elected:  President,  Chas.  B.  Dud- 
ley; General  Secretary,  Albert  C.  Hale.;  Treasurer,  Chas.  F. 
McKenna;  Librarian,  Prank  E.  Dodge.  Directors  to  serve  two 
years:  Chas.  F.  Chandler,  Peter  T.  Austen,  Chas.  E.  Munroe, 
Albert  B.  Prescott.  Councilors  to  serve  three  years:  J.  W. 
Mallet,  Albert  B.  Prescott,  T.  H.  Norton,  G.  C.  Caldwell. 

The  retiring  President,  E.  F.  Smith,  then  introduced  the 
President  elect,  Chas.  B.  Dudley,  with  a  few  congratulatory 
words  to  the  Society  in  having  secured  a  man  so  worthy  to 
occupy  the  position.  After  a  brief  and  appropriate  respbnse  by 
Dr.  Dudley,  he  was  requested  to  occupy  the  chair  while  the 
retiring  President  presented  his  address  on  "A  Glance  at  the 
Field' of  Electro-Chemistry." 

On  motion  of  Prof.  Sabin,  the  Society  passed  a  vote  of  thanks 
to  the  Mayor  and  the  Cleveland  Chamber  of  Commerce,  the 
Western  Reserve  University,  Case  School  of  Applied  Science 
and  the  Cleveland  Chemical  Society  for  their  kind  invitation  to 
hold  the  Twelfth  General  Meeting  of  the  American  Chemical 
Society  in  Cleveland,  and  for  the  courtesies  extended  to  the 
Society  during  their  meeting.  The  thanks  of  the  Society  were 
also  voted  to  the  members  of  the  Local  Committee  on  Arrange- 
ments, to  the  proprietors  and  managers  of  the  various  works 
visited,  to  those  who  received  the  chemists  and  conducted  them 
through  the  works,  and  to  the  persons  who  conducted  the 
various  excursions  and  visits. 

Upon  motion  of  Dr.  Dudley,  a  vote  of  thanks  was  g^ven  to 
the  retiring  President,  the  General  Secretary  and  the  Editor  for 
the  highly  satisfactory  manner  in  which  they  had  discharged 
the  duties  of  their  respectives  offices  and  to  those  who  had  pre- 
pared papers  for  the  meeting. 

Dr.  Mabery,  President  of  the  Cleveland  Chemical  Society 
expressed  the  great  pleasure  felt  by  the  people  of  Cleveland  at 
the  honor  the  Society  had  conferred  upon  them  in  visiting  their 
city,  and  also  the  appreciation  which  they  felt  of  the  advantages 
this  visit  had  conferred  upon  them. 

Dr.  A.  B.  Prescott,  one  of  the  Ex-Presidents  of  the  Society 
was  called  upon  by  President  Smith  for  some  remarks,  and 
spoke  briefly  of  the  rapid  growth  of  the  Society,  not  only  in 


(24) 

numbers  but  also  in  general  tone  and  character  of  its  work. 
The  Twelfth  General  Meeting  of  the  American  Chemical  Soci- 
ety was  then  adjourned. 

Albbrt  C.  Hale, 

General  Secretary. 

EXCURSION  TO  THE  WORKS  OP  THE  GRASSELLI   CHEM.  CO.* 

This  was  the  only  excursion  scheduled  for  Tuesday  afternoon » 
December  31.  It  was  joined  by  nearly  every  visiting  and  local 
chemist  and  was  in  charge  of  Mr.  Edwin  P.  Cone,  experimental 
and  research  chemist  of  the  company.  Chemists  to  the  number 
of  seventy-five  assembled  at  a  convenient  locality  and  were  trans- 
ported by  electric  cars  to  the  plant  of  the  company  located  in 
the  southern  part  of  the  city.  Here  they  were  met  by  Messrs. 
E.  R.  Grasselli,  T.  S.  Grasselli,  J.  P.  Lihme,  gentle- 
men of  the  operating  department,  and  others,  and  were 
escorted  through  the  plant.  This  company  operates  ten  different 
large  chemical  plants  in  various  parts  of  the  country,  one  of  the 
largest  being  the  works  visited  in  Cleveland. 

The  following  were  the  points  of  interest  that  were  inspected : 

Sulphuric  Acid. — Several  systems  are  operated  here  for  burn- 
ing lump  and  fine  ore,  the  latter  being  especially  adapted  for 
such  work.  Only  pirites  is  burned  obtained  from  different 
parts  of  this  country  and  abroad.  The  construction  of  these  plants 
was  found  to  be  modem  and  the  equipment  equal  to  the  best. 

In  connection  with  these  systems  are  the  concentrating  plants 
where  sulphuric  acid  in  large  quantities  is  concentrated  to  its 
various  commercial  strengths. 

Nitric  Acid, — In  this  plant  nitric  acid  was  seen  in  process 
of  manufacture  from  Chile  saltpeter  on  a  large  scale.  In  con- 
nection with  this  was  a  plant  for  the  production  of  different 
grades  of  acid  for  the  trade. 

Hydrochloric  Acid, — This  plant  comprises  various  modem 
devices  for  the  manufacture  of  numerous  qualities  of  muriatic 
acid.  Sodium  chloride  and  nitre-cake  are  used  to  a  large 
extent.  Salt-cake  from  these  plants  is  worked  up  in  large 
quantities  and  sold  to  glass  manufacturers. 

1  This  description,  written  by  B.  P.  Cone,  was  received  too  late  for  insertion  in  the 
report  of  the  General  Secretary. 


(25) 

Mixed  Acids, — In  this  department  sulphuric  and  nitric  acids 
of  pnq)er  strengths  are  mixed  in  such  proportions  as  the  trade 
demands  and  sold  in  large  quantities  to  dynamite  manufacturers. 

Glycerol. — ^This  plant  is  adapted  to  the  manufacture  of  chem- 
ically pure  glycerol,  which  is  obtained  on  the  large  scale  from 
crude  glycerol  by  distillation  with  steam.  This  department 
has  achieved  considerable  reputation  for  the  quality  of  the  prod- 
uct, which  is  equal  in  every  respect  to  any  in  the  market. 
Great  care  is  exercised  in  its  manufacture  and  many  chemical 
tests  made  to  insure  a  high-grade  article.  A  beautiful  product 
is  msrde  and  each  visitor  was  presented  with  a  small  bottle  as  a 
souvenir. 

Ammonia. — In  this  extensive  plant  large  quantities  of  ammo- 
niacal  liquor  are  worked  up  into  all  grades  of  aqua  ammonia, 
ammonium  sulphate  and  other  ammonia  products. 

Laboratories. — The  different  laboratories  were  visited  and 
chemists  were  found  busy  in  many  operations  of  interest  to  the 
analytical  and  research  chemist. 

The  extensive  shops  of  the  company  as  well  as  the  sal-soda 
and  Glauber's  salt  plants  were  also  visited. 

A  pleasant  and  agreeable  surprise  awaited  the  party  after  the 
tour  of  inspection.  In  the  work's  office  of  the  company  a  spread 
was  served,  in  every  way  adapted  to  appease  the  hunger  and 
quench  the  thirst  caused  by  the  long  tour  of  tlie  afternoon. 
After  the  cigars  had  been  passed  and  a  social  chat  indulged  in, 
the  party  were  transported  back  to  the  city  by  cars.  The  excur- 
sion was  voted  by  one  and  all  a  most  delightful  and  instructive 
one. 


BOARD  OF  DIRECTORS. 

Resolved,  That  the  Editor  be  and  he  is  hereby  instructed  to  mail 
regularly  to  the  Secretary  of  each  Local  Section  of  the  American 
Chemical  Society  a  copy  of  the  Journal  for  the  use  of  the  sec- 
tion, upon  written  request  of  the  Chairman  and  Secretary  of  the 
Section. 

NBW   MBMBBRS   ELECTED   DECEMBER    26,    1 895. 

Bartlett,  Edwin  J.,  Dartmouth  College,  Hanover,  N.  H. 
Bomberger,  F.  B.,  College  Park,  Md. 


(26) 

Booty  Johannes  Cornelius,  24  East  20th  St.,  N.  Y.  Citj. 
Gray,  Marietta,  care  of  University  of  Nebraska,  Lincoli,  Neb. 
Milliard,  H.  J.,  204  Columbia  Heights,  Brooklyn,  N.  V. 
Hollinger,  Myroen  John,  Sharpsville,  Mercer  Co.,  Pa. 
Hunicke,  H.  Aug.,  1219  Mississippi  Ave.,  St.  Louis,  Mo. 
Skinner,  W.  W.,  College  Park,  Md. 

Summers,  Bertrand  S.,  Western  Electric  Co.,  Chicago,  HI. 
Wigfall,  Edward  Newton,  1822  Arch  St.,  Phila.,  Pa. 

ASSOCIATES  KLECTBP  DECBMBBR  26,    1 895. 

AUison,  WilUam  O.,  William  St.,  N.  Y.  City. 

White,  Richard  A.,  Grand  Central  Station,  N.  Y.  City. 

NEW  MEMBERS  ELECTED  JANUARY  l8,   1 896. 

Bartow,  Edward,  Williamstown,  Mass. 
.  Battle,  H.  B.,  Ph.D.,  Raleigh,  N.  C. 

Poulk,  Chas.  W.,  B.A.,  Ohio  State  Univ.,  Columbus,  Ohio. 

Pox,  H.,  1224  Rookery  Building,  Chicago,  111. 

Graves,  George  H.,  358  State  St.,  Bridgeport,  Conn. 

Hall,  Clarence,  Aetna,  Lake  Co.,  Indiana. 

Hartwell,  Burt  L.,  B.Sc.,  Kingston,  R.  I. 

Hicks,  Edwin  P.,  52  Beaver  St.,  N.  Y.  City. 

Hopkins,  Cyril  George,  204  So.  4th  St.,  Champaign,  111. 

Magrunder,  E.  W.,  Johns  Hopkins  Univ.,  Baltimore,  Md. 

McGeorge,  Arthur,  205  West  78th  St.,  N.  Y.  City. 

Pickering,  Oscar  W,,  2  Milk  St.,  Newburyport,  Mass. 

Pitman,  John  R.,  Prankford  Arsenal,  Phila.,  Pa. 
'  Rhodes,  Edward,  Highfields,  Pordsham,  Cheshire,  Eng. 

Sargent,  Chas.  S.,  B.Sc,  Peace  Dale,  R.  I. 

Seal,  Alfred  Newlin,  1418  BouvierSt.,  Phila.,  Pa. 

Warwick,  Arthur  William,  Wickes,  Mont. 

Williams,  Charles  B.,  B.S.,  Raleigh,  N.  C. 

Woodcock,  Reginald  C,  636  West  55th  St.,  N.  Y.  City. 

Tennille,  Geo.  P.,  Ph.D.,  519  West  33rd  St.»  N.  Y.  City. 

ASSOCIATES  ELECTED  JANUARY  18,  1896. 

Brenke,  Wm.  Chas.,  506  South  5th  St.,  Champaign,  111. 
Gazzolo,  Frank  Henry,  930  West  Green  St.,  Urbana,  Ohio. 
Keeler,  Harry,  506  South  5th  St.,  Champaign,  111. 

CHANGES  OF  ADDRESS. 

Grosvenor,  W.  M.,  Jr.,   New  J.  H.   Wolfe  Hotel,  Cripple 
Creek,  Colo. 
Guild,  P.  N.,  College  of  Mont.,  De^  Lodge,  Mont. 


(27) 

V 

Johts,  John,  care  of  The  Guppinheimer  Smelting^  Co.,  Perth 
Amboj,  N.  J. 
Jones,  L.  J.  W.,  2126  High  St.,  Denver,  Colo. 
Maury,  Geo.  P.,  Braddock,  Pa. 
Mtinsell,  C.  E.,  100  Horatio  St.,  N.  Y.  City. 
Pamiiy,  Dalton,  9123  Ontario  Ave.,  Chicago,  111. 
Prochazka,  G.  A.,  138  W.  13th  St.,  N.  Y.  City. 
Rosel!,  C.  A.  O.,  841  Broadway,  N.  Y.  City. 
Townsend,  Clinton,  U.  S.  Patent  Office,  Washington,  D.  C. 
Voorhees,  S.  S.,  2101  G  St.,  N.  W.,  Washington,  D.  C. 


MEETINGS  OP  THE  SECTIONS. 

WASHINGTON    SECTION. 

A  meeting  was  held  November  14th,  1895,  President  Mun- 
roe  in  the  chair,  with  thirty-five  members  present. 

Messrs.  H.  B.  Hodges  and  Allan  Wade  Dow  were  elected  as 
members,  and  Messrs.  W.  W.  Skinner  and  J*\  B.  Bomberger  as 
local  associates. 

Dr.  Marcus  Benjamin  read  a  paper  on  *'  The  Smithsonian  In- 
stitution's Contributions  to  Chemisry  from  1846  to  1896."  He 
recalled  the  fact  that  Smithson  was  regarded  as  one  of  the  most 
expert  chemists  in  elegant  analysis  and  thought  this  fact  had 
much  to  do  with  the  provision  made  for  a  chemical  laboratory 
in  the  original  program  of  the  Smithsonian  Institution.  He 
then  traced  the  history  of  the  laboratory  of  the  institution,  men- 
tioning the  many  chemists  who  have  occupied  it  and  whose  work 
has  been  published  by  the  institution.  Among  these  were  J. 
Lawrence  Smith,  Dr.  Robert  Hare,  Edward  W.  Morley,  Genth, 
Gibbs,  Booth,  Carrington  Bolton,  Clarke,  Traphagen,  Magee, 
and  Tuckerman.  The  paper  was  concluded  with  a  bibliography 
of  the  chemical  papers  published  by  the  Smithsonian  Institution. 

Mr.  Cabell  Whitehead  presented  **Some  Notes  of  a  Recent 
Visit  to  European  Mints."  In  the  discussion  of  this  paper  ref- 
erence was  made  to  the  explosions  so  common  in  the  lighting  of 
a  Buffalo  Dental  Company's  muffle  furnace.  Mr.  Dewey  said 
that  these  explosions  can  be  avoided  by  raising  the  whole  body 
of  the  furnace  by  a  simple  arrangement  of  movable  levers  and 
then  slipping  a  lighted  paper  over  the  burner. 


(28) 

Under  the- title  "Calcium  Phosphide/*  Prof.  Chas.  E  Man- 
roe  described  the  process  of  manufacture  which  he  invented  and 
carried  into  operation  at  the  U.  S.  Naval  Torpedo  Staion  in 
1 89 1.  Iron  crucibles  were  employed  in  which  quickline  was 
heated  to  redness,  when  white  phosphorus  in  sticks  wai  added 
through  an  iron  tube  which  passed  thtough  the  cover.  The 
process  was  so  simple  that  it  was  eventually  carried  on  by 
unskilled  laborers.  The  phosphide  was  produced  at  a  cost  of 
twenty  cents  per  pound,  while  in  the  market  it  was  selling  for 
$2.25  per  pound.  It  was  manufactured  for  use  in  at±omobile 
torpedoes  while  at  practice,  and  was  found  so  efficient  that  when 
a  pound  in  its  container  was  submerged  in  eighteen  feet  of  water 
it  gave  a  flame  on  the  surface  two  feet  in  height,  wiich  con- 
tinned  to  bum  intermittently  for  three  hours. 

Discussion  was  by  Messrs.  Whitehead,  Stokes,  Kelly  and 
Fireman. 

NEW  YORK  SECTION. 

The  regular  monthly  meeting  of  the  New  York  Section  was 
held  at  the  College  of  the  City  of  New  York,  23d  street  and  Lex- 
ington avenue,  on  Friday  evening,  January  loth.  The  usual 
informal  dinner  preceded  the  meeting. 

The  meeting  was  called  to  order  at  8:30.  Prof.  P.  T.  Austin 
in  the  chair;  about  seventy  members  present.  After  the  reading 
of  the  minutes,  Mr.  Eimer  was  asked  to  describe  some  improved 
and  novel  apparatus  which  had  been  placed  on  exhibition  by 
Messrs.  Eimer  &  Amend. 

Mr.  G.  C.  Henning,  M.E.,  delegate  for  the  American  Society 
of  Mechanical  Engineers  to  the  International  Conference  at 
Zurich,  1895,  reviewed  the  **  Present  Status  of  Iron  and  Steel 
Analysis, ' '  calling  attention  to  the  discrepancies  in  some  recent 
work  of  different  chemists  in  determining  the  constituents  of 
the  same  quality  of  steel,  with  special  reference  to  carbon  and 
phosphorus,  and  to  the  omission  of  the  direct  determination  of 
iron,  which  he  thinks  conducive  to  overlooking  such  elements 
as  titanium,  tungsten  and  others,  which  are  more  often  present 
than  the  usual  iron  analysis  would  indicate,  as  they  are  bnt 
infrequently  determined  directly. 

He  reviewed  papers  by  German,  French  and  English  authors, 


(29)     . 

giving  results  of  microscopic  examination  of  iron,  and  methods 
of  preparing  the  samples  for  examination,  and  described  the 
group  of  carbon  "compounds'*  recognizable  under  the  micro- 
scope by  suitable  methods  of  surface  etching. 

He  considers  rhat  the  microscope  has  opened  a  field  which 
marks  a  great  advance  in  methods  of  determining  the  condition 
and  quality  of  iron  and  steel,  and  thinks  that  chemical  methods 
need  great  improvement  to  distinguish  the  conditions  in  which 
the  carbon  exists. 

Mr.  Rossi  in  discussing  Mr.  Henning's  paper  thought  it  would ' 
be  very  difficult,  if  not  impossible,  to  recognize  the  different 
combinations  of  iron  and  carbon  \}y  chemical  means,  at  least  in 
the  present  state  of  chemical  science,  since  there  is  so  little  out- 
side of  physical  characteristics  to  distinguish  them. 

Prof.  Breneman  asked  whether  in  a  *'  burned"  iron  the  micro- 
scope would  show  an  amount  of  magnetic  oxide  proportionate  to 
the  degree  of  deterioration  of  the  iron.  Mr.  Henning  replied 
that  this  was  practically  so  ;  that  the  oxidation  progressed  from 
the  surface  inward,  and  a  properly  polished  and  etched  specimen 
piece  would  show,  when  examined  by  powers  over  800  diame- 
ters, the  grains  of  oxide  interlaced  with  the  iron,  in  a  form 
readily  distinguishable  from  the  iron. 

Dr.  McKenna,  while  admitting  the  need  for  chemical  methods 
of  determining  the  number  and  kind  of  compounds,  is  of  the 
opinion  that  physical  methods  must  be  employed  in  conjunction 
with  chemical  methods,  and  that  while  chemical  methods  ma^*^ 
advance  greatly,  the  physical  methods  ought  never  to  be 
omitted  or  displaced.  Prof.  Breneman  suggested  that  the  manu- 
facturers would  contribute  greatly  to  the  advancement  of  the 
matter  by  having  the  expensive  chemical  investigations  required 
conducted  in  their  own  laboratories  at  the  iron  works,  where  the 
practical  side  is  already  highly  developed  and  the  material  for 
research  abundantly  supplied  ;  and  where  the  results  are  most 
wanted  and  can  be  instantly  applied.  He  also  brought  out  the 
looseness  of  the  term  "compound"  as  used  by  the  physicist, 
and  urged  the  importance  of  keeping  a  clear  distinction  between 
the  true  chemical  compound  and  the  mixtures  which  were  inac- 
curately termed  compounds.     In  reply  to  these  remarks,  Mr. 


(30) 

Henaing  said  that  several  steel  andiron  companies  in  this  cosn- 
try  have  already  established  very  complete  micrographic  labora- 
tories, where'in  three  hours  an  accurate  determination  of  the 
condition  of  any  specimen  of  the  daily  output  may  be  secured. 
Mr.  George  C.  Stone  read  a  "  Note  on  the  Piobable  Produc- 
tion of  Permanganate  1^  Direct  Combustion  of  Manganese." 

In  discussing  this  note,  Dr.  Rosell  called  attention  to  the  fact 
that  potassium  permanganate,  when  heated  to  a  red  heat,  will 
decompose,  and  that  the  other  permanganates  behave  in  the 
same  way.  In  fact,  the  permanganates  can  only  be  made  in  the 
wet  way.  On  the  other  hand,  manganates  are  generally  pro- 
duced in  the  dry  way,  and  the^  will  stand  a  very  high  temper- 
ature. 

If,  therefore,  a  substance  after  having  been  heated  to  the  tem- 
perature of  the  blast  furnace,  would  dissolve  in  pure  water  witb 
the  well-known  rich  purple  color  of  a  permanganate  solu- 
tion, it  seems  almost  certain  that  such  a  substance  could  not  be 
a  permanganate,  but  it  could  be  a  solution  of  a  ferrate. 

It  is,  of  course,  also  possible  that  the  water  used  for  dissolv- 
ing the  substance  in  question  was  not  pure,  but  accidentally 
contained  some  acid,  whereby,  on  dissolving,  the  manganate 
was  converted  into  permanganate. 

A  second  paper  by  Aft.  Stone  was  entitled  "  Remarks  on  Mr. 
Auchy's  Paper  on  the  Volumetric  Determination  of  Manganese." 
He  reviewed  the  Volhard  method  and  described  the  conditions 
under  which  he  obtained  the  most  satisfactory  results.  He 
found,  that,  provided  all  the  iron  was  oxidized,  it  made  no  dif- 
ference whether  nitric,  sulphuric,  or  hydrochloric  acid  were 
used.  The  only  difficulty  occurred  when  the  amonnt  of  manga- 
nese was  extremely  small,  in  which  case  it  was  extremely  diffi- 
cult to  get  the  precipitate  to  cohere  and  give  a  clear  solution  in 
"  ■  '   "  "     end  reaction. 

resented  a  paper  on  the  "  Manufacture  of 
Chloroform  from  Acetic  Acid,"  in  which 
y  of  acetone  from  its  first  mention  to  the 
:  was  quoted  as  mentioning  acetic  acid  as 
one,  its  vapor  being  passed  through  a  red- 
.mice  stone.     It  was  shown  that  this  sub- 


(30 

stance  was  well  known  prior  to  1848  and  had  been  made  in  large 
quantity  prior  to  1882. 

Dr.  Squibb  described  his  method  of  .preparing  acetone  by 
destructive  distillation  of  acetic  acid,  with  water  vapor  in  a 
rotary  still. 

In  regard  to  acetone-chloroform  he  quotes  Liebig  as  giving 
the  preference  to  acetone  as  the  most  suitable  compound  for  the 
preparation  of  chloroform. 

The  work  of  Bottger  and  Siemerling  was  described  and  the 
results  obtained  by  them  were  reviewed.  One-third  of  the  ace- 
tone used  was  the  largest  yield  of  chloroform  obtained  by  Bott- 
ger, its  specific  gravity  was  1. 3 1  and  it  always  contained  ace- 
tone. 

The  misleading  results  of  Siemerling's  work  were  accepted  so 
implicitly  and  quoted  so  definitely  in  standard  works  of  refer- 
ence that  the  further  prog^ss  of  the  manufacture  of  chloroform 
from  acetone  was  for  many  years  obstructed,  and  patents  have 
been  issued  in  which  the  claims  were  based  on  supposed  im- 
provements on  these  erroneous  results. 

The  last  paper  of  the  evening,  '*  Some  Notes  on  Highly  Com- 
pressed Gases,"  was  read  by  Mr.  J.  S.  Stillwell.  He  described 
some  investigations  which  had  been  made  of  certain  explosions 
of  the  containing  cylinders. 

Some  investigators  had  claimed  that  the  passage  through  a 
minute  orifice  of  light  under  high  pressure,  2,500  pounds  to  the 
square  inch,  >vould  create  sufficient  friction  and  consequent 
heating  to  cause  explosive  union  with  any  oils  or  fat  which 
might  be  present,  and  which  might  be  volatilized  by  the  men- 
tioned source  of  heat.  The  author  had,  in  the  course  of  prac- 
tical experience,  tested  this  point  over  a  hundred  thousand 
times,  and  was  satisfied  that  the  heat  never  rose  to  the  danger 
point  under  normal  conditions  of  working,  and  that  a  heat 
approaching  400"*  P.  was  necessary  before  danger  of  explosion 
need  be  feared.  This  high  temperature  of  the  comjpressed  gas 
was  never  reached,  except  through  some  careless  or  accidental 
want  of  properly  cooling  the  compressor  cylinder. 

The  meeting  adjourned  at  11 :  15. 


(32) 

LBHIGH  VALLEY  SSCTION. 

The  Annual  Meeting  of  the  Section  was  held  in  the  laboratory 
of  I^afayette  College,  Thursday,  Jan.  i6,  at  3  p.  m.  The  ballots 
for  the  election  were  opened  and  Counted  according  to  the  con- 
stitution, and  the  following  were  found  to  be  elected  for  the  en- 
suing year : 

Presiding  Officer :  Albert  Ladd  Colby. 
Secretary  and  Treasurer ;  Albert  H.  Welles. 
Member  of  Executive  Committee  ;  Edward  Hart. 

A  letter  from  the  General  Secretary  of  the  Society  of  Chemi- 
cal Industry  was  read,  thanking  the  Section  for  their  kindness 
to  Thomas  Tyrer  and  Ludwig  Mond. 

The  Secretary  was  instructed  to  furnish  abstracts  of  the  pro- 
ceedings, to  such  journals  as  might  ask  for  them. 

The  following  papers  were  presented  by  Edward  Hart :  "Note 
on  Some  Curious  Specimens  of  Zinc  Oxid" ;  '*  Note  on  a  Barium 
Blast  Furnace  Slag.'* 

He  explained  that  some  granulated  zinc  having  been  acd- 
dently  left  in  an  earthenware  crucible  in  a  muffle  over  night,  led 
to  the  discovery  of  a  most  curious  formation  of  zinc  oxide,  and 
having  designedly  repeated  the  experiments,  the  results  weit! 
exhibited. 

The  barium  blast  furnace  slag  was  from  Nova  Scotia.  The 
ore  contained  6.30  percent,  barium  stdphate,  and  the  slag  3.46 
per  cent,  barium  oxide,  as  the  chemist  of  the  company  reported 
it.  The  question  was  referred  to  Prof.  Hart  how  to  calculate 
the  barium  in  the  slag,  and  from  the  data  which  he  gave  he  con- 
cluded it  was  neutral,  the  barium  sulphate  being  reduced  to 
barium  sulphide,  and  existing  as  such  in  the  slag. 

Prof.  Richards  called  attention  to  the  notable  amount  of  alu- 
minum, w>.,  sixty-three  per  cent,  in  the  pig  iron  produced  at 
the  furnace  mentioned.  He  cited  a  case  of  a  furnace  in  the  Ju- 
niata Valley,  which,  under  abnormal  conditions  had  produced, 
as  high  as  one  per  cent,  aluminum,  although,  as  is  well  known, 
the  presence  of  metallic  aluminum  in  pig  iron  is  considered  in- 
admissible by  some  authorities. 

Mr.  Colby  suggested  that  hereafter  a  topic  be  chosen  for  the 


(33) 

evenings  discussion,  and  a  leader  be  appointed  to  open  the  dis- 
cussion and  it  was  decided  to  adopt  the  plan  at  the  next  meeting. 

Albbrt  H.  Welles, 

Secretary. 

NEBRASKA  SECTION. 

A  meeting  of  the  Nebraska  Section  was  held  on  Thursday, 
Dec.  19,  in  the  Chemical  Laboratory  of  the  University  of  Ne- 
braska.    The  meeting  was  a  pronounced  success  in  every  way. 

The  following  papers  were  read :  ( i )  *  *The  Occurrence  of  Na- 
tive Iron  in  Nebraska,'*  by  Prof.  H.  H.  Nicholson.*'  (2)  "The 
Effect  of  Freezing  on  the  Salts  in  Solution  in  Spring  and  Well 
Waters.  Preliminary  Notice,**  by  Prof.  H.  H.  Nicholson.  (3) 
••  The  Description  of  a  Shaking  Apparatus  for  Laboratory  Use,** 
by  Mr.  R.  S.  Hiltner. 

RHODE  ISLAND  SECTION. 

The  December  meeting  of  the  Rhode  Island  section  was  held 
at  Providence  on  Thursday  evening,  December  12,  1895,  Chair- 
man, Mr.  C.  A.  Catlin  presiding. 

Mr.  J.  C.  Hebden  read  a  paper  upon  *'  The  Relation  of  Acid 
and  Basic  Properties  of  the  Artificial  Dyes  to  their  Dyeing  Pro- 
perties.** 

The  paper  was  illustrated  with  diagrams  and  dyed  samples  of 
wool. 

The  January  meeting  was  held  at  Providence,  Thursday  even- 
ing, January  16,  1896,  C.  A.  Catlin  in  the  chair. 

A  paper  upon  "  Amphoteric  Reaction  of  Milk  **  was  read  by 
W.  M.  Saunders.  After  mentioning  the  results  obtained  by 
various  investigators  upon  the  subject,  the  reader  described  the 
experiments  performed  by  himself.  The  milk  of  about  seventy- 
five  cows  was  examined  as  to  the  reaction  to  litmus  paper.  The 
larger  number  gave  a  neutral  reaction  to  litmus,  the  remainder 
an  acid  or  alkaline  reaction  in  about  equal  proportion.  Cows 
giving  milk  with  an  alkaline  reaction  to  litmus  on  one  day  gave 
the  acid  reaction  a  few  days  later. 

CINCINNATI  SECTION. 

The  Section  met  in  regular  session  Wednesday,  January  15, 
1896,  President  Twitchell  presiding. 


(34) 

Mr.  P.  Homburg,  chairman  of  the  committee  appointed  to 
draft  resolutions  on  the  death  of  Prof.  C.  R.  Stuntz,  reported  the 
following : 

"  Since  it  has  pleased  Providence  to  call  from  his  labors  to 
rest,  Prof.  C.  R.  Stuntz,  we,  the  members  of  the  Cincinnati  Sec- 
tion of  the  American  Chemical  Society,  through  our  committee, 
desire  to  express  our  deep  sorrow  at  the  loss  of  our  esteemed 
friend  and  colleague,  and  also  our  sincere  sympathy  with  all  who 
mourn  his  death. 

*'  His  genial  disposition,  his  courteous  bearing,  his  devotion 
to  science  and  learning  in  general,  and  to  the  success  of  our  or- 
ganization in  particular,  we  keenly  appreciate. 

*' The  legacy  of  his  noble  example  will  tend  to  alleviate  the 
distress  caused  by  his  departure. 

**p.  homburg, 
"Dr.  S.  p.  Kramer, 

"E.  TWITCHBLL, 

**H.  B.  FooTK, 

**  Committee." 

On  motion  of  Dr.  Springer,  the  resolutions  were  adopted,  and 
the  committee  was  instructed  to  send  a  copy  to  the  family  of 
the  deceased. 

Mr.  H.  B.  Schmidt,  of  Cincinnati,  was  elected  a  member  of 
the  Section. 

Papers  were  read  on  * 'Mercury:  Its  Occurrence  and  Produc- 
tion," by  Frank  I.  Shepherd;  and  "A  Few  Noteson  the  Deter- 
mination of  Ircad,''  by  J.  Hayes-Campbell. 


iMQcd  with  March  Nnmber,  1896. 


Proceedings, 


COUNCIL. 

The  following  persons  have  been  elected  as  members  of  the 
standing  committees  for  one  year  : 

Committee  on  Papers  and  Publications — ^J.  H.  Long  and 
Thomas  B.  Osborne. 

Committee  on  Nominations  to  Membership — A.  A.  Breneman, 
P.  T.  Austen,  and  C.  A.  Doremus. 

Finance  Committee — Durand  Woodman,  A.  P.  Hallock,  and 
A.  H.  Sabin. 

C.  F.  Mabery  has  been  elected  a  member  of  the  Council  for 
1896  in  place  of  Charles  B.  Dudley,  President. 

The  bills  of  the  Chemical  Publishing  Co.  for  $289.24  for  the 
January  number  and  $272.43  for  the  February  number  of  the 
Journal,  have  been  approved. 

NEW  MBMBBRS  BLBCTBD  FEBRUARY  I,  1 896. 

Bookman,  Samuel,  9  East  62nd  St.,  N.  Y.  City. 

FuUam,  Frank  L.,  cor.  Gold  and  John  Sts.,  Brooklyn,  N.  Y. 

Hanks,  Abbot  A.,  718  Montgomery  St.,  San  Francisco,  Cal. 

Lihme,  I.  P.,  27  Tift  Ave.,  Cleveland,  O. 

Lippincott,  Warren  B.,  3179  Ashland  Ave.,  Chicago,  111. 

Maywald,  F.  J.,  592  Kosciusko  St.,  Brooklyn,  N.  Y. 

Leret,  Fred.,  Virginia,  St.  Louis  Co.,  Minn. 

Sharpless,  Fred.  F.,  811  Wright  Block,  Minneapolis,  Minn. 

Steams,  F.  C,  M.D.,  44  Montgomery  St.,  Jersey  City,  N.  J. 

ASSOCIATE  ELECTED  FEBRUARY  I,  1 896. 

Gordon,  Alexander,  44  Montgomery  St.,  Jersey  City,  N.  J. 

NEW  MEMBERS  ELECTED  FEBRUARY  24,  1 896. 

Baker,  Theodore,  Box  97,  Belford,  N.  J. 
Barrett,  Jesse  M.,  Purdue  University,  Lafayette,  Ind. 
Borland,  Chas.  R.,  E.  C.  Powder  Co.,  Oakland,  Bergen  Co., N.J. 
Cheney,  John  P.,  So.  Manchester,  Conn. 
Christiansen,  H.  B.,  Hermitage,  Floyd  Co.,  Ga. 
Jones,  Wm.  J.,  Jr.,  Purdue  University,  Lafayette,  Ind. 
Martin,  Alex.  M.,  F.C.S.,  Douglas  Villa,  Dunbeth,  Road, 
Coatbridge,  Scotland. 


