Mr. Hassall on the Structure of the Pollen Granule. 565 



orders of dicotyledons the pollen does not indicate any regu- 

 lar gradation of structure, although in many of the higher 

 families it is surprisingly complex. 



So great is the difference between the pollen of Exogens 

 and Endogens, that it alone furnishes a character by which 

 it may be at once determined to which class any plant be- 

 longs. The pollen granule of an endogen may thus be cha- 

 racterized. It is either spherical, oval, or elliptical ; generally, 

 if not always, composed of two membranes, rarely possessing 

 more than one pollen tube, and, w ith a single exception, never 

 more than two. This exception occurs in Limnocharis Hum- 

 boldtii, in which the granule is spherical, and the extine per- 

 forated with six or seven apertures for the passage of the pol- 

 len tubes. The elliptical formation of granule prevails much 

 among monocotyledons, and has been met with in forty-four 

 of the seventy-three genera of Endogens submitted to the mi- 

 croscope. An exogenous pollen granule may be thus defined : 

 it generally presents a more complicated organization ; the 

 number of enveloping membranes is either two, three, or four ; 

 its form is various, being most commonly either three-lobed, 

 spherical, or triangular ; and it is furnished with pollen tubes 

 varying in number, exclusive of three exceptions, from three 

 to upw ards of fifty. Of these forms the three-lobed type is 

 of the most frequent occurrence, and is absolutely character- 

 istic, so far as my experience goes, of an exogen, being found 

 in 187 of the 332 genera of dicotyledons examined by me. 



The exceptions occur in the genera Acanthus (see fig. 44.), 

 Dryandra, and Magnolia, the last a genus so evidently dico- 

 tyledonous as not to admit a doubt of its real nature ; and 

 yet here, strange to say, the granule, so far as has been ob- 

 served, perfectly resembles the elliptical form so prevalent 

 among monocotyledons. (See fig. 150.) The pollen granule 

 of Dryandra, although provided with but two pollen tubes, is 

 curved and has three distinct tunics. (See fig. 93.) 



We come now to the more immediate purpose of this pa- 

 per, viz. to consider how far the pollen granule may be relied 

 on as a means of classification. It has already been said that 

 by it alone a monocotyledon may be distinguished from a di- 

 cotyledon, and hence should be carefully consulted when any 

 doubts are entertained of the position of a plant in the 

 vegetable kingdom. Much has been written upon the posi- 

 tion which the NymphaeacecB occupy in the vegetable kingdom ; 

 some arguing that they ought to be placed among Endogens, 

 to which they are united by so many evident affinities, — 

 others referring them to Exogens ; but the question is as yet 

 an undecided one among botanists, although admitting, in my 



