similar to that which we obtained in 1938 for breakwater dikes. 

 In effect, the American formula isj 

 P+ = Rt 5H 3 



(3-1)3 (A _ r) 3 



and ours: 



P = 



NA 3 d 



(cosoc - sinoc)3 (d-1)^ 



in which 



p^ = p = weight of the individual stones 



H = A = 2h = wave height 



a = d = relative density of the material 



/tl = natural batter of the rock fill e£ 1 



r = tan ex = slope of the rock fill 

 Rt_, and N being the coefficients. 



.Expressing the American formula in our notation, it becomes; 



P = E+ cos-'oc 



A J d 



_-n3 



(cos a - sin a ) (d-1)- 



which is ours except only that it includes the factor cos~oc in the 

 coefficient. 



It is a matter of record that on establishing our formula it 

 was already indicated , that the coefficient should be able to vary 

 with the data of the problem. 



Practically the angle oc , which varies most for the upper 

 part of the dike, will not vary much, for from cotana,ci3, correspond- 

 ing to present rock fill dikes, it cannot get much steeper than cotan 

 (X z - 2, even in the reflecting dikes, because of the enormous 

 weight of the stones this necessitates. Between those maximum 

 limits the relation is: 



nos^OCi ~ 



1 ~ 1.2 



cos 



3oc 



which would represent only a small difference in the weight of 

 th e sto nes and even less in their size, whose relation would be 

 -?/l.2 = 1.06. Only direct observation can determine properly N 

 or Rl, including cases of very steep batters. 



Concerning this last coefficient, it should be indicated 

 that the expression for it determined in the American report can 

 have reality only when the protecting blanket is made up of 

 parallelopiped blocks perfectly aligned, with their three 

 dimensions horizontal, normal to the slope, and following the line 



22 



