IN NATURAL CLASSIFICATION. 589 
many pores; when translucent, we find but three openings. 
Lastly, the observation of neighbouring genera will often 
enable us to decide; thus Geranium commonly bursts by 
three pores, but in Pelargonium we find many. When the 
pollen of this last genus is placed in water, the intine, be- 
coming swollen by absorption, protrudes under the form of 
numerous vesicles, through the openings of the exine, but 
shortly some two or three become much enlarged, the others 
disappear, and the large ones (if acid be added) becoming 
Opaque, eventually burst, permitting the escape of the fo- 
villa. Sometimes pollen with indefinite, and other with 
definite orifices, are found in the same family, as in the Ra- 
nunculacee ; and it is remarkable, that the Alismaceæe, so 
similar in many respects to Ranunculacee, should agree with 
it in the mode of dehiscence of its pollen, while it differs, in 
the same respect, from all other monocotyledonous plants. 
We have now considered the pollen in its most remarkable 
points of view. We have found that its most important 
modes are those connected with dehiscence, complexity or 
simplicity, opacity or translucency of eine, and figure. 
€ have not taken into account the curious modifications 
Which the appendages of the eine present. ‘The variously 
Organized spines and reticulations on the surfaces of 
pollen, so beautifully depicted by Fritzche and Móhl, re- 
quire an exceedingly high magnifying power for their exami- 
nation, and are useless for the purposes of classification. 
The characters we have studied, have not been found inva- 
riable: on the contrary, they are each subject to exceptions. 
Are we therefore to discard them from Taxonomy? No, for 
in this respect, the pollen does not differ from other organs. 
If we examine the different characters hitherto employed in 
classification, we shall see them all liable to variation. Thus, 
amongst plants classed as embryonate, the Rhizanthee have 
no embryo, and Caladium has not, more than the nucleus of 
Chara, a distinct point of germination. If we divide em- 
bryonate plants by the number of their cotyledons, we 
nd wheat with two cotyledons, amongst Monocotyle- 
