I'OKTO KICO EXPERIMENT STATION. 431 



afford siilliciciit siijidc. it lias hccii on the wliolo a failiii'c. Moreover, 

 tlic atiiiospli(M-(' within iKMni-- slio-litly more liuiuid and waiinci-, con- 

 duces to funj^us diseases anion*;- the seedlings. Previous to this sev- 

 eral small palm leaf sheds were used successfully. 



PLANT COLLECTIONS. 



The roll(nviu<;' collections of I'conomic ])lants have i)eeii assembled: 



HAN AN A J'LAT. 



About 3 acn's of wind-sheltered hillside having a western exposure 

 were set aside for the collection of bananas and plantains. This plat 

 is intended for a variet}' test, a source for stock distribution of the 

 more valuable kinds, and as an experiment in methods of planting, 

 cultivation, and fertilizing- (PI. XVIII, lig. 2). In regard to the methods 

 of planting, it has been found that the "name" (shoi-t portion of stem 

 base with corm-like rootstock) was preferable for planting in dry soil; 

 that the '" tallo" (8-foot section of stem with root) gave best results in 

 wet soil, and that the "pichon" (:^-3-foot sucker or offshoot from 

 stem ])ase) was the most convenient for general purposes. The num- 

 ber of offshoots produced bv each of these methods was nearly the same 

 after eight months. Contrary to the popular belief it was indicated 

 that drj'ing the roots in the sun for several days previous to planting 

 was injurious to the vitality of the plant and considerably retarded its 

 sprouting. It is possibl(\ however, that sun-baking may be more or 

 less effective in preventing decay of the root when planted in wet soil. 

 Very few of the "pichones'' failed to start into growth within one 

 month after setting, but about 5 per cent of the roots planted by the 

 other methods rotted. Liberal quantities of wood ashes, phosphate 

 rock, nitrate of soda, rotted coffee pulp, and stable manure were used 

 without appai'ent effect. The plants which have made the best growth 

 thus far are those growing in a heavy red clay which apparently con- 

 tains very little hunuis though nuich moisture. 



An experiment to determine the effect of allowing few and many 

 offshoots from the parent plant to remain is in progress; three years 

 will be required to terminate this experiment. 



Although but five varieties of this most important fruit are com- 

 inoidy offered for sale in the markets of thi' island, an unexpected!}' 

 larg(> number of the following native varieties has been procured from 

 various ])arts of the island: 



Enamt. Coloratle Blanco. IMiitano Ilarton. 



Kiiano I)ol>U'. Tiosa, or Datyl. Plataiio Tn-scientos. 



Chaiiialiico. Dnrninico. Plalnno Kiiami. 



Chamaliiio I'at<». Maii/aiio. I'i;itaiio CiKireiiteno. 



Gnariin, or <ii>jant<'. I'rii'to. Platano Morado. 



Guanin Dr>l)i«'. Inj,'l(''!<. Giiayaho. 



Morado, or Colniudo. Coufjo. Ceiiizo. 



Morado Doble. Congo Morado. 



