4 BULLETIN 379, U. S. DEPARTMENT OF AGRICULTURE. 
middle of September. They increased in frequency as the season 
progressed until it was not unusual to have from 6 to 10 a day. On 
both August 11 and August 16 there were 11 explosions and fires 
reported, and the total for the week from August 9 to 14 reached 46, 
an average of between 6 and 7a day. The explosions were clasied 
by days and weeks as follows: 
ifriny We ee | Se) A nesses Fis | 
Siphiy OS OF ote hs Be ee Ae 3 eerie: 
Ulva l= See Siege i wae 6 166 
AS 2 oe Ss Baas oS Se ae 32 Fulyy lke lei ee 10 
PAS gO WA oe ee Se a SAG August 2 ene ae: 49 
Aue 16-21 ce 2 ee a eae fo 
Aug. 22-28_________------------- 29 Maximum, one week22 22 == 46 
Aug. 30-Sept. 4------___-------- 18 Maximum, one day, Aug. 10,16___ 11 
SepivG il G2 S220 ee ee es 2 
CLASSIFICATION BY TIME OF DAY. 
The explosions and fires occurred at all times during the day. The 
following tabulation shows the time of 128 occurrences: 
On Geasennes saree: * xix Lee D2 Oss Dr TE See ee ee 14 
EEL ORY 7 eg 14 Vi alc ee ae etl a 3 ee Oud COA pe Sc) ee BSS ee 13 
CgbORSy Aq Me Soe Se we F = bt SES HCO ypc 22s ee ore eee eee 12 
Set OnO vas Th ste Seen Sa ye eee 3 1 D5t0. Gipy  See ee Cees LE 
Satomi 4 nee Wes 2 lars ake, 9:56 to: Tp. eee i eee ay 
OMRON EE sae eee eae ge Foo Se ee 4.008 pects! 2 See eee 5 
LO p02 OOM eee eee e 2 ee 10 
TEER | Mod BEN 75 6 Peep 2 fac RE eR 3 128 
TE Re fly OR 0 eae ge Per a 8 
From this it will be noted that the greatest number of explosions 
occurred between the hours of 1 and 7 p. m. 
MAKE OF MACHINES. 
The explosions were generally distributed among all types of 
separators operated in the territory and were not confined to any one 
make or type. It was thought that possibly the explosions and fires 
were confined to stationary machines, but investigators learned that 
there were at least three occurrences in the “ combine” type. 
The investigation shows that the explosions occurred in both steel 
and wooden separators. In view of the fact that data are not avail- 
able to determine the number of machines of each type in the terri- 
tory, it is impossible to state the percentage of each affected. 
