e.g., (1) In pure Dravidian such as Tamil, a vowel must come before the initial র and ল, and according to this rule, we find that নগ্ন reduced to লঙ্গ has taken the vowel উ before it, in the formation of the word উলঙ্গ (cf. Oriya, উলগন্); (2) in Tamil 'শ' and 'চ' are not different letters and so the word শালা (house) has been reduced to চালা and in this very fashion we have formed the word চালা or চাল (roof or thatch) in Bengal. (3) কুঠার has been reduced to koḍāl and koḍāli in some Dravidian speeches, and we too, have given currency to exactly similar forms কোদাল and কোদালি independently of the Dravidians; (4) we get পেল from বল (force) in Tamil (and so also পলি from বলী, strong), and it is striking that our obsolete Bengali form পেলা পেলি (cf. বলাবল as a method in chess-playing in Mārāṭhi), which still exists in Oriya, and of which our mordern form is ঠেলাঠেলি, has been formed according to Dravidian method. I have given previously a list of Dravidian words as are in use in Bengali; I add a few more examples of those words which have been wrongly considered by some to be Sanskrit derivatives; they are; (1) Kaṭāl as a variant of খাড়ল or খাড়ি, indicates sea in Tamil, and it is this word which is used in Bengali to signify the swelling of the sea, as অমবস্যার কটাল; (2) নড় (to move) of Tamil is exactly the word which is in use in Bengali and Oriya; (3) Pala (pronounced in Tamil almost as পাল) signifies many in Tamil, and it is this word which is in use in Bengali to signify a flock or herd, as in একপাল গরু; (4) the Bengali word মাগি has no doubt come from মাউগি (wife), still in use in Behar, but the original word is Dravidian মুক্কণ or মোক্কন্ or মোগ্গন্ (Kôta dialect); the Oriya form মাইকিনা is closer to the original; I should also note that our next-door neighbours, the Oraons of Chutia Nagpur, who have given us the words খোকা and খুকি (কোক্কাই হাদু, কুক্কি হাদু), have the use of the word মুক্কা in their speech.
