We present neutron powder diffraction results which give unambiguous evidence for the formation of the recently identified new crystalline ice phase [4], labeled ice XII, at completely different conditions. Ice XII is produced here by compressing hexagonal ice I h at T = 77, 100, 140 and 160 K up to 1.8 GPa. It can be maintained at ambient pressure in the temperature range 1.5 < T < 135 K. High resolution diffraction is carried out at T = 1.5 K and ambient pressure on ice XII and accurate structural properties are obtained from Rietveld refinement. At T = 140 and 160 K additionally ice III/IX is formed. The increasing amount of ice III/IX with increasing temperature gives an upper limit of T ≈ 150 K for the successful formation of ice XII with the presented procedure.Although, water has been extensively studied both experimentally and theoretically it still rewards us with new and unexpected properties. This has been demonstrated recently by (i) the detection of polyamorphism [1][2][3] and (ii) the discovery of a new crystalline phase (ice XII) [4,5]. Having been observed in different regions of water's phase diagram the two phenomena were originally thought to be disconnected (Figure 1). The phenomenon of polyamorphism, i.e. the existence of two distinct amorphous phases, is still lacking a comprehensive understanding [6]. Although polyamorphism is equally observed in other substances particularly interesting explanations have been put forward for water. These ideas are based on computer simulations [7] and link polyamorphism to particularities of the supercooled liquid like phase segregation and a second critical point. Due to homogeneous crystallization supercooled water is experimentally inaccessible in the region where these phenomena are expected. The hypotheses must, therefore, be checked indirectly, e.g. by establishing the glassy character of the amorphous phases. The formation of the highdensity amorphous ice (HDA) -achieved by compressing crystalline hexagonal ice I h at temperatures below 150 K to pressures exceeding 1 GPa (Fig. 1) -has received particular attention in this context [7][8][9].So far, the formation of HDA from ice I h has been reported as a well-defined transition channel. The contamination of the amorphous samples by crystalline impurities has been granted little attention [10][11][12]. Only recently [13][14][15] strong experimental indications have become available which imply that all these contaminations correspond to ice XII. As ice XII was originally observed in a completely different region of water's phase diagram this shows that it is a rather prolific phase of water. Moreover, the co-production of ice XII has, as we will argue in the conclusions, far reaching implications for the I h to HDA transformation and, thus, on the origin of water's polyamorphism.In this letter we show that the structure of ice XII produced at low temperatures is definitely identical with the phase characterized by Lobban et al.[4] at higher temperatures. Furthermore, we find that no continuous connec...