574 V EMBRYOLOGY OF THE LOWER VERTEBRATES APP. 



some particular group say Amphioocus, or the Mammalia with whose 

 structure they happen to be thoroughly familiar, they assume its general 

 organization to be primitive. As a matter of fact it may be assumed with 

 considerable probability that every existing vertebrate is to a certain 

 I'xti'iit a mixture of primitive features and specialized. It is only by 

 careful comparative study that it can be decided which features an 

 probably primitive and it is quite certain that these will not be found all 

 within one group. Consequently speculations based upon the intensive 

 study of one particular group are to be distrusted, though there is always 

 less ground for distrust if the group is one which is recognized for reasons 

 other than embryological, as being on the whole archaic. 



When minute histological details are concerned another qualification 

 which should be possessed by the animal chosen for investigation is large 

 size of its cell units. 



The material should be abundant. Not only should there be a con- 

 tinuous series of stages but there should be numerous specimens of each 

 stage. There is no such thing as an absolutely normal individual : the 

 conception " normal " is an abstraction based upon the observation of 

 numerous individuals. Only by observing numerous individuals can we 

 therefore arrive at a knowledge of normal development. Work carried out 

 on a few specimens may of course provide isolated observations of much 

 interest and value but it is inadequate to serve as a basis for general 

 conclusions. 



In all descriptive embryology it is necessary to have some method of 

 specifying the stage of development of individual embryos. Unfortunately 

 there has been a great lack of uniformity as to the particular method of 

 doing this. One of the most frequently used is that of specifying the 

 period of time during which development has been going on as for example 

 a "chick embryo of 40 hours' incubation." This method is quite un- 

 satisfactory, owing to the fact that the actual stage of development of any 

 individual embryo is a function of other factors in addition to mere time, 

 such as temperature and individual idiosyncrasy. Thus in many tropical 

 freshwater animals a statement of the age of the embryo is practically 

 worthless unless accompanied by a record of the temperature, and even 

 then there remains the unknown element of individual peculiarity such as 

 is for example illustrated by Fig. 251 where a number of sister eggs of a Frog 

 are seen to have " lost step " with one another to a marked extent even 

 at a comparatively early stage of development. In other words 

 or embryos of the same age are liable to vary greatly in their de-ree 

 of development, and a statement of their age is not adequate as a precise 

 indication of tin- stage of development. The want of precision varies in 

 dillerent cases: it is less for example in a Kul herian mammal where 

 development take- |)l;iee ;it a fa i rl y < I' .'li 1 1 i t e t einj irrat lire t han it is in a Kish 

 or Amphibian inhabiting a tropical pool or swamp where the tern pera lure 

 is li;ihle to -ji-.-.-it variation. 



It i- nere^.iry then m referring to particular stages of development t<> 

 define them by structural features. Here however a new difficulty presents 

 in the I'aet that ! he relative rate of development of ditlerent 

 Of i- not the lame in dill'erent individuals. It follows that if a 

 number nf indi viduals he grouped together as hem- at tin- same st:. 



judged b\ a particular orpin A it will he found that other 



