1953} 



BOTANY OF THE GUAYANA HIGHLAND 



137 



Salpinga maguirei Gl. sp. now. 



Caulis herbaceus 4-angulatus fere 4-alatus, densissime brunneo-lepidotus; 

 folia petiolata ovata membranacea subacuminata; spica secunda longe peduncu- 

 lata, floribus subsessilibus. 



Stems herbaceous, up to 3 dm. long, densely covered with pale brown, sessile, 

 peltate, circular, scarious scales about 0.7 mm. wide; similar scales occur on 

 the petioles and very sparsely on the primary veins. Leaves very thin, ovate, up 

 to 75 mm. long by 48 mm. wide, subacuminate, obscurely crenate and with a short 

 bristle in each sinus, at base rounded or subcordate to a triangular petiole- 

 summit, 3-5-nerved, the nerves narrowly 2-winged beneath, naked at base of the 

 blade, glabrous on both sides, beneath very minutely white-pustulate. Peduncle 

 3-5 cm. long, pale yellowish-gray, above glabrous, at base with a very few 

 scales. Flowers as many as 7, 5-merous, apparently sessile but actually on a 10- 

 winged pedicel about 2 mm. long and merging gradually into the hypanthium. Ky- 

 panthium narrowly obconic s glabrous, pale, about 5 mm. long, conspicuously but 

 obtusely 10-winged c Calyx-tube prolonged about 0.5 mm.; sepals very thin, spread- 

 ing, triangular-acuminate, 4.7 mm. long from the torus. Petals pink, at least 5 

 mm. long. Stamens dimorphic; filaments flat, glabrous; anthers slender, the larger 

 about 5 mm. long, the smaller about half as large; connective prolonged at base 

 into an antrorse obtuse appendage about 1.2 mm. long and a very minute, retrorse, 

 flattened scale. 



TYPE: from wet cliffs on the escarpment of Cerro Sipapo, altitude 900-1200 

 m., Basse tt Maguire & Louis Politi 27503; New York Botanical Garden. Dis- 

 tinguished immediately from the few other known species of the genus by its 

 lepidote indument. 



In respect to its high degree of endemism, the genus Macrocentrum is without 

 doubt remarkable. For about a century it was known by three species only, and of 

 these one was represented by a single specimen collected by Appun on his trip to 

 Roraima. Within the last three decades exploration has penetrated far into the 

 mountains of southern Venezuela, British Guiana, and Surinam, Appun's species 

 has again been collected and no fewer than eleven others have been discovered. 

 Most of them grow on wet rocks in the immediate vicinity of waterfalls. At pres- 

 ent it would seem that every mountain explored has its endemic species on its 

 own numerous cataracts, but, as a matter of fact, various species are now be- 

 ing found on a second, or even a third mountain, although additional local 

 species continue to appear. The few species of the Surinam mountains are still 

 unknown in British Guiana or southern Venezuela. Dr. Maguire collected four 

 species on Cerro Sipapo, one of which was previously known from Duida, while 

 three were undescribed. Two of these have a habit totally unlike any other 

 known species, so unlike that I was unable to refer them to a genus on my first 

 inspection. When the Duida species were first described, I drew up a key to 

 distinguish all the species known at that time. A similar key to the fourteen 

 species now known is presented below and reference to it is suggested for the 

 diagnostic characters and supposed relationships of the species. It seems that 

 the number of parts in the flower may be of little taxonomic significance, inas- 

 much as Dr. Maguire found 4-merous and 5-merous flowers in the same colony. 

 The key below is based primarily on the three very different types of inflores- 

 cence found in the genus. 



Macrocentrum angustifolium Gl., sp. nov. 



Frutex humilis; folia isomorpha, anguste elliptica, incurvo-denticulata, glabra; 

 flores 5-meri, breviter pedicellati in axillis foliorum superiorum, ad basim a jugo 