(36) 

Myers,  H.  Ely,  Riddlesburg,  Bedford  Co.,  Pa. 
Slagle,  Robert  Lincoln,  Brookings,  S.  D. 
Smyth,  Dr.  Geo.  A.,  900  South  Boulevard,  Oak  Park,  III. 
Tidball,  Walton  C,  care  of  E.  R.  Squibb  &  Sons,  Gold  and 
John  streets,  Brooklyn,  N.  Y. 

ASSOCIATES  BLBCTBD  FEBRUARY  24,  1896. 

Pomeroy,  Thomas  W.,  Lafayette  College,  Easton,  Pa. 
Stover,  Edward  C,  Trenton  Potteries  Co.,  Trenton,  N.  J. 

CHANGES  OP   ADDRESS. 

Atkinson,  Elizabeth  A.,  Three  Tons,  Pa. 

Baekeland,  Leo.,  care  Nepera  Chem.  Co.,  Nepera  Park,  N.Y. 

Blalock,  Thos.  L.,  3106  O'Donnell  St.,  Baltimore,  Md. 

Bromwell,  Wm.,  Ph.D.,  care  Tenn.  C.  I.  and  R.  Co.,  1918- 
1920  Morris  Ave.,  Birmingham,  Ala. 

Campbell,  Geo.  F.,  80  Bristol  St.,  New  Haven,  Conn. 

Chamberlain,  G.  D.,  care  N.  W.  Mall  Iron  Co.,  Milwaukee, 
Wis. 

Comelison,  R.  W.,  care  McKenzie  Bros.  &  Hill,  Bloomfield. 
N.J. 

Doremus,  Dr.  C.  A.,  17  Lexington  Ave.,  N.  Y.  City. 

Foote,  Henry  B.,  241  Walnut  St.,  Cleveland,  Ohio. 

Graves,  W.  G.,  1661  Huron  St.,  Cleveland,  Ohio. 

Kenan,  Wm.  R.,  Jr.,  care  Carbide  Mfg.  Co.,  box  45,  Niagara 
Falls,  N.  Y. 

Kiefer,  H.  E.,  16  W.  4th  St.,  South  Bethlehem,  Pa. 

Morse,  Fred.  W.,  Lock  Box  30,  Durham,  N.  H. 

Spencer,  G.  L.,  Centralia,  Wood  Co.,  Wis. 

Trubek,  M.,  325  Academy  St.,  Newark,  N.  J. 

Walker,  Henry  V.,  38-40  Clinton  St.,  Brooklyn,  N.  Y. 

ADDRESS  WANTED. 

Johnson,  Jesse,  last  address  Augusta,  Ga. 


MEETINGS  OF  THE  SECTIONS. 

WASHINGTON    SECTION. 

The  regular  monthly  meeting  of  the  Washington  Section  was 
held  December  12,  1895,  President  Munroe  m  the  chair,  with 
thirty-six  members  president.  In  the  absence  of  the  Secretar3', 
W.  D.  Bigelow  was  elected  Secretary,  pro  tempore.  The  follow- 
ing were  elected  to  membership  :  W.  W.  Skinner,  F.  B.  Bone- 
berger,  and  H.  Carrington  Bolton.      A  committee  was  appointed 


(37) 

to  arrange  for  a  social  meeting  of  the  Section  to  report  at  the 
Jafiiiary  meeting. 

The  first  paper  of  the  evening  was  '*  Exhibition  of  Argon  and 
Helium,"  by  Dr.  W.  F.  Hillebrand.  He  discussed  concisely 
the  spectra  of  argon  and  helium  and  closed  by  exhibiting  the 
spectra  to  the  Society. 

The  second  paper  was  by  Dr.  H.  W.  Wiley,  on  the  **  Use  of 
Acetylene  Illumination  in  Polariscope  Work,  with  Illustrations." 
Dr.  Wiley  stated  that  acetylene,  while  not  inferior  in  point  of 
accuracy  to  other  forms  of  illumination,  is  so  intense  as  to  per- 
mit accurate  polarization  with  solutions  so  dark  in  color  that 
they  cannot  be  polarized  with  lights  commonly  used  for  this 
purpose.  He  called  attention  to  the  **  Schmidt  and  Haensch 
Triple  Field  Polariscope,"  which  was  said  to  be  a  great  assis- 
tance in  both  rapid  and  accurate  work.  The  paper  was  illus- 
trated with  the  acetylene  light  and  the  polariscope  referred  to. 

Mr.  F.  P.  Dewey  read  a  paper  on  **  The  Early  History  of 
Electric  Heating  for  Metallurgical  Purposes."  The  paper  was 
comprehensive,  embracing  the  various  patents  relating  to  elec- 
tric heating  for  metallurgical  purposes  and  also  many  relating  to 
electric  reduction.  It  was  illustrated  by  photographs  and  draw- 
ings of  the  various  forms  of  apparatus  described. 

The  last  paper  of  the  evening  was  **  A  Tribute  to  the  Memory 
of  Josiah  P.  Cooke,"  by  Dr.  Marcus  Benjamin.  An  excellent 
portrait  of  Prof.  Cooke  was  exhibited  and  the  sketch  of  his  life 
was  of  special  interest  from  the  fact  that  the  statements  made 
were  from  a  manuscript  sent  to  Dr.  Benjamin  some  years  ago  by 
Prof.  Cooke.  After  discussion  by  Messrs.  Munroe,  Tassin,  and 
Wiley,  the  Section  adjourned. 

NEW  YORK   SECTION. 

The  regular  meeting  of  the  New  York  Section  was  held  at  the 
College  of  the  City  of  New  York  on  Friday  evening,  Feb.  7,  at 
8.30  p.  M.  The  following  papers  were  read  :  **  New  Facts  about 
Calycanthus,"  by  Dr.  R.  G.  Eccles,  and  **  Items  of  Interest 
from  the  Cleveland  Meeting,"  by  A.  A.  Breneman. 

Dr.  Eccles  described  his  work  and  also  that  of  Dr.  H.  W. 
Wiley   on  the  calycanthus  seeds  and  the  alkaloids  obtained 


(38) 

therefrom ;  exhibiting  the  seeds,  the  principal  alkaloid  obtained, 
its  salts,  the  color  reactions  of  both,  and  the  crystalline  forms*of 
the  salts. 

Prof.  Breneman  described  the  features  of  the  Cleveland  meet- 
meeting,  which  were  of  particular  interest  to  industrial  chemists, 
referring  especially  to  the  low  pressure  distillation  of  light  pe- 
troleum oils  as  conducted  in  Prof.  Mabery's  specially  equipped 
laboratory. 

Dr.  Durand  Woodman  exhibited  a  simple  lecture  table  appa- 
ratus for  experimentally  demonstrating  the  luminosity  of  the 
acetylene  flame,  generating  the  gas  from  calcium  carbide. 

The  meeting  was  adjourned  at  10.45  p.  m. 


ANNUAL  REPORTS  OP  THE  SECTIONS. 

The  following  annual  reports  from  the  secretaries  of  the  sec- 
tions were  received  by  the  General  Secretary  too  late  for  inser- 
tion in  their  proper  place : 

WASHINGTON    SECTION. 

Seven  meetings  have  been  held  and  an  abstract  appended 
gives  the  list  of  papers  read  and  topics  discussed  at  these  meet* 
ings.    The  following  is  a  list  of  the  present  officers : 

President — Charles  E.  Munroe. 

Vice  Presidents — E.  A.  de  Schweinitz  and  W.  D.  Bigelow. 

Treasurer— W.  P.  Cutter. 

Secretary — A.  C.  Peak. 

The  officers  as  above  with  the  following  constitute  the  Execu- 
tive Committee :  H.  W.  Wiley,  F.  P.  Dewey,  F.  W.  Clarke,  and 
W.  H.  Seaman.     There  are  no  other  standing  committees. 

The  secretary  of  the  local  section  has  no  way  of  determining 
the  standing  of  members.  According  to  a  statement  made  by 
the  General  Secretary,  December  8,  1894,  the  membership  of 
the  Washington  section  was  sixty-four.  As  it  now  appears  to 
be  seventy -four  the  gain  during  the  year  is  ten. 

November  8,  1894. — President  W.  H.  Seaman  in  the  chair. 
Ten  members  present.  Resignation  of  Prof.  J.  C.  Gordon  read 
and  accepted.     Cooperation  of  the  Society  asked  by  John  W. 


(39) 

Hoyt  in  the  formation  of  a  "  National  Posi-Graduate  Univer* 
sity."  Prof.  H.  W.  Wiley  made  a  report  on  the  •*  First  Congress 
of  Chemists,"  at  the  San  Francisco  exposition.  Paper  read  by 
W.  D.  Biglow  on  the  "  Coloring-Matter  in  California  Red 
Wines." 

December  ij,  18^4. — President  W.  H.  Seaman  in  the  chair 
Twenty  members  present.  Paper  by  Oma  Carr  and  J.  F.  San- 
bom  on  the  "  Dehydration  of  Viscous  Organic  I^iqnids,"  read 
by  Mr.  Carr.  Mr.  W.  D.  Bigelow  and  E.  £.  Ewell  described  a 
continuous  extractor  for  large  quantities  of  material. 

January  10^  /<^95. — President  W.  H.  Seaman  in  the  chair. 
Fourteen  members  present.  The  following  officers  were  elected : 
President,  Charles  B.  Munroe;  Vice  Presidents,  E.  A.  de 
Schweinitz  and  W.  D.  Bigelow;  Treasurer,  W.  P.  Cutter; 
Secretary,  A.  C.  Peale.  Additional  members  of  the  Executive 
Committee,  H.  W.  Wiley,  F.  P.  Dewey,  F.  W.  Clarke,  and  W. 
H.  Seaman.  H.  C.  Sherman,  F.  P.  Veitch,  W.  G.  Brown,  and 
V,  K.  Chesnut  were  elected  to  membership. 

February  14,  18^3, — ^The  meeting  was  devoted  to  the  annual 
address  of  the  retiring  president,  W.  H.  Seaman,  upon  ''  Chem- 
istry in  Education."  President  Charles  E.  Munroe  in  the  chair, 
with  members  of  the  Society  and  invited  guests  from  the  Socie- 
ties of  Washington  present. 

March  14,  J8gs. — President  Charles  E.  Munroe  in  the  chair. 
Thirty-five  members  present.  Dr.  J.  E.  Blom€n  and  G.  E.  Bar» 
ton  elected  to  membership.  The  following  papers  were  read : 
**The  Constitution  of  the  Silicates,"  by  F.  W.  Clarke.  **  On 
the  Chloronitrites  of  Phosphorus  and  the  Metaphosphinic  Acids,  * ' 
by  Dr.  H.  N.  Stokes;  **  The  Manufacture  of  Soluble  Nitrocel- 
lulose for  Nitrogelatin  and  Plastic  Dynamite,"  by  Dr.  J.  E. 
Blom6n. 

April  II y  i8g5, — President  Charles  E.  Munroe  in  the  chair. 
Fifty-three  members  present.  The  following  papers  were  read : 
**  The  Determination  of  Nitrogen  in  Fertilizers,"  by  H.  C.  Sher- 
man; **  Exhibition  of  Calcium  Carbide,"  by  Charles  E.  Mun- 
roe ;  "Precipitation  of  Small  Quantities  of  Phosphoric  Acid  by 
Ammoniacal  Citrate  of  Magnesia,"  by  E.  G.  Runyan  and  H.  W. 
Wiley.     The  subject  for  discussion  was  *  *  Can  Argon  be  Accepted 


(40) 

as  a  New  Element?"  Discussion  was  by  Charles  K.  Munroe, 
F.  W.  Clarke,  T.  M.  Chatard,  and  H.  N.  Stokes. 

May  p,  18^5, — President  Charles  E.  Munroe  in  the  chair. 
Forty  members  present.  Messrs.  Marion  Dorset  and  S.  C.  Mil- 
ler elected  to  membership.  The  following  papers  were  read  : 
**ANew  Meteorite  from  Forsyth  Co.,  N.  C,"  by  E.  A.  de 
Schweinitz ;  **  Hydrogen  Fluoride  Poisoning,"  by  Peter  Fire- 
man; *'  Progress  in  the  Manufacture  of  Artificial  Musk,"  by 
W.  H.  Seaman.  The  subject  for  discussion  was.  "The  Chemi- 
cal Action  of  Micro-organisms."  and  was  participated  in  by  H. 
A.  de  Schweinitz,  Surgeon  General  Sternberg,  H.  W.  Wiley, 
Prof.  George  P.  Merrill,  and  R.  B.  Warder. 

The  Society  adjourned  until  November. 

CINCINNATI   SECTION. 

The  annual  election  held  December  i8th,  1894,  resulted  as 
follows  : 

President,  Karl  Langenbeck  ;  Vice-Presidents,  B.  D.  Westen- 
felder  and  I.  J.  Smith  ;  Treasurer,  Henry  B.  Foote  ;  Secretary, 
E.  C.  Wallace ;  Directors,  Dr.  S.  P.  Kramer,  Prof.  O.  W.  Martin, 
H.  L.  Nickel. 

The  following  were  elected  chairmen  of  the  standing  commit- 
tees for  the  year : 

1.  Didactic  Physical  and  Inorganic  Chemistry,  Dr.  Alfred 
Springer. 

2.  Organic  Chemistry,  Prof.  T.  H.  Norton. 

3.  Analytical  Chemistry,  Lewis  William  Hoffmann. 

4.  Medical,  Physiological  and  Biological  Chemistry,  Dr.  S.  P. 
Kramer. 

5.  Technical  and  Pharmaceutical  Chemistry,  Prof.  J.  U. 
Lloyd. 

The  following  named  persons  have  been  elected  members  of 
this  Section  since  October  31,  1894  :  W.  G.  Wallace,  Richard 
W.  Proctor,  Charles  E.  Jackson  and  George  F.  Feid,  elected 
December  18,  1894;  F.  Homburg,  E.  D.  Frohman,  elected  Jan- 
uary 15,  1895  ;  Harry  L.  Lowenstein,  elected  February  15,  1895  ; 
Prof.  A.  F.  Linn,  and  Dr.  JohnMcCrae,  elected  October  15, 1895. 

In  the  death  of  Lewis  William  Hoffmann  and  W.  G.  Wallace 


(41) 

the  Cincinnati  Section  sustained  a  loss  of  two  of  its  popular 
younger  members,  who  were  highly  esteemed  by  their  associates. 

Eight  meetings  were  held  during  the  year,  at  which  the  follow- 
ing papers  were  presented.  Special  meeting  held  November  7, 
1894,  addressed  by  Dr.  H.  Hensoldt.  The  subject  announced, 
* 'Occult  Science  in  the  Orient.** 

Stated  meeting  December  18,  1894  :  **  Diphtheria  Antitoxin,*' 
Dr.  S.  P.  Kramer;  **  Elective  Fermentation  in  Diabetes,**  Dr. 
Alfred  Springer. 

.Stated  meeting  January  15,  1895:  *' Separation  of  the  Solid 
and  Liquid  Fatty  Acids,**  E.  Twitchell ;  *' Report  of  Progress 
in  Organic  Chemistry,**  Dr.  H.  E.  Newman. 

Stated  meeting  February  15,  1895  :  Papers  announced  were 
postponed  on  account  of  sickness  of  the  essayists.  '*  The  Diffi- 
culty of  Obtaining  Distilled  Water  to  Meet  Pharmacopeial  Re- 
quirements,** was  discussed  by  Prof.  Lloyd,  Dr.  Springer  and 
Prof.  Norton. 

Stated  meeting  March  15,  1895:  ''Determination  of  Phos- 
phorus in  Ferrosilicon,**  John  H.  Westenhoff ;  **  The  Souring 
of  Milk,**  Robert  W.  Hochstetter. 

Meeting  April  16,  1895  :  "  Recent  Important  Discoveries  in 
Chemistry,**  Prof.  T.  H.  Norton. 

Meeting  May  15,  1895:  "Adulteration  of  Powdered  Elm 
Bark,**  Henry  B.  Foote;  "Ammonium  Thioacetate,'*  Prof.  T. 
H.  Norton. 

Stated  meeting  October  15,  1895:  **A  Tribute  to  Pasteur,** 
Dr.  Alfred  Springer  ;  "  Laboratory  Uses  of  Aluminum  and  Re- 
cent Progress  in  Theoretical  Chemistry,*'  Prof.  T.  H.  Norton. 

A  pamphlet  issued  by  the  Executive  Committee  gives  names, 
occupation  and  addresses  of  members  of  the  Section,  the  names 
of  the  authors  and  titles  of  papers  read  during  1894. 

NEBRASKA  SECTION. 

The  Nebraska  Section  was  organized  at  a  meeting  held  in 
Lincoln,  June  14,  at  which  meeting  officers  were  selected  for  the 
ensuing  year,  as  follows  : 

President,  H.  H.  Nicholson  ;  Secretary  and  Treasurer,  John 


(42) 

White ;  Executive  Committee,  H.  H.  Nicholson,  John  White, 
Rosa  Bouton,  T.  L.  I^yon,  W.  S.  Robinson. 

The  first  regular  meeting  of  the  Section  was  held  in  the  Chemi- 
cal Laboratory  of  the  University  of  Nebraska  on  October  30, 
with  a  good  attendance  of  members  and  a  number  of  invited 
guests. 

Papers  were  read  as  follows  : 

By  Prof.  T.  L.  Lyon  :  **  The  Source  of  Error  in  the  Estima- 
tion of  Sugar  in  Beet  Juice  by  Means  of  the  Sucrose  Pipette.'* 

By  Mr.  Samuel  Avery  :  **  Notes  on  the  Electrolytic  Determi- 
nation of  Iron,  Nickel  and  Zinc." 

Mr.  C.  H.  Suveau,  of  the  Department  of  Motive  Power  of  the 
Burlington  and  Missouri  River  Railroad,  was  elected  a  local  as- 
sociate member. 

Other  meetings  will  be  held  in  December,  March  and  Jun^. 

Our  present  membership  is  thirteen. 

CHICAGO  SBCTION. 

The  Chicago  Section  has  held  but  two  meetings,  one  for 
organization,  and  the  other  just  reported,  at  which  papers  were 
presented. 

The  membership  is  twenty-five. 

The  officers  are  as  follows  : 

President,  Frank  Julian  ;  Vice-President,  J.  C.  Foye  ;  Secre- 
tary, P.  B.  Dains  ;  Treasurer,  J.  H.  Long ;  Executive  Commit- 
tee, Frank  Julian,  A.  L.  Smith,  F.  B.  Dains. 

NEW  YORK  SECTION. 

Meetings  were  held  and  papers  read  as  follows  : 

November  g^  18^4:  **The  Rapid  and  Accurate  Analysis  of 
Bone-black,"  by  William  D.  Home;  "Recent  Progress  in 
Physiological  Chemistry,**  by  Dr.  E.  E.  Smith. 

December  ij J  1894:  The  Chemical  Nature  of  Diastase,"  by 
Thomas  B.  Osborne,  of  New  Haven ;  "  Glucose  from  a  Sani- 
tary  Standpoint,"  by  E.  H.  Bartley,  M.D. ;  ''Indiscriminate 
Taking,"  by  P.  T.  Austen. 

January  10^  ^^95  '  "Improvement  in  the  Manufacture  of 
Acetone,"  by  Dr.  E.  R.  Squibb  ;  **  Recent  Progress  in  Photo- 
graphic Chemistry,"  by  Dr.  J.  H.  Stebbins. 


(43) 

February  i8y  i8g^  :  No  quorum. 

March  S,  iSpj:  **  The  Late  Prof.  Henry  B.  Nason,"  by  W.  P. 
Mason;  ''Elective  Fermentation  in  Diabetes,"  by  Alfred 
Springer ;  **  Note  on  Absorbent  Blocks,"  by  W.  H.  6n>ad- 
hurst ;  **  Note  on  the  Precipitation  of  Iron  by  Alkali  Nitrites," 
by  GiUett  Wynkoop ;  '' Volumetric  Determination  of  Zinc  and 
Manganese,  and  a  New  Indicator  for  Ferrocyanides,"  by  G.  C. 
Stone;  **  Note  on  the  Reduction  of  Nitrates  by  Ferrous 
Hydroxid,"  by  P.  T.  Austen  ;  "  Stability  to  Light  of  Haematon- 
ylin  Blacks  on  Wool,"  by  P.  T.  Austen. 

May  ij,  iSpj:  **  Recent  Progress  in  Analysis  of  Soils,"  byH. 
W.  Wiley  ;  **  Tribute  to  the  Memory  of  Dr.  Gideon  Moore,"  by 
C.  P.  McKenna  ;  "  Chemical  History  of  a  case  of  Arsenical  and 
Antimonial  Poisoning,"  by  C.  A.  Doremus  ;  "The  Estimation  of 
Acetic  Acid  in  Vinegar,"  by  A.  R.  Leeds. 

/une  14,  i8gs  '-  **  Determination  of  Nitrogen  by  the  Gunning 
Method,"  by  W.  D.  Field  ;  "On  Asbestos  and  its  Commercial 
Application,"  by  G.  C.  Stone  ;  **  Examination  of  Lard  for  Im- 
purities," by  David  Wesson  ;  "On  Commercial  Argol  and  its 
Products,"  by  Wm.  McMurtrie ;  "A  Modem  View  of  Electro- 
Chemical  Action,"  by  C.  L.  Speyers ;  "  On  the  Relation  of  the 
Chemical  Engineer  to  Factory  Management,  "by  John  Enequist. 

The  informal  dinners  preceding  the  meetings  have  been  con- 
tinued at  a  majority  of  the  meetings.  The  total  expenditures  of 
the  section  have  amounted  to  I157.98  ;  those  of  the  preceding 
year  were  $160.82.  The  largest  item  in  each  case  is  the  con- 
tribution to  the  Treasury  of  the  Scientific  Alliance. 

The  following  officers  have  been  elected  for  the  current  year  ; 
Chairman,  P.  T.  Austen;  Secretary  and  Treasurer,  Durand 
Woodman  ;  Executive  Committee,  A.  H.  Sabin,  A.  C.  Hale, 
A.  R.  Leeds ;  Delegates  to  Council  of  Scientific  Alliance,  P.  T. 
Austen,  C.  F.  McKenna,  A.  C.  Hale. 

The  list  of  members  is  annexed  hereto,  and  shows  a  total 
membership  of  234  as  compared  with  183  last  year,  or  a  gain  of 
fifty-one  members. 

RHODE  ISLAND  SBCTION. 

The  Rhode  Island  Section  of  the  American  Chemical  Society 


(44) 

respectfully  transmits  the  following  general  report  of  the  busi- 
ness of  the  Section  for  the  year  September  i,  1894,  to  Septem- 
ber I,  1895. 

The  work  of  the  Rhode  Island  Section  for  the  past  year  may 
be  described  in  brief  as  follows,  all  the  meetings  having  been 
held  in  Providence : 

September  ^7,  18^4  :  A  paper  prepared  by  Prof.  E.  E.  Calder 
upon  the  **Chemistry  of  Albuminurea,"  was  read  by  Mr.  W.  M. 
Saunders,  the  author  being  unable  to  be  present. 

October  18 ^  18^4 :  A  paper  was  read  by  Mr.  W.  M.  Saunders 
upon  '*  Lantern  Slides  and  their  Preparation,"  illustrated  by  the 
stereopticon. 

December  ij,  18^4  :  A  paper  was  read  by  Charles  A.  Catlin, 
upon  **  Bread  and  Bread  Stuffs." 

January  77,  i8g§ :  A  paper  was  read  by  Mr.  Geo.  F. 
Andrews  upon  **  The  Accuracy  of  the  fire  assay  of  Silver." 

February  2j,  1893  :  A  paper  was  read  by  Prof.  J.  H.  Appleton 
upon  **  Argon." 

March  21,  i8gs  '  A  paper  was  read  by  Mr.  J.  P.  Famsworth 
upon  *'  Selection  of  water  for  Bleaching  and  other  manufactur- 
ing purposes." 

Aprit  24,  i8g§  :  A  paper  was  read  by  Mr.  H.  C.  Burgess  upon 

'*  A  Resum^  of  the  methods  of  Bleaching  Cotton  Piece  Goods." 

May  2j^  189s  :  A  paper  was  read  by  Mr.  E.  D.  Pearce  upon 
**  The  coloring-matter  of  Pollens,"  with  illustrations  under  the 
microscope." 

June  i^^  i8gs:  The  Annual  Meeting  was  held  at  the  Hope 
Club  House  where  the  members  were  entertained  at  dinner  as 
the  guests  of  the  Chairman,  Mr.  Charles  A.  Catlin,  who  pre- 
sented a  paper  upon  Chemical-Laboratory  Microscopy,  illus- 
trated by  the  microscope. 

The  interest  in  the  local  section  still  continues  to  be  well  sus- 
tained, and  already  another  new  year  of  its  work  has  begun 
with  very  flattering  prospects  for  the  future. 

At  present  date  the  names  of  the  officers  of  the  Section  are  : 
Chairman,  Charles  A.  Catlin ;  Secretary  and  Treasurer,  Wal- 
ter M.  Saunders  ;  Executive  Committee,  Chairman,  ex-officio, 
Secretary  and  Treasurer  ex-officio,  George  F.  Andrews. 


(45) 

Number  of  members  belonging  to  the  Rhode  Island  Section 
at  the  present  time,  twenty  (20).  Net  increase  over  last  year 
two  (2). 

LEHIGH  VALLEY  SECTION. 

It  has  been  found  desirable,  as  the  Section  is  limited  in  num- 
ber, to  hold  fewer  meetings.  Four  meetings  have  been  held 
during  the  past  year;  viz,^  November  i,  1894,  January  17,  May 
2,  and  October  10,  1895.  The  October  meeting  was  the  most 
successful  in  the  history  of  the  Section,  the  Society  entertaining 
at  that  time  Thomas  Tyrer,  Esq.,  President  of  the  Society  of 
Chemical  Industry  of  England,  and  invited  representatives  of 
the  New  York  Section  of  the  same  Society  and  the  New  York 
Section  of  the  American  Chemical  Society.  An  inspection  of 
the  large  government  plant  of  the  Bethlehem  Iroti  Co.  was 
made,  followed  by  an  elegant  dinner  tendered  the  visiting  chem- 
ists, while  the  stated  meeting  was  held  in  the  afternoon  at 
Lehigh  University. 

The  following  papers  have  been  presented:  **  Helmholz's 
Contributions  to  Science,"  George  P.  Scholl;  '*  Castner's  Elec- 
trolytic Process  for  Production  of  Caustic  Soda,"  W.  H.  Chand- 
ler; "A  New  Ammonia  Condenser,"  Edward  Hart;  **The 
Determination  of  Graphite  in  Pig  Iron,"  P.  W.  Shimer  ;  **  The 
Selection  of  Samples  for  Analysis,"  A.  L.  Colby ;  **  On  Stand- 
ardization of  Iodine  Solution,"  G.  H.  Meeker;  **  A  Device  for 
Sampling  Metals,"  P.  W.  Shimer  ;  **  The  Rapid  Methods  Used 
in  the  Bethlehem  Iron  Co.'s  Laboratory,"  A  L.  Colby. 

Mention  should  also  be  made  of  the  interesting  paper  read  by 
Dr.  William  McMurtrie  at  the  October  meeting,  on  **  Chemical 
vs.  Bacteriological  Examination  of  Water,"  written  by  Prof.  W. 
P.  Mason. 

We  close  the  year  with  the  same  number  as  we  began  ;  w>., 
21,  losses  having  been  made  up  by  new  members. 

Our  annual  meeting  will  be  held  the  third  Thursday  in  Janu- 
ary.    The  ofl&cers  of  the  Section  for  the  year  1895  are  as  follows : 

Presiding  OflScer — Edward  Hart. 

Treasurer — Albert  L.  Colby. 

Secretary— Albert  H.  Welles, 


(46) 

Executive  Committee — Edward  Hart,  Albert  L.  Colby,  Albert 
H.  Welles,  and  J.  W.  Richards. 

THE  CLEVELAND  EXCURSIONS. 

Owing  to  delay  in  the  receipt  of  copy  it  was  impossible  to  give 
a  full  account  of  the  meeting  at  Cleveland  in  last  month's  issue. 
The  following  additional  matter  has  since  been  received  by  the 
editor : 

One  very  pleasing  feature  of  the  meeting  in  Cleveland  was  the 
excursions  to  various  works  and  other  places  of  interest.  Both 
afternoons  were  set  apart  for  this  purpose,  one  special  part^*^  made 
a  trip  on  Tuesday  morHiog,  and  on  Monday  evening  the  Case 
School  of  Applied  Science  and  Adelbert  College  were  visited. 
Such  a  large  number  of  excursions  were  planned,  and  so  many 
works  were  freely  opened  for  inspection,  that  it  was  impossible 
to  visit  them  all.  Routes  Nos.  i,  6,  and  7,  as  given  by  the 
local  comrilittee's  program,  were  omitted,  and  all  the  time 
available  was  devoted  to  the  others. 

Route  No.  2  was  led  by  Mr.  D.  B.  Cleveland,  chemist  of  the 
American  Wire  Works. 

The  first  place  visited  was  the  plant  of  the  Otis  Steel  Co. 
Mr.  Bartol,  the  Superintendent,  took  charge  of  the  visitors  and 
showed  them  the  plate  mill,  the  basic  open  hearth  furnaces,  the 
steel  foundry,  the  car  axle  foundry,  the  machine  shop  and  the 
laboratory. 

The  Continental  Chemical  Co;  was  next  visited.  They  make 
red  pigments  and  fuming  sulphuric  acid  from  copperas,  but  had 
unfortunately  been  recently  burned  out,  so  that  it  was  impos* 
sible  to  see  the  works. 

The  manager,  Dr.  Ramage,  exhibited  an  apparatus  for  mak- 
ing ozone,  which  he  said  brought  its  cost  down  to  such  a  point 
as  to  warrant  its  being  used  for  disinfecting  garbage,  refuse 
from  stock-yards,  fats,  etc. ;  for  bleaching,  for  oxidizing  sul- 
phurous to  sulphuric  oxide,  thus  doing  away  with  the  lead 
chambers  in  the  manufacture  of  sulphuric  acid,  and  for  many 
other  purposes. 

He  claimed  that  treatment  with  ozone  is  an  almost  sure  cure 
for  consumption  in  the  first  two  stages  and  for S3rphilitic  diseases. 


(47) 

He  stated  that  with  one-fifth  to  one-seventh  of  a  horse  power 
he  can  change  one  hundred  and  twenty  cubic  feet  of  atmos- 
pheric air  a  minute  to  a  mixture  containing  fifteen  to  eighteen 
per  cent,  of  ozone,  all  the  oxygen  present  being  changed.  He 
uses  a  fifty  volt  and  two  to  three  ampere  alternating  current, 
which  he  converts  to  a  current  of  fifty  thousands  volts. 

The  Cleveland  Nitrous  Oxide  Co.  was  next  visited,  and  Mr. 
Clark  and  Mr.  Hatch  showed  the  party  around. 

They  sell  oxygen  and  hydrogen,  epsom  salts,  liquid 
nitrous  oxide  and  carbonic  acid  gas.  They  also  make  nitrate 
of  ammonia,  as  the  commercial  salt  is  too  impure  for  their  use. 

They  exhibited  some  acetylene  gas  burning  from  an  ordinary 
gas  jet  to  show  the  character  and  illuminating  power  of  this 
much  talked-of  new  illuminant. 

By  this  time  it  was  too  late  to  visit  other  places  of  interest, 
and  the  party  returned  to  the  hotel. 

Route  No.  3  was  taken  on  Tuesday  morning,  the  mem* 
bers  of  this  party  thus  missing  the  regular  morning  session. 
This  was  done  at  the  suggestion  of  the  Managers  of  the  Varnish 
works,  as  their  work  of  boiling  varnish  could  only  be  seen  in  the 
morning.  The  party  was  guided  by  Mr.  George  Marshall,  and 
included  the  following  places  :  Cleveland  Varnish  Co.,  Cleve- 
land Rubber  Co.,  Glidden  Varnish  Co.,  and  Warner  &  Swasey, 
instrument  makers. 

