466 Holden and Daniels. — Observations on the 
and from that provided by other species to suppose that extremely intimate 
syncotyly may lead to the suppression of all evidence of double origin in 
the composite petiole, and, if carried farther, in the cotyledon lamina also. 
The evidence yielded by the behaviour of cotyledonary buds, though 
different in character, is of no less interest in this connexion. The buds in 
the axils of the cotyledons are not developed very early, and in much of the 
material studied they are either not evident or are doubtfully represented by 
small groups of meristem. In the older syncotylous seedlings an excellent 
series of forms showing progressively closer fusion of the buds has been 
secured. In the simplest type the double origin of the bud is very evident, 
as it consists of a tangentially elongated ovoid mass of meristem which 
bifurcates near the apex (Fig. 82). Other stages show a rather massive bud 
with apical splitting, and others again show what to all appearance is 
a single normal bud (Fig. 35). In addition to this series a solitary specimen 
was discovered among the ‘ Type 2 ’ seedlings which showed an axillary bud 
which was quite evidently of double origin. There was thus in this type of 
seedling indications of the double nature of the cotyledonary bud, whilst the 
‘ Type 1 ’ series shows that, as is the case with the midribs, a fusion so 
intimate may occur that all evidence of dual origin is suppressed. When 
these facts are considered, together with the striking similarity in the modi- 
fications of the epicotyl, the view that the whole of the abnormal seedlings 
must be regarded as syncotylous in origin is materially strengthened. 
Further collateral evidence is provided by a similarity in the behaviour of 
the vascular strands during their passage from the cotyledon to the root in 
the two types of seedling, and also by certain new features in the transition 
region which exhibit a close parallelism. Considerable numbers of modified 
seedlings were obtained which, although their appearance would suggest 
a triarch and diarch condition of the hypocotyl, nevertheless possessed a root- 
whorl with four members (Figs. 3, 4), and therefore presumably four xylem 
poles. An investigation of these revealed the fact that the additional poles 
may arise in one of two ways. One has already been indicated in the 
previous paper dealing with these seedlings ( 18 ), and is due to the combina- 
tion of two features, namely, the growth in size and importance of the 
marginal as distinct from the lateral strands, and the remaining independent 
of the latter so that the marginals alone contribute to the root-pole opposite 
that formed by the midrib. The independent laterals are often continued 
into the hypocotyl as endarch strands, which usually soon lose their proto- 
xylem and ultimately die out. Prior to this they may become mesarch and 
occasionally, near the base of the hypocotyl, exarch, though the position of 
the protoxylem, where persistent, is inconstant. It is from this type of 
persistent lateral strand that a new member of the lateral root-whorl and 
subsequently a root-pole originate. Neither the fact of the protoxylem 
remaining endarch nor its disappearance at a lower level seem to affect 
