is practically impossible to ascertain as we do not know with certainty 
when ovulation takes place, and should we know this it remains to 
determine how soon after ovulation the sperm cell enters the ege. 
We do know in a general way that conception probably begins near 
a menstrual period, but here again there is a difference of opinion 
of nearly a month as it is uncertain whether we should count from 
the last period or from the first lapsed period. 
I have brought together the evidence which bears upon this 
subject in Chapter Eight of the Manual of Human Embryology edited 
by Kemet and myself. In order to compare the obtainable data, it 
was necessary to consider anew the question of measuring human 
embryos. The material at my disposal was not as satisfactory as it 
should be as the records in the literature are not always carefully 
given, and those found with various collections are subject to con- 
siderable variation due to inaccuracy, and lack of uniformity, in making 
measurements, not to speak of the shrinkage of the specimens in the 
fluids in which they have been preserved. Often the measurement 
is made after the specimen has shrunken greatly or it is determined 
after it has been cut into serial sections. It is clear therefore that 
the measurements of the small preserved specimens may fall short of 
those of the fresh specimen fully 50 per cent. Thus, an eight milli- 
meter embryo may shrink to four which, according to my curve in 
the Manual (Fig. 147), is thirty-three days old while according to the 
convention of His it is only twenty-one days old. In all probability 
a specimen that measures eight millimeters fresh is nearer forty days 
old. Here then there is a discrepancy of three weeks to be accounted for. 
Von Bagr has stated that the evidence obtained from studying other 
mammalian embryos is probably the best criterion in determining the 
age of the human. Unfortunately the facts obtained from the lower 
animals leave much to be desired, but this much is true; the time 
of fertilization can be determined with considerable precision. Were 
it not for this evidence, a probable error of three weeks in human 
embryos under ten millimeters long would still remain which could 
not be corrected. An error as large as this could not possibly be 
observed at the end of pregnancy, but one of three weeks is very 
conspicuous when the embryo is small. 
In order to construct the curves given in the “Manual” there 
were measured one thousand specimens; one-half of which were ac- 
companied with menstrual data. The great fluctuation in the measure- 
