First, let me make a small digression and compare tram to drum (Ger. Trommel) and to Greek δρόμος (course, race-course, lap, public walk, road, track, way)… and to Serbian drum (road) in addition. Vasmer mentions the Russian words ударить, деру́, драть, раздор (from O.Slav. оударити ’slam, strike, hit, ram’; Serb. udariti, udar, udaram ‘blow, stroke, strike, beat, kick, knock, push’; s-udar-anje ‘clash, shock, collision, crash, impact, smash’) and compares it to Greek δῆρις, -ἡ ‘battle, contest’. Here it seems also to be interesting to mention that Old Slavic дерѫ, дьрати (d’rati ‘tear, flay’) is a cognate to Sanskrit dr̥ṇā́ti (tear, split) and Greek δέρω (separate by avulsion, skin, flay; hence the Greek noun δέρμα ’skin, hide’). All these words are derived from the same proto-form similar to *xər-bhl-gən.; i.e. from its contracted basis – *xər-gon.
English track is probably the source of Serbian traka (band, ribbon) and it is farther (reversibly) related to Serbian trag (trace, footprint, mark, track) as well as to the Russian word дорога (road, path). One of the first thing we must have in mind here is that the most of those *thər- words are derived from an earlier agglutinated form, which sounded similar to *thuer-bəl–gon (Eng. turbid, turbulence, turbine, curving, trouble, Gr. τύρβη ‘tumult, disorder’, Lat. turbidus; cf. Serb. turoban ‘mirky, gloomy’). In this case, the semantic shifts are remarkable: for instance, German trommeln (drum, beat out, pound, batter) is the word from the same “arsenal” as Serbian grmljavina (thunder, thundering; hence Serb. grom ‘thunder’; OSlav. громъ; Gr. χρόμᾰδος ‘crashing sound’, τρίβω ‘to rub, thresh’; τριβόμενος ‘frictional’; Serb. trljanje, trenje ‘friction’, Russ. трение ‘friction’; Cz. tření).
Another word, that should be thoroughly examened, is English group (Ita. gruppo). This word has been developed following the same logic as the above-mentioned “turbulence”. As a matter of fact, turbulence must be a “descendent” word in comparison to group, because the grouping is a precondition for a certain kind of “turbulant activities” in the future. Group is closely related to grab (Ger. graben ‘dig, grub, burrow’, Serb. grabiti ‘catch, grab’, OSlav. грабити, Serb. grablje ‘rake’) and grave (Ger. Grab, Serb. grob; OSlav. гробъ). Serbian krupno (big) and ukrupnjenje (making bigger) are the close relatives to Serbian ogroman (large, big, enormous, giant, huge), and all of these words are derived from the above-mentioned ur-basis – *xər-bhl-gən.
The protoform *xər-bhl-gən is in fact a “twofold or doubled roundness”. It entails that the primary meaning of such agglutinated word was ring + ball : Gr. κρίκος ‘ring’ + ἀμφέλκω ‘be surrounded by’; Serb. krug ‘circle’ + oblo ’round’; Ger. Ring + Ball, like in kurbeln ‘wind’, Kurbel ‘winder’; Lat. circus + oval or bulla ’round swelling’, OE hring ‘circle’ + bolla ’round object, cup, bowl’; cf. OE hring-boga ’snake’ and Icel. hring-laginn ‘coiled up’, both from hypothetical *hring-bləg(n), i.e. in sense ‘bent in circle’ or ‘coiled being’ (hence probably the Serbian n. rugoba ‘a monstrous creature’ and adj. rugoban ‘monstrous, ugly’). Greek Charybdis (Χάρυβδις) is a word equal to Serbian grdoba ‘monstrous creature’, while Serbian grdoba is just a variant of the above-mentioned rugoba (from *h/rugoba => *gruhoba => grdoba). The name crocodile appears to be obtained in a similar way –*hro(ho)b(d)ile; (cf. Ir. crogall, Sc. eireallach ‘a monster’, eirbleach ’slack-jointed or crippled person, Sc. hirplock ‘lame creature, cripple‘, OSl. хромъ, Serb. hrom, hramati ‘lame, to limp’, Skt. स्राम srāmá ‘lame’, Eng. horrible ‘horrifying, ugly’). Alslo there is Gaelic rìbhinn ‘maid, nymph, serpent, beautifull female’, which might be related to Old Slavic рыба ‘fish’ (Serb. riba; cf. Lat. rubeta ‘toad’) and German Raupe ‘caterpillar’.
