164 BACTERIOLOGY OF THE EYE 



On bullock's or sheep's blood-serum, prepared in Loffler's method, an unevenness 

 of the surface develops in twenty-four hours, and takes the form of small, moist, 

 slightly sunken, transparent areas, which slowly deepen, so that the medium is gradu- 

 ally liquefied, mostly without any change in its colour. With some strains the fluid 

 material becomes brown. This has something to do with the serum, as in cooked 

 serum it occurs much more rarely. If the milky and slightly viscous fluid mass be 

 removed with a loop, the surface of the serum appears gouged out. When the 

 colonies are widely separated, deep pits are formed, but when close together, 

 they coalesce on the surface, and their viscous contents flow down into the con- 

 densed water ; the peculiarly excavated surface is then very obvious. In about 

 fourteen days almost the whole of the medium is liquefied down to the depths of the 

 tube. The liquefaction, as a rule, is not complete, as the Diplobacilli die out, and 

 generally hardly anything but involution forms is found in the liquefied mass. In 

 cooked serum the liquefaction is neither so intense nor so rapid as it is in serum 

 prepared by fractional sterilization. 



This action is extremely characteristic of the Diplobacillus. Of all the pathogenic 

 bacteria which are commonly found on the conjunctiva only the allied Petit type of 

 Diplobacillus possesses this power. It should be noticed, by the way, that many 

 sterile secretions have a certain power of dissolving or liquefying serum media. 

 Every hollow on the surface of Lofner's serum should not, therefore, immediately be 

 considered as a colony of Diplobacilli; a round excavation becoming progressively 

 deeper, however, certainly indicates Bacteria. We must also remember that serum 

 which has been imperfectly sterilized may contain the spores of liquefying organisms 

 which may develop in the incubator. 



On serum agar the Diplobacillus grows in the form of small, translucent, flat 

 pale grey drops, resembling colonies of Pneumococci. Under higher magnification 

 the colonies appear to be finely granulated, round in shape, with an even, slightly 

 prominent margin, and show very little tendency to coalesce ; individually they are 

 smaller than when grown on Loffler's blood-serum. 



In serum bouillon a faint but quite marked cloudiness occurs in twenty-four to 

 forty-eight hours ; a slight deposit sinks to the bottom, and can easily be shaken up. 

 On agar, free from serum, the Diplobacilli from the conjunctiva usually fail to 

 grow, and when occasionally they do develop their growth is scanty, they die out 

 readily, and cannot be propagated. In this they differ from the Petit type, which 

 grows freely on ordinary agar. Growth on agar is quite exceptional. (I have 

 recently grown one or two strains from conjunctivitis cases which at first showed 

 a fair growth on agar, but on cultivation lost it. In gelatine stab cultures at room 

 temperature slow growth and liquefaction occurred in these particular strains, just 

 as in the Petit type. This was lost later.) 



In milk, or potatoes, or sloped bullock's blood, and blood media they do not grow. 

 The media should always have an alkaline reaction ; when neutral, the organisms 

 grow badly, and acid media are unsuitable. This explains why the Diplobacilli 

 in the presence of the Staphylococcus pyogenes aureus sometimes grow badly on 

 serum, although they were present in excess in the secretion ; indeed, the bacilli 

 may not grow at all when the rapid growth of the cocci has produced an acid 

 reaction in the medium. On the other hand, the Diplol>acilli grow very well along 

 with xerose bacilli, as these latter do not cause any change in alkalinity, but rather 

 favour the growth of the former. They grow well with the ordinary white Stapliy- 

 lococci of the conjunctiva. 



The Diplobacilli can often be obtained in pure culture from cases 

 with profuse secretion, provided that the lid margins and angles are 

 not touched. 



