on subjects pertaining to his craft—­his comments on play-making most especially,—­was illuminating and judicious.  I have been privileged to read the comments sent by him to Professor Matthews during the period of their collaboration together over “Peter Stuyvesant;” they are practical suggestions, revealing the peculiar way in which a dramatist’s mind shapes material for a three hours’ traffic of the stage—­the willingness to sacrifice situation, expression—­any detail, in fact, that clogs the action.  Through the years of their acquaintance, Howard and Matthews were continually wrangling good-naturedly about the relation of drama to literature.  Apropos of an article by Matthews in The Forum, Howard once wrote: 

I note that you regard the ‘divorce’ of the drama from literature as unfortunate.  I think the divorce should be made absolute and final; that the Drama should no more be wedded to literature, on one hand, than it is to the art of painting on the other, or to music or mechanical science.  Rather, perhaps, I should say, we should recognize poligamy for the Drama; and all the arts, with literature, its Harem.  Literature may be Chief Sultana—­but not too jealous.  She is always claiming too large a share of her master’s attention, and turning up her nose at the rest.  I have felt this so strongly, at times, as to warmly deny that I was a ‘literary man’, insisting on being a ‘dramatist’.

Then, in the same note, he adds in pencil:  “Saw ‘Ghosts’ last night. 
Great work of art!  Ibsen a brute, personally, for writing it.”

In one of the “Stuyvesant” communications, Howard is calculating on the cumulative value of interest; and he analyzes it in this mathematical way: 

So far as the important act is concerned, I have felt that this part of it was the hardest part of the problem before us.  We were certain of a good beginning of the act and a good, rapid, dramatic end; but the middle and body of it I felt needed much attention to make the act substantial and satisfactory.  To tell the truth, I was quietly worrying a bit over this part of the play, while you were expressing your anxiety about the 2nd act—­which never bothered me.  There must be 2nd acts and there must be last acts—­audiences resign themselves to them; but 3rd acts—­in 4 and 5 act plays—­they insist on, and will have them good.  The only exception is where you astonish them with a good 2nd act—­then they’ll take their siesta in the 3rd—­and wake up for the 4th.

This psychological time-table shows how calculating the dramatist has to be, how precise in his framework, how sparing of his number of words.  In another note, Howard says: 

This would leave the acts squeezed “dry”, about as follows:—­Act I, 35 minutes; Act 2, 30; Act 3, 45; Act 4, 20—­total, 130—­2 hrs., 10 min., curtain up:  entr’acts, 25 min.  Total—­2 hrs., 35 min.—­8:20 to 10:55.

There are a thousand extraneous considerations bothering a play that never enter into the evolution of any other form of art.  After seeing W.H.  Crane, who played “Peter Stuyvesant” when it was given, Howard writes Matthews of the wisdom shown by the actor in his criticism of “points” to be changed and strengthened in the manuscript.