1890.] On the Question of the Fixation of Free Nitrogen. 93 



growth, and better colour, than those in either pot 2 or pot 3 with the 

 soil-extract seeding. Indeed, it was not until about the middle of 

 September, that is four or five weeks after the seeding with soil- 

 extract, that the plants in pots 2 and 3 began to show a darker green 

 colour than those in pot 1 without the soil-extract. The indication 

 was, however, so striking, that on September 25 it was decided to 

 count the leaves, and to estimate the relative area of leaf-surface, on 

 the plants in the different pots. For this purpose, the leaves were 

 classified into those which were dead, those that were dving, those 

 which were changing colour, and those which were still bright green. 

 It must suffice here to show the number, and the estimated relative 

 area, of the total leaves in each case, on September 25, when the first 

 counting and estimates were made, on October 17, on November 14, 

 and on December 14, when the plants were cut. The following table 

 summarises these results. The first four columns show the total 

 number of leaves, and the second four the estimated relative leaf- 

 surface, that of the plants of pot 1 (without soil seeding) on Septem- 

 ber 25, being taken as 100. 



Peas, 1888. 





Number of leaves. 



Estimated relative leaf- 

 surface. 



Pot 1. 



Pot 2. 



Pot 3. 



Pot 4. 



Potl. 



Pot 2. 



Pot 3 



Pot 4. 





144 



140 



120 



164 



100 



67 



58 



128 





188 



200 



184 



216 



143 



172 



158 



242 





244 



300 



244 



280 



170 



249 



245 



328 





382 



540 



390 



434 



267 



481 



434 



463 



It is thus seen that, on September 25, after it had been observed 

 that the plants in pots 2 and 3, with the soil-extract seeding, had 

 begun to show a darker green colour than those in pot 1 without the 

 soil-extract, they nevertheless, up to that date, showed both a less 

 number of leaves, and considerably less leaf-surface, than the plants 

 in pot 1. It is not very clear why the plants with the soil-extract seeding 

 should have remained so long in a comparatively backward condition. 

 It may be that the result was only accidental, depending on the 

 character of the seeds, or on the fact that pot 1 stood at the southern 

 end of the row, and nearest the glass. The alternative is that, in the 

 early stages of development of the organisms supplied in the soil- 

 extract, and of the resulting nodules, the growth of the main plant 

 was, in some way, retarded. The figures show, however, that, from 



