Lyon : OBSERVATIONS ON EMBRYOGENY OF XELUMBO. 645 



the embryo-sac, Trecul ('54, 2) a concretion of a homogeneous 

 substance containing numerous small, acicular crystals, and 

 finally Wigand-Dennert ('87) in a monograph on Nehunbnini 

 speciosum wrote : " Es ist unzweifelhaft der Ueberrest von Endo- 

 spermzellen, welche sich in der Hohlung zwischen den Cotyle- 

 donen erhalten und gleichsam auf der Oberflache der Plumula 

 niedergeschlagen und dadurch das Ansehen einer Membran 

 erhalten haben." 



From the above survey of literature it will be seen that mod- 

 ern knowledge of the seed is based on investigations made upon 

 mature or nearly mature fruits. That alone which will deter- 

 mine the correct interpretation of the structures (Y. e., a knowl- 

 edge of their origin) is entirely wanting. It was in recognition 

 of this fact that the present embryological study was undertaken. 

 In a preliminary note, published in Science, the more important 

 conclusions were announced, i. The embryo of Nelumbo is 

 genuinely monocotyledonous in its development. The plumule 

 arises laterally and at first there is but one cotyledon which later 

 bifurcates to form the two fleshy bodies. 2. The membrane 

 surrounding the plumule is, as conjectured by Wigand, a true 

 endosperm arising within the embryo-sac. 



COLLECTION AND METHODS. 



The material for the investigation in hand was collected by 

 the author in August, 1899, and August, 1900, in southeastern 

 Minnesota while working on the botanical survey of that region. 

 There, Nelumbo lutca grows in great luxuriance in the sloughs 

 and bayous of the Mississippi river, the soft muddy bottoms 

 and quiet waters of these bayous affording an ideal habitat for 

 the plant, where it often forms beds many acres in extent to the 

 exclusion of other vegetation. 



In collecting, each carpel was removed from the torus, the 

 lower end cut away and the upper portion with the attached 

 ovule placed in the fixing fluid. One-half and one per cent, 

 chromic acid and chrom-acetic acid were used as killing re- 

 agents. The material was thoroughly washed and passed gradu- 

 ally into 70 per cent, alcohol in which condition it was brought 

 into the laboratory. The ovules were imbedded in paraffine and 

 serial sections cut with a Minot microtome. The sections were 

 stained on the slide, a variety of stains being used ; the photo- 

 micrographs and drawings, however, which accompany the 