We can see that Gaelic cruimhean ‘worms’ sounds close to rìbhinn and cruimh ‘worm’, and Latin vermis (worm) must be obtained from the same language “depot” of “curved creatures” (Ger. gekrümmt, krumm ‘bent, curved, curled’; Lith. kirminas, kirmėlė ‘worm’; Skt. कृमिन् kṛmin ‘worm’) as it happened to PSl. *čьrvь (OSl. чрьвь ‘worm’, Russ. червь, Serb. crv). In fact, in this case, the “curved line” is a basic “motto” in depicting some unusual or even unnatural characteristic (Serb. kriv-ljenje ‘bending, distortion’, Russ. кривой ‘distorted’, ис–крив–ление ‘distortion, deformation, curvature’).
In order to understand how the initial *kərb- “root” (from proto-form *xər-bhl-gən) changed its appearance and became *kərp-, *kərv-, *kərm- or *kərf let us compare Latin crimen ‘a judgment, charge, accusation, reproach’ (Gr. κρι̂μα ‘a decision, judgment’) with the South-Slavic krivica (guilt, culpability; krivina ‘curve’; OSl. кривъ ‘crooked’, Serb. kriv ‘crooked, incorrect, guilty’; Cz. křivý ‘crooked, awry, incorrect, improper, deceitful’; Gr. κροιός ‘having deformity’; Lith. kreivas ‘crooked, curved, wry, wrong, false, unfair’). The crucial question here is whether the Latin word crimen is or not related to Serbian krivnja ‘culpability’.
Of course, worm (Serb. crv ‘worm’) is not guilty (Serb. kriv ‘guilty’) for his curly or curved (Serb. kriv ‘curved, crooked, bent’) appearance. Similar can be spoken about the relation among Slavic crevo (OSl. чрѣво, Serbian crevo; Cz. střevo), which has two meanings: intestine and hosepipe. In both cases, we have to deal with the curved/hollowed – cablelike objects. The same ‘curved’ logic is well visible among the Greek words κοιλία ‘cavity of the body, belly, abdomen, intestines’, κοῖλος ‘hollow, concave, cavernous’ and σκολιός ‘curved, winding, twisted, tangled’ (cf. Slav. kolo ‘wheel, circle’; OSl. коло). Even the word for worm followed the same pattern as it was seen above (Gr. σκώληξ ‘worm’; literary, a twisted, curved creature) in Slavic and Germanic languages.
In Old Prussian, intestine was named grābs, phonetically very close to German Grab ‘grave’ and Slavic grab (hornbeam; genus Carpinus). Is there anything at all that would be able to connect those three words? On the other hand, OPr. kērmens ‘body’ (Latv. ķermenis ‘body’) appears to be related to Ger. Körper and Lat. corpus, while OPr. kīrms ‘worm’ (Latv. cērme ‘worm’) seems to be related to ‘body’ (corpus) in a certain, maybe unusual way? Latin trabs ‘a beam of wood, tree-trunk’ is a possible cognate to the Serbian word trup ‘body, trunk, torso, corpus’ (OSl. троупъ, Cz. trup; Russ. труп) and there is another Serbian word (trupac ‘beam of wood, tree-trunk’) – whose meaning is exactly the same as the meaning of Latin trabs.
One thing is very interesting here. Namely, the English words tramp, trample (MLG trampen ‘walk heavily, stamp’) and stamp (OE stempan ’stamp’) are the cognates of the Serbian verbs trapati ‘walk heavily’ (n. trapanje; adj. trapav ‘awkward, clumsy’) and stupati ‘march’ (n. stupanje ‘marching’). Both of these Serbian words are related to the tree stem (Serb. stablo ’stem, bole, trunk’; Russ. стебель, ствол; Cz. stéblо) or tree-trunk (Serb. trupac, truplo ‘tree-trunk’). Not casually, Serbian stablo ’stem, stalk’ is used in naming of the most distal part of human leg – stopalo ‘foot’ (OSl. стопа; Russ. ступня, стопа; Cz. stopa), because of the similarity between tree stem and human leg. The almost same relation could be seen in Germanic: Eng. stem (from *stebh-l-; Ger. Stiel, Stamm, Stengel ), step, Ger. Stufe ’step, stair, grade, degree; cf. Serb. stepen, stupanj ‘grade, degree’, stepenice ’steps, stairs’.
