922 



JOURNAI, OF FORESTRY 



Table 4. — Percentage of Possible Number of Times Fires Could Occur. Basis, 

 12 Forests — 6 Years. 



Forests, but we can only conclude that a difference in number of fires 

 as great as 62 per cent is a significant difference. This is due. not 

 only to the fact that fires do not occur so often on north as on south 

 slopes, but probably also to the fact that a considerable number of 

 fires burn out of themselves on the north slopes while this is a com- 

 paratively rare occurence on the south slopes. At any rate the wide 

 divergence in the values for the two is too great to be explained entire- 

 ly on the grounds of coincidence. 



In this connection it is interesting to refer back to the paper cited 

 earlier. It was found that fires can not spread if moisture content of 

 litter is over about 8 per cent, and that one season on typical north 

 and south slopes the moisture content was above danger point one- 

 fifth of the time on south slopes and nearly one-half of the time on 

 north slopes. In other words the two entirely independent lines of in- 

 vestigation check in at least a qualitative measure. 



RELATION OF RATE OF SPREAD TO COVER CONDITIONS 



One exceedingly important control of rate of spread has not yet 

 been discussed; namely, the influence of cover conditions. In the an- 

 alysis of the fire statistics all fires occurring for three years were seg- 

 regated into two classes, those occurring in the timber and those in 

 brush. Among the latter were included some fires which were es- 

 sentially brush fires, although strictly speaking they occurred in tim- 

 ber. In so far as suppression is concerned, a fire in dense under- 

 brush in an open timber stand should be classed with brush fires 

 rather than timber fires and this practice has been followed. As can 

 be seen from the following table, very striking differences in rate of 

 spread were found for the two sets of conditions. 



