The subgenus Cycas consists entirely of small trees reaching to several meters in height, with aerial stems often enlarged at the base.  They have large ovoid pollen cones at least 8 cm in diameter.  The seed bearing aggregations form a loose globular apical cluster, often spread into a pendant ring by re-growth of the apex.  The pinnate leaves grow in large clusters which may die before the next growth episode.  The subgenus consists of thirty species which can be divided into three sections (or subgenera) based on differences in the microsporophyll apex.  Less successful has been the proposal to divide the subgenus based on the structure of the fruits which, like subterranean bulbs, does not correlate with anything else and would create a great deal of geographic overlap. 

The type section of Cycas occurs in India and across the Indian Ocean, with an outlier in southern Thailand.  It is distinguished by a gradual upward bend of the apical sterile part of the microsporophyll.  Pinnae are flat with the midvein strongly marked above, less so or not at all below.  Six species have been described with the type being C. circinalis.

 Cycas circinalis L., 1753 (Sp. Pl.: 1188)
The leaves are up to 3 meters long about a quarter of which is petiole most of which has 2-3 mm thorns.  The pinnae are 15-30 cm x 7-10 mm.  The apex of the microsporophyll is tapering and more or less broadly acuminate 20-45 mm long.  The pollen cone is up to 70 x 18 cm with a 5 cm peduncle.  The apex of the megasporophyll is up to about 9 cm x 25-38 mm, the terminal spine is up to 3 cm long with teeth 1-4 mm long along the sides.  The fruit is 25-38 x 21-31 mm with a fibrous layer but no spongy layer.  Found in southwestern India .  In most ways C.circinalis resembles a majority of the species in its subgenus, indicating that these slowly growing plants are closely related.

Cycas pranburiensis Yang et al, 1999 (Brittonia 51: 44) 
Recently discovered in a limestone area of southern Thailand.  Morphologically it is essentially identical with C. circinalis except for a smaller pollen cone (28 x 10 cm), the much smaller leaves (to 120 cm), and the wider pinnae (to 16 mm).  The petiole is reported to have few thorns. 

Cycas sphaerica Roxb., 1832 (Fl. Ind. 3: 147)  
C. nathorstii Schust., 1932 (in Engler, Pflanzenr. 99: 76)  
Differs from C. circinalis in the somewhat wider  pinnae (9-14 mm), somewhat longer teeth on the megasporophyll (aabout 6 mm), and the narrowly acuminate apex of the microsporophyll.  Lindstrom & Hill (2002, Novon 12: 237) insist that nathorstii differs from sphaerica especially in that the apical spine of the megasporophyll bears teeth but, in fact, the type of sphaerica has the very same character.  They say the pinnae are wider (9-14 vs 9-12 mm), the megasporophyll narrower (25 vs 30 mm), and the apical spine of the microsporophyll is longer but give its length as 10 mm while that of sphaerica has been reported to at least 17 mm.  They say that the male cone is larger but give no dimensions and male cone sizes are notoriously fugitive.  That leaves only the megasporophyll teeth which they report to 4 mm while those of sphaerica reach at least 6 mm.  Can this even be considered a variety?  C. sphaerica is native to eastern India and nathorstii to Shri Lanka. 

Cycas beddomei Dyer, 1883 (in Hook. f., Fl. Brit. India 5: 658)  
Quite distinctive because of the narrow revolute pinnae (9-18 cm x 2-4 mm), not flat, which, unlike the rest of the section, have spinescent  tips.  Leaves reach 120 cm about 15% of which is a petiole with a few thorns.  The apex of the microsporophyll is narrowly acuminate at least half of which is a 10-18 mm spine.  The pollen cone is up to 33 x 9 cm.  The teeth on the megasporophyll are thorns up to 15 mm long and there is a terminal spine 2-4 cm long.  The fruit is slightly larger than that of C. circinalis.  C. beddomei comes from an area in southern India. 

Cycas annaikalensis Singh & Radha, 2006 (Brittonia 58: 119)  
Recently found in the C. circinalis zone and more or less intermediate between that and C. sphaerica but distinguished by the fat pollen cone (30-50 x 15-21 cm) and the large fruit (38-49 x 35-43 mm).  The fruit, like that of C. thouarsii, has no fibrous layer and, being nearly of the same large size, may well also have a spongy layer in the seed.  The apex of the microsporophyll is tapering or slightly acuminate, 38-47 mm long.  The teeth on the megasporophyll are up to 8 mm long and the terminal spine  up to 25 mm.  The pinnae are 26-35 cm x 8-12 mm. 

Cycas thouarsii Desf., 1826 (ex Gaudich in Freyc., Voy. Autour du Monde 19: 434)     
C. zeylanica (Schust.) Lindstrom & Hill, 2002 (Novon 12: 238); C. rumphii var 
zeylanica Schust., 1932 (in Engler, Pflanzenr. 99: 75)  
Distinguished by the spongy layer in the seed but no fibrous layer in the fruit as well as by the long lanceolate apex of the megasporophyll (60-120 x 17-25 mm) with weakly developed small teeth to 2 mm and the triangular sterile apex of the microsporophyll to 20 mm long with a narrow acumen to 10 mm.  The leaves are the same as those of C. circinalis except the pinnae are 8-15 mm wide.  The fruit is large (40-60 x 42-50 mm).  The pollen cone reaches 60 x 15 mm with a 5 cm peduncle.  Found in and around Madagascar and in Shri Lanca.  Hill claims the cataphylls in Shri Lanka are longer (10-18 vs 7 cm) but that is hardly the basis for a distinct species and calling attention to the commonality across the Indian Ocean is an important consideration.  Lindstrom & Hill claim to have found their zeylanica in the Andaman area but I have seen material from there (including some of their citations) and it is most definitely C. edentata, what Hill calls literalis.  Because Lindstrom & Hill do not report  the presence of the latter there, I have to presume there are not two species in the Andaman area and that zeylanica is not there.