442 Scientific Proceedi?)qs, Royal Dublin Society. 



dimensions of the others are 0'35 mm. by 0"016 mm. With this exception the 

 rays of mahoganies vary from 2 mm. in height downwards. The proportions, 

 however, of different sizes within these limits may be very different. Thus 

 in the Khayas the number of small rays approximates to the number of large 

 rays (PI. XXX., fig. 50, PI. XXXL, figs. 56, 59, PL XXXIL, figs. 62, 63), 

 while in the Swietenias (PI. XXII., figs. 2, 5, 6, PL XXIIL, fig. 8) the 

 number of larger rays greatly predominates over that of the smaller ones. 

 In the former consequently the rays, seen in a tangential section, appear of 

 irregular sizes, while in the latter they seem much more uniform in size. 



The marginal cells of the rays of different mahoganies present differences 

 of diagnostic value. Often in radial sections they appear roughly pentagonal, 

 in tangential sections triangular, and usually much deeper and wider, and 

 less extended in a radial direction than the middle cells of the rays 

 (PL XXIIL, fig. 9, PL XXIV., fig. 18, PL XXXIL, fig. 66), which are narrow 

 cylinders or prisms with their axes in a radial direction. On the other hand, 

 the marginal cells of Colombian mahogany (PL XXV., fig. 24), Pterocarpus 

 (PL XXXV., fig. 84, PL XL., fig. 114), and of Entandrophragma (PL XXXIII., 

 fig. 72) and others can scarcely be distinguished in shape from the middle cells 

 of the rays. Where the marginal cells are larger and well differentiated from 

 the middle cells they appear in the cross-section as radial series of large 

 parenchymatous cells, and give the section a very characteristic appearance. 

 Sometimes {e.g. Khaya, PL XXX., fig. 50) a proportion of the large rays have 

 margins several cells deep but only one cell thick. In tangential sections 

 this gives a striking appearance which reminds one of the same shaped rays 

 (rat-tailedj in Walnut. Generally the walls of the marginal cells are no thicker 

 than those of the middle cells ; but in Gaboon mahogany they are strongly 

 thickened and pitted and in radial section approximate to rectangles or 

 rhomboids. Guatemalan mahogany also has marginal cells of this shape 

 (PL XXVIII. , fig. 42). The walls of the middle cells are seldom very thick, 

 but in some of the dense woods they are sufficiently thickened to render the 

 lumen seen in tangential section rounded and so contrast with the angular 

 outline of the middle cells of other wood viewed from the same aspect. 

 Observation of this point is sometimes useful. 



Apparently little reliance can be placed on the presence or absence of 

 septa in the fibres. Thus they occur in some specimens of "Spanish" and 

 Cuban mahoganies, but cannot be detected in others. The same appears to 

 be the case with certain African woods otherwise similar to one another. In 

 certain mahoganies they form a striking feature, e.g. in Carapa guianensis 

 (PL XXIV., figs. 14, 17) and Aucoumea Maineana (PL XXXV., fig. 80). 



The presence or absence of pore-rings, annual rings, and zonation of the 

 fibres has already been commented upon. 



