THE EMBRYO 343 



has been used to support this theory. But this anatomical structure is 

 found also in many dicotyledons and is characteristic of cotyledons in 

 other seed plants. On the existence of a paired vascular system in the 

 cotyledon of many monocot}4edons, the relationship of cotyledon types 

 in the angiosperms has been read in two directions: the single cotyledon 

 has been considered a double structure, the result of the fusion of two; 

 the two cotyledons have been considered formed by the splitting of the 

 single cotyledon. Here also, the prevalence of a double vascular system 

 in cotyledons of both major groups of angiosperms and of some gymno- 

 sperms renders both theories untenable. 



The origin of the solitary cotyledon by the suppression or loss of one 

 of a pair is supported by much evidence. All stages of reduction in size 

 and loss of function of one cotyledon are present in the dicotyledons. In 

 many taxa, there is some difference in size in the cotyledons, and, in a 

 few taxa, the differences reach an extreme, with one nearly or wholly 

 lost — Cyclotnen, Ranunculus Ficaria, species of Conjdalis, Claijfonia, 

 Peperomia (Fig. 136). Stages in the specialization of the one-cotyledon 

 embryo are also found in some monocotyledons, where differentiation in 

 form and function occurs in a pair of cot\'ledons. The tip of one of the 

 pair may develop as a suctorial organ — TriUium, Arisaema, Arum. A 

 second cotyledon is present in vestigial form in primitive genera in the 

 Dioscoreaceae and Commelinaceae — Tamus, Dioscorea, Commelina, 

 Tinantia (Fig. 125). 



Ontogeny likewise seems to show evidence that the solitary cotyledon 

 has been derived from two by the loss of one. During early differentia- 

 tion of organs in the embryo of many taxa, a ridge appears around one 

 end of the flattened proembryo. On one side of this ridge, the cotyledon 

 develops; on the other, growth ceases early. In the dicotyledons, the 

 two cotyledons develop on opposite sides of similar annular ridges. 



The solitary cotyledon undoubtedly represents one of an ancestral 

 pair; the other has been gradually reduced and lost. It should be 

 remembered that, in the period when the solitary cotyledon was con- 

 sidered the ancestral condition, the monocotyledons were generally 

 accepted as the primitive angiosperms. This concept doubtless formed, 

 at that time, the background for theories of the relationship of the 

 cotyledons of the two major groups. 



Variations in Cotyledon Number. Within the dicotyledons, number of 

 cotyledons varies in some genera: Centranthus has three, two, or one; 

 Calendula, Dimorphofheca, Ambrosia, Impatient, Raphanus, two or 

 one. In these genera, the solitary cotyledon occurs only sporadically, and 

 a vestige of the second cotyledon may be present. A third cotyledon is 

 frequently present in many common genera — Acer, Jnghns. Polycotyle- 

 dony has been claimed to exist in some dicotyledons, especially in 

 parasitic genera — Loranthus, Nuijtsia, Persoonia. The cotyledons of the 



