result that neither party is
entirely satisfied. Our experience during the year has brought to
our attention some defects in all of the specifications now before
us, and acting under the impression that there is a distinct
feeling that we should revise our specifications, we offer the
attached specifications for your consideration. Our Association has
no specification for Open-Hearth Steel Rails, and in order to
comply with the instructions, a specification for Open-Hearth Steel
Rails is included.

"We believe it necessary to submit a sliding scale for the
percentages of carbon and phosphorus, which provides for increasing
the carbon as the phosphorus decreases. The fixing of this scale
properly is a matter requiring care, and we admit that our
knowledge on the subject is limited. The American Railway
Association specification calls attention to this matter in the
following words: 'When lower phosphorus can be secured, a proper
proportionate increase in carbon should be made.' The amount of
increase is not provided for in the specifications, and this
appears to us to be necessary in order to secure uniformity of
practice; otherwise, the fixing of these percentages becomes a
matter of special arrangement. Bessemer rails are being furnished
regularly with phosphorus under the maximum allowed, and where this
is done, the carbon should be raised above the higher limit now
fixed in our specifications, or a soft and poor wearing rail will
result; yet this condition has not been fully guarded against in
rails furnished under existing specifications. The lower and upper
limits for carbon have heretofore been fixed with the intention
that the mills furnish rails with a composition as near between the
two limits as possible. The mills, however, in order to meet the
prescribed drop tests with the least difficulty, keep both the
carbon and manganese as nearly as possible to the lower limits,
with the corresponding result that a generally poor-wearing rail is
furnished.

"Some roads have prescribed the limits of deflection to be allowed
under the drop test. With our present knowledge, we believe that we
should fix a minimum deflection to eliminate brittle rails and to
secure greater uniformity of product; also maximum deflection to
eliminate soft ra