A  party  of  six  started  Tuesday  at  9  a.m.  from  the  Hollenden 
and  proceeded  to  the  Cleveland  Varnish  Co.,  where  Mr.  Stark, 
the  chemist,  conducted  them  ^rough  the  works  and  explained 
the  different  processes.  The  store-room  was  first  visited,  wherein 
large  boxes  of  rosin  are  stored,  some  from  New  Zealand  and  some 
from  Zanzibar,  the  first-named  place  being  the  chief  source.  The 
gums  or  rosins  used  for  varnish  are  amber,  fossil  and  annual ; 
amber  and  fossil  are  best,  as  the  thousands  of  years  they  have 
lain  in  the  groCind  seems  to  have  cured  them.  These  gums  are 
assorted  according  to  color,  the  lightest  being  the  most  valuable. 
They  are  found  by  probing  in  the  sand,  in  lumps  from  the  size 
of  a  bean  to  that  of  a  wash  tub.  The  largest  piece  of  Kauri 
gum  found  weighs  250  pounds.  This  is  the  chief  gum,  but  Zan- 
zibar is  also  used  for  fine  varnishes.     These  gums  are  insoluble 


(48) 

in  common  solvents  or  oil,  and  must  be  melted  at  560°  P.  in 
order  to  decompose  partially,  so  that  they  may  unite  with  the 
oil.  'The  chemists  visited  the  boilers  where  the  operation  of 
dissolving  the  gum  was  in  progress.  These  boilers  are  la]:]ge 
vessels,  in  shape  somewhat  like  the  farmer's  sap  or  soap 
kettles,  placed  on  a  small  buggy  to  facilitate  removal  from  the 
fire.  From  twenty  to  twenty*five  per  cent,  is  driven  off  as  water 
and  non-drying  oils  on  heating.  The  latter  are  unsta* 
ble,  and  have  not  been  investigated  or  utilized,  except 
in  connection  with  the  lampblack  industry.  The  gums  are  pre- 
pared for  the  boilers  by  hand,  as  the  best  size  for  melting  (about 
that  of  a  hen*s  egg),  is  obtained  by  chopping  with  a  small 
hatchet.  The  machines  crush  them  too  fine.  Linseed  oil  does 
not  dry  quickly  unless  boiled,  and  dryers  are  made  by  adding  a 
lead  or  manganese  salt  to  boiled  oil.  These  combine  with  the 
oil,  forming  a  compound  which  absorbs  oxygen  quickly  and 
hardens.  The  boilers  for  oil  are  the  same  as  for  melting,  gen* 
erally  larger,  with  hoods  to  carry  off  volatile  products.  The 
gum  boilers  are  fitted  with  covers  to  decrease  loss  from  spatter- 
ing and  to  control  the  irritating  jEumes  which  are  carried  off  by 
means  of  tall  chimneys.  Different  solvents,  as  turpentine,  for 
instance,  are  used  to  give  required  consistency  to  the  finished 
varnish,  and  are  added  to  the  cooled  mixture  formed  by  adding 
boiled  oil  to  the  melted  gum.  The  varnish  is  then  allowed  to 
settle  in  large  tanks  for  from  two  months  to  as  many  years.  The 
longer  the  time  the  finer  the  varnish  made.  This  is  necessary, 
as  filtering  does  not  remove  the  sediments.  The  top  is  siphoned 
off,  filtered  through  filter  presses  and  run  into  tanks  for  ageing. 
The  ageing  room  is  kept  at  a  constant  temperature,  such  that 
the  varnish  is  fluid,  for  from  nine  to  eleven  months.  The  party 
visited  the  cooper  shop,  storage  tanks,  shellac  mixers,  (which 
are  nothing  more  than  a  barrel  fastened  on  a  shaft  to  rotate  in 
the  direction  of  its  circumference) ,  and  a  paint  mill  of  the  latest 
pattern,  which  does  not  materially  differ  from  the  earliest  ones 
made.  The  filters  were  of  the  well  known  Johnson's  make,  and 
the  cloths  may  be  used  one  day,  then  removed  and  agitated  in 
a  shellac  mixer  with  turpentine  until  the  gummy  sediment  is 
washed  out.     Each  lot  of  varnish,  japan,  or  enamel  after  ageing 


(49) 

is  tried,  X.  ^.,  some  substance  for  which  it  will  eventually  be 
used  as  a  cover,  is  coated  with  it  and  dried  or  baked. 

The  most  noticeable  feature  at  the  Varnish  Co.,  sets  aside  the 
old  adage  about  the  shoemaker's  wife  and  blacksmith's  colt,  for 
paint  and  varnish  had  been  used  in  a  very  neat  and  tasty  man- 
ner throughout  the  establishment,  and  what  was  particularly 
prominent,  was  the  absolute  cleanliness  that  pervaded  even  the 
store-rooms,  the  settling  and  ageing  rooms,  where  long  lines  of 
tastefully  painted  tanks  pleased  the  eye,  and  the  excellent 
arrangement  of  drafts,  hoods,  covers  and  stacks  to  carry  off  the 
offensive  odors  detrimental  to  the  workman's  health. 

After  visiting  the  laboratories  and  drying  rooms,  the  offices 
were  next  in  order,  where  a  most  pleasant  surprise  was  in  store. 
At  the  invitation  of  the  president,  Mr.  Tyler,  the  chemists  sat 
down  to  a  most  elaborate  luncheon,  which  was  very  acceptable. 
After  a  vote  of  thanks  had  been  given  their  hospitable  host,  the 
visitors  wended  their  way  to  the  Cleveland  Rubber  Co. 

Notwithstanding  the  torn  up  condition  of  the  rubber  company 
on  account  of  inventory,  every  courtesy  was  shown  by  the  fore- 
man of  each  department,  who  personally  explained  his  part  of 
the  work.  The  first  room  visited  was  the  calendering  room, 
where  the  crude  exudings  of  the  South  American  rubber  trees 
are  mangled  into  workable  shape.  The  rubber  gum  as  received 
contains  fourteen  to  eighteen  per  cent,  water,  and  ten  to  twelve 
per  cent,  extraneous  matter,  such  as  stones^  twigs,  bark,  etc. 
Stones  are  the  favorite  adulterants,  as  the  crude  gum  is  bought 
by  weight.  The  gum  is  first  run  through  rollers  upon  which 
water  plays,  to  cleanse  it.  It  comes  from  the  rollers  looking  like 
a  sheet  of  thin  cork.  It  is  then  allowed  to  cure  for  three  or  four 
months  in  a  dry  room  at  about  70®  F.  The  South  American 
natives  cure  rubber  by  exposure  to  air  and  sun,  so  that  vulcani- 
zation is  not  necessary,  and  they  make  very  durable  shoes  from 
the  product  so  cured.  Modern  science  has  not  yet  perfected  a 
system  for  hastening  this  curing  operation,  and  it  is  necessary 
to  cure  at  70**  three  or  four  months  since  higher  heat  has  a 
decomposing  effect.  From  the  curing  room  the  rubber  is  passed 
through  slightly  heated  rollers  again  and  again  until  it  becomes 
a  compact,  yielding,  non-porous  mass.     A  little  vaseline  added 


(so) 

in  thisjworking  serves  to  make  the  mass  more  pliable  and  softens 
it.  From  the  first  rollers  it  passes  to  others,  where  different 
colored  powders  are  added  and  worked  into  the  body  of  the  rub- 
ber. Pure  rubber  is  black  and  is  useless  for  many  purposes,  but 
by  mixing  with  various  constituents  it  may  be  impressed  into  the 
pores  of  cloth,  making  a  covering  impervious  to  moisture, 
rolled  into  sheets  capable  of  holding  gases  or  liquids,  moulded 
into  any  shape,  hardened  by  heat,  and  welded.  The  fillers,  as 
they  are  called,  are  zinc  oxide,  whiting,  lampblack,  litharge, 
sulphide  of  iron  and  antimony,  and  many  other  substances 
known  only  to  the  trade.  Each  filler  has  a  different  effect  on 
the  rubber.  Sulphide  of  antimony  is  used  for  fine  elastic  rub- 
ber, such  as  is  used  in  dental  operations  and  for  marine  valves. 
In  general,  if  rubber  is  to  be  hardened,  as  when  used  for  door 
mats,  or  if  the  surface  is  made  impervious  to  moisture  or  air 
for  gas  bags,  water  bottles,  mackintoshes,  hose,  bicycle  tires, 
etc.,  sulphides  are  used  in  order  that  vulcanization  may  take 
place  after  the  article  is  formed.  I/itharge,  lampblack  and  zinc 
oxide  give  color  as '  well  as  body  to  the  rubber.  In  order  to 
cover  cloth  the  rubber  is  wound  on  rollers  and  fed  through  cal- 
enders or  heated  rolls  set  to  such  a  size  that  rubber  introduced 
between  the  rolls  in  mass  is  forced  through  the  cloth  and  becomes 
part  of  it ;  at  the  same  time,  by  heating  the  rolls  vulcanization 
also  takes  place.  Vulcanization  consists  of  forming  a  compound 
of  sulphur  and  rubber  in  air  by  heating,  and  although  it  takes 
much  of  the  elasticity  from  the  rubber,  renders  it  impervious  to 
liquids  or  gases.  Cloth  was  shown  prepared  in  this  way  for 
mackintoshes,  hose,  etc.  Cloth  so  prepared  may  be  sewed 
together  ;  a  solution  of  rubber  in  any  volatile  solvent  applied  to 
the  stitches,  or  the  seam  covered  with  a  strip  of  rubber  wet  with 
this  solution,  the  whole  placed  in  a  steam-heated  oven,  heated 
and  when  withdrawn  it  is  found  the  cement  has  welded  the  holes 
left  by  stitches,  or  the  strip  upon  the  original  pieces,  so  that  it 
becomes  a  compact  mass. 

The  party  next  visited  the  molding  department.  Here  the 
rubber  is  prepared  as  above  with  a  filler  of  some  sulphide  and 
then  pressed  like  dough  into  heated  molds.  It  is  very  plastic, 
filling  all  crevices  readily  and  taking  every  impression.     It  does 


(5t) 

not  melt,  but  the  first  heat  softens  it  to  the  consistency  of  soft  t^fi3% 
while  a  greater  heat  vulcanizes  it.  In  order  to  keep  surfaces  in 
contact  from  welding  together,  a  little  whiting  is  sprinkled 
between  them.  Scraps  are  reworked,  and  old  rubber,  such  as 
boots,  is  reclaimed  and  made  into  coarse  articles,  such  as  mats. 
The  mechanical  arrangements  were  (some  of  them)  wonderful, 
but  space  cannot  be  given  to  a  detailed  description  of  them  all. 
As  the  time  for  the  morning  excursion  had  now  been  expanded, 
the  party  were  obliged  to  forego  the  pleasure  of  visiting  the 
other  places  of  interest  on  the  list. 

It  is  to  be  regretted  that  more  did  not  avail  themselves  of  the 
opportunity  extended  to  them  to  visit  these  works. 

Route  No.  4  was  for  convenience,  divided  into  two  excursions, 
the  Grasselli  Chemical  Works  being  visited  on  Tuesday  and  the 
other  places  on  Monday. 

The  trip  to  the  Oil  works  was  under  the  guidance  of  Mr. 
H.  L.  Payne,  chemical  engineer,  and  the  visitors  were  intro- 
duced to  the  practical  side  of  a  subject,  whose  chemical  side 
was  so  interestingly  presented  by  Dr.  Mabery  in  his  address  on 
the  same  evening.     Only  one  refinery  was  visited. 

The  one  selected  is  not  under  the  control  of  the  Standard  Oil 
Co.,  and  works  up  all  of  its  own  product.  The  visitors  were 
therefore  fortunate  in  being  able  to  see  within  the  confines  of 
one  muddy  hillside  all  the  branches  of  this  vast  industry.  The 
members  of  this  firm,  Messrs.  Scofield,  Shurmer  &  Teagel,  were 
all  very  obliging,  and  Mr.  Daniel  Shurmer*  and  his  son  person- 
ally conducted  the  party  through  the  extensive  works. 

The  stills  are  sheet  steel  tanks  set  in  brick  work  and  have  an 
open  coal  fire  under  them.  Some  are  set  on  end  and  others  lie 
on  the  side.  They  are  not  usually  covered  in  like  a  boiler,  but 
exposed  to  the  air  on  top  and  sides,  and  only  protected  from  the 
weather  by  a  rough  shed.  This  circumstance  would  seem  to 
cause  a  waste  of  fuel,  but  it  probably  assists  in  the  proper 
"cracking**  of  the  oil.  The  distillate  is  condensed  in  a  series 
of  wrought  iron  pipes,  which  are  kept  cool  by  immersion  in  a 
long  wooden  trough.  The  trough  is  kept  filled  with  cold  water 
from  springs  in  the  side  hills.  Various  methods  of  procedure 
are  adopted  in  the  distillation,  depending  upon  the  character  of 


(52) 

the  product  desired,  for  not  all  of  the  various  oil  products  are  or 
can  be  made  in  one  distillation.  The  lighter  and  more  volatile 
products  are  usually  collected  together  and  then  redistilled  with 
steam  heat  in  order  to  separate  them  into  commercial  '  *  naphtha , ' ' 
"benzine,"  ** gasoline,"  etc.  In  this  distillation  the  steam  is 
introduced  directly  into  the  oil  and  the  condensed  water  is  drawn 
off  at  the  bottom  of  the  still'tank.  The  very  light  vapors  and 
gases  which  come  off  first  are  sucked  down  by  a  steam  syphon 
and  burned  under  the  boilers. 

The  residue  left  in  the  crude  oil  stills  may  be  either  a  thick 
tarry  pitch  or  the  distillation  can  be  carried  so  far  that  only  a 
porous  coke  is  left.  This  coke  contains  only  o.oi  to  0.02  per 
cent,  ash,  and  is  in  great  demand  for  the  manufacture  of  elecrtric 
light  carbons. 

The  burning  oils  are  all  purified  by  washing  in  a  huge  lead' 
lined  separatory  funnel,  called  an  agitator.  They  are  first 
washed  with  concentrated  sulphuric  acid  which  unites  with  and 
precipitates  the  basic  matters,  such  as  phenol.  The  oil  after 
this  washing  contains  less  oxygen  than  before.  The  principal 
object  of  this  treatment  is  to  remove  those  bodies  which  would 
cause  a  coloration  of  the  oil ;  a  water^white  kerosene  is  sup^ 
posed  to  best  suit  the  consumer.  The  excess  of  acid  is  washed 
out  with  water  or  steam  and  the  oil  is  completely  neutralized  by 
a  littl  e  caustic  soda.  This  refinery  is  using  some  Lima  oil,  and 
they  treat  it  with  litharge  in  the  agitator  to  remove  the  sulphur. 
Other  refineries  use  copper  oxide  in  their  stills  for  the  same 
purpose. 

One  of  the  most  interesting  products  of  crude  oil  is  paraffin. 
It  distills  over  as  a  greenish  oil,  and  the  visitors  were  permitted 
to  examine  the  crystallizing  process  where  the  solid  wax  is 
frozen  out  and  filtered  from  the  more  liquid  part  of  the  oil.  A 
second  or  third  treatment  of  the  crude  wax  with  solvent  naph- 
tha and  this  freezing  process,  turns  out  the  pure  white  chewing 
gum.     Every  paraffin  works  has  its  own  ammonia  freezing  plant. 

Most  of  the  refineries  have  no  chemist  in  their  employ.  They 
are  mainly  concerned  with  the  physical  properties  of  their  pro- 
ducts, and  a  boy  soon  learns  to  make  the  test  for  specific 
gravity,  flashing  point,  burning  point,  freezing  point  and  vis- 
cosity. 


(53) 

The  cooperage  and  shipping  departments  were  large  and 
interesting,  but  when  the  party  arrived  at  that  portion  of  the 
works  they  were  ready  to  return  to  the  hotel. 

Tf  he  other  excursion  of  Route  No.  4  was  taken  by  all  on  Tues- 
day morning.  The  Grasselli  Company's  works  are  of  such 
magnitude  that  a  description  is  impossible.  They  make  sul- 
phuric acid,  nitric  acid,  hydrochloric  acid,  mixed  acid  fornitro- 
glycerol  factories,  ammonia,  glycerol,  and  all  the  various  salts 
and  bye  products  of  such  an  industry.  Mr.  E.  F.  Cone,  chief 
chemist  of  the  experimental  laboratory,  was  the  conductor  on 
this  occasion,  and  Mr.  C.  A.  Grasselli,  president,  and  his  staff 
of  superintendents  were  instrumental  in  showing  everything  in 
the  works.  None  of  the  chemists  will  forget  the  pleasant  little 
lunch  which  awaited  them  in  the  company's  office. 

Route  No.  5  was  of  interest  to  iron  works  chemists,  and  com- 
prised the  Cleveland  Rolling  Mill  and  the  Blast  Furnace  depart- 
ment and  mill  of  the  Union  Rolling  Mill  Co.  It  was  in  charge 
of  Mr.  F.  E.  Hall. 

The  party  left  the  Square  at  1:30  p.m.  Monday  and  proceeded 
first  to  the  Crescent  Sheet  and  Tin  Plate  Co.'s  Works  on  Besse- 
mer Avenue,  near  the  N.  Y.  P.  &  O.  R.R.  This  company 
employs  two  hundred  men  and  has  a  capacity  of  thirty  tons  per 
day  of  sheet  iron  and  tin  plate.  The  plant  is  new,  having  been 
in  operation  about  one  year,  and  is  fitted  up  with  all  the  latest 
improvements  in  boilers,  engines  and  all  machinery,  including 
an  electric  plant  for  lighting,  and  operating  the  electric  cranes 
for  handling  rolls  and  heavy  materials. 

From  here  the  party  proceeded  to  the  Emma  Blast  Furnace  at 
the  intersection  of  the  N.  Y.  P.  &  O.  and  C.  &  P.  Railroads. 
This  furnace  is  operated  by  the  Union  Rolling  Mill  Co.,  whose 
rolling  mills  were  visited  later.  The  furnace  department  employs 
one  hundred  men  and  has  a  capacity  of  two  hundred  tons  of  pig 
iron  daily.  The  plant  is  modern  in  every  respect,  has  three 
blowing  engines  and  three  brick  hot  blast  stoves.  The  company's 
chemical  laboratory  is  located  at  the  furnace  plant  and  is  in 
charge  of  Mr.  Frank  E.  Hall. 

The  next  place  visited  was  the  rolling  mill  of  the  same  com- 
pany, situated  some  distance  farther  up  the  C.  &  P.  R.R.    Here 


(54) 

are  employed  about  350  men,  the  daily  product  of  the  works 
being  about  150  tons  of  merchant  iron.  No  steel  is  made,  the 
company  making  a  specialty  of  high  grade  wrought  iron. 

After  inspecting  this  plant  the  party  proceeded  to  the  large 
works  of  the  Cleveland  Rolling  Mill  Co.  This  works  is  the 
largest  in  the  city.  The  Bessemer  Steel  plant  has  a  capacity  of 
1 ,000  tons  daily,  about  seventy-five  per  cent,  of  which  is  made 
up  into  finished  products,  consisting  of  rails,  shafts  and  wire, 
in  the  various  departments  of  the  company's  works.  In  all  about 
3.500  men  are  employed. 

One  blast  furnace  is  located  at  this  plant.  The  company's 
two  larger  furnaces  are  located  about  three  miles  farther  down 
on  the  N.  Y.  P.  &  O.  R.  R.,  from  which  place  **  direct  metal  '* 
is  run  to  the  converters,  the  molten  metal  being  carried  this  dis- 
tance over  the  N.  Y.  P.  &  O.  R.  R. 

The  chemical  laboratories  of  the  company  are  in  charge  of 
Mr.  Chaddock.     Ten  chemists  are  employed. 

Route  No.  8,  Adelbert  College  and  Case  School  of  Applied 
Science  were  appointed  for  Monday  evening  Dr.  Gruener, 
instructor  of  chemistry  in  Adelbert  College,  was  absent  on  his 
vacation,  and  Mr.  H.  L.  Paj'ne,  a  graduate  of  the  first  chemis- 
try class  in  Case  School,  was  appointed  to  head  this  excursion. 

The  new  physical  laboratory  of  Adelbert  College  excited  the 
greatest  admiration.  The  chemical  laboratory  lacked  its  most 
interesting  feature — Dr.  E.  W.  Morley. 

At  Case  School  the  visitors  saw  the  marks  of  an  active  and 
well  equipped  institution  for  the  study  of  practical  applied 
science.  The  entertainment  of  the  evening,  of  course,  was 
Dr.  Mabery's  talk  on  **  Petroleum.** 

Route  No.  9  was  to  the  Steel  Works  at  Lorain,  O.  The  trip 
was  in  charge  of  Mr.  Hugo  Carlsson,  chief  chemist  of  the  John- 
son Co. 

These  works  occupy  about  eighty-six  acres^of  land  lying  along 
the  Black  river,  two  miles  south  of  Lorain.  The  plant  was  first 
put  in  operation  April  i,  1895.  Their  finished  products  are 
billets  and  special  rails  for  street  railways.  The  plant  consi.sts 
of  the  Bessemer  department,  blooming  mill,  shape  mill,  engine 
and  boiler  houses,  etc.     Plans  for  the  erection  of  six  bla.st  fur- 


(55) 

naces  have  been  completed  and  work  will  begin  on  their  erec- 
tion soon. 

The  engine  house  contains  the  blowing  engine  for  the  Besse* 
mer  department ;  this  was  made  by  the  Southwark  Company » 
Philadelphia.  It  is  a  horizontal  double  expansion  engine  with 
blowing  cylinder  sixty  inches  diameter  and  sixty  inches  stroke. 
In  the  same  building  are  the  dynamos  and  straight  line  engines 
which  furnish  power  for  the  Johnson  Electric  Railway  between 
Lorain  and  Elyria.  There  are  two  boiler  houses  containing 
National  Water  Tube  Boilers  with  Murphy  Automatic  Stokers, 
the  combined  capacity  of  the  two  batteries  being  about  6,000 
horse  power.  The  feed  water  is  heated  and  purified  before 
it  enters  the  boilers.  Gas  for  the  heating  furnaces  is  furnished 
by  a  plant  of  Duff  gas  producers.  Four  ten-foot  cupolas  melt  the 
pig  metal,  and  there  are  two  eight-foot  cupolas  for  spiegel.  The 
converting  department  contains  two  ten  ton  vessels.  Ingot 
molds  are  arranged,  on  small  cars  in  pairs.  There  is  no 
casting  pit.  The  ingots  are  transferred  from  the  converter  house 
to  the  soaking  pits,  in  which  they  are  kept  until  ready  for  roll- 
ing. The  blooming  mill  is.  drived  by  a  very  powerful  reversing 
engine,  built  by  the  Galloways,  Manchester,  England.  It  has 
double  cylinders  each  fifty-five  by  sixty  inches.  The  shape  mill 
is  driven  by  two  engines,  also  of  the  Galloways'  make.  The 
heating  furnaces  are  provided  with  two  cranes,  one  for  charging 
and  one  for  drawing. 

An  important  feature  of  the  works  is  the  special  machinery  for 
straightening  rails.  The  laboratory  occupies  a  two-story  brick 
building,  and  contains  a  250,000  pound  Olsen  testing  machine. 

THE  RECEPTION  BY  THE  CHAMBER  OF  COMMERCE. 

The  various  hosts  who  had  vied  with  one  another  in  extend- 
ing courtesies  and  hospitalities  to  the  visiting  chemists  while  in 
Cleveland,  not  being  satisfied  with  the  royal  welcome  which 
they  had  already  given,  tendered  their  guests  a  most  delightful 
reception  in  the  rooms  of  the  Chamber  of  Commerce,  Tuesday 
evening,  Dec.  31.  We  quote  from  a  Cleveland  paper  the  follow- 
ing description  of  the  elaborate  decoration  of  the  rooms : 

**  The  decorations  were  to  a  certain  extent  paradoxical,  being 


(56) 

emblematic  of  both  summer  and  winter.  The  walls  of  the  ceil- 
ing were  draped  with  holly,  heavily  laden  with  red  berries.  This 
drapery  came  down  to  within  a  few  feet  of  the  floor,  where  it 
was  met  by  banks  and  screens  of  palms  and  other  tropical  plants. 
This  apparent  paradox  was  a  pretty  tribute  to  the  various  sec- 
tions of  the  country,  north  and  south,  from  which  theguestscame. 
The  floral  designs  added  to  the  beauty  of  the  scene.  One,  a 
large  shield  of  white,  studded  with  roses,  bore  the  following  in- 
scription : 

*  Welcome,  1895 — American  Chemical  Society.' 

In -the  center  of  the  assembly  room,  on  a  table,  stood  an  im- 
mense design  which  was  a  tribute  to  Dr.  E.  W.  Morley  and  took 
the  form  of  a  reference  to  his  great  genius  in  determining  the 
atomic  weight  of  oxygen.  It  was  a  huge  balance  erected  on  a 
base  of  American  beauty  roses.  From  the  arms  of  the  balance 
hung  globes  of  white  flowers.  On  one  in  purple  was  the  simple 
capital  letter  *0,'  representing  oxygen,  and  on  the  other  the 
purple  figures  *  15.879,'  the  atomic  weight  of  oxygen." 

On  entering  the  Chamber  the  guests  were  welcomed  by  the 
Reception  Committee,  and  introduced  to  the  Cleveland  gentle- 
men who  were  present.  From  early  in  the  evening  until  mid- 
night the  visiting  chemists  had  the  pleasure  of  meeting  in  social 
intercourse  those  who  had  already  done  everything  in  their  power 
to  welcome  and  entertain  the  American  Chemical  Society,  and 
to  furnish  them  opportunities  for  a  successful  meeting. 

During  the  evening  the  Chamber  of  Commerce  Musical  Club 
and  the  Schubert  Mandolin  Club  rendered  various  choice  selec- 
tions, which  were  enthusiastically  received.  The  evening  was 
also  enlivened  by  the  humorous  recitations  and  impersonations  of 
Mr.  J.  E.  V.  Cooke. 

Refreshments  were  served  during  the  evening  in  the  commit- 
tee room,  and  it  seemed  as  though  nothing  was  wanting  to  make 
the  occasion  one  long  to  be  remembered.  To  cap  the  climax, 
however,  and  to  show  their  appreciation  of  one  of  Cleveland's 
most  distinguished  scientists,  a  message  of  greeting  and  compli- 
mentary reference  to  his  labors  in  determining  the  atomic  weight 
of  oxygen,  was  sent  by  cable  to  Dr.  E.  W.  Morley,  who  is  spend- 
ing the  winter  in  Europe. 

Some  of  the  chemists  were  obliged  to  leave  on  early  trains  and 
were  thus  unable  to  enjoy  the  whole  of  the  evening.  But  they, 
as  well  as  those  who  remained,  took  with  them  the  pleasantest 
recollections  of  their  visit  to  Cleveland,  and  the  Twelfth  Gen- 
eral Meeting  of  the  American  Chemical  Society. 

Erratum, — Page  32,  sixth  line  from  bottom  of  page y^r  sixty- 
three  per  cent,  read  sixty-three  hundredths  of  one  per  ceiit. 


Iftsned  with  April  Number,  1896. 


Proceedings. 


COUNCIL. 

The  Council  have  approved  the  nomination  of  Edward  Hart 
as  Editor  for  1896. 

CHANGES  OP  ADDRESS. 

Bloomfield,  L.  M.,  1239  Harrison  Ave.,  Cleveland*  Ohio. 
Furman,  H.  Van  F.,  Room  118,  Boston  Building,  Denver,  Col. 
Koebig,  Dr.  Julius,  306  Market  St.,  San  Francisco,  Cal. 
Penberton,  H.,  Jr.,  1008  Clinton  St.,  Philadelphia,  Pa. 
Phillips,  Francis  C,  P.  O.  Box  126,  Allegheny,  Pa. 
Sherman,  H.  C,  Columbia  University,  New  .York  City. 
Spencer,  G.  L.,  134  Rich  Ave.,  Mt.  Vernon,  N.  Y. 

ADDRESSES   WANTED. 

Bachman,  Irving  A.,  formerly  of  Augusta,  Ga. 
Jones,  Dr.  Walter,  formerly  of  Lafayette,  Ind. 


MEETINGS  OF  THE  SECTIONS. 

WASHINGTON    SECTION. 

The  annual  meeting  was  held  January  9,  1896,  and  wascalled 
to  order  by  the  President,  Charles  E.  Munroe,  at  8:00  p.  m. 

The  following  persons  were  elected  to  membership  :  Messrs. 
E.  W.  Magruder,  C.  C.  Moore,  and  E.  C.  Wilson. 

The  publication  of  Bulletin  No.  9  was  announced  and  arrange- 
ments reported  by  a  committee  for  a  social  meeting  to  be  heldin 
February. 

The  reports  of  the  Treasurer  and  Secretary  were  read  and 
adopted,  after  which  the  election  of  officers  for  the  ensuing  year 
was  held  with  the  following  result : 

President — E.  A.  de  Schweinitz. 

Vice  Presidents — W.  D.  Bigelow  and  W.  G.  Brown. 

Treasurer— W.  P.  Cutler. 

Secretary — A.  C.  Peale. 

Additional  Members  of  the  Executive  Committee — Charles  E. 
Munroe,  V.  K.  Chestnut,  F.  P.  Dewey,  and  H.  N.  Stokes. 


(58) 

The  first  paper  of  the  evening  was  read  by  H.  W.  Wiley  on  a 
'*  Steam- Jacketed  Drying  Oven."  **In  order  to  surround  the 
drying  space  of  an  oven  entirely  with  steam,  the  door  of  inordi- 
nary steam-jacketed  drying  oven  is  made  with  double  wdls,  into 
which  the  steam  from  the  oven  is  conducted  by  two  metal  flexi- 
ble tubes  inserted  at  the  top  and  bottom  of  the  door.  They  are 
so  arranged  as  not  to  interfere  with  opening  the  door.  By 
this  method  the  entire  drying  space  of  the  apparatus  is  sur- 
rounded with  steam,  easily  securing  a  constant  and  even  tem- 
perature. 

The  temperature  is  regulated  by  a  pressure  gauge  in  which 
the  steam,  by  acting  on  a  column  of  mercury,  cuts  off  the  gas  when 
a  given  pressure  is  reached.  A  steam  pressure  of  two  inches 
will  cause  a  temperature  of  about  102**  in  the  drying  space  of  the 
oven.  By  setting  the  gauge  at  any  position  desired,  the  temper- 
ature can  be  regulated,  when  steam  is  used,  to  read  from  the 
boiling  point  of  water  up  to  los"*.  For  other  temperatures  other 
liquids  can  be  used.  For  instance,  alcohol,  or  amyl  alcohol  for 
still  higher  temperatures,  and  so  on.  Ether  cannot  be  employed 
with  safety  on  account  of  the  danger  of  explosion  in  case  of 
leakage.'* 

Dr.  Wiley  exhibited  the  drying  oven  in  actual  operation. 

The  second  paper,  also  by  Dr.  Wiley,  was  on  the  **Heat  of 
Bromination  of  Oils.'*  **  The  method  of  determining  the  heat 
of  bromination  of  oils,  as  proposed  by  Hehner  and  Mitchell,  in  a 
recent  number  of  the  Analyst,  is  very  difficult  to  work  from 
the  meager  directions  given  by  the  authors.  The  especial  dif- 
ficulty in  the  process  is  in  handling  the  liquid  bromine  in  quan- 
tities of  one  cc.  at  a  time.  I  find  that  the  process  is  made  prac- 
ticable by  dissolving  both  the  oil  or  fat  and  the  bromine  in  chlo- 
roform, in  which  condition  the  bromine  solution  is  easily  handled 
by  means  of  a  special  pipette. 

In  order  to  make  a  number  of  analyses  of  the  same  sample, 
five  grams  of  the  fat  may  be  dissolved  in  chloroform  and  the  vol- 
ume completed  to  fifty  cc.  Ten  cc.  of  this  solution  will  contain 
one  gram  of  the  fat.  In  like  manner  five  cc.  of  bromine  may  be 
dissolved  in  chloroform  and  the  volume  completed  to  fifty  cc,  or 
larger  quantities  in  the  same  proportion  may  be  used.     The 


(59) 

gradual  evolution  of  hydrobromic  acid  from  a  mixture  does  not 
interfere  with  the  analytical  process,  as  the  amount  of  broinine 
used  is  always  largely  in  excess.  Ten  cc.  of  the  bromine  solu* 
tion  containing  one  cc.  of  the  liquid  bromine  are  used  for  each 
ten  cc.  of  fat  solution. 

The  pipette  for  handling  the  bromine  solution  is  so  arranged 
as  to  be  filled  by  the  pressure  of  a  rubber  bulb,  thus  avoiding  the 
danger  of  sucking  the  bromine  vapor  into  the  mouth.  The  solu- 
tion is  poured  upon  the  chloroform  solution  held  in  a  long  nar- 
row tube,  in  which  a  delicate  thermometer,  capable  of  being  read 
to  tenths  of  a  degree,  by* means  of  a  magnifying  glass,  isplaced. 
This  tube  is  held  in  a  large  cylinder,  from  which  the  air  can  be 
removed,  thus  affording  a  good  insulation  in  respect  to  heat. 

The  determinations  should  be  conducted  in  a  room  where  the 
temperature  is  as  constant  as  possible  and  the  pieces  of  the  appa- 
ratus should  be  exposed  to  the  open  air  for  at  least  half  an  hour 
after  completing  one  determination  before  beginning  another,  in 
order  to  be  restored  to  the  standard  room  temperature.  Dupli- 
cates usually  agree  within  one  or  two-tenths  of  a  degree,  though 
sometimes  the  variations  are  greater. 

The  ratio  of  the  heat  of  bromination  to  the  ordinary  number 
must  be  established  for  each  system  of  apparatus  employed. 
The  heat  of  bromination  of  various  oils  was  determined  by  the 
method  and  apparatus  described  above,  and  the  process  seems  to 
be  one  of  considerable  analytical  value.  For  exact  scientific 
purposes,  calorimetric  measurements  of  the  degree  of  heat  pro- 
duced must  be  made." 

Discussion  was  by  Messrs.  Warder,  Freeman  and  Munroe. 

Professor  Charles  E.  Munroe  then  made  some  remarks  upon 
the  *•  Corrosion  of  Electric  Mains."  He  exhibited  sections  of 
electric  light  cables  in  which  the  lead  coating  had  become  so 
corroded  that  in  some  places  the  interior  conductor  was  exposed, 
while  at  others  the  cable  was  coated  with  nodular  earthy-looking 
masses.  The  cables  were  parts  of  and  arranged  on  the  three 
wire  system,  which  carried  a  direct  current  of  i  lo  volts  on  each 
wire,  and  which  had  been  laid  underground  in  the  upper  com- 
partment of  a  terra  cotta  conduit.  The  corroded  main  was  a 
branch  in  an  alley.     The  principal  main  in  the  street  was  not 


(6o) 

attacked  in  the  least.  Analysis  showed  the  incrustation  tc  con- 
tain  nitrate,  chloride,  carbonate,  oxide  of  lead,  water  and  strace 
of  organic  matter.  Surrounding  the  alley  were  stables,  aid  the 
author  found  in  the  salts  in  the  soil  produced  by  the  excieta  all 
the  necessary  materials  and  conditions  for  effecting  chemical 
corrosion /^r  se.^  without  resorting  to  any  electrolytic  theory. 

Dr.  Wiley,  in  discussing  the  paper,  said  he  thought  there 
might  have  been  a  denitrifying  process. 

