2g6 DISCOVERY REPORTS 



the porpoise. These Hmited curves may be extrapolated forwards to birth and backwards to conception. 

 It is usually possible to estimate the average neonatal length with some accuracy, but the effect of 

 extrapolation on the estimated time of birth can be greatly influenced by uneven sampling caused by 

 differential migration of pregnant females (see below). The difficulty about finding the average date 

 of conception by this method is that the rate of growth in the first 2 months is very slow. One approach 

 is to take a species such as the humpback whale for which the average dates of conception and calving 

 have been established, as well as the growth of the foetus over part of gestation. Then by analogy 

 the growth rate in the early months can be estimated for other species. Freehand extrapolation and 

 analogy do not provide a very firm basis, and the detailed conclusions drawn from growth curves 

 constructed in this way may be greatly in error. 



A better approach is to see whether it is possible to make any mathematical generalizations about 

 foetal growth by combining the data which are available from all sources, for a variety of species, 

 concerning foetal lengths, and mating and calving seasons. 



GESTATION PERIOD 



Text-fig. I. Diagram illustrating relation between foetal weight (or length), gestation period, /„ and a 



(after Huggett and Widdas, 1951). 



In this respect the work of Huggett and Widdas (1951) provides a starting point. These authors 

 showed that for a variety of mammals of widely different mammalian orders the cube root of the foetal 

 weight gave a linear plot with age for all except the first part of pregnancy. They suggested that the 

 beginning of this steady state of growth might be correlated with the full establishment of the placental 

 circulation. Previously to this work they had shown that foetal length increases linearly with age. 



Their hypothesis is expressed in the general formula W^ = a{t—t^ and since WccL^ it follows that 

 L=a{t — t^. The term a is constant in respect of any particular species and is called by them the 

 ' specific foetal growth velocity '. It is the slope of the line relating foetal weight, W^ (or foetal length, L) 

 to the age after conception {t) in that segment of the growth curve where the relationship is linear. 

 The term t^ ' is the intercept where the linear part of the plot, if produced backwards, cuts the time 

 axis' (see Text-fig. i). This term has 'no clear biological significance in foetal development, but if 

 the numerical value of t^ can be estimated by analogy with other mammals . . . then one known weight 

 {W) and time from conception {t) would be sufficient to determine the value of a for the mammal 

 concerned '. Their estimate of t^ for different animals is based on the observation that ^0 increases as 

 the gestation time lengthens but forms a decreasing fraction of the total gestation time. Huggett and 

 Widdas use arbitrary estimates of t^^: for gestation times from o to 50 days f,,^ 0-4 x (gestation time); 

 50-100 days, fo — 0"3X (gestation time); 100-400 days ^q^ 0-2 x (gestation time); and over 400 days 

 fo-o-i X (gestation time). These estimates of ^o apply to weight data and for length t^ is slightly less. 



