COTTON BALES IX THE BRAZILS. 320 



follows that the outline presented by each cortical mass is that of a 

 pyraniitl, whose sides are formed of a series of rectangular steps like 

 an ordinary staircase. It is difficult to account for this singular ap- 

 pearance, but the more probable explanation seems to be that a layer 

 of wood is deposited for every layer of bark, so that by the time a new 

 deposition of the bark is about to take place the wood has already sur- 

 rounded the extremity of the previous plate, in consequence of which 

 the progress of the new plate inward is barred by the previous season's 

 layer of wood. If this explanation be sound the number of cortical 

 plates on one side of, and including the primitive cortical ray, in- 

 dicates the age of the stem under examination. Each annular layer of 

 wood is thus broken up into four distinct portions by the projecting 

 bark, each portion filling up one of the spaces enclosed by two of the 

 arms of the cross. The number of plates formed on each side of the 

 four primitive cortical rays rarely exceeds six. The peculiarity of ar- 

 rangement, to which I here draw attention, is so striking, that it is a 

 matter of surprise to see this feature so badly represented in Gaudi- 

 chaud's plates; it is fairly drawn by Schleiden, in his 'Principles of 

 Botany' ())p. 251-252 of the English translation), but a better figure 

 is given by Dnchartre, in his ' Elements de Botanique,' p. 167. 



Another arrangement of the cortical portion is also common. It 

 commences, as in the last method, with tlie projection of four slender 

 rays into the midst of the woody fibres, reaching about halfway to the 

 pith ; but the next additions which take place are not found by the side 

 of the four primitive rays, as in the first-noticed arrangement, but occur 

 as four new projections placed exactly midway between tiie first four, so 

 that the stem now exhibits eight of these rays arranged like the spokes 

 of a cart-wheel. At first, the four secondary rnys are very much shorter 

 in length than the four primitive rays, but as the stem increases in age, 

 all the eight rays become of equal lcn<ith. Even in this type some spe- 

 cies exhibit an approach to the first type, by some of the primitive ra\ s 

 in the older stems having one or two lateral plates lying alongside theui. 



Perhaps the most striking form of all the Bic/notiiacere I have hitherto 

 examined is one which unites the peculiarities of both the preceding 

 arrangements, carried to such an excess that the cortical portion at last 

 forms one-half the bulk of the stem. Originally, the woody portion is 

 arranged in the form of a cross, the bark filling up the whole space en- 

 closed by till' four arms of the cro-^s. As the stem increases in diameter, 



