l8 9 I J Abnormal Phyllotaxy. 163 



The buds should be studied at as early a stage as possible. 

 Fig. 11 represents a plant, as it was seen on Feb. 15th, during 

 a very cold winter. The protecting scales have been removed 

 and the larger leaf forcibly expanded. Fig. 12 shows the 

 petals and stamens, partly enclosed by one sepal, the other 

 having been cut away. Fig. 13 gives a slightly magnified 

 view of the pistil at this period. It is evident from the size 

 already attained by the organs, that flower buds should be 

 studied for phyllotactic purposes during the summer previous 

 to their flowering season. 



Trillium sessile. — In the spring of 1882 a four leaved Tril- 

 lium was found which illustrated admirably the attempt of 

 succeeding whorls to continue or accommodate themselves to 

 the abnormal conditions presented by previous whorls. This 

 case is illustrated by fig. 14; the letters serve to designate the 

 flower members but have no other significance. In the normal 

 development of this plant there should be three sepals and 

 three petals, in decussating whorls. The outer set of floral 

 envelopes, marked a, l? f r, should have been sepals, and the in- 



/ 



Instead of 



that, sepal e occupies the position of a petal, and together with 

 sepal c attempts to attach itself to the whorl of four leaves as 

 though it were an independent dimerous whorl of sepals. 

 This attempt, however, is spoiled by sepal a in its usual posi- 

 tion, but as if to maintain at least the semblance of a dimerous 



/ 



/ 



a and e. Further b, which should be a sepal, is developed as 

 a petal, and only two of the first whorl of stamens are in ex- 

 istence, b and c, the third a being entirely absent. The sec- 

 ond whorl of stamens and the leaves present the normal 

 conditions. If now this case be considered as the attempt of 

 a trimerous plant to accommodate itself to the dimerous con- 

 ditions presented by the leaves, then the absence of a stamen 

 •n the first whorl can be explained. The dimerous whorl of 

 leaves^, A, is succeeded in decussating order by a second 

 similar whorl B, B; and this in a similar way by a, c and c 

 With the adjacent part of /, all acting together as a simple 

 dimerous whorl, decussating with B, B\ and all having a green- 

 ish color. With this in turn the purple petals d and b are 

 found decussating, occupying a position almost directly in line 

 with the leaves, B, B. The trimerous character of the plant 



