162 THE EATE OF GROWTH [ch. 



after. At six weeks old there is another well-marked retardation; 

 it follows on a rapid spurt, and coincides with the epoch of puberty *. 



In arthropod animals growth is apt to be especially discontinuous, 

 for their bodies are more or less closely confined until released by 

 the casting of the skin. The blowfly has its striking metamorphoses, 

 yet its growth is wellnigh continuous; for its larval skin is too thin 

 and dehcate to impede growth in the usual arthropod way. But 

 in a thick-skinned grasshopper or hard-shelled crab growth goes by 

 fits and starts, by steps and stairs, as Reaumur was the first to shew ; 

 for, speaking of insects f, he says: "Peut-etre est-il vrai generale- 

 ment que leur accroissement, ou au moins leur plus considerable 

 accroissement, ne se fait que dans le temps qu'ils muent, ou pendant 

 im temps assez court apres la mue. lis ne sont obHges de quitter 

 leur enveloppe que parce qu'elle ne prend pas un accroissement 

 proportionne a celui que prennent les parties qu'elle couvre." 

 All the visible growth of the lobster takes place once a year at 

 moulting-time, but he is growing in weight, more or less, all along. 

 He stores up material for months together; then comes a sudden 

 rush of water to the tissues, the carapace sphts asunder, the lobster 

 issues forth, devours his own exuviae, and lies, low for a month while 

 his new shell hardens. 



The silkworm moults four 'times, about once a week, beginning 

 on the sixth or seventh day after hatching. There is an arrest or 

 retardation of growth before each moult, but our diagram (i^ig. 35) 

 is too small to shew the sHght ones which precede the first and 



problem, in Ergebnisse d. Physiologie, xxxiii, pp. 883-973, 1931. These two cases 

 of Lupinus and Vallisneria, a^e among the many which lend themselves easily to 

 Backman's growth -formula, viz. Lupinus, \ogp= - 2-40 + 1-48 log T - 6-61 log^ T and 

 Vallisneria, log p = + 1-28 + 4-51 log T - 2-62 log^ T. See for an admirable resume 

 of facts, Wolfgang Ostwald, Ueher die zeitliche Eigenschaften der Entwicklungsvorgdnge 

 (71 pp.), 1908 (in Roux's Vortrdge, Heft v); and many later works. 



* Cf. R. Robertson, Analysis of the growth of the white mouse into its con- 

 stituent processes, Journ. Gen. Physiology, vin, p. 463, 1926. Also Gustav 

 Backman, Wachstum d. w. Maus, Li^nds Univ. Arsskrift, xxxv, Nr. 12, 1939, 

 with copious bibliography. Backman analyses the complicated growth- curve of 

 the mouse into one main and three subordinate cycles, two df which are embryonic. 

 Cf. St Loup, Vitesse de croissance chez les souris. Bull. Soc. Zool. Fr. xviii, 

 p. 242, 1893; E. Le Breton and G. Schafer, Trav. Inst. Physiol. Strasburg, 1923; 

 E. C. MacDowell, Growth-curve of the suckling mouse. Science, Lxvin, p. 650, 

 1928; cf. Journ. Gen. Physiol, xi, p. 57, 1927; Ph. THeritier, Croissance. . .dfeins les 

 souris, Ann. Physiol, et Phys. Chemie, v, p. i, 1929. 



I Memoires, iv, p. 191. 



i 



