There is agreement among scholars that Latin strāgulum ‘covering, blanket’ [Var.+] (cf. in-strāgulum ‘bed-covering’ [Cato], (vestis) strāgulus ‘cloth used as bedding’ [Cic.+]) are based on a seṭ root *sterh3-‘hinbreiten, ausbreiten’ (LIV 2001: 599–600), whose nasal-infixed form continues in Latin as sternō ‘cover, spread out’. The zero-grade of the root, *strh3-(traditionally *st-), well explains the Latin forms with strā-as well as Latin strāmen(tum) ‘straw laid for bedding’ and Greek στρῶμα ‘mattress, bed’ (see Beekes 2010: 1409–1410). We can see its full-grade form in Vedic starī-man-‘Ausbreitung, Ausstreuung’ (Grassmann 1873: 1589), whose -ī-indicates the presence of a laryngeal. Latin strāgēs ‘destruction, ruin’ [Acc.+] also seems to be built upon the same root. On the basis of the semantics, it is possible that the underlying root was ultimately an aniṭ root *ster-‘niederstrecken’ (LIV 2001: 597–598), which Narten (1967: 57–63, 1968: 131–134) suggests should be separated from *sterh3-. However, the seṭ root in the zero grade, *strh3-, is required to explain the -ā-of strāgēs, which echoes the comment in LIV: “In lat. sternere sind die beiden Wurzel *ster-und *sterh3-… zusammengefallen…” (see also Narten 1967: 65 n. 9, 1968: 133 n. 119).
Our attention is now drawn to the -g-of strāgulum and strāgēs. There have been some scholars who regard this -g-as a root-enlargement, such as IEW (2005: 1030), Ernout and Meillet (1985: 647), and Zucchelli (1970: 31 n. 7, 176). One might support this view by referring to forms in some other Indo-European languages, but they cannot be explained by an extended root like *sterh3-g-. For example, OHG strach, MHG strac ‘ausgestreckt’ and OHG strecchen, strecken ‘ausgestreckt machen’ (< PGmc. *strak(k)janan ← *strak(k)az [Norw. strak ‘tight’, etc.]; see Orel 2003: 380), cited by Walde and Hofmann (1965: 600), seem to continue something like *strog-with no laryngeal cf. PGmc.