What's so difficult about $\pi_{15}(SO)$?

Regarding the table of $SO(n)$s-of-origin in Davis+Mahowald (if you can get MathSciNet), is there a good reason that it should take longer for $\pi_{15}(SO)$ to be representable than $\pi_{19}(SO)$, which is already fully representable in $SO(11)$?

-