PAIRING OF SALMON. ,57 



the impregnation of the roe takes place after ex- 

 clusion ; and this seems reasonable, considering 

 the immense number produced from one female, 

 amounting, as I have stated before, to about 

 600,000. It appears indeed almost impossible, that 

 so large and compact a mass can be impregnated 

 in the body of the female, the male having no or- 

 gan with which to do it ; and no part of the roe 

 can be impregnated without an actual contact with 

 the seminalis masculinum, either before or after ex- 

 clusion. As the male has been seen to cover the 

 pit which has been dug for the female, and the 

 water has been observed to be immediately co- 

 loured, the probability is, that the natural office 

 of the male is performed on the ovaria, whilst they 

 are extended on a flat surface on the sand, and 

 when every egg is so placed as to be able to receive 

 the masculine influence. Nor is it at all improbable 

 that the female should make a second and a third, 

 or more beds, for the reception of the whole of the 

 pea, which it is hardly possible or likely that she can 

 discharge at one time or at one place, for the oper- 

 ation takes a considerable time in performing. I 

 am warranted in this conclusion from what follows 

 we must make a comparison to something : taking 

 it as a datum that the roe consists of 600,000 pea, 

 now, as the natural orifice is small, only one pea 

 can be emitted at a time, and these must be ad- 

 justed and deposited in proper and regular order 

 and situations. They are not let out of her body 

 all at once, as at a sluice, in a promiscuous, irregu- 



