1886.] 



MICROSCOPICAL JOURNAL. 



133 



development can thus be kept well 

 under control, and if it be found that 

 the exposure is about right, the bro- 

 mide can be ^^'ashed out and develop- 

 ment begvm anew. 



The novice will be puzzled at times 

 to know whether a finished negative 

 is over or vmder timed, although it 

 may be evident that something is the 

 matter. Usually, the question can be 

 definiteh^ settled by examining the 

 shadows. If the detail is all visible and 

 the negative is thin, the margins of the 

 plate where it was protected by the 

 holder remaining clear, the exposui'e 

 was too long. If the margins are not 

 clear it is an indication of a foggy 

 plate. The fog ma}' be due to the 

 emulsion, or to accidental exposure 

 to light. Fog may also be readily 

 produced by using too much alkali 

 in the developer. A fogg}' or light 

 struck plate will give a weak and flat 

 negative, just as an over-exposed 

 plate. An under-timed plate will 

 not show the details in the shadows, 

 while in a landscape the sky will 

 probably develop dense, black, and 

 perfectly opaque. 



We have still to consider a few 

 methods of reducing and intensify- 

 ing negatives, but these must be de- 

 ferred until next month. 



[To be continued.^ 

 O 



Provisional Key to Classification of 

 Alg« of Fresh Water.— IX. 



BY THE EDITOR. 



S^Continued from page 97.] 



Fa7nily XIV. Nostocace^. 



Filamentous, simple or branched 

 trichomes. Resting spores observed 

 in many genera, also peculiar hetero- 

 cysts — colorless cells apparently with- 

 out contents — interspersed in the 

 series of vegetative cells, the function 

 of which is unknown. 



Groups. 

 Filaments usually branched, rarely 

 simple, provided with a sheath, taper- 

 ing to a hair-like end ; with hetero- 



CystS. (RlVUI.ARIE^.) 



Filaments not tapering to a hair- 



like end, sheathed, branched, cell 

 division only at right angles to the 

 length of the filament, branches 

 formed by lateral outgrowths of the 

 filaments breaking through the 

 sheath. Usually with heterocysts. 



(SCYTONEME.E.) 



Filaments not tapering to a point, 

 sheathed, branched, cell division also 

 parallel to the length of the filament, 

 whereby branching results and the 

 filament itself includes series of cells 

 13'ing side by side. (Stigoneme^e.) 



Simple, unbranched filaments, with 

 or without sheaths, never tapering to 

 a point ; heterocysts always present, 

 and resting cells (spores) usually 

 observed. 



Propagation in two ways : i . by 

 development of resting cells after a 

 period of repose, giving rise to new 

 filaments by repeated division; 3. 

 by multicellular, germinating fila- 

 ments (hormogonia) , separated por- 

 tions of the trichome. which grow 

 into new plants. (Nostoce^e.) 



Simple, unbranched filaments, with 

 or without sheaths, single or forming 

 extended layers, without heterocysts 

 and resting cells, never tapering to a 

 hair poiiit. 



Propagation: i. bv the breaking 

 up (disarticulation) of filaments (hor- 

 mogonia) the single pieces growing 

 into new filaments i^O s cilia ria ^ 

 Lyngbya^ Symploca) ; 2. by uni- 

 cellular g o n i d i a , which may be 

 either the separated, terminal cells 

 of filaments ( Chaniccsiphon^ Lepto- 

 thrix?) or special cells developed in 

 the course of vegetative division 

 ( Crenot/irix) . 



A number of the genera manifest, 

 either always or under certain con- 

 ditions, a turning around the longi- 

 tudinal axis of the filaments, and a 

 forward, creeping movement ( Os- 

 cillaria^ Beggiatoa^ Sp irulina ^ 

 Spirochceta). (Oscillarie^.) 



a. RivuLARiE.E. Group i. 

 Synopsis of Genera. 

 Growing in tufts. Heterocysts at 

 base of branches. Calothrix^ 105. 