Latin ramus ‘bough, branch, twig’ looks as if it has nothing to do with Serbian grana ‘branch’. Nevertheless, both words are probably derived from the same “root” (*grəbh-gn-), which is, in fact, one of the variations of the primeval form *xər-bhl-gən (”a curved line”). Perhaps this would be clearer if we compared Latin ramale -is ‘brushwood’ with the Serbian grmlje ‘brushwood’. The Greek word for thicket is δρυμός; as we can see, phonetically very close to δρόμος (a public walk, colonnade, from δραμει̂ν , aor. inf. of τρέχω ‘to run’). Now it becomes possible that Lat. ramus is a cognate to above-mentioned trabs ‘beam of wood’. In Russian, the noun хребет ‘ridge, edge, spine’ (Serb. hrbat ‘back’, greben ‘ridge’) is the antecedent of грань ‘verge, edge, margin, boundary’ (Serb. granica ‘boundary, border’, grana ‘twig, branch, limb’).
Semantic shifts in these cases seem to be unexpected, but if we supposed that thousands and thousands of words were sprang from the same source, then such shifts should not be seen as startling. For instance, Serbian grumen ‘lump of soil, clod, clump, turf’ (Russ. грум) is probably the same word as Latin gramen ‘grass, turf, herb’ and these two words may be distantly related to Serbian trava ‘grass’ (OSl. трава, Pol. trawa) and Latin herba ‘herb, grass’. In this context, the Slavic appelative d(e)revo (OSl. дрѣво; Russ. дерево; Pol. drzewo) might be from the same origin as trava ‘grass’. As it is quite obvious, now we are entering the field of pure guessing.
Maybe it would be interesting to mention the Czech words křoví, křoviny ‘bush, shrub, scrub’, travina ‘grass’ and dřevo ‘wood, timber’, i.e. dřevina ‘woody species’. Czech křoví ‘bush’ is probably a false friend to Serbian korov ‘weed’ (cf. Serb. adj. za-korovljeno = za-travljeno ‘weedy’; Russian сорная трава ‘drossy grass, weed’). The OE syntagm grēne græs might be an indication that green is related to grass. Swedish gren has the same mening as Serbian grana (branch; Dan. grene sig ‘branch out’), and both of these words must be derived from the same proto-word (*grebhn-; Serb. greben, hrbat, hrid ‘ridge, cliff’). The OE compound word hrycg-weg ‘an elevated piece of ground, ridge-way’ (Serb. greben ‘ridge’, hrid-ina ‘cliff’) is composed of two words: hrycg ( from hring ‘ring’) and weg ‘road, way’(from IE *bhelg-; Russ. бежать, убегать ‘run’, Cz. běh, běžet ‘run’; Latin pello, pellere, pepuli, pulsus ‘drive out’, veho, vehere ‘carry, ride, sail’; also Serb. beg, bežati ‘running away’, voziti ‘drive’; cf. Serb. vozilo – from *vogi-kolo ‘drive the wheel’ or *begi-kolo ‘the running wheel, cart, wagon’ – and Lat. vehiculum, vehi-culi).
Let us now try to resume what we have been talking about until now:
1) German trommeln is equal to Serbian grmljenje (grmeti ‘to thunder, grom ‘thunder’); while Eng. thunder appears to be related to Serb. tandaranje ‘rattle’ (Serb. tandarati from *hun-dar => udar ‘blow, bump, kick, knock, strike, hit’, udaranje (cf. OE þunrian ’striking, beating, kicking’; probably a metathesis from *hundrian). These sound transpositions are the reason that there are different Germanic words for thunder: Ger. Donner, Dan. torden ‘thunder’ (cf. an ancient Germanic thunder deity – Donar, Thor; OE Þunor, Þūr; Celtic Taranis). Latin tono -are ‘to thunder’ also came from the protoform similar to *hundrian. In reality tono -are is a distant cousin to Latin torno -are ‘turn’ and these two words may be compared to Germ. Donar and Thor. Hence Spanish tronada ‘thunderstorm, tornado’; i.e. from Latin turbedo, turbedinis ’storm’.
2) On the other hand, Serbian drmnuti, drmati ’shake, quake, strike’ is clearly related to grom ‘thunder’ (OSlav. громъ; Serb. grom je drmnuo ‘the thunder struck’) and trema ’stage fright, shaking, shudder’ (Lat. tremo -are ‘tremble, shake, shudder’). There is another Serbian word – trepet ‘tremble’ (OSlav. трепетъ; Russ. трепетать Gr. τρόμος ‘trembling’, τρέμω ‘tremble, trepidation’), a cognate to Latin trepido -are ‘tremble, be afraid, waver’ – which shows that the above mentioned trema and drmati are derived from the “root” *trebh-; i.e. from a certain ur-word similar to the Latin adj. turbatus -a -um ‘disturbed, disordered, restless, troubled’.
