236 THE MOSAIC STAGE OF DIFFERENTIATION 



pressed in amphibian embryos by the direction of beat (mostly 

 antero-posterior) of the ciHa which it bears. The ciUa arise at the 

 early neurula stage, and if a piece of epidermis of Amblystoma is 

 rotated through 180° and replanted at this stage, the cilia beat in the 

 normal direction. If, however, the epidermis is rotated at the late 

 neurula stage, its polarity is then fixed, and the cilia beat in the 

 reversed direction.^ 



The so-called balancer, present in some Urodela larvae {Triton^ 

 Diemyctyliis, Amblystoma punctatum, but absent or extremely rudi- 

 mentary in Amblystoma tigrinum'^), is an organ of attachment in the 

 form of a cylindrical projection of ectoderm with a mesenchymal 

 core. Balancer-forming potencies occupy fields in the ventral ecto- 

 derm of the head, beneath the eyes ; they are at a maximum at a 

 central point, and decrease with increasing distance from it.^ If a 

 part of the rudiment is extirpated at an early stage, a balancer will 

 be formed from the neighbouring regions.* A balancer rudiment 

 grafted into other positions is self-diiTerentiating, and induces the 

 formation of the mesenchymal core from host-tissue. A rudiment 

 from Amblystoma punctatum can be grafted on to Amblystoma 

 tigrinum, or even on to the anuran Rana sylvatica, and develop into 

 a balancer with induced core, although these hosts normally possess 

 no balancer. As mentioned in Chap, vi (p. 177), a single balancer- 

 field can give rise to as many as four balancers.^ (See also p. 327.) 



The fully formed balancer has radial but not bilateral symmetry, 

 and it does not appear that the balancer-field is polarised. The facts 

 already recorded in Chap, vi (p. 177), viz. that balancer-forming 

 potencies can be evoked by neural crest cells, neural fold cells, and 

 fore-gut-wall cells, even belonging to Urodeles or Anura which 

 normally possess no balancer, serve as a further illustration of the 

 principle enunciated above (p. 23 1), that the quality of the structure 

 produced depends on intrinsic properties of the field, and not on 

 specific stimuli of the releasing mechanism. At the same time, the 

 fact that tissue from species which possess no balancer is capable of 

 evoking balancer-forming potencies, shows that the absence of a 

 balancer in these species is due to the absence of such a field in 



^ Twitty, 1928. ^ Nicholas, 1924 a. 



^ Harrison, 1925 b. ^ Bell, 1907. 



^ Raven, 193 1 a. 



