Vaccination against smallpox was a time-sensitive, sequential procedure that crossed the boundary between animals and humans. The goal was to immunize the human body against smallpox by implanting another virus, Variolae vaccinae, which belonged to the same family as the smallpox virus but resulted in a much milder case of illness. Although Variolae vaccinae was long identified with the cowpox virus, recent genome sequencing has shown that it was in fact a form of horsepox capable of infecting both cows and humans [Damaso 2017; Hirokawa 2017]. Moreover, even in the early days of vaccination there were a variety of strains circulating under the label of Variolae vaccinae. However, this module uses the term "cowpox vaccine" for Variolae vaccinae because the smallpox vaccine was assumed to be cowpox (in Japanese: gyūtō) in both Europe and in Japan and strongly associated with cows in the popular imagination [Rusnock 2009, p. 20]. 

In the late eighteenth century, several physicians across Europe proved the protective properties of Variolae vaccinae, and British doctor Edward Jenner pioneered a method that involved extracting lymph from a liquid-filled pustule on the arm of a cowpox patient and transferring it to the arm of an uninfected person. Because Variolae vaccinae was not contagious among humans, physicians needed to intervene. The vaccinator typically made several shallow incisions with a lancet on the upper arm of the receiving child to imbed the vaccine into the skin. The number of incisions depended on the age of the child. Before the procedure, physicians had to confirm that the child was not currently suffering from any other illness such as scabies that might interact with the vaccine and interfere with immunization.

Next, vaccinators had to monitor the child's reaction. After three or four days, they checked whether pocks had appeared at the incision site. Between six and eight days after transfer, the pocks were considered ripe for extraction, and vaccinators could decide to use the child as a supplier for vaccinating other children. This process consisted of puncturing a pustule, removing lymph with a lancet, and transferring the lymph to an incision in the arm of an unvaccinated child. Such extraction was necessary in the nineteenth century because cowpox and other similar pox forms, unlike human smallpox, were rarely encountered in nature [Rusnock 2009, p. 18-22]. In Japan, physicians used the term tōbo ("pox base;" literally: "pox mother") for children who provided lymph for further vaccinations. The amount of lymph that could be extracted from one "pox base" was small. ... The author of a Japanese vaccination textbook estimated that virus reconstituted from one scab was sufficient to vaccinate about four or five children [Naitō hiroku]. This number imposed another limitation on the work of vaccinators.

Part of the vaccinator's job was to detect so-called "false pocks" (or "spurious pocks"). In some cases, vaccinated children developed pocks that superficially resembled genuine cowpox but in fact constituted reactions to other infections such as syphilis or bacterial contamination. Proper classification of "false pocks" was crucial because the children who developed them remained unprotected from the smallpox virus and might spread diseases to others if used as "pox bases." If a child was found to display false pocks, vaccinators usually decided to revaccinate and repeat all the necessary steps. The need for a follow-up visit to detect the quality of pocks also made it easier for vaccinators to recruit some children as "pox bases" for the perpetuation of the vaccine.

Vaccination required careful attention to time and space. To succeed, physicians had to prepare and expose a human body to the cowpox virus and then transfer the virus to another suitable host at just the right point in time. Ideally, both bodies would be present at the same place at time of transfer so that the lymph would not be exposed to the elements for too long. To perpetuate the vaccine in one place, physicians needed to be careful not to vaccinate too many people at once, because only unvaccinated bodies were capable of producing a reaction and the vaccine could be lost if such bodies became unavailable at any point during the year [Rusnock 2009, p. 30].

Physicians also experimented with glass and other containers to preserve scab and lymph for later use and transport them over long distances. Although they sometimes succeeded, they considered these methods as far less reliable than arm-to-arm transfer between children.

In 1874, the new Meiji government established a state-funded institution to harvest vaccines directly from cows by injecting cows with Variolae vaccinae (i.e. horsepox) [Soekawa, p. 83; Jannetta 2009, p. 135-137]. Until that point, the smallpox vaccine in Japan was reproduced exclusively through person-to-person transmission.