The second section of subgenus Cycas is section Rumphiae (Hill) Wang, 1996 (in Wang & Liang, Cycads in China: 28; subsection Rumphiae Hill 1994, Austr. Syst. Bot. 7: 530; subgenus Truncata de Laub., 1998, Blumea 43: 367), which occurs across Malesia from the Andaman Islands to Tonga.  It is distinguished by the more or less truncated to blunt apical margins of the microsporophylls, sharply bent upwards, usually with a straight apical spike in the center.  The sterile part below the bend is 4-10 mm high.  In most cases the leaves are up 3 meters long while the pinnae midrib is raised below, less so or not at all above with the margins slightly bent. In most cases the fruit is large, up to 6 cm long.  Most seeds have a spongy layer (buoyant) but some do not and all those with a spongy layer in the seed do not have a fibrous layer in the fruit (easily shed seed coat), which usually becomes orange.  It is notable that C. thouarsii in section Cycas also has a spongy layer but in other ways does not conform to section Rumphiae.  Eight species are included in the section with the type species C. rumphii. 

Cycas rumphii Miq., 1839 (Bull. Sci. Phys. et Nat. Nearl. 2: 45) 
C. bougainvilleana Hill, 1994 (Aust. Syst. Bot. 7: 557) 
Distinguished by the prominent crest or wings on the apex of the seed with the fruit up to 60 x 45 mm and by pinnae usually more than 15 mm wide (to 18 mm) and distinctly acuminate.  It is a coastal plant and often branched.  The petiole is about one fifth of the leaf and usually fully equipped with small (3-5 mm) thorns.  The pollen cones reach 55 x 29 cm with a long peduncle.  The spine on the microsporophyll is only 2-5 mm long.  The teeth on the megasporophyll reach at least 5 mm in length and the apical spine is 20-35 mm long.  Hill distinguished C. rumphii from bougainvilleana by the irregular or indistinct teeth on the megasporophyll and the interrupted hypodermis in the pinnae.  Irregular or indistinct teeth occur all over the genus Cycas on older battered megasporophylls but do not characterize rumphii, even for the material cited by Hill thereon.  The leaf specimen, which I have seen, illustrated and likely the basis for characterizing the hypodermis distinction is (unlike the type specimen) a sample of C. celebica (seemanii).  C. rumphii occurs  near the coast from the Moluccas along the north coast of New Guinea to the Solomon Islands. 

Cycas celebica Miq., 1839 (Bull. Sci. Phys. et Nat. Neerl. 2: 45)
C. seemanii A. Braun, 1876 (Sitzungsber. Ges. Naturf. Fr. Berlin 268)  
C. micronesica Hill, 1994 (Austr. Syst. Bot. 7: 554) 
Closely resembles C. rumphii but differs in the weak apical crest on the seed, narrower pinnae mostly less than 15 mm wide, hardly acuminate, and continuous hypoderm, with a pollen cone more slender (15 cm).  The petiole tends to have few or no thorns.  Lindstrom (Brittonia 54: 308, 2002) is wrong  that I based the distinction between C. celebica and C. rumphii merely on the lack of spines on the petiole.  I took note of the fact that Rumphius specifically observed that examples on Celebes were different, including the lack of spines, and selected a specimen from Ambon to illustrate the difference (plate 21), carefully drawing a pinnule that is 15 mm wide.  I have further noted several specimens from Celebes (where C. rumphii is hardly represented) that conform exactly to the observation by Rumphius.  After examining abundant material from the Marianas, New Caledonia, and Fiji variously characterized by Hill as C. micronesica and C. seemanii, I can find no difference from the original C. celebes material.  I consider it likely that the broad current distribution of the species has resulted from the dispersal by natives who have certainly similarly dispersed other useful plants.  Besides outlying islands, C. celebica is found along coasts from Borneo to western New Guinea. 

