358 



This grouping stands rather for cuhltral differences, however, 

 than for fundamental distinctions between varieties. It frequently 

 happens that the same variety may be raised in one region as a 

 wet-land crop, and in an adjacent region as a dry-land crop. It 

 is probable that kalo was first grown entirely without irrigation, in 

 lands naturally moist. The evolution of extensive irrigation sys- 

 tems was a later and lengthy agricultural development. 



Descriptive data relative to the varieties of kalo have been 

 collected by the authors, through extensive conferences with na- 

 tive Hawaiian kalo planters. This data is necessarily fragmen- 

 tary, but it is believed to possess intrinsic value. 



{To be continued.) 



GOVERNMENT REPORTS LOWEST FIRE LOSS. 



With the middle of September the fire season on the national 

 forests came practically to an end with less damage than ever re- 

 corded. There is always some danger from carelessness of 

 camp or of settlers burning brush and clearing land, but the real 

 danger season extends only from the middle of June until the 

 middle of September. 



Forest officers throughout the west are congratulating them- 

 selves on a season so markedly free from heavy losses. They 

 feel that the immunity from loss has been due to two principal 

 causes: partly due to a favorable season, but largely to a much 

 better organization for fire protection than has been attained here- 

 tofore. The effectiveness of the organization is shown particu- 

 larly by the fact that while there were in all approximately 2260 

 fires, as against 2470 last year, yet the area burned so far this 

 year is only about 60,000 acres, as against 230,000 acres in 1912, 

 and 781,000 in 1911. California, Arizona and New Mexico have 

 suffered most during the past season. 



The various causes of fires have not changed greatly in their 

 relative proportions. Railroads and lightning head the list, with 

 campers next. There has been, however, a marked decrease in 

 the number of fires caused by burning brush, which, according to 

 the forest officers, indicates a closer cooperation with the settlers 

 in and near the forests and with timbcrland owners in fire protec- 

 tion and control. It is still true, nevertheless, that a large pro- 

 portion of all fires started are due to human agencies and may 

 generally be charged against carelessness, l-'ires caused by light- 

 ning are, of course, not preventable, but the .system of lookouts 

 by which they may be detected immediately after being set is 

 greatly lessening the loss from this source. 



