THE MOSAIC STAGE OF DIFFERENTIATION 231 



the mid-brain exert a specific ''negative" trophic eflfect in the 

 hind-Hmbs, but the question cannot be regarded as settled. (See 



P- 430-) 



Before leaving the limb-field, there is a further point which re- 

 quires consideration. As early as the neurula stage, the mesoderm 

 of the limb-field is found to be self-diflFerentiating. At the same 

 time, it is clear, if only from the fact that the limbs are sym- 

 metrically placed with regard to the plane of bilateral symmetry, 

 that the localisation of limb-forming potencies is in some way de- 

 pendent on something else, which we may at present call the general 

 gradient-field of the organism (see Chap. ix). We are ignorant as 

 to the causes which are normally operative in calling forth these 

 limb-forming potencies in ordinary development, but we do know 

 that these potencies may be experimentally called forth by a variety 

 of agents. Grafted ear-vesicles,^ celloidin beads, or the free termina- 

 tion of various nerves deflected so as to underlie the tissues of the 

 field,- all result in the formation of limbs. The quality of the struc- 

 ture produced is therefore a specific property of the field, the 

 activities of which may be "released" by a variety of non-specific 

 agents. As we shall see, the same is true of other fields, and 

 probably of all. 



§6 

 A curious contrast to the regulative capacity of the limb-field is 

 the mosaic nature of the rudiment of the shoulder-girdle. This 

 rudiment consists of three centres of chondrification, representing 

 the coracoid, precoracoid, and scapular elements, but they are not 

 all contained within the 3 J-somite limb-disc, for when the latter is 

 grafted it will give rise to a shoulder-girdle of about one-third 

 normal size.^ Conversely, after extirpation of a limb-disc, portions 

 of the shoulder-girdle rudiment are left in sitii.^ Removal of, or 

 grafting of, portions of the limb-girdle rudiment at the early tail- 

 bud stage in Amblystoma results in the development of partial 

 structures, and regulation to form a complete girdle does not take 

 place, while regulation does take place to form a perfect limb.^ 



In some experiments on Amblystoma in which limb-discs 



^ Balinsky, 1925, 1926, 1927; Filatow, 1927. - Detwiler, 1918. 



^ Locatelli, 1925; Guyenot and Schotte, 1926; Bovet, 1930. 

 ^ Harrison, 1918. 



