634 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1907. 
1876 to 1886, since that time a reduction of more than 30 per cent has 
been noted. These figures show what may be accomplished in reduc- 
ing the dangers of infection with tuberculosis by a régime of educa- 
tion, improved conditions of living for the poorer classes, and the 
segregation in hospitals and sanitoria of any considerable number of 
the infective tuberculous during the most dangerous period of the 
disease. 
The discovery of the microbic agent of tuberculosis naturally 
awakened the hope that a specific means of treating and, possibly, of 
preventing tuberculosis might now be found. The early years follow- 
ing the cultivation of the tubercle bacillus saw no realization of this 
hope, and to-day we are still far from the desired goal. However, the 
prodigious labor which has been expended in the search for a means 
of protection against infection with the tubercle poison has not been 
wholly devoid of results. 
In an address of this kind it is not practicable to deal with the 
separate contributions, in detail, which the many workers have made 
to the subject of immunity in tuberculosis. The most that can be 
accomplished is to bring together the more important results of all 
the workers and, after having assembled them, to judge of their value 
and to consider, possibly, in what important respects they are still 
imperfect. I can not do better, at the beginriing, than to remind you 
that the successful point of departure has been the discovery that 
variations in type and in virulence exist among tubercle bacilli. The 
earlier view which taught that the tubercle bacillus is a micro- 
organism of uniform, and fixed virulence has been shown to be erro- 
neous, first by the discovery of variations according to certain origins, 
and second by a gradual decline in pathogenic power suffered by 
certain strains through long cultivation outside the animal body. 
The animals which have been of special use for tests of immunity 
are rabbits, cattle, and goats. The guinea pig, which furnishes an 
almost ideal animal for the detection of tuberculosis, because of the 
sensitiveness of its reaction to inoculations with tubercular material, 
fails, for the same reason, to be a highly suitable animal in which to 
carry out tests of immunity; and yet it has been employed with some 
success. 
The first important contribution to the subject of experimental 
immunity in tuberculosis was made by Koch in connection with his 
researches on tuberculin—a product of the growth in broth of tubercle 
bacilli, freed from the bacilli and concentrated. In spite of the 
failure of tuberculin to bring about a favorable issue in all cases of 
human tuberculosis in which it is administered, it still remains a 
useful, perhaps the most useful, strictly medicinal agent employed for 
the treatment of tuberculosis. But the sum of its useful properties is 
not embraced in its employment as a therapeutic substance; it is also 
