XXXV111 PROCEEDINGS OF THE ROYAL HORTICULTURAL SOCIETY. 



Their arguments are in many cases carefully thought out, but they 

 have failed to carry conviction upon some of the more important points ; 

 and I seek in this contribution to show why I have still cause to doubt 

 the conclusions they arrived at. 



The scientific world of to-day accepts the belief that all organic life had 

 been evolved from one, or at least very few, ancestral types or stocks, and 

 it is of great importance to trace back, correctly if possible, these primary 

 distinctions in germination. This is as near the fountain-head as 

 we may go, and from whence the most valuable deductions may be 

 drawn. 



We may take it that, in comparative anatomy, function is a safer 

 guide than locality. That is to say that when we seek to trace the 

 occurrence of one organ in many different complex organisms, we should 

 not rely upon superficial appearance, mere locality or contiguity, but 

 should search for the function performed in each case. The organ per- 

 forming a similar function in each case is that which we seek. 



In short, the corresponding organ in a number of widely differing 

 complex organisms is not the one which happens to occur in the same 

 relative position of contiguity, but it is the one which performs the 

 corresponding function. This may be contiguous in one case and widely 

 removed in another. 



I am quite prepared to admit the existence of many apparent exceptions 

 or contradictions to this rule. In simpler forms of life organs are few, and 

 each organ serves a variety of functions. In more complex life organs 

 are more numerous and more specialised, but are in most cases able to 

 perform not only their special function, but also (to some extent at least) 

 the more general functions they performed in the long-distant past. 



I do not therefore wish to place too much importance upon any 

 deduction I have drawn from the function of the original process 

 emanating from the seed of certain Amaryllids, yet we must not alto- 

 gether overlook it, for it would appear as though the problem presented 

 to many authors dealing with seed germination had not been to discover 

 the sequence of events in the process of germination, but rather, having 

 decided beforehand that certain organs must exist, to find them. 



To such persons it was impossible that any difficulty should arise. 

 The contiguity of the cotyledons to the "lead" in the dicotyledonous 

 orders pointed unmistakably to the contiguous process in the Amaryllidexe 

 being in that case the cotyledon. 



But when we consider the function served by the cotyledons in 

 dicotyledonous plants and then seek for some organ performing a corre- 

 sponding function in these Amaryllids, we shall certainly not find it in 

 this original process issuing from the embryo. 



The cotyledons, or seed-leaves, have, as their main function, the 

 support of the embryo (and subsequently of the growing point) from the 

 moment of parturition until the process of weaning is perfected. 



During this, sometimes prolonged, period, the tissue of the cotyledon 

 becomes gradually wasted or atrophied. Shortly after weaning the tissue 

 of the cotyledon dies. Its function is completed, and the plant is now 

 drawing sustenance from its own roots. 



For purposes of experiment, or demonstration of the processes of 



