GENERAL OBSERVATIONS. 17 



II. General Observations. 



1. Method of Determining Oethostichies. 



As the difficulty of determining a vertical series increases with 

 the length of the internodes, and may be complicated by possible 

 torsion and irregular growth curvatures, a plant may be selected in 

 which growth and subsequent displacements are obviously at a 

 minimum. 



A typical " rosette-plant," such as Sempervivum, affords suitable 

 material : the leaves develop symmetrically and retain the positions 

 impressed on them in the bud. 



The plant presents, as a whole, the usual curved systems of inter- 

 secting parastichies, the most obvious being five in one direction and 

 eight in the other (fig. 2 ; also fig. 4). 



By a horizontal cut, the whole plant may be reduced to one 

 plane, and, commencing at the centre, the leaves may be readily 

 numbered in serial order of development from one to about fifty 

 (fig. 3). A vertical row of leaves, or an orthostichy, should clearly 

 appear in the section as a radius passing through the centre of the 

 leaves, differing in number by the denominator of the fractional 

 expression. The leaves 1, 6, 11, 16, etc., form a very obvious curve, 

 so that I is rejected ; the leaves 1, 9, 17, 25, etc., form an equally 

 obvious curve in the opposite direction, hence | is rejected. For 

 the same reason, 1, 14, 27, 40 form a lesser curve, but still distinct 

 enough to disqualify ■^. In 1, 22, 43 the curve hardly appears, 

 and 1, 35 is possibly straighter, but further investigation is limited 

 by the number of leaves. 



A comparison of the closer and closer proximity of these curves 