Professor  Munroe  said  there  could  have  been  no  constant 
moisture  present,  that  is,  there  was  no  submergence,  but  there 
must  have  been  water  passing  through  the  conduit. 

CINCINNATI  SECTION. 

The  Section  met  in  regular  session  Saturday  evening,  Feb- 
ruary 15,  1896.     Vice  President  Martin  presided. 

Dr.  Alfred  Springer  read  a  paper  on  **The  Characteristics  of 
Illuminates,"  and  exhibited  a  photograph  of  the  bones  of  the 
hand  made  by  means  of  the  Roentgen  X  rays.  The  picture  was 
kindly  loaned  for  the  purpose  by  Mr.  G.  W.  Zwick,  of  Coving- 
ton, Ky.,  who  had  recently  brought  it  from  Germany. 

**  Notes  on  Helium  and  Argon  "  was  read  by  Professor  T.  H. 
Norton. 

Dr.  S.  Waldbott  showed  how  the  value  of  litmus  paper  as  an 
indicator  could  be  enhanced.  His  method  was  to  use  a  capillary 
pipette  instead  of  an  ordinary  stirring  rod,  and  to  hold  the  point 
of  the  pipette  containing  a  drop  of  the  solution  upon  the  litmus 
paper ;  a  bright  red  spot  would  be  seen  at  the  point  of  contact, 
even  in  very  dilute  acid  solutions.  The  Doctor's  paper  on  "The 
Assay  of  Ipecac,"  announced  for  the  evening,  was  postponed 
till  next  meeting. 

NORTH   CAROLINA  SECTION. 

On  February  22nd  about  a  dozen  chemists  met  in  the  office  of 
the  Experiment  Station  in  Raleigh  to  organize  the  North  Caro- 
lina Section.    The  following  officers  were  elected  : 

President — F.  P.  Venable,  University  of  North  Carolina, 
Chapel  Hill. 

Vice  President — Charles  K.  Brewer,  Wake  Forest,  N.  C. 

Secretary  and  Treasurer— W.  A.  Withers,  Raleigh,  N.  C. 


(6i) 

The  following  papers  were  read  : 

''Absorptive  Power  of  the  Soil  for  Bases  and  Its  Relation  to 
FertiUty,"  by  Prof.  Withers. 

"A  Study  of  the  Zirconates,"  by  Dr.  Venable  and  Mr.  Clarke. 

**  Notes  on  the  Reduction  of  Methylenedi-<?-/-*«-nitraniline," 
by  Dr.  Baskerville. 

NEW  YORK  SECTION. 

The  regular  meeting  of  the  New  York  Section  was  held  at  the 
College  of  the  City  of  New  York,  on  Friday  evening,  March  6th, 
at  8:30  o'clock.     Professor  P.  T.  Austen  in  the  chair. 

The  following  papers  were  read  : 

**The  Cassel-Hinman  Gold  and  Bromine  Process,"  by  P.  C. 
Mcllhiney." 

**The  Specific  Gravity  of  Glue  Solutions,"  by  E.  R.  Hewitt. 

**  Investigations  in  the  Chemistry  of  Nutrition,"  by  Dr.  W.  O. 
Atwater. 

Mr.  Mcllhiney  enumerated  the  advantages  of  bromine  over 
chlorine  in  the  gold  extraction  process  as  (i)  greater  solubility 
of  bromine,  as  three  and  two-tenths  per  cent,  against  0.76  per 
cent.;  (2)  lesser  oxidizing  pywer,  whereby  the  iron  pyrites  is 
less  acted  upon  ;   (3)  greater  solvent  power  for  gold. 

The  bromine  is  recovered  by  distillation  with  live  steam  in 
stone  tanks,  after  addition  of  sulphuric  acid  and  an  oxidizing 
agent. 

The  process  is  especially  adapted  to  low  grade  telluride  ores, 
which  have  not  hitherto  been  profitably  worked. 

Mr.  Cassel,  being  present,  was  asked  to  what  extent  the  pro- 
cess had  been  worked,  and  whether  ores  containing  sulphides 
could  be  treated.  He  replied  that  fifty  tons  per  diem  had  been 
treated  since  January  ist,  and  the  capacity  was  to  be  increased  ; 
that  ores  containing  small  amounts  of  sulphides  had  been  suc- 
cessfully treated,  using  very  weak  solution  of  bromine,  and 
eighty  per  cent,  of  the  bromine  had  been  recovered  ;  but  it  was 
best  to  roast  sulphide  ores.  The  cost,  including  roasting,  was 
$1.75  per  ton. 

Mr.  Hewitt,  in  his  work  on  the  **  Specific  Gravity  of  Glue 
Solutions,"  had  obtained  his  results  from  experiments  on  all 
grades  of  glue  from  the  best  photographic  gelatine,  to  the  dark- 
est and  poorest  grades  in  the  market. 


(62) 

He  found  the  expansion  of  glue  solutions  to  be  the  same  as 
water  alone  ;  that  the  specific  gravity  of  glue  containing  water 
was  less  than  in  the  dry  state  ;  that  the  hydrometer  could  not 
be  used  in  solutions  containing  over  sixty-five  per  cent.,  and  that 
the  quality  of  the  glue  had  no  effect  on  the  specific  gravity  of 
the  solutions. 

He  concludes  that  there  is  a  series  of  distinct  chemical  com- 
binations of  glue  with  water. 

In  the  discussion  of  the  paper,  Dr.  Home  asked  if  the  specific 
gravity  of  a  glue  solution  could  be  determined  by  dropping  it 
into  some  solution  of  known  density,  not  acting  on  the  glue  solu- 
tion. 

Mr.  Hewitt  replied  that  this  method  had  been  tried,  using 
xylol,  chloroform,  and  some  other  liquids,  but  the  results  were 
not  as  satisfactory  as  could  be  obtained  by  the  hydrometer. 

The  presence  of  Dr.  C.  B.  Dudley,  President  of  the  Society, 
was  then  announced,  and  Dr.  Dudley  addressed  the  meeting  in 
part,  as  follows : 

*  *  Gentlemen  of  the  New  York  Section  :  It  has  been  a  rare 
pleasure  to  attend  this  meeting  of  the  New  York  Section,  and  I 
would  like  to  congratulate  yoii  on  one  or  two  points.  First,  the 
advantage  that  comes  to  you  from  being  able  to  meet  together, 
read  papers,  shake  hands  and  dine  together.  I  am  so  far  away 
from  the  chemists  that  it  does  me  good  to  meet  and  shake  hands 
with  a  chemist.  In  the  early  days  of  the  Pittsburg  Society  I 
tried  to  meet  with  them  and  have  been  present  on  many  enjoya- 
ble occasions,  but  having  joined  the  Society  when  there  was 
only  a  New  York  Section,  I  have  felt  at  home  with  you  and 
have  wished  I  could  meet  with  you  oftener. 

** Another  thing  on  which  I  wish  to  congratulate  you.  Our 
General  Secretary  informs  me  that  we  have  a  good  round  thou- 
sand now  in  our  membership.  There  are  those  of  you  who  have 
stood  by  the  Society  when  it  was  not  as  prosperous  as  it  is  now, 
who  can  appreciate  this. 

**  Now  as  to  what  is  to  be  done  in  the  field  of  our  labors.  My 
daily  work,  or  a  great  part  of  it,  is  with  iron  and  steel,  and  if  I 
could,  I  would  give  all  my  time  to  the  study  of  pig  iron. 

**  There  are  many  problems  yet  to  be  solved  in  regard  to  it, 


(63) 

and  of  which  a  great  deal  might  be  said,  but  as  there  are  other 
papers  to  come  before  you  this  evening,  I  will  not  detain  you 
longer.     I  am  very  glad  to  have  been  able  to  meet  with  you.*' 

Dr.  W.  O.  Atwater  was  then  introduced,  and  after  giving  a 
synopsis  of  the  work  which  had  been  done  in  other  countries, 
especially  in  Germany,  on  the  chemistry  of  food  and  nutrition, 
he  described  the  progress  which  had  been  made  in  this  country, 
beginning  with  the  early  work  of  Professor  Baird,  then  of  the 
Smithsonian  Institute,  who  gave  the  first  impulse  to  this  work 
by  his  studies  of  the  food  value  of  a  number  of  varieties  of  fish. 
He  then  passed  to  a  description  of  the  work  recently  done  under 
his  direction  and  that  now  in  progress  in  determining  the  heats 
of  combustion,  or  fuel  values  of  food.  He  said  that  we  know  the 
laws  of  conservation  of  energy  hold  good  in  the  living  organism, 
but  we  do  not  yet  know  haw  they  hold  good.  We  must  study 
these  things  in  the  living  organism,  and  for  this  paper  a  respir- 
atory calorimeter  has  been  constructed  at  Middletown  by  which 
the  experimental  determination  of  heat  of  radiation,  energy  of 
food  consumed,  etc.,  is  to  be  obtained.  A  man  had  been  kept 
in  this  apparatus  for  four  days,  and  it  was  expected  to  arrange 
to  extend  the  experiment  to  a  week  or  even  several  weeks. 

Eight  attendants  were  require  to  run  these  experiments. 

Dr.  Dudley  asked  whether  Professor  Atwater  had  used  a  cur- 
rent of  oxygen  instead  of  potassium  chlorate  in  his  experiments 
on  the  heats  of  combustion  of  foods,  and  stated  that  he  had  used 
the  oxygen  with  very  satisfactory  results  in  determinations  of 
calorific  value  of  coal. 

Dr.  Dudley  also  asked  whether  the  quality  of  the  fat  of  ani- 
mals was  dependent  on  the  food. 

Professor  Atwater  replied  that  the  fat  formation  is  a  function 
of  both  the  organism  and  the  food. 

On  motion  of  the  Secretary,  a  vote  of  thanks  was  passed  to 
Professor  Atwater  for  his  interesting  report  on  the  progress  of 
the  chemistry  of  nutrition. 

Professor  Breneman  moved  that  a  committee  be  appointed  to 
make  a  report  at  the  next  meeting  on  the  feasibility  of  organiz- 
ing a  chemical  club  from  the  members  of  the  New  York  Section. 
Seconded  and  carried. 


(64) 

The  Chair  appointed  Messrs.  Breneman,  McMortrie,  and 
Hallock. 

The  Librarian  announced  the  receipt  of  a  bequest  from  Dr.  A. 
A.  Pesquet,  of  two  microscopes  and  accessories.  The  Chair 
directed  that  a  suitable  recognition  of  the  gift  be  made. 

RHODE  ISLAND  SECTION. 

The  regular  meeting  of  the  Rhode  Island  Section  was  held  at 
Providence,  Thursday  evening,  Feb.  13,  1896.  Mr.  Chas.  S. 
Bush  in  the  chair. 

A  paper  was  read  by  Mr.  Charles  E.  Swett.  Subject,  "Ultra- 
marine." 

Th^  reader  presented  the  results  of  a  few  experiments  he  had 
performed  upon  ultramarine,  with  some  of  the  more  common  re- 
agents. 

The  March  meeting  was  held  on  the  19th  inst.,  at  Providence. 
Chairman  C.  A.  Catlin,  presiding. 

Mr.  Walter  E.  Smith  read  a  paper  upon  **The  Origin  of  Pe- 
troleum." 

In  brief,  the  paper  was  as  follows : 

The  theories  given  for  the  origin  of  petroleum  are  in  general 
divided  into  three  classes : 

1.  The  chemical  theories  advanced  by  Berthelot  and  Mende- 
l^efiF,  that  water  on  metallic  carbides  forms  acetylene,  which  is 
further  changed. 

2.  The  theory  that  it  is  indigenous  to  the  rocks  in  which  it  is 
found. 

3.  The  theory  that  it  is  a  distillate  formed  from  highly  organ- 
ized substances. 


Issued  with  May  Number,  x 

Proceedings. 

COUNCIL. 

The  Council  has  decided  to  hold  the  summer  meeting  at 
Buffalo,  August  21  and  22. 

NEW  MEMBERS  BISECTED  MARCH  26,   1896. 

Brown,  Thomas,  Jr.,  M.S.,  Princeton,  N.  J. 

Banner,  W'.  E.,  441  Green  St.,  Philadelphia. 

LaWall,  Charles  H.,  305  Cherry  St.,  Philadelphia. 

Nagelvoort,  J.  B.,  3237  Michigan  Ave.,  Chicago,  111. 

Sprout,  Louis  P.,  Scotia,  Pa.,  P.  O.  Benore. 

Stewart,  Dr.  Andrew,  1420  Q  St.,  N.  W.,  Washington,  D.  C. 

Wagner,  John  R.,  Drilton,  Pa. 

ASSOCIATES  EI.ECTED  MARCH  26,   1876. 

Caldwell,  Thomas  O.,  Agr.  Exp.  Sta.,  Bozeman,  Mont. 

Flowers,  John,  Agr.  Exp.  Sta.,  Bozeman,  Mont. 

Pilgrim,  Heber  B.,  Lafayette  College,  Easton,  Pa. 

Sieb,  Peter,  Agr.  Exp.  Sta.,  Bozeman,  Mont. 

Twitchell,  Mayville  W..  7098th  St.,  N.E.,  Washington,  D.C. 

Walter,  Charles  Albert,  506  South  5th  St.,  Champaign,  111. 

CHANGES  OP  ADDRESS. 

Barton,  G.  E.,  care  Whitall,  Tatum  &  Co.,  Flint  Glass 
Works,  Millville,  N.  J. 

Benjamin,  Dr.  Marcus,  Smithsonian  Institute,  Washington, 
D.  C. 

Berry,  W.  G.,  26  Whitehall  St.,  N.  Y.  City. 

Breyer,  Theo.,  P.  O.  box  112,  Peoria,  111. 

Brown,  H.  F.,  113  West  Central  St.,  Natick,  Mass. 

Fields,  J.  W.,  Stillwater,  Okla. 

Johns,  John,  306  Toone  St.,  Baltimore,  Md. 

Kelley,  J.  H.,  26  Snell  Hall,  Univ.  of  Chicago,  Chicago,  111. 

Lloyd,  Rachael,  care  R.  L.  Lloyd,  Lansdowne,  Pa. 

Low,  A.  H.,  P.  O.  drawer  1537,  Denver,  Colo. 

Maury,  George  P.,  care  Edgar  Thompson  Steel  Works,  Brad- 
dock,  Pa. 

Nickel,  Herman  L.,  care  N.  K.  Fairbank  Co.,  St.  Louis,  Mo. 

Pomeroy,  Charles  T.,  190  Mt.  Pleasant  Ave.,  Newark,  N.  J. 

Rosengarten,  F.  H.,  care  Photographic  Society,  10  So.  i8th 
St.,  Philadelphia,  Pa. 

Steiger,  Geo.,  1425  Corcoran  St.,  N.W.,  Washington,  D.  C. 


(66) 
MEETINGS  OF  THE  SECTIONS. 

WASHINGTON    SECTION. 

A  regular  meeting  was  held  February  i3tli,  1896.  As  the 
meeting  was  devoted  mainly  to  social  purposes  and  the  inaugu- 
ration of  the  newly  elected  president,  Dr.  E.  A.  de  Schweinitz, 
it  was  held  at  the  rooms  of  The  Washington  Down  Town  Lunch 
Club.  After  the  transaction  of  necessary  business  a  lunch  was 
served  which  was  enjoyed  by  thirty-one  members.  The  follow- 
ing persons  were  elected  to  membership :  Clinton  P.  Townsend, 
S.  S.  Voorhees,  and  Dr.  F.  K.  Cameron. 

The  Presidential  address  before  the  Washington  Section  was 
delivered  by  the  retiring  President,  Professor  Charles  E.  Mun- 
roe,  at  a  special  meeting  held  Friday,  February  21,  the  subject 
being  **  The  Development  of  Smokeless  Powders."  The  lecturer 
sought  to  show  that  the  necessity  for  a  high-power,  smokeless 
propellent  had  been  created  by  the  mechanical  perfection  to 
which  ordnance  had  attained  and  the  precision  of  the  weapons 
and  the  instruments  by  which  they  were  directed  ;  that  the 
possible  production  of  such  propellent  was  dependent  on  the 
discovery  of  guncotton,  nitroglycerol,  and  certain  nitro-substi- 
tution  compounds,  and  the  improvements  in  their  manufacture  ; 
that  the  possibility  of  producing  uniform  and  reliable  propel- 
lents was  dependent  on  the  invention  of  pressure  gauges  and 
velocimeters  ;  and  that  the  possibility  of  their  economical  pro- 
duction was  dependent  on  the  invention  of  mechanical  mixers 
and  formers  applied  in  other  arts.  In  a  historical  rdsum6  it  was 
shown  how  very  recent  most  prior  inventions  and  discov- 
eries were,  *and  it  was  pointed  out  that  a  very  large  propor- 
tion of  the  inventions  were  made  by  American  scientific  men. 

The  many  smokeless  powders  manufactured  or  prepared  were 
then  described  or  enumerated  and  classified  into  mixtures  of 
different  cellulose  nitrates  with  oxidizing  agents ;  mixtures  of 
soluble  or  insoluble  cellulose  nitrates  with  oxidizing  agents ; 
mixtures  of  soluble  or  insoluble  cellulose  nitrates  with  nitro- 
glycerol ;  mixtures  of  cellulose  nitrates  with  nitro-substitution 
compounds  ;  and  pure  cellulose  nitrate  powders ;  and  the  meth- 
ods of  manufacture  were  briefly  stated. 


(6?) 

The  lecturer  then  related  his  own  experience  in  inventing 
a  smokeless  powder.  Recognizing  at  the  outset  the  necessity 
for  the  closest  approximation  to  absolute  chemical  and  physical 
tiniformity  in  a  high-powered  powder,  and  being  familiar  with 
the  difficulty  of  securing  such  constancy  in  a  physical  mixture, 
he  set  about  producing  a  powder  from  carefully  purified  cellu- 
lose nitrate  of  the  highest  degree  of  nitration.  This  was  the  first 
and  only  attempt  made,  90  far  as  the  lecturer  was  aware,  to  pro- 
duce a  powder  which  consisted  of  a  single  substance  in  its  pure 
state. 

A  factory  was  erected  at  the  Torpedo  Station,  prior  to  his 
resignation  of  his  position  there,  and  the  powder  manufactured 
was  proved  at  Indian  Head  •  by  Ordnance  Officers  of  the  Navy. 
Secretary  Tracy  said  of  this  powder,  **  Report  of  the  Secretary  of 
the  Navy,  1892,  page  25.'*  **  It  became  apparent  to  the  Depart- 
ment early  in  this  administration  that  unless  it  was  content  to 
pass  behind  the  standard  of  military  and  naval  progress  abroad 
in  respect  to  powder,  it  must  take  some  steps  to  develop  and  to 
provide  for  the  manufacture  in  this  country  of  the  new  smoke- 
less powder,  from  which  extraordinary  results  had  been  obtained 
in  Europe."  With  this  object  negotiations  were  at  first  attempted 
looking  to  the  acquisition  of  the  secret  of  its  composition  and 
manufacture.  Finding  itself  unable  to  accomplish  this  the 
Department  turned  its  attention  to  the  development  of  a  similar 
product  from  independent  investigation.  The  history  of  these 
investigations  and  of  the  successful  work  performed  in  this 
direction  at  the  Torpedo  Station  has  been  recited  in  previous 
reports.  It  is  a  gratifying  fact  to  be  able  to  show  that  what  we 
could  not  obtain  through  the  assistance  of  others  we  succeeded 
in  accomplishing  ourselves,  and  that  the  results  are  considera- 
bly in  advance  of  those  hitherto  obtained  in  foreign  countries.'' 

The  conditions  that  a  smokeless  powder  should  fulfill  and  the 
tests  prescribed  by  the  lecturer  were  then  set  forth,  and  in 
closing  he  pointed  out  that  the  powder  was  now  developed  to  a 
higher  degree  than  the  gun  and  that  changes  in  the  latter  to 
render  it  more  efficient  were  being  considered  by  ordnance 
experts. 


(68) 

CINCINNATI  SECTION. 

The  Section  met  in  regular  session,  Tuesday,  March  17,  1896. 

President  Twitchell  presided. 

The  discussion  of  '*The  Scientific  Concepts  of  Etidorhpa" 
was  announced  for  the  evening.  The  popularity  of  the  book 
was  evidenced  by  the  presence  of  many  friends  of  the  author 
and  of  the  other  members  of  the  Section. 

Dr.  Alfred  Springer  read  extracts  from  the  book  and  took 
issue  with  the  author  on  some  of  the  statements.  Prof.  Lloyd 
re-affirmed  his  belief  in  the  theories  advanced  and  referred  the 
Doctor  to  the  preface  to  the  author's  edition,  in  which  he  had 
stated  he  would  decline  "  to  make  any  subsequent  comments  on 
the  work."  The  Professor  then  read  three  chapters  in  the  orig- 
inal manuscript,  which  had  been  omitted  from  the  published 
work.  He  now  regrets  the  omission,  as  the  continuity  of  the 
narrative  is  somewhat  impaired  thereby. 


RHODE  ISLAND  SECTION. 

The  regular  meeting  of  the  Rhode  Island  Section  was  held  at 
Providence,  Thursday  evening,  April  16, 1896,  Chairman,  Charles 
A.  Catlin,  presiding. 

Mr.  Charles  S.  Bush  read  a  paper  on  **  Petroleum  Products." 
The  following  is  an  outline  of  the  paper  : 

1.  Discovery  of  petroleum. 

2.  Brief  history  of  the  petroleum  industry  in  the  United 
States. 

3.  Outline  of  the  distilling  and  refining  process  now  in  general 
use  in  the  United  States. 

4.  The  importance  of  petroleum  as  a  means  of  reducing  fric- 
tion to  a  minimum. 

5.  New  methods  compared  with  old  ones,  especially  referring 
to  **  petroleum  products"  used  for  lubricating  purposes. 


NEBRASKA  SECTION. 

The  regular  meeting  of  the  Nebraska  Section  ^**s  held  at  the 
University  of  Nebraska,  on  Tuesday  evening,  March  31,  at 
eight  o'clock. 


(69) 

The  president  being  absent,  Mr.  Samuel  Avery  was  elected 
chairman /ri9  tern.,  and  called  the  meeting  to  order. 

The  following  papers  were  read  : 

"Recent  Work  on  the  Roentgen  Rays,"  by  Prof.  D.  B. 
Brace,  of  the  Department  of  Physics,  University  of  Nebraska. 

**  Report  on  Argon,"  by  Miss  Rosa  Bouton. 

**  Calcium  Carbide  and  Acetylene,"  by  Dr.  John  White. 

Prof.  Brace  exhibited  some  Crookes'  tubes  prepared  in  his 
laboratory,  made  a  general  statement  of  the  manner  in  which 
these  were  prepared  and  used,  of  the  effect  of  the  X  or  Roent- 
gen rays,  and  exhibited  some  photogrraphs  taken  by  their  use. 
Of  these  one  was  of  special  interest ;  it  represented  a  shado- 
graph  of  a  metal  object  taken  by  the  cathode  and  anode  rays. 
There  was  no  appreciable  distinction  between  them. 

Miss  Bouton's  paper  gave  a  very  thorough  and  clear  account 
of  argon,  from  the  very  earliest  experiments  of  Lord  Rayleigh 
on  the  density  of  nitrogen  down  to  and  including  the  present 
state  of  our  knowledge  of  argon,  its  chemical  and  physical  prop- 
erties. 

Dr.  White  exhibited  a  specimen  of  calcium  carbide,  which 
had  been  prepared  in  the  electrical  laboratory  of  the  University, 
gave  a  brief  historical  statement  of  the  carbides  in  general,  and  of 
their  use  in  the  preparation  of  acetylene.  He  then  prepared 
some  acetylene  by  treatment  of  the  carbide  with  water,  and  by 
burning  the  gas  under  proper  conditions  showed  how  it  may  be 
used  as  an  illuminant.  He  followed  this  by  a  short  lecture,  in 
which  the  economic  use  of  acetylene  as  an  illuminant  was  dealt 
with,  laying  speqial  stress  upon  its  advantages  and  disad- 
vantages. 

Owing  to  the  lateness  of  the  hour,  Dr.  White's  paper  on 
**  Metallic  Suboxides"  was  postponed. 

At  the  business  meeting  which  followed,  Mr.  E.  C.  Ellioet 
and  Miss  Marietta  Gray  were  elected  members  of  the  Section. 


NEW  YORK  SECTION. 

Minutes  of  the  meeting  of  April  lo. 

A  report  was  made  by  the  chairman  of  the  committee  appointed 


(70) 

to  consider  the  organization  of  a  chemical  club  in  New  York. 
Out  of  eighty-two  replies  already  received,  sixty  were  uncondi- 
tionally in  favor  of  the  project. 

It  was  further  stated  that  as  there  had  evidently  been  some 
misunderstanding  as  to  the  intended  membership »  it  should  be 
known  that  there  is  no  intention  of  limiting  the  membership  to 
any  section  of  the  chemical  fraternity,  but  to  include  chemists 
and  chemical  manufacturers  generally. 

Dr.  Albert  R.  Leads  read  a  paper  on  *'  Standard  Prispis  in 
Water  Analysis,  and  the  Valuation  of  Color  in  Potable  Waters.*' 

In  the  discussion  of  Dr.  Leed's  paper,  Prof.  Birchmore  ex- 
plained an  arrangement  of  adjustible  colored  prisms  projecting 
inside  a  glass  cylinder,  one  over  the  other,  by  which  the  Nessler 
reagent  colors  could  be  matched  and  recorded.  The  cylinder  is 
to  be  filled  with  a  liquid  having  the  same  refractive  index  as 
glass  ;  oil  of  juniper  was  mentioned  as  suitable  ;  and  the  record 
is  made  by  readings  on  the  milled  heads  of  the  screws  by  which 
the  Qverlapping  of  the  prisms  is  regulated. 

Dr.  I/ceds  moved  that  a  committee  be  appointed  to  unify  the 
methods  of  color  comparison  and  report  upon  a  standard  of 
measurement  of  color  in  potable  waters. 

Prof.  McMurtrie  thought  that  such  committee  should  be 
appointed  by  the  council,  and  that  the  secretary  should  com- 
municate the  resolution  to  the  President  of  the  Society. 

Dr.  Leeds'  motion,  as  amended  by  Prof.  McMurtrie,  was 
seconded  and  carried. 

A  paper  was  read  by  C.  L.  Speyers  on  '*  Matter  and  Energy." 

Dr.  E.  G.  Love  exhibited  some  fine  photomicrographs  of 
starches. 

Di;.  L.  Saarbach  exhibited  and  described  an  improved  form  of 
**  Laboratory  Temperature  Regulator,'*  which  he  had  found  sen- 
sitive, reliable,  adjustable,  and  easily  taken  apart  for  cleaning. 


Issued  with  June  Number,  1896. 

Proceedings. 

COUNCIL. 

Prof.  H.  H.  •Nicholson,  Lincoln,  Neb.,  has  been  elected  a 
member  of  the  Council,  to  take  the  place  left  vacant  by  the  elec- 
tion of  Dr.  C.  B.  Dudley  to  the  presidency  of  the  Society. 

The  Council  has  voted  to  accept  the  invitation  to  hold  the 
-winter  meeting  at  Troy,  N.  Y.,  on  Tuesday  and  Wednesday, 
December  29  and  30. 

The  New  York  Section  has  asked  that  a  committee  be  ap- 
pointed to  unify  the  methods  of  color  comparison  and  report  on 
a  standard  for  measurement  of  color  in  potable  waters.  The 
Council  has  agreed  to  the  formation  of  such  a  committee  and 
named  the  following  persons  to  act  as  members  :  A.  R.  Leeds, 
Wm.  P.  Mason,  Thomas  M.  Drown. 

NEW  MEMBERS  ELECTED  MAY  II,  1 896. 

Bowman,  J.  W.,  Green  Island,  N.  Y. 

Hunziger,  Dr.  Aug.,  care  Weidman  Silk  Dyeing  Co.,  Pater- 
son,  N.  J. 

Yates,  J.  A.,  Williamsburg,  Ky. 

ASSOCIATES  ELECTED  MAY  II,  1 896. 

Meade,  Richard  K.,  Longdale,  Va. 

Pilhashy,  Benjamin  M.,  1058  Cutter  St.,  Cincinnati,  O. 

CHANGES  OP  ADDRESS. 

Dodge,  F.  E.,  316  Bowne  Ave,  Flushing,  N.  Y. 
Hays,  Joseph  A.,  147  So.  i8th  St.,  Pittsburg,  Pa. 
Hopkins,  Cyril  G.,  409  W.  Main  St.,  Urbana,  111. 
Lord,  N.  W.,  338  W.  Eighth  Ave.,  Columbus,  O. 
Peale,  A.  C,  box  2043,  Station  A,  Philadelphia,  Pa. 
Power,  Frederick  B.,  535  Warren  St.,  Hudson,  N.  Y. 
Shepherd,  Frank  I.,  Kyle,  Ohio. 
Stillwell,  J.  S.,  box  3015,  N.  Y.  City. 
Tonceda,  Enrique,  care  Troy  Steel  Co.,  Troy,  N.  Y. 

ADDRESSES  WANTED. 

Gallaher,  Phil.  C,  formerly  of  Leadville,  Colo. 


(72) 
MEETINGS  OP  THE  SECTIONS. 

CINCINNATI  SECTION. 

The  regular  meeting  of  the  Section  was  held  Wednesday 
evening,  April  15th. 

Dr.  S.  Waldbott  presented  a  paper  on  **  The  Assay  of  Ipecac," 
in  which  he  outlined  the  various  methods  for  the  alkaloidal  assay 
of  crude  drugs  and  gave  some  results  obtained  by  applying  the 
Lloyd  method  for  the  assay  of  fluid  extracts,  to  the  determina- 
tion of  emetine  in  ipecac  root;  with  some  slight  modification, 
Dr.  Waldbott  thinks  good  results  may  be  obtained. 

In  a  paper  on  **  The  iodoso-  and  iodo-compounds  and  iodonium 
bases,  Dr.  John  McCrae  gave  an  interesting  account  of  some  of 
the  work  he  had  done  on  these  compounds,  under  the  instruc- 
tion of  Victor  Meyer. 

NEW  YORK  SECTION. 

The  New  York  Section  held  its  usual  monthly  meeting  in  the 
chemical  lecture  room  in  the  College  of  the  City  of  New  York  on 
Friday  evening.  May  8,  with  about  fifty  members  present.  Dr. 
Peter  T.  Austen,  presiding.  In  response  to  inquiries  regarding 
the  progress  made  by  the  committee  appointed  to  canvass  the 
matter  of  the  organization  of  a  chemical  club.  Prof.  Austen 
stated  that  in  accordance  with  the  instructious  given,  it  had 
increased  its  numbers  to  fifteen  and  had  held  several  meeting^, 
to  one  of  which  the  members  of  the  New  York  sections  of  the 
American  Chemical  Society  and  of  the  Society  of  Chemical 
Industry,  as  well  as  manufacturers  and  gentlemen  interested  in 
the  science  and  art  of  chemistry,  business  men  and  friends  of 
chemistry  were  invited.  The  meeting  was  full  and  enthusiastic. 
The  committee  was  instructed  to  increase  its  number  to  fifty  or 
more  and  to  push  the  organization  of  the  club  as  rapidly  as  pos- 
sible. The  committee  had  held  another  meeting  and  added  a 
large  number  of  names  of  prominent  chemists,  manufacturers, 
and  business  men  to  the  list.  The  general  opinion  seems  to  be 
that  the  initiation  fee  should  be  fixed  at  $25,  and  yearly  dues  at 
$25.  It  is  the  intention,  while  in  no  way  hampering  or  restrict- 
ing the  evolution  of  the  Chemical  Club,  which  many  of  the  more 
enthusiastic  supporters  of  the  movement  predict,  to  start  the 
club  in  a  conservative  and  economical  way,  and  not  to  exceed 


(73) 

the  pecuniary  limit  which  shall  be  decided  upon  after  careful 
deliberation.  It  appears  that  there  is  not  in  existence  in  this  or 
any  foreign  country  any  real  chemical  club,  as  differentiated 
from  a  chemical  society.  It  is  believed  that  the  science  and  art 
of  chemistry  furnish  so  much  that  is  characteristic  that  a  chem- 
ical club  may  easily  be  made  a  unique  organization.  The  mem- 
bers of  the  committee  of  fifteen  are  Prof.  A.  A.  Breneman,  Dr. 
A.  P.  Hallock,  Prof.  Peter  T.  Austen,  Dr.  W.  McMurtrie,  Prof. 
Morris  Loeb,  Prof.  C.  A.  Doremus,  Dr.  E.  R.  Squibb,  Dr.  J.  H. 
Wainwright,  Mr.  A.  H.  Mason,  Mr.  S.  W.  Fairchild,  Mr.  W. 
H.  Nichols,  Mr.  W.  J.  Matheson,  Mr.  T.  F.  Main,  Prof.  A.  H. 
Sabin,  and  Dr.  C.  P.  Chandler. 

Dr.  A.  R.  Leeds,  of  Stevens  Institute,  read  a  paper  on  the 
"Bacteriaof  Milk  Sugar.''  The  author  finds  that  the  morphology, 
classification,  physiology,  and  botany  of  bacteria  are  so  rudi- 
mentary and  unsatisfactory  that  the  most  valuable  methods  of 
bacteriological  investigation  are  still  of  a  chemical  nature,  and 
the  advances  to  be  made  in  the  near  future  are  to  be  looked  for 
mainly  on  the  chemical  sides  of  the  subject. 

The  author  was  interested  to  note  in  the  progress  of  his  work 
that  precipitated  zinc  hydroxide,  which  is  generally  considered 
amorphous  or  gelatinous,  is  really  crystalline. 

