Resting -spores 27 



papillate exterior. The details of the development of the spores have been 

 described by various authors in a number of genera. In the Hormogonese 

 any vegetative cell of the filament may become a spore 1 , or, as in the 

 Rivulariacese, only the cell next the basal heterocyst may become thus 

 differentiated. In some species (about 30 per cent.) of the genus Anabtena 

 the spores arise in relation to the heterocysts, being developed from those 

 cells adjacent to the heterocysts on either one or both sides 2 . In Hapalosiphon 

 and Scytonenia, in which the vegetative filaments are enclosed in strong 

 sheaths, there is no enlargement of a vegetative cell on its conversion into 

 a spore. The only obvious change is the development of a thick brown wall 

 (W. & G. S. W., '97). In these genera spore-formation is in no way related 

 to the heterocysts; nor, of course, can it be in the five genera of the 

 Homocystese in which resting-spores are known to occur. 



During the development of the spore, particularly if there is any 

 considerable increase in size, much reserve material accumulates in it. 

 According to Phillips ('04) this is largely passed into the spore from the adjacent 

 ' nurse ' cells by way of the intercellular pores. The two spore-membranes 

 can be seen at a very early stage, and as development proceeds they 

 gradually become more and more differentiated. The inner membrane 

 (endospore), which from the first completely envelops the protoplast, remains 

 colourless. The outer membrane (exospore) is at first only a cylindrical 

 sheath, but subsequently it extends around the two ends of the spore 

 and becomes a strong membrane. In the mature spores the endospore 

 and exospore are separated by a very thin layer of mucilage. Fritsch ('05) 

 states that in Anab&na azollte the endospore and exospore are merely 

 the fully-developed inner investment and cell-sheath respectively, both 

 of which in the mature condition completely envelop the protoplast. The 

 resting-spores vary much in shape in the different genera and species. 

 They may be spherical, ellipsoid, cylindrical, or more -rarely sublimate. 



1 Phillips describes the spores of Oscillatoria as often arising from more than one cell, 

 sometimes as many as four cells fusing together by the absorption of the intervening cell-walls. 



' This development has been referred to as ' centripetal ' in contrast to the so-called ' centri- 

 fugal ' development in which spore-formation begins in cells distant from the heterocysts and 

 gradually advances towards them. It would, however, be wise completely to discard the use of 

 these terms in relation to spore-formation in Anabicna, and in any case the application of the 

 terms should be reversed since the only obvious fixed points are the heterocysts. Even in those 

 species of the genus which were at one time referred to Spli&rozyga (Agardh, 1827 ; Kalfs, 1850) 

 spore-formation does not always begin in the cells adjacent to the heterocysts and gradually 

 recede from them [the method which would be more correctly termed ' centrifugal ' !], but is 

 frequently mixed in the same spore-aggregate. The terms, as used, would be to some extent 

 correct if the filaments of Anab&iia consisted of a chain of vegetative cells terminated at each 

 end by a heterocyst, but this is not the case except in certain very specialized species, such as 

 the plankton-forms Anabicna circularis G. S. W. and A. Tanganyika; G. S. W. (vide G. S. W. 

 '07; and fig. 19 A E). 



