One of continuing vulnerability of the No. 2 drill was the way main handle was attached to the frame.  This vulnerability manifested itself in cracks of the handle at the point of attachment to the frame.  It is especially notorious on the drills introduced sometime after 1906 when thicker handles were implemented, following changes in the frame design.  The new handle was much thicker at the point of attachment to the frame and carried wider, stepped ferrule.

Here are some pictures showing the problem on drill from c. 1938.

There are a variety of reasons for this persistent vulnerability and they are analyzed in depth by George Langford in his No. 2 drill type study.  Spend some time on his web site at GeorgesBasement.com - well worth it!  In addition to his finding, I believe there is one more reason for this problem.

The outside diameter of the frame end, the tenon, is 20/32 inch but the inside diameter of the ferrule is 24/32 inch.  In addition, the casting of this tenon does not have a sharp, 90 degrees angle to the edge of the frame and existing slope prevents a ferrule to fit snugly. 

Consequently, the ferrule is a free-floating part that does not serve its purpose.  I believe these parts should fit tight to each other and do not leave any wiggle room.  However, it is not the case on drills I have examined.  This creates overall pressures distribution problem during the drill use.

Here is a cut-out of the assembly.   As you can see, the work of controlling the drill is performed only by the wooden handle attached to the threaded rod, without any support from the ferrule or surfaces of the tenon and wood of the handle touching together.  In correct design and execution these gaps would be eliminated and provide for much better distribution of pressure on wooden handle.