318 CALIFORNIA ACADEMY OF SCIENCES. [Proc. 3D Ser. 



which become very much swollen, and form two caps, one 

 above the other. The upper part of the nucellus is also 

 included, but the cap is principally derived from the integ- 

 uments (fig. 59). The margins of the inner integument 

 are prolonged upward into a conical beak which fits closely 

 into the opening in the outer integument. The whole 

 structure is enclosed in the space at the top of the ovary 

 which remains as a conical cavity in the pericarp. 



The fruit of Sfarganium is not very unlike the caryopsis 

 of the grasses, but differs in the persistence of the integu- 

 ments and probably also of a portion of the nucellus. By a 

 further compression and disorganization of these a typical 

 caryopsis would be formed. 



VI. Abnormal Embryo-sacs. 



In Sfarganium, as in most other low Monocotyledons 

 hitherto examined, there are occasional deviations from the 

 normal development in the embryo-sac. Some of these 

 have already been noted by the writer (Campbell, 1897) in 

 the case of S. eurycarfium ; indeed, it was these abnormali- 

 ties which directed attention to Sfarganmrn. In both S. 

 simflex and S. Greenii similar deviations from the type 

 were noted. 



The most usual form is a multiplication of the nuclei 

 within the embryo-sac without any noticeable increase in 

 its size. As in such cases (fig. 68) the embryo cannot be 

 detected, and the egg-apparatus has either not been formed, 

 or has disappeared, the most probable explanation is that 

 the sac had not been fertilized, but that the vegetative tissue 

 of the gametophyte, i. e., the antipodal cells and the endo- 

 sperm, are capable of a limited growth. This is quite com- 

 prehensible if Sfarganium is, as it seems to be, a very low 

 type of Angiosperm, and the case might be very well com- 

 pared to the limited growth of the prothallium in the hetero- 

 sporous Pteridophytes when fertilization is not effected. 



Other abnormal embryo-sacs are shown in figs. 65 and 

 69. In the first case, taken from an ovule of S. simplex ■, 

 the embryo-sac was very broad and divided longitudinally 



