Morphological Studies in the Life- Histories of Bacteria. 479 



number, however, coming to maturity on solid media, the majority fading 

 and disappearing. 



16. The appearance of the very minute forms, seen in figs. 1, 3, and 4, 

 and Plates 16 and 20, makes it impossible to be certain, without 

 prolonged observation on the warm stage, that, in attempting to obtain 

 cultures from single individuals of normal size by Barber's method, or 

 by the fragmented slip method of isolation, one is not in reality culti- 

 vating from several individuals. Unless therefore the presence of these 

 minute forms can be excluded, the use of these two methods for obtaining 

 cultures in liquid media from single organisms cannot be relied on. 



17. The presence of these minute forms is probably the explanation of 

 the apparent filterability through Chamberland filters of such relatively 

 large organisms as the Bacillus bronchisepticus, and is perhaps responsible 

 for the general view that even well-made Berkefeld filters are not suitable 

 for bacteriological work. 



18. By the dark-ground method of illumination many of these small 

 detached buds appear as minute bacilli in the act of undergoing binary 

 fission. Not infrequently these appear as coccoid bodies, if binary fission 

 has not begun. In the study of aberrant bacterial forms with dark-ground 

 illumination the use of the hanging drop method, apart from the inherent 

 fallacies of dark-ground work, is fatal to correct interpretation unless 

 streaming movements have first been reduced to a minimum. For example, 

 it is frequently stated that apparent branching in bacteria is, in reality, 

 simulated by mere apposition, and that observation of a hanging drop with 

 dark-ground illumination will soon dispel the illusion, separation of apposed 

 organisms sooner or later always taking place. On casual inspection of 

 dark-ground hanging drops this statement appears to represent the truth, 

 especially if streaming movements are still free. If, however, a drop of 

 emulsion be firmly pressed under a cover-slip and then examined, it will be 

 found, streaming movements now being reduced to a minimum, that detach- 

 ment does not invariably take place. That this is not the result of pressure 

 is shown by the fact that in favourable cases long lateral buds will exhibit wide 

 lateral movements, whilst the base, or point of attachment to the parent stem, 

 remains fixed. In other cases short lateral buds retain their relative 

 position to the parent stem, itself exhibiting unfettered rotatory movements. 

 The accuracy of these observations is confirmed in Plate 20 of warm stage 

 studies. 



In concluding it is perhaps unnecessary to point out that no claim 

 whatever is here made that the complete life-histories of the bacteria of the 

 vol. lxxxix. — B. 2 s 



