26 BULLETIN 1115, U. S. DEPARTMENT OF AGRICULTURE. 
Fic. 48.—Average results with AS-3, BS-3, CS-3, FS-3, IS-3 (paper-lined sugar barrels). 
All of the sugar barrels were in bad condition at the end of the test | 
because of handling incident to sampling. The results of the tests | 
show the unfitness of this type of container as a package for com- 
mercial purposes. 
It is plainly shown that the containers are rated in the same order 
by both the carbon dioxid and the water-soluble arsenic oxid results. } 
The metal drums show no change and the sugar barrels show the } 
most rapid change, the other two being intermediate. Little dif- 
ference between the lined and unlined containers of the same type 
is observed. The calcium arsenate stored in the sugar barrels } 
apparently reached equilibrium in most cases in approximately | 
eight months. No seasonal variation in the rate of change is 
apparent. 
Very little information concerning the effect of locality was ob- 
tained. Strictly comparable results were possible only with samples 
AW and AT, and samples HW and HT. In both cases the rate of 
absorption of carbon dioxid was higher at Tallulah than at Wash- 
ington. The authors are not in a position to state whether or not 
this was due to a higher carbon dioxid content in the air at Tallulah 
or to some other condition. 
VY WA ®AAN 
S 96O-9-9--S45-S 
FER CLIVI™ 
~ 
Fig. 49.—Average results with AS-4, BS-4, CS-4, FS-4, HS-4, IS-4 (hardwood barrels), 
