59 



in youth and would find more moisture and better soil than in 

 pure o-rass areas. The haole koa would prevent their overtopping 

 by the grass and would also stimulate them to rapid height 

 growth. 



The area in the Honolulu Watershed Forest Reserve, chosen 

 for the experiment is situated on a gently sloping flat-topped ridge 

 facing the north-w^est and exposed to strong trade winds at an 

 elevation of 760 feet above sea-level, 3.5 miles from the Honolulu 

 Post Office and about 450 feet south-east of No. 2 reservoir in 

 Nuuanu Valley, where the average annual rainfall is approxi- 

 mately, 110 inches. 



The area was covered by a pure stand of Hilo grass, matted 

 thick and growing very close and rank and about two feet high. 

 All vestige of former forest cover had disappeared but in the 

 neighboring gulches and on the mountain sides above are found 

 the native koa {Acacia koa) and ahakea (Bobea- elatior) on the 

 well-drained soil, the kukui or candlenut tree {Alcurifes inoliic- 

 cana) in the gulch bottoms, and the introduced waiawi guava 

 {Psidium Guayaza pyrifentm ). The smaller plants in the region 

 are the common introduced guava {Psidium Guayava) , naupaka 

 {ScaevoJa Chaiuissouiaua) and the ti plant {Cordyliiic fcniii- 

 nalis ) . 



Nine plots, each 20.8 feet on a side and containing one-hun- 

 dredth of an acre each, separated by a three-foot path, were 

 laid out on the ground, as shown in the accompanying diagram. 

 and marked at the corners with 2'' x 2" x 4' painted white stakes. 

 The grass on plots 1, 2, and 3 was left as it was found. The grass 

 on plots 4, 5. and 6 was burned. The grass on plots 7, 8, and 9 

 was burned and the ground then dug lip with mattocks, thus 

 receiving the same treatment as if plowed. 



After the plots had been thus treated, seed of the haole koa. 

 picked two weeks previously in Makiki \'alley. was. on March 

 17, 1921, sown broadcast on the plots in the following manner. 

 On plots 1, 4, and 7, 1.6 ounce of seed was sown on each or at the 

 rate of 10 pounds per acre. On plots 2, 5, and 8, 3.2 ounces of 

 seed was sown on each or at the rate of 20 pounds per acre. On 

 plots 3, 6, and 9, 4.8 ounces of seed was sown on each or at the 

 rate of 30 pounds per acre. 



After the seed was sow^n on plots 1. 2, and 3, the grass was 

 trampled down. On the day following the sowing of the seed 

 there came a soaking rain which lasted two days. 



The purpose of the experiment is to determine : 



1. Whether the haole koa wall germinate and establish itself 

 if the seed is merely sow^i on Hilo grass areas. 



2. Whether seed sown broadcast on Hilo grass areas which 

 have been burned w^ill readily germinate and form a forest. 



3. Whether a better stand wih be secured if the grass is 

 burned and the ground plowed. 



4. WHiich amount of seed. 10, 20, or 30 pounds per acre, gives 

 the most satisfactorv results. 



