Not only is the general average in this book low, but the lowest chapters are the most dramatic, contain the most dialogue and the most character-photography. The higher chapters are those in which this fades away into a burlesque monologue; and the highest chapter in the book (Chapter XIV, average .364) is one of the few chapters in the whole work which do not contain a single word of dialogue, but resembles rather the mock-heroic tone of The Gull's Hornbook.

To conclude, then, the further we examine the works of Webster and Dekker, the more evidence we find that the former almost invariably1 has a high word-average, and that the latter, as a playwright at least, consistently has a very low one. Consequently, when we find five or six scenes in the citizen-comedies

1 The following table shows the word-averages of the few short poems which Webster has left us:

with Webster's high average, nineteen or twenty with Dekker's low average, and a wide, almost empty gap between, we are justified in assuming that Webster wrote-not all perhaps-but certainly the larger part of the former scenes; and that Dekker wrote all or nearly all of the latter. We might expect to find some casual trace of Dekker's revising touches in the higher scenes, or of Webster's in the lower; but we certainly should not expect that either of these writers would have any considerable part in such scenes. The word-test, then, would seem to indicate the following division of scenes:

Webster's Part.

Westward Ho, I. 1 and III. 3.

Northward Ho, I. 1; II. 2; V. A.

Dekker's Part.

Westward Ho, I. 2; II.; III. 2; III. 4; IV; V. Northward Ho, I. 2; I. 3; II. 1; III. 2; IV; V. B.

Westward Ho, III. 1.

Northward Ho, III. 1.

Uncertain.

It will be noticed that this test, while not going quite so far as Mr. Stoll, in the main agrees with him in giving by far the larger part of the two plays to Dekker.

Of course we realize that this test, like any test, is not absolutely infallible. But when we consider that the plays of Dekker which we have discussed range through every walk of life and every phase of human emotion, from the king's palace in Match Me in London to the home of the day laborer in the Shoemaker's Holiday, from the pure presence of Jane to the reeking atmosphere of the brothel, from the tragedy of Hippolito's grief and the Duke's dignified

anger to the vulgar mirth of prentices and bawds, and that through all this he steadily keeps a low word-average, while Webster has a high one, we must admit that this test has a decided value.1 And

if we can support this test by other evidence agreeing with it, we may hope, in part of the scenes at least, to arrive at what a court would call reasonable proof. A considerable amount of such supporting evidence will be found in the following chapters.

1 One interesting proof of the value of this test is found by tracing the word-average of a single character from scene to scene. If a single character is the work of one author, we should certainly expect such a character to be reasonably consistent in vocabulary. On the contrary, the language of Birdlime and Justiniano in Westward Ho, and of Bellamont and Mayberry in Northward Ho, varies remarkably.

Birdlime

Westward Ho. Solid lines. Words. Average.

Bellamont's rather high average in IV. 1 may possibly be due merely to his affected style, since he is talking here in poetry; it is probably a burlesque on Chapman's grandiloquence. But the radical extremes in the other scenes cannot be explained.

Mayberry.

Northward Ho. Solid lines. Words. Average.

TTIL

II. 2

57

IV. 1

62

IV. 3

15

A

37

V. B

51

7

5

.130 .083

20

.351
.145

1

8

(too short)
.216 (short)

5

.098

In I. 1 Mayberry's low average in a high-average scene seems rather suspicious, especially as his total part is nearly 60 lines in length. His change of heart in II. 2 is certainly peculiar.

CHAPTER III.

THE PARALLEL-PASSAGE TEST.

WESTWARD HO.

The value of parallel passages as a test of authorship has been almost universally recognized. Now if this is true of authors in general, it certainly is true of Dekker and Webster; for each of these men has the habit of repeating his own phrases from play to play, until it becomes a positive mannerism. Mr. Stoll has already published a long list of almost verbatim parallels from the various plays of Webster,1 and a still longer list of similar parallels between his own plays might easily be gathered from the works of Dekker. On the other hand, while each of these writers constantly repeats himself, neither shows any tendency in later plays to repeat the phraseology of the other. Five or six parallelisms 2 may be pointed

Our weak safety

Runs upon enginous wheels.

Whore of Babylon, 1607, Sig. C. 2:

For that one act gives, like an enginous wheel,

Motion to all.

I have been unable to find more than two or three other parallels in carefully reading all the works of both authors.