310 Beer, Development of tlic ])ullcii yniiii and antlier of sonie Ona^raceae. 



has only l)een found to occmii- in ilio few isolated cases mentio- 

 ned above, otlier less extreme instanccs belonging to the same 

 categorv of ]iliGnomcna are not iinknown. Whcrovor \ve Hiul 

 that a iiew lamella is interpolalcd hetween the protuplast and 

 an older lamella and the latter still continues to grow in thick- 

 ness or in siirface it does so whilst it is neither in union or in 

 contact with the living element of the cell. 



Meanwhile changes are taking ])lace in the protoplast and 

 \vc' lind that a Üuid is .forming in the cytoplasm, partly at the 

 expense of carbohydrates which have reached it fromwithout 

 and partly at the expense of the cytoplasm itself. This fluid is 

 formed in vacuoles whieh gradually run together until the proto- 

 plast is reduced to a hoUow sphere enclosing a single, large, 

 central vacuole. "VVe have reason to believe that this fluid dif- 

 fuses out from the protoplast and furnishes material for the 

 growtli of the pollen-membrane. 



The three most important features in the formation and 

 development of the layers of the pollen wall may, therefore, be 

 summarised as follows: 



I. Both the primary pollen wall and the secondary thickening 

 layer originate in intimate connection with the plasmoderma 

 (Hautschicht). 

 II. The greater portion of the sul)sequent growtli of both 

 these niembranes takes place by intussusception whilst they 

 are completely separated from the protoplast. 

 III. The material requii'ed for the growth of the membranes 

 is derived from the tecretory activity of the poUen-proto- 

 plast. 

 We can at present only vaguely guess at the most pro- 

 bable way in which the growth of these membranes takes place. 

 There are some facts, such as Ambro nn 's work^) uj^on 

 the optical properties of the cuticulariced walls, which indicate 

 that the cell-wall may be underlaid i^y a crystalline structure 

 and it is possible that when the membrane is first formed the 

 protoplast (to which it is then firmly fixed) determines the 

 character and the arrangement of these crystals. 



The later growtli of the membrane, even after it has become 

 separated from the living element of the cell, liiay be considered 

 to take place in a manner which depends upon the nature and 

 relative positions of its crystalline components. 



8. After the poUen protoplast has become almost com- 

 pletely exhausted by its secretory activity its substance is once 

 more repenished by the material derived from the disintegration 

 of the tapetum. 



The very 3^oung tapetum contains a rather scanty cytoplasm 

 and only a single nucleus. Later the tapetal cells are furnished 

 with a denser cytoplasm and nuclei which may vary in number 

 from one to eight. Until the end of the special-mother-cell 



1) Ambro nn, H. Ber. d. Deutscli. bot. Gesell. 1888. p. 226. 



