in one direction, then at right angles to the first division, and a third division follows at right angles to the plane of the first two, thus producing solid groups of fours, eights, or sixteens (Fig 5), called Sarcina.  Each different species of bacteria is uniform in its method of division, and these differences are therefore indications of differences in species, or, according to our present method of classification, the different methods of division represent different genera.  All bacteria producing Streptococcus chains form a single genus Streptococcus, and all which divide in three division planes form another genus, Sarcina, etc.

The rod-shaped bacteria also differ somewhat, but to a less extent.  They almost always divide in a plane at right angles to their longest dimension.  But here again we find some species separating immediately after division, and thus always appearing as short rods (Fig. 6), while others remain attached after division and form long chains.  Sometimes they appear to continue to increase in length without showing any signs of division, and in this way long threads are formed (Fig. 7).  These threads are, however, potentially at least, long chains of short rods, and under proper conditions they will break up into such short rods, as shown in Fig. 7a.  Occasionally a rod species may divide lengthwise, but this is rare.  Exactly the same may be said of the spiral forms.  Here, too, we find short rods and long chains, or long spiral filaments in which can be seen no division into shorter elements, but which, under certain conditions, break up into short sections.

Rapidity of multiplication.

It is this power of multiplication by division that makes bacteria agents of such significance.  Their minute size would make them harmless enough if it were not for an extraordinary power of multiplication.  This power of growth and division is almost incredible.  Some of the species which have been carefully watched under the microscope have been found under favourable conditions to grow so rapidly as to divide every half hour, or even less.  The number of offspring that would result in the course of twenty-four hours at this rate is of course easily computed.  In one day each bacterium would produce over 16,500,000 descendants, and in two days about 281,500,000,000.  It has been further calculated that these 281,500,000,000 would form about a solid pint of bacteria and weigh about a pound.  At the end of the third day the total descendants would amount to 47,000,000,000,000, and would weigh about 16,000,000 pounds.  Of course these numbers have no significance, for they are never actual or even possible numbers.  Long before the offspring reach even into the millions their rate of multiplication is checked either by lack of food or by the accumulation of their own excreted products, which are injurious to them.  But the figures do have interest since they show faintly what an unlimited power of multiplication these organisms have, and thus show us that in dealing with bacteria we are dealing with forces of almost infinite extent.