f 



* 



I 



1 



t 

 * 



1902] OJV THE DEVELOPMENT OF CERTAIN PIPERACEAE 337 



As to the systematic position of Peperomia, the writer sees at 

 present no reason to doubt that it is })roperly placed among the 

 Piperaceae. Whether the Piperales as an order are to be regarded 

 as very primitive forms is not so clear. It is, of course, possible 

 that the absence of floral envelops here is primitive. The pres- 

 ence of pcrispcrm in the seed is a character found in many of 

 \ the simpler orders of Angiosperms, but it has probably been 



* secondarily acquired, since we find nothing in the higher Ptcri- 



dophytes or Gymnosperms to suggest that it is primitive. A 

 survey of the Dicotyledons shows that perisperm is at present 

 known in the seeds of Piperales, Aristolochiales, Polygonales, 

 Centrospermae, and Ranales. All of these orders, as has been 

 pointed out by Schimper and Lesquereux, are old geologically, 

 and possibly represent branches of a single stock or phylum of 

 the Dicotyledons (see Bessey, 1897, P- 33)- ^ '^'^ inclined to 

 believe, as many of the older systematic waiters have held, that 

 the Piperales are much more closely related to the Polygonales 

 than the position assigned them by Engler (1S98) indicates. 



Within the Piperales it is evident that the flowers of Piper and 

 Heckeria are more complex in structure (probably primitively 

 so) than those of Peperomia. This is indicated by the syncar- 

 pous ovary and the presence of two integuments. It is not clear 

 to the writer that Saururus is " distinctly more specialized" than 

 some of these higher Piperaceae, as suggested by Campbell 

 (1901, p. 114), nor that the Saururaceae are derived from the 

 Piperaceae. 



The morphological features of the germination of the seed in 

 Peperomia and Heckeria indicate that the aleurone-containing 

 endosperm of these forms acts as a digesting and absorbing appa- 

 ratus for transferring the starch stored in the perisperm to the 

 embryo. 



In several other genera, of the Cannaceae, Polygonaceae, 

 Phytolaccaceae, Caryophyllaceae, and others, a thin layer of 

 endosperm separates perisperm and embryo and seems to serve 

 the function above mentioned. The embryo sporophyte is per- 

 haps everywhere nourished through the gametophyte and not by 

 the parent sporophyte directly. 



..1 



