30 INVEETEBEATA CHAP. 



will be made. Here we have divergences as great as those which 

 exist between any upholders of rival phylogenetic theories. 



The real truth is that Experimental Embryology is an adjunct 

 and not an alternative to Comparative Embryology. It is a new and 

 refined instrument of dissection : instead, for instance, of separating 

 the blastomeres of a segmented egg by optical differences they are 

 actually separated and their values tested by their powers of develop- 

 ment. But the difference between the Echinoderm egg where any of 

 the first eight blastomeres will develop into a whole larva, and the 

 Annelid egg where the loss of a blastomere means the loss of a portion 

 of the larva, still requires for its explanation the principle of affinity, 

 that is to say that the ultimate explanation of the specific peculiarities 

 of development is found in the chemical nature of the hereditary 

 substance. 



So, before the future student of embryology stretches an almost 

 limitless field of research. We must ultimately find out not only 

 how the chemical quality of the germ-plasm determines the growth of 

 the formless egg into the highly complex adult, but we must also find 

 out howthis chemical quality can be altered, so that variations can occur 

 and evolution can take place, and this is the root-problem of biology. 



In order to make any attempt to solve this root problem we must, 

 however, be able to control the whole life -cycle of the animal 

 experimented on, and this is precisely where the work of Loeb, 

 Herbst, and Driesch breaks down. All these workers have chosen as 

 experimental objects the eggs of Echinoderms. These eggs are 

 produced in enormous numbers and are easy to rear through the first 

 stages of their development, but to reach even the adult form to say 

 nothing of the adult dimensions they have to pass through a 

 prolonged larval life during which there is an enormous mortality, 

 and even when this metamorphoses into the adult shape is success- 

 fully accomplished they have less than the millionth of the bulk of the 

 fully ripe sexual form. Under the most favourable circumstances a 

 year or two must elapse before they could produce germs and, therefore, 

 before it could be possible to say whether the experiments had really 

 altered the hereditary potentialities, or whether the distorted larva 

 is merely the resultant of the new force applied and of the unaltered 

 hereditary potentiality of the germ. The method of rearing these 

 larvae until they attain sexual maturity has now been elucidated, but 

 only a small proportion of the fertilized eggs can so far be reared ; and 

 none of the workers mentioned above have attempted to rear the 

 larvae beyond the earliest stages of their development. 



We pass now to consider the special embryology of the different 

 groups of Invertebrata. In every case we shall, so far as possible, give 

 examples of the fundamental laws of development laid down in this 

 introductory chapter, and we shall indicate also what solid results 

 have been gained from experiments performed on the developing eggs 

 of animals belonging to each group. 



The group Protozoa are excluded, not because they do not 



