Occasional references are found to the existence of the tubular system in a few fish enamels as early as in the latter half of the 19th century [1, 2]. In recent published researches, OCKERSE in 1961  by the use of replica method light microscopically and ISOKAWA et al. in 1964  historadiographically and telemicroscopically respectively reported the tubular structure found in a fish enamel to be a true tubule. The latter investigators predicted that the inner wall of this enamel tubule might consist of irregularities full of concavities and convexities. The present report, which serves as a follow-up study of a previous publication, makes use of scanning electron microscope.
Since 1954, ORLAND et al. [1, 2] have carried out a series of experiments by the use of germ-free animals  and reported the facts that the caries in a rodent family was bacterially caused and that enterococcus was one of the bacterial strains responsible. In their wake, FITZGERALD et al. succeeded in inducing the caries by a slightly aerobic and non-proteolytic streptococcal strain which was isolated from the agnotobiotic rat . However, both ORLAND and FITZGERALD were cognizant of the fact that this animal existed in somewhat unnatural conditions. In 1960, therefore, FITZGERALD et al.  removed 5 streptococcal strains from the carious cavity of hamsters fed on a carious-inducing diet and, by planting the strains either singly or in combination in caries-inactive rats, they were successful in this experimental induction of the caries. At the same time, they inoculated lactobacillus and diphtheroid bacillus which were isolated from the carious cavity singly into caries-free animals. However, this attempt proved a failure. The authors had isolated about 330 streptococcal strains from the oral cavities or carious dentin of Wister rats and of them, they selected Str. bovis which was consiered to be closely linked to the cariogenecity. Interestingly enough, there was found none of these strains that corresponded to 5 streptococcal strains used by FITZGERALD or to those which ORLAND and FITZGERALD employed to induce successfully the caries in germ-free animals. On the strength of this fact, the authors were inclined to think that streptococcus responsible for the caries in a rodent was not of a single kind but rather there must be streptococcal strains responsible for the caries which were proper to the species of animals and their growth environment. The present experimental study was conducted so as to substantiate this inference above.