ORIGIN OP METAMEEIO SEGMENTATION. 115 



I do not think any really satisfactory explanation can be 

 offered at present of these facts. I venture, however, to sug- 

 gest the following as an attempt at an explanation. 



In many living Triploblastica the embryo leaves the egg at 

 a very early stage as a larva ; at a stage in which it is little 

 more than a gastrula. Inasmuch as the parent of this ancestor 

 has differentiated nephridia and muscles, &c., it is easily con- 

 ceivable that the larva should precociously acquire as much of 

 these organs as it requires. Hence mesenchyme. This larva 

 is a small animal, and does not require a pouched gut ; its 

 hypoblast becomes specialised for digestion ; now it would 

 obviously hamper these exceedingly active larvae if the gut 

 repeated the phylogeny ; at any rate, it is easily conceivable 

 that it would be more advantageous if it were possible that the 

 digestive cells should not have to undergo active developmental 

 changes. Hence the mesoblast has to be formed in another 

 way. The methods in which it is formed are, as is well known, 

 various ; it nearly always, however, originates at the lips of the 

 blastopore, as the result of the proliferation of a cell, or cells 

 which do nothing else but divide and give origin to the 

 mesoblastic bands. This, as I have suggested above, may be 

 looked upon as a modified development of that of the ancestral 

 archenteron, which became pouched, and gave rise to somites 

 (secondary invagination) . 



In those animals in which this larval phase has become 

 merged in the embryonic development, this process is con- 

 tinued ; but the area from which the major part of the meso- 

 blast arises, i. e. from which the secondary invagination takes 

 place, is larger. This may obviously be explained as being due 

 to the fact that, the development being protected, it is not im- 

 portant that the amount of growing tissue present at any given 

 moment should be as small as possible, in order not to hamper 

 the larva. 



On this view Araphioxus presents a most surprisingly primi- 

 tive development, so far as its somites are concerned. 



I need hardly point out that the prevailing order of develop- 



