2 BULLETIN 445, U. S. DEPAETMENT OF AGKICULTUEE. 



orange-growing district of Bahia (PL I), though most of its com- 

 mercial plantations do not date back more than 40 or 45 years. The 

 name of the originator does not appear to be known at the present 

 day, or the exact location of the property on which the variety origi- 

 nated. Only the most fragmentary accounts are given by the orange 

 growers, who should and probably do know more about the subject 

 than most others. The most complete and probably the most ac- 

 curate statement is that furnished by the Rev. W. A. Waddell, a 

 Presbyterian missionary, who has lived for years in the vicinity of 

 Bahia and has been much interested in this subject, as follows: 



Twenty years ago an old man, a very intelligent cabinetmaker, told me that 

 in his youth, before the independence of Brazil, the laranja de umbigo (navel 

 orange) was found only in some groves in Cabulla. He, as a boy soldier, in 

 company with his comrades, " chupou muitas " (ate many) during the siege 

 of Bahia, being stationed in a grove that contained some trees. Most of his 

 comrades had never seen them before, but he had seen them sold by the slaves 

 of a Portuguese. He had heard that a " mandinga " woman charmed a seed 

 and made the first tree yield " umbigoed " fruit. This was information gathered 

 Avhen he was young, say, 1816 or 1818. I came to the conclusion that the 

 seedling tree originated in Cabulla in 1810-1820, or perhaps even earlier, and 

 was first propagated by a Portuguese grower, and that in 1822, the year of 

 Brazilian independence, there was quite a lot of trees. Of course, the produc- 

 tion of any odd-shaped fruit would be explained by fetichism among the lower 

 classes. 



It will be noted that Dr. Waddell speaks of the " seedling tree " 

 which originated in Cabulla. All the evidence, however, indicates 

 that the variety originated as a sport, or mutation, upon a Selecta 

 orange tree, laranja selecta, as it is known in Brazil. The Selecta is 

 almost identical with the navel orange in many characters and fre- 

 quently shows a marked tendency to produce navel fruits, even 

 though it is normally without any vestige of a navel. The Bahians 

 themselves recognize the similarity between these two varieties and 

 call the navel orange " Selecta de umbigo," or navel Selecta. This 

 name may, in fact, have been given to the first navel tree to indicate 

 its origin. 



The Selecta orange, while rarely seen at Bahia, is still cultivated 

 commercially in the vicinity of Rio de Janeiro, especially at Sao 

 Gongalo, a suburb of Nicthero}^ In one of the groves of this section, 

 that of Joao Elias Esteres, the presence of occasional fruits with 

 well-defined navels was observed on trees which normally produced 

 typical Selecta fruits. The navels in these fruits were in same cases 

 as large and well developed as in the typical navel orange, although 

 they did not protrude through an opening in the apical end of the 

 fruit as commonly as in the latter variety. 



The typical Selecta orange (PI. II) is slightly oblate in form and 

 contains 15 to 20 seeds. In bud-sport fruits with navels (PL III), 