Dr.  H.  W.  Wiley,  of  the  United  States  Department  of  Agri- 
culture in  Washington,  offered  a  paper  entitied  **  Recent 
Advances  in  Milk  Investigations. ' '  In  the  absence  of  the  author 
the  paper  was  read  by  Dr.  William  McMurtrie.  It  treated  of 
the  bacterial  theory  of  milk  decomposition,  the  composition  of 
woman's  milk  as  compared  with  cow's  milk,  and  the  relative 
value  of  the  two  for  infant  food,  and  of  the  commercial  standards 
which  should  be  fixed  for  the  milks  sent  to  the  city  markets. 

The  author  reviewed  the  work  of  Soldner  regarding  the  pro- 
teid  content  of  human  milk,  and  quoted  the  figures  given  by 
authority  for  the  average  composition  of  human  milk,  as  follows : 

Per  cent. 

Proteids 1.52 

Fat 3.28 

Sugar 6.50 

Ash 0.27 

Citric  acid ^ 0.05 

Undetermined 0.78 

Total  dry  substance 12.40 


(74) 

The  undetermined  substances,  0.78  per  cent.,  are  mostly 
nitrogenous  bodies  not  generally  found  in  cow's  milk,  and  for 
this  reason  cow's  milk  can  never  be  so  diluted  or  altered  as 
to  properly  supply  the  natural  nutriment  of  the  infant. 

Soldner  follows  the  method  of  Munk  for  determination  of  pro- 
teids,  regarding  as  non-proteid  jnatter  those  nitrogenous  bodies 
not  precipitated  by  tannin  in  presence  of  common  salt.  In 
woman's  milk  these  amount  to  nine  per  cent,  of  the  total  nitro- 
genous constituents,  and  in  cow's  milk  to  about  six  per  cent. 

The  author  then  discussed  the  view  of  Bechamp  that  milk  de- 
rived from  healthy  animals  is  capable  of  spontaneous  alteration, 
which  consists  in  the  development  of  lactic  acid  and  alcohol  and 
the  development  of  curds  in  those  milks  which  contain  casein- 
ates  produced  by  the  precipitating  action  of  the  acids  formed. 
Oxygen  and  the  germs  present  in  the  air  are  held  to  have  noth- 
ing to  do  with  this  alteration  of  the  properties  of  milk.  The 
general  conclusion  reached  is  that  microqrganisms,  such  as 
vibriones  and  bacteria,  are  developed  by  a  natural  evolution 
from  the  microzymes,  even  in  milk  which  has  been  boiled. 

The  surprising  results  of  Soldner  and  Bechamp  should  lead  to 
new  studies  of  bacterial  action  in  milk.  If  it  should  prove  true 
that  milk  contains  autogenetic  germs  for  its  own  change,  and  that 
by  the  development  of  these  germs  into  vibriones  and  bacteria, 
the  natural  souring  takes  place,  it  will  be  necessary  to  change 
completely  the  common  view  respecting  these  processes. 

The  author  further  discussed  the  commercial  standards  for  the 
composition  of  milk,  declaring  that  the  value  of  milk,  both  for 
butter  and  cheese  making,  should  be  gauged  by  its  content  of 
butter  fat,  denouncing  the  claim  of  dealers  that  any  milk  from  a 
healthy  cow  should  be  sold  without  legal  restriction,  no  differ- 
ence what  its  content  of  fat  may  be,  and  recommending  that 
the  minimum  standard  for  fat  content  of  milk  supplied  for  human 
consumption  should  be  placed  at  three  per  cent,  or  higher. 

Dr.  Leeds  considers  that  in  judging  of  the  figures  of  Soldner 
presented  it  is  important  to  be  informed  of  the  conditions  under 
which  the  samples  for  analysis  were  taken  and  the  quantity  used 
for  analysis,  particularly  for  the  determination  of  such  constitu- 
ents as  citric  acid  and  the  undetermined  substances.     Samples 


(75) 

of  woman's  milk  usually  available  are  too  small  for  such  minute 
determinations.  Regarding  the  content  of  proteids,  the  figures  * 
of  Soldner  do  not  vary  widely  from  those  previously  found  and 
reported.  One  hundred  samples  of  woman's  milk  examined  in 
New  York  gave  an  average  of  less  than  two  per  cent.,  with 
variations  of  0.75  per  cent,  to  4.75  per  cent.  The  only  explana- 
tion of  the  very  low  figure  of  Soldner  is  that  only  partial  secre- 
tion was  available.     The  figure  1.52  given  is  not  surprising. 

That  the  various  bodies  secreted  from  the  blood  should  be 
present  is  generally  accepted.  Variations  in  the  composition  of 
the  milk,  due  to  emotional  influences,  such  as  nervousness, 
excitement,  fatigue,  fright,  anger,  etc.,.  are  well  known. 

The  fat  and  total  solids  given  in  the  analysis  are  surprisingly 
low. 

Dr.  Bccles  questioned  the  declaration  that  modified  cow's 
milk  was  not  a  proper  food  for  infants.  Constant  experience, 
forced  by  necessity,  shows  that  it  supplies  excellent  nutrition  for 
infants. 

Prof.  Marston  Bogert,  of  Columbia  College,  read  a  paper  on 
*'  Normal  Heptyl  Thiocyanate." 

The  steps  followed  by  the  author  in  preparation  of  heptyl 
thiocyanate  are  as  follows  :  Production  of  heptyl  alcohol  from 
oenanthol  by  reduction  with  zinc  dust  and  acetic  acid,  conver- 
sion of  the  heptyl  alcohol  into  the  bromide,  and  addition  of  the 
bromide  to  boiling  alcoholic  solution  of  potassium  thiocyanate. 
The  yellow  oil  finally  obtained  washed  free  from  potassium^  thi- 
ocyanate, dried  with  calcium  chloride  and  distilled,  all  passed 
over  between  230°  and  234*  C. 

Normal  heptyl  thiocyanate  is  a  colorless,  mobile  liquid,  hav- 
ing a  slightly  alliaceous  but  rather  pleasant  odor  and  a  specific 
gravity  of  0.931  at  is"*  C. 

Dr.  Austen  exhibited  an  apparatus  for  lecture  demonstration 
of  the  properties  of  the  heavier  gases. 


WASHINGTON    SECTION. 

A  regular  meeting  was  held  Thursday,  March  12th,   1896, 
with  President  Dr.  de  Schweinitz  in  the  chair.     There  were 


(76) 

thirty-five  members  present,  and  Dr.  Andrew  Stewart  was 
elected  to  membership. 

Mr.  F.  P.  Dewey  read  a  paper  on  **  The  Refining  of  Lixivat- 
ing  Sulphides.''  Dr.  Dewey's  paper  reviewed  the  leaching^ 
process  and  the  treatment  of  sulphide  precipitates  produced.  He 
described  the  sulphuric  acid  process  of  treating  the  sulphides,  in 
which  they  are  treated  in  strong  sulphuric  acid  to  convert  the 
sulphides  into  sulphates,  after  the  charge  is  treated  with  water, 
the  silver  precipitated  by  copper  and  melted,  the  copper  sul- 
phate crystallized.  In  the  1894  run  of  the  Marsac  Refinery » 
116,519^  pounds  of  sulphides,  carrying  572,544.4  ounces  of  sil- 
ver by  the  corrected  assay  were  treated,  and  574,623.26  ounces 
of  silver  were  returned,  showing  a  plus  clean  up  of  2,073.81 
ounces,  or  0.36  per  cent.  96.29  per  cent,  of  the  product  was  in 
the  form  of  bars,  averaging  999.4  fine  in  silver,  no  gold. 

Professor  H.  W,  Wiley  and  E.  E.  Ewell  read  a  paper*  on 
"The  Determination  of  Lactose  in  Milks  by  Double  Dilution 
and  Polarization." 

Professor  H.  Carrington  Bolton  read  a  paper  on  "  Berthelot's 
Contributions  to  the  History  of  Chemistry,"  reviewing  his 
**  Collection  des  AlchimistesGrecs,"  (Paris,  1887  ;  3  Vol.  4to), 
and  his  **  La  Chimie  an  Moyen  Age,"  (Paris,  1893 ;  3  Vol.  4to), 
showing  their  scope,  analyzing  their  contents  and  indicating  the 
important  changes  in  (Chemical  history  resulting  from  Berthelot's 
studies.' 

In  the  discussion  of  Dr.  Bolton*s  paper.  Dr.  Wiley  referred  to 
the  fact  that  the  Phoenicians,  as  early  as  1200  years  before 
Christ,  became  famous  by  reason  of  the  remarkable  dyes  which 
they  produced,  and  that  they  were  derived  from  a  colorless  sub- 
stance found  in  certain  moUusks,  which,  when  exposed  on  fibers 
to  the  light,  turned  green,  then  red  and  purple.  He  referred  to 
the  fact  that  on  the  continent  of  Europe  many  scientific  men  had 
also  become  famous  in  politics,  •  and  among  them  preeminently 
Berthelot  and  Virchow.  Berthelot  was  at  least  one  official 
chemist  who  had  attained  political  distinction,  and  his  career 
might  be  imitated  with  advantage  to  the  public  service  by  some 

1  This  Journal,  x8, 438. 
s  This  Journal,  18, 466. 


(77) 

American  scientists;  and  we  should  not  despair  of  looking  for- 
ward to  the  day  when  chemists  should  at  least  be  members  of 
the  Cabinet,  if  not  Presidents  of  the  United  States.  He  thought 
Berthelot  was  particularly  well  suited  to  write  of  the  alchemists, 
because  some  of  his  views  would  do  credit  to  the  wildest  vagar- 
ies of  the  alchemists,  especially  his  notion  that  the  art  of  the 
chemist  in  the  synthesis  of  foods,  would  in  the  near  future  ren- 
der the  practice  of  agriculture  unnecessary.  But  we  should  not 
criticise  a  great  man  because  of  his  vagaries,  and  after  all  it 
may  be  true  that  insanity  is  the  highest  type  of  genius. 

The  topic  of  discussion  for  the  evening  was  *  *  Style  in  Chemi- 
cal Books  and  Papers."  Dr.  Wiley  opened  the  discussion  by 
saying  there  are  many  problems  that  present  themselves  to 
authors  of  scientific  work.  Some  of  these  are  of  vital  import- 
ance, while  others  are  mere  matters  of  taste.  Not  having  ex- 
pected to  discuss  the  question  on  this  occasion,  he  would  confine 
himself  to  the  minor  topics.  He  suggested  that  there  should  be 
some  uniform  systsm  of  abbreviation  employed,  and  for  his  part 
preferred  very  much  small  letters  without  periods.  The  intro- 
duction of  capital  letters  in  abbreviations  marred  the  appearance 
of  a  book,  and  appeared  to  be  entirely  unnecessary.  He  thought 
perhaps  some  abbreviation  of  common  metric  terms  would  prove 
advantageous;  for  instance,  the  writing  of  the  words  cubic  cen- 
timeters repeatedly  not  only  requires  a  great  deal  of  space,  but 
the  repetition  of  the  term  becomes  tiresome.  Some  short 
word  might  be  used  to  represent  this  magnitude,  as,  for  instance, 
cubics.  It  would  be  well  for  chemists  to  agree  upon  some  such 
system,  provided  the  system  were  rational  and  easy  of  applica- 
tion. 

Another  question  which  often  arises  is  in  regard  to  the  agree- 
ment between  the  noun  and  the  verb,  as,  for  instance,  should  we 
say  ICO  grams  of  iron  are,  or  loo  grams  of  iron  is  ?  Another 
minor  point  is  in  the  writing  of  numerals  ;  whether  they  should 
be  written  out  or  the  Arabic  numerals  used.  He  has  adopted 
the  plan  of  writing  out  in  full  numbers  below  loo,  and  placing 
the  Arabic  numerals  for  loo  and  above.  This  was  an  arbitrar>' 
division,  however,  which  might  well  be  changed  if  some  agree- 
ment could  be  reached  in  regard  to  it. 


(78) 

It  appears  that  most  scientific  writers  are  so  eager  to  express 
the  truth  or  fact  which  they  wish  to  convey  that  they  lose  sight 
altogether  of  the  style  in  which  the  expression  is  made,  and,  as 
a  result,  their  sentences  become  involved,  and  their  meaning  far 
from  clear. 

Another  point  which  merits  discussion  is  in  the  use  of  proper 
names  to  indicate  any  apparatus  or  process  known  by  the  name 
of  the  inventor.  He  preferred  in  such  cases,  where  the  personal 
idea  had  been  lost,  to  use  no  capital,  but  to  write  the  name  of 
the  apparatus  or  process  with  a  small  letter,  as,  for  instance,  a 
gooch  or  an  erlenmeyer.  The  same  is  true  of  materials  or 
reagents  with  geographical  adjectives,  as  for  instance,  german 
silver  and  canada  balsam,  both  of  which  should  be  written  with 
small  letters,  just  as  the  French,  without  disparaging  the  great 
Emperor,  write  the  name  of  the  coin  napoleon  with  a  little  n  or 
as  we  write  telford  or  macadam  for  the  name  of  a  road. 

Chemists  should  be  careful  about  '*take."  It  is  not  elegant 
to  say  ''take  five  grams  and  place  in  a  dish."  "  Place  five 
grams  in  a  dish  "  is  entirely  sufficient.  In  one  work  on  chem- 
istry he  found  the  author  directing  the  analyst  to  *'  take  twenty- 
five  grams  of  Glauber  salts  "  with  a  big  g.  *' Weigh  out"  is 
inadmissable.  Wedo  not  weigh  out,  norin,  noron,  nor  under.  We 
simply  weigh.  In  measuring  it  is  not  necessary  to  say  weigh, 
as  the  chemist  knows  enough  to  use  the  balance  without  specific 
directions.  A  typically  unnecessary  form  of  expression  and  one 
not  impossible  to  find  is  ''take barium  chloride,  weigh  out  five 
grams,  dissolve  in  water  and  filter  ofiF  the  insoluble  residue." 

Above  all,  the  scientific  writer  should  avoid  indulging  in  fine 
writing.  Plain,  unvarnished  statement  of  fact  in  a  clear,  lucid 
manner  is  what  we  should  strive  for.  An  example  of  how  not 
to  write  is  the  following: 

* '  For  not  only  does  the  soil  make  possible  a  very  much  greater 
profusion  of  land  life  than  could  otherwise  exist,  but  it  has  also 
played  an  extremely  important  part  in  that  long-continued  never- 
ending,  and  sublime  process  of  evolution  whereby,  as  lands  have 
insensibly  changed  into  sea  and  seas  into  land,  as  mountains 
have  risen  so  slowly  and  silently  out  of  level  plains  as  to  spring 
their  broad  arches  directly  across  wide  rivers  to  the  height  of  a 


(79) 

mile  and  yet  leave  their  course  unaltered,  as  climates  have 
changed  from  cold  to  warm  or  from  wet  to  dry,  both  plants  and 
animals  in  this  great  drama  of  the  world  action  have  been 
enabled  to  change,  not  simply  their  costumes,  but,  if  the  exigen- 
cies of  the  new  scene  demanded  it,  legs  for  fins  or  even  abandon 
them  altogether  and  crawl  upon  their  bellies  through  the 
grass." 

Professor  Bolton  said  he  thought  one*s  grammar  school  edu- 
cation must  have  something  to  do  with  style  in  writing.  Great 
labor  is  expended  nowadays  upon  abstracts  from  foreign  publi- 
cations. He  thought  there  should  be  a  different  method  of 
treating  them.  The  results  might  be  presented  without  giving 
the  steps  by  which  they  are  reached.  He  thought  that  no  one 
could  depend  entirely  upon  the  abstracts,  but  have  to  refer  to 
the  original  papers  for  details. 

Professor  Seaman  said  he  was  glad  the  subject  was  brought 
up  ;  he  thought  some  chemists  had  an  idea  that  the  English 
language  is  sacred,  and  that  no  changes  should  be  made  in  it. 
This  feeling  must  be  met.  He  said  that  most  persons  read  by 
words,  and  not  by  letters,  and  if  a  word  has  not  the  usual 
appearance  to  which  we  are  accustomed,  our  first  impression  is 
that  it  is  wrong,  and  hence  he  feared  that  no  changes,  however 
judicious,  would  seem  agreeable.  As  to  the  agreement  of  col- 
lective nouns  with  verbs,  Goold  Brown  concluded  that  the  only 
principle  to  be  followed  is  euphony.  When  the  best  grammar- 
ians cannot  formulate  rules,  uniformity  can  hardly  be  expected. 
As  to  the  use  of  small  letters  instead  of  capitals,  the  various 
changes  that  are  going  on  in  the  language  generally  ought  to 
be  considered.  The  councils  of  biological  societies  have  agreed 
that  small  letters  should  be  used.  Up  to  a  few  years  ago  specific 
names  derived  from  proper  names  were  begun  with  capitals,  but 
now  the  small  letters  are  used.  As  to  the  abbreviations  used  for 
weights  and  measures,  he  said  no  system  had  universal  assent. 
Physicists  and  mechanicians  have  in  use  a  long  series  of  abbre- 
viations for  linear,  areal  and  cubic  measures.  In  three  of  the 
most  important  German  chemical  works,  including  the  **  Be- 
richte,'*  small  letters  are  used,  for  the  liters  (1)  ;  for  the  grams 
(g)  ;  and  for  cubic  centimeters  (cc),  and  he  would  be  in  favor 


(8o) 

of  adopting  these,  but  unfortunately  the  pharmacists  and  physi- 
cians who  are  endeavoring  to  introduce  them  into  their  arts, 
have  agreed  upon  the  capital  G  and  small  r  for  gram,  and  capi- 
tal C  and  small  c  without  a  point  between  them  for  cubic  centi- 
meter. Remsen  in  his  first  edition  used  cc.,  g.,  and  1.  Which 
are  chemists  to  follow  ?  The  French  do  not  use  habitually  any 
abbreviations.  Some  chemists,  unfortunately,  have  not  adopted 
the  new  spelling. 

Professor  Clarke  said  that  a  friend  who  wished  to  become  a 
journalist,  had  consulted  a  newspaper  man,  and  the  advice  he 
received  was  simple — *'Have  something  to  say  and  say  it." 
Sometimes  the  writer  is  not  sure  as  to  what  he  wishes  to  say, 
and  he  tries  to  say  something  else,  and  his  writing  becomes 
involved.  Another  fault  which  was  observed,  especially  in  those 
who  have  just  returned  from  abroad,  was  that  everything  that 
has  ever  been  written  upon  the  subject  is  given  in  an  article, 
and  the  discovery  of  the  authors  is  either  buried  in  the  mass  or 
occupies  a  very  small  place  at  the  end  of  the  article.  He  thought 
a  logical  order  should  be  followed,  and  an  effort  should  be  made 
to  state  what  is  said,  simply  and  clearly.  He  thought  Steele's 
**  Fourteen  Weeks  in  Chemistry'*  was  a  model  of  bad  style. 

Professor  Munroe  said  that  so  far  in  the  discussion  there  was 
apparently  very  little  difference  of  opinion.  He  read  the  fol- 
lowing from  an  article  in  **  Science.**  **  If  we  hear  a  baby  cry- 
ing with  two  ears,  are  we  to  think  it  is  twins?  '*  as  an  example 
of  style  in  a  scientific  article.  He  thought  this  illustrated 
Professor  Clarke's  remarks  about  having  something  to  say  and 
saying  it.  Professor  Munroe  thought  that  the  question  of  style 
had  to  be  considered  from  two  standpoints  :  that  of  the  manual 
or  text-book,  and  that  of  the  technical  or  scientific  paper. 
Abbreviations  that  might  be  properly  included  in  the  latter, 
should  not  be  introduced  in  the  former  until  they  have  long 
been  used  in  technical  literature.  He  was  especially  doubtful 
as  to  the  advisibility  of  changing  the  adjectives  to  the  substan- 
tive as  a  '*gooch,  a  bunsen,  a  ruhmkorff,  or  a  wiley,"  and  the 
latter  suggests  that  where  one  is  as  fertile  as  our  distinguished 
associate,  there  may  be  a  difficulty  in  determining  which  one  of 
his  many  devices  shall  be  called  a  ** wiley." 


(8i) 

Dr.  Fireman  said  he  did  not  agree  with  what  had  been  said  as 
to  abstracts.  Many  papers  are  not  accessible,  and  possibly  only 
one  journal  could  be  obtained.  With  a  good  abstract  the 
description  of  a  process  may  be  of  use.  Neither  did  he  agree 
with  the  idea  that  an  historical  sketch  should  be  introduced. 
He  thought  a  summary  was  frequently  of  greater  use.  They 
are  generally  brief  and  give  valuable  references. 

Dr.  de  Schweinitz  closed  the  discussion  by  saying  that  he 
agreed  with  Professor  Munroe  that  the  style  should  differ  in 
text-books  and  in  technical  papers.  What  is  proper  in  one  is 
not  so  in  the  other.  He  thought  that  abbreviations  should  be 
dropped  as  the  purpose  is  to  make  what  is  written  useful  to  all, 
and  he  thought  the  ideas  and  the  statements  should  be  expressed 
as  simply  and  clearly  as  possible. 

A  regular  meeting  Was  held  Thursday,  April  9, 1896,  with  the 
the  President,  Dr.  K.  A.  de  Schweinitz  in  the  chair,  and  thirty 
members  and  ten  guests  present. 

The  minutes  of  the  eighty-seventh  meeting  were  read  and 
approved. 

A  letter  from  Dr.  Salmon,  inclosing  a  circular  letter  from  the 
Director  of  the  Pasteur  Institute  in  Paris  was  read,  asking  the 
society  to  appoint  a  member  to  represent  it  upon  the  committee 
to  raise  funds  for  the  erection  of  a  monument  in  Paris  to  Pas- 
teur. The  President,  Dr.  de  Schweinitz,  was  unanimously 
elected  to  represent  the  Chemical  Society  upon  this  committee. 

There  being  no  further  business  the  reading  of  papers  was 
proceeded  with. 

The  first  paper  of  the  evening  was  by  Mr,  V.  K.  Chestnut 
upon  *'  Some  Vegetable  Skin  Irritants  and  their  Chemical  Com- 
position." The  paper  consisted  of  a  review  of  the  work  of  Dun- 
stan  and  Miss  Boole  on  Croton  Oil,  and  of  Pfaff  on  Toxicoden- 
drol,  a  new  oil-like  body  from  the  poison  ivy,  Rhus  radicans; 
together  with  an  account  of  some  vesicating  plants  which  have 
been  but  little  studied.  Specimens  of  this  plant  were  exhibited 
and  the  effect  of  an  alcoholic  solution  of  lead  acetate  as  an  anti- 
dote to  Rhus  poisoning  was  illustrated  by  experiments  carried 


(82) 

out  by  the  writer  on  himself.  These  experiments  also  showed 
conclusively  that  toxicodendrol  was  the  vesicating  principle  of 
the  poisonous  species  of  Rhus.  Discussion  was  by  Messrs. 
Tassin,  Munroe,  Cutter,  Stewart,  Fireman  and  de  Schweinitz. 
Mr.  Tassin  asked  whether  it  was  the  lead  acetate  or  the  alco- 
hol that  is  the  antidote.  Mr.  Chestnut  answered  that  the  alco- 
holic solution  of  lead  acetate  is  the  best  remedial  agent.  If  the 
oil  is  kept  long  enough  on  persons  supposed  not  to  be  suscepti- 
ble, they  will  be  poisoned;  the  poisoning  may  take  place  at  the 
end  of  twelve  hours,  or  not  for  five  days.  Portions  of  the  skin 
that  are  thick  are  not  so  easily  affected  as  are  those  where  it  is 
thin.  Professor  Munroe  gave  his  experience  as  to  nitrobenzol, 
which  he  had  used  in  considerable  quantity,  and  to  which  he 
and  the  workmen  were  exposed  ;  it  was  inhaled  as  vapor,  and 
came  in  contact  with  the  skin,  but  no  one  was  poisoned.  The 
vapor  is  suffocating,  but  the  workmen  soon  became  accustomed 
to  it.  All  the  books,  however,  state  that  it  is  poisonous.  Mr. 
Cutter  said  that  he  could  uphold  the  books,  as  he  had  experi- 
enced its  poisonous  effects ;  he  had  rigor,  fever,  chills,  and  pal- 
pitation of  the  heart,  and  was  unconscious  afterwards ;  the  effects 
lasted  for  three  days,  and  the  smell  even  now  would  affect  him. 
Dr.  Stewart  gave  his  opinion  of  its  poisonous  effects  upon  the 
skin  ;  in  his  own  case  it  had  caused  an  eruption  that  lasted  three 
or  four  hours.  Dr.  Fireman  thought  that  different  effects  might 
be  produced  by  vapors  and  by  the  liquid;  he  referred  to  the  effect 
of  hydrofluoric  acid  vapors,  which  are  not  poisonous  in  any 
degree,  although  the  liquid  was  well  known  to  be  very  poison- 
ous. Professor  Munroe  thought  there  might  be  differences  in 
the  substance.  Dr.  de  Schweinitz  thought  that  possibly  it  was 
impure  in  the  cases  cited  by  Dr.  Cutter  and  Mr.  Stewart. 

Mr.  Ewell  read  the  second  paper  of  the  evening  on  **  The 
Effect  of  Acidity  on  the  Development  of  the  Nitrifying 
Organisms,'*  by  E.  E.  Ewell  and  H.  W.  Wiley. 

*'  While  it  has  been  known  for  many  years  that  active  nitrifi- 
cation occurs  only  in  the  presence  of  some  basic  substance 
capable  of  neutralizing  the  free  acid  as  fast  as  it  can  be  formed, 
very  little  time  has  been  devoted  to  the  study  of  the  exact  degree 
of  acidity   that  the   nitrifying  organisms   can    endure.       As 


(83) 

the  authors  had  some  forty  samples  of  soil  at  their  disposal  during 
the  last  year  for  other  purposes,  it  seemed  wise  to  improve  the 
opportunity  to  test  the  influence  of  acidity  on  the  nitrifying  or- 
ganism contained  in  the  soils  from  various  parts  of  the  country. 
Tests  were  made  with  forty-four  different  soils,  from  twenty-two 
states  and  territories.  The  results  showed  great  uniformity  in 
the  relation  to  acidity  of  the  organisms  contained  in  the  various 
soils.     Excluding  five  tests  in  which  no  nitrification  exists,  and 

m 

five  tests  in  which  it  was  excessive  because  of  the  calcareous 
nature  of  the  soils  used  for  the  seeding  of  the  cultures,  the 
average  amount  of  nitrogen  nitrified  was  twenty  parts  per  mil- 
lion ;  the  minimum  result  of  the  thirty-four  tests  included  in 
this  average  was  eleven,  and  the  maximum  twenty-five  parts  per 
million.  The  tests  are  to  be  repeated  with  pure  cultures  of  the 
nitrifying  organisms  of  the  same  soils.  This  series  of  experi- 
ments was  made  as  a  study  of  the  nitrous  organisms  only,  but 
the  results  show  that  the  nitric  organisms  are  not  more  sentitive 
to  acidity  than  the  nitrous  organisms,  the  final  product  being 
nitrate  in  nearly  every  case. 

Dr.  de  Schweinitz,  referring  to  the  action  of  acids  on  the 
growth  of  bacteria,  said  they  seemed  to  be  able  to  accommodate 
themselves  to  their  environment,  especially  in  the  case  of  the 
tuberculous  bacillus,  and  after  a  time  they  seemed  to  grow  bet- 
ter in  an  acid  medium  than  in  any  other,  though  at  first  they 
needed  coaxing. 

The  third  paper  of  the  evening  was  on  '  *  The  Chemistry  of  the 
Cactacese,'*  by  E.  E.  Ewell. 

Until  very  recently  other  species  of  cacti  than  Cereus  grandi- 
florus  and  a  few  related  species  have  generally  been  regarded  as 
devoid  of  constituents  of  pharmacological  value.  These  and 
other  species,  have  been  used  in  medical  practice  in  the  coun- 
tries in  which  they  grow,  but  their  use  has  rarely  extended  to 
the  more  civilized  nations.  Species  of  the  genus  Anhalonium 
have  long  been  used  for  curative  and  ceremonial  purposes  by  the 
Indians  of  Mexico,  and  the  southwestern  parts  of  our  own  coun- 
try. They  found  places  in  the  Mexican  pharmacopoeia  of  1842, 
under  the  name  of  "pellote,"  or  **Peyotl,"  but  have  been 
omitted  from  the  later  editions.     The  dried  aerial  portions  of 


(84) 

species  Anhahnium  figure  in  the  commerce  of  our  southwestern 
border  under  the  name  "  mescal  buttons."  The  species  of  this 
genus  have  been  the  subject  of  scientific  investigation  by  at 
least  three  groups  of  persons  during  recent  years :  First,  a  group 
of  persons  at  Berlin,  where  the  work  was  beg^nby  Dr.  L.  Lewin, 
the  crude  material  being  supplied  to  him  by  Messrs.  Parke, 
Davis  &  Co.,  of  Detroit;  second,  a  group  of  persons  at  the 
Pharmacological  Institute  at  Leipsic,  where  the  work  has  been 
conducted  by  Dr.  Arthur  Hefifter ;  third,  a  group  of  persons  in 
this  country,  centering  in  the  Bureau  of  American  Ethnology, 
and  including  as  associates  the  Division  of  Chemistry  of  the 
United  States  Department  of  Agriculture  for  Chemical  studies, 
Drs.  Prentiss  and  Morgan  for  a  study  of  physiological  proper- 
ties, and  the  Botanical  Division  of  the  United  States  Depart- 
ment of  Agriculture  for  the  settlement  of  botanical  questions. 

Lewin  reported  the  presence  of  an  alkaloid  in  AnhaUmium 
lewinii  in  1888.  He  has  given  this  the  name  of  anhalonin  and 
made  an  extended  report  on  its  physical,  chemical  and  physio- 
logical properties  in  December,  1894.  He  has  also  found  evi- 
dence of  physiologically  active  substances  in  the  related  species. 

In  August,  1894,  HefiFter  reported  the  presence  of  a  poisonous 
alkaloid  ivLA./issuratum,  to  which  he  gave  the  name  anhalin; 
he  found  an  extract  of  A,  prismaHcum  to  be  physiologically 
active,  but  did  not  have  sufficient  material  for  a  more  extended 
study  ;  he  separated  an  alkaloid  that  he  named  pellotin  from^^. 
Tvilliamsi;  in  A,  lewinii  h^  found  evidence  of  the  presence  of 
three  alkaloids,  the  description  of  the  first  of  which  accords  with 
the  description  of  Lewin's  anhalonin.  He  made  an  extended 
study  of  the  chemical,  physical  and  physiological  properties  of 
anhalin  and  pellotin. 

In  this  country,  the  separation  of  the  constituents  of  these 
plants,  and  the  study  of  the  action  of  the  substances  thus 
obtained  as  well  as  of  the  crude  materials,  upon  men  and  the 
lower  animals,  were  begun  in  the  autumn  of  1894,  but  before 
receiving  the  paper  of  Heffter.  A,  lewinii^  in  the  form  of 
"mescal  buttons,''  has  served  as  the  material  for  these  studies. 
Anhalonin  and  a  second  alkaloid  have  been  separated  in  con- 
siderable quantity.     These,  as  well  as  other  constituents  of  the 


(85) 

<lrug,  including  one  or  more  resins,  are  turned  over  to  Drs. 
Prentiss  and  Morgan  for  physiological  experiments,  as  rapidly 
as  they  are  obtained  in  an  appropriate  state  of  purity*  A  com- 
plete chemical  study  of  the  constituents  of  the  plant  is  in  pro- 
cess, including  those  substances  of  interest  to  the  vegetable 
physiologist  as  well  as  those  of  interest  to  the  therapeutist. 

The  paper  was  illustrated  with  specimens  of  the  cactus  of  dif- 
ferent varieties  from  the  Botanical  Gardens  and  the  Department 
of  Agriculture.  Mr.  Mooney  followed  with  a  paper  on  ^*The 
.Mescal  Ceremony  among  the  Indians.*' 

The  mescal  plant  is  a  small  variety  of  cactus  native  to  the 
lower  Rio  Grande  Region,  and  about  the  Pecos  River  in  East- 
em  New  Mexico.  The  botanical  name  has  finally  been  fixed 
by  Professor  Coulter  as  Lophophora  ztnUtamst.  Mescal  is  the 
name  by  which  it  is  known  to  the  Indian  traders,  but  it  is  not  to  be 
confounded  with  the  other  mescal  (Maguey)  of  Arizona.  The 
local  Mexican  name  ispeyote,  a  corruption  of  the  original  Aztec 
name,  from  which  it  would  seem  that  the  plant  and  ceremony 
were  known  as  far  south  as  the  valley  of  Mexico,  at  a  period 
antedating  the  Spanish  conquest.  Several  closely  related 
species  are  described  by  Lumholtz  as  being  used  with  ceremo- 
nial rites  among  the  tribes  of  the  Sierra  Madre, 

The  dry  tops,  when  eaten,  produce  such  marked  stimulating 
and  medicinal  results  and  such  wonderiuUy  beautiful  psycho- 
logic effects,  without  any  injurious  reaction,  that  the  tribes  of 
the  region  regard  the  plant  as  the  vegetable  incarnation  of  the 
deity,  and  eat  it  at  regular  intervals  with  solemn  religious  cere- 
mony of  song,  prayer  and  ritual.  The  ceremonial  and  medi- 
cinal use  of  the  plant  was  first  brought  to  public  notice  by  James 
Mooney  in  a  lecture  delivered  before  the  Anthropological  Society 
of  Washington  in  1891,  as  the  result  of  studies  made  among  the 
Kiowas  and  associated  tribes  of  Western  Oklahoma.  As  the 
ceremony  is  forbidden,  and  the  trade  in  the  plant  made  contra- 
band upon  the  reservations,  the  investigation  was  a  matter  of 
some  difficulty.  In  1894  Mr.  Mooney  brought  back  a  large 
quantity  of  the  dried  mescal,  which  was  turned  over  to  the 
chemists  of  the  Agricultural  Department  for  analysis,  and  to 
Drs.  W.  P.   Prentiss  and  P.  P.   Morgan,  of  Washington,  for 


(86) 

medical  experimentation.  The  results  thus  far  would  seem  to 
indicate  that  the  Indians  are  right  in  asserting  that  they  have 
discovered  in  the  mescal  a  valuable  medicine  entirely  unknown 
to  science,  and  which  will  probably  take  its  place  in  our  pharm- 
acopoeia along  with  those  other  Indian  remedies,  quinine  and 
coca.  The  ceremony  amd  songs  are  briefly  described  by  Dr. 
Mooney,  whose  full  investigation  of  the  subject  will  ultimately 
appear  in  one  of  the  publications  of  the  Bureau  of  American 
Ethnology. 

