1908] COULTER— MEG ASPO RES AND EMBRYO SACS 365 



directly as the megaspore," as in Peperomia, and in the majority of 

 embryo sacs there are eight nuclei (the normal number of nuclei, but 

 an elimination of two divisions); but sometimes there are "twelve 

 or more nuclei" (meaning another division, but still one less than the 

 usual number). 



The situation among the Araceae deserves attention, for great 

 irregularities in the embryo sac have been observed. 9 In Aglaonema 

 the nuclei are four to twelve; while in Nephthytis they are reported 

 as ranging in number from two to twelve or thirteen. Unfortunately, 

 in these cases we are not sure of the behavior of the mother cell ; but 

 so far as the records for Araceae go, the mother cell divides once or 

 not at all. It is needless to draw any conclusions as to the amount 

 of reduction involved in the four nuclei of Aglaonema or the two 

 nuclei of Nephthytis until we know something of their history. At all 

 events, there is a reduction of the ordinary number of divisions in these 

 forms; and this is probably true even in the cases where twelve or 

 thirteen nuclei appear. 



The only case reported, in which there remains some evidence 

 that an additional free nuclear division may occur, is that of the 

 Penaeaceae, a small African group of shrubby xerophytes restricted 

 to southwestern Cape Colony. In a preliminary note, Stephens 10 

 has announced his results with five species, representing three genera 

 (Sarcocolla, Penaea, and Brachysiphon) , and his account applies to 

 all of them. There are sixteen free nuclei formed in the sac before 

 fertilization, which are said to be derived from the innermost one of 

 a row of three megaspores. In such a case three "megaspores" are 

 as good as four, except that the outermost cell is not a megaspore. 

 The testimony breaks down, however, at the critical stage, for the 

 investigator is not certain as to the "row of three." Of course if 

 it should turn out to be a "row of two," the megaspore condition 

 would be that of Cypripedium, followed by the usual number of 

 divisions. The behavior of the nuclei, however, suggests strongly 



9 Campbell, Douglas H., Studies on the Araceae. The embryo sac and embryo 

 of Aglaonema and Spathicarpa. Annals of Botany 17:665-687. pis. 30-32. 1903. 



Studies in the Araceae. III. Annals of Botany 19:329-349. pis. 14-17. 



1905. 



10 Stephens, E. L., A preliminary note on the embryo sac of certain Penaeaceae. 

 Annals of Botany 22:329, 330. 1908. 



