31 



Plants with the above qualities should not he topped, but left for 

 seed production. All suckers should be removed and only the crown 

 flower stalk left. It is also advisable to remove none of the leaves, 

 as experiments have shown that the greater the number of leaves 

 left on the seed plant the better tlie quality of seed. Great care 

 should be taken to prevent pollen from neighborin.t;: varieties being 

 carried to the selected plants, and consequent crossing of varieties. 

 In case of doubt in this regard it is best to reject the seed. The 

 results of careful seed selection will be very clearly shown on the 

 field the next j'ear. For types of leaves see Plate Y, figures 1 to 4. 



The Plant Breeding Laboratory of the United States Department 

 of Agriculture has begun selecting and cross-breeding tobacco with 

 reference to improving the leaves, and eighteen selections and crosses 

 in duplicate were sent to the writer to be planted during the winter 

 of 1903-4. One set was planted at Aguas Buenas and the other at 

 Aibonito. At the last-named place the set was destroyed bj' the 

 "leaf miner," or " split worm," after the plants had been set in the 

 field, so no .seeds were secured. (PI. IV, fig. 4.) 



At Aguas Buenas the plants did well. (PI. IV, fig. 3.) Of each 

 selection the best and most uniform plants were selected and 30 to 40 

 capsules only left on the main flower stalks. These were covered 

 with manihi paper bags to prevent crossing with other plants. Seeds 

 of these selections have been saved for further work along this line. 

 It may be said that as a i-esult of this selection the plants have shown 

 great improvement in uniformity of leaf, and the results are very 

 promising. Tiie most impoi'tant characteristics of the mother plants 

 were all transmitted in the seed, as was proved bj' a comparison of 

 each plant with the notes which had been carefully made of the 

 appearance of the mother plant. It is necessary for each tobacco 

 grower to make his own selections in the field, and upon deciding 

 which type or types of plants he wishes to perpetuate to bag such 

 selected plants after the flowering begins. 



SUGGESTIONS FOR THE IMPROVEMENT OF PORTO RICAN 



TOBACCO. 



The investigations carried on during the year are not in all respects 

 conclusive. It is recognized that further work is desirable, and it is 

 hoped that such may be continued in the near future. 



In tobacco culture, as with any other branch of agriculture, it is 

 recognized that practice must conform to the conditions with which 

 the planter has to contend. However, there ai-e certain general prin- 

 ciples that are in all cases applicable. Endeavoring to keep within 

 the bounds of these, the following suggestions are made with refer- 

 ence to improving the methods of growing, curing, and fermenting 

 tobacco in Porto Rico. 



