A spinal cord segment consists of a portion of the spinal cord proper with all of its rootlets that join to form the associated pairof spinal nerves. Number of rootlets that forms a spinal nerve and the length of the cord segments vary from one level of the cord to the next. There is no clear demarcation between segments of the spinal cord except for the interval between root fibres of theadjacent nerves. Each spinal cord segmentis numbered as per the attachment of the paired spinal nerves and the cord consists of as many segments, as there were pairs of spinal nerves. The distinct differences in the number of spinal cord segments between species reflected individual morphology and suited the functional requirements of the animal. Since the information about the attachment of spinal nerve rootletsin the caprine spinal cord is very scanty,this study was undertaken to elucidate the same during prenatal development in goat.

The study was conducted on 52 goat foetuses of different ages. Specimens were fixed by 10 per cent neutral buffered formalin. Root emergence length (REL) was measured as the extent of the dural surface occupied by the respective spinal nerve roots at the point of their emergence. Root attachment length (RAL) was the length of the cord to which the rootlets of each spinal nerve were attached, viz. distance from most rostral to the most caudal rootlet of both the dorsal and ventral roots of each spinal nerve. Inter root length (IRL) was the length of the cord surface devoid of rootlets,lying between the attachment of the roots of adjacent spinal nerves, viz. the distance between the most caudal and the most rostral rootlet of adjacent spinal nerves.

There were 36 pairs of spinal nerves in the foetuses under study: 8 cervical, 13 thoracic,6 lumbar,4 sacral and 5 coccygeal.According to Ghoshal (1975), the number of spinal nerves varied depending on the thoracic, lumbar and sacral vertebrae present in sheep and goat.

These spinal nerves had a common pattern of structure and were segmentally arranged. Each arose from the spinal cord by a dorsal root and a ventral root. The cervical nerve roots were shorter but nerve roots further became longer caudally.The dorsal roots had a small number of large coarse,rootlets whereas the ventral roots consisted of many small rootlets with less tensile strength. This was in accordance with findings of Thomas and Combs (1962) in cat.

The emergence length of the dorsal and ventral roots of spinal nerves was measured in the last two months of gestation. The root emergence of the dorsal roots was always greater than that of the ventral roots. So, the dorsal roots pierced the duramater over a larger area compared to the ventral roots. Histological observations of Sanomiya (1927)indicated that there were more of glial tissues and a greater number of nerve fibres in the dorsal roots,which might account for this gross difference.

The root emergence length was slightly greater in cervical, rostral thoracic and lumbar regions of the cord. It decreased rapidly through the sacral region. The greatest root emergence length was measured dorsally at 8th cervical segment as 1.980+or-0.020 mm at fourth and as 4.662+or-0.078 mm fifth month of gestation. Ventrally, the greatest root emergence length was observed at first, second, third and eighth cervical segment as 1.050+or-0.085 mm at fourth month and at third cervical segment as 3.554+or-0.258 mm at fifthmonth of gestation.

In the thoracic region, dorsal root emergence length was greatest at first thoracic segment in fourth and fifth month. In lumbar region, the length was less variable at fourth month but in fifth month, fifth lumbar segment had the greatest length. In the sacral region, first sacral segment had the greatest length in both the agegroups.

Ventrally, the same was true with first thoracic and first sacral segments, but it varied at the lumbar region. Greatest root emergence length was recorded at third cervical, first thoracic, fourth lumbar and first sacral segments in sheep (Rao, 1990).This observation is identical with those recorded in thoracic and sacral regions in the present study. Ventral root emergence legth remained less variable in the thoracic region. This is in accordance with the findings of Sharma and Rao (1971)inbuffaloes and Sharma et al. (1973) in goats.

In general, ventral roots originated over greater area when compared to corresponding dorsal roots. This is in accordance with the study by Thomas and Combs in cat and monkey (1962 and 1965 respectively) and Sharma and Rao (1971) in buffalo. Sharma et al. (1973) reported that this observation was true in goats in prelumbar segments. In the present study, the lumbar and sacral segments presented ventral root attachment length either greater than or equal to those at the dorsal aspect.

The maximum root attachment length was associated with the longestspinal cord segment,second cervical, dorsally in fourth and fifth month. It measured 2.444+or-0.176 mm and 7.443+or-1.060 mm respectively at fourth and fifth month. On ventral aspect it was at third cervical segment as 2.600+or-0.191mm in fourth and second cervical segment as 7.571+or-1.295 mm at fifth month of gestation. The root attachment length decreased after sixth lumbar segment towards the coccygeal region.

The minimum dorsal and ventral root attachment length were observed in the coccygeal region in both the age groups studied. In general,the ventral roots originated over greater area compared with corresponding dorsal roots. 

A comparison between dorsal root attachment length and segment length at fifth month of gestation showed that both these parameters followed a similar pattern showing a relation between the two (Fig. 1). So there stands a relation between the root attachment length and segment length. Being greater inthe longer ones and decreased in the shorter ones. So as stated by Sharma etal. (1973), the root attachment length contributes to the segment length. Fig.1 Segment length, dorsal root attachment length and inter root length of spinal cord at last month of gestation.

The greatest root attachment length was recorded at third cervical segment in dogs (Fletcher and Kitchell, 1966), at second cervical segment through thirdcervical segment in buffalo (Sharma and Rao,1971) and at third cervical segment in goats (Sharma et al., 1973) and Sheep (Rao, 1990)In difference, Taluja et al. (1983) reported maximum RAL at seventh cervical segment in all age groups in goat foetuses. Taluja et al. (1999) also reported maximum RAL at fifth, seventh and eighth cervical segments in calves.

RAL increased from eleventh thoracic to first lumbar in fourth month and from eighth thoracic to firstlumbar segments in fifth month of gestation in the present study. Sharma et al. (1973) reported that this second increment of RAL lay at eleventh thoracic to first lumbar dorsally in goats.

The maximum IRL was noticed at second cervical segment in the fourth and fifth month as 4.500+or-0.129 and 6.490+or-0.643 mm respectively. A similar condition of maximum IRL wasrecorded at second cervical segment in goat foetuses (Taluja et al.,1983),third cervical segment in sheep (Rao, 1990) and second cervical segment in calves (Taluja et al., 1999; Parmar etal.,2000).

Cervical enlargement region showed a decrease in inter root length in both fourth and fifth month (Fig. 1)rarely with, no measurable distance at eighth cervical segment. In the fifth month, lumbarenlargement region also showed a decrease in inter root length after sixth lumbar segment. Longest inter root length was at elevenththoracic segment at fourth month and at twelfth thoracic segment in fifth month. Minimum inter root length was at first thoracic segment. Similar pattern occurred in camel (Gholami et al., 1998).

The segments of cervical and lumbar enlargements had shorter length for root attachment and shorter interval between adjacent roots thanthe segments in other regions of cord in the present study.This isfound true with the findings of Fletcher and Kitchell(1966) in dog