266 METAMORPHOSIS 



volume takes place in lamellae laid down by apposition, and also in cases where 

 these lamellae are separated from protoplasm by other layers. Growth in 

 thickness can thus be produced by intussusception also. We will content our- 

 selves with one example, viz. the cell -wall of Gloecapsa alpina, which has been 

 carefully investigated by NAGELI (1858) and CORRENS (1889). When the cells 

 of this alga divide each daughter-cell forms a new cell-wall of its own. The 

 original wall, however, now separated from the protoplasm by these new walls, 

 increases also both in thickness and in extent as shown in Fig. 55, //. CORRENS' s 

 measurements and arguments show that this growth is explicable only on the 

 assumption of an interpolation of organic material. 



Not infrequently the greatly thickened membrane remains completely 

 homogeneous, but most usually we can make out a concentric lamination in it 

 which may be due to different causes (CORRENS, 1891). The successive layers 

 may be differentiated chemically, or only by the amount of water they contain. In 

 the latter case the lamination is due to the same cause as in starch grains ; 

 but the causal connexion is no more clearly explained in the one case than in 

 the other. We need not, however, enter in further detail here as to the struc- 

 ture and growth of starch grains, although these bodies have taken a foremost 

 place in all discussions on growth, and will always have a historic interest as 

 being the basis of NAGELI' s (1858) theory of intus- 

 susception. Thanks to the investigations of MEYER (1881 

 and 1895) and of SCHIMPER (1881) we now know that the 

 growth of a starch grain is effected by the external apposi- 

 tion of new material from without, and that the increase in 

 size of the grain takes place in the same way as in- 

 crease m s ^ ze f a crystal or a spherocrystal. In 

 a certain sense starch grains are rather subjects for 

 crystallographical than physiological study. We say ' in 

 a certain sense ' intentionally, since although growth by 

 apposition has been definitely proved to occur in them 

 thick- the occurrence of intussusception also is not completely 



excluded. 



One point of similarity between the starch grain and the cell-wall must be 

 noted. Many cases in which we were in the past accustomed to assume growth 

 in the cell-wall by intussusception have now been shown to be due to apposition ; 

 similarly cases of increase occur due to supplementary intussusception, and 

 hence it was customary in any generalization to give ' apposition ' or ' intus- 

 susception ' as alternatives. At the time when there was a reaction against the 

 NAGELI-HOFMEISTER theory of intussusception one had become accustomed to 

 regard the cell-wall as a non-living structure, and to compare it with the shell of 

 a mollusc. HOFMEISTER (1867), on the other hand, held that the membrane was 

 alive and ascribed to it all the characters we now associate with protoplasm only. 

 When cases became known which militated against the universality of apposition, 

 and which rather suggested the activity of a living structure, attempts were 

 made to explain the situation by assuming an infiltration of the protoplasm 

 into the cell-wall, and so to account for its apparent vitality (WIESNER, 1886 ; 

 STRASBURGER, 1889). The existence of this hypothetical plasma in the wall 

 has never as yet been demonstrated (CORRENS, 1894), and so perhaps we may 

 come back once more to HOFMEISTER' s conception and regard the membrane 

 itself as living, although certainly not independent life as we understand it in 

 protoplasm, since there is no doubt whatever that protoplasm may grow without 

 a cell-wall, but not a cell-wall without protoplasm. But the membrane is 

 perhaps alive in so far as it can exhibit the power of ' assimilation in the real 

 sense of the word ' (p. 259), being able itself to form new wall material out of 

 definite but as yet unknown substances provided by the protoplasm. If this be 



