110 CARNEGIE INSTITUTION OF WASHINGTON. 



The material and data at our disposal are by no means entirely satis- 

 factory, inasmuch as the records in the literature are not always care- 

 fully given, and those found among the records of various collections 

 are subject to considerable variation, owing to inaccuracy and lack of 

 uniformit}^ in making measurements, not to speak of the shrinkage of 

 the specimens in the fluids in which they have been preserved. Thus 

 the measurement was often made after the specimen had shrunken 

 greatly or was determined after the embryo had been cut into serial 

 sections. Hence it can readily be understood why measurements of 

 the small preserved specimens sometimes fall short of those of the fresh 

 specimens by fully 50 per cent. Thus, an 8-millimeter embryo may 

 shrink to 4 millimeters, which, according to the curve of growth given by 

 Mall, would belong to a 33-days-old pregnancy, whereas, according to 

 the convention of His, it would be 21 days old. In all probability a 

 specimen that, when fresh, measures 8 millimeters is nearer 40 days 

 old. Here there is a discrepancy of three weeks to be accounted for. 



In measuring young human embryos it must be emphasized again 

 that the crown-rump measurement, or sitting height, should be made 

 from the crown, over the middle of the mid-brain, to the lowest point 

 of the rump while the embryo is fresh, or, what amounts to the same 

 thing, after it has been fixed in formalin. This measurement should 

 be exact, as it represents the living and not the shrunken embryo. 



In recent years we have measured a large number of embryos, first, 

 fresh or in formalin, and again after dehydration; finally they will be 

 measured in paraffin. Nevertheless, if all the suitable specimens in 

 the literature are compared with the best in our collection, the varia- 

 tion in size at a given stage is still very pronounced. In order to give 

 the question a test, we photographed all of the good profile illustrations 

 given in the literature, so that the pictures of the embryos gave a 

 crown-rump measurement of about 50 millimeters. Then we had 

 prints, enlarged to this size, made from our own negatives. These 

 various photographs could now be compared from different stand- 

 points — that is, they could be arranged according to their menstrual 

 age, according to their greatest length, or according to their degree of 

 development. When the photographs were arranged according to 

 their menstrual age, no satisfactory classification could be reached. 

 When considered according to size the results were no better. Finally, 

 when grouped according to their stage of development, a more satis- 

 factory result was obtained. 



At first we arranged the 266 photographs of embryos, ranging from 

 2 to 25 millimeters, in 20 stages, taking the external features (branchial 

 arches, arms, and legs) as our guide. Since all of the photographs had 

 been marked arbitrarily to conceal the length of the specimens, and as 

 there were numerous duplicates in the collection to confuse us, we 

 continued arranging the photographs until the stages became thor- 



