A bit of a conundrum, if you get my 'Drift'.

I am inclined to agree with "Madsenshooter" and Dick Hosmer, that with U.S. rifles and their right-hand rifling twist, 'drift' must be to the right.

The model 1901 rifle sight was definitely built to correct for drift by moving the eye-piece to the left, as range increases, which would indicate drift is to the right.

By contrast, the short-careered, model 1898 rifle rear-sight, a.k.a. 'The Dickson sight', had its three notches slightly off-set to the Right of the center-line, (which would indicate a correction for bullet drift to the left).

The model 1902 sight, a rendition of the Dickson sight, has its single notch on the centerline, hence no built in correction for bullet drift. 

As Mr. Hosmer might say, "With Krags, we are often presented with an inconvenient truth". But, I am not sure what the Truth is.

The 'range testing' for drift with The United States Magazine Rifle caliber .30, that was performed in 1895, 1896, and 1897, appears to have been under the supervision of Lt. T.C. Dickson. His name is the signature submitting the report.

It does sound to me from the content of the report, that they were having difficulty completing these 'shooting tests' and may have been getting surprising results.

The bullet drift tables submitted in the 1897 report were repeated in 1904 and 1917 government publications (attached pages).

Also, attached are a couple of pages from Dr. Hudson's practical work on rifle shooting from 1902. The Good Doctor understood sight picture, bullet drift, and trusted his 1901 sight!

I suppose it is possible the old drift tables were flawed, caused some embarrassment, were eventually ignored and worked around, but, they sure got reprinted a lot??? That's the government for you!

---