Dr.  Francis  P.  Morgan  followed  with  a  paper  on  the  **  Physi- 
ological Action  and  Medicinal  Value  of  Anhalonium  lewinii. 
(**  Mescal  Buttons.")  '*  Dr.  Morgan  stated  that  the  investiga- 
tion had  been  intrusted  to  Dr.  D.  W.  Prentiss,  with  whom  he 
was  associated.  Experiments  were  tried  and  observations  taken 
at  regular  intervals  to  determine  the  action  of  the  entire  button 
on  the  system.  The  most  striking  result  was  the  production  of 
visions  of  the  most  remarkable  kind  with  e^^es  closed,  and 
especially  so  in  the  dark.  Changes  of  color  were  character- 
istic ;  tubes  of  shining  light,  figures,  cubes,  balls,  faces,  land- 
scapes, dances  and  designs  of  changing  colors  were  among  the 
most  persistent  visions.  They  were  hardly  seen  with  the  eyes 
open  ;  in  full  dose  no  effect  on  the  reason  or  will  is  noticed  in 
most  cases.  There  was  direct  stimulation  of  the  centers  of 
vision  and  dilatation  of  the  pupils.  About  one-quarter  of  the 
quantity  or  three  buttons,  are  sufficient  to  give  the  visions  in  the 
case  of  white  men.  Dr.  Morgan  detailed  the  experiences  of  dif- 
ferent persons  who  had  tried  the  experiments.  In  some  cases 
there  was  slowing  of  the  heart,  from  seventy-five  to  forty-five 
beats,  followed  by  a  risQ  to  normal ;  there  is  also  inability  to 
sleep,  and  a  loss  of  the  sense  of  time — hours  seem  to  intervene 
between  words.  The  physiological  action  is  not  identical  with 
that  of  any  known  drug,  it  is  unlike  cannabis  indica,  cocaine, 
etc.  The  constituents  of  the  mescal  buttons  are  being  experi- 
mented with,  but  the  investigations  are  still  incomplete. 
Anhalonin  causes  increased  reflex  irritability  and  convulsions, 
like  strychnine.  It  is  evidently  not  the  active  principle ; 
another  constituent  has  been  isolated  whose  action  is  widely 
different.     It  does  not  cause  opisthotonos,  nor  tetanus,  and  has 


(8?) 

no  action  like  that  of  strychnine.  A  third  principle  has  also 
been  isolated.  The  resin  is  supposed  to  be  the  active  principle 
and  will  probably  be  of  use  in  medicine.  The  experiments  are 
still  being  conducted  and  will  be  detailed  later  on. 

Dr.  de  Schweinitz  expressed  the  indebtedness  of  the  Society 
to  Mr,  Mooney  and  to  Dr.  Morgan,  and  said  that  the  further 
results  would  be  of  interest  to  the  Society. 

A  regular  meeting  was  held  Thursday,  May  14,  1896.  The 
president,  Dr.  de  Schweinitz  in  the  chair,  with  twenty-three 
members  present.  Mr.  Mayville  W.  Twitchell  was  elected  as 
associate  member  and  Mr.  Charles  N.  Forrest  as  member.  The 
president  presented  the  following  resolutions,  endorsed  by  the 
executive  committee,  and  the  Society  adopted  them;  the 
president  and  secretary  were  instructed  to  sign  and  transmit 
them. 

Washington,  D.  C,  May  14,  1896. 

To  THE  Honorable,  The  President  of  the  United  States 
Senate. 

Dear  Sir  : — In  view  of  the  proposed  legislation  now  before 
the  Senate  in  the  form  of  a  bill  entitled  **  An  Act  for  the  further 
prevention  of  cruelty  to  animals  in  the  District  of  Columbia,** 
which,  however,  is  practically  an  act  to  limit,  and  eventually 
stop,  all  experiments  upon  animals  in  the  District  of  Columbia, 
the  Chemical  Spciety  of  Washington,  including  among  its  mem- 
bers a  number  of  the  most  prominent  chemists  in  the  country, 
desires  to  present  to  the  Senate  of  the  United  States  a  formal 
and  positive  protest  against  the  enactment  of  any  legislation 
upon  the  subject  of  vivisection. 

The  laws  at  present  on  the  Statute  books  of  the  District  of 
Columbia,  if  properly  carried  out,  will  apply  to  all  cases  of 
cruelty  to  animals  which  exist  in  this  District.  The  proposed 
bill  is  objectionable  for  very  many  reasons.  The  penalties  pre- 
scribed for  the  infraction  of  .the  law  are  preposterous.  An 
expert  who  did  not  happen  to  possess  a  permit  from  the  District 
Commissioners  for  the  performance  of  experiments  upon  animals 
might  suddenly  have  placed  in  his  hands  material,  the  danger- 
ous character  of  which  could  only  be  determined  by  an  imme- 
diate experiment  upon  an  animal.  Should  such  a  test  be  made 
without  a  license,  though  possibly  the  lives  of  hundreds  of  peo- 


(88) 

pie  were  involved,  the  experimenter  would  be  subject  to  an 
enormous  fine  and  imprisonment,  for  having  in  the  interests  of 
humanity  inoculated  a  guinea  pig,  or  a  rabbit,  or  some  other 
animal,  without  a  formal  permit  from  the  District  Commission- 
ers. 

While  the  majority  of*  the  members  of  our  Society  are  not 
directly  engaged  in  experiments  in  which  animals  are  used,  we 
know  that  in  certain  lines  of  work,  toxicology,  materia  medica, 
biochemistry,  and  the  like,  animal  experimentation  is  absolutely 
necessary  for  the  advancement  of  knowledge. 

The  agitators  of  the  proposed  legislation  have  not  been  able 
to  show  a  single  instance  of  cruel  experiments  conducted  in  the 
District  of  Columbia,  either  in  any  of  the  laboratories,  or  med- 
ical colleges,  or  public  schools,  consequently  there  is  no  need 
for  any  law  on  the  subject.  Furthermore,  Washington  is 
becoming  the  center  of  education  for  the  entire  United  States. 
Four  large  universities  are  located  here ;  several  more^  are  in 
prospect,  and  the  proposed  legislation  would  hamper  and  event- 
ually destroy  all  possibility  for  advanced  postgraduate  work  in 
the  biological  science,  and  indirectly  in  all  allied  branches. 

We  therefore,  collectively  as  a  Society,  and  individually  as 
members,  desire  to  protest  strenuously  against  any  legislation  on 
the  subject  of  vivisection,  deeming  it  to  be  unwise,  unnecessary, 
and  in  direct  opposition  to  the  spirit  which  has  for  a  number  of 
years  actuated  the  United  States  government  to  encourage  the 
advance  of  science.  We  hold  further  that  such  legislation  would 
be  a  direct  contradiction  of  the  well-known  practical  results  that 
have  already  been  obtained  by  scientific  investigations  con- 
ducted under  the  government,  which  have  made  possible  the 
saving  of  many  thousand  dollars  worth  of  property  and  many 
human  lives. 

Yours  very  respectfully, 

[Signed]  E.  A.  de  Schweinitz, 

Pres.  Wash.  Chem.  Soc. 
A.  C.  Peale, 

Secretary. 

The  president,  representing  the  Society,  as  a  member  of  the 
Pasteur  Committee,  reported  that  the  committee  had  organized 
and  was  ready  to  receive  subscriptions. 

There  being  no  further  business  the  reading  of  papers  was 
proceeded  with,  Vice  President  Bigelow  taking  the  chair. 

The  first  paper  was  by  Mr.  Frederick  P.  Dewey,  on  **  Practi- 
cal Analytical  Accuracy." 


(89) 

The  paper  did  not  go  into  the  m^ans  of  securing  accuracy, 
but  dealt  entirely  with  the  results  actually  obtained  when  a 
number  of  chemists  worked  upon  the  same  sample.  Not  very 
much  has  been  published  in  this  line,  but  sufficient  has  been 
done  to  show  that  the  ordinary  accuracy  of  analytical  work  is 
not  what  it  ought  to  be  and  that  there  is  room  for  much  im- 
provement. 

The  paper  gave  results  from  analytical  symposiums  published 
in  the  Transactions  of  the  American  Institute  of  Mining  Engi- 
neers and  the  Proceedings  of  the  Association  of  Official  Agri- 
cultural Chemists.  It  was  also  somewhat  historical  in  charac* 
ter  in  tracing  the  development  of  accuracy  in  some  determina- 
tions. 

The  discussion  of  Mr.  Dewey's  paper  was  by  Messrs.  Bigelow 
and  Clarke. 

Mr.  Bigelow  said  that  Mr.  Dewey  had  selectedthe  most  accu- 
rate of  the  determinations  by  the  Official  Agricultural  Chemists, 
but  Dr.  Dewey  said  he  had  simply  taken  those  most  nearly  in 
his  own  line  of  work.  He  referred  to  Campbell's  tables  in  the 
Journal  and  thought  it  was  unfortunate  that  nothing  was  said  as 
to  the  way  they  were  obtained  nor  how  he  was  led  to  adopt  the 
various  figures.  He  allows  usually  a  small  variation,  but  with 
silica  he  allows  a  variation  of  over  one-half  per  cent,  when  large 
quantities  are  present. 

Prof.  Clarke  thought  that  a  great  source  of  variation  between 
different  observers  was  due  to  the  fact  that  too  great  faith  was 
placed  in  the  reagents.  The  work  was  not  done  with  the  same 
reagents  and  reagents  are  not  always  the  same.  Here,  there- 
fore, is  a  source  of  error. 

Mr.  Bigelow  said  he  thought  another  source  of  error  was  due 
to  the  fact  that  the  work  was  often  done  by  students  or  subordi- 
nates and  these  results  were  published  with  the  others. 

Mr.  Dewey  said  his  paper  was  intended  to  exhibit  what  was 
actually  obtained  every  day.  He  thought  the  assistant's  work 
was  not  always  the  poorest.  The  principal  very  often  was  out  of 
practice. 

Dr.  P.  Fireman  read  a  paper  on  **  A  New  Mode  of  Formation 
of  Tertiary  and  Quartemary  Phosphines."      When  phosphoni- 


(90) 

um  iodide  is  heated  with  ether  in  a  sealed  tube  at  160'',  for  six 
hours,  both  ethyl  groups  of  the  ether  became  available  for  sub- 
stitution, and  these  form  the  hydriodic  salts  of  triethj'l-  and 
tetraethylphosphine  according  to  these  equations  : 

2PH,I  +  3(C,H.),0  =  2P(C.H.)..HI  +  3H.O. 
2PH,I  +  4(C.H,),0  =  2P(C,HJ,I  +  4H,0. 

The  author  is  at  present  occupied  with  the  preparation  of  the 
homologous  phosphines  by  the  action  of  PH  J  and  homologous 
ethers.  He  is  also  experimenting  with  a  view  of  obtaining 
amines  by  the  action  of  ammonium  iodide  or  ammonium  bromide 
on  ethers  or  alcohols. 

In  the  discussion  of  Dr.  Fireman's  paper  Dr.  Stokes  asked  if 
he  had  obtained  any  traces  of  primary  or  secondar>'  phosphines. 
Dr.  Fireman  answered  that  primary  phosphines  are  excluded 
and  as  to  secondary  phosphines  he  had  at  one  time  hoped  he 
had  obtained  them,  but  he  could  not  say  with  any  certainty  that 
he  had.  Dr.  Stokes  thought  they  might  be  recognized  bj-  the 
odor.     Tertiary  phosphines  have  the  odor  of  hyacinths. 

Dr.  Stokes  then  read  a  paper  on  **  Metaphosphinic  Acids.*' 

Dr.  H.N.  Stokes  spoke  on  his  investigation  of  the  metaphos- 
phinic acids,  a  series  of  acids  having  the  general  formula 
(PNO,H,)n,  i'  ^.i  metaphosphoric  acijds  in  which  one-third  of 
the  oxygien  is  replaced  by  NH.  They  are  not,  however,  stricth' 
speaking,  derivatives  of  metaphosphoric  acids,  for  while  these 
contain  a  nucleus  consisting  of  phosphorus  atoms  united  by 
oxygen,  the  metaphosphinic  acids,  as  proved  by  their  forma- 
tion from  the  chloronitrides(PNCl,)„,  have  a  nucleus  consisting 
of  alternate  phosphorus  and  nitrogen  atoms.  Two  members  of 
the  series  have  been  studied,  viz,,  trimetaphosphinic  and  tetra- 
metaphosphinic  acids,  PjNjOjHo  and  P^N^O.H^,  derived  from 
PjNjCl^  and  P,N,C1,.  Trimetaphosphinic  acid  apparently  has 
the  tautomeric  formulas 

P(OH),  PO.OH 

/  ^                                                     /    \ 

N  N                                            HN           NH 

II  I  and                      II 

(HO),P  P(OH),                          HO.PO          PO.OH 

\    ^  W   / 

N  NH 


(91) 

Salts  of  both  forms  have  been  obtained.     Under  the  action  of 

acids,   a    successive  decomposition  is  effected  into  P,N,0,H,, 

P.NO.H,,  H,P,0„  and  H.PO,.     The  second  of  these  may  be 

regarded  either  as  PO(OH),.NH.PO(OH),  or  PO(OH),.0. 

NH 
PO  <Cr)TT''  ^^^  former  being  supported  bj"  its  derivation  from 

P,N,OeH„  the  latter  by  its  easy  conversion  into  pyrophosphoric 
acid.  It  presents  a  peculiar  case  of  intra-molecular  wandering  of  the 
nitrogen  atom,  possibly  to  be  explained  by  a  process  analogous 
to  Beckmann's  transformation.  (Particulars  mil  appear  in  the 
American  Chemical  Journal. ) 

Dr.  Fireman  said  the  compounds  appeared  to  him  to  be  of 
similar  constitution  as  cyanic  acid  and  its  derivatives ;  and  also 
in  regard  to  the  tendency  to  polymerize  and  to  appear  in  iso- 
meric forms,  there  is  a  striking  resemblance  between  both  classes 
of  compounds.  He  thought  the  results  would  be  of  interest  in 
theoretical  chemistry.  As  to  the  silver  salts  he  asked  Dr. 
Stokes  if  he  had  tried  to  prepare  the  esters. 

Dr.  Stokes  said  that  a  number  of  lactams  are  known  that  are 
stable,  that  open  out  or  have  open  rings  in  which  the  tendency 
to  break  up  is  not  marked.  The  tri  acid  is  easily  broken  up  but 
the  tetra  acid  is  not.  In  regard  to  the  esters,  he  had  tried  to 
get  them  from  the  silver  salts,  but  thej''  are  not  like  the  organic 
ethers.  They  are  very  unpleasant  to  deal  with  and  are  coupled 
mixtures  with  which  little  can  be  done.  He  had  thought  of  the 
analogy  with  cyanic  acid  and  especially  with  the  cyanuric  acid 
compounds.  Another  theoretical  point  is  the  possibility  of 
stereo-isomeric  forms.  When  you  have  an  oxime  there  are  two 
isomers  known.  They  split  up  differently  — cis  and  trans  bodies. 
He  had  not  been  able  to  find  triphosphinic  acid  in  anj^  but  two 
tautomeric  forms. 

Prof.  Munroe  said  he  was  glad  to  see  that  the  methods  of 
organic  chemistry  were  being  applied  to  inorganic  chemistrj^ 
but  he  would  like  to  know  why  the  linear  form  of  the  salts 
was  written  one  way,  and  read  in  the  reverse  way  in  or- 
ganic chemistry.     He  thought  it  was  very  confusing  to  students. 

Dr.  Stokes  thought  it  was  a  matter  of  custom. 


(92) 

Dr.  Fireman  said  he  thought  that  the  reason  for  writing  the 
formula  of  the  organic  acid  first  and  then  that  of  the  metal  was 
due  to  the  fact  that  the  formulas  of  the  organic  acids  were  usu- 
ally of  a  complicated  nature  and  therefore  it  is  natural  to  dis- 
pose of  them  first  and  that  afterwards  it  is  an  easy  task  to  fit  in 
the  symbol  of  the  metal. 

Prof.  Munroe  thought  this  was  not  the  explanation. 
«  Dr.  Stokes  said  different  men  developed  the  two  methods 
working  from  two  different  sides  when  a  series  of  homologous 
compounds  is  written  out.  The  constants  are  put  down  first 
and  then  those  that  vary.  This  was  why  he  wrote  them  this 
way. 

Prof.  Seaman  thought  this  the  correct  explanation .  Ideas  have 
a  different  arrangement  with  different  people,  normal  to  each 
one.  He  thought  there  was  more  uniformity  in  this  country 
than  anywhere  else. 

The  Society  adjourned  until  November. 


Issued  with  July  Number,  1896. 


Proceedings. 


NBW  MEMBERS  ELECTED  JUNE  25,  1 896. 

Hanna,  Prof.  Geo.  P.,  Charlotte,  N.  C. 

Harsh,  S.  A.,  Revenue  Gold  Mining  &  Milling  Co.,  Norris, 
Mont. 

Lederle,  Ernest  Joseph,  Ph.D.,  Health  Department,  N.  Y. 
City. 

Ludwig,  H.  T.  J.,  Mount  Pleasant,  N.  C. 

McFetridge.  Joseph,  Natrona,  Pa. 

Melville,  W.,  Woodmere,  Mich. 

Mewbome,  R.  G.,  Raleigh,  N.  C. 

Miller,  H.  K.,  Raleigh,  N.  C. 

Parmelee,  CuUen  W.,  io8  Tenth  St.,Greenpoint,  Long  Island. 

Pegram,  W.  H.,  Durham,  N.  C. 

Smalley,  Frank  W.,  University  of  Cincinnati,  Cincinnati,  O. 

Thompson,  F.,  102  East  Seventh  St.,  Covington,  Ky. 

Uhlig,  E.  C,  care  of  Whitall,  Tatum  &  Co.,  46-48  Barclay 
St.,  N.  Y.  City. 

Wood,  Joseph  R.,  240  Green  Ave.,  Brooklyn,  N.  Y. 

ASSOCIATES  ELECTED  JUNE  25,  1896. 

Howell,  John  W.,  Edison  I^amp  Works,  Newark,  N.  J. 
Twining,  T.  E.,  Newark,  Ohio. 

CHANGES  OP  ADDRESS. 

Behr,  Amo,  P.  O.  Box  I,  Jersey  City. 
Eakins,  I^.  G.,  Box  434,  Florence,  Colo. 
Lane,  H.  M.,  care  of  Great  Falls  Iron  Works,  Great  Falls, 
Mont. 

Lenher,  V.,  Mechanicsburg,  Pa. 

Peter,  Alfred  M.,  236  East  Maxwell  St.,  Lexington,  Ky. 

Sargent,  Geo.  W.,  Bellwood,  Blair  Co.,  Pa. 


MEETINGS  OF  THE  SECTIONS. 

CINCINNATI  SECTION. 

The  regular  meeting  of  the  Section  was  held  Saturday  even- 
ing, May  1 6th. 

Dr.  S.  P.  Kramer  presented  **  Some  New  Facts  Concerning  X 
Rays,"  giving  an  interesting  account  of  some  experiments,  and 


(94) 

exhibiting  some  negatives  made  by  means  of  his  new  five  plate 
Topler-Holtz  machine. 

Mr.  B.  M.  Pilhashy,  of  Cincinnati  and  Mr.  J.  N.  Hurty,  of 
Indianapolis,  were  elected  members  of  the  Section. 

The  meeting  adjourned  until  October  15th. 

NEBRASKA  SBCTION. 

The  Nebraska  Section  held  its  fourth  regular  meeting  on  June 
5th,  at  8:00  p.  M. 

The  meeting  was  called  to  order  by  the  president.  In  the 
absence  of  the  secretary,  Mr.  J.  B.  Becher  was  elected  Secretary 
pro  tern. 

The  minutes  of  the  last  meeting  were  read  and  adopted. 

The  following  officers  were  elected  for  the  ensuing  year  : 
President,  H.  H.  Nicholson;  Secretary  aud  Treasurer,  Dr. 
John  White ;  Executive  Committee,  Samuel  Avery,  R.  S.  Hilt- 
ner  and  J.  F.  Becher. 

The  Secretary's  report  was  read  and  approved. 

The  Treasurer's  report  having  been  read,  Mr.  E.  C.  Elliott 
and  Miss  Rosa  Bouton  were  appointed  an  auditing  committee. 

The  committee  pronounced  the  report  correct,  whereupon  it 
was  approved. 

A  letter  from  J.  Stanley  Brown,  Secretary  of  the  Joint  Com- 
mission of  the  Scientific  Societies  of  Washington,  was  then  read, 
calling  attention  to  the  anti-vivisection  bill  now  pending  before 
Congress. 

The  President  appointed  a  committee  to  draft  suitable  resolu- 
tions, which  were  adopted. 

**  Whereas,  There  is  now  pending  before  the  Congress  of 
the  United  States  a  bill  entitled  *A  bill  for  the  further  preven- 
tion of  cruelty  to  animals  in  the  District  of  Columbia  ; '  and 

Whereas,  In  our  opinion  such  legislation  is  opposed  to  the 
proper  development  of  biological  and  medical  science  ;  and 

Whereas,  It  is  feared  that  such  legislation  may  be  further 
extended  to  the  several  states  and  territories,  thereby  very  seri- 
ously restricting  the  progress  of  scientific  investigation  ;  be  it 

Resolved^  That  the  Nebraska  Section  of  the  American  Chemi- 
cal Society  most  earnestly  protests  against  the  enactment  of  such 


(95) 

legislation  as  being  unnecessary  and  prejudicial  to  the  best 
interests  of  mankind. 

Signed, 

H.  H.  NiCHOMON,  President. 
Rosa  Bouton, 
John  Whitb, 
Edward  Elliott, 

Committee, 

Mr.  Benton  Dales  was  elected  an  associate  member. 

In  the  absence  of  Dr.  White,  his  paper  entitled  ** Contributions 
to  the  Chemistry  of  the  Suboxides,"  Iwas  read  by  Mr,  E.  E. 
Nicholson. 

NBW   YORK   SBCTION. 

The  June  meeting  of  the  New  York  Section  was  held  on  Fri- 
day evening,  June  5th,  at  the  College  of  the  City  of  New  York, 
Prof.  A.  A.  Breneman  presiding. 

After  the  reading  of  the  minutes  the  chairman  of  the  commit- 
tee on  Organization  of  the  Chemical  Club  reported  that  at  a 
recent  meeting  of  the  committee,  held  at  the  Board  of  Trade, 
much  enthusiasm  ws  shown,  and  the  movement  was  making 
good  progress. 

A  communication  from  the  Joint  Commission  of  the  Scientific 
Societies  of  Washington  in  regard  to  the  Senate  bill  1552, 
intended  to  restrict,  if  not  prohibit,  vivisection,  was  taken  up 
and  acted  upon. 

The  sentiment  of  the  meeting  was  unanimous  in  the  direction 
of  preventing  affirmative  action  by  Congress  on  the  said  bill ; 
and  the  following  resolutions  were  unanimously  adopted,  after  a 
full  discussion,  in  which  Profs.  Sabin,  Breneman,  Doremus, 
Hale,  and  McMurtrie  participated. 

Resolved,  That  the  New  York  Section  of  the  American  Chem- 
cal  Society  most  earnestly  opposes  the  legislation  proposed  by 
Senate  bill  1552,  entitled  **A  bill  for  the  further  prevention  of 
cruelty  to  animals  in  the  District  of  Columbia." 

Resolved,  That  the  proposed  legislation  is  unnecessary  and 
would  seriously  interfere  with  the  advancement  of  biological 
science  in  that  district ;  that  it  would  be  especially  harmful  in 
its  restriction  of  experiments  relating  to  the  cause,  prevention, 
and  cure  of  the  infectious  diseases  of  man  and  of  the  lower  ani- 
mals ;  that  the  researches  made  in  this  department  of  biological 


(96) 

and  medical  science  have  been  of  immense  benefit  to  the  human 
race  ;  and  that,  in  general,  our  knowledge  of  physiology,  of 
toxicology,  and  of  pathology,  forming  the  basis  of  scientific 
medicine,  has  been  largely  obtained  by  experiments  upon  living 
animals,  and  could  have  been  obtained  in  no  other  way. 

Resolved,  That  physicians  and  others  who  are  engaged  in 
research  work  having  for  its  object  the  extension  of  human 
knowledge  and  the  prevention  and  cure  of  disease  are  the  best 
judges  of  the  character  of  the  experiments*  required  and  of  the 
necessity  of  using  anesthetics,  and  that  in  our  judgment  they 
may  be  trusted  to  conduct  such  experiments  in  a  humane  man- 
ner, and  to  give  anesthetics  when  required  to  prevent  pain.  To 
subject  them  to  penalties  and  to  espionage,  as  is  proposed  by 
the  bill  under  consideration,  would,  we  think,  be  an  unjust  and 
unmerited  reflection  upon  a  class  of  men  who  are  entitled  to  our 
highest  consideration. 

Dx.  C.  A.  Doremus  read  a  **  Note  on  the  Presence  of  Oil  in 
Boiler  Scale.'* 

Mr.  J.  A.  Matthews  described  **  A  New  Method  of  Preparing 
Phthalimide.'' 

The  ehair  announced  this  as  the  last  meeting  of  the  season, 
and  stated  that  the  fall  and  winter  meetings  would  probably  be 
held  in  the  same  rooms. 


iHBued  with  August  Kumber,  1896. 


Proceedings. 


COUNCIL. 

Dr.  Drown  having  resigned  from  the  Committee  **to  unify  the 
methods  of  color  comparison  and  report  on  a  standard  for  meas- 
urement of  color  in  potable  waters/*  the  President  of  the  Amer- 
ican Chemical  Society  has  appointed  in  his  place  Mr.  Allen 
Hazen,  85  Water  St.,  Boston,  Mass.  and  Mr.  Hazen  has  accepted 
the  appointment. 

CHANGES  OF    XdDRKSS. 

Bloomfield,  L.  M.,  Ohio  Experiment  Station,  Wooster,  Ohio. 

Booraem,  J.  V.  V.,  Box  190,  Glen  Cove,  N.  Y. 

Fuller,  Fred.  D.,  Durham,  N.  H. 

Lippincott.  Warren  B.,  North  Western  Iron  Co.,  Mayville, 
Wis. 

Mar,  F.  W.,  138  First  Ave.,  West  Haven,  Conn. 

Myers,  H.  Ely,  Carnegie  Steel  Co.  Ltd.,  Lucy  Furnace,  Pitts- 
burg, Pa. 


MEETINGS  OF  THE  SECTIONS. 

NORTH    CAROLINA   SECTION. 

The  summer  meeting  was  called  to  order  at  3.30  p.  m.,  July 
7th,  in  the  Chemical  Lecture  Room  of  the  University  of  North 
Carolina.  There  were  ten  members  in  attendance.  The  sec- 
retary reported  ten  new  applicants  for  membership,  who  were 
duly  elected.  There  were  other  applicants  who  had  not  yet  con- 
formed to  the  condition  of  becoming  members  of  the  American 
Chemical  Society.  The  membership  roll  has  doubled  in  less 
than  a  half-year. 

Resolutions  were  offered  and  adopted  with  regard  to  the  vivi- 
section bill  and  the  appointment  of  a  director-in-chief  of  the  Sci- 
entific Bureaus  of  the  Department  of  Agriculture.  The  follow- 
ing papers  were  then  read  : 

^'Crystallized  Aluminum,"  by  F.  P.  Venable;  "Detection 
and  Purification  of  Saccharin,"  by  B.  W.  Kilgore  :  "Reduction 


98 

of  Sulphuric  Acid/'  by  Chas.  Baskerville  ;  * 'Comparison  of  Di- 
gestibility of  Raw  and  Steamed  Cotton  Seed,"  by  J.  A.  Bizzell; 
**An  Attempt  to  Form  some  Organic  Compounds  of  Zirconium," 
by  ThomasClarke ;  *  'Determination  of  Sulphur  in  the  Presence  of 
Iron/'  by  W.  A.  Withers  and  R.  G.  Mewbome ;  "Action,  of 
Phosphorus  Trichloride  on  an  Ethereal  Solution  of  Hydrogen 
Dioxide,"  by  W.  A.  Withers  and  G.  S.  Fraps ;  "Some  Diffi- 
culties  in  the  Way  of  the  Periodic  Law,"  by  F.  P.  Venable. 
The  section  then  adjourned. 


iMued  with  8e|>teinber  Number,  1896. 

Proceedings. 

CHANGBS  OP   ADDRESS. 

Beeson»  J.  L.,  Bethel  College,  Russellville,  Ky. 
Best,  Dr.  Otto,  care  of  Fritzche  Bros.,  Garfield,  N.  J. 
Haines,  Reuben,  Haines  and  Chew  streets,  Germantown,  Pa. 
Emmens,  Stephen  H.,  179  Washington  Building,  N.  Y.  City. 
Kelley,  J.  H.,  Bentonville,  Ark. 
Moale,  Philip  R.,  82  Chestnut  street,  Asheville,  N.  C. 
Mumper,  W.  N.,  823  W.  State  street.  Trenton,  N.  J. 
Nipholsi  Wm.  H.,  32  Liberty  street,  N.  Y.  City. 
Potter,  Wm.  R.,  100  Broad  street,  Providence,  R.  I. 

DBCBASED. 

Bower,  Henry,  Gray's  Ferry  Road,  Philadelphia,  Pa. 
MEETINGS  OF  THE  SECTIONS. 

RHODE   ISLAND  SECTION. 

The  regular  monthly  meeting  of  the  Rhode  Island  Section 
was  held  at  Providence,  May  21,  1896,  Mr.  Charles  S.  Bush  in 
the  chair. 

A  paper  was  presented  by  Mr.  William  R.  Potter,  upon 
"Fallacies  in  Urine  Analysis  due  to  the  Presence  of  Salicylic 
Acid  and  its  Compounds." 

After  a  brief  introduction  of  the  value  of  urine  examination 
as  an  aid  to  the  physician  in  his  diagnosis,  the  reader  described 
in  detail  the  influence  salicylic  acid  had  upon  the  albumin  test 
and  the  glucose  test  as  commonly  practiced.  The  chief  source 
of  error  was  pointed  out  to  lie  in  the  sparing  solubility  of  sali- 
C3^1ic  acid  in  an  aqueous  solution,  and  the  ease  with  which  its 
compounds  were  decomposed  by  other  acids. 

The  annual  meeting  of  the  Rhode  Island  Section  was  held  at 
Pawtucket,  R.  I.,  on  Thursday  afternoon,  June  11,  1896. 

Members  met  upon  the  invitation  of  the  presiding  officer,  Mr. 
C.  A.  Catlin,  at  the  Country  Club.  After  dinner  the  annual 
election  of  officers  took  place.  The  following  were  elected  for 
the  ensuing  year: 

Presiding  Officer,  Mr.  Edward   D.    Pearce  :    Secretary  and 


(lOo) 

Treasurer,  Mr.  Walter  M.  Saunders  ;  Member  of  the  Executive 
Committee,  Mr.  George  F.  Andrews. 

The  retiring  chairman  then  presented  his  annual  address, 
taking  for  his  topic,  the  subject  of  chemically-applied  mechanics, 
introducing  it  by  brief  historical  reference  to  the  progress  ^f 
chemistry,  more  particularly  to  the  development  of  apparatus 
and  improvements  along  the  line  of  chemical  manipulation.  It 
appears  that  after  all  the  same  forms  of  retort  and  crucible  that 
did  service  in  the  time  of  Zosimus,  are  still  the  stereotyped  forms 
of  the  chemical  supply  house. 

Great  progress  in  recent  years  in  arts  dependent  upon  chemistry, 
has,  however  forced  the  development  of  apparatus,  and  the 
application  of  mechanical  expedients,  until  there  has  grown  up 
what  may  in  all  truth  be  called  the  science  of  chemically-applied 
mechanics,  practically  unrecognized  as  yet  by  the  training  schools 
of  the  profession,  though  covered  in  a  general  way  by  courses 
offered  in  chemical  engineering.  These  courses  do  not  meet 
the  case.  The  real  demand  is,  that  chemical  students  shall  have 
opportunity  in  training,  to  pursue  the  study  of  practical  mechanics 
as  applied  to  chemical  manipulation,  simply  a  new  study  added 
to  the  old  curriculum.  Illustrating  something  of  the  scope  of  this 
chemically-applied  mechanics,  may  be  cited,  hydraulics,  for  in- 
stance, as  applied  to  the  handling  of  liquids,  the  speaker  show- 
ing from  his  own  practical  experience  how  the  attendant  phenom- 
ena of  the  various  filtering  methods,  may,  by  a  general  classi- 
fication, be  brought  to  useful  presentation  of  the  whole  subject. 
Further  may  be  cited  pneumatics  as  applied  to  the  manipulation 
of* air  draughts,  the  charging  of  liquids  with  gases,  etc.,  strength 
and  adaptability  of  materials  to  the  construction  of  apparatus, 
their  acid  or  alkaline  resisting  qualities,  with  effects  of  saline 
solutions  upon  them  ;  heat  in  its  varied  application  ;  light  in  the 
practical  application  of  its  actinic  properties  ;  and  the  milling  of 
materials.  Again  illustrating  particularly,  the  speaker  showed 
how  the  attendant  phenomena  of  the  various  milling  processes 
may  be  brought  under  general  statements  for  better  considera- 
tion. 