3) Serbian trpljenje ’suffering, troubled by pain or loss’ appears to be logically related to trljanje ‘rubbing’ and trenje ‘rubbing, friction’ (Lat. tero, terere, trivi, tritus ‘rub’; Gr. τρίβος ‘rubbing, attrition, worn truck’; τριμμός ‘beaten truck/path’; Serb. trošan put ‘a worn path’); all probably related to Latin turbulentus -a -um ‘turbulent, stormy, tumultuous, be agitated’. In fact, the English noun storm (Ger. Sturm ’storm’, Strom ‘currennt, stream’; OE styrian ‘disturb’) should be in a close relation to Latin turbulentus as well as to Serbian strujanje ‘current, flux, stream, circulation’ (being stormed = being in trouble).
– Of course, there are a lot of different IE words, which are derived from the same *xər-bhl-gən protoform. One of the interesting words is German treiben (OHG trīban: Goth. dreiban; OE drīfan)’propel, drive, push out, force, pursue’; it seems to be an undisputable cognate of Serbian teranje ‘driving’. On the other hand, Serbian isterati ‘oust, expel, drive out’ looks as if it is a cognate of Serbian istrebiti (Russ. истребить, истреблять) ‘exterminate, annihilate, wipe out, root out’. Serbian trebiti ‘root out, exterminate, eradicate’ can be compared to the above-mentioned words, trljanje/trenje ‘rubbing’ and another Serbian word that links trljanje and trebljenje ‘extermination, eradication’ – it is trvenje ‘a process of mutual annihilation’ (hence Serb. po-tiranje, za-tiranje’annihilation’, sturiti, štrojiti ‘make sterile’). Now, with a high probability, we can say that Serbian teranje, trebljenje and trvenje are derived from the same source as English drive (Ger. treiben, ver-treiben = Serb. pro-terati ‘oust, expel, banish’).
4) Latin termen, terminis ‘limit, boundary, end, terminus’ (Gr. τέρμων ‘boundary’) may be related to Latin trabs ‘beam, tree-trunk’ and this one to Serbian taraba (from truplo, trupac ‘tree-trunk’) in a similar way as Latin palus, pali, which, at the same time, has the meaning ’stake, pole’ as well as the meaning ‘fence’ (cf. Eng. palisade, Serb. palica ’stake, stick, pole, club, rod’, from oblica ‘a round shaped pole’; Serb. oblo ’round’, obala ‘coast’, belega ‘mark, boundary’; hence Serb. obeležiti zemlju ‘to mark the land boundaries’). Although it is said that Serbian taraba ‘fence’ is a Turcism (P. Skok, Etimologijski riječnik hrvatskog ili srpskog jezika, Zagreb 1971; book III p. 443), in reality, it can be compared with Latin trabs (the fence made of treetrunks?; Gr. τράφηξ/τράπηξ ‘beam’) or Serbian trupac ‘treetrunk’ (from trubac; cf. Serb. truba ‘trump, trumpet’; OSlav. трѫба). Old English trem, trym ’step’ is probably related to German Treppen ’stairs’ (MLG trame, treme ’stair, rung of a ladder’; M.Du. trame ‘balk, beam’) and, indirectly, to the above mentioned Serbian taraba ‘a fence made of wooden logs’ (cf. Gr. τράφος ‘ditch, trench, fence’). A similar logic might be used to explain the origin of the Serbian word trem (or trijem) ‘portico, lobby, porch’; in fact, it is a porch made of wooden beams (Gr. τέρεμνον).
5) It seems that Sanskrit करभ karabha ‘trunk of an elephant, wal, metacarpus’ might be of a great help here. Namely, Sanskrit karabha seems to be related to Serbian truplo ‘trunk’, grana ‘branch’ and surla ‘proboscis’ (all from *hər-bhl-gn-; cf. Serb. osoran ‘with a big snout, with a nose turned up (contemptuously), surly’ (from *osurlen). If we compare Lithuanian straublys, Serbian surla and Dutch slurf, all with the meaning ‘trunk, proboscis, virile-member’ (Serb. surlav ‘having a big snout’) we may possible be able to understand the process of sound changes, which started from the above-mentioned *hər-bhl-gn agglutinated ur-form, whose original meaning was close to “curved line” or ’rounded object”.
6) The Old English verb treppan ‘to tread, trample‘ appears to be the same word as Serbian noun trapanje ‘a heavy walk’, trampling’. Hence the Serbian adjectives trapav ‘awkward, clumsy’ (Cz. trapný), traljav ’sloppy’, noun truljenje ‘decay’, trulja ‘rag’, dronjak ‘rag’ (from *dro[b]ljnjak), droplja ‘bustard, wading bird, slow bird’ (Russ. дрофа; Cz. drop; Ger. Trappen) and even drolja ‘a tramp, prostitute’. According to Vasmer, Russian трамбовать ’stamp, trample’ is a borrowing from German trampeln ‘trample’