Nature of Heterocysts 19 



of the genus Anabfena the filaments break readily at all points, and this 

 fracture is in no way controlled by the heterocysts. Moreover, in Nodularia, 

 the fracture of the filaments is almost invariably between the heterocysts. 

 Neither does the structure of a filament of Stigonema or Scytonema support 

 this view, although in the former genus heterocysts always limit the hor- 

 mogones. Hieronymus ('92), Hegler ('01), and later, Fritsch ('04), have 

 regarded the heterocysts as storehouses for reserve substances, the latter 

 passing into the heterocyst along the protoplasmic threads which com- 

 municate with the adjoining cells. Phillips ('04) has suggested that as 

 the heterocyst is usually next or near to the spores it might possibly be 

 considered as a storehouse of food for them. He also states that the 

 heterocyst of Cylindrospermum will develop into a spore if it gets sufficient 

 nutriment and hereditary material passed into it from the other cells. The 

 fact that spores occur abundantly quite away from the heterocysts, as in 

 Hapalosiphon, Scyionema, and certain species of Anabwna, does not lend 

 much support to this view ; nor does the fact that in one filament there may 

 be twenty, or even thirty, times as many spores as heterocysts. 



Fig. 13. A Z>, formation of gonidia by the heterocysts of the aberrant Anabsena cycadearum 

 Reinke. In B D, a number of gonidia are being formed ; in A, only three. E, young 

 filament developed from a gonidium. All x 2200 (after Spratt). 



The remark made by Spratt ('11) that in Anabsena detached heterocysts ' certainly 

 occur under both natural and artificial conditions ' is erroneous as a general statement. 

 Detached heterocysts never occur in the heterocystous Myxophyceae living under normal 

 natural conditions. The species examined by Spratt was Anabsena cycadearum Reiuke, 

 which in addition to being the most specialized species of the genus, having adapted 

 itself to conditions of environment totally unlike those under which species of Anab&na 

 normally live, is also a degenerate form. 



Brand ('01) observed the contents of the heterocysts in Nostoc commune 

 and N. microscopicum set free as gonidia, which subsequently developed into 

 new filaments. Fritsch has observed states which he thought might be due 

 to arrest of gonidia-formation in the heterocysts of a species of Anabtena. 

 These observations were made on material in cultures. Spratt (11) has also 

 observed the formation of gonidia in the heterocysts of Anabtena cycadearum, 

 'in material from old nodules and hanging-drop cultures.' The contents 

 divide first into two, and eventually into a number of small spherical gonidia 



22 



