144 THE BEGINNINGS OF LIFE. 



more frequent in saline solutions than Bacteria, and 

 in some of these after they have been boiled no Bacteria 

 ever present themselves. This, for instance, has been 

 found to be the case with ammonic tartrate and sodic 

 phosphate solutions; although when the former salt 

 has been replaced by ammonic carbonate, such a so- 

 lution has yielded Bacteria after it has been boiled 1 . 

 Because in some solutions Bacteria reproduce Bacteria, 

 and in other solutions Torulte reproduce Torul<e because 

 each of these forms ( breeds true ' they have been re- 

 garded by many as distinct 'species/ This, however, is 

 not altogether conclusive. A fragment which detaches 

 itself from one of the lowest living things has just 

 as much tendency to grow into the form of its parent, 

 as the fragment detached from a given crystal has to 

 reproduce a similar crystalline form. In each case, 

 however, the parent form is reproduced only so long 

 as the conditions remain the same. Placed under 

 new conditions the crystalline fragment may grow up 

 with a modified form, and, similarly, a change may 

 overtake a portion of matter thrown off from a pre- 

 existing living form. In order that the crystal may 

 lapse into another form, it seems necessary that the 

 new conditions shall be capable of bringing about a 

 new molecular arrangement (or allotropic state) of the 



trop leve, la germination se fait d'une manikre confuse ou ne se pro- 

 duit pas ; les spores spontanees s'alterent avant qu'elle ait lieu, et le 

 liquide se remplit de Bacteriums, de Monades, ou de Vibrions.' (Loc. 

 cit., p. 175, note i.) 



1 Compare Exps. n and x with Exps.y and z (vol. i. p. 462). 





