6\2 SECONDARF CHAXGES. 



paragraphs, the rings arc here also on the whole concentric, but with the same 

 irregularities as have been repeatedly mentioned above. Usually the different 

 segments of one ring are of unequal strength. Here and there the margin of a 

 segment curves towards the next inner ring, and becomes continuous with it. The 

 successive zones are therefore in direct connection, in not exactly defined longitudinal 

 bands. The thickness which the successive rings attain decreases the further they 

 are from the middle ; they thus remain narrow in comparison with the cortex and 

 pith. For instance, in Miquel's old stem, which has been repeatedly mentioned, the 

 whole thickness of the 6-8 rings is hardly one-third of the radius of transverse 

 section ; the absolute thickness of the single rings varies between o-^<"^ and o-2<"^. 

 A different relation is found at the inverted-conical base of stems which are derived 

 from lateral buds. A piece of Cycas revoluta before me has, for instance, at the 

 bottom a radius of transverse section of about 2 2i">», of which 5™™ are referable to 

 pith, five to the cortex, and twelve to the four almost equally strong rings of wood ; 

 130™™ higher up in a radius of transverse section of 29i»"i, 15™^ are referable to the 

 pith, nine to the cortex, and only five to the two rings of wood, of which the outer is 

 just beginning to appear. 



Both the genera with successively renewed rings have, in the second place, in 

 addition to those described a further system of accessory bundles, which is in Cjycas 

 a cortical system, in Eiicephalarios medullary. 



The latter consists in the mature stem of numerous bundles distributed through 

 the whole pith ; these run in an undulating longitudinal course, and are connected 

 one with another, and with the inner surface of the ring of wood by branches in all 

 directions. They form an elaborate plexus, which is peculiar to the pith and gives off no 

 branches into the cortex. According to Mettenius the bundles appear at a late stage, 

 since they were not present in the whole upper half of a stem of E. horridus of the size 

 of the fist, which he investigated. They may therefore, for the present, be ascribed 

 to the category of cauline medullary bundles, which were described in other cases on 

 p. 253. The medullary bundles are collateral; their xylem, according to Mettenius, 

 never has spiral or annular tracheides, it has (according to incomplete investigation 

 of E. Caffer) fundamentally the structure of the secondary strands of wood of the 

 rings of the same plant, and seems to have a long continued, though very slow, 

 growth in thickness. The stronger bundles in the stem of the last-named species 

 are about i"i"^ in thickness. 



The accessory cortical system of Cycas, which jNIiquel found in C. circinalis, 

 and INIettenius in C. revoluta, arises, according to the description of the latter, from 

 strands of secondary meristem in the cortical parenchyma. Longitudinal rows of 

 cells of the latter ' undergo a division into smaller cells, and become transformed into 

 cambial strands, which gradually develope into small masses of wood.' It appears as 

 though, as the stem grows old, new bundles of this order may be formed successively 

 for a time ; at least they were found in small numbers in young individuals, but in 

 old ones in large numbers. Still this difference may be purely individual, and all 

 the genetic conditions require still further investigation. 



The arrangement of the bundles in the old stem investigated, in which they 

 were numerous and strongly developed, is described by Mettenius in the main 

 as follows. They are arranged in the transverse section in several irregular 



