WIDTH OF FIRE-TRACES. 427 



It is evident that the more a fire-trice deviates from the per- 

 pendicular to the direction of a constant wind, the easier, more 

 economical and less fraught with risk does its clearing hecome. 

 But in observing this obvious rule, there is one danger that has to 

 be avoided. It must first be clearly ascertained that the wind in 

 question is really a constant one, and that it has no tendency to 

 veer completely round all of a sudden. Should the wind change 

 in this way while the trace was being fired, the flames would rush 

 along in the direction of the uuburut length of the trace, and in a 

 few minutes, perhaps even in a few seconds, extend so far as to get 

 entirely out of hand and then spread away into the forest. Such 

 an eventuality is most to be feared in dry, short wiry grass, in 

 which the fire may travel along at the rate of four miles an hour 

 before a very moderate wind. Hence, unless there is complete 

 certainty that the prevailing wind is not subject to sudden changes, 

 and there is some latitude for choice, it is always advisable to let 

 the fire-trace make a small angle (10 to 15 deg.) with the direction 

 in which it blows. The full force of this recommendation will bo 

 perceived when we come to treat of the method of firing traces. 



II. Width of fire-traces. 



A trace may be intended to serve as a complete fire-break in it- 

 self, or its object may be simply to arrest to a certain extent the 

 progress of a forest conflagration and to serve as a base from which 

 to counterfire. In the first case, it should be broad enough to 

 prevent the most violent fire from crossing it in any circumstance ; 

 in the other, its width should be merely sufficient to arrest the pro- 

 gress of a fire at most points and to make it easy to light rapidly, 

 and with the minimum of labour to control, a counterfire in order 

 to meet an advancing conflagration. It is evident that all boundary 

 traces should be of the former class, while internal ones may be 

 either the one or the other, according as the staff of watchers is or is 

 not sufficiently numerous and well-distributed to start counterfires 

 promptly before the conflagration to be restricted can reach trie- 

 trace in question. 



It may be objected that all traces without exception should bo 

 effective fire-breaks ; but it must be remembered that the broader 

 the traces are, the more expensive are they to clear, and, what 

 is a far more important consideration, the larger is the area ab- 

 stracted from protection, for as the aggregate length of internal 

 traces in a valuable and well-exploited forest is always great, n, 

 few additional feet of width would necessarily involve a consider 



