118 GIBBS, 



reached in the best managed nipales. It should, however, be exceeded. Some 

 oijerators obtain only 4,000 liters per hectare during a full season. 



In the year 1909 one of the largest producers obtained 8,700,000 liters of tuba 

 from an estimated area of 1,000 hectares which is at the rate of 8,700 liters 

 per hectare. Less than one-third of this area was producing, partly from lack 

 of labor and partly because the greater portion of the estate had not been made 

 accessible to the distillery. The alcohol production from this yield was about 

 5 per cent. 



The atmospheric and climatic conditions affect the flow of sap. The 

 natives working among the plants are positive that they can predict the 

 approach of a storm by a sudden checking, in some cases amounting to 

 almost a stoppage, in the flow, occuring about three days before the 

 arrival of the tj^hoon. It is stated that after a temporary stoppage has 

 been produced by meteorologic conditions an increased flow will start 

 which will balance the period of inactivity. 



The seasonal variation in the composition of the fresh tuba must be 

 slight, although when received at the distilleries the differences, both 

 in quality and quantity, are very great. This is undoubtedly due to the 

 fact that rains dilute the juice after it has flowed from the flower or fruit 

 stalk. The gain in volume is balanced by the loss in sugar and alcoholic 

 content. 



As a result of the experimental work performed by myself and others 

 associated with me in this work, I believe that the average conditions in 

 a nipale cared for according to present methods will result in a yield 

 of 43 liters for each plant during the season, an average daily yield 

 during the period of production of 0.58 liter and a total yield of 87,000 

 liters per hectare per year. 



The numler of plants per hectare. — Mr. E. E. Williams, of the divi- 

 sion of organic chemistry of the Bureau of Science, at my request made 

 a careful estimate of the nmnber of plants per hectare and he writes as 

 follows concerning his investigations : 



The swamps of Bulacan and Pampanga average about 2,000 to 2,250 nipa 

 plants per hectare, including both water and land surface. This estimate is based 

 on actual counts on measured areas, after having explored a hundred kilometers 

 or more of the channels and traversing afoot the interior portions of the swamps 

 in numerous places. 



A strip 1 kilometer long and 10 meters wide through the unworked swamps 

 of Ayala and Company near Consuelo showed 2,170 plants per hectare. The 

 first 500 meters of this strip gave an average of 2,600 plants, while the second 

 tallied only 1,750 per hectare. This difi'erence is due to many low spots and 

 ponds covered or surrounded by gro^vth other than nipa. The entire swamp is 

 foul and not cared for. A similar strip through the Buencamiuo swamps near 

 the iarrio of Song Fo, aggregating an area of over 7,000 square meters, averaged 

 2,410 plants per hectare. This strip also traversed small areas where nipa was 

 largely replaced by mangrove or grasses. A rectangular area of 3,200 meters 

 in the immediate vicinity tallied 2,600 trees per hectare. This plot embraced 

 only clean, well kept swamp, unbroken by water courses. A similar strip 500 

 meters in length through swamps exploited by Ynchausti and Company south of 



