70 
Transactions of the Roydl Society of South Africa. 
jouag seed of G. Chiemon were provisionally interpreted as vestiges of the 
bundle supply of an aborted ring of male flowers.'' 
(7) In an injured spike, one of the female flowers of the nodal ring was 
replaced by a spike. f 
Evidence of a more convincing character supporting the same con- 
clusion is furnished by an abnormal female flower of a Grnetum identified 
as G. scandens, described by MM. Lignier and Tison.J This flower possessed 
four envelopes. § A transverse section shows in addition a group of normal 
A 
: Diagrammatic transverse section of an abnormal female flower of G. scandem 
(after Lignier and Tison). J^.fif. = embryo sac ; iV=nucellus ; I, II, III, IV = 
successive envelopes; Zo. = lateral lobe of envelope I ; 3f=male flowers; 
?i=ihairs. 
male flowers (M) with hairs {h) of the. usual type between envelopes I-II; 
and groups of similar hairs, without flowei^s, between envelopes II-III aiid 
III^IY. Envelope T is without vascular suj)ply ; it terminates above in a 
* Berridge, 1912, p. 990. This interpretation is not accepted by Lignier and 
Tison (1913, p. 69). 
.t Pearson, loc. cU., pi. 25, fig. 7. , 
X Lignier and Tison, 1913. I am indebted to the courtesy of the authors for the 
only copy of this pap&r, a vail able here. Of the date of its arrival I have no record. 
I do not doubt that it was in my possession when my own paper (1915, C) was in 
preparation. I reg^ret that I have completely overlooked it until noAv (April, 1916). 
§ Cf. §4 above. 