Generally  the  scope  of  chemically-applied  mechanics  may  be 
stated  as  the  application  of  mechanical  principles  to  chemical 
manipulations.  The  whole  subject. should  be  presented  to  the 
chemical  student  in  a  general  way,  setting  forth  at  least  the 
fundamentals  along  these  and  related  lines,  expanding  into 
particular  detail  in  the  more  important,  thus  broadly  laying  the 
foundation  for  the  development  and  exercise  of  the  inventive 
faculty  in  applying  mechanical  means  to  special  chemical  re- 
quirements, whether  it  be  in  the  factory  or  the  research  labora- 
tory. 


Issued  with  October  Number.  1896. 


Proceedings. 


THIRTEENTH  GENERAL  MEETING  OF  THE 
AMERICAN  CHEMICAL  SOCIETY. 

Buffalo,  N.  Y.,  August  21,  1896. 

President  Dr.  Charles  B.  Dudley  called  the  meeting  to  order. 
Dr.  Roswell  Park,  President  of  the  Buffalo  Society  of  Natural 
Sciences,  welcomed  the  visiting  chemists  as  follows  : 

Mr.  President  and  Gentlemen  of  the  Chemical  Society  :  I  am  very  glad 
to  join  wtth  my  friends  in  the  City  of  Buffalo  in  welcoming  you  h«re.  My 
idea  of  what  should  be  said  on  such  an  occasion,  is  that  it  should  be  char- 
acterized by  genuineness  rather  than  eloquence,  by  brevity  rather  than 
length.  I  am  sure  we  are  very  glad  to  see  you  here.  I  know  that  this 
is  the  first  time  the  Chemical  Society  ever  met  in  Buffalo,  aud  we  hope 
that  you  will  like  us  so  well  that  you  w^ill  come  again.  I  have  on  several 
occasions  in  time  past  welcomed  associations  of  citizens  here.  I  tell . 
them  we  have  pleasant  weather  here  always.  I  was  sure  in  my  own  mind 
that  you  would  have  it.  It  is  a  promise  we  can  safely  make  anytime  in 
the  summer.  Buffalo  seems  to  be  generally  regarded  now  as  the  ideal 
convention  city.  We  have  had  a  convention  here  of  some  kind  almost 
every  week,  and  this  will  continue  through  September.  There  are  meet- 
ings here  almost  all  the  time.  If  you  study  our  statistics  you  will  find 
us  the  healthiest  city  in  the  country.  If  you  will  travel  around  our 
streets  you  will  discover  we  have  the  mos^  attractive  residential  city  in 
the  country.  Aud  in  every  way,  both  from  our  treatment  of  you  and 
what  you  see  for  yourselves,  we  hope  you  will  feel  thoroughly  welcome, 
and  thoroughly  at  home. 

It  is  always  proper,  I  think,  on  such  occasions,  to  blow  our  own  horn 
a  little  bit.  I  have  found  so  little  appreciation  of  Buffalo  abroad,  of  what 
Buffalo  is,  that  I  am  going  to  say  a  little  to  you  about  Buffalo.  It  is  the 
sixth  commercial  city  in  the  world. '  That  is  not  generally  appreciated. 
A  friend  of  mine  went  to  Boston  and  while  there  was  talking  to  a  friend 
of  his  in  that  city  about  Buffalo.  The  Boston  man  said,  "  Buffalo?  Is 
that  on  Lake  Erie  or  Lake  Ontario?"  At  the  same  time,  we  have  a 
much  greater  tonnage  coming  into  our  harbor  in  one  year  than  comes 
into  Boston  harbor.  That  he  overlooked.  About  5,000,000  tons  of  ton- 
nage enter  our  harbor,  aud  about  the  same  leave  our  harbor  every  year. 
There  is  only  one  other  city  can  say  this,  and  that  is  Liverpool.  That  is 
not  generally  appreciated.  Thirteen  years  ago,  when  I  moved  here,  the 
city  had  about  125,000  inhabitants;   now  it  has  a  third  of  a  million,  so 


(I02) 

you  can  get  an  idea  of  how  it  is  growing.  We  who  live  here  and  aeeithe 
trend  of  affairs,  look  forward  to  a  time  when  there  will  be  but  one  city  to 
Niagara  Palls.  We  are  coming  nearer  and  nearer  to  that  all  the  time. 
It  is  .not  far  off,  I  assure  you. 

Now,  with  all  we  have  and  all  we  can  do  for  yon,  gentlemen,  you  cer- 
tainly are  cordially  welcome.  You  will  hear  more  of  this,  as  I  expect 
you  will  attend  the  meetings  next  week,  and  perhaps  be  more  formally 
welcomed  by  the  city  officials  on  other  occasions ;  but  our  homes  are 
opened  to  yon,  and  everything  we  can  do  in  any  way  for  you,  is  cordially 
placed  before  you. 

To  refer  just  a  moment  to  the  scientific  aspect  of  this  gathering,  I  have 
never  had  a  chance  to  talk  to  professional  chemists  before,  and  there  is 
one  appeal  I  want  to  make  to  you  as  coming  from  our  profession  to  yours. 
Of  course  we  are  working  in  large  measure  on  common  ground ;  espe- 
cially when  it  comes  to  physiological  chemistry,  and  in  the  chemistry  of 
the  fluids,  etc.,  of  the  body,  we  are  on  absolutely  common  ground;  but 
there  is  very  much  we  have  to  rely  upon  you  for,  in  order  to  help  our- 
selves forward  ;  and,  as  one  who  is  eagerly  anxious  for  the  discovery  of  a 
particular  substance,  an  ideal  in  our  business,  which  you  only  can  prob- 
ably furnish,  I  will  make  this  scientific  appeal  to  you.  We  have  been 
working  for  years  to  find  a  substance  which  shall  have  a  germicidal  prop- 
erty so  far  as  deleterious  agents  are  concerned,  and  yet  which  will  not  be 
toxic  with  the  tissues  of  the  human  body  ;  a  chemical  substance  whose 
relative  and  absolute  toxicity  are  far  enough  apart  to  make  it  a  safe  sub- 
stance to  use.  When  we  have  that,  we  hope  to  saturate  the  human  body 
with  the  substance  which  will  be  at  the  same  time  not  toxic  with  the  tis- 
sues of  the  larger  organs.  I  do  not  know  whether  that  time  will  ever 
Come.  It  seems  to  me  an  ideal  substance.  I  do  not  know  how  you,  who 
are  so  interested  in  the  affairs  of  the  world  at  large,  as  well  as  humani- 
tarians, can  make  a  better  discovery  than  one  along  the  lines  I  have  sug- 
gested. It  is  not  for  commercial  purposes,  but  purely  for  the  benefit  of 
humanity.  You  will  pardon  this  little  appeal  to  your  chemical  abilities ; 
it  is  the  only  chance  I  have  ever  had  to  make  it. 

Permit  me  only  to  reiterate  what  I  have  said  to  you  about  the  cordial- 
ity of  our  welcome,  our  earnest  endeavor  to  extend  to  you  our  hospitality, 
our  earnest  hope  that  your  first  meeting  will  not  only  be  so  successful 
that  you  will  look  back  to  it  hereafter,  but  will  be  so  pleasant  to  you 
that  you  will  want  to  come  here  again  quite  often.     (Applause.) 

The  President. 

Dr.  Park,  The  Committee  of  Arrangements,  and  Fellow  Members  of 
the  American  Chemical  Society :  I  am  sure  I  voiee  the  sentiment  of  those 
who  are  present  when  I  say  that  we  appreciate  this  kind  welcome,  and 
we  thank  you  for  it.  I  doubt  not  there  are  a  good  many  present  who  can 
well  remember  when  chemical  analysis,  except  for  purely  scientific  pur- 


(I03) 

poses,  was  a  rarity.  In  my  early  student  days  the  chemical  analyses 
that  were  made,  except  as  I  say,  for  such  purely  scientific  purposes,  were 
largely  made  by  the  professors  in  colleges.  They  were  slow.  They  were 
very  expensive  and  any  business  that  wanted  a  chemical  analysis,  studied 
quite  a  while  before  employing  a  chemist  to  make  it.  That  state  of  affairs 
is  now  changed.  With  the  growth  of  the  technical  school  there  has  come 
forward  each  year  a  large  crop  of  young,  enthusiastic  chemists,  and  with 
this  supply,  if  I  may  use  the  word,  has  come  likewise  the  necessity  for 
their  existence  and  the  work  for  them.  I  am  not  saying  anything  more 
than  is  known  to  you  all  when  I  say  that  a  very  large  number  of  com- 
mercial ventures  and  enterprises  to-day  cannot  live  without  their  chem- 
ist. The  steel  works  would  not  be  able  to  maintain  themselves  a  month 
without  a  chemist.  The  sugar  industry  needs  the  chemist,  the  brewing 
industry,  the  textile  industry,  and,  indeed,  I  might  go  on  and  enumerate 
occupation  after  occupation  which  is  based  largely  upon  the  chemist's 
work.  The  railroads,  as  you  know,  are  using  chemists,  and  the  cities 
begin  to  have  their  chemists  to  protect  people  against  fraud  and  adulter- 
ation in  products  which  are  for  sale.  As  we  all  know,  agriculture  is  more 
and  more  every  day  becoming  based  on  chemistry,  and  our  government 
itself  supports  one  of  the  best  chemical  establishments  in  the  world. 
Now,  this  increase  in  chemists,  this  increase  in  their  work,  this  demand 
for  them,  has  led  to  another  necessity;  namely,  that  the  chemists  should 
occasionally  look  each  other  in  the  face,  that  they  should  talk  things 
over  with  each  other,  that  they  should  profit  by  each  other's  work,  and 
that  brings  us  to  state  what  the  organization  of  the  Chemical  Society  is, 
an  organization  with  something  like  i,ooo  members,  an  organization 
which  supports  a  Journal  that  is  published  every  month,  and  with  some 
eight  or  nine  local  sections  located  in  different  parts  of  the  country. 
This  organization  must,  as  we  all  know,  have  a  place  for  meetings.  We 
are  already  having  two  meetings  every  year  and  this  year  we  come  to 
Buffalo,  and  I  may  say,  that  this  city  is  the  Mecca  to  which  all  scientific 
men  are  travelling  this  year — this  city  which  may  almost  be  called  the 
mother  of  scientific  organizations.  I  believe  that  the  reorganization  of 
the  American  Association  for  the  Advancement  of  Science,  one  of  the 
oldest  scientific  organizations  in  the  country  as  we  all  know,  took  {>lace 
here  in  Buffalo  in  1866,  after  the  war.  It  had  previously  had  existence 
but  the  war  injured  it,  or  caused  a  temporary  cessation  and  the  reorgan- 
ization took  place  in  Buffalo.  Thus  much  for  our  reason  for  existence 
and  thus  much  for  our  coming  here.  We  appreciate  very  greatly  your 
kind  and  gracious  welcome.  We  look  forward  to  an  interesting  and  profit- 
able time.    We  thank  you.    (Applause.) 

The  following  papers  were  then  read  and  discussed  : 

**  Mercuric  Chlorothiocyanate,'*  by  Charles  H.  Herty  and  J. 
G.  Smith.  Read  by  Dr.  Herty.  Discussed  by  Messrs.  Hart, 
Prescott  and  Frankforter. 


(i04) 

*'  The  Reduction  of  Concentrated  Sulphuric  Acid  by  Copper," 
by  Charles  Baskerville.     Read  by  the  author. 

**  Notes  on  the  Preparation  of  Glucinum,**  by  Edward  Hart. 
(An  informal  description  of  the  progress  of  work  on  the  prepa- 
ration of  glucinum  and  its  alloys.  A  glucina  crucible  was  ex- 
hibited and  also  some  nearly  pure  glucina  prepared  by  the 
method  already  described  in  the  Journal,  17,  604.  This  glu- 
cina apparently  contains  the  same  unknown  substance  already 
detected  by  Kruss,  and  as  200  pounds  of  beryl  are  being  opera- 
ted on  it  is  hoped  that  enough  may  be  obtained  for  its  identifi- 
cation.) The  paper  was  discussed  by  Messrs  C.  B.  Dudley  and 
Hart. 

"The  Inspection  and  Sanitary  Analysis  of  Ice,*'  by  C.  L. 
Kennicott.  Read  by  the  author.  Discussed  by  Messrs.  W.  P. 
Mason,  Cochran,  McKenna,  \V.  A.  Noyes,  Breneman,  Miller, 
Phillips,  Robbins,  Frankforter  and  C.  B.  Dudley. 

•*  A  New  Form  of  Potash  Bulb,"  by  M.  Gomberg.  Read  by 
Dr.  Prescott.     Discussed  by  Mr.  Phillips. 

'*  Morphine  in  Puirefactive  Tissue,"  by  H.  T.  Smith.  Read 
by  Dr.  Prescott.     Discussed  by  Mr.  Miller. 

•'  Some  New  Compounds  of  Thallium,"  by  L.  M.  Dennis  and 
Martha  Doan,  with  crystallographic  notes  by  A.  C.  Gill.  Read 
by  Dr.  Dennis.  Discussed  by  Messrs.  Prescott,  Hart,  Frank- 
forter and  Mason. 

The  President :  It  has  reached  pretty  nearly  the  hour  of  ad- 
journment and  I  would  like  to  make  an  announcement  or  two 
as  to  the  work  of  the  Society  during  the  interim.  Early  in  the 
spring  a  letter  was  received  by  the  Society  stating  that  Canniz- 
zaro's  seventieth  birthday  was  to  occur  on  the  nth  of  July, 
and  it  was  proposed  to  make  a  testimonial  to  him  in  some  way. 
This  letter  asked  the  cooperation  of  the  American  Chemical 
Society.  After  talking  tKe  matter  over  it  was  decided  since 
Cannizzaro  was  already  an  honorary  member  that  we  should  send 
him  a  testimonial  engrossed  oq  parchment.  This  was  duly  pre- 
pared and  was  sent  in  time  to  reach  Rome  some  two  or  three  weeks 
before  his  birthday.  However,  since  that  time  we  have  received 
a  second  letter  stating  that  owing  to  the  fact  that  most  of  the  pro- 
fessional people  who  were  interested  in  Cannizzaro  were  out  of 
town  during  the  very  warm  season,  it  has  been  decided  to  post- 
pone the  public  recognition  of  the  occasion  until  later  in  the  fall, 
I  think  some  time  in  October.     So  we  have  not  as  yet  heard 


(i05) 

from  the  other  side  as  to  what  has  been  done  with  the  testimonial. 
I  would  say  likewise  that  in  this  letter  there  was  a  statement 
to  the  effect  that  the  form,  which  recognition  was  taking  on 
the  other  side  was  that  of  accumulating  a  fund  to  be  used  for 
some  scientific  purpose. 

I  would  also  state  that  at  the  last  meeting  of  the  Society  in 
Cleveland  a  committee  of  three  was  appointed  to  take  up  the 
question  of  coal  analysis.  That  committee  consisted  of  Mr. 
Hillebrand  of  the  Coast  Survey,  Chairman,  Prof.  W.  A.  Noyes 
and  the  President  of  the  Society.  The  committee  has  been  able  to 
do  very  little  thus  far  except  to  get  ready.  They  are  not  prepared 
to  make  any  formal  report  at  this  meeting,  partly,  I  think  due  to 
my  own  fault  in  the  matter.  Our  progress  has  not  been  suffi- 
cient to  make  a  formal  report.  This  is  simply  to  let  you  know 
that  the  subject  has  not  been  dropped. 

About  the  beginning  of  the  summer  a  paper  was  read  in  the 
New  York  Section  by  Prof.  Leeds  on  the  color  of  water,  and  at 
his  suggestion  a  Committee  was  appointed  to  report  to  the 
Society,  a  standard  to  be  used  for  determining  the  color  of  water 
and  a  method.  That  committee  consists  of  Prof.  Leeds,  Chair- 
man, Prof.  Mason  and  Mr.  Allen  Hazen,  formerly  connected 
with  the  State  Board  of  Health  of  Massachusetts,  who  has  done 
a  good  deal  of  work  on  water  analysis.  We  have  some  regular 
or  standing  committees  ;  I  have  not  been  able  to  get  in  communi- 
cation with  the  Chairmen  of  all  of  them  as  yet,  and  we  will  try 
to-morrow  to  see  whether  we  can  get  information  from  them  on 
the  state  of  the  subjects  committed  to  them.  After  some  an- 
nouncements by  the  general  secretary  and  the  local  committee 
of  arrangements,  the  session  adjourned. 

Saturday  Morning,  August  22,  1896. 

The  President  called  the  Convention  to  order  at  9:40  o'clock. 

The  President :  As  we  have  considerable  to  get  through  with 
to-day  I  think  we  had  better  start  as  soon  as  possible  and  first 
of  all  I  will  ask  the  Society  to  give  two  or  three  minutes  to  Dr. 
de  Schweinitz  who  wants  to  present  the  matter  of  the  Pasteur 
monument. 

Dr.  de  Schweinitz:   Mr.   President    and    gentlemen  of  the 


(io6) 

Society :  I  only  desire  to  detain  you  for  a  moment  to  ask  for 
subscriptions  towards  the  erection  of  an  international  monument 
in  Paris  to  Pasteur.  The  French  Government  has  organized 
this  movement  and  requested  the  cooperation  of  all  scientists, 
or  I  should  say,  rather,  of  all  members  of  the  different  branches 
of  science  in  the  United  States.  As  Pasteur  was  a  chemist,  the 
chemists  of  the  United  States  should  be  the  first  to  respond  to 
this  request.  Printed  blanks  of  a  general  announcement,  giving 
the  names  of  the  members  of  the  French  committee,  and  also  of 
the  organizing  committee  of  Washington,  which  has  been 
started »  will  be  distributed,  and  in  addition  to  this  subscription 
blanks,  as  you  see  here,  upon  which  you  are  requested  to  place 
the  amount,  however  small,  it  does  not  make  any  difference, 
and  however  large,  the  larger  the  better  and  the  more  will  the 
contribution  be  appreciated,  to  be  forwarded  to  Washington. 
These  blanks  with  the  names  and  the  amounts  will  be  preserved 
and  will  be  deposited  in  Paris  in  the  archives  in  connection  with 
this  Pasteur  monument.  I  will  distribute  these  blanks  and  be 
greatly  obliged  to  the  members  of  this  Society  if  they  will  join 
in  the  contribution  at  as  early  a  date  as  possible  and  to  the 
largest  amount  that  they  feel  able  to  give. 

The  President :  I  am  sure  the  appeal  is  one  that  we  are  all 
interested  in,  and  if  chemists  can  see  their  way  to  subscribe  for 
this  purpose,  we  shall  be  very  glad.  We  all  feel  willing  un- 
doubtedly, but  possibly  not  all  of  us  are  able. 

Dr,  de  Schweinitz :  Mr.  President,  I  might  add  that  the 
subscriptions  so  far  received  have  varied  in  amount  from  twenty- 
five  cents  up,  so  that  no  one  need  have  any  hesitancy  on  that 
subject. 

The  President :  I  presume  there  is  no  one  can  not  subscribe 
at  least  the  minimum  amount. 

I  wish  to  say  that  Dr.  Levi  has  brought  up  a  few  samples  of 
aniline  colors  made  at  the  aniline  works,  which  there  was  no  op- 
portunity to  distribute  yesterday,  and  anyone  here  can  avail 
himself  of  the  samples  if  he  so  desires. 

The  following  papers  were  then  read  and  discussed  : 

**  Contribution  to  the  Knowledge  of  Rutheno  Cyanides,"  by 
James  Lewis  Howe.     Read  by  the  author. 


(xo7) 

•'  Analytical  Methods  Involving  the  Use  of  Hydrogen  Diox- 
ide," by  B.  B.  Ross.     Read  by  the  author. 

Prof,  Hart :  Mr.  Chairman,  while  we  are  waiting  for  Prof. 
Ross  to  place  these  figures  on  the  blackboard,  there  is  a  matter 
that  has  been  called  to  my  attention  which  I  would  like  to  pre- 
sent to  you ;  it  will  only  take  a  half  minute  ;  this  is  connected 
with  the  subject  of  advertising  for  the  Journal.  By  resolution 
of  the  Board  of  Directors  I  was  appointed  a  committee  of  one  to 
secure  advertisements  for  the  Journal.  This  is  an  important 
source  of  revenue,  and  we  have  derived  considerable  money  to 
be  applied  to  the  publication  of  the  Journal  in  this  way.  It  is 
believed  that  with  some  additional  assistance  this  source  of  rev- 
enue can  be  still  further  increased.  We  have  to  depend  for  this 
assistance  on  voluntary  aid,  and  I  wish  to  acknowledge  the  great 
assistance  I  have  already  received  from  Dr.  McMurtrie  in  this 
direction.  The  Society  is  indebted  to  him  more  than  is  perhaps 
generally  known.  It  has  been  suggested  to  me  that  a  number  of 
members  of  the  Society  would  be  willing  to  assist  in  the  matter 
of  procuring  advertisements,  and  that  it  would  be  well  to  increase 
the  committee  to  ten  members.  These  members  would  then  feel 
that  it  was  their  duty  to  assist  in  securing  the  advertisements, 
and  it  is  believed  that  this  will  result  in  securing  considerable 
additional  patronage.  I  therefore  move  that  the  President  have 
power  to  increase  the  committee  to  not  more  than  ten  members. 

Dr.  Hale:  I  second  the  motion. 

Dr,  McMurtrie :  I  think  it  might  be  well  further  to  give  the 
committee  power  to  extend  its  membership  in  case  that  appears 
desirable.     I  would  move  to  amend  in  that  manner. 

Prof,  Hart :  I  accept  the  amendment. 

President  Dudley  put  the  motion  as  amended,  and  it  was  duly 
carried. 

The  following  papers  were  then  read  : 

*'The  Estimation  of  Thoria  ;  Chemical  Analysis  of  Monazite 
Sand,**  by  Charles  Glaser.     Read  by  Dr.  Hart. 

''  The  Estimation  of  Thorium  and  its  Separation  from  Other 
Rare  Earths,"  by  L.  M.  Dennis.  Read  by  the  author.  These 
two  papers  were  discussed  by  F.  W.  Clarke  and  L.  M.  Dennis. 


(io8) 

**  A  Complete  Analysis  of  Phytolacca  Decandra,"  by  G.  B. 
Frankforter  and  Francis  Ramaley    Read  by  Mr.  Frankforter. 

**  The  Cr>'stallized  Salts  of  Phytolacca  Decandra/'  by  G.  B. 
Frankforter  and  Francis  Ramaley.  Read  by  Mr.  Frankforter. 
Discussed  by  A.  B.  Prescott. 

**  The  By- Products  formed  in  the  Conversion  of  Narcoline  into 
Narceine/'  by  G.  B.  Frankforter.     Read  by  the  author. 

**  The  Composition  of  American  Kaolins,"  by  C.  F.  Mabery 
and  Otis  T.  Klooz.  Read  by  Dr.  Hart.  Discussed  by  Messrs. 
Dudley,  Baskerville,  McMurtrie,  Noyes,  Prochazka,  Breneman 
and  Patrick. 

The  following  papers  were  read  by  title  : 

**  Composition  of  Certain  Mineral  Waters  in  Northwestern 
Pennsylvania,*'  by  A.  E.  Robinson  and  Charles  F.  Mabery. 

**  Zirconium  Oxalates,"  by  F.  P.  Venable  and  Charles  Bas- 
kerville. 

**  Aluminum  Analysis,"  by  James  Otis  Handy. 

*  *  An  Analytical  Investigation  of  the  Hydrolysis  of  Starch  by 
Acids,"  by  George  W.  Rolfe  and  George  Defren. 

* '  The  ££fect  of  an  Excess  of  Reagent  in  the  Precipitation  of 
Barium  Sulphate,"  by  C.  W.  Foulk.  Discussion  by  T.  M. 
Gladding. 

**  Determination  of  Reducing  Sugars  in  Terms  of  Cupric  Ox- 
ide," by  George  Defren. 

**  Acidity  of  Milk  Increased  by  Boracic  Acid,"  by  E.  H. 
Farrington. 

*'  The  Actual  Accuracy  of  Chemical  Analysis,"  by  Frederic 
P.  Dewey. 

'  *  Some  Extensions  of  the  Plaster  of  Paris  Method  in  Blowpipe 
Analysis,"  by  W.  W.  Andrews. 

**  Device  for  Rapidly  Measuring  and  Discharging  a  Definite 
Amount  of  Liquid,"  by  Edward  L.  Smith. 

**  Table  of  Factors,"  by  E.  H.  Miller. 

**  A  Modified  Form  of  the  EbuUioscope,"  by  H.  W.  Wiley. 

'*  The  Signification  of  Soil  Analysis,"  by  H.  W.  Wiley. 

**  Notes  on  the  Determination  of  Phosphorus  in  Steel  and  Cast 
Iron,"  by  George  Auchy. 

•*  The  Development  of  Smokeless  Powder,"  by  C.  E.  Munroe. 

The  President:  I  would  like  to  announce  that  the  winter  meet- 
ing will  be  held  in  Troy,  it  having  been  decided  by  the  Council, 


(i09) 

on  the  invitation  of  our  membership  in  Troy  to  hold  the  meeting 
at  that  place.  We  are  hoping  to  make  that  meeting  one  of  the 
best  the  Society  has  ever  had  and  I  would  like  to  ask  Prof. 
Mason  to  give  us  a  word  or  two  in  regard  to  our  meeting  next 
winter  at  Troy. 

Prof,  Mason  :  Mr.  President  and  fellow  members,  it  has  been 
very  gratifying  to  me  to  learn  that  you  have  decided  to  come  to 
Troy.  We  are  not  a  large  city,  but  we  will  do  our  very  best  to 
make  your  stay  agreeable.  There  are  some  things  there  that 
are  worth  seeing.  We  will  be  able  to  show  you  the  largest  gun 
plant  in  the  world,  much  larger  than  Krupp's.  Of  course  when 
you  speak  about  Krupp's  plant  it  means  his  whole  concern, 
the  gun  plant  and  that  for  other  varieties  of  iron  and  steel 
manufacture  as  well,  but  the  gun  portion  of  his  plant  would  go 
into  a  small  part  of  the  United  States  gun  plant  which  you  will 
see  at  Troy.  As  you  know,  all  the  artillery  now  used  by  the 
army  is  made  there,  practically  ;  I  believe  there  are  a  few  un- 
finished contracts  out,  but  I  am  not  positive  about  that.  You 
will  be  able  to  see  electric  cranes  that  I  think  are  larger  than 
you  have  ever  seen  elsewhere.  You  will  be  able  to  see  guns  in 
all  stages  of  manufacture.  I  hope  you  will  be  able  to  see  an  old- 
fashioned  smooth  bore  of  fifteen  or  twenty  inches  caliber  lying 
along  side  of  a  modern  twelve  or  thirteen.  It  looks  like  a  soda 
water  bottle.  We  have  some  other  institutions  there  that  we  are 
proud  of,  for  instance  the  new  basic  steel  plant,  which  will  be  in 
full  operation  by  the  time  you  get  there,  the  Burden  Iron  Works 
where  they  make  Burden's  best  iron,  which  you  have  often  heard 
of.  The  shirt  foundries  and  collar  smelting  works  with  their 
attendants  are  well  worth  seeing.  (Laughter.)  More  particu- 
larly the  E  &  W  Collar.  You  have  probably  heard  of  them. 
They  have  sent  you  a  special  invitation.  We  have  N  + 1 
breweries  in  Troy.  We  can  take  care  of  the  N  and  we  have  as- 
signed the  I  to  our  President.     (Laughter.) 

It  will  give  us  great  pleasure  to  see  you  and  I  am  heartily 
glad  that  you  are  coming  and  the  Mayor  of  the  city  sends  his 
especial  invitation. 

The  President  :    I  am  sure  we  will  all  look  forward  to  this 


(no) 

meeting  with  a  great  deal  of  interest,  and  as  I  said  at  the  very 
outset  we  hope  to  make  this  the  most  important  meeting  the 
Sochety  has  ever  had.  At  this  point  and  a  propos  here  I  want 
to  give  3*ou  a  word  of  exhortation  in  regard  to  the  condition  of 
the  society.  As  everybody  knows  the  most  important  thing  in 
the  Society  is  the  Journal.  The  Journal  is  impossible  without 
money.  Our  rates  are  low,  our  annual  dues  being  only  $5.00. 
The  Society  of  Civil  Engineers  in  this  country  charges  $15.00, 
the  Mechanical  Engineers  $15.00,  the  Mining  Engineers  $10.00, 
the  Mining  Institute  of  Great  Britain  two  guineas;  the  German 
Mining  and  Steel  Institute  charges  $10.00.  We  are  trying  to 
run  a  Society  on  $5.00  and  the  management  does  not  think  at 
present  that  it  would  be  advisable  to  raise  that  figure.  But  we 
want  more  money.  How  can  we  get  more  money  ?  Obviously 
by  getting  more  members.  If  every  member  of  the  Society  would 
get  one,  think  what  would  happen  the  doubling  of  our  member- 
ship. It  is  believed  there  are  something  like  5,000  chemists  in 
the  United  States  who  are  eligible,  either  as  full  members  or 
associates.  We  have  practically  now  about  1,000.  Your 
management  has  in  mind  plans  in  regard  to  the  advancement  of 
the  Journal  to  make  it  still  more  representative,  having  it  cover 
wider  fields,  but  for  this  purpose  money  is  necessary,  and  money 
with  our  present  ideas  in  regard  to  our  present  society  can  only 
come  to  us,  at  least  as  far  as  we  can  see,  through  increase  in 
membership.  Will  not  every  member  of  the  societ}-  do  something 
in  the  next  four  or  five  months  to  increase  our  membership. 
We  certainly  are  well  established  on  a  good  foundation.  It  is 
an  honor  to  be  a  member  of  our  society.  We  give  a  full  requital 
for  everything  we  get  from  our  membership,  and  certainly  the 
time  is  fast  approaching  when  any  American  chemist  who  ex- 
pects to  keep  up  with  his  profession  cannot  afford  to  be  outside 
of  the  Society.  Let  every  member  bring  one  member  with  him 
and  more  if  possible,  at  the  Troy  meeting  or  bring  them  in  be- 
tween now  and  then.  I  will  call  upon  the  Secretary  for  a  few 
announcements  connected  with  the  Society. 

The  Secretary  :  Perhaps  I  might  say,  Mr.  President,  that  Dr. 
Mason  with  becoming  modesty  has  failed  to  remind  you  that  the 
oldest  institution,  if  I  am  not  mistaken,  for  the  education  of 


(Ill) 

civil  engineers,  is  in  Troy,  and  as  a  representative  of  the  institu- 
tion, he  has  some  modesty  in  speaking  of  it.  Allow  me  to  call 
attention  to  one  point  in  reference  to  increase  in  member- 
ship ;  there  is  provided  in  the  constitution  a  class  of  mem- 
bers who  are  not  necessarily  chemists,  but  who  are  interested 
in  chemistry,  the  associates;  and  it  would  seem  as  though  there 
might  be  a  large  amount  of  recruiting  from  this  source.  Tliere 
are  very  many  people  who  do  not  feel  themselves  distinctively 
chemists  and  yet  they  are  interested  either  through  their  busi- 
ness or  by  their  inclination  in  the  development  of  chemistry ; 
and  it  would  seem  possible  to  have  as  large  a  membership  of  as- 
sociates as  of  active  members.  We  can  do  a  good  work  in  that 
way,  and  the  $5.00  of  an  associate  is  worth  just  as  much  as  the 
$5.00  of  an  active  member. 

In  regard  to  the  membership  of  the  society,  I  would  say  that 
last  spring,  somewhere  about  March,  I  think,  for  the  first  time 
in  the  history  of  the  Society,  we  struck  a  membership  of  a  full 
round  1000  in  number.     (Applause.) 

The  President :  I  am  sure  those  of  us  who  have  the  pleasure 
of  being  at  this  meeting  can  not  fail  to  have  recognized  that  there 
Jias  been  at  the  helm  some  guiding  hands,  and  I  am  going  to  say 
for  your  information  that  those  guiding  hands  are  not  the  officers 
of  the  Society  but  the  local  committee.  I  feel  that  it 
would  be  improper  for  us  to  close  the  meeting  without  some  rec- 
ognition of  the  kindness  we  have  received  at  the  hands  of  our 
members  here  and  also  those  who  have  contributed  to  our  hap- 
piness during  this  visit.  I  will  call  upon  Prof.  Mason  to  propose 
due  recognition. 

Prof,  Mason  :  Mr.  President  and  Gentlemen  ;  it  seems  to  me 
entirely  fitting  that  we  should  pass  a  vote  of  thanks  to  those  who 
have  so  kindly  looked  after  our  pleasure  and  interest,  and  I  will 
therefore  move  you  that  the  thanks  of  this  society  are  due  to  the 
local  committee  of  arrangements,  Drs.  H.  M.  Hill,  J.  A.  Miller, 
T.  B.  Carpenter,  L.  E.  Levi,  also  to  the  local  committee  of  the 
American  Association  for  the  Advancement  of  Science,  especially 
Mr.  Eben  p.  Dorr,  Secretary,  also  to  the  local  press  and  to  the 
managers  and  directors  of  the  various  works  visited;  namely,  the 
Milsom  Rendering  and  Fertilizing  Works,  Garbage  Reduction 


(H2) 

Works,  Lang's  Brewery,  Buffalo  Reduction  Company,  Calcium 
Carbide  Works  (Niagara  Works),  Cataract  Construction  Com- 
pany, Clifi  Paper  Mill,  Tonawanda  Iron  and  Steel  Company, 
Schoellkopf   Aniline  and  Chemical  Company,  Crystal  Water 
Company,  also  Jaeger's  Roof  Garden  and  Cai£. 

