life leads before long to
a consuming disease; and as Agnivarna is even then unable to resist
the pleasures of the senses, he dies. His queen is pregnant, and she
mounts the throne as regent in behalf of her unborn son. With this
strange scene, half tragic, half vulgar, the epic, in the form in
which it has come down to us, abruptly ends.

If we now endeavour to form some critical estimate of the poem, we are
met at the outset by this strangely unnatural termination. We cannot
avoid wondering whether the poem as we have it is complete. And we
shall find that there are good reasons for believing that Kalidasa did
not let the glorious solar line end in the person of the voluptuous
Agnivarna and his unborn child. In the first place, there is a
constant tradition which affirms that _The Dynasty of Raghu_
originally consisted of twenty-five cantos. A similar tradition
concerning Kalidasa's second epic has justified itself; for some time
only seven cantos were known; then more were discovered, and we now
have seventeen. Again, there is a rhetorical rule, almost never
disregarded, which requires a literary work to end with an epilogue in
the form of a little prayer for the welfare of readers or auditors.
Kalidasa himself complies with this rule, certainly in five of his
other six books. Once again, Kalidasa has nothing of the tragedian in
his soul; his works, without exception, end happily. In the drama
_Urvashi_ he seriously injures a splendid old tragic story for the
sake of a happy ending. These facts all point to the probability that
the conclusion of the epic has been lost. We may even assign a
natural, though conjectural, reason for this. _The Dynasty of Raghu_
has been used for centuries as a text-book in India, so that
manuscripts abound, and commentaries are very numerous. Now if the
concluding cantos were unfitted for use as a text-book, they might
very easily be lost during the centuries before the introduction of
printing-presses into India. Indeed, this very unfitness for use as a
school text seems to be the explanation of the temporary loss of
several cantos of Kalidasa's second epic.

On the other hand, we are met by the fact that numerous commentators,
living in different parts of India, know the text of only nineteen
cantos. Furthermore, it is unlikely that Kalidasa left the poem
incomplete at his death; for it was, without serious question, one of
his earlier works. Apart from evidences of style, there is t