362 ANNUAL REPORT SMITHSONIAN INSTITUTION,, 1942 



places in Brazil, and this was one of them. Instead of the usual moist, 

 humid forest on the mountainsides, we found a dry, dusty jungle of 

 bamboos and dwarf trees through which for hours we hacked our way 

 with sharp facao. It was here that we found Deuterocohnia Meziana, 

 that unique bromeliad whose 5- to 7-foot flower stem continues to 

 bloom for years from the same stalk. It grew as well on limestone 

 rocks overhanging the Paraguay River as on the manganese rocks. 

 Unlike most of the bromeliads, this plant is caulescent. I have seen 

 overhanging the high rocky ledges specimens probably 50 years old, 

 with large, ridged trunks that gave them the appearance of prostrate 

 yuccas. 



In evolutionary development Deuterocohnia is so close to Puya that 

 it seems to be but an advanced form of that genus "distinguished by 

 the advanced characters of appendaged petals and woody habit." ^ 



In central Brazil, on the edge of the high plateau in the state of 

 Minas Geraes, we found other terrestrials that were relatives of the 

 pineapple and close to the primitive form of Puya. This is a section 

 of mines — ^gold, iron, and diamond. In every direction we could see 

 the effect of the vast deposits of ores in the soil, and at evening the 

 purple haze, mingled with red and yellow glints from the sun, made a 

 glowing spectacle radiating earth colors seldom seen outside of a 

 mining district. 



It was in this section that Glaziou, the French botanist who spent 

 the latter part of his life in Brazil, did considerable collecting. He 

 was a bromeliad enthusiast, and in his years of collecting he discovered 

 some 65 new species in this family, a greater number than any other 

 collector had ever found. It was interesting to find many of his 

 species, and in a number of cases our specimens were the first found 

 since the type was named. It was a keen satisfaction to be able to 

 collect two new Dyckias and three new Vriesias in the rather arid 

 rocky areas of Minas Geraes (with promise of still other undescribed 

 species in the material obtained). 



In most of these rocky areas one would expect to find cacti, as in 

 Mexico, but in parts of Brazil most of the soil is acid, whereas the 

 regions of Mexico in which cacti thrive are alkaline. So in Minas 

 Geraes cacti were the exception rather than the rule. We seldom found 

 bromeliads and cacti together. 



In southern Brazil our collecting was confined to Parana, where 

 Dyckia enchoHAoides was typical of the primitive terrestrials near 

 Puya. Unlike other Dyckias, it grew on bare granite rocks on the 

 Atlantic coast at as low an altitude as 6 feet above sea level. Its 

 species name indicates how close are the Dyckias and Encholiriums. 

 Dyckia cncholiHoides is one of the few Dyckias that have developed 



* Smith, L. B., Geographical evidpnce on the lines of evolution in the Bromeliaeeae. 

 Sonderdr. Hot. Jahrb., Bd. 66, Heft 4, p. 460, 1934. 