The  President  put  the  question  on  the  adoption  of  the  motion, 
which  was  carried  unanimously. 

The  President :  Is  there  any  further  information  desired  or  any 
further  question  to  come  up  ? 

Prof,  Mason  :  May  I  ask  this  question  :  Is  it  possible  to  so  ar- 
range matters  as  to  consolidate  the  summer  meetings  of  the 
Chemical  Society  and  Section  C  ?  I  ask  it  because  I  personally 
can  be  away  but  a  week.  The  two  meetings  occupy  more  than 
a  week.  I  should  like  to  attend  the  two  meetings  in  full,  but  I 
can  not  do  it.  My  position  is  such  that  I  am  obliged  to  return 
next  Wednesday  night.  The  result  is  I  cut  off  half  nearly  of  the 
American  Association  meeting.  Inasmuch  as  it  is  a  meeting  of 
almost  the  same  men  under  different  names,  is  it  not  possible  to 
so  arrange  matters  as  to  have  them  all  together. 

Dr.  Norton :  I  feel  very  much  as  Dr.  Mason  does.  In  order 
to  bring  this  to  decisive  action  I  move  you  that  the  Council  be 
authorized  to  use  its  discretion  in  arranging  for  a  joint  meeting 
of  this  society  and  Section  C  of  the  American  Association  next 
year.  I  think  this  will  enable  its  to  give  an  expression  to  our 
feelings  and  leave  the  Council  free  to  take  the  proper  measures. 
I  know  a  number  of  our  members  are  coming  on  next  week. 
They  do  not  feel  as  though  they  could  give  nine  or  ten  days  to 
the  meeting  of  both  societies.  There  are  a  number  present  in 
the  room  who  will  have  to  leave  next  Monday  or  Tuesday.  By 
a  little  careful  study  we  can  arrange  to  have  the  whole  chemical 
work  that  would  come  before  the  Society  and  before  Section  C 
of  the  American  Association,  carried  on  in  the  sessions  of  the 
live  days  which  are  given  up  for  that  purpose.  I  think  it  would 
be  much  more  desirable  because  we  do  not  want  our  membership 
stringing  along  through  some  seven  days,  part  of  us  listening  to 
papers  now,  and  part  at  the  end  of  the  week.  I  feel  from  con- 
versation with  a  number  of  our  members  that  there  is  a  general 
belief  that  we  ought  to  have  some  simple  arrangement  for  joint 


("3) 

meetings,  and  tliey  can  be  presided  over  alternately  by  the  Pres- 
ident of  our  Society  and  the  vice-president  of  Section  C. 

Mr,  Prescott :  I  second  the  motion,  and  I  think  at  the  present 
time  when  the  meetings  of  the  Association  are  as  the}'  are,  that 
the  plan  can  be  carried  out  much  better  than  it  would  have  been 
before  the  present  arrangement  had  taken  place. 

Prof.  Hart :  I  second  the  motion,  Mr.  President,  but  I  wish  to 
point  out  one  matter  that  ought  to  be  thought  of,  that  is,  the  in- 
creasing number  of  papers.  We  have  ten  more  papers  at  this 
meeting  than  we  had  last  year,  and  most  of  you  have  already 
received  programs  of  Section  C  of  the  American  Association  and 
can  see  what  an  enormous  program  that  is.  People  who  take 
the  trouble  and  pains  to  prepare  papers  for  these  meetings  nat- 
urally feel  that  they  would  like  to  have  the  papers  read.  That 
is  a  thing  to  which  we  should  give  careful  recognition.  If  any- 
thing of  the  kind  is  done  it  is  not  possible,  I  think,  to  secure  any 
more  time  in  Section  C  than  we  have  now,  and  according  to  the 
printed  program  that  time  is  already'  taken  up.  We  have  not 
read  more  than  one-half  the  papers. 

Prof,  Kcnnicoit :  It  does  not  seem  to  me  it  would  be  a  wise 
thing  to  sink  our  identity  in  any  other  society.  Simply  to  meet 
with  Section  C  would  seem  to  me  to  be  loss  of  identitv. 

Dr,  Hale :  Mr.  President,  it  seems  to  me  the  motion  that  has 
been  made  is  eminently  a  proper  one.  The  various  points  one 
way  or  the  other  of  difficulty  or  ease  of  adjustment  would  come 
properly  before  the  Council  for  consideration  and  they  would 
have  plenty  of  time  to  confer  with  one  another  and  consider  the 
subject.  Certain  it  is  that  we  have  a  large  number  of  chemists 
who  are  increasingly  loyal  and  devoted  both  to  the  American 
Chemical  Society  and  Section  C,  and  by  bringing  the  chemists 
together  at  this  time  we  have  undoubtedly  added  to  the  attend- 
ance and  the  interest  and  the  number  of  papers  of  both.  It  seems 
to  me  that  the  whole  subject  is  wisely  referred  to  the  Council  of 
the  Society,  and  of  course  Section  C  can  take  whatever  similar 
action  it  chooses. 

A/r,  Breneman :  I  am  quite  in  accord  with  the  resolution, 
but  it  seems  to  me  it  would  simplify  matters  very  much  if  we 
should  simply  decide  to  abolish  the  summer  meeting  and  let  the 


(114) 

winter  meeting  be  the  only  one.  Tliat  is  the  annual  meeting ;  it 
is  the  meeting  where  the  election  occurs  and  the  one  of  greatest 
interest.  I  do  not  see  any  reason  for  a  joint  meeting.  If  the 
arrangement  suggested  is  made,  the  winter  meeting  will  be  dis- 
tinctive and  the  only  annual  meeting  of  the  society. 

Prof.  Kennicott :  I  see  that  my  predictions  are  to  be  verified. 
We  have  already  started  to  sink  the  identity  of  the  Society.  A 
great  many  members  would  be  unable  to  attend  any  meeting  in 
the  winter. 

Dr,  PrescoH :  Mr.  President,  I  think  the  Council  would  be 
very  glad  if  we  could  have  a  general  expression  of  opinion  like 
that  of  Prof.  Kennicott  and  others  very  briefly  at  this  time. 

Dr,  Howe:  The  suggestion  that  has  been  made  is  one  I 
remember  when  the  original  discussion  took  place  in  regard  to 
the  reorganization  of  the  Chemical  Society.  It  was  proposed  at 
that  time  that  the  American  Chemical  Society  should  have  its 
winter  meeting,  but  that  the  summer  meeting  should  not  be  for 
the  reading  of  papers  ;  that  the  papers  then  should  be  read  at 
the  meeting  of  the  American  Association.  It  certainly  is  not 
desirable  to  carry  on  any  merging  of  identity,  at  the  same  time 
it  seems  to  me  that  the  plan  suggested  would  be  a  valuable  one 
to  those  of  us  who  are  present  here  as  chemists,  and  more  valu- 
able than  the  present  plan,  if  we  can  mass  together  all  the  papers 
and  have  all  the  members  present  in  a  four  days'  session,  so  that 
we  could  have  the  fullest  and  most  helpful  discussion.  Some  of 
us  are  unfortunately  unable  to  be  present  at  the  winter  meeting, 
but  I  think  even  for  us  it  would  be  better  if  all  the  papers  were 
presented  together  in  the  meeting  of  the  American  Association 
in  the  summer.  It  does  not  seem  to  me  we  want  to  do  anything 
to  injure  the  American  Association  or  have  things  in  such  a 
situation  that  we  feel  obliged  to  come  here  this  week  and  go  ofi 
next  week  and  miss  everything  that  goes  on  in  the  Association. 
I  think  the  Association  owes  a  great  deal  to  the  Chemical 
Society  for  what  it  has  done  in  stirring  up  an  interest  again  in 
Section  C.  I  think  there  should  be  some  amicable  arrangement 
of  this  matter. 

Prof.  Mason :  Just  one  word  I  would  like  to  say.  We  come 
here,  it  is  true,  to  listen  to  chemical  papers,  but  we  also  come  to 


(115) 

meet  chemists,  and  if  we  have  an  opportunity  of  meeting  all  the 
members  of  the  American  Chemical  Society  and  the  members  of 
Section  C  as  well,  we  fulfil  the  second  object  we  came  for  better 
than  if  we  should  string  the  meeting  over  so  many  days,  and  as 
a  result  one  man  goes  before  another  arrives  and  perhaps  they 
want  to  see  each  other. 

Dr,  McMurtrie :  Mr.  Chairman,  there  are  some  difficulties 
that  occur  to  me  in  this  connection.  The  matter  has  been  of 
course  discussed  a  good  deal  during  the  past  three  or  four  or 
five  years  ;  it  had  been  when  the  reorganization  of  the  Chemical 
Society  occurred,  and  one  of  the  important  difficulties  has  arisen 
to  my  mind  during  the  last  year  or  so  when  I  have  been  more  or 
less  active  in  the  work  of  Section  C  of  the  American  Association. 
In  preparing  the  program  of  proceedings  for  last  year  I  was 
reminded  that  it  was  impossible  to  have  papers  presented  in  the 
meetings  of  Section  C  by  any  other  than  members  of  the  Ameri- 
can Association.  Now  there  is  a  larger  portion  of  the  member- 
ship of  the  American  Chemical  Society  who  are  not  members  of 
Section  C.  The  consolidation  of  the  meetings  will  necessarily 
rule  out  those  men  from  participation  in  the  meeting.  This  point 
would  of  course  be  brought  before  the  Council  in  a  discussion  of 
the  matter,  and  would,  I  suppose,  have  weight.  There  are  a 
good  many  members  of  the  Chemical  Society,  T  know,  who  feel 
that  they  do  not  care  to  have  membership  in  the  American  Asso- 
ciation, and  it  seems  to  me  that  in  any  action  we  take  in  this 
regard  their  wishes  should  be  carefully  considered.  I  think  it 
might  be  possible  to  arrange  to  have  the  meetings  succeed  each 
other  in  the  same  week,  but  that  arrangement  which  has  been 
followed  in  the  past  year,  has  been  objected  to  by  the  officers  of 
the  American  Association,  holding  that  it  interfered  in  a  large 
measure  with  the  work  of  the  Association.  It  was  in  a  measure 
on  this  account  that  the  meetings  of  the  Association  are  fixed 
for  the  week  continuously  ;  that  is  beginning  with  Monday.  So 
that  those  societies  which  are  called  by  the  officers  of  the  Asso- 
ciation, affiliated  societies  might  have  their  meeting  either  in 
the  week  preceding  or  succeeding  the  meeting  of  the  Associa- 
tion. There  seems  to  be  a  feeling,  I  think,  among  a  good  many 
of  the  officers  of  the  Association  that  it  would  be  in  a  measure 


(116) 

imp^^ible  to  secure  the  coalescence  of  the  different  societies 
with  the  similar  sections.  As  I  say,  all  these  points  will  be 
brought  necessarily  before  the  Council  in  the  discussion  of  the 
matter,  and  it  is  the  only  way  in  which  it  can  be  determined 
after  all.  It  must,  under  the  constitution,  go  before  the  Council 
before  it  is  open  again  to  be  brought  before  the  Society. 

Mr,  Cochran  :  Before  the  question  is  put  I  would  like  to  ask 
one  question,  and  that  is  this  :  I  myself  see  some  objections  to  it, 
but  I  do  want  to  attain  it ;  I  would  like  to  attend  both  meet- 
ings if  I  could;  I  would  like  to  be  here  when  all  the  chemists  are 
here.  This  year  particularly,  my  time  is  very  limited  ;  I  shall 
leave  Buffalo  this  evening  and  be  cut  out  of  the  meetings  next 
week  entirely.  The  question  I  wanted  to  ask  is  this :  Is  it 
impossible  that  both  meetings  should  run  on  at  the  same  time? 
Could  we  not  have  a  section  meeting  or  the  meeting  of  the 
American  Chemical  Society  conducted  at  the  same  time 
that  the  meeting  of  Section  C  of  the  American  Association  is 
conducted*?  The  programs  are  large.  Some  of  us  would  desire 
to  hear  papers  in  one  section  one  week  and  some  in  the  other, 
and  in  that  way  we  could  save  our  time  and  get  the  papers  pre- 
sented so  that  we  could  all  hear  them.  I  know  there  are  objec- 
tions to  it,  but  at  the  same  time  I  desire  to  have  the  subject  con- 
sidered. 

Dr,  McMurtrie :  We  are  not  alone  in  this  matter.  Nearly  all 
the  other  sections  of  the  Association  are  in  about  the  same  posi- 
tion, and  it  is  coming  to  be  a  serious  question  as  to  what  shall 
be  done  in  this  matter,  whether  the  American  Association  shall 
be  taken  into  a  confederation  of  scientific  societies,  or  whether 
some  such  plan  as  is  suggested  now  shall  be  carried  out.  It 
seems  to  me  that  this  might  be  permitted  to  grow  into  a  confed- 
eration. 

The  President :  I  was  about  to  remark  on  that  same  subject 
that  there  are  othei-  affiliated  bodies  exactly  in  the  same  position 
as  Dr.  McMurtrie  has  said,  so  that  it  is  obvious  this  question  is 
a  serious  one. 

The  President  put  the  question  and  it  was  adopted. 


("7) 

Dr.  Norton :  Mr.  President,  I  would  like  to  say  a  few  words 
as  to  what  has  been  said  in  regard  to  the  pleasure  and  profit  we 
have  all  had  in  meeting  together  as  a  Society  during  the  past 
few  days,  and  I  feel  that  our  success,  which  is  actually  a  marked 
one  this  year  in  point  of  attendance  and  interest,  is  due  not  only 
to  the  efforts  of  our  Local  Committee,  but  also  to  the  able  prepa- 
ration made  in  advance  for  the  meeting  by  the  officers  of  the 
Society,  and  I  would  therefore  like  to  move  before  we  separate 
to-day,  that  the  cordial  thanks  of  the  Society  be  expressed  to 
the  President  and  Secretary  for  the  measures  which  they  have 
taken  to  render  this  meeting  so  successful,  and  I  would  like  to 
ask  the  Nestor  of  the  Society,  Dr.  Prescott,  to  put  that  motion. 

Dr,  Prescott:  I  am  very  glad  to  place  this  motion  before  you 
and  have  an  opportunity  to  vote  for  it. 

Dr.  Prescott  put  the  motion  which  was  unanimously  carried. 

The  President :  In  behalf  of  the  ofiicers,  I  will  only  say  that 
it  is  a  regret  on  their  part  that  most  of  us  are  so  busy  with  our 
daily  life  that  we  can  not  give  all  that  is  in  our  hearts  and 
minds  to  do  for  the  inteiiests  of  the  Society,  and  we  thank  you 
for  your  vote.     ( Applause) . 

The  President :  I  declare  then  the  meeting  adjourned  until  the 
Troy  meeting. 


ANNOUNCEMENT. 

All  persons  who  have  papers  to  offer  for  the  next  general  meet- 
ing, which  will  be  held  the  latter  part  of  December  in  Troy, 
N.  Y.,  are  requested  to  forward  at  their  earliest  opportunity  an 
abstract,  or  the  full  manuscript  of  their  papers  together  with 
titles  and  names  of  authors  to  the  General  Secretary,  Albert  C. 
Hale,  551  Putnam  Ave.,  Brooklyn,  N.  Y.^  so  that  the  papers 
may  all  be  passed  upon  by  the  committee  on  papers  and  publica- 
tions in  time  for  announcement  upon  the  program  which  must 
be  in  print  before  the  meeting. 


(ii8) 

CHANGES  OP   ADDRBSS. 

Burt,  M.  C,  io6  Chestnut  St.,  Springfield,  Mass. 

Conradson,  P.  H.,  Franklin,  Pa. 

Bakins,  L.  O.,  care  of  Guppenheim  Smelting  Co.,  Perth  Am- 
boy,  N.  J. 

Lane,  Henry  M.,  care  of  Washington  Agricultural  College, 
Pullman,  Wash. 

Mar,  P.  W.,  32  McDonough  St.,  Brooklyn,  N.  Y. 

McCrae,  John,  7  Kirklee  Gardens,  Kelvinside,  Glasgow, 
Scotland. 

Welles,  Albert  H.,  635  Quincy  Ave.,  Scranton,  Pa. 

Whitehead,  Robt.  L.,  Box  142,  Mt.  Washington,  Md. 


iMued  with  VcnrttobcT  Number,  .1896. 


Proceedings. 


COUNCIL. 

By  direction  of  the  Council  a  congratulatory  address  was  for- 
warded to  Stanislas  Canizzaro,  an  honorary  member  of  this 
Society,  upon  his  seventieth  birthday. 

December  29  and  30  has  been  selected  as  the  date  for  the 
annual  meeting  at  Troy,  N.  Y. 

MEMBERS  ELECTED  SEPTEMBER  21,  1 896. 

Belden,  A.  W.,  Chapel  Hill,  N.  C. 
Blair,  Augustine  W.,  Guilford  College,  N.  C. 
Chamot,  Emile  M.,  Cornell  Univ.,  Ithaca,  N.  Y. 
Davis,  Dr.  Floyd,  Des  Moines,  Iowa. 

HaUer,  H.  Loft,  F.C.S.,  27  Hilda  St.,  Beverly  Road,  HuU, 
England. 

Hotopp,  C.  H.,  Stroudsburg,  Pa. 

Kruskal,  Dr.  Nicholas,  72  Delancy  St.,  N.  Y.  City. 

Marlatt,  Miss  Abby  L.,  Providence,  R.  I. 

Meade,  Richard  K.,  Longdale,  Alleghany  Co.,  Va. 

Patrick,  George  E.,  Dept.  of  Agr.,  Washington,  D.  C. 

Slosson,  E.  E.,  Laramie,  Wyo. 

Smith,  Prof.  E.  G.,  Beloit  College,  Beloit,  Wis. 

Stahl,  Dr.  Karl  P.,  57th  St.  and  A.  V.  R.  R.,  Pittsburg,  Pa. 

Tolman,  Prank  L.,  U.  S.  Naval  Lab.,  Brooklyn,  N.  Y. 

ASSOCIATE  ELECTED  SEPTEMBER  21,  1896. 

Brinton,  C.  S.,  West  Chester,  Pa. 

CHANGES  OP   ADDRESS. 

Bachman,  Irving  A.,  Allentown,  Pa. 
Behr,  Amo,  17  Lawn  Ridge,  Orange,  N.  J. 
Doerflinger,  Wm.  P.,  85  Lafayette  Ave..  Brooklyn,  N.  Y. 
Boot,  J.  C,  Brooklyn  Distilling  Co.,  Kent  Ave.,  Brooklyn, 
N.  Y. 
Dal  Molin,  A.  A.,  30  E.  i8th  St.,  N.  Y.  City. 
Davidson,  Geo.  H.,  28  Woodbine  St.,  Brooklyn,  N.  Y. 
Fuller,  Fred.  D..  Agr.  Expt.  Sta.,  Geneva,  N.  Y. 
Habirshaw,  William  M.,  Glenwood  Works,  Yonkers,  N.  Y. 


(I20) 

Hollick,  Herbert,  Post  OflSce,  New  York  City. 
Kutroff,  Adolph,  128  Ouane  St.,  New  York  City. 
Loeb,  Morris,  118  W.  72nd  St.,  New  York  City. 
Munsell,  C.  E.,  no  Horatio  St.,  New  York  City. 
Sargent,  Geo.  W.,  Univ.  of  Pa.,  Dormitories,  37th  and  Spruce 
Sts.,  Philadelphia,  Pa. 
Thompson,  P.,  11  Willmot  St.,  Ann  Arbor,  Mich. 
Tidball,  Walton  C,  291  Prospect  PI.,  Brooklyn,  N.  Y. 
Volckening,  G.  J.,  65  Van  Buren  St.,  Brookl3m,  N.  Y. 


MEETINGS  OF  THE  SECTIONS. 

RHODE   ISLAND  SECTION. 

The  first  meeting  for  the  year  1896-97  was  held  at  Provi- 
dence, on  Thursday  evening,  September  24th,  Chairman  K.  D. 
Pearce  presiding. 

Prof.  J.  H.  Appleton  read  a  paper  upon  the  **  Electrolysis  of 
Salt." 

The  introduction  to  this  paper  was  a  brief  discussion  of  the 
present  chemical  application  of  electricity.  First  in  importance 
at  present  is  the  preparation  of  metals,  copper,  gold  from  cya- 
nide solutions,  zinc,  glucinum,  and  even  some  more  difficultly 
reducible  metals  or  non-metals  ;  sodium,  lithium,  cadmium, 
cobalt,  nickel,  and  phosphorus.  Next,  reference  was  made  to 
the  production  of  certain  compounds  in  which  primarily  the 
heat  of  the  current  is  involved :  silicon  carbide,  calcium  car- 
bide, as  well  as  those  metallic  carbides,  produced  by  Moissan, 
which,  with  water,  yield  such  varied  hydrocarbons  (very  sug- 
gestive in  relation  to  the  origin  of  petroleum). 

The  electrolysis  of  salt  by  several  methods,  notably  Castner's, 
was  next  taken  up. 

In  conclusion,  there  were  presented  some  comments  on  the 
probable  influence  of  the  electrolysis  of  salt  on  the  alkali  in- 
dustry. 


new  york  section. — annual  meeting. 

October  9,  1896. 

The  meeting  was  called  to  order  at  8.20  p.  m.,  by  Dr.  P.  T. 
Austen,  Chairman. 


(121) 

In  the  absence  of  the  secretary,  Dr.  A.  C.  Hale  was  appointed 
secretary /n?  tern. 

The  minutes  of  the  meeting  held  June  8th,  1896,  were  read 
and  approved. 

Reports  of  officers  and  committees  being  in  .order,  the  chair- 
man called  upon  Dr.  Hale  to  report  for  the  delegates  to  the 
Scientific  Alliance  of  New  York.  Dr.  Hale  made  a  brief  oral 
report,  which  was  accepted. 

Dr.  P.  T.  Austen,  the  retiring  chairman  of  the  section, 
reported  on  the  work  of  the  year  and  the  general  condition  and 
prospects  of  the  section  and  the  society  as  well  as  the  outlook 
for  American  chemists  generally. 

After  these  remarks  by  the  retiring  chairman  the  election  of 
officers  of  the  section  for  the  ensuing  year  was  held. 

Dr.  Durand  Woodman  was  unanimously  elected  secretary  and 
treasurer.     Other  officers  were  elected  unanimously,  as  follows  : 

Chairman — Dr.  Wm.  McMurtrie. 

Executive  Committee — Dr.  Charles  A.  Doremus,  Prof.  A.  A. 
Breneman,  Dr.  Albert  C.  Hale. 

Delegates  to  the  Scientific  Alliance  of  New  York — Dr.  Wm. 
McMurtrie,  Dr.  C.  F.  McKenna,  Dr.  C.  A.  Doremus. 

Dr.  Wm.  McMurtrie,  chairman-elect,  then  took  the  chair,  and 
upon  motion  of  Dr.  Doremus,  a  vote  of  thanks  to  the  retiring 
chairman  was  passed  unanimously. 

Prof.  Breneman  reported  very  encouraging  progress  in  refer- 
ence to  the  proposed  chemical  club. 

Papers  were  read  and  discussed  as  follows  :  **  Some  Disputed 
Points  about  the  Light  of  Carbon,**  by  Woodbridge  H.  Birch- 
more  ;  discussed  by  Prof.  Speyers,  Mr.  Birchmore,  and  Mr.  Still- 
well.  **The  Conversion  of  Cow  Milk  into  a  Substitute  for 
Human  Milk,"  by  Henry  A.  Bunker;  discussed  by  Dr.  Eccles, 
Dr.  Bunker,  and  Dr.  McMurtrie. 

Upon  motion  of  Dr.  Doremus,  seconded  by  Dr.  Squibb,  the 
following  named  persons  were  appointed  a  committee  to  cooperate 
with  other  scientific  bodies  in  New  York  for  the  purpose  of 
securing  a  lecture  from  Prof.  Henri  Moissan  before  his  return  to 


(122) 

Prance :  C.  A.  Doremus,  A.  A.  Breneman,  M.  Loeb,  and  Wm. 
McMurtrie. 

Upon  motion  of  Dr.  Hale,  the  chairman  of  the  section  was 
authorized  and  requested  to  appoint  a  committee,  with  him- 
self as  chairman,  to  arrange  the  programs  for  the  meetings  of 
the  section  during  the  year. 

The  meeting  then  adjourned. 


Issued  with  December  Nnmber,  1896. 

Proceedings. 

COUNCIL. 

At  the  Btt£Falo  meeting  of  the  American  Chemical  Society  it 
was  voted  that  the  Council  be  requested  to  take  into  considera- 
tion ways  and  means  for  bringing  the  Summer  meeting  of  the 
Society  into  closer  relation  with  that  of  Section  C  of  the  A.  A. 
A.  S.,  so  that  both  meetings,  if  possible,  may  be  held  within  the 
same  week,  thus  affording  the  opportunity  for  all  chemists  to 
attend  both  meetings. 

Inasmuch  as  both  these  bodies  were  well*  represented  at  the 
meeting  referred  to,  it  was  suggested  that  a  good  deal  of  time 
could  be  gained  by  appointing  then  and  there  a  committee  of 
conference  from  each.  This  was  accordingly  done,  the  commit- 
tee on  the  part  of  the  American  Chemical  Society  being  Messrs. 
W.  P.  Mason,  W.  McMurtrie,  Edward  Hart,  T.  H.  Norton  and 
A.  B.  Prescott. 

A  joint  meeting  of  this  Committee  was  held  with  a  Committee 
of  Section  C,  and  the  following  recommendations  were  agreed 
upon  : 

ist.  Section  C  to  have  a  business  meeting  for  purposes  of 
organization  on  Monday  of  the  week  of  meeting,  and  the  Vice- 
President's  address  to  take  place  late  in  the  afternoon  of  that 
day. 

2nd.  The  American  Chemical  Society  to  be  given  Monday 
and  Tuesday  for  their  work. 

3rd.  Section  C  of  the  A.  A.  A.  S.  to  be  given  the  balance  of 
the  week. 

4th.  The  arrangement  of  the  program  for  the  reading  of  papers 
before  the  two  bodies  to  be  left  to  the  discretion  of  the  President 
of  the  American  Chemical  Society  and  the  Vice-President  of 
Section  C  of  the  A.  A.  A.  S. 

These  recommendations  were  approved  by  Council  Oct.  27, 
1896. 


(124) 

In  view  of  the  increasing  number  of  papers  presented  at  the 
meetings,  the  Council  has  decided  that  the  Troy  meeting  shall 
extend  over  three  davs  if  this  shall  be  found  necessarv. 


NEW  MEMBERS  ELECTED  XOVEMBEK  5,   1896. 

Andrews,  Prof.  W.  W.,  Sackville,  New  Brunswick. 
Burner,  Prof.  N.  L.,  Ohio  Med.  Univ.,  Columbus,  Ohio. 
Case,  Wm.  A.,  Mt.  Washington,  Baltimore  Co.,  Md. 
Clark,  Arthur  W.,  Conshohocken,  Pa. 
Evans,  Wm.  Lloyd,  Ohio  State  Univ.,  Columbus,  O. 
Fossler,  Miss  Mary  L.,  734  N.  9th  St.,  Lincoln,  Nebr. 
Hochstetter,  Robert  W.,  Oak  St.  and  Bellevue  Ave.,  Cincin 
nati,  O. 

Levi,  Louis  E.,  Ph.D.,  548  Franklin  St.,  Buffalo,  N.  Y. 
Mathews,  John  Alex.,  Columbia  Univ.,  N.  Y.  City.* 
Mooers,  Chas.  A.,  Agr.  Exp.  Sta.,  Knoxville.  Tenn. 
Schoen,  Joseph,  2317  Indiana  Ave.,  Chicago,  111. 
Schroeder,  J.  Henry,  Grand  and  Nassau  Sts.,  Cincinnati,  O. 
Sturcke,  H.  E.,  284  Pearl  St.,  N.  Y.  Citv. 
Sy,  Albert  P.,  Univ.  of  Buffalo,  24  High  St.,  Buffalo,   N.  Y. 
Wessling,  Prof.  Hannah  L.,  147  Milton  St.,  Cincinnati,  O. 

ASSOCIATES  ELECTED  NOVEMBER  5,   1 896. 

Cooley,  Fred.  C.  1029  L  St.,  Lincoln,  Nebr. 
Culver,  Frank  S.,  1610.K  St.,  Lincoln,  Nebr. 
Dales,  Benton,  1242  P  St.,  Lincoln,  Nebr. 
Himrod,  George,  1446  Q  St.,  Lincoln,  Nebr. 
Hiltner,  Martin  E.,  1301  N  St.,  Lincoln,  Nebr. 
Lange,  Miss  Helen  P.,  346  N.  17th  St.,  Lincoln,  Nebr. 
O'Sullivan,  Miss  Eva,  445  N.  13th  St..  Lincoln,  Nebr. 
Pharmelee,  Howard  C,  care  of  Cooperative  Book  Co.,  Lin- 
coln, Nebr. 

Thatcher,  Roscoe  W.,  540  N.  15th  St.,  Lincoln  Nebr. 

CHANGES  OF    ADDRESS. 

Allen,  Walter  S.,  34  So.  6th  St.,  New  Bedford,  Mass. 
Cushman,  Allerton  S.,  Chemical' Laboratory,  Harvard  Univ., 
Cambridge,  Mass. 

Hancock,  David,  1720  Fifth  Ave.,  Birmingham,  Ala. 
Reese,  Chas.  L.,  1801  Linden  Ave.,  Baltimore,  Md. 
Sturm,  Arthur  B.,  Box  92,  May  wood.  111. 

ADDRESS   WANTED. 

Bradley,  Edson,  formerly  of  35  Broadway,  New  York  City. 


(125) 

MEETINGS  OF  THE  SECTIONS. 

RHODB  ISLAND  SECTION. 

A  meeting  of  the  Rhode  Island  Section  was  held  at  Provi- 
dence, on  Thursday  evening,  October  29,  1896. 

Mr.  E.  D.  Pearce  mentioned  the  results  of  his  experiments  in 
bleaching  brown  tower  acid.  Samples  pf  acid  taken  before  and 
after  bleaching  were  exhibited.  Mr.  Pearce  also  stated  that 
the  coloring  matter  of  the  anthers  of  the  wild  evening  prim- 
rose was  altered  by  acids  and  alkalies  in  the  same  way  as 
turmeric. 

A  paper  was  read  by  W.  M.  Saunders  upon  **  The  Determina- 
tion of  Sulphur  in  Iron."  The  reader  described  briefly  the  del- 
eterious effect  of  sulphur  in  iron.  The  small  amount  permitted 
in  foundry  work,  and  the  diflSculty  of  determining  this  amount 
was  mentioned. 

Next  a  description  of  methods  of  analysis  was  given.  The 
reader  considered  the  evolution  methods,  although  not  in  every 
case  giving  the  full  sulphur  contents  of  the  iron,  to  be  accurate 
enough  for  practical  purposes.  The  results  compare  favorably 
with  the  oxidation  method. 


NEW  YORK  SECTION. 

The  November  meeting  of  the  New  York  Section  was  held  on 
the  6th,  Professor  McMurtrie  in  the  chair,  and  fifty-one  mem- 
bers present. 

The  chair  announced  the  acceptance  by  the  executive  com- 
mittee of  an  invitation  from  Drs.  Morton  and  Leeds  to  hold  the 
December  meeting  at  the  Stevens  Institute  of  Technology. 

The  death  of  Mr.  Alfred  H.  Mason  was  announced  and  a 
sketch  of  his  life  was  read. 

A  motion  was  made  and  seconded  that  the  executive  commit- 
tee be  recommended  to  authorize  the  secretary  to  employ  a 
stenographer  to  report  the  discussions  of  papers  presented  at  the 
meetings  ;  such  report,  when  properly  edited,  to  be  sent  to  the 
committee  on  papers,  for  publication  in  the  Journal. 

The  following  papers  were  read  : 


(126) 

"  The  Volumetric  Determination  of  Acetone,"  by  Dr.  E.  R- 
Squibb. 

**  Notes  on  a  Chemist's  Trip  Abroad,"  by  C.  A.  Doremus. 

*•  A  New  Form  of  Pyknometer,"  by  J.  C.  Boot. 

**  Improvements  in  the  Colorimetric  Tests  for  Copper,"  by 
Geo.  I^.  Heath. 

*' Note  on  Solubility  of  Bismuth  Sulphide  in  Alkaline  Sul- 
phides," by  Geo.  C.  Stone. 

The  meeting  then  adjourned. 

CINCINNATI   SECTION. 

The  meeting  was  held  on  November  17,  in  the  Woyd  Library. 
After  welcoming  the  Society^  Prof.  J.  U.  Lloyd  read  a  paper  en- 
titled "Bibliography  of  American  Pharmacy,"  giving  a  concise 
history  of  the  different  editions  of  the  U.S.  Pharmacopeia  and 
its  commentaries,  the  various  dispensatories  and  formularies. 
This  paper  was  rendered  doubly  interesting  by  the  exhibition  of 
the  rare  old  editions  of  these  works  from  the  well-filled  shelves 
of  the  Lloyd  Library.  Prof.  O.  W.  Martin  read  a  paper  opening 
the  discussion  on  the  **  Teaching  of  Elementary  Chemistr>-."  A 
contribution  on  this  subject  by  Dr.  James  Lewis  Howe  was  read 
by  Mr.  H.  B.  Poote.  Excerpts  from  paper  which  Prof.  Paul 
Freer  presented  at  the  summer  meeting  of  American  Association 
for  the  Advancement  of  Science,  elicited  considerable  discussion 
by  Dr.  Springer,  Profs.  Norton,  Martin  and  Homburg. 

Dr.  William  H.  Crane,  of  Cincinnati  was  elected  a  member  of 
the  Section.