248 
Ellis . — The Life-history of Bacillus hirtus. 
Cell-Division. 
Migula, in his System der Bakterien, p. 140, makes the following 
assertion with reference to the mode of division of the cells of Bacteriaceae : 
c Gegentiber den Coccaceen ist diese Teilung dadurch charakterisiert, dass 
ihr eine zur Teilungswand senkrechte Streckung der Zelle vorhergeht. 
Es ist hier also ein entwickelungsgeschichtliches Merkmal gegeben, durch 
welches die Coccaceen scharf von den iibrigen Bakterien getrennt werden.’ 
My observations confirm this statement so far as the majority are con- 
cerned. In most cases, the individual showing two cells, in which the 
division wall is not fully formed, is longer than any one-celled individual. 
This, however, is not universal, as is seen in Fig. 23 where division has 
taken place without previous elongation, and in Fig. 24 is shown a case 
where very slight elongation has preceded the act of division. Hence, 
though Migula’s statement is true for the majority, its application does not 
appear to me, from my observations, to be universal. All we are entitled 
to assert is that elongation in length preceding division is much more 
common in the Bacteriaceae than the converse process, and we may also 
assert that this precedence of elongation is much more common than 
is increase of volume preceding the division of the Coccaceae. The two 
orders cannot be absolutely distinguished by this characteristic ; all we are 
entitled to say is, that in one order one kind of division is much more 
common than it is in the other. 
The second point to determine is whether a division-wall is formed 
before or after the constriction at the point where the wall appears. 
Meyer ( 9 ) figures a case (Fig. 9 d) in which there appears to be a very 
slight constriction preceding the formation of the division-wall. I have not 
noticed this stage in this Bacillus, and in another form of Bacillus, in which 
I was able to follow the division from beginning to end, no constriction was to 
be seen preceding the appearance of the division-wall. I was able to observe 
a single cell dividing, and then the daughter-cells repeating the process, 
until the genesis of more than a hundred individuals from a single cell had 
been directly observed : in this case the daughter-cells remained attached, so 
that I was able to observe the process of division several times over. In no 
case did a constriction precede the formation of a wall. If this were the 
usual mode of procedure, it should not be difficult to find stained specimens 
exhibiting this stage in the division, but I have never been able to see this 
stage in any of my stained preparations. On the other hand, by careful 
staining, numerous cases are seen in which individuals without the slightest 
constriction possess division-walls (Figs. 9, 11, 17, 30). The method 
of staining, which gave the best results with me, consisted in treating the 
cells with about 50 °/ o Formalin for 3-4 minutes, then adding to the liquid 
by means of a platinum-loop, two or three loopfuls of a somewhat dilute 
