178 H. B. GUPPY^ M.B., F.R.S.E.^ OX PLANT-DISTRIBUTION 



the four closely connected faniilies, the ZingiberacCcT, the 

 Cannace^e, the JMarantacese, and the Musaceie, which behave as 

 tribes of the great plant-group of the Scitaminete, once 

 distrilnited (as assumed by the tlieory) as a generalised type 

 over the Avarm regions of the earth and now represented by its 

 original tribes as separate " families " in the different parts of 

 its area.* 



As another illustration of the working of the differentiation 

 theory, I will take a family like that of the Pandanacere, that is 

 restricted to the warm regions of the Old World, and displays 

 but slight relationship to other families, excepting, perhaps, to 

 the Sparganiace;e of temperate latitudes. Here we would suppose 

 that the differentiation of the original world- ranging type has 

 advanced so far that the type has been lost seemingly beyond 

 recognition. Until we can discover the representatives of the 

 Pandanaceie in tropical America we can only frame guesses as 

 to the original type. That they exist there w^e are compelled 

 to assume, but the primitive cliaracters have been obscured in 

 the differentiating process. 



It will be thus perceived that the differentiating process has 

 been by no means uniform in its results, and we will now 

 proceed to look a little more closely into its working. 



On the Determining Causes of the Irregular PcATe of 

 THE Differentiation of Primitive Plant-Types. 



It is assumed that the differentiation of plant-forms is a 

 response to the secular differentiation of the conditions of 

 existence, l)eginning with a time when uniform conditions and 

 uiidirrtMeiitiated ly])es prevailed. This, however, will be 

 disLMi>-SL'd when we (leal with the physical side of the theory. 

 Here we are especially concerned with the unequal rate of the 

 change. 



It would not be possible to frame a scale connecting the 

 degree of diflerentiation of a family with its relative antiquity, 

 for the sufficient reason that there is no indication of any 

 uniformity in the rate of the process. As far as the geological 

 record can at present guide us, we seem to be justified in 

 assuming an equally great antiquity for all primitive 

 phanerogamous types. But how great is the contrast in the 



The Zii)giberace;e are in;tiiil y Old World, the Caunace:e and Maran- 

 tace;e are mainly Aiiieiican, w hilst the Musacea' are fairly well shared 

 between the two hemispheres. 



