II GENERAL NATURE OF REGENERATION 595 



of "growth" (Table 3). A similar indifference is often recorded between regenera- 

 tive and embryonic "growths", and between regenerating and normal tissues 

 {Table 3), even when grafted in close proximity (Needham, 1950a). Two "or- 

 ganisers" in one embryo behave similarly. The qualitative interactions recorded 

 by others (Table 3) usually have been slight and might be spontaneous reactions 

 by each growth. In Asellus (Needham, 1950a) the epidermis and exoskeleton of 

 approximated normal, embryonic and regenerating tissues became continuous 

 throughout, yet they remained a complete mosaic anatomically (Fig. i). In view 

 of their contrasted metabolisms (p. 609), some degree of independence between 

 regenerating and local, parental tissues seems essential, and actual mechanical 

 barriers have been recorded (Needham, 1952 p. 22; Singer, 1954). 



III. THE STAGES IN AN ACT OF REGENERATION 



For the present purpose a detailed analysis of the causal sequence of events in 

 typical epimorphic regeneration (Needham, 1952) is not required, but only of the 

 chronical and other relationships of growth-components to qualitative components 

 and to the process as a whole. Particularly important is the biphasic nature of the 

 whole process, an initial phase of regression (R-phase) involving demolition, 

 dedifferentiation and segregation, followed by a P-phase of progressive processes, 

 aggregation, cell-proliferation, growth and differentiation. The R-phase appears 

 to be more than merely the clearing-up of damaged material, and is probably 

 necessary to restore an "embryonic" condition favourable to new growth, — "reculer 

 pour mieux sauter", — -particularly in the higher animals. Metabolism locally, and 

 to some extent systemically, shows a corresponding biphasic change. 



The cytolytic demolition of cells and intercellular material, damaged by in- 

 jury and subsequently by foreign organisms and other secondary causes, is the 

 first morphogenetic process and is often quite completed before the histolytic 

 dedifferentiation of intact cells and material begins (Thornton, 1938). The latter 

 is mainly a chemical process (Butler and Schotte, 1 94 1 ) as in the dedifferentiation 

 of metamorphosis in insects, whereas phagocytosis plays a major role in demolition. 

 Dedifferentiation does occasionally (Kamrin and Singer, 1955; Butler and Blum, 

 1955) but not usually (Thornton, 1954), occur without ti^auma, but it can then 

 spread far from the site of injury. It is enhanced by a number of experimental 

 procedures, e.g. denervation, irradiation and colchicine-treatment and halted by a 

 regenerating nerve-supply and by a living, grafted blastema. It is not yet proved 

 that all dedifferentiated cells are reused (Butler and Schotte 1941; Thornton, 

 1953) and under certain experimental conditions certainly many are not. 



However most of the cells of the blastema have a local origin, in Amphibia 

 (Butler, 1935; Butler and O'Brien, 1942). In the lower Metazoa (Wolff and 

 Dubois, 1947, 1948; Evlakhova, 1946) and to some extent in other Metazoa 

 (Wolff and Wey-Schue, 1953) immigrant, undifferentiated neoblasts contribute 

 to the blastema, but there is at present no definite evidence that this affects the 

 extent of dedifferentiation. Also there is no evidence that the extent of dedifferen- 

 tiation is proportional to the amount of tissues to be regenerated: like nitrogen 



Literature p. 64^ 