Cycas scratchleyana F. Muell., 1885 (Victorian Naturalist 2: 18)
C. apoa Hill, 1994 (Aust. Syst. Bot. 7: 553) 
An inland species that strongly resembles C. celebica in almost every way except for the fibrous layer in the fruit, the lack of a spongy layer in the seed, and the short to essentially lacking apical spine on the megasporophyll.  Obviously, specimens lacking the female structures could be confused.  An abundance of thorns on the petioles is reported.  Hill distinguishes C. apoa on the basis of undulate pinnae margins and no teeth on the megasporophyll, neither character to be found on two isotypes which I have seen (see comments under C. rumphii).  His illustration of undulate pinnae margins clearly shows a specimen with pinnae only 6 mm wide (half that reported for the species) and must be considered to be a half developed specimen.  C. scratchleyana, though inland, otherwise has essentially the same distribution as that of C. rumphii.       

Cycas edentata de Laub., 1998 (Blumea 43: 373)
C. litoralis Hill, 1999 (in Hill & Yang, Brittonia 51: 70)
A species exceedlingly similar to C. rumphii (with which it has long been confused) in almost all ways except for the apex of the megasporophyll which is broadly acuminate and toothless, for the lack of wings on the seed, and for a long (12-24 mm) spine on the microsporophyll.  Found along coasts from the Andaman Islands to the Philippines and to Timor. 

Cycas truncata de Laub. 2007 (Encephalartos 92: 17) 
C. inermis auct. non Lour., Hill et al, 2004 (Bot. Rev. 70: 182)
An inland species strongly resembling both C. celebica and C. scratchleyana, with C. celebica the spike on the megasporophyll, with C. scratchleyana the lack of a spongy layer in the seed and the thorny petiole, but differing in the 14-23 mm spine on the microsporophyll and the lack of any wings on the seed apex (note that these differences also apply to C. edentata).  It has been confused with C. circinalis and variously with other species.  It is found from southern Viet Nam to eastern Sumatra and the Philippines to Timor. 

Cycas javana (Miq.) de Laub., 1996 (in Wang & Liang, Cycads of China: 65)
C. circinalis L var javana Miq., 1842 (Monogr. Cycad.: 28)
A species resembling C. celebica except for the apex of the megasporophyll which is narrow (16-21 mm), broadly acuminate, and has teeth to 9 mm long, for the lack of wings on the seed, and for a 10-15 mm spine on the microsporophyll (note the similarity with the differences between C. edentata and C. rumphii).  C. javana has petioles well supplied with thorns.  Found along coasts from eastern Sumatra across Java and nearby islands. 

Cycas campestris Hill, 1994 (Aust. Syst. Bot. 7: 538)
C. falcata Hill, 1999 (Kew Bull. 54: 209)
An inland species resembling C. scratchleyana but differing in the long and spreading teeth on the megasporophyll to at least 9 mm and continuing onto an apical spine, in no wings on the apex of the seed, and in the obtuse apex of the microsporophyll (no apical spine).  Hill included material of C. schumanniana in his description.  In New Guinea the fruits become red.  Hill described C. falcata for southern Celebes and reported orange-yellow fruits and a spongy layer in the seed but saw no male material.  The limited information for C. falcata leaves open the possibility that it could be distinct.  Besides falcata, C. campestris extends across the Moluccas and the length of New Guinea. 

Cycas schumanniana Lauterb., 1901 (in Schum. & Lauterb., Pl. Deutsch. Schutzgeb. Sudsee: 154) 
A small tree of grasslands in eastern New Guinea.  The slightly keeled leaves reach only 150 cm in length, about one third petiole which may bear a few small thorns.  The pinnae are up to 23 cm long by 6-10 mm wide with a spinescent apex unlike most or all other species of the section.  The pollen cone reaches 30 x 12 cm with a 4 cm peduncle.  Microsporophylls have an apical spine to 11 mm long.  The megasporophyll has an apical spine 10-25 mm long and lateral teeth to at least 4 mm.  The fruit is up to 42 x 34 mm with no fibrous layer and the seed has no spongy layer.  Hill grouped C. schumanniana with the Australian species but the microsporophyll is sharply different.