^"""iTi?;?™"}       Barhadoes  and  Socotrine  Aloes,  549 
line  principles  by  its  inferior  solubility  and  especially  by  the  circum- 
stance that  it  yields  no  chrysammic  acid  nor  definite  chloro-  or  bromo- 
substitution  derivatives. 
Taking  all  these  circumstances  into  consideration,  1  am  unable  to 
adopt  the  suggestion  of  Rochleder  that  these  bodies  constitute  three 
successive  terms  of  a  homologous  series.  On  the  contrary,  the  analyt- 
ical results  obtained  by  different  experiments  indicate  that  barbaloin  and 
zanaloin  have  the  same  composition.  They  must,  therefore,  be  repre- 
sented by  the  same  formula.  I  propose  for  them  both  in  the  anhy- 
drous state  the  symbols  C^gH^gO^,  which,  as  w\\\  be  seen  by  the  an- 
nexed statement  of  percentages,  agrees  satisfactorily  with  all  the  ana- 
lytical numbers.  This,  which  is  simpler  than  the  formula  hitherto  re- 
ceived, is  also  conformable  with  the  statement  of  Graebe  and  Lieber- 
mann  that  aloin  yields  anthracene  or  some  closely  allied  hydrocarbon, 
perhaps  methylanthracene,  when  heated  with  zinc. 
^16^20^7  ^le^^lS^T  ^16^16^7 
requires  requires  requires 
C    50-25  C    59'62  C  60*00 
H     6-17  H     5-59  H  5*00 
Bromo-derivative 
Chloro-derivative 
CieHigBrgO^ 
requires 
^  34*34 
H  2-68 
Br  42-93 
C16H15CI3O7 
requires 
C  45-12 
H  3-52 
CI  25-03 
Acetyl-derivative 
B,,H,(C,H30)P, 
requires 
C  58-92 
H  5-35 
This  formula,  Cj^H^gO^,  has  been  proposed  by  Ro:hleder  for  nata- 
loin  and  it  agrees  closely  with  the  results  of  my  analyses  of  that  sub- 
stance, but  the  discussion  of  this  part  of  the  question  must  hz  deferred 
till  further  experimental  results  have  been  accu nulated. — Phar.  Jour, 
and  Trans, ^  1875,  Sept.  11. 
