There is no doubt that Lith. spalvà (4) ‘colour’, Latv. spalva ‘feather, plumage; hair of quadrupeds, fur; bristle; colour of fur, colour of bristles’ and their cognate Latv. spilva ‘cotton grass, seed wool (e. g of cotton grass); [pl.] down’ is derived from the IE. root *(s)pelh- ‘to split, separate, sever’. Alternative suggestions, e. g. a connexion with IE. pel- ‘to cover’, which at first sight might seem more attractive from a semantic point of view, are doomed to fail because the underlying root must have ended in a laryngeal. This is indicated by the intonation of Latv. spal-va. There are several cases where accent class 4 of Lithuanian disyllabic ā-stems matches Latvian level or broken tone. It can be demonstrated that in these cases Latvian is more conservative than Lithuanian (e. g. Lith. kalvà [4] and Latv. kalva < *kolh-ṷah₂-). Already in the IE. parent language the root (s)pelh- must have developed the special meanings: 1. ‘to separate the useless from the usable parts’; 2. ‘to remove the skin from the body of an animal’, cf. on the one hand OLith. pẽlūs pl. ‘chaff’< *pelh-u-, OInd. palāva- pl. ‘chaff’ < *pelh-oṷ- (acc. sg. *pelh-oṷ-ṃ< *pélh-oṷ-ṃ), Pruss. pelwo ‘chaff’, Sl. *'pélva ‘chaff’ < *pélh-ṷah₂-; Lith. spãlis (2) ‘shive’ < *spolh-iḭo-; on the other hand Lat. spolium ‘the skin removed from the body of an animal’ < *spolh-iḭo n. The latter meaning first resulted in ‘(generally) skin’, whence ‘parts covering the skin’, ‘hair, bristle, feathers’ and finnaly also ‘colour of the bristles of animals’. The same holds true of Latv. spilva. It seems possible that in the case of Lithuanian the last stage of the semantic development, the transition to the abstract meaning ‘colour’ is artificial.