ROOT-NODULES OF LEGUMINOSAE. 1 09 



land which probably had never borne lupins. The only satisfactory explanation which he finds for 

 this is the long adaptation of the bacteria to soil having the same reaction. Those which live in 

 alkaline soils are capable, he thinks, of invading all plants indigenous to such soils while those living 

 in acid (non-calcareous) soils attach themselves indifferently to the lupin, the furze, and the broom. 



These facts led him to undertake new experiments. He says that if the reaction of the soil is 

 the essential reason for the existence of two great physiological groups of nodule bacteria it should 

 be sufficient to accustom to acid media a bacterium from alkaline soil, in order to render it capable 

 of producing nodules on the roots of the lupin. This he says he succeeded in demonstrating : Bacteria 

 cultivated 8 months on media of gradually increasing acidity produced nodules on all the lupins 

 inoculated with them. The nodules appeared on the first lateral roots. There were none on the tap 

 roots. Five checks gave only i nodule. The same experiments on white lupin gave negative results 

 when grown in mineral solutions, but results were positive on plants normally developed in sand. 



In explanation of the influence of the acidity of the soil on the penetration of roots Mazd says : 



" If the soil is alkaline, the acidity of the secretions of the roots is neutralized to a certain depth 

 in the tissues. The bacteria, very sensitive to the action of acids, penetrate this layer, attracted by 

 the diffused sugar, but are not able to go farther into the roots." 



There must, therefore, he thinks, be forms especially adapted to acid soils. 



From his results he concludes that Nobbe's hypothesis is not confirmed either by cultural 

 experiments, or by the physiology of the bacteria. He says that the bacteria which are free in the 

 soil may be grouped according to the reaction of the soil, into two great categories, and that the 

 forms which are found in acid soils are capable of invading only those plants which avoid alkaline 

 soils such as the lupin, furze, and broom. 



Concerning the prevalence of the bacteria he says that when they do not manifest themselves 

 by the production of root-nodules it is not because they are absent from the soil but because the 

 conditions for their development are lacking. These conditions are obtained by proper treatment of 

 the soil, and certainly no one would attempt to inoculate with pure cultures a soil that had not been 

 so ameliorated. Hence he thinks that the use of pure cultures does not greatly aid agriculture. This 

 opinion applies to the nodule bacteria of the Leguminosae. It remains to be seen whether the 

 bacteria of alinite are of as little value. 



In 1899, Maria Dawson contributed an interesting paper on nitragin and the nodules of legu- 

 minous plants. Her investigations were carried on in England in the laboratory of H. Marshall 

 Ward and were suggested by the commercial introduction of nitragin by Nobbe and Hiltner. 



Her studies were confined principally to Viola hirsuta and Pisum sativum. Each showed pal- 

 mately branched nodules within 14 days of sowing. 



Various fixatives were tried. The most satisfactory results were obtained by using Flemming's 

 more concentrated solution or absolute alcohol. Hand sections served better than microtome ones 

 for examination of the bacterial filaments within the cells. She found abundant evidence of the 

 parasitic nature of the organism. In fresh material the infection tubes were made visible by treating 

 with Eau de Javelle or potash. In all cases a bright spot of infection was seen either at the tip or 

 at the side of the hair, accompanied by a bladder-like swelling of the hair at the point of attack. 

 Hand sections of fresh material treated as above showed the course of the infection-tube across the 

 cortex and its branching into the deeper cells of the nodule. Trumpet-like swellings where the 

 tubes cross the cell-walls and numerous spherical or pear-shaped swellings on the tube within the 

 cells (previously described by Marshall Ward) were clearly seen, as well as breaks in the tube, each 

 portion ending in a fine point, the points directed toward each other (fig. 36). This also has been 

 seen by others. 



The author next attempted to obtain a reagent which would stain the filaments but not the 

 bacteroids. Gold chloride (0.5 per cent) used on fresh material gave some help. Fresh material 

 en masse was left in the stain from i to 24 hours for microtome sections, while hand sections were 

 stained for 10 or 15 minutes. In either case the material was quickly washed with water and trans- 

 ferred to a solution of formic acid (0.25 per cent) in the dark for 24 hours, or for the same time to 

 water acidulated with acetic acid, in the light. The sections were then washed well in water and 

 placed in formic glycerin, or if intended for imbedding, the material was transferred gradually to 

 absolute alcohol and thence to paraffin. By this method the contents of the filament stained deeply 

 and had the vague appearance of being made up of numerous short rodlets. The limiting layer 

 remained colorless. 



This hint as to the nature of the filament was successfully supplemented by the use of Stras- 

 burger's method for differentiating fungous hyphae in the tissues of the host. She found the best 

 method of treatment to be as follows : Sections hardened in alcohol (best without previous treatment 

 with chromic or osmic acid) are placed for about two hours in alcoholic potash (one part 5 per cent 

 potash to three parts absolute alcohol) and then passed into Eau de Javelle for 10 minutes. From 



