No. 3.] AMPHIOXUS AND THE MOSAIC THEORY. 587 



to reach conclusive results on a number of important points 

 which it is to be hoped may be cleared up by other investi- 

 gators. ^ 



A. General Stmwtary. 



I. An isolated ^ blastomere undergoes a cleavage identical 

 with, or approximating to, that of a normal embryo. It pro- 

 duces a normally-formed blastula and gastrula of half the 

 normal size, and finally may give rise to a half-sized dwarf 



1 The eggs should be gently shaken, in a small glass tube about half filled with 

 water, and then poured into a larger vessel of fresh water. Violent shaking, such 

 as is necessary in the case of Echinus, completely disintegrates the blastomeres of 

 Aniphioxics. The blastomeres can easily be isolated and develop well in the 

 hanging drop for about twelve hours. For later stages they should be placed in 

 a small covered vessel not more than half filled with water. 



I have experimented with 2-, 4-, 8-, and i6-celled stages, giving most attention 

 to the 2-celled and 8-celled, with the main object of determining first, the limit of 

 regenerative i^ower, and second, its, form of action as shown in the mode of cleavage 

 of the isolated blastomeres. In weighing the results certain sources of error should 

 be carefully borne in mind. First, normal eggs and embryos vary considerably in 

 size. Second, they do not develop precisely at the same rate, so that it is im- 

 possible to experiment with any stage absolutely unmixed (e.g., a batch of 8-celled 

 embryos almost invariably contains a certain number of 4-celled). Third, a certain 

 number of the eggs always develop abnormally, whether shaken or not. Fourth, 



— perhaps most important of all — the rude op-^iration of shaking very often causes 

 the isolated blastomeres, as well as the U7iseg7nented ova, to break into still sfnaller 

 fragments. Some of these (those containing a nucleus no doubt) continue to 

 develop and may thus give deceptive results — a blastomere of the 2-celled stage, 

 for example, producing a gastrula of only ^3 or }( the normal size, or (?) even 

 less. The only absolutely certain results, therefore, are in cases where the egg- 

 membrane remains unbroken ; and results derived from naked blastomeres, as 

 well as those based on the study of preserved material, must be judged with a 

 certain reserve. 



It is necessary also to define clearly the terms to be employed. For the sake 

 of brevity one of the blastomeres of the 2-celled stage may be called a ^ blasto- 

 mere, one of the 4-celled stage a X blastomere, etc., and the later stages may be 

 correspondingly formed, yi embryos, J4 blastulas, }4 gastrulas, etc., ^ embryos, 

 etc. For comparison with the normal (total) development, we may distinguish the 

 }4 embryo (derived from a )4 blastomere) from the embryo-half {i.e., half of a 

 normal embryo derived from an unmodified egg), the X embryo from an embryo- 

 fourth, and so on. By a double, triple, quadruple or multiple embryo, we designate 

 one consisting of two, three, four, or several incompletely separated bodies. Mul- 

 tiple development is the origin from a single egg of more than one body, whether 

 separate or partially united. Partial development gives rise to a partial embryo 



— that is, one which (as in the case of % or Vie embryo) is not a dwarf entire 

 body but represents a fragment of a normal body. 



