217 



p-5 mm. long; style altogether 0*6-0-75 mm. long, tlie undivided 

 part as long as the arms, rarely twice as long; ovules 15-18; 

 leaves long and gradually acuminate, acute, obtuse or rounded 

 at the base. 



Colombia (Smith 789) . Pedicels 7 mm. long ; staminodes 

 0-8 mm. long; style altogether 14-1-3 mm. long, the undivided 

 part two to three times as long as the arms ; ovules 21-23 ; leaves 

 shortly acuminate or cuspidate, obtuse or rounded at the base. 



It will be seen that the Venezuelan plant has shorter staminodes 

 and style than the Cuban or Colombian specimens, and that it 

 usually has the undivided part of the style no longer than the 

 style-arms. In the length of the pedicels, however, it agrees with 

 Cuban specimens, and in the number of ovules it is intermediate 

 between the Cuban and the Colombian. The shape of the leaves is 

 rather variable. It is evident that, on the characters given above, 

 it is undesirable to separate the Cuban, Venezuelan and Colombian 

 plants, even as distinct varieties. 



The Wood. — As the twigs which bore the leaves and flowers 

 were separate from the samples of wood, it was thought advisable 

 to compare the microscopic structure of the twigs with that of 

 the wood-specimens, in case there should have been any mistake 

 on the part of the collector. The result of the comparison is quite 

 satisfactory as regards agreement in structure, and proves that 

 the twigs and wood belong to the same species, or to two closely 

 related species. Under the circumstances they may be accepted 

 as belonging to the same species. 



The specimens of wood referred to are a small block and a 

 section of the stem (or short log with bark) about 12 inches in 

 diameter and 2^ inches thick. The wood is close-grained and 

 yellowish, and bears a considerable resemblance to true boxwood. 

 The presence of a secretion in the inner bark (secondary bast) is 

 indicated in the log by a resinous or oily stain on the transverse 

 surface. The secretion has oozed out of the bark, and has then 

 soaked into the peripheral part of the wood, so that the stain 

 extends a short distance (1-3 mm.) on either side of the cambium. 



The structure of the wood is as follows. The vessels are small, 

 rounded or elliptical (with greater diameter usually less than 

 50 fx)y solitary or in snort radial rows, fairly evenly distributed, 

 and numerous, e.g.y 240 to the square millimeter. The perfora- 

 tions of the vessels are simple. The pits between the vessels and 

 the medullary ray-cells are small and bordered, like those on the 

 other parts of the walls of the vessels. The medullary rays are 

 numerous ; some are uniseriate, but most are multiseriate, being 

 usually three or sometimes four cells broad in their middle region. 

 Many of the multiseriate rays have an upward or downward 

 uniseriate prolongation of considerable length (as seen in 

 tangential section). The cells in the multiseriate portion are 

 mostly rounded, and 10-15 fx in diameter, while the cells of the 

 uniseriate part are quadrangular, and may measure from 25 to 

 60 fx by 12 to 18 //. Solitary crystals of oxalate of lime occur in 

 many of the cells of the uniseriate portions of the rays. The 

 height of some of the medullary rays is rather more than 1 mm. 



H^ 



B 



