In the beginning of the $20$th century, the periodic table had $8$ columns, and not $18$ as today. Electrons and of course $spdf$ electrons were unknown. The numerous transition elements were included in this table by doubling the line between Argon and Krypton, then between Krypton and Xenon. It was going this way.
The $4$th period was made of two lines. So each square (each box) contained two elements written on two lines. The first seven atoms $\ce{_{19}K - _{25}Mn}$ occupied the columns $\ce{I - VII}$, and their oxidation state may always have been equal to the number of the column. Then the three atoms $\ce{_{26}Fe}$, $\ce{_{27}Co}$ and $\ce{_{28}Ni}$ did not belong to any column. We will come back to them later on. Then the seven next atoms $\ce{_{29}Cu - _{35}Br}$ had to be put in columns $1$ to $\ce{VII}$, under the corresponding elements : $\ce{_{29}Cu}$ in the same box and under $\ce{_{19}K}$, $\ce{_{30}Zn}$ in the same box and under $\ce{_{20}Ca}$, etc. up to $\ce{_{25}Mn}$ above $\ce{_{35}Br}$. With this system all atoms had at least one oxidation state equal to the number of their column. The same procedure was adopted for $5$th period, with $\ce{_{37}Rb}$ and $\ce{_{47}Ag}$ in the first column,$\ce{_{38}Sr}$ above $\ce{_{48}Cd}$, etc.
This table was useful, except nobody knew what to do with the three atoms $\ce{_{26}Fe}$, $\ce{_{27}Co}$ and $\ce{_{28}Ni}$ having all valence $2$, and of course of the six corresponding atoms of the platinum group, immediately under them : $\ce{_{44}Ru}$ to $\ce{_{46}Pd}$ and $\ce{_{76}Os}$ to $\ce{_{78}Pt}$. There was plenty of discussion at that time about what to do with them. Usually these elements were "rejected" in the 8th column, with the noble gases. or they were put in a separate column made specially for them. This problem was finally solved with the discovery of the $spdf$ electrons.