856 REPORT— 1903. 



Embryological evidence of another kind has recently been brought forward 

 to show that dicotyledons are the elder branch. 



In the following paragraphs this evidence is set forth at some length under 

 two heads : 



(A) The nature and strength of the evidence itself. 



(B) The support which the view it suggests receives from the mature 

 characters defining monocotyledons. 



(A) 1. The anatomy of the seedling soon after germination is often of value in 

 the study of afiinities. 



My own observations have convinced me that within the Liliaceae 



the symmetry of the vascular system in the cotyledon, hypocotyl, 



and primary root, is commonly characteristic of the genus, and is a 



valuable guide to the affinities of genera with each other. 



(A) 2. The vascular symmetry of the single cotyledon in monocotyledons forms 



a sharp contrast with that of the first leaf. The bundles of the first leaf and 



those which succeed it are symmetrical about a true midrib. The bundle-system 



of the cotyledon is symmetrical about a pair of bundles, which may be distinct 



or partially united. 



To this rule there are a few exceptions. But most of the cases in 



which the cotyledon seems to have a midrib can be connected through 



allied species with forms in which the bundle-system is bisymmetrical, 



and further work will probably link the few exceptional forms which 



remain with the double type. 



(A) 3. The bisymmetrical vascular structure of the ' cotyledon ' in monocoty- 



ledonous seedlings may be interpreted as the last trace of the double structure 



which arose from the gradual union of two ancestral cotyledons to form a single 



member. 



This view is supported by : 



(i) Comparative study of the Liliaceae, which shows that — exclud- 

 ing species of exceptional habit — the many divergent schemes of 

 vascular symmetry found among the seedlings of this family can be 

 referred to a single type, in all probability the most ancient examined. 

 In this type the two massive bundles of the cotyledon are quite 

 distinct, and are symmetrically placed near the foci of its elliptical 

 transverse section. 



(ii) Evidence from the Amaryllidacese, Iridacese, and Aroidese 

 which tends to show that the structure of their seedlings is derived 

 from a Liliaceous type. 



(iii) Corroborative evidence from Palmse and Scitaminete to show 

 the general absence of a midrib from the cotyledon. Several herbaceous 

 species of Scitaminese have two distinct and opposite bundles in their 

 cotyledon. 



(iv) Comparison with seedlings of Ranales, and particularly with 

 the species, fairly numerous in this alliance, which have their coty- 

 ledons partially united A single trace from each cotyledon enters 

 the hypocotyl in the Eanal type. Each trace is clearly double, but 

 the root is diarch. In the primitive Liliaceous type the two traces 

 from the cotyledon enter the hypocotyl from opposite sides. Each 

 opens out into a double structure during the transition. The root 

 is tetrarch. Besides the resemblance in dual symmetry, I have per- 

 ceived in Eranthis &nA. Podophyllum ?t, similarity to the Liliaceous type 

 in the method of transition from stem to root, and in Eranthis the 

 temporary appearance of a tetrarch xylem plate at the base of the 

 tuber. The resemblance between the Ranal and Liliaceous type of 

 seedling structure supports our hypothesis whether we consider it as 

 homologous or homoplastic. 



