produced in Panjabi, and the difference marked by an extra curve in the ḍh, while the ph is indicated by the simple ढ. Kashmiri squares the corners of the ḍh, and exactly follows the Gupta in its ph. Kutila has adopted a very different type in its फ, which has been followed by all the other alphabets. The Tibetan ཕ shows how by a different order of making the strokes the ढ of the earlier alphabets might pass into the Bengali ফ, and thence into Nagari फ; so much depends upon the order followed by the pen in forming the letter. Let any one who doubts this try the experiment of forming the Sanskrit letters backwards, beginning where the pen generally leaves off, and after writing the letter quickly half a dozen times he will be surprised to see how far it has deviated from its original shape.
Bh is the Kutila form rounded and written as though the central curve were a loop; m differs only in having the top open, to distinguish it from स, which in Panjabi is written like the Nagari म, owing to the Gupta character from which it is derived not having the little tail which marks the स.
R also lacks the tail, and thus approaches the Gupta rather than the Kutila type.
The v assimilates more to the Vallabhi form than any other; and the h is Gupta.
In a large number of instances the Kutila differs from the Gupta type only by the addition of a little tail at the right-hand lower corner. This tail being regarded as the continuation of the right-hand line of a letter has resulted in the vertical straight line so characteristic of Devanagari letters, such as प, थ, य, व, in none of which has the Gupta character any tail, or consequently anything to give rise to a straight stroke. The Panjabi character probably took its rise from the Gupta, or it might be more accurate to say that the earlier character of Asoka underwent modifications, the type of which is uniform throughout India, down to the Gupta era, but